Roiz Michael The Queen 39 S Indian Defence 2019 PDF [PDF]

Zitiervorschau

Table of Contents Title page Key To Symbols Preface Bibliography Rare Options Chapter 1 - 4.Bg5 and 4.Bf4 Chapter 2 - 4.e3 Introduction Chapter 3 - 4.e3 and 5.Bd3 Chapter 4 - 4.Nc3 Petrosian System Chapter 5 - 4.Bg5 and 4.Bf4 Chapter 6 - 4.e3 Introduction Chapter 7 - 4.e3 and 5.Bd3 Chapter 8 - 4.Nc3 4.g3 Chapter 9 - 5.Nbd2 Chapter 10 - 5.Qa4 Chapter 11 - 5.Qb3 Chapter 12 - 5.Qc2 Sidelines Chapter 13 - 5.Qc2 and 10.Rd1 Chapter 14 - 5.b3 Introduction Chapter 15 - 9.cxd5 Catalan Chapter 16 - Sidelines Chapter 17 - 4.Bd2 Chapter 18 - 8.Qc2 Chapter 19 - 9.Bf4 Variation Index

2

Grandmaster Repertoire The Queen’s Indian Defence By

Michael Roiz

Quality Chess http://www.qualitychess.co.uk

First edition 2018 by Quality Chess UK Ltd Copyright © 2018 Michael Roiz

Grandmaster Repertoire – The Queen’s Indian Defence All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, electrostatic, magnetic tape, photocopying, recording or otherwise, without prior permission of the publisher.. Paperback ISBN 978-1-78483-056-4 Hardcover ISBN 978-1-78483-057-1 All sales or enquiries should be directed to Quality Chess UK Ltd, Suite 247, Central Chambers, 11 Bothwell Street, Glasgow G2 6LY, United Kingdom Phone +44 141 204 2073 e-mail: [email protected] website: www.qualitychess.co.uk Distributed in North and South America by National Book Network Distributed in Rest of the World by Quality Chess UK Ltd through Sunrise Handicrafts, ul. Szarugi 59, 21-002 Marysin, Poland Typeset by Jacob Aagaard 3

Proofreading by Colin McNab & John Shaw Edited by Andrew Greet Cover design by adamsondesign.com

4

Key to symbols used ² ³ ± µ +– –+ = © „ ƒ ÷ ? ?? ! !! !? ?! ™ #

White is slightly better Black is slightly better White is better Black is better White has a decisive advantage Black has a decisive advantage equality with compensation with counterplay with an initiative unclear a weak move a blunder a good move an excellent move a move worth considering a move of doubtful value only move mate

5

Preface Following the success of my previous book Grandmaster Repertoire – The Nimzo-Indian Defence, I was delighted when Jacob Aagaard and John Shaw offered me the opportunity to write a companion volume covering the Queen’s Indian and Catalan from Black’s perspective, making for a complete repertoire after 1.d4 Nf6 2.c4 e6. As a long-time 1.d4 player, I have used a variety of weapons against the Queen’s Indian and have also incorporated it into my Black repertoire. Despite all this experience in my playing career, I could never have imagined how strategically rich this opening is, until I analysed it for this book. Even though the Queen’s Indian has a reputation for solidity, certain variations can lead to extremely sharp, double-edged play. I believe the Queen’s Indian and Nimzo-Indian combine perfectly, since Black’s strategic goals are similar in both openings: he develops quickly and aims to control the centre with pieces initially, while keeping a flexible pawn structure. In this book we will encounter a few lines where an early Nc3 allows Black to transpose to a pleasant version of a Nimzo-Indian with ...Bb4. Moreover, I have endeavoured to make our complete repertoire as compact as possible, which is one reason why I opted to meet 3.g3 with 3...Bb4†, after which 4.Nc3 would lead straight to Chapter 7 of my Nimzo-Indian book. Apart from making the repertoire theoretically robust and sharing numerous theoretical novelties, one of my main goals in this book has been to share my knowledge of certain thematic pawn structures such as hanging pawns, isolated d5-pawn, Hedgehog structure and more. It is worth mentioning that it is mostly Black who gets to choose which structure to enter, and the correct decision will depend on how well his pieces will coordinate in the resultant positions, as well as taking into account the opponent’s set-up. For instance, after 4.g3 Ba6 5.Nbd2 White’s ability to exert pressure on the centre is limited, so 5...d5 becomes more appealing. If, on the other hand, White goes for some other 5th-move option which enables his knight to go to the more active c3-square, then I would refrain from fixing Black’s central structure so soon. The Queen’s Indian is justifiably popular at all levels and has proven its reliability even in World Championship matches. I hope the readers will enjoy this book and find many useful things for their chess education. Michael Roiz Rishon LeZion, October 2018 The author can be contacted via chessbasic.com.

6

Bibliography Avrukh: Grandmaster Repertoire 1: 1.d4 Volume 1, Quality Chess 2008 Avrukh: Grandmaster Repertoire 1A: The Catalan, Quality Chess 2015 Beliavsky & Mikhalchishin: The Petrosian System Against the QID, Chess Stars 2008 Bologan: The Powerful Catalan, New in Chess 2012 Greet: Play the Queen’s Indian, Everyman Chess 2009 Hera & Tuncer: A Cutting-Edge Gambit against the Queen’s Indian, New in Chess 2014 Hilton & Ippolito: Wojo’s Weapons Volume 1, Mongoose 2010 Palliser: Play 1 d4!, Batsford 2003 Roiz: Grandmaster Repertoire – The Nimzo-Indian Defence, Quality Chess 2017 Electronic/Internet Resources ChessPublishing ChessBase Magazine

7

A) 4.Bg5 Bb7 8 A1) 5.Nc3 h6 6.Bh4 Be7 7.Qc2 c5 8 A11) 8.Rd1 10 A12) 8.dxc5 12 A2) 5.e3 h6 6.Bh4 Be7 7.Nc3 0-0 13 A21) 8.Be2 14 A22) 8.Bd3 15 B) 4.Bf4 Bb4†!? 16 B1) 5.Nbd2 17 B2) 5.Nfd2 Nh5! 19 B21) 6.a3!? 19 B22) 6.Bg3 20

8

1.d4 Nf6 2.c4 e6 3.Nf3 b6 This chapter will deal with two early bishop developments: A) 4.Bg5 and B) 4.Bf4. A) 4.Bg5 This resembles the well-known Nimzo/Queen’s Indian hybrid set-up with 4.Nc3 Bb4 (rather than 4...Bb7, which will be our repertoire choice in Chapter 4) 5.Bg5, which leads to a full-blooded battle. As we will see though, the present line is less dangerous.

The database reveals that the text move dates back to 1887 and was employed by several great players of the past. However, modern practice has demonstrated that this method of development carries some strategic risk, as Black may be able to hunt down the dark-squared bishop and exchange it for a knight at a suitable moment. 4...Bb7 We will analyse A1) 5.Nc3 and A2) 5.e3. For our purposes, I will take the early Nc3 development as a 9

sign that White wants to keep the option of e2-e4. All set-ups involving Nc3 and e2-e3 will be examined under variation A2. A1) 5.Nc3 h6 5...Bb4 is a valid option but not our repertoire choice. 5...Be7 is a popular move which could easily transpose to our main lines after a subsequent ...h6 and Bh4. However, I am happy to nudge the bishop immediately. 6.Bh4 A weaker alternative is: 6.Bxf6 Qxf6 7.e4 The more modest 7.e3 is safer, but after 7...g6 8.Be2 Bg7 9.0-0 0-0 Black had comfortable equality in Rivera – Tiviakov, Neiva 2005. In the long run, the absence of the dark-squared bishop might cause White problems, especially if Black carries out the ...c5 break at an opportune moment. The text move sees White establish a big pawn centre, which seems like the most principled attempt to justify the early exchange of the dark-squared bishop. The database contains close to a hundred games from this position, so I was surprised to see that the following strong move was played in only a few of them:

7...g5! It makes sense to put pressure on the d4-pawn as soon as possible. The text move does this in two ways, as it makes room for ...Bg7 while also preparing to drive the f3-knight away. 8.h3 8.e5 was played in Vidalis – Bousios, Athens 1976, when 8...Qg7!N would have been strong. Black intends some combination of ...g4, ...Nc6, ...d6 and ...0-0-0; one way or another, White will have 10

serious problems on the central dark squares. 8...h5!?N I like this move the most, although 8...Nc6 was also quite promising for Black in Mohanadasan – Almarza Mato, corr. 2011.

9.d5 Making room for the knight to come to d4 if it is attacked. 9.e5 Qg7 is similar to the above note on 8.e5. 9...Na6³ The vulnerability of the dark-squares causes White significant problems.

6...Be7 I like this simple method of breaking the pin. 11

7.Qc2 This plan of preparing e2-e4 is what gives 5.Nc3 an independent character. The more modest 7.e3 transposes to variation A2. 7.d5? is not a move that White particularly wants to play anyway, but I mention it here just to show a simple tactical trick: 7...Nxd5! 8.Bxe7 Nxe7 and Black is a pawn up. This may have been a silly example, but we will see some lines below where Black can make proper use of this motif.

7...c5 Naturally, Black challenges the d4-pawn before White can establish a perfect pawn centre. I analysed two main options: A11) 8.Rd1 and A12) 8.dxc5. 8.e4 has been tried by some strong players, but exposing the d4-square is a risky decision. In almost all of the games from here, Black exchanged on d4 and reached a decent version of a Sicilian structure. Although there is nothing wrong with this, I am attracted by an interesting new idea: 8...Nc6!?N

12

9.d5 (in the event of 9.0-0-0 Nxd4 10.Nxd4 cxd4 11.Rxd4 Qc7 12.Bg3 Qc5 13.Rd2 b5! White’s position starts to feel shaky) 9...Nd4 10.Nxd4 cxd4 11.Nb5 a6!? 12.Nxd4 Nxd5! Making use of the tactical motif mentioned earlier. 13.Bxe7 Nxe7 14.Bd3 0-0 15.0-0 Rc8 Black has comfortable play. A11) 8.Rd1

White mobilizes the last of his queenside pieces. 8...0-0 Black has a few solid options but castling seems most flexible.

13

9.e4 9.dxc5 bxc5 10.e3 transposes to 10.Rd1 in the notes to variation A12 below. 9.d5? Nxd5 10.cxd5 Bxh4µ leaves White with no compensation for the pawn. 9.e3 is harmless in view of 9...cxd4 10.Nxd4 Nc6, when White has problems completing development because of the pressure on the g2-pawn. The text move seems most natural, and it was played in Ajeet Singh – Grib, Maribor 2012. Here I found an interesting new idea:

9...Nc6!?N 9...Nh5!?N 10.Bxe7 Qxe7 is another idea but I prefer to target the d4-square. 10.Be2 After 10.d5? exd5 11.cxd5 Nb4 12.Qb1 Nfxd5 White does not have much for the pawn. 10...Nxd4 11.Nxd4 cxd4 12.Rxd4 Qc7! Controlling the e5-square is extremely important in Hedgehog pawn structures. 13.0-0 13.Bg3 gives Black more than one decent option: 13...Qc6 (if a draw is an acceptable result, Black can play 13...Qc5!? 14.Rd2 Rfd8 15.0-0 d5 16.cxd5 exd5 17.Nxd5 Nxd5 18.Qxc5 Bxc5 19.Rfd1 Bb4 20.Rd4 Bc5 when White has nothing better than repeating moves: 21.R4d2 Bb4=) 14.Bf3

14

14...e5! 15.Bxe5 Qe6 16.Bg3 Bc5 Black manages to regain the pawn and solve all his problems. 13...Qe5 14.Rd3 The somewhat more natural 14.Rfd1 Bc5 15.Bg3 runs into: 15...Qg5! (But not 15...Qxd4?! 16.Rxd4 Bxd4 17.e5 Bxc3 18.Qxc3 Ne4 19.Qd4 Nxg3 20.hxg3 when Black is doomed to passive defence; the rooks have no access to open files.)

16.e5 (After 16.R4d2 e5 Black has good counter-chances thanks to his control over the d4-spot and the poor location of the bishop on g3.) 16...Nh5 17.Rg4 Qf5 18.Qxf5 exf5 19.Rh4 Nxg3 20.hxg3 Bc6 The position is roughly equal but practically more dangerous for White, whose rook is stuck on the edge of the board.

15

14...b5! This thematic break is perfectly timed! It makes complete sense to unblock the c-file and exploit the unfortunate location of White’s queen. Black must obviously avoid 14...Bxe4? 15.Nxe4 Qxe4 16.Bxf6 Bxf6 17.Bf3+– when he loses the exchange for a pawn, and gets no other form of compensation. 15.cxb5 In the event of 15.Bg3?! Qc5 16.Bf4 bxc4 17.Rg3 Bd6 18.Bxd6 Qxd6 19.Bxc4 Qf4³ White suffers from a lack of harmony in his camp. 15...Bxe4 16.Nxe4 Qxe4 Liquidating White’s central pawn is an undeniable achievement for Black. 17.Bxf6 Bxf6 18.Bf3 Rac8! This tactical resource enables Black to maintain the material balance. 19.Qb3 Qc4

16

20.Qxc4 Rxc4 21.Rxd7 Ra4! White is unable to hang on to his extra pawn, and will have to take care to maintain the balance. A12) 8.dxc5 bxc5

9.e3 0-0 10.Be2 The position resembles the 10.dxc5 bxc5 line in the notes to variation A22 on page 15, but here White is leaving the d-file unblocked in the hope of developing pressure along it later. 10.Rd1 Nc6 11.Be2 Nh5 12.Bxe7 Qxe7 13.0-0 transposes to our main line, and was the actual move order of the Jianu – Kamsky game quoted below. 17

10...Nc6! Keeping the pawn structure flexible. 10...d6 has been the most common choice but I am not so keen on it. A good example continued 11.0-0 Nbd7 12.Rfd1 Qb6 13.Rd2 Rfd8 14.Rad1 a6 15.h3 Bc6 16.Ne1 when Black had a solid but rather passive position in Ivanchuk – Anand, Sao Paulo/Bilbao 2011. 11.0-0 Now I like the following way of handling the position: 11...Nh5!N

18

The exchange of bishops makes Black’s position more harmonious. The idea is new in this particular position, but has occurred in similar situations, and after White’s 13th move below we will transpose to some existing games. 12.Bxe7 Qxe7 13.Rad1 f5!? 13...Nf6= has been played a few times and gives Black a solid position, but I think we can be a bit more adventurous on the kingside. 14.Rd2 I also examined: 14.Qd2N 14...Nf6 15.Qd6 Qxd6 16.Rxd6 Nd8! This excellent manoeuvre allows Black to relieve the pressure on the d7-pawn. 17.Ne5 Nf7 18.Nxf7 Kxf7= 14...Nf6 15.Rfd1 Rad8 Black has a harmonious position and may begin thinking about aggressive measures on the kingside, such as ...g5-g4. In the following game, White started to drift into difficulties.

19

16.Ne1 Ne5 17.Qa4 Ne4 18.Nxe4 fxe4ƒ Black had excellent prospects in Jianu – Kamsky, Brasov 2011. A2) 5.e3 This is the clear main line, especially if we include the many games where White starts with 5.Nc3 and follows up with e2-e3.

5...h6 6.Bh4 6.Bxf6 Qxf6 is unimpressive for White; he has nothing better than 7.Nc3, transposing to the Rivera – Tiviakov game, as mentioned in the note to White’s 6th move in variation A1. 20

6...Be7 7.Nc3 This is clearly the best square for the knight. Delaying the knight’s development brings no benefits. For instance, after 7.Bd3 c5 8.0-0 0-0 White has made all the obvious developing moves and has nothing better than transposing to normal paths with 9.Nc3, which transposes to variation A22 on page 15.

7...0-0 White’s next job is to develop his king’s bishop; both A21) 8.Be2 and A22) 8.Bd3 are reasonable. A21) 8.Be2 This move is well met by: 8...Ne4 9.Bxe7 Qxe7 10.Nxe4 Bxe4

21

The simplifications favour Black, who will soon complete development and play ...c5, rendering White’s small space advantage useless. 11.0-0 d6 12.Nd2 Bb7 13.Bf3 13.e4? would be premature in view of 13...Qf6! and White loses a pawn. 13...c5

14.Bxb7 Another game continued 14.Ne4 cxd4 15.exd4 d5 16.Nc3? (White should settle for equality with 16.cxd5N 16...Bxd5 17.Nc3 Bxf3 18.Qxf3 Nd7=) 16...dxc4 17.d5 and now in Abdulla – Pranavananda, 22

New Delhi 2017, Black could have refuted his opponent’s dubious gambit with:

17...Na6!N 18.dxe6 Rad8 19.Qe2 Bxf3 20.Qxf3 Qxe6 21.Rfe1 Qc8µ White remains a pawn down. 14...Qxb7 15.Nf3 15.Qf3 looks to be a drawing attempt but after 15...Nc6! White may regret leaving his knight short of good squares. An instructive example continued: 16.Rfd1?! cxd4 17.exd4

17...d5! 18.cxd5 Nxd4 19.Qg4 e5µ Black successfully transformed the pawn structure and established a clear advantage in Salama – Olsen, Copenhagen 2007.

23

15...Nd7 16.dxc5 dxc5= Black was fine in Spassky – Tal, Tbilisi (8) 1965, and several subsequent games. A22) 8.Bd3

This is the most ambitious way of developing – White now has more control over the centre. 8...c5 9.0-0 9.dxc5 bxc5 10.0-0 Nc6 transposes to 10.dxc5 bxc5 in the note to White’s next move. 9...Nc6 24

Another decent choice is 9...cxd4 10.Nxd4 (10.exd4 Nc6 11.Rc1 reaches our main line below) 10...Nc6= as in Azmaiparashvili – Navara, Ermioni 2006. 10.Rc1 10.d5?! is never played, and indeed after 10...exd5 11.Bxf6 (11.cxd5 Nxd5! wins a pawn by exploiting White’s loose bishop) 11...Bxf6 12.cxd5 Ne5 Black has excellent play in the arising Benoni structure due to his strong dark-squared bishop. 10.dxc5 looks like a concession. Indeed, after 10...bxc5 11.a3 Nh5! the exchange of dark-squared bishops makes Black’s position more harmonious. 12.Bxe7 Qxe7 13.Ne4 was seen in Y. Vovk – Leko, Berlin (rapid) 2015, when Black could have played aggressively with:

13...f5!N 14.Ng3 Nf6 15.Qd2 g5 With good chances to develop an initiative on the kingside.

25

10...cxd4 11.exd4 d5 Black is aiming for a good version of an IQP structure. 12.cxd5 12.Bxf6 Bxf6 13.cxd5 was played in Murali Krishnan – Short, Nagpur 2008. My preference is:

13...Nxd4!?N 14.Nxd4 Bxd4 15.dxe6 fxe6„ Black’s powerful bishops are at least as important as the isolated e-pawn. 12...Nxd5

13.Bg3 26

After 13.Bxe7 Ncxe7³ White had no real compensation for his inferior pawn structure in G. Wilson – Prieckaerts, email 2016. After the text move Black has a choice of good possibilities, the following being my favourite: 13...Bd6 Intending to trade bishops while also preparing ...Nce7. 14.Nxd5 Bxg3 15.Ne3 Bf4 16.Be4 We have been following Gelfand – Leko, Monte Carlo (rapid) 2007. A natural and promising continuation would have been:

16...Rc8N 17.g3 Bxe3 18.fxe3 f5! 19.Bb1 Qd5³ White is clearly under pressure, as the long diagonal is too exposed. B) 4.Bf4

27

This move dates back to 1920 but it is justifiably known as the Miles System after the first British GM, who started using it in 1978 to score brilliant victories over Spassky and Larsen, among others. 4...Bb4†!? In my opinion, this response is the most precise. It is worth comparing the alternative: 4...Bb7 5.e3 Bb4† Black may also try 5...Be7 6.h3 and now: a) Spassky and Larsen both went for 6...0-0 followed by ...d5, which is playable, but not our choice for this repertoire. b) 6...c5 7.Nc3 cxd4 8.Nxd4 leads to Hedgehog-style positions in which Miles’s results were not so impressive. Nevertheless, it seems to me that the f4-bishop exerts annoying pressure on the dark squares in Black’s camp. 6.Nfd2 0-0 7.a3 Be7 8.Nc3 c5?! 8...d5 is a better option although 9.cxd5 has led to good results for White after both 9...exd5 and 9...Nxd5.

28

9.d5! The merits of having the bishop on f4 were well illustrated after: 9...exd5 10.cxd5 Bxd5 10...Nxd5 11.Qf3 Nxf4 12.Qxb7 Nc6 13.exf4 leaves Black with inadequate compensation for a piece. 11.Nxd5 Nxd5 12.Qf3 Nc7 13.Qb7! d6 14.Ne4 Qd7 15.0-0-0ƒ White went on to win another excellent game in Miles – Timman, Wijk aan Zee 1979. White may respond to the bishop check with B1) 5.Nbd2 or B2) 5.Nfd2. 5.Nc3 Bb7 transposes to variation B of Chapter 4 on page 58. B1) 5.Nbd2 This has been slightly the more popular move but it reduces White’s control over the centre, as the knight would ideally prefer to go to c3 once the bishop has been chased away from b4. 5...Bb7 6.e3 6.a3 makes little sense, as after 6...Be7 the e4-square is still under Black’s control. 7.e3 c5 8.Bd3 We have been following Nikolaidis – Schmaltz, Fuerth 2002. It seems to me that Black could have pushed for more than equality with:

29

8...Nh5!N 9.Bg3 d6 10.0-0 0-0³ White will lose the bishop pair without obtaining anything in return.

6...0-0 7.Bd3 7.h3 is more prudent although Black gets easy equality after: 7...d5 8.Be2 (I also examined 8.a3N 8...Be7 9.cxd5 Nxd5 10.Bh2 c5 11.dxc5 Bxc5 12.Rc1 Bd6=) 8...dxc4 9.Bxc4 c5 10.dxc5 Bxc5 11.0-0 Bd6 Black was fine in Chirila – Li Ruifeng, Philadelphia 2016. The text move has been the most popular choice but Black gets an excellent game with a familiar plan:

30

7...Nh5! 8.Bg3 8.Bg5!? invites complications which Black should not shy away from. The critical line continues: 8...Bxf3! 9.Qxf3 Qxg5 10.Qxa8 Nc6 11.Qb7 Qxg2 12.Rf1

12...Bxd2† 13.Kxd2 Qxh2³ Reaching an unbalanced position where Black’s play seems much easier, and indeed he eventually prevailed in Rumyantsev – Schakel, corr. 2012. 8...g6 9.0-0 Nxg3 10.hxg3 We have been following Shushpanov – Khalifman, St Petersburg 1996. I like the following plan:

31

10...Be7!N The bishop will be ideally placed on f6 or g7. A secondary benefit of this move is that ...c5 can be played without leaving the bishop sidelined on b4. 11.e4 It is hard to suggest another constructive idea for White. 11...d6 12.Qe2 c5 13.d5 Bf6 14.Rab1 Re8³ The absence of White’s dark-squared bishop starts to tell. B2) 5.Nfd2

32

This retreat may appear artificial but it is White’s most ambitious way of meeting the bishop check, and is quite thematic. (See the Miles – Timman game, in the note to Black’s 4th move, for another example of it.) White intends to chase the bishop away from b4 and then develop his other knight to c3. 5...Nh5! This is the point behind our early bishop check on the previous move. Unlike the analogous position after 4...Bb7 5.e3 Bb4† 6.Nfd2, White’s queen is not controlling h5, which enables Black to hunt down the dark-squared bishop. White may react with B21) 6.a3!? or B22) 6.Bg3. The awkward-looking 6.Be5 Nc6 7.e3 Nxe5 8.Qxh5 Ng6 9.a3 Be7 10.Qf3 took place in Bermejo Collado – Ionov, Alcobendas 1993. A simple and strong continuation is:

33

10...Rb8N 11.Nc3 Bb7 12.Qh3 0-0 13.Bd3 f5³ White finds himself in a difficult situation. B21) 6.a3!? This move is clever but it doesn’t bother Black after the simple reply: 6...Be7 7.Bg3 Nxg3 8.hxg3 This position was reached in Seunnenga – Van der Wiel, Netherlands 1993. My suggestion for Black is:

8...c5!N 9.e3 34

9.dxc5 bxc5 10.Nc3 Nc6 11.e3 Rb8 12.Qc2 g6 13.Be2 0-0³ offers Black excellent prospects due to the strong dark-squared bishop. 9...cxd4 10.exd4 Bb7 11.Nc3 g6! Covering the h5-spot while vacating g7 for the bishop. 12.Qc2 Bf6 13.Nde4

13...Bxe4! Our strategy involves fighting for the dark squares, so the light-squared bishop is expendable, especially as we want to play ...Nc6 next, which would have blocked the bishop on b7. 14.Qxe4 After 14.Nxe4?! Nc6 15.d5 Nd4 16.Nxf6† Qxf6 17.Qc3 e5³ the superiority of the centralized knight over White’s bishop is indisputable. 14...Nc6 15.0-0-0 0-0„ Black is by no means worse. B22) 6.Bg3

35

This is the usual move. As we will see, the slightly different timing of the a2-a3 move will lead to a significant change in Black’s set-up. 6...Bb7 7.a3 I also considered: 7.e3 g6 8.Qb3!?N (a few games have continued 8.a3 Nxg3 9.hxg3 Bf8, later transposing to our main line below) 8...Bf8! The dark-squared bishop will be ideally placed on g7. 9.Be5 This is not forced, but it is logical to try and provoke a weakening pawn move; otherwise Black will clearly have a comfortable position. 9...f6 10.Bg3 Nxg3 11.hxg3 f5

12.Nc3 Bg7 13.0-0-0 Qe7 14.Nf3 Nc6!³ It is hard to suggest a constructive plan for White. If he advances any of his central pawns, Black’s dark-squared bishop may become extremely powerful.

36

7...Nxg3 8.hxg3 Bf8! Most of White’s pieces are on their initial squares, so Black can afford the time needed to reroute the prized bishop to g7. Instead, 8...Be7N 9.Qc2 g6 10.Nc3 0-0 11.0-0-0 promises White decent attacking chances on the kingside. 9.Nc3 g6

10.Qc2!? This ambitious move prepares long castling and has yielded White two wins so far. It is risky though, 37

as his king will not be safe on the queenside and the d4-pawn becomes more vulnerable. A calmer alternative is 10.e3 Bg7 11.Nf3 (11.Qc2 transposes to our main line below) 11...d6 and now a good example continued: 12.Qc2 Nd7 13.Be2 a6 14.Rd1 Qe7 With two bishops and a flexible position, Black had slightly better prospects in Richter – Wells, Würzburg 1996. 10.e4!?N may be the best way of handling White’s position, although weakening the dark squares carries an obvious risk. After 10...Bg7 11.Nf3 d6 12.Qd2 Nd7 13.0-0-0 Qe7„ Black benefits from having a flexible position with chances to seize the initiative in the long run. 10...Bg7 11.e3 0-0 Despite the semi-open h-file, Black should not hesitate to remove his king from the centre. 12.g4!? Another game continued: 12.Rd1 d6 13.Nf3 Nd7 14.Be2 c5 15.dxc5 Nxc5 16.b4

16...Qf6! A small finesse to solve the problem of the knight. 17.Rd4 Nd7 18.Nb5 d5³ Despite Black’s eventual defeat, his chances were better at this stage in Erdos – Babula, Prague 2016.

38

12...d5!? With this move Black allows the b7-bishop to be blocked in for a while, the justification being that the d4-pawn is now fixed in place, in preparation for ...c5. Another point is that White will have to spend time improving the knight on d2. Another attractive option is 12...c5!?N 13.d5 exd5 14.cxd5 d6, when the powerful bishop on g7 offers Black excellent play. 13.cxd5 exd5 14.Nf3 c5 15.0-0-0 This position was reached in I. Sokolov – Adams, Prague (rapid) 2002, and a couple of subsequent games. In all three cases, Black played 15...cxd4 16.exd4 Nc6 and went on to lose, although his position is not so bad at this stage. However, this is all immaterial as we have a clear improvement available:

39

15...Nd7!N 16.Kb1 a6‚ Black has the makings of a dangerous attack on the queenside. Conclusion This chapter dealt with two early bishop developments, both of which are sensible, but ultimately not too dangerous for Black. After 4.Bg5 Bb7 White has a few options. In the event of 5.Nc3 h6 6.Bh4 Be7 7.Qc2 we play 7...c5 to challenge the enemy centre. In both this position and after 8.Rd1 0-0, I came up with the interesting new idea of meeting the e2-e4 advance with ...Nc6!?, maintaining the tension and aiming to capitalize on the weakened d4-square. A safer plan of development is 5.e3 h6 6.Bh4 Be7 7.Nc3 0-0, but we saw that Black is doing well after both 8.Be2 Ne4 and 8.Bd3 c5. 4.Bf4 is a tricky move, despite its innocuous appearance. (If you doubt this, just look again at the list of players Miles defeated with it back in the day!) I find 4...Bb4†!? to be the best move order, the main point being to meet the thematic 5.Nfd2 with 5...Nh5!, hunting down the dark-squared bishop. We analysed two main options in 6.a3!? and 6.Bg3. The specific details will vary from one line to the next, but the general assessment is clear: Black obtains at least equal chances by focusing his counterplay on the dark squares.

40

A) 5.a3 24 B) 5.Be2 26 C) 5.Nbd2 c5 29 C1) 6.b3 30 C2) 6.Bd3 31 D) 5.Nc3 d5 32 D1) 6.a3 32 D2) 6.cxd5 exd5 7.Bb5† c6 35 D21) 8.Ba4!? 35 D22) 8.Bd3 36

41

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.e3 Although this advance may appear rather modest, White’s attacking abilities should not be underestimated. His bishops will often be deployed on d3 and b2, in Colle-Zukertort style, followed by a knight hop to e5. This method of development was successfully employed by many strong players of the past, such as Bogoljubov and Tartakower. Nowadays, it is mostly used a surprise weapon, but even top players like Kramnik and Mamedyarov have tried it and managed to find fresh ideas. 4...Bb7

We have reached our first branching point. In this chapter, we will consider the sidelines A) 5.a3, B) 5.Be2, C) 5.Nbd2, followed by the somewhat more popular D) 5.Nc3. 5.Bd3 is the main line, and will be discussed separately in the next chapter beginning on page 39. 5.b3 42

The immediate attempt to develop the bishop on b2 is inaccurate in view of: 5...Bb4†! Like in many Queen’s Indian lines, this check interrupts White’s natural development. 6.Bd2 6.Nbd2? loses material to 6...Bc3 7.Rb1 Be4 8.Bb2 Bxb1 9.Bxc3 Bg6µ and White had no real compensation for the exchange in Wang Chen – Li Di, Zhongshan 2014. 6...Be7! Both 6...Bxd2† and 6...Qe7 are perfectly sensible and playable, but I would prefer to avoid simplifications while leaving White’s bishop misplaced on d2.

7.Nc3 0-0 8.Bd3 c5! This move highlights another drawback of the bishop on d2: the d-file is blocked, so White cannot play d4-d5. 9.0-0 cxd4N 9...d6 gave Black a perfectly decent position in Kavinda – Snehal, Moscow 2017, but I would prefer to play more actively. 10.exd4 d5 Black has a comfortable position with pressure on the opponent’s centre. A) 5.a3 This prophylactic move looks too passive and is rarely seen at GM level. 5...g6!? There is obviously nothing wrong with normal developing moves such as 5...Be7 or 5...d5. In the latter case, the likely reply of 6.Nc3 leads directly to variation D1 on page 32. Nevertheless, the slowness of White’s last move offers Black a bit more scope for creativity in his opening play. I find the double 43

fianchetto quite appealing, as it offers a lot of potential for counterplay and White’s last move may turn out to be a loss of time. 6.Nc3 Bg7 7.Bd3 7.Be2 is more timid, and 7...0-0 8.0-0 d5 9.cxd5 Nxd5 10.Bd2 Nd7 11.Rc1 c5 was fine for Black in Li Wenliang – Miles, Beijing 1996. 7...c5 It is important to challenge White’s pawn centre before e3-e4 comes. 7...d5 is playable, but after 8.cxd5 Nxd5 we have a surprising transposition to a well-explored theoretical line of the Petrosian System(!), which lies outside of our repertoire.

8.0-0 8.e4N 8...cxd4 9.Nxd4 Nc6 is similar to our main line and is likely to transpose after both sides castle. 8.d5!? is thematic but Black should meet it resolutely with: 8...exd5 9.cxd5 Nxd5! (9...0-0?! 10.e4 was pleasant for White in Eingorn – Loginov, Oberwart 2001) 10.Nxd5 Bxd5 11.Bxg6 hxg6 12.Qxd5 Nc6 Black has an active position and White will not be able to maintain his blockade on d5. A game continued:

44

13.Rb1 (13.0-0N 13...Rh5 is similar) 13...Rh5! 14.Qd3 d5³ White’s opening play had clearly not been a success in F. Gheorghiu – Ovsejevitsch, Switzerland 2016. 8...0-0 9.e4 After 9.dxc5 bxc5 10.e4 Nc6 11.Be3 d6 12.h3 Qe7 13.Qd2 Nd7³ Black enjoyed full control over the dark squares in Annurudha – Negi, Tabriz 2014. 9...cxd4 10.Nxd4 Nc6 This improves over the more passive 10...d6, as seen in Fish – Volke, Berlin 2007. 11.Nxc6 dxc6! Favourably changing the pawn structure.

45

12.Bg5 12.e5 Ng4 13.Qxg4 Qxd3 14.Rd1 Qf5 15.Qxf5 exf5³ gives Black the better endgame thanks to the bishop pair. 12...h6 13.Bf4 After 13.Bh4 Qd4! the pressure along the h8-a1 diagonal starts to tell, so Black’s chances are preferable. 13...c5 14.Qe2 Nh5 15.Bd2 Qh4 16.Bc2 Rad8³ Black had a lot of activity and excellent control over the dark squares in Stieger – Zenner, email 2010. B) 5.Be2

46

This method of development looks timid and cannot bother Black, although it has been tested in many hundreds of games. 5...d5 This will be our usual scheme of development in the main lines of this and the next chapter, and I see no reason to deviate from it here. With that being said, White’s unambitious set-up allows Black to develop his pieces in practically any way he wishes. 5...Bb4†!? is a decent alternative, and after 6.Nbd2 (or 6.Bd2 Be7 7.0-0 0-0 8.Nc3 c5=) 6...0-0 7.0-0 Be7 8.b3 c5 9.Bb2 Nc6 Black’s position is perfectly playable. 6.0-0 Bd6 7.b3 0-0 8.Bb2 I also considered: 8.cxd5 exd5 9.Ba3 Black can exchange bishops to reach a roughly equal position, but I prefer the more ambitious alternative:

47

9...c5!? Hoping to show that the bishop will be misplaced on a3, whereas the one on d6 can be used for attacking functions. 10.Nc3 a6 11.dxc5 11.Qc2 Nbd7 12.Rad1 Qe7 was Boensch – Schussler, Tallinn 1983, where a draw was agreed, just as an interesting middlegame was being reached. It seems to me that White cannot put serious pressure on the hanging pawns. For instance, 13.dxc5 bxc5 14.Nh4 g6 15.g3 Ne5 with excellent piece play for Black. 11...bxc5 12.Rc1 Nbd7 Black’s active pieces should fully compensate for the vulnerability of the hanging pawns. This is really all you need to know to get through the opening; however, the position is so interesting that I could not resist extending the analysis a little further. What follows is not meant to be exhaustive and you certainly don’t need to memorize any of the lines, but I hope you will find it interesting to see the dynamic possibilities which exist in these hanging-pawns positions.

48

13.Nh4!? I regard this as White’s most ambitious way of handling the position. The knight heads for f5, while vacating the f3-square for the bishop. 13...Re8 14.Bf3 After 14.Nf5 Bf8 15.Na4 g6 16.Ng3 Rc8 17.Bf3 Bc6! 18.Nxc5 Bxc5 19.Bxc5 Bb5 20.Bd4 Bxf1 21.Qxf1 h5 White has compensation for the exchange, but Black is fine. 14...Ne5 15.Na4 It would be inaccurate to insert 15.Nf5?! Bf8 before playing 16.Na4, as 16...Qd7 17.Ng3 Nxf3† 18.Qxf3 d4ƒ gives Black the upper hand.

15...Bc8! This great tactical resource enables Black to exploit the poor location of the knight on h4. 49

16.Bxc5 16.Bxd5? Nxd5 17.Qxd5 Qxh4 18.Qxd6 Bb7! yields Black a decisive attack. 16...Bxc5 17.Rxc5 17.Nxc5? g5–+ traps the knight. The text move avoids an immediate loss of material, but Black can renew the threat with: 17...Be6! But not 17...g5?? 18.Bxd5+– of course.

White is in trouble, for instance: 18.Qd4 18.Be2 Ne4µ 18...Ned7 19.Rc6 g5µ

50

8...Nbd7 9.Nc3 I also checked the alternative knight development: 9.Nbd2 Qe7 Most of the pieces are developed symmetrically but the d6-bishop is better than its counterpart: not only is it more active, but it also leaves the e7-square conveniently free for the queen. 10.Ne5 Rfd8

11.Qc2 After 11.Nxd7 Rxd7 12.Rc1 c5 13.cxd5 Nxd5 14.Bb5 Rdd8 Black was slightly better and eventually won in Donchenko – Roiz, Biel 2014. The text move is hardly a significant improvement, since the queen is still placed rather vulnerably. 11...c5 12.Nxd7 Rxd7 13.dxc5 Bxc5 14.Rad1 We have been following the game Kempinski – Cornette, Germany 2014. My suggested improvement is:

51

14...Rc8N 15.Qb1 Bb4³ Black’s active pieces should ensure a small but lasting initiative.

9...a6 This is a typical prophylactic measure to safeguard the position of the bishop. If Black wants a fighting game, then this is the way to play. Alternatively, 9...c5 looks like a safe equalizer: 10.cxd5 Nxd5! (It seems to me that the only real merit of having the light-squared bishop on e2 is that White avoids blocking the d-file, which might become relevant after 10...exd5. Therefore I prefer to avoid entering this type of structure at the moment.) 11.Nxd5 Bxd5 12.dxc5 Bxc5 13.b4 (13.Rc1 Qe7=) 13...Be7= Black had no problems in Krabbe – Stefanov, Oerebro 1966. 52

10.Rc1 10.Qc2 Qe7 11.a3 was played in Alekseev – Khismatullin, Sochi 2016. I recommend:

11...Rac8!?N Shadowing the white queen. 12.cxd5 exd5 13.b4 appears consistent, but then 13...c5! 14.bxc5 bxc5 15.dxc5 Rxc5 16.Nd4 g6 offers Black excellent piece play and pressure along the c-file. 10...Qe7 Once again, we should improve the queen before doing anything committal such as advancing the cpawn. 11.Qc2

53

11...dxc4!? A normal move such as 11...Rac8 is fine too, but I’d like to show this as an example where Black can get a good game after changing the central structure. 12.bxc4 c5 13.Rfd1 cxd4 14.exd4 Rac8 15.Nd2 In Nabavi – Seifert, Nuremberg 2010, Black’s best way forward would have been to improve the pieces as follows:

15...Rfd8N 16.a4 Bf4„ With good prospects against the hanging pawns.

54

C) 5.Nbd2

The knight often goes to d2 in this structure, but it is hard to see the reason for committing to it so soon. Now it becomes easier for Black to challenge the centre. 5...c5 There is nothing wrong with 5...d5 intending ...Bd6, which may well transpose to variation B31 of the next chapter after Bd3, b2-b3 and so on. However, I like the text move because White’s active possibilities against it are greatly diminished with the knight on d2 instead of c3. White’s most natural developing moves are C1) 6.b3 and C2) 6.Bd3. C1) 6.b3 g6!? A more conventional way of handling the position would be 6...cxd4 7.exd4 d5, when 8.Bb2 Be7 9.Bd3 0-0 10.0-0 reaches a popular position which can arise via various moves orders. Black has a sound set-up and will attempt to put pressure on White’s hanging pawns. Nevertheless, I prefer the text move against the particular move order chosen by White. With the knight already on d2, there is no chance of d4-d5 being a good move, so Black can opt for a double fianchetto with a view to exerting pressure on the d4-pawn.

55

7.Bb2 Bg7 8.Bd3 0-0 9.0-0 Nc6 10.Rc1 10.dxc5 bxc5 11.Qe2 was seen almost a century ago in Euwe – Selezniev, Bad Pistyan 1922. A logical continuation would be:

11...d6N 12.Rad1 Qe7 13.Ne4 Nxe4 14.Bxe4 a5 15.Bxg7 Kxg7 16.Rd3 a4 17.Rfd1 Rfd8 Black’s play seems more comfortable, as b3 is a slightly more significant weakness than d6. 10...Rc8 11.Qe2 This position was reached in Liebhart – Seifert, Schwaebisch Gmuend 2014, and now Black missed a powerful idea:

56

11...Nh5!N Illustrating a major strategic idea behind Black’s set-up: the d4-pawn is in real danger! Play might continue: 12.Ne4 Nb4! 13.Bb1 Nxa2 14.Rcd1 Bxe4 15.Bxe4 f5 16.Bb1 Nb4 Black has won a pawn, and the bishop pair does not provide White with full compensation. C2) 6.Bd3

6...cxd4 6...g6!? is playable here too, but 7.e4!? cxd4 8.e5 is rather double-edged. 57

7.exd4 Nc6 8.b3 It is less precise to start with 8.0-0?! in view of: 8...d5 9.b3 (9.cxd5 is well met by 9...Qxd5!N 10.Bc4 Qd8 11.Re1 Be7³) 9...dxc4 10.Nxc4

10...Be7!N (10...Nxd4 11.Nxd4 Qxd4 12.Bb2 gave White enough activity for the pawn in Stanton – Nagley, email 1999) 11.Nce5 0-0³ White is suffering from the weakness of the isolated pawn. 8...d5 9.Bb2 White has managed to protect the d4-pawn, but Black can once again benefit from delaying the development of the dark-squared bishop.

58

9...g6! 9...Be7 10.0-0 0-0 reaches a position which has been tested in hundreds of games. The kingside fianchetto has two advantages: first Black bolsters his kingside, making it harder for White to build an attack; and secondly, the bishop will be ideally placed to put pressure on White’s centre from g7. 10.0-0 Bg7 11.Rc1 0-0 12.Qe2 The ambitious 12.c5?! runs into 12...Nd7! when White loses his stability in the centre: 13.Bc3 bxc5 14.dxc5 e5µ 12.Re1 is hardly an improvement. 12...Rc8 13.Qe2 Re8 14.a3 occurred in Ftacnik – Tibensky, Martin 2003, when Black should have played:

14...Re7!N 15.h3 dxc4 16.bxc4 Rd7³ White is under considerable pressure.

59

12...Rc8 13.Rfd1 Re8 14.Nf1 Nh5 15.Qe3 dxc4 16.bxc4 Ne7³ White’s pieces were tied to the defence of the weak central pawns in Dizdar – Timman, Sarajevo 1984. D) 5.Nc3

5...d5 This is our standard set-up against the 4.e3 system. 5...Bb4 6.Bd3 takes the game into Nimzo paths, which lie outside of my proposed repertoire from Grandmaster Repertoire – The Nimzo-Indian Defence. We will analyse D1) 6.a3 followed by D2) 6.cxd5.

60

6.Bd3 reaches a position mentioned briefly on page 40 – see the note on 6.Nc3 in the next chapter. 6.b3 will almost certainly transpose elsewhere. For instance, 6...Nbd7 7.Bb2 Bd6 and now 8.Be2 0-0 9.00 leads back to variation B, while the more popular 8.Bd3 0-0 9.0-0 transposes to variation B32 of the next chapter on page 50. D1) 6.a3

This position can also be reached via the Petrosian System. The combination of e2-e3 and a2-a3 may appear modest, but White’s set-up is not without merits. 6...Nbd7 7.b4 This move aims to seize space on the queenside while limiting Black’s counterplay in the centre. White has also fixed the central structure with: 7.cxd5 exd5 7...Nxd5?! leads to a version of the 6.cxd5 Nxd5 7.e3 line of the Petrosian System, as covered in variation B of Chapter 7, where Black has played 7...Nd7 instead of my recommended 7...Be7. 8.Ne5 8.Bd3 Bd6 9.0-0 0-0 10.b4 leads to a well-known Nimzo-Indian line where Black has gained a tempo, as his bishop has gone to d6 in one move rather than retreating from b4. The value of the extra tempo is significant, as evidenced after 10...a6 11.Qb3 Qe7, putting pressure on the b4-pawn, so that White cannot employ the usual squeezing plan with a3-a4 and b4-b5. Play may continue 12.Bb2 b5! followed by bringing the knight to c4, with excellent play for Black. After the text move I found a natural improvement over an old game.

61

8...Bd6N The risky 8...Nxe5?! 9.dxe5 Ne4 10.Nxe4 dxe4 was seen in Puc – Gligoric, Zagreb 1949, when 11.Bd2!N 11...Qd7 12.Bc4ƒ would have been promising for White. 9.f4 0-0 10.Bd3 Ne4! Control over the e4-square is vital in such positions. Instead after something like 10...c5 11.0-0 a6 12.Qf3 b5 13.Bd2 White has decent attacking prospects, helped by the powerful knight on e5. 11.Nxe4 dxe4 12.Bc4 Qh4† 13.g3 Qh3 14.Bf1

14...Qf5!? If a draw is an acceptable result then 14...Qe6 15.Bc4 Qh3= gives White nothing better than repeating moves. The text move is an attempt to keep the game going. 62

15.g4 Qe6 16.Bc4 Qe7 17.0-0 Rad8„ I like Black’s chances; White’s pawn advances have left some holes in his kingside which may be exploited later.

7...dxc4 This exchange isn’t the only option, but I prefer to avoid the somewhat passive positions which arise after White advances with c4-c5. 8.Bxc4 c5 9.bxc5 bxc5 10.d5N White needs to improve on 10.Qa4? Bxf3 11.gxf3 cxd4 12.exd4 Qc7 when Black was clearly better in Czimer – Hudecz, Budapest 2002. The text move is the only challenging continuation, and it demands an accurate response:

63

10...exd5 11.Nxd5 Rb8! 12.0-0 12.Nxf6† Qxf6 13.Rb1 is risky for White, although he is still okay after accurate play. The critical line continues: 13...Qc3† 14.Nd2 Bxg2 15.Rxb8† Nxb8 16.Qg4! This excellent tactical resource enables White to maintain the balance. 16...Bxh1 17.Qc8† Ke7 18.Qxb8 Qxc1† 19.Ke2

19...Bf3†!? Several other moves lead to a draw, but this is the most stylish. 20.Kxf3 Qxd2 21.Qe5† Kd8= and White must take the perpetual.

12...Bd6 Black has enough activity to compensate for his extra pawn island. Many continuations are possible but the following looks most critical:

64

13.Nxf6† Qxf6 14.Ra2 0-0 15.Rd2 Ne5! An important resource to solve the problems along the d-file. 16.Rxd6 Bxf3 17.gxf3 Qg5† 18.Kh1 Nxc4

19.Rd5 Qf6 20.Rxc5 Rfd8 Black has full compensation for the pawn, thanks to the activity of his pieces and the vulnerability of White’s king. D2) 6.cxd5 This is the main option which gives the 5.Nc3 move order a distinct character. 6...exd5 White’s last move was a double-edged decision. By exchanging on d5 so early, White loses some flexibility and gives Black chances to secure control over the e4-square, as a future ...Qe7 or ...Re8 can be played without a pawn on e6 blocking the e-file. On the other hand, White gets the additional possibility of a check on b5, which he hopes will disrupt Black’s coordination. 6...Nxd5 has also been tested at a high level but I do not like it, since 7.Bb5† c6 8.Bd3 gives White the upper hand in the centre.

65

7.Bb5† This check is White’s only way to make serious use of the early exchange on d5. 7.Bd3 Bd6 8.0-0 0-0 transposes to variation B22 of the next chapter on page 44. 7.Ne5 Bd6 8.Bb5† (8.f4 0-0 9.Bd3 c5 10.0-0 leads to variation B221 of the next chapter on page 44) 8...c6 and now 9.Ba4!?N leads straight to variation D21 while 9.Bd3 is the note to White’s 9th move in variation D22. 7...c6 7...Nbd7? proved to be a serious mistake after 8.Ne5 Bd6 9.Bc6 Qc8 10.Qa4± in Chernyshov – Karayev, Prague 2017. We will consider the rare but interesting D21) 8.Ba4!? followed by the more common D22) 8.Bd3. D21) 8.Ba4!?

66

This retreat is not without merits: the c6-square will not be available for our knight, so there will be less pressure on the d4-pawn. 8...Bd6 9.Ne5 0-0 10.0-0 c5 Another attractive way of handling the position might be 10...b5!?N 11.Bc2 Nbd7 12.f4 b4 13.Na4 c5, obtaining some initiative on the queenside.

White can grab a pawn by means of 14.Nxd7 Nxd7 15.dxc5 Nxc5 16.Nxc5 Bxc5 17.Bxh7† Kxh7 18.Qc2† Kg8 19.Qxc5 but after 19...a5© the lack of harmony in White’s camp and the vulnerability of his king provide Black with sufficient compensation. 11.Bd2 a6 67

12.Rc1 I also examined 12.f4N 12...b5 13.Bc2 cxd4 14.exd4 Nc6 15.Be3 Na5„ when Black has promising counterplay on the light squares. 12...Nbd7 13.f4 b5 14.Bc2 cxd4 15.exd4 Qb6 16.Kh1 Qxd4 17.Ne2 Qb6 18.Bc3 We have been following Salem – Kryvoruchko, Abu Dhabi 2015. Here Black can improve with:

18...Bc5!N 19.Bf5 Nxe5 20.fxe5 Ne4 21.Bd4 Bxd4 22.Qxd4 Qd8³ White does not have sufficient compensation for a pawn. D22) 8.Bd3 68

8...Bd6 Greet prefers 8...Be7 in Play the Queen’s Indian, so that the bishop will be better placed in the event that e3-e4 is played soon. However, I see no reason to fear that move, so I prefer the more active method of development. 9.0-0 After 9.Ne5 0-0 White hardly has anything better than 10.0-0, when 10...c5 reaches variation B221 of the next chapter on page 44. I also considered 9.e4 dxe4 10.Nxe4 Nxe4 11.Bxe4 0-0 and now it is important to check that the thematic piece sacrifice doesn’t work: 12.Bxh7†? (White should prefer 12.0-0, which transposes to the main line below) 12...Kxh7 13.Ng5† Kg8 14.Qh5 Bb4†! 15.Kf1 Ba6† 16.Kg1 Bd3 and Black wins. 9...0-0 10.e4 10.Ne5 c5 is another transposition to variation B221 of the next chapter. 10...dxe4 11.Nxe4 Nxe4 12.Bxe4 Nd7 13.Bg5 Nf6 My analysis indicates that the pawns on d4 and c6 are equally weak, while the pin on the knight does not cause Black any problems.

69

14.Bc2 This move has achieved a heavy plus score so we should take it seriously. 14.Rc1 is fairly harmless in view of 14...h6 15.Bxf6 Qxf6 16.Qa4 b5 17.Qc2 Rac8 18.Rfe1 Rc7= as in Provotorov – Tarasov, Kaluga 1997. 14.Re1 h6 15.Bxf6 (15.Bh4 occurred in Malinik – Mullner, Hungary 2008, when 15...Re8!N would have equalized easily, for instance: 16.Bc2 Rxe1† 17.Qxe1 c5 18.dxc5 Bxc5 19.Rd1 Qe7=) 15...Qxf6 16.g3 Rfe8 17.Qd3 Rad8 Black was fine in Zvjaginsev – S. Ivanov, Moscow 2002. 14...h6 As we will see, it is useful to nudge the bishop back to h4 in order to have the option of breaking the pin with ...g5 at the right moment. 15.Bh4 c5! Surprisingly, this has only been played once. Please note that 15...g5? 16.Nxg5! is too dangerous for Black. 16.Re1 This was played in the game. I also considered the following alternative, which the computer rates as White’s best: 16.Ne5N 16...Re8 17.f4 This ambitious move has an obvious drawback: the long diagonal is seriously exposed. 17...Bf8 17...cxd4? 18.Qd3 is an example of what Black must avoid. 70

The text move is natural and strong. Now the d5-spot is vacant for the queen, while the pressure on the d4-pawn becomes more significant. 18.Kh1!? Rc8! Bringing the last piece into play. The rook is active along the c-file, and can also go to c7 to defend f7 if required. 18...Qxd4 19.Qe2! offers White definite compensation for the pawn, although 19...Rxe5! 20.Qxe5 Ng4 21.Qxd4 cxd4„ should be okay for Black. 19.Ba4 19.Bb3 Rc7 20.dxc5 Qxd1 21.Raxd1 Bxc5= is no better.

19...Qd5 20.Qe2 Re7 21.dxc5 Rxc5 22.Rad1 Qe4 23.Qxe4 Bxe4= The chances remain balanced in this lively queenless position.

71

16...g5! 17.Bg3 Obviously we must also check the thematic piece sacrifice: 17.Nxg5!? hxg5 18.Bxg5

18...Bc7! Threatening ...Qd5. White has nothing better than 19.dxc5 (19.Bb3?! Kg7 20.dxc5 bxc5µ) 19...Qxd1 20.Raxd1 Kg7 21.Re7 Rac8³ when Black’s extra piece is worth more than the three enemy pawns. 17...Bxg3 18.hxg3 We have been following the game Ionov – S. Ivanov, St Petersburg (rapid) 2016. Here Black should have accepted the challenge with:

72

18...Bxf3!N 19.Qxf3 cxd4 20.Qd3 Qd6 21.f4 g4³ White has insufficient compensation for a pawn. Conclusion This chapter has introduced the important option of 4.e3, which remains a popular choice at all levels. The natural 4...Bb7 is our reply, after which we looked at several options, apart from the main line of 5.Bd3 which will be discussed in the next chapter. The sidelines of 5.a3, 5.Be2 and 5.Nbd2 all have their merits, but also certain drawbacks which we can look to exploit. Against the last of them, 5...c5 is an attractive option, which may be followed by a kingside fianchetto. 5.Nc3 is the most important option discussed in this chapter, when 5...d5 is our move, inviting a transposition to the next chapter in the event that White develops his bishop straight to d3. If he wishes to avoid this, 6.cxd5 exd5 7.Bb5† is the most significant way of doing so, but 7...c6 is not too much of a concession as Black can generally liberate his position with a timely ...c5 after completing development.

73

A) 6.cxd5 40 B) 6.0-0 Bd6 41 B1) 7.b4!? 41 B2) 7.Nc3 0-0 43 B21) 8.Qe2 43 B22) 8.cxd5 exd5 44 B221) 9.Ne5 44 B222) 9.Qc2 45 B3) 7.b3 0-0 8.Bb2 Nbd7 47 B31) 9.Nbd2 Ne4 48 B311) 10.Rc1 48 B312) 10.Qc2 49 B32) 9.Nc3 Re8!? 50 B321) 10.cxd5 51 B322) 10.Rc1 53

74

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.e3 Bb7 5.Bd3 This is by far the most common choice. White is not yet obliged to choose the set-up for his queenside pieces, whereas the d3-square is clearly the most active spot for the light-squared bishop. 5...d5 This is our standard set-up, intending to put the bishop on d6. We should pay attention to A) 6.cxd5, as this has been tried by some strong players with an interesting follow-up in mind. However, in the great majority of games White castles on the kingside before developing his queenside pieces in one way or another, so we will take B) 6.0-0 as our main line. 6.b3 Nbd7 7.Bb2 Bd6 8.0-0 0-0 is covered under variation B3 on page 47. 6.Nc3 also has little independent significance after: 6...Bd6 (I have to mention that 6...Nbd7 7.0-0 Bd6 seems less precise in view of 8.cxd5 exd5 9.Qc2! a6 10.e4 dxe4 11.Nxe4 0-0 12.Nxd6 cxd6 13.Bf4² when White is better due to the powerful bishops.) 7.0-0 0-0 We have transposed to variation B2. A) 6.cxd5 exd5 7.b3!? This unusual move order has been played with the idea of exchanging the dark-squared bishops. Black’s most challenging reply is:

75

7...Bb4†! The following game illustrates the potential strategic danger of allowing White to carry out his plan: 7...Bd6 8.Ba3 0-0 9.0-0 Re8 10.Bxd6 Qxd6 11.Nc3 a6 12.Rc1² Black was doomed to a passive defence in Kramnik – Pelletier, Geneva (rapid) 2013. 8.Nbd2 8.Bd2 Bd6 9.Nc3 a6 leaves White’s dark-squared bishop misplaced. Play may continue 10.0-0 0-0 11.Ne5 c5 12.f4 Nc6„ and Black is doing well. 8...Nbd7 9.Bb2 0-0 10.0-0

76

10...Ne4 11.Rc1 11.a3 Bd6 12.Ne5 doesn’t bother Black either: 12...Qe7 13.Ndf3 This position was reached in Elguezabal Varela – Prieto Martin, Madrid 2011. Black has a few good possibilities at his disposal, but 13...f6N 14.Nxd7 Qxd7= would be my preference. 11...Re8 12.Qe2 Around here, White starts to drift into a somewhat passive position. 12.Bb5!?N looks more purposeful although 12...a6 13.Bc6 Bxc6 14.Rxc6 Ndf6 seems a good reply. For instance, 15.Qc2 Nxd2 16.Nxd2 Bd6= and Black does not face any problems.

12...a5 13.a3 Bd6 14.a4 Re6 15.Rfd1 Qe7 16.Nf1 f5³ Black had the more comfortable game in Berkes – Janssen, Schwarzach 2015. B) 6.0-0

77

6...Bd6 As usual, we develop the bishop to its most active square. Once again, there are many possible move orders, but White will usually play b2-b3 and Bb2 at some point, and develop the b1-knight, either to d2 or, more commonly, c3. To make the web of move orders even more complicated, White may exchange on d5 at any moment, or he may maintain the central tension indefinitely. We will start by analysing B1) 7.b4!?, followed by lines after B2) 7.Nc3 where White does not follow up with an immediate 8.b3, before finally moving on to the most common set-ups with B3) 7.b3. As mentioned earlier, 7.cxd5 exd5 could be played at any moment. Black can usually proceed in the same manner as in variations B2 and B3 below, where we will cover a number of lines with this pawn structure. 7.Nbd2 has no real independent value after 7...0-0, for instance: 8.Qc2 (or 8.b3 Nbd7 9.Bb2, transposing to variation B31) 8...Nbd7 9.e4 dxe4 10.Nxe4 Bxe4 11.Bxe4 Nxe4 12.Qxe4 We have transposed to variation B21; see page 44 for the continuation, beginning with 12...c5. B1) 7.b4!?

78

This ambitious move should be met by: 7...dxc4 Obviously we must avoid 7...Bxb4?? on account of 8.Qa4† Nc6 9.Ne5 Qd6 10.c5! bxc5 11.Bb5 and White wins. Black should not allow his opponent to expand freely on the queenside: 7...0-0?! 8.c5 Be7 9.Bb2² 8.Bxc4 0-0 9.a3 9.b5 seems worse in view of 9...a6 when White cannot maintain his advanced pawn. 10.Nc3 axb5 11.Nxb5 Be7 12.Bb2 occurred in Moskalenko – Brodsky, Simferopol 1990. Black has few attractive possibilities, but I prefer:

79

12...Nbd7N 13.a4 (13.Qe2?! is weaker in view of 13...c6 14.Nc3 b5! 15.Bd3 b4 16.Ne4 Nxe4 17.Bxe4 Ba6 18.Bd3 Bxd3 19.Qxd3 Qa5³) 13...c6 14.Nc3 c5 With comfortable play for Black.

9...Nbd7 10.Bb2 Qe7 11.Nbd2 This modest move is the best that White can do. Delaying queenside development with 11.Ne5?! is dubious: 11...c5 12.Nxd7 Nxd7 13.dxc5 bxc5 14.b5 Rad8 15.Qe2 Nb6 Black had the initiative in Danner – Razuvaev, Dortmund 1993. 11.Nc3?! is also inaccurate. 11...c5! 12.dxc5 bxc5 13.b5 occurred in Rashkovsky – Sultanov, Samara 2000, when Black could have played more energetically with:

13...Ng4!N 14.e4 Nge5 15.Nxe5 Nxe5 16.Be2 c4ƒ White suffers from a lack of harmony in his camp.

80

11...a5!N This is the easiest way of solving Black’s opening problems. 11...c5?! does not work here; after 12.dxc5 bxc5 13.b5 e5 14.Nb3 Black had no real compensation for his inferior pawn structure in Spassov – Ornstein, Pamporovo 1981. 12.b5 e5 13.dxe5 Nxe5 14.Be2 Rfd8

15.Qc2 15.Nxe5 Bxe5 16.Bxe5 Qxe5 17.Qc2 Ne4 is absolutely fine for Black. 15...Rac8 16.Nd4 c6! 17.Nf5 Qe6„ 81

Black has good prospects along the c-file. B2) 7.Nc3 0-0

Now we have a further split between B21) 8.Qe2 and B22) 8.cxd5. 8.b3 Nbd7 9.Bb2 reaches the main line of the chapter: see variation B32 on page 50. 8.Nb5?! cannot be recommended. 8...Be7 9.Qe2 was seen in Stocek – Duda, Litovel 2013, and now Black could have exploited the lack of coordination among the enemy pieces by means of:

9...dxc4!N 10.Bxc4 a6 11.Nc3 b5 12.Bd3 A number of games have reached this position via different move orders, although in many of them it was White to play from here. 12...Nbd7 Black has an excellent version of a Queen’s Gambit Accepted, for instance: 13.e4 c5 14.e5 Nd5³ 82

B21) 8.Qe2 Nbd7 I have to mention that 8...Ne4 seems like a slight inaccuracy in view of 9.Qc2! f5 10.cxd5 exd5 11.Nb5, when Black cannot avoid trading his important dark-squared bishop.

9.e4 The most consistent. I suggest meeting 9.Rd1 with the novelty 9...dxc4N, and after 10.Bxc4 c5 we have transposed to a few games:

83

11.dxc5 (instead after 11.e4 cxd4 12.Nxd4 Qb8 13.g3 a6³ White’s set-up was quite shaky in Aymard – Laurenc, corr. 2013) 11...Bxc5 12.b3 a6 13.a4 Qc7 Black had a comfortable game in Schmidt – Chuchelov, Dresden 1995. 9...dxe4 10.Nxe4 Nxe4 11.Bxe4 Bxe4 12.Qxe4 c5 White has no advantage, as the following examples demonstrate.

13.Rd1 I also checked 13.Bg5 Qc7 14.Rad1 h6 15.Bh4 and now 15...Rae8! prepares to push some pawns on the kingside; Black was doing fine in Spacek – Klimes, Prague 1990. 13...Qc7 14.Bg5 Rae8 15.Rac1 h6 16.Be3 Rd8= Black has successfully neutralized the pressure along the d-file, and a draw was soon agreed in Mikhalchishin – Stohl, Germany 1993. B22) 8.cxd5 exd5

84

Following this pawn exchange, White’s most important options are B221) 9.Ne5 and B222) 9.Qc2. 9.b3 is actually the most common continuation of all, but 9...Re8 (or 9...Nbd7 first) 10.Bb2 Nbd7 leads straight to variation B321 on page 51. B221) 9.Ne5 c5 10.f4 This aggressive set-up is well met by: 10...Nc6! Putting pressure on the d4-pawn and highlighting White’s lack of central stability. 11.Qf3 The passive 11.Ne2?! is not in the spirit of White’s previous few moves. 11...Ne4 12.Bxe4 dxe4 13.Ng3 occurred in Adly – S. Zhigalko, Dubai (blitz) 2014, when Black should have continued:

85

13...cxd4!N 14.Nxc6 Bxc6 15.Qxd4 Bc5µ With a big positional advantage. 11...cxd4 12.Nxc6 Bxc6 13.exd4 Re8 14.f5 We have been following Doettling – Iljushin, Patras 1999. Here Black missed a nice idea:

14...b5!N Intending to drive the knight away before plonking the knight on e4. 15.Bxb5 15.Bg5 b4 16.Ne2 h6 17.Bh4 Be7 18.Rac1 Qb6³ works out nicely for Black.

86

15...Bxb5 16.Nxb5

16...Bxh2† A simple trick to regain the pawn. 17.Kxh2 Qb8† 18.Bf4 Qxb5 19.Be5 Qb6 Black is comfortably equal, and it is White who has to be slightly careful due to his exposed king. B222) 9.Qc2

This move introduces a positional threat of Nb5 to force the exchange of the d6-bishop, while also preparing e3-e4 in some lines. 87

9...c5! This seems like the most logical way to deal with both of the aforementioned ideas. Moreover, White’s queen is not well placed in the ensuing pawn structure. 10.dxc5 10.Rd1 Nc6 11.dxc5 (11.Be2 looks ugly and leaves White in a passive position after 11...Re8 12.Qf5 Qe7³) 11...bxc5 transposes to our main line, and was the actual move order of the Mamedyarov – Kramnik game noted below. White has also tried maintaining the tension with 10.b3 Nc6 11.a3 in several games. I like 11...Re8 12.Bb2 as in Perch – Kochukova, Moscow 2017, at which point Black should continue:

12...cxd4N 13.exd4 Rc8 In the arising symmetrical structure, Black’s pieces are placed more harmoniously. 10...bxc5 11.Rd1 Improving the rook appears logical, although Black’s response to it in our main line is quite convincing. 11.e4!? occurred in Schweer – Wiesner, Neuenbürg 2008. Here Black should invite complications with 11...c4!N, when the critical continuation is: 12.e5 cxd3

88

13.Qb3! Bxe5! 14.Nxe5 (or 14.Qxb7 Nbd7 and White has to struggle for equality) 14...Nc6 15.Nxd3 Ba6„ Black has at least enough activity to maintain the balance. 11...Nc6 12.a3 Ne5! Exchanging the main defender of White’s kingside.

13.Nxe5 Bxe5 14.e4 14.f4 Bd6 15.b4 was played in Parker – Adams, England 2011, when Black missed the strongest reply: 15...c4!N 16.Be2 a5! 17.b5 Re8 18.Bf3 Bc5³ 14...d4 15.Ne2 89

We have been the encounter between top players Mamedyarov – Kramnik, Dortmund 2007. At this moment I suggest the following improvement over Vladimir’s play:

15...Rc8!N Black’s development advantage is getting obvious. The following beautiful lines illustrate his dynamic potential: 16.f4 16.b3 is no better in view of: 16...Ng4! 17.f4

17...Qh4! 18.h3 Qf2† 19.Kh1 f5!ƒ White comes under an enormous attack.

90

16...Bc7! This is the right square for the bishop, as it can cause massive trouble on the a7-g1 diagonal. 17.b3 17.Ng3? allows 17...c4! 18.Bxc4 Bb6–+ when White’s position is collapsing. 17...Bb6 18.e5 18.Ng3 c4 19.bxc4 Nxe4µ 18...Ng4 19.h3

19...Qh4! 20.hxg4 Qxg4 21.Ng3 Qxg3 22.Bxh7† Kh8µ White is in real trouble. B3) 7.b3

91

I will take this as the main line; White prepares the queenside fianchetto before deciding how to develop the b1-knight. 7...0-0 8.Bb2 Nbd7 Both B31) 9.Nbd2 and B32) 9.Nc3 have been tried by strong players, with the latter being more popular. 9.Ne5 seems somewhat premature. 9...c5 10.f4 (10.Nd2 cxd4 11.exd4 dxc4 12.Ndxc4 Be7 13.Re1 Nd5 left White with no real compensation for his inferior pawn structure in Oswald – Chuchelov, Cologne 1994)

10...cxd4 11.exd4 Ne4 12.Qe2 f5!?N (12...f6 is another reasonable move, which was played in Brousek – Javorsky, corr. 2005) Both sides have secured the central knights, but Black’s set-up looks more 92

harmonious to me. B31) 9.Nbd2

Compared to positions with the knight on c3, here White has more control over the e5-square. On the other hand, there is no Nb5 threat or pressure on the d5-pawn. 9...Ne4 White has two main options: B311) 10.Rc1 and B312) 10.Qc2. 10.Ne5?! is premature in view of: 10...Nxd2 11.Qxd2 Bxe5 12.dxe5 Nc5

93

13.Rfd1 Nxd3 14.Qxd3 dxc4 15.Qxc4 Qe7 Black had the better pawn structure for free in Qaddara – Jasim, Abu Dhabi 2006. B311) 10.Rc1 Qe7

11.Ne5 The following alternative are harmless. 11.Qc2 f5 12.Rfd1 was seen in Leitao – Ionov, Wijk aan Zee 1999, when Black should have played:

12...c5!N 13.cxd5 exd5 14.Nf1 Ndf6 15.h3 Rae8³ Black has the more active play. 94

11.Qe2 f5 12.cxd5!?N (after 12.Ne5 Nxe5 13.dxe5 Bc5 14.Nf3 f4³ Black was better in Guttridge – Van Dijk, Guernsey 2009) 12...exd5 13.Ba6 Bxa6 14.Qxa6

White is trying to exploit the light-square weaknesses on the queenside, but the remoteness of his queen enables Black to seize the initiative: 14...Rae8! 15.Qb5 Qe6 16.Rc2 Nxd2 17.Rxd2 f4³ 11...Nxd2 12.Qxd2 Nxe5 13.dxe5 We have been following the top-level game Nakamura – Nepomniachtchi, Saint Louis 2017. The following natural continuation looks best to me:

13...Bc5N 14.cxd5 Bxd5 95

The position is close to equal but White should be careful, since Black’s pawn structure is better and he has the potential to create a passer on the queenside. B312) 10.Qc2

10...Ndf6 There is nothing wrong with 10...f5, other than the fact that 11.cxd5 exd5 12.Ne5 Qe7 13.f4 secures the e5-outpost for the knight, thus reaching an extremely solid position for White. 11.Rad1 11.Ne5 was seen in Uzunoglu – Atalik, Serik 2017. Black could have comfortably met this move with: 11...c5!N 12.cxd5 Nxd2 13.Qxd2 cxd4

96

14.Bxd4 (14.Nc6 Bxc6 15.dxc6 Ng4 16.h3 dxe3 17.fxe3 Nxe3 18.Rf3 Nd5³ does not provide White with sufficient compensation for the pawn) 14...Bxe5 15.Bxe5 Qxd5 16.f4 Ba6 17.Rfd1 Rfd8 18.Bd4 Ne4 Black’s control over the e4-square gives him excellent prospects. 11...Qe7 12.Ne5 c5 13.cxd5 We have been following the game Prusikin – Naiditsch, Pulvermuehle 2004. It seems to me that there was no good reason for Black to reject the most natural recapture:

13...exd5N 14.Bxe4 14.Ndf3 is well met by 14...Ne8!³, highlighting the vulnerability of White’s centralized knight.

97

14...dxe4 15.Ndc4 cxd4 16.Nxd6 Qxd6 17.Bxd4

17...Qe6 18.Nc4 Nd5= Both sides have solid positions and neither can claim the upper hand. B32) 9.Nc3

This popular move reaches a common position which may arise via many different move orders, not all of which begin with the Queen’s Indian. The knight is moving towards the centre, so Black’s potential hanging pawns (in the event that he plays ...c5 and central exchanges ensue) could come under pressure. 9...Re8!? 98

The prophylactic 9...a6 is much more common, but I prefer not to alter Black’s flexible pawn chain on the queenside. The text move is unusual but it has been used successfully by a number of strong GMs. Black improves his rook in anticipation of the e-file being opened. I considered two main continuations: B321) 10.cxd5 and B322) 10.Rc1. A harmless alternative is: 10.Qc2N 10...dxc4 11.bxc4 (White also gets no advantage with 11.Bxc4 c5 12.Rad1 cxd4 13.Nb5 Bc5 14.Bxd4 a6 15.Bxf6 gxf6 16.Nd6 Bxd6 17.Rxd6 Qc7 18.Rd4 b5=)

11...e5 Obviously, this thematic break is the point behind Black’s somewhat unusual 9th move. 12.d5 Nc5 13.e4 c6„ Black obtains excellent play along the c-file. The other natural possibility is: 10.Nb5 Bf8 11.Rc1 11.Ne5 is well met by 11...dxc4! 12.Nxc4 (12.bxc4!?N is an improvement which maintains the balance, although 12...Nxe5 13.dxe5 Nd7 still leaves White with some long-term worries due to his inferior pawn structure12...a6 13.Nc3 c5 and Black was better in Rosko – Mista, Czech Republic 2018. This position was reached in Radulski – Markus, Vrnjacka Banja 2005. I suggest the following new idea:

99

11...dxc4!N 12.bxc4 c5! 13.Re1 a6 14.Nc3 g6! I like this prophylactic concept: Black is fighting against the d4-d5 advance. 15.Bf1 Bg7„ In the arising complex struggle, White’s central pawns might be vulnerable. B321) 10.cxd5 exd5

11.Rc1 11.Nb5 Bf8 12.Rc1 has occasionally been played. Black replies with 12...c6 (12...c5!? is also possible) 13.Nc3 and now 13...Bd6 would transpose to a bunch of existing games, but the bishop is not badly placed tucked away on f8, so Black could also play something like 13...Ne4 or 13...g6!?, with no 100

problems in either case. 11...a6 12.Ne2 The knight is heading for g3 while unblocking the c-file. This plan is thematic for this line, but it does not trouble Black when the e5-square is under control. 12...Ne4 13.Ng3 g6 14.Qc2 Another typical plan is: 14.Rc2 White intends Qe2 and Rfc1, perhaps followed by sacrificing an exchange by taking the pawn on c7 (after ...Qe7), or perhaps on c6 if Black moves it there. 14...Qe7 15.Qe2 f5 Black shows his aggressive intentions on the kingside and makes the knight on e4 more stable. On the flip side, he has to keep an eye on the e5-square and avoid opening the long diagonal.

16.Rfc1 c6 We have been following the game Kosic – Prohaszka, Budapest 2007, where a draw was prematurely agreed. Black is at least equal, as the following lines indicate. 17.Rxc6?! The exchange sac isn’t correct, but it’s hard to see what else White should do next. After something like 17.Qd1 a5 18.Ne2 a4 White’s set-up looks rather passive. 17...Bxc6 18.Rxc6 b5

101

White has less than full compensation. For instance: 19.Qc2 Nb8 20.Rc8 Qe6 21.Rxe8† Qxe8µ

14...Rc8! Black is overprotecting the c7-pawn and heavily restricting the mobility of White’s pieces. 14...Qe7?! allows 15.Ne5! when White manages to fix the centralized knight, since 15...Nxe5? 16.dxe5 Bxe5 17.Bxe5 Qxe5 18.Bxe4 dxe4 19.Qxc7± leads to a difficult endgame. 15.Qe2 h5 16.Rfd1 This position was reached in Janssen – Rotstein, Germany 2007, and a few other games. On each occasion, Black continued with 16...h4, but driving the knight back to f1 is not much of an achievement 102

and Black has more options if he leaves it on g3 for the time being. Therefore I suggest:

16...a5!N 17.Rc2 After 17.Ba6 Bxa6 18.Qxa6 Nxg3 19.hxg3 Nf6 20.Rc6 Qd7 21.Rdc1 Ra8 22.Qb5 Kg7 White cannot make further progress on the queenside, whereas Black still has various ideas on the opposite flank. 17...Re6 18.Rdc1 c6 Covering the weak c6-square. Now White’s only constructive plan is to exchange the light-squared bishops.

19.Ba6 19.Nf1?! allows Black to prevent the aforementioned plan with 19...Ra8! 20.N1d2 Qe7, after which 103

only Black’s position might be improved. 19...Bxa6 20.Qxa6 Bxg3! 21.hxg3

21...g5!„ Black is by no means worse. For instance, White may push too hard with: 22.Ne5? Nxe5 23.dxe5 g4µ And suddenly the superiority of the strong knight over the passive bishop starts to tell. B322) 10.Rc1

104

This natural move was played in Ioseliani – Cmilyte, Batumi 1999, and a few other games. Surprisingly, the following reply has not yet been played. 10...dxc4!?N This method of challenging the enemy centre is a logical follow-up to Black’s last move. 11.bxc4 11.Bxc4 a6 is fine for Black, for instance: 12.a4 c5 13.d5 exd5 14.Nxd5 Bxd5 15.Bxd5 Nxd5 16.Qxd5 Nf6 17.Qc4 (17.Qg5 Re6 18.Rfd1 Qe8=)

17...Ne4 18.Rfd1 Qb8! This precise move enables Black to mobilize his queenside pawns. After the text move we briefly transpose to a game, but deviate with another novelty.

105

11...c5!N 11...e5 also led to interesting play in Gerl – Kornitzky, Fuerth 2014, but I prefer to start with the other pawn move. 12.d5!? This ambitious advance is White’s only serious attempt to fight for the initiative. After 12.Re1 e5! Black benefits from having the rook on e8. Play may continue 13.Nb5 Bb8 14.dxe5 Bxf3 15.Qxf3 Nxe5 16.Bxe5 Bxe5= and Black has no problems.

12...exd5 13.Nxd5 106

White can also consider 13.cxd5 to keep the extra pair of knights on the board. A logical continuation is 13...a6 14.a4 Ne5 15.e4 Nxd3 16.Qxd3 Rc8„ when Black’s counterplay will be based on the ...c4 advance. 13...Nxd5 13...Ne5!? also looks perfectly playable, for instance: 14.e4 14...Nxd5 15.cxd5 Nxd3 16.Qxd3 f6„ White’s central pawns are stopped. 14.cxd5

14...Nf6 15.Qc2 15.Bb5 can be met by: 15...Nxd5!? (if a draw is an acceptable outcome, there is nothing wrong with 15...Re4 16.Bd3 Re8=) 16.Bxe8 Qxe8

107

17.Qd3 f6 18.Rfd1 Bf8 In the arising unbalanced position, Black’s extra pawn and strong bishops provide him with sufficient compensation for the exchange. 15...Bxd5 16.Rfd1 The pressure along the d-file is annoying, so it makes sense to simplify matters.

16...Bxf3 17.gxf3 Qe7 18.f4 Rad8 White certainly has some play for the sacrificed pawn, but Black is not worse at all. Conclusion

108

The position after 4.e3 Bb7 5.Bd3 d5 remains a popular occurrence at all levels of play. The multitude of move orders and transpositions makes this variation quite tricky to analyse in a systematic way. Black’s general plan involves developing with ...Bd6, ...0-0 and ...Nbd7, but we saw a few specific move orders where it makes sense to tweak the plan with ...Bb4†, or an early ...c5 followed by ...Nc6. From the perspective of adding to existing opening theory, the most interesting part of the chapter was our main line of 6.0-0 Bd6 7.b3 0-0 8.Bb2 Nbd7 9.Nc3, when 9...Re8!? is a rare but attractive option, which I believe to be at least as effective as the more common 9...a6. The natural 10.Rc1 even allows us to respond with a novelty; and even if White responds optimally, he gets no advantage according to my analysis.

109

A) 5.Qd3 57 B) 5.Bf4 Bb4! 58 B1) 6.e3 58 B2) 6.Qb3 60 C) 5.Qc2 c5 61 C1) 6.e4 62 C2) 6.dxc5 bxc5 65 C21) 7.Bg5 65 C22) 7.Bf4 67 D) 5.g3 Bb4 68 D1) 6.Bd2 68 D2) 6.Bg2 70

110

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.Nc3 This move invites 4...Bb4 with a Nimzo/Queen’s Indian hybrid, but we will stick to the ‘pure’ QID alternative. We may, however, consider ...Bb4 in the near future, depending on how White proceeds. 4...Bb7 White’s most important option here is 5.a3, reaching the Petrosian System of Chapters 5-8. Before then, this chapter will cover the relatively minor options of A) 5.Qd3, B) 5.Bf4, C) 5.Qc2 and D) 5.g3. 5.Bg5 has already been covered in variation A of Chapter 1. 5.e3 takes us back to variation D of Chapter 2 on page 32. A) 5.Qd3

This move is similar to 5.Qc2, in the sense that White is preparing e2-e4. The advantage of having the 111

queen on d3 instead of c2 is that White is ready to meet ...c5 with d4-d5, but the drawbacks are that the queen blocks the f1-bishop and is generally more exposed in the centre. 5...d5 5...Bb4 6.a3 Bxc3† 7.Qxc3 transposes to a version of the Classical Nimzo-Indian where Black should be theoretically fine, but this is not compatible with the 4.Qc2 d5 variation which I advocate in Grandmaster Repertoire – The Nimzo-Indian Defence. 6.cxd5 Nxd5 7.e4 Nxc3 8.bxc3 c5 This position is similar to variation B of Chapter 8, but the presence of the queen in the centre causes White definite problems. True, White appears to have saved time by omitting a2-a3, but that move is far from useless, as the following note demonstrates. 9.Qb5† If White goes for natural development with 9.Be2N, then 9...cxd4 10.cxd4 Bb4†! exploits the absence of a2-a3. Following 11.Bd2 Bxd2† 12.Nxd2 0-0 Black has easy play, with pressure on White’s central pawns. The Ukrainian GM Ovsejevitsch has had the present position twice with White, scoring 2/2. However, neither of his opponents found the best reply:

9...Nc6!N The knight is more active here than on d7. 10.Bg5 Be7 11.Bxe7 Qxe7 12.dxc5 0-0! Black can afford to offer a pawn sac, having the following justification in mind. 112

13.cxb6?!

13...Qa3! 14.Qb3 axb6 15.Qxa3 Rxa3³ Black will regain the pawn in a favourable situation. B) 5.Bf4

This mixing of the Nimzo-Indian and Miles System should not cause Black any problems, although it has been tried by such strong players as Korchnoi, Mamedyarov, Navara and others. 5...Bb4! 113

Now that White has avoided the more active Bg5 development, converting to a Nimzo-Indian becomes extremely effective, since Black has total control over the e4-square. We will consider B1) 6.e3 and B2) 6.Qb3. 6.Rc1 is hardly an attractive option6...Ne4 7.Nd2 Bxc3 8.bxc3 In Franks – Melzer, Baden 1992, simple and strong would have been:

8...Nxd2N 9.Qxd2 d6 10.f3 f5 11.e3 Qe7 12.Bg3 Nc6 13.Bd3 0-0-0ƒ Black enjoys a clear positional advantage, and White’s king will not find safety anywhere on the board. B1) 6.e3

114

This looks consistent, but the following damage to White’s pawn structure puts him in serious strategic risk, since his dynamic resources are quite limited. 6...Ne4 7.Qc2 7.Qb3 Bxc3† 8.bxc3 g5 9.Be5 f6 10.Bg3 h5 11.h4 g4 12.Nd2 Nxg3 13.fxg3 occurred in Franco Ocampos – Seirawan, Los Polvorines 1981. A natural improvement is:

13...f5!N The structural drawbacks of White’s position are becoming obvious. The continuation might be: 14.0-0-0 Qe7 15.Bd3 Nc6 16.e4 Na5 17.Qc2 fxe4 18.Bxe4 0-0-0µ 7...0-0 8.Bd3 The timid 8.Be2 f5 9.0-0 Bxc3 10.bxc3 d6 11.Rab1 Nd7³ also left White in a worse position in Aronian – Nakamura, Beijing (rapid) 2012.

115

8...f5 9.0-0 Bxc3 10.bxc3 d6 11.Nd2 11.c5!? was an attempt by White, in Spassky – Lutikov, Leningrad 1960, to change the unfavourable character of the position. I see no reason for Black to avoid the ‘greedy’ option:

11...bxc5N 12.Rab1 Bc6 13.Nd2 Nxd2 14.Qxd2 Nd7 15.f3 Rb8 16.Rxb8 Qxb8 17.Rb1 Qa8³ The text move is a sensible option to swap off the strong knight on e4, but White will still have to contend with some strategic problems.

116

11...Nxd2 12.Qxd2 Nd7 13.f3 e5 14.Bg3 Qe7 15.Rae1 Rae8 16.Qc2 In Lasinskas – Henrichs, Differdange 2008, Black should have taken the opportunity to clarify the central structure with:

16...e4!N 17.fxe4 17.Be2 exf3 18.Bxf3 Bxf3 19.gxf3 Qe6 20.Qd3 Rf7µ leaves White without any compensation for his weaknesses. 17...Bxe4 18.Rf4 Nf6 19.Bh4 Qe6 20.Bxf6 Bxd3 21.Qxd3 Rxf6³ White is fighting for a draw.

117

B2) 6.Qb3

This move has been played in more than a hundred games, including high-level examples such as Mamedyarov – Karjakin, Navara – Polgar and more. 6...Qe7! Amazingly, the move which I believe to be the strongest remains virtually untested; the only example I found was a correspondence game from 1993. Black is protecting the bishop in the most useful way possible, by developing the queen and avoiding any structural concessions. Presumably, the hanging c7pawn is the reason why it hasn’t been played more often; but as we will soon see, this is not a real problem. I also spent some time looking at 6...c5 7.dxc5 Bxc5 and concluded that Black can equalize, but my recommended move is simply stronger. 7.e3 The immediate 7.Bxc7? is punished by 7...Nc6 8.Be5 (otherwise ...d6 traps the bishop) 8...Na5 9.Bxf6 gxf6µ when White loses the extra pawn and finds himself in a difficult position. 7.a3 provokes an exchange on c3 in a favourable situation for Black, who will get a serious development advantage. 7...Bxc3† 8.Qxc3 d6 9.e3 (also after 9.Nd2 Nh5 10.Be3 0-0 11.f3 e5³ White’s set-up looks awkward) 9...Ne4 10.Qc2

118

10...g5! 11.Bg3 Nd7 12.Nd2 Nxg3 13.hxg3 f5 14.0-0-0 0-0-0³ Both kings are safe for now, but Black has the better pawn structure and his bishop exerts pressure on the long diagonal.

7...Ne4!N 7...Bxc3† 8.Qxc3 d6 was played in the stem game, Sward – Sundstrom, corr. 1993, but the text move is much more challenging. Black insists on damaging his opponent’s pawn structure, while offering the c7pawn as a temporary sacrifice. 8.Bxc7 After 8.Bd3 Bxc3† 9.bxc3 d6 10.h3 Nd7 11.0-0 0-0 the bishops are not effective, so White has no compensation for the doubled pawns.

119

8...Nc6 9.Qc2 The knight is coming to a5 anyway, so White may as well move his queen out of harm’s way. 9...Na5 10.Bg3 The same can be said about the bishop!

10...Rc8 11.Bd3 f5! This cool move enables Black to regain the pawn in a better situation than 11...Nxg3 12.hxg3 Nxc4 13.Rc1, when White’s position is quite solid. 12.0-0 Bxc3 13.bxc3 Nxc4³ Black takes full control over the light squares, so White’s position is not pleasant at all. C) 5.Qc2

120

Like in variation A with 5.Qd3, the queen is supporting the e2-e4 advance, but it is generally better placed on c2 where it avoids blocking the light-squared bishop. However, there is also a clear drawback: White has less control over the d5-square, so Black’s reply is self-explanatory. 5...c5 White has two main replies: C1) 6.e4 and C2) 6.dxc5. 6.d5? doesn’t work: 6...exd5 7.cxd5 Nxd5 8.Bg5 (after 8.Qe4† Qe7µ Black remained a pawn up in Sucher – Diermair, St Veit 2009) In Sabrowski – Schilling, Berlin 2006, Black could have obtained a big advantage with:

121

8...f6!N 9.Nxd5 Bxd5 10.Rd1 Be6 11.Bf4 Nc6 12.e4 b5! 13.Be2 Qb6 14.0-0 Be7µ White has no real compensation for the pawn. C1) 6.e4 cxd4 7.Nxd4 This resembles another well-known theoretical line: 4.a3 Ba6 (rather than 4...Bb7 which I recommend) 5.Qc2 Bb7 6.Nc3 c5 7.e4 cxd4 8.Nxd4, which my editor Andrew Greet recommended for Black in Play the Queen’s Indian. The difference is that in our line the white pawn is on a2 instead of a3. Overall this should help Black, although he needs to make specific use of the b4-square, rather than simply reproducing the theory of the 4.a3 Ba6 variation.

7...Bc5 We should start by chasing away the centralized knight. The immediate 7...Bb4 seems less precise. 8.f3 0-0 9.Bg5 Qc7 10.a3 Be7 In Jakovenko – Vallejo Pons, Khanty-Mansiysk 2013, White missed a chance to secure a pleasant space advantage with: 11.Ndb5!N 11...Qc5 12.Qd2 a6 13.Be3 Qc6 14.Nd4 Qc7 15.Rc1² 8.Nb3 The knight is placed a bit awkwardly here, but White wants to release the pressure on f2. 8.Be3? Ng4 is terrible for White. I also examined: 8.Nf3!?N Black has a few sensible replies but I like the following idea: 8...Ng4!? 9.Bg5! This active move looks to be White’s only way to maintain the balance. 122

9.Nd1?! is too passive, and 9...0-0 10.h3 Nf6 11.Nc3 d5ƒ gives Black great prospects. 9.h3? Nxf2 10.Rh2 is a mistaken attempt to trap the knight. This would actually be a decent option in the analogous position with the pawn already on a3, but the difference here is that White lacks the important resource of b2-b4 in some key lines. For instance: 10...Qc7! (10...Qf6! is another strong move, and after 11.e5 Qg6 12.Qxg6 hxg6µ the knight escape) 11.g4 Nh1! 12.Bg2 Qg3† 13.Kf1 Qf2† 14.Qxf2 Nxf2µ Black escapes with the knight and keeps an extra pawn.

9...f6 9...Bxf2†?! 10.Qxf2 Nxf2 11.Bxd8 Nxh1 12.Bg5 leads to an unbalanced position, where the vulnerability of Black’s dark squares might tell. 10.Bh4 Nc6 11.0-0-0 0-0 12.h3 Nge5 13.Nxe5 Nxe5 14.Kb1 Rc8„

8...Bb4! 123

8...Nc6 is the main line in the analogous position with the pawn on a3 (as could occur in the 4.a3 Ba6 line, as noted earlier). However, if Black plays in this manner, it is not clear how he can benefit from the absence of a2-a3. On the contrary: there are certain lines where White’s queenside structure will be more secure with the pawn back on a2. 9.Bd3 A poor alternative is: 9.f3?! This causes White to fall too far behind in development. 9...0-0 10.Be3?! 10.Bd3 is preferable; but if that’s the case, then it’s obvious that the bishop should have gone to d3 on the previous move. After 10...Nc6 11.0-0 Ne5 Black is more than okay. 10...d5 11.0-0-0 In Makhnev – Rumjantsev, Kimry 2014, Black should have continued:

11...Nbd7!N 12.exd5 exd5ƒ Followed by ...Rc8, putting White’s king in serious danger. 9...Nc6 10.a3 10.Bf4 was seen almost a century ago, and it’s quite impressive to see how energetically Black reacted: 10...Nh5! 11.Bd2 Qf6 12.0-0 (12.a3N should be met by 12...Nf4!. Play may continue 13.Bf1 Bxc3 14.Bxc3 e5 15.g3 Ne6 16.f4 Qe7 17.f5 Nc5„ and we reach a closed type of position, where White’s bishop pair is of little benefit.)

124

12...Nf4 13.Bxf4 Qxf4„ Black had comfortable play in Pokorny – Gilg, Trencianske Teplice 1926. 10.0-0N is another natural continuation, when 10...Ne5! is a good answer. Best play looks to be 11.Bf4!? Bxc3 12.Qxc3 Nxd3 13.Qxd3 Bxe4 when White has barely enough compensation to cling to equality. For instance:

14.Qd6 Qe7 15.Qd4 d5 16.f3 Bg6 17.cxd5 Nxd5 18.Qxg7 0-0-0 In the resulting position, Black has the more comfortable play due to the powerful knight on d5. 10...Bxc3† Black retreated the bishop in a few games but I would rather not waste time, and the loss of the bishop pair is of little consequence.

125

11.Qxc3 In Stanisz – Chilton, London 2016, Black should have continued with:

11...d6N Now the knight can go to its perfect home on e5. Opening up the position would definitely suit White: 11...d5?! 12.exd5 exd5 13.0-0 0-0 14.Bg5 dxc4 15.Bxc4² 12.Bf4 Ne5 13.f3 13.Bxe5 dxe5 14.0-0 0-0 is also pleasant for Black, for instance:

15.Rfd1 Qc7 16.Nd2 Rfd8 17.b4 a5 18.f3 Bc6= Black’s control over d4 fully compensates for White’s 126

preferable pawn structure. 13...0-0 14.0-0 Rc8 15.Rfd1 Nfd7 16.Be2 Black’s most straightforward continuation is:

16...Nxc4! 17.Bxc4 d5 18.exd5 exd5 19.Rd4 Rxc4 20.Rxc4 dxc4 21.Qxc4 Nf6= Black has no problems in this simplified position. C2) 6.dxc5 bxc5 6...Bxc5 has been the more popular recapture, but Black has limited prospects of taking over the initiative in the ensuing Hedgehog structure. In my opinion, the presence of a pawn on c5 makes White’s position less safe from a strategic point of view.

127

We will consider C21) 7.Bg5 and C22) 7.Bf4. The overambitious 7.e4?!N weakens the key d4-spot, so after 7...Nc6 8.Be2 (or 8.Bd3 d6 9.0-0 Be7 10.Rd1 Nd4 11.Nxd4 cxd4 12.Ne2 e5³) 8...d6 9.0-0 Be7 10.Rd1 e5 Black gets an easy game. C21) 7.Bg5 This has been the most common continuation, but the pin doesn’t bother Black and the bishop will soon be liquidated. 7...h6 8.Bh4

128

8...Nc6!N This is not just a natural developing move; it is also connected with a concrete idea. 8...Be7 9.Rd1 0-0 10.e3 Nh5 11.Bxe7 Qxe7 12.Be2 Nf6 13.0-0 Rd8= was perfectly playable in Flear – Shvidler, Tel Aviv 1989, but White’s position is also extremely safe. 9.e3 It doesn’t make much sense to start with 9.Rd1, as Black can play 9...g5! 10.Bg3 Nh5 just as in the main line. 11.Bd6? only worsens White’s problems: 11...Bxd6 12.Rxd6

12...g4! 13.Ng1 Nd4 14.Qd2 Qf6 15.e3 Qe5–+

129

9.h3 This attempt to create an escape route for the bishop leads to trouble after: 9...Nd4! 9...Bd6 10.Qd2 Bb8 11.0-0-0 g5 is interesting but the text move is more convincing.

10.Qd3 10.Qd2 runs into 10...Ne4! 11.Nxe4 Nxf3† 12.exf3 Qxh4 13.Rd1 Bc6³ when the pressure along the d-file has been neutralized, so Black is firmly in control. 10...Qb6 11.0-0-0 Rb8 12.Rd2 Bd6! Once again, the bishop on h4 is a source of concern for White.

13.e3 13.Bxf6 gxf6 14.Ne1 Be5 15.e3 Nc6ƒ is excellent for Black. 130

13...Nxf3 14.gxf3 Bxf3 15.Rg1 g5 16.Bg3 Bxg3 17.fxg3 d5³ White has minimal compensation for the pawn.

9...g5! 10.Bg3 Nh5 Now we see the importance of the knight on c6: the bishop is unable to escape to e5, so Black manages to take control over the dark squares. 11.Be2 Nxg3 12.hxg3 Bg7

13.g4 0-0 14.Rd1 Qe7³ White fails to put any pressure on the d-pawn, whereas Black’s bishops have a great deal of potential. Our main plan for the middlegame should be to open the position with a suitably prepared ...d5 advance. 131

C22) 7.Bf4

This seems a better choice than the previous line, although Black still has no real trouble obtaining a good position. 7...Be7 8.Rd1 8.Nb5?! was played in Burmakin – Aseev, St Petersburg 1995. Apparently both players underestimated the strength of the simple reply 8...0-0!N, when it turns out that the rook is poisoned:

9.Nc7?! (9.Bc7 Qc8 10.Bd6 is the lesser evil, although 10...Bxd6 11.Nxd6 Qc7 12.Nxb7 Qxb7 13.e3 Nc6 14.Be2 Rfb8 15.b3 a5³ offers Black easy play and pressure along the b-file) 9...Nh5 10.Bg3 f5! 132

11.Nxa8 Be4 12.Qd2 f4µ White is about to lose both of the trapped pieces. This position was reached in Sapis – Appolonov, Wisla 1992, and one subsequent game. I like the following idea:

8...Qb6!?N Black covers the d6-square while removing his queen from the file of the enemy rook. 9.e3 The prophylactic 9.h3 0-0 10.e3 delays White’s development, and after 10...Bxf3! 11.gxf3 Nc6 12.Be2 d5 the king on e1 starts to feel uncomfortable. 9...Nh5 10.Bg3 White aims to make use of the h-file when his bishop is exchanged. 10...d6 11.Be2

133

11...Nxg3 12.hxg3 g6 13.g4 Nc6 Black has a flexible set-up, with good long-term chances thanks to the bishop pair and extra central pawn. His only slight weaknesses are the d6- and h7-pawns, neither of which can easily be attacked. Moreover, it is hard to find a constructive plan for White. D) 5.g3

This set-up was first employed by Efim Bogoljubov in 1920. In the event of 5...Be7 6.Bg2 the play would transpose to a main line which lies outside of our repertoire, but fortunately we have a more convenient solution.

134

5...Bb4 This is a good moment to convert to a Nimzo set-up. In the Nimzo-Indian move order after 4.g3, Black does not have time to fianchetto his queen’s bishop because White is first to occupy the long diagonal and the knight has not yet gone to f3. Here, our bishop is already on b7, so White’s set-up loses much of its sting. Many moves have been tried, but we will focus on the most popular D1) 6.Bd2 and D2) 6.Bg2. D1) 6.Bd2 This move has become surprisingly popular recently. White avoids any damage to his queenside structure and hopes to make use of his space advantage. 6...c5 In my opinion, the slowness of White’s previous move gives us good reason to strike at the centre. 6...Bxc3 7.Bxc3 Ne4 was a solid alternative in Gelfand – Leko, Odessa 2007. 7.dxc5 A harmless alternative is: 7.a3 Bxc3 8.Bxc3 Ne4 9.Rc1 0-0 10.Bg2 d6 11.0-0 Nd7 12.Qd3 I also examined 12.d5!?N, a thematic attempt to fight against the b7-bishop. In my opinion, the most effective way of developing Black’s counterplay is: 12...exd5 13.cxd5 f5! 14.Nd2 Nxc3 15.Rxc3 Nf6 16.e4 Qd7 17.Re3 Rae8„ The pressure on the d5-pawn offers Black comfortable play. 12...Nxc3 13.Rxc3 Nf6 14.Rd1

14...Be4!N 135

14...Qc7 is a bit too passive for my taste, and after 15.Ne1 Bxg2 16.Nxg2 Rfd8 17.e4² White had a pleasant space advantage in Thorsteinsson – Wedberg, Stockholm 1990. 15.Qd2 15.Qe3 Qc7= is similar. 15...Qe7= Black has no problems at all. 7...Bxc5

8.Bg2 0-0 9.0-0 d5 Exchanging a couple of pawns seems like the simplest among many playable options. 10.cxd5 Nxd5 11.Qc2 Other possibilities do not offer White much either. For instance: 11.Ne5N 11...Nd7 12.Nd3 Nxc3 13.Bxc3 Bxg2 14.Kxg2 Be7= is level. 11.Nxd5 Bxd5 12.Bc3 Nd7 White has no advantage in the symmetrical structure. Perhaps his best try is:

136

13.Qa4!?N Black had no problems after 13.b4 Be7 14.a3 Nf6= in Papineau – Lacourse, Hull 2001. 13...Nf6 14.Rfd1 Qe7 15.Bxf6 Qxf6 16.b4 16.e4 Bb7 17.e5 Qf5 18.Nd4 Bxd4 19.Qxd4 Bxg2 20.Kxg2 Rac8 is basically equal, although White has to be ever so slightly careful due to the overextended e5-pawn. 16...Bd6 17.Nd2 Bxg2 18.Kxg2 Qe7 19.Ne4 Be5 20.Rac1 Rad8= The position is balanced, though plenty of play remains. This position was reached in Mamedyarov – Karjakin, Baku 2015. White’s last move connected his rooks, while also setting up some tactical possibilities involving Ng5. Therefore I like the following prophylactic approach:

137

11...Nf6!N 12.Rfd1 Nbd7 Black’s position lacks any weaknesses. To equalize fully, he just needs to find a safe spot for the queen. 13.Ng5 This is White’s most ambitious try. In the event of 13.Rac1 h6 14.a3 Rc8 15.b4 Be7 16.Qb2 a6 17.Bf4 Qe8= Black reaches a harmonious set-up; his queen is only temporarily misplaced, and the d7-knight will soon be heading for c4. 13...Bxg2 14.Kxg2 Qc8 15.Nge4 Be7 16.Bf4

16...Nd5! This is the most precise route to equality. Instead after 16...Qb7 17.f3 Rfc8 18.Rac1 b5 19.Qd3 a6 20.Bd6 Bd8 21.Bb4 Black still has some defending to do. 17.Qd2 Nxf4† 18.Qxf4 Qb7= D2) 6.Bg2

138

This is the main line in terms of popularity, although Black gets a comfortable game with ease. 6...Bxc3† 7.bxc3 0-0 8.0-0 d6 Black has an extremely harmonious and flexible position. The doubled c-pawns are not really vulnerable for the moment, but they do make it hard for White to advance his central pawns without creating weaknesses. 9.a4 This has been the most popular choice. White has tried all kinds of moves and there’s little point in analysing all of them, so I will just mention a few noteworthy lines to present some of Black’s main ideas and resources. 9.d5?! is thematic but it does more harm than good from White’s perspective: 9...Nbd7 10.Nd4 (10.Bg5 h6 11.Bxf6 Qxf6 12.Nd4 Nc5 13.Nb5 Qe7µ was great for Black in Battaglini – Issel, Berlin 2007)

139

10...Ne5! 11.dxe6 Bxg2 12.Kxg2 Qe8! 13.exf7† Qxf7µ In Kazhgaleyev – Ivanchuk, Almaty (blitz) 2008, Black soon regained the pawn while keeping a big positional advantage. 9.Re1 h6!? 9...Nbd7 has been much more popular but the text move is always useful, and it keeps the useful option of ...Nc6-a5 in some lines. 10.Bf1 10.Bh3 Re8 11.Qc2 occurred in Bluebaum – Meier, Baden-Baden 2014, when 11...e5!N would have been most consistent. Play may continue 12.Nd2 e4! 13.Nf1 Nc6 14.Ne3 Na5³ when Black has clearly benefited from his flexible 9th move, which postponed the development of the b8knight.

140

10...Re8 10...Ne4 can be met by 11.Nh4!? Nxc3 12.Qc2 Ne4 13.f3© with a decent initiative for the sacrificed pawn. 11.a4 a5 12.Nh4 e5 13.dxe5 dxe5 14.Qxd8 Rxd8 15.Ba3 Nbd7³ Black had won the opening battle in Harikrishna – Grischuk, Foros 2006.

9...Nc6! 9...a5 has been tried by some strong players but it is preferable to develop a piece and send the knight towards a5, where it puts pressure on the c4-pawn. 10.Nd2 Na5 11.e4 11.Bxb7 Nxb7 leaves White’s monarch less comfortable. A good example continued 12.e4 e5 13.f4 Re8 14.Qf3 and now in Khurtsidze – Vakhania, Tbilisi 2015, Black missed a powerful idea:

141

14...Na5!N White has no immediate threats and so 15.f5 is the natural attacking move, but then 15...c6! gives Black the better chances, as the ...d5 break will come before White has time for g4-g5. 11...c5 12.Re1 Here I found a useful improvement over the two existing games:

12...Rc8!N 12...cxd4 13.cxd4 Rc8 was played in Leiber – Erturan, Legnica 2013, when 14.Ba3!?N 14...Nxc4 15.Nxc4 Rxc4 16.Bf1 Rc8 17.f3© would have brought White promising compensation for the sacrificed pawn.

142

13.e5!? This is the most energetic option, although it is positionally risky, as the doubled c-pawns will now become isolated. 13.d5 Re8 14.f3 Qd7 15.Bf1 Ba6³ gives Black an excellent position. 13...dxe5 14.Bxb7 Nxb7 15.dxe5 Nd7 16.Qg4 White is striving for a kingside attack but his plan can be thwarted.

16...f5! 17.exf6 Qxf6 18.Ne4 Qf5³ Both sides have pawn weaknesses but those on White’s side are clearly more serious. Conclusion After 4.Nc3 Bb7 White’s most popular continuations are 5.a3, 5.Bg5 and 5.e3, all of which transpose to other parts of the book. This chapter has dealt with an assortment of alternatives. 5.Qd3 d5 resembles a line of the Petrosian System which will be analysed later. Black is doing fine, as White’s queen is exposed in the centre and the absence of the pawn on a3 makes ...Bb4† a useful resource. 5.Bf4 Bb4! sees Black convert to a promising version of a Nimzo-Indian. 6.Qb3 has been tried by some strong players but the almost untested 6...Qe7! gives Black excellent chances of taking over the initiative. 5.Qc2 supports a possible e2-e4 advance, but 5...c5 is a good reaction to challenge White’s pawn centre. We analysed 6.e4 and 6.dxc5, both of which seem fine for Black.

143

Finally, 5.g3 is a modest version of the Fianchetto System, after which the Nimzo treatment with 5...Bb4 works well. White can either avoid doubled c-pawns with 6.Bd2 or allow them with 6.Bg2, but Black can count on equal chances in either case, with the main battle lying ahead in the middlegame.

144

A) 6.Qa4† 75 B) 6.Bf4 77 C) 6.Qc2 dxc4! 7.e4 c5 79 C1) 8.Bf4!? 82 C2) 8.d5 83 D) 6.Bg5 Be7 87 D1) 7.cxd5 88 D2) 7.Qa4† Qd7! 89 D21) 8.Qxd7† 89 D22) 8.Qc2 91 D3) 7.Bxf6 Bxf6 8.cxd5 exd5 93 D31) 9.Qb3 94 D32) 9.Qa4† 95 D33) 9.g3 96 D4) 7.e3 0-0 98 D41) 8.Rc1 98 145

D42) 8.Qc2 100 1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.a3 Even though this move was first played as far back as 1923, it gained both its popularity and its name due to the efforts of Tigran Petrosian who mostly used it in the 60s and 70s – although the database shows that it was just one of several systems that the 9th World Champion used against the Queen’s Indian. The obvious strategic idea behind White’s last move is to restrict the mobility of Black’s dark-squared bishop and secure the future of the knight when it arrives on c3. Since the battle for control of the e4square is extremely important in so many Queen’s Indian lines, one can hardly call a2-a3 a loss of a tempo. 4...Bb7 4...Ba6 is a major alternative but I prefer the text move, which is the most natural and popular. 5.Nc3 The main purpose of White’s last move was to prepare this natural development of the knight, so I will not spend time analysing any rare alternatives. 5...d5 This is definitely the most challenging reply. Black fights for the central squares and aims for quick development. Other continuations such as 5...Ne4 6.Nxe4 Bxe4 7.Nd2 and 5...g6 6.Qc2 can be seen from time to time in grandmaster praxis, but in both cases I like White’s chances due to his superior central control.

146

White’s most popular continuation by a long way is 6.cxd5, and it will be given special attention in Chapters 6-8. Before then, the current chapter will deal with the alternatives A) 6.Qa4†, B) 6.Bf4, C) 6.Qc2 and D) 6.Bg5. 6.e3 has already been covered under 4.e3 Bb7 5.Nc3 d5 6.a3; see variation D1 of Chapter 2 on page 32. 6.Ne5?! is too slow; after 6...Be7 White risks falling behind in development. For instance: 7.cxd5 Nxd5 8.Qa4† c6 9.Nxd5 Qxd5

10.f3 0-0 11.e4 Qd8 12.Be3 c5 13.0-0-0 Qc7³ White’s experimental opening play was not a success in Piket – Langeweg, Lugano 1989. A) 6.Qa4†

147

The idea behind this check is to interrupt Black’s natural development. White’s scheme may appear artificial but it has been tried by many strong players, including Karpov, Gelfand, Svidler and Dreev, so it should be taken seriously. 6...Qd7!? Three other moves have been tried but I prefer the text. If White can afford to move his queen at such an early stage, why shouldn’t Black do the same? 7.Qc2 White has also tried: 7.Qxd7† Nbxd7 8.Nb5 If White does not go for this move, Black will have an easy game as the queen exchange only facilitated his development. 8...Bd6 This is forced. White is pinning his hopes on the bishop pair (once he exchanges on d6), but he is using a lot of time. 9.e3 0-0 9...dxc4?! led to a fairly short draw in Dreev – Gelfand, Linares 1997, but White missed a chance to improve his pieces to good effect with: 10.Nd2!N 10...a6 11.Nxd6† cxd6 12.Nxc4 Ke7 13.a4² 10.Nxd6 10.cxd5 Bxd5 11.Nxd6 cxd6 12.Bd2 a5! transposes. 10...cxd6 11.Bd2

148

11...a5!N It is useful to limit the scope of the d2-bishop. 11...Ne4 12.cxd5 Bxd5 occurred in Vinogradov – Kostin, Voronezh 2015, when 13.Bd3N would have offered White a tiny but risk-free edge. 12.cxd5 Bxd5 13.Bb5 Rfc8 14.Ke2 Rc2 15.Rhb1 e5„ The activity of Black’s pieces fully compensates for White’s pair of bishops.

7...dxc4 8.e3 White had better play it safe with this move. 8.e4?! This advance has proven to be overambitious; White will simply be left a pawn down. 149

8...b5 9.Bf4 a6 Black has plenty of time to secure the b5-pawn. 10.0-0-0 Be7 11.g4 White is trying to play creatively, but this move only invites additional trouble.

11...Bd6! 12.Bg5 h6 13.Bxf6 An earlier game continued 13.Bh4 Bf4† 14.Kb1 Qe7 15.e5 Bxf3 16.exf6 gxf6 17.Rg1 Nd7 and White did not have much for two pawns in Greenfeld – Korchnoi, Beersheba 1992. 13...gxf6 14.d5

14...Qe7! Finally vacating the d7-square for the knight. 15.Bg2 Nd7 150

Black was already completely winning in Ivanisevic – Tiviakov, Leon 2001.

8...c5 9.Bxc4 After the last few natural moves, Black’s next priority should be to improve his misplaced queen. 9...Qc8! This seems most precise. I also considered 9...Qc7 but was not so happy after 10.e4! a6 (10...cxd4? 11.Nb5 Qd7 12.Bf4 Bxe4 13.Nc7† Kd8 14.Qd1±) 11.d5 exd5 12.Nxd5 Nxd5 13.Bxd5 Bxd5 14.exd5 Bd6 15.Be3 when White has interesting play, with active pieces and a potentially strong passed pawn. 10.Qe2 Here I found a useful improvement over a high-level game.

151

10...Be7!N 10...Nbd7 11.0-0 a6 12.Rd1 b5 13.Ba2 Be7 14.e4 b4 was seen in Khalifman – Ivanchuk, Reykjavik 1991, and now after 15.axb4!N 15...cxb4 16.e5! bxc3 17.exf6 Nxf6 18.bxc3 0-0 (18...Qxc3 seems too risky after 19.Bb2 Qc7 20.d5!ƒ) 19.c4² I prefer White’s position due to his strong pawn centre. 11.0-0 0-0 12.dxc5 bxc5! In the event of 12...Qxc5 13.e4 Nbd7 14.Bf4 White should not have anything special. However, I prefer the text move, which weakens Black’s pawn structure but offers other benefits in return.

13.e4 Nc6„ Black’s central control and potential pressure along the b-file compensate for his damaged pawn 152

structure, as is often the case in positions of this type. B) 6.Bf4 This way of handling the position is rather harmless, as White fails to put much pressure on the opponent’s centre. 6...Bd6! 6...Be7 7.e3 0-0 reaches a line of the Bf4 Queen’s Gambit Declined, where Black should be doing okay. However, it is worth taking advantage of our Queen’s Indian move order to challenge White’s strong bishop.

7.Bg3 This has been the most common reply. White hopes to trade bishops in a favourable situation, where the open h-file makes a difference. In the event of 7.Bg5 Black can, at the very least, transpose to variation D with 7...Be7. A more important alternative is: 7.Bxd6 cxd6! 7...Qxd6 is less precise on account of 8.cxd5 exd5 9.Rc1 a6 10.g3 0-0 11.Bg2 Nbd7 12.0-0² when Black was doomed to a passive defence in Nakamura – Dubov, Doha 2016.

153

8.e3 8.cxd5 Nxd5 9.Nxd5 Bxd5 10.e3 0-0 transposes. 8...0-0 9.cxd5 White generally makes this exchange before developing his light-squared bishop, so as to avoid losing a tempo after ...dxc4. 9...Nxd5 10.Nxd5 Bxd5 11.Bd3

11...f5!? Control over the e4-square is one of the main motifs in the Queen’s Indian, so this move feels most thematic. That said, there is nothing particularly wrong with 11...Nd7 12.e4 Bb7. Black has a solid position and, with two sets of minor pieces exchanged, his slight spatial disadvantage is no real problem. 154

12.0-0 Nd7 13.Qe2 Nf6= Black had a comfortable game in Salomon – Urkedal, Fagernes 2016.

7...0-0 8.e3 Nbd7 8...c5 9.dxc5 bxc5 10.Be2 is perfectly playable as well. 9.cxd5 It is no surprise that 9.Bxd6?! has yet to be played, as 9...cxd6 simply leaves White a tempo down on the 7.Bxd6 line noted above. 9.Bd3 This move proves to be harmless at best, provided Black responds correctly with: 9...Bxg3! 9...a6 occurred in Cramling – Dzagnidze Beijing (rapid) 2013, when 10.cxd5!N would have given White a slightly improved version of our main line, as the bishop belongs on d3 anyway whereas the ...a6 move is hardly essential for Black. 10.hxg3 dxc4 11.Bxc4 c5 12.0-0 White cannot profit from the open h-file. The attempt to prove otherwise with 12.Qc2 runs into 12...cxd4 13.exd4 Rc8 14.Bd3 h6³ when White has an unfavourable IQP position with no real chances on the kingside. After the text move, I found a simple improvement.

155

12...Qe7!N 12...cxd4 occurred in Riederer – Weber, Oberhof 2012, when 13.Qxd4 would have been about equal. So why help White to mobilize his queen? 13.Qe2 a6 14.Ba2 b5³ Black’s set-up is more harmonious. 9...Nxd5

10.Nxd5 After 10.Rc1 Nxc3 11.Rxc3 Bxg3 12.hxg3 c5= the open h-file did not bother Black at all in Wirig – A. Kovacevic, Aix-les-Bains 2011.

156

10...Bxd5 11.Bd3 11.Rc1 Bxg3 12.hxg3 c5= was the same story as the above note in Voiculescu – Horvat, corr. 2015. After the text move White is ready to advance his e-pawn, potentially disturbing both of Black’s bishops. A natural solution is:

11...Bxg3!N It is best to deviate from 11...f5, as was played in Bossenbroek – M. Hendriks, corr. 2006. Unlike the Salomon – Urkedal game in the note to move 7, Black has not been able to improve his pawn structure with ...cxd6 here, so the weak e5-square might become a problem. 12.hxg3 h6 13.0-0 c5= Black has nothing to worry about. Moreover, the doubled g-pawns might invite a knight foray to g4 at some point in the future. C) 6.Qc2

157

This is quite an ambitious way of maintaining the tension in the centre. Like in the main line with 6.cxd5 Nxd5 7.Qc2 (as featured in Chapter 8), the queen is supporting the potential e2-e4 advance. The text move was first tested at the elite level in 1992 by Boris Gelfand, who used it to defeat both Karpov and Korchnoi that year. It was not long before several other strong GMs started using it, most notably Dreev, who has played it many times. 6...dxc4! I believe this to be the most accurate response. 6...Nbd7, 6...c5 and 6...Be7 are popular alternatives, each of which suffers from the same drawback: after 7.cxd5 Nxd5 Black has been move-ordered into the 6.cxd5 Nxd5 7.Qc2 variation, having chosen a different option from the 7...Nxc3 move which is our repertoire choice in Chapter 8. The text move avoids that problem and thus is the move which should be preferred on principle, provided that its correctness can be backed up by concrete analysis, which does indeed appear to be the case. It is worth including a couple of historically important examples after: 6...Nbd7 7.cxd5 Nxd5 (7...exd5 8.Bf4 has proved to be quite passive for Black) 8.Nxd5

158

Compared with variation C of Chapter 6, Black is unable to recapture with the queen here, which makes this exchange more effective. Let me show two examples in which one of the all-time greats was unable to solve Black’s problems: a) 8...exd5 9.Bg5 f6 10.Bf4 c5 11.g3 g6 12.h4!ƒ made it hard for Black to cope with his weaknesses in Kasparov – Karpov, Moscow (32) 1984. b) 8...Bxd5?! was an unsuccessful attempt to improve, and after 9.e4 Bb7 10.Bb5 c6 11.Bxc6 Rc8 12.d5 Qc7 13.Nd4± things had clearly gone wrong for Black in Gelfand – Karpov, Moscow 1992. 7.e4 This is clearly the most logical move. 7.Bg5?! This move has been employed by several strong players including Dreev, but it proves to be dubious in view of the following accurate response.

159

7...a6! 8.e4 b5 Black will maintain a healthy extra pawn. 9.d5 This is White’s trickiest attempt. 9.e5?! h6 10.Bh4 g5 resembles the Botvinnik Variation of the Semi-Slav, but here the b7-bishop is not blocked. 11.Nxg5 hxg5 12.Bxg5 Be7 13.exf6 Bxf6 14.Bxf6 Qxf6 15.Rd1 Nc6 16.Ne4 Qg7–+ Nestorovic – Timoshenko, Belgrade 1995. Despite the material balance, White’s position is absolutely lost due to his inability to complete development and remove his king from the centre.

9...Be7! Black does best to ignore the d-pawn and focus on development. 10.dxe6 160

10.0-0-0? exd5 11.Bxf6 Bxf6 12.Nxd5 0-0! refutes White’s opening play: after the further 13.Nb6 cxb6 14.Rxd8 Rxd8 15.Be2 Nd7 followed by ...Nc5 his position was ruined in Bacrot – Rausis, Enghien-les-Bains 1997. 10...fxe6 11.Rd1 White’s hopes are mainly connected with the weakness of the e6-pawn, but Black’s next precise move neutralizes the enemy initiative. 11...Nfd7! Beliavsky & Mikhalchishin overlook this move, only mentioning 11...Qc8?! 12.a4, after which Black soon fell into trouble in Tkachiev – Pelletier, France 2003. The text move is much better, as the attack on the g5-bishop yields Black an important tempo for consolidating.

12.Bf4 0-0 13.Nd4 Rf6 14.Bg3 c5 15.Nf3 Nc6 White had no real compensation for the pawn in Zakharevich – Ionov, St Petersburg 2001. 7...c5 Striking at the centre is clearly the right choice, especially as White’s queen is no longer defending the d4-point.

161

We will analyse C1) 8.Bf4!?, which was played for the first time in 2017, followed by the more established main line of C2) 8.d5. 8.dxc5 This move has been chosen by some strong players but I consider it completely harmless. 8...Bxc5 9.Bxc4 0-0

10.0-0 10.Bg5 has occurred in four games but so far nobody has tested the most logical continuation for Black: 10...Nc6!N 11.Rd1 (even worse is 11.e5? Nd4! 12.Nxd4 Qxd4 13.exf6 Qxc4µ) 11...Qc7 12.Bxf6 gxf6 13.0-0 Rfd8 Black has the better chances due to his control over the dark squares. 10...Nbd7 11.b4 162

After 11.Bg5 Qc7 12.Qe2 Ne5 Black enjoyed excellent piece play in Vaganian – A. Sokolov, Minsk (3) 1986. 11...Be7 12.Bb2 In Cmilyte – Javakhishvili, Predeal 2007, Black missed the strongest continuation:

12...Qc7!N 13.Bd3 a5! White will suffer from her exposed queenside pawn structure. C1) 8.Bf4!?

This fresh idea was introduced by Li Chao in 2017.

163

8...a6! 8...cxd4 was tried some months later in Farago – Petrik, Hungary 2017. For some reason White avoided the critical 9.Nb5!N, which the text move is obviously intended to prevent. 9.Bxc4 b5 10.Bb3 I also examined another natural retreat: 10.Ba2N 10...cxd4 11.Rd1 Nc6 11...Qb6?! 12.Nxd4 Nc6 would invite trouble after 13.Nb3! Be7 14.0-0ƒ.

12.0-0 Be7 13.Qe2 13.e5 Nh5 14.Be3 Qc7 15.Nxd4 Nxd4 16.Bxd4 Nf4 17.f3 Rd8„ gives Black plenty of activity. 13...b4! The immediate 13...Qb6? runs into 14.Nd5 exd5 15.exd5 when Black is in trouble, as his knight has no good squares available. 13...0-0 is playable, although 14.e5 Nh5 15.Be3 Qc7 16.Nxd4 Nxd4 17.Rxd4 g6 reaches a position where the placement of the knight on h5 will not be to everyone’s taste.

164

14.axb4 Qb6 15.Nd5 exd5 16.exd5 Nxb4 That’s the point of Black’s 13th move! 17.d6 0-0 18.dxe7 Rfe8

19.Bc4 d3 20.Qe5 Nfd5= Black will liquidate the e7-pawn, so his problems are solved. 10...cxd4 11.Rd1 We have been following the game Li Chao – Eljanov, Geneva 2017. Black’s most challenging continuation would have been:

165

11...Nc6!N Keeping the extra pawn for the time being. 12.0-0 The ambitious 12.e5? Nh5 13.Bg5 Be7 14.Ne4 runs into 14...Nxe5! 15.Nxd4 Qc8µ when Black is simply a pawn up. The text move gives rise to a complicated position where White’s compensation is based on his better development and the pin along the d-file. Therefore my suggestion is: 12...Qc8!? Inviting simplifications. 12...Be7!? is also possible, for instance: 13.Qe2 Qb6 14.Nd5 exd5 15.exd5 Na5 16.d6 0-0 17.dxe7 Rfe8 18.Bc2 (18.Nxd4 Nxb3 19.Nxb3 Nd5 20.Qe5 Nxf4 21.Qxf4 Rxe7=) 18...Nc4 With mutual chances. 13.Nxd4 13.e5?! Nh5 14.Ne2 Nxf4 15.Nxf4 g6!³ does not offer White adequate compensation for the pawn. 13...Nxd4 14.Rxd4 White has regained the pawn but Black’s pieces are getting more space. 14...Bc5 15.Rd2 0-0= Black has successfully completed his development and has no problems. C2) 8.d5 166

This is the main line. White invests his hopes in the power of the passed pawn, even though Black can blockade it easily in the short term. 8...exd5 9.exd5 Bd6 9...a6 10.Bxc4 b5 has been tried few times by strong players, including Vishy Anand. However, it seems risky in view of the improvement 11.0-0!, which was pointed out by Beliavsky & Mikhalchishin and subsequently tested in a correspondence game. 11...Be7N seems like Black’s best chance. (The game continued 11...bxc4 12.Qa4† Qd7 13.Re1† Kd8 14.Qxc4‚ and White eventually prevailed in Schulz – Irmer, email 2014.) 12.Bf4 0-0 13.d6 Bxf3 14.dxe7 Qxe7 15.Ba2 Bc6 16.Rfe1© White has promising play for the pawn, thanks to his lead in development and powerful bishop pair. 10.Bxc4 White has also tried: 10.Bg5!? 0-0 11.0-0-0 This is a more aggressive set-up, but Black should be fine with careful play. 11...Nbd7 12.Bxc4 Qb8

167

13.Bd3!?N This new move is the best idea I could find for White. 13.Ne4?! only invites unfavourable simplifications: 13...Bf4† 14.Kb1 Nxe4 15.Qxe4 Bxg5 16.Nxg5 Nf6 17.Qh4 h6 18.Ne4 Nxe4 19.Qxe4 Re8 20.Qf3 Qd6³ The d-pawn was more of a weakness than a strength in Pinter – Palac, Tucepi 1996. 13.Kb1 a6 14.Ne4 Nxe4 15.Qxe4 Re8 16.Qg4 occurred in Lautier – Karpov, Monte Carlo (blindfold) 1996, and now with the natural 16...Qc7!N 17.Bd3 g6 18.h4 h5 Black could have seized the initiative. 13...a6 14.Bf5 White intends Ne4 to develop his initiative on the kingside. Play might continue: 14...Qc7 15.Ne4 Nxe4 16.Qxe4 g6 17.Bxd7 Qxd7 18.Ne5 Qe8 19.Bf6

168

19...Bxe5 20.Qxe5 Qxe5 21.Bxe5 Rfd8 22.Bc7 Rxd5 23.Rxd5 Bxd5 24.Bxb6= The resulting endgame is rather drawish.

10...0-0 11.0-0 h6! I agree with Beliavsky & Mikhalchishin’s assessment that preventing the Bg5 pin is the best way to limit White’s attacking abilities. I spent quite a lot of time analysing 11...Nbd7 (or 11...a6, which is likely to transpose after a few moves) 12.Bg5. My eventual conclusion is that Black has no serious problems and can demonstrate equality here as well, but overall his task is easier after my recommended move. 12.Re1 Nbd7!? 12...a6 is a much more common continuation, with most games continuing 13.a4 Nbd7 followed by a quick transposition to one of the lines below where ...a6 and a2-a4 are played. There is not much difference between the two move orders, but I did find an interesting extra option for Black, which is shown in the note to the next move in the main line below.

169

13.b3 This seems like the most logical choice, and is the overwhelmingly most popular choice in the similar position with pawns on a6 and a4. Since the g5-square is no longer available to the bishop, White intends to place it on b2 and develop pressure on the long diagonal. The ambitious 13.Qf5 should not bother Black, as White is unable to create serious threats on the kingside. 13...a6 14.Ba2?! (14.a4N 14...Re8 15.Bd2 Qc7 transposes to the 13.Bd2 line below) 14...b5 15.Be3 Re8 Black’s expansion on the queenside gave White problems in Moes – Ortiz de Latierro, corr. 2010. Another game continued 13.Bd2 a6 14.a4 Qc7 15.Qf5 Rfe8 16.Rxe8† Rxe8 17.h3, and here I found a natural improvement over Black’s play in Timoscenko – Ionov, Vendrell 1996:

170

17...Ne5N 18.Nxe5 Bxe5 Black is doing well, for instance: 19.Re1 Qd6 20.Qd3 Bxc3!? 21.Rxe8† Nxe8 22.bxc3 Nc7= The d5-pawn comes under fire and White has to be accurate to maintain the balance. 13...a6 13...Re8!?N is a natural move which has yet to be tested. After 14.Bb2 I found a nice idea for Black:

14...Ne5! 15.Nd2 Ng6! Rerouting the knight makes Black’s set-up more harmonious. In particular, now there is more pressure on the d5-pawn. 16.Rad1 Rxe1† 17.Rxe1 Qd7 18.Nde4 Be5= 14.a4

14...Qc7 171

Connecting the rooks and taking control over e5. 15.h3 This is definitely a worthwhile prophylactic move. 15.Bb2 This natural move is well met by: 15...Ne5 16.Ne4 Nxf3† 17.gxf3 Nxd5 Now White needs an improvement.

18.Ng3!N 18.Kh1? Be5 19.Ng5 hxg5 20.Rxe5 Nf4 21.Qc3 Qc6 was bad for White in Bologan – Macieja, Stepanakert 2005. 18.Nxd6N 18...Qxd6 19.Rad1 Rfe8 20.Rxe8† Rxe8 21.Qf5 Rd8 also leaves White fighting to equalize. 18...Bxg3 19.hxg3

172

19...Qc6 20.Qf5 Nf6 The bishop pair offers White just enough play for the pawn, but Black has no problems. 15...Ne5 16.Nh4 Rfe8 17.Bb2 This position was reached in Lautier – Karpov, Groningen 1995, and a couple of later games. Considering that the knight may go to f5 at any moment, the following retreat makes a lot of sense to me:

17...Bf8N I will include a few illustrative lines to show that the position is balanced. 18.Nf5 Nxc4 19.bxc4 Qf4! 173

The activation of the queen offers Black decent counterplay. 20.Qd3 20.Ne2 Qe4 21.Qxe4 Nxe4 is fine for Black. 20...Bc8 Depriving the knight of the f5-square is undeniably an achievement for Black.

21.Rxe8 Nxe8 22.Ng3 Nd6 23.Nce4 Nxe4 24.Qxe4 Qxe4 25.Nxe4 Bd7= White’s rook is tied to the defence of the a4-pawn and cannot support the advance of the central passer, so Black has no problems. D) 6.Bg5

174

By pinning the knight, White increases the pressure on the d5-pawn. However, this natural move is not without drawbacks. For instance, the probable exchange of dark-squared bishops should generally make Black’s position easier to handle. 6...Be7 White has a wide choice of possibilities; we will focus on the main options of D1) 7.cxd5, D2) 7.Qa4†, D3) 7.Bxf6 and D4) 7.e3. 7.Rc1 0-0 8.cxd5 Nxd5 9.Bxe7 Qxe7 10.Nxd5 Bxd5 11.e3 c5 (11...Rc8 is also fine) reaches a position covered on page 98; see the M. Gurevich – Aronian game, as given in the note to White’s 9th move in variation D41. 7.Qc2 This has been tried by strong players such as Krasenkow and Korchnoi, but it hardly has any independent value. 7...0-0 8.cxd5 8.e3 has been the most common continuation, leading straight to variation D42. Another transpositional possibility is 8.Rd1 Nbd7 9.e3; see 9.Rd1 in the notes to variation D42 on page 100. 8...Nxd5 9.Nxd5 Bxd5 9...exd5 10.Bxe7 Qxe7 is also perfectly playable, but I would prefer to avoid blocking the b7bishop behind a static central pawn.

175

10.h4!?N This aggressive move seems like an interesting attempt to justify White’s delay in developing his kingside pieces. After 10.Bxe7 Qxe7 11.e4 Bb7 12.Bd3 c5 13.dxc5 Rc8 14.0-0 Rxc5 15.Qe2 Nd7 White was suffering from a slightly vulnerable e-pawn in Cruz – Bregadze, Barbera del Valles 2009. 10...Nd7 11.e4 Bb7 12.Bd3 12.0-0-0?! would be overoptimistic in view of 12...h6 13.Bf4 Nf6 14.Bd3 c5 15.dxc5 Rc8 when White’s king is more exposed. 12...c5 13.e5 h6

14.Bh7† Kh8 15.Be4 Bxe4 16.Qxe4 cxd4 17.Qxd4 Nc5 Black’s king is safe, so White can hardly hope for any advantage. 176

D1) 7.cxd5 Nxd5 8.Bxe7 Qxe7

As I mentioned earlier, exchanging the dark-squared bishops tends to give Black an easy game with no problems equalizing. Several strong GMs have played this way with White, but they must have been aiming for a safe position rather than an opening advantage. 9.Nxd5 Here are a couple more examples where Black equalized with ease: 9.Qa4† Qd7 10.Qc2 Nxc3 11.bxc3 occurred in Andersson – Makarichev, Hastings 1979, when 11...0-0N would have been simplest. Play may continue 12.e3 c5 13.Bd3 h6 14.0-0 Qc7= when Black has no problems at all. 9.Rc1 0-0 10.e3 Nf6 11.Be2 Nbd7 12.0-0 c5 13.dxc5 Nxc5 14.b4 Rfd8 15.Qc2 Nce4= and the players soon agreed a draw in Laznicka – Sargissian, Turin (ol) 2006. 9...Bxd5 10.e3 0-0 11.Bd3 Rc8 12.Rc1 c5 Once again, Black frees his position and is easily equal. A recent high-level game continued:

177

13.dxc5 Rxc5 14.Rxc5 Qxc5 15.Ng5 h6 16.Ne4 Qe7= Dreev – Matlakov, Sochi 2017. D2) 7.Qa4†

Just as in variation A, White can insert this check as an attempt to disrupt Black’s natural development or provoke the ...c6 advance, which would block the Queen’s Indian bishop. 7...Qd7! Both 7...Bc6 and 7...c6 are perfectly playable but I believe the text move to be the best option to maintain the harmony in Black’s set-up. 178

White has two serious replies: D21) 8.Qxd7† and D22) 8.Qc2. 8.Qb3 dxc4 9.Qxc4 0-0 gives Black easy development, for instance:

10.e3 (10.Rd1 Rc8! 11.e3 Qe8! is a plan worth remembering; after 12.Be2 c5 13.0-0 Nbd7 14.h3 a6 15.Qb3 b5³ Black had the more harmonious position in Kukhmazov – Matlakov, St Petersburg 2016) 10...Rc8 11.Be2 c5 12.0-0 This was Dobrov – Tkachiev, Nancy 2013. Black has a number of playable moves but my favourite is:

12...Qe8!N Once again, this regrouping move works well. Sometimes it is used to remove the queen from the d-file; on other occasions, such as here, it vacates the d7-square for the knight. 13.Rac1 Nbd7 14.Qd3 a6 Black is at least equal. 179

D21) 8.Qxd7† Nbxd7 9.Nb5 This is White’s only challenging idea. Instead after something like 9.e3?! c5 Black is already fighting for an edge due to his speedier development.

9...Bd8! Beliavsky & Mikhalchishin only consider 9...Kd8 and conclude that Black can equalize. This assessment seems reasonable, but the text move is a more harmonious way of protecting the pawn. 10.cxd5 Bxd5! 10...Nxd5 has been much more popular, probably because it was the choice of both Gelfand and Anand in a couple of the earliest games in this line. However, I see no reason to allow e2-e4. The text move keeps control over the central light squares and ensures Black of a comfortable game, as the following lines demonstrate. 11.Rc1 11.e3N 11...c5 12.Nc3 Bb7 is perfectly comfortable for Black. 11.Bf4 was seen in Slisser – Nagel, Hoogeveen 2003, when Black should have played:

180

11...c5!N 12.dxc5 Nxc5 13.b4 Ncd7 14.Nd6† Ke7 Black enjoys excellent piece play. 11...c5 12.Bxf6!? 12.e3 occurred in Even – Molzahn, corr. 2006, when 12...a6!N 13.Nc3 Bxf3 14.gxf3 cxd4 15.exd4 b5³ would have reached a position where White’s rotten pawn structure is more relevant than his bishop pair. We have been following Lorincz – Somogyi, Budapest 2001. White’s last move was a reasonable attempt to disrupt Black’s coordination, and in the game it succeeded, as 12...gxf6 13.e4 led to a roughly equal position. Instead, Black could have caused more problems with:

12...Nxf6!N Offering a pawn sacrifice.

181

13.dxc5 bxc5 14.Rxc5 0-0 Black’s lead in development more than makes up for the small material investment. For instance: 15.e3 Ne4 16.Rc2 16.Rc1 Bf6 is also troublesome for White.

16...Bb3 17.Rc1 Ba5† 18.Nc3 Rfd8ƒ Black will easily be able to restore material equality; the only real question is whether or not White will be able to neutralize the pressure and equalize. D22) 8.Qc2 Beliavsky & Mikhalchishin prefer this move, which has been tried by several of the world’s leading players. Black has more than one decent response though.

182

8...c5!? I was surprised to discover that this natural move has yet to be tested at GM level. Indeed, it has only been played in a mere two games out of nearly a hundred from this position. Beliavsky & Mikhalchishin don’t consider it either. Despite all that, attacking White’s pawn centre is an entirely normal plan for this variation, and there are no clear drawbacks to it. The two most popular alternatives are 8...0-0, aiming for solidity, and 8...dxc4 9.e3 Bxf3 10.gxf3 b5, when White has full compensation for a pawn but no real advantage. 9.Rd1!?N There is no real theory to Black’s last move, so I had to develop my own analysis almost from scratch. The text move seems like White’s most interesting attempt to exploit the awkward location of Black’s queen. 9.e3N is playable but harmless: 9...cxd4 10.Nxd4 dxc4 11.Bxc4 0-0 12.0-0 Nc6 13.Nf3 Rfd8 14.Rfd1 Qe8= The first game to feature Black’s last move continued: 9.cxd5 Nxd5 10.Ne5 In Meinhardt – Bungter, Dresden 2006, the best reply would have been:

183

10...Qd8N 11.Bxe7 Qxe7 12.Nxd5 exd5 But not 12...Bxd5? 13.e4 when White has an easy advantage. 13.Qa4† Nc6! Black uses some simple tactics to develop conveniently.

14.Nxc6 14.e3 can be met by 14...0-0 15.Nxc6 Qe8! (but not 15...Qd7? 16.Bb5 a6 17.Ne5 Qxb5 18.Qxb5 axb5 19.Nd7±) 16.Bb5 a6 and Black is fine. 14...Qd7 15.dxc5 bxc5 16.e3 Bxc6 17.Qa5 Qd6= Black’s hanging pawns are not really a problem, as White is not sufficiently well developed to arrange an effective attack on them. An important point is that 18.Rc1 can easily be met by 18...c4, when the backward d5-pawn is certainly no weaker than the one on b2. 184

The other game continued: 9.Ne5 Qc8 And here White needs an improvement.

10.cxd5N 10.Qa4†?! was played in Swapnil – Ramakrishna, New Delhi 2010, but after the correct reply of 10...Nc6!N 11.cxd5 Nxd5 12.Bxe7 Nxc3 13.bxc3 Kxe7 14.e3 Rd8³ White’s scheme has only succeeded in supporting Black’s development. 10...cxd4! Things may appear messy but the position soon clarifies. 11.Qa4† Nbd7 12.Qxd4 exd5 13.Nxd7 Qxd7 14.e3 0-0 Black’s lead in development compensates for the isolated pawn. For instance:

185

15.Bb5 Qf5 16.Bh4 Bc5 17.Qd3 Qxd3 18.Bxd3 d4= Leading to simplifications and easy equality.

9...cxd4 10.Nxd4 0-0 11.e4!? The quiet 11.e3 does not pose many problems. Play may continue 11...dxc4 12.Bxc4 Qc8 13.Bd3 (13.Qe2 Nc6=) 13...h6 14.Bh4 Nbd7 15.0-0 a6 with equal chances. 11...Nc6! 11...dxe4?! is inferior in view of 12.Nf5 Qc7 13.Nxe7† Qxe7 14.Nxe4 Nbd7 15.Nxf6† Nxf6 16.Be2² when White has a powerful pair of bishops. The text move is better; Black offers a pawn sacrifice but, in return, his development advantage starts to 186

tell. 12.Nxc6 Bxc6 13.cxd5 exd5 14.Bxf6 Bxf6 15.Nxd5 Bxd5 16.Rxd5

16...Rac8 17.Qd2 Qc6= Black is poised to regain the pawn, which will result in an equal endgame. D3) 7.Bxf6 Bxf6 8.cxd5 exd5

This approach has been tried by many strong players, including the young Garry Kasparov back in 1979. White has given up the bishop pair in order to close the diagonal of the b7-bishop. His general plan will be to keep the position closed and eventually exert pressure along the c-file. To avoid this, we will 187

almost always play a quick ...c5, even at the expense of a pawn in some cases. We will analyse D31) 9.Qb3, D32) 9.Qa4† and D33) 9.g3. Simple development does not bother Black at all: 9.e3 0-0 The position resembles one of the main lines of the Tartakower-Makogonov-Bondarevsky (TMB) system of the Queen’s Gambit Declined. The only difference on Black’s side is that his pawn is on h7 instead of h6, which is a tiny negative. For White, on the other hand, the developing move Rc1 has been replaced by the far less useful a2-a3, so Black is the clear ‘winner’ of the comparison.

10.Bd3 In the analogous TMB position, 10.b4?! is one of White’s best tries, but here the absence of a rook on c1 means that 10...c5 is a strong reply, after which Black is already better. White can also develop more modestly with 10.Be2, hoping to keep the d-file clear for attacking purposes. Nevertheless, 10...c5 still gives Black a pleasant game. For instance: 11.0-0 Qe7 12.dxc5 (The artificial 12.Qb1 Rd8 13.b4?! Nc6 14.bxc5 bxc5 15.Rd1 occurred in Kozul – Cvitan, Kastel Stari 1997, when the players agreed a draw. However, Black could and perhaps should have played on with 15...Rac8N 16.Qb2 Na5³.) 12...Bxc3 13.bxc3 We have been following Ghorbani – Taleb, Tagaytay City 2004. Here I prefer the natural 13...Qxc5N in order to keep the c-file open. The arising position is balanced, as the isolated pawns on c3 and d5 are equally weak. The text move is White’s most ambitious method of development. On the other hand, the bishop blocks the d-file and thus decreases White’s control over the centre.

188

10...c5 11.0-0 g6 12.Rc1 Nc6 13.Bb1 c4!? 13...cxd4 14.exd4 Rc8 was good enough to maintain the balance in Keuter – Schuetze, email 2007. The text move is more ambitious. 14.Re1 14.b3N looks like an improvement although 14...Na5 15.bxc4 dxc4!? (or 15...Nxc4) leaves Black with the easier position to handle. 14...Bg7 15.b3 In Adianto – Karpov, Cap D’Agde (rapid) 1998, Black should have continued with:

15...Na5!N 16.b4 16.bxc4 dxc4³ is also promising for Black. 16...Nb3 17.Rc2 a5 18.b5 a4 189

White is doomed to passive defence in an inferior position. D31) 9.Qb3 0-0 10.Rd1 White’s plan is clear: he is playing prophylactically to prevent Black from attacking the centre.

10...c5! It turns out that Black can play the thematic freeing break, regardless of the measures White has taken. 11.dxc5 This is the only consistent move. In several games White has backed out of the challenge with 11.e3, effectively admitting that his 9th and 10th moves were misguided. Black has several decent continuations but the most straightforward is 11...c4 12.Qc2 Nd7 (or 12...a6N immediately) 13.g3 a6 14.Bg2 b5 15.0-0 Re8 when Black had a pleasant game in Krasenkow – Beliavsky, Kallithea 2008. 11...Nd7! Black develops another piece and prepares to gain further time against the enemy queen. 11...d4 12.cxb6 Na6 was given by Krasenkow in his annotations to the aforementioned game, but I prefer the text move. 12.cxb6!N This brave decision seems to be White’s only chance to maintain the balance. After 12.c6? Bxc6 13.Nxd5 Bxd5 14.Rxd5 Rc8 15.Rd1 Nc5µ White suffered from a severe lack of 190

development in Boros – Friedel, Saint Louis 2014, and his extra pawn was hardly relevant.

12...Nc5!? This is the most dynamic continuation. The simpler 12...Qxb6 13.Qxb6 Bxc3† 14.bxc3 axb6= invites major simplifications and would most likely lead to a draw. 13.Qc2 Qxb6 14.e3 d4! 15.Nxd4 Rad8 16.Bc4 Bxg2 17.Rg1 Bb7 Black has at least enough compensation for a pawn, and White faces the greater practical challenges due to his uncastled king. D32) 9.Qa4†

191

Yet again, we see that White may throw in this queen check in order to interrupt Black’s natural development. 9...Qd7 This is my preference, although there is nothing wrong with 9...c6. 10.Qb3!? So far this move has only been played once, but I find it to be White’s most interesting try. 10.Qxd7† Nxd7 11.Nb5 Bd8 12.g3 a6 13.Nc3 Bf6= was comfortable for Black in Cu. Hansen – Tiviakov, Esbjerg 2002. 10.Qc2 0-0 11.e3 c5 12.0-0-0 cxd4 13.Nxd4 Nc6 was promising for Black in Kozul – Sax, Bled 1999. Finally, 10.e3 0-0 11.Qxd7 Nxd7= was harmless in Sargissian – Macieja, Batumi 2002. 10...c5 11.Rd1 We have been following Bacrot – Grischuk, Cap d’Agde (rapid) 2003. In the game Black played 11...c4 and eventually won, but I prefer a more flexible and dynamic way of handling the position:

192

11...0-0!N 12.dxc5 12.e3 c4 13.Qc2 Na6 14.Be2 b5 gives Black the upper hand on the queenside. 12...d4 13.cxb6 Rd8 14.e3 The greedy 14.bxa7?! Rxa7 15.e3 (even worse is 15.Nb5? Bd5 16.Qd3 Rb7 17.Nfxd4 Nc6–+) 15...Bxf3 16.gxf3 Rb7 17.Nb5 Na6 offers Black a powerful initiative.

14...Qe7 15.Nb5 Bxf3 16.gxf3 axb6 Black’s development advantage fully compensates for the missing pawn, to say the least. D33) 9.g3 193

This was Kasparov’s choice in 1979 and it seems like White’s best. We have already seen that White’s attempts to prevent ...c5 are doomed to fail, so White may as well focus on kingside development. 9...0-0 10.Bg2 c5 11.0-0 Na6! Transferring the knight to e6 has proven to be the most effective way of handling the position. 11...Nc6?! 12.dxc5 Bxc3 13.bxc3 bxc5 14.Rb1² gave White strong pressure against Black’s pawns in Sulava – Olivier, Geneva 1996. The merits of White’s set-up are also well-illustrated in the following case: 11...Nd7 12.e3 Rc8 Now in Cramling – Zhao Xue, Beijing (blitz) 2012, White could have provoked a favourable change in the pawn structure with 13.Qb3!N 13...c4 14.Qc2. The resulting position resembles the Krasenkow – Beliavsky game, as given in the notes to variation D31, but here Black is too slow to get coordinated. For instance, 14...a6 is met by 15.Nd2! with an awkward attack on d5; and if 14...Be7 (preparing ...Nf6) 15.Ne5 White seizes the initiative.

194

12.e3 This has been by far the most popular choice; indeed, control over the d4-square is crucial in positions of this type. After 12.Qd2 Qe7 White found nothing better than 13.e3 anyway, and 13...Rfd8 14.Rfe1 Nc7 15.h4 h5 16.Kh2 g6 was fine for Black in Korchnoi – Ionov, Ohrid 2001. Another game continued: 12.Rc1 Qe7 13.e3 I also examined: 13.dxc5N 13...Nxc5 14.Nxd5 (14.e3 Ne6 15.Qd2 Rfd8=) 14...Bxd5 15.Qxd5 Rad8 (15...Qxe2 is also fine) 16.Qc4 Bxb2 17.Rc2 Bf6= 13...Rfd8 14.Re1 Nc7 15.Re2 This was Almeida – Li Shilong, Barcelona 2004. Out of many playable moves, I like the following ambitious approach:

195

15...c4!?N Hoping to mobilize the queenside majority. Play might continue: 16.e4 dxe4 17.Nxe4 Ne6 18.Nxf6† Qxf6 19.Rd2 b5 Black is by no means worse. 12...Qe7 13.Re1 13.Qd2 and 13.Rc1 are covered above, via 12.Qd2 and 12.Rc1 respectively. White’s last move anticipates the opening the e-file. Moreover, the rook may be heading for d2 in order to put more pressure on the d5-pawn. 13...Rfd8 14.Qa4 14.Qb3 is no better; after 14...Rab8 15.Rad1 Nc7 16.Rd2 g6 17.h4 Bg7 Black had excellent play in Cu. Hansen – Bruzon, Esbjerg 2003. 14...Nc7 15.Rad1 Ne6 16.h4 We have been following Kuljasevic – Gabrielsen, Oslo 2004. Here I prefer the typical prophylactic measure:

196

16...g6N 17.Nh2 h5„ Securing the dark-squared bishop and intending to push the queenside pawns. D4) 7.e3 0-0

Once again, we have something resembling the Tartakower-Makogonov-Bondarevsky system. Here too, Black has missed out on the chance to include ...h6, but the more relevant detail is that White has spent a whole tempo on a2-a3. Therefore Black should have an easier time equalizing here than in the TMB proper. We will analyse D41) 8.Rc1 and D42) 8.Qc2. 197

8.Bxf6 Bxf6 9.cxd5 exd5 has already been covered on page 93 – see 9.e3 0-0 in the notes to variation D3. 8.cxd5 Nxd5 9.Bxe7 Qxe7 10.Nxd5 Bxd5 11.Rc1 c5 (11...Rc8 is also fine) transposes to M. Gurevich – Aronian, as referenced in the note to White’s 9th move in variation D41 below. 8.Bd3 (or 8.Be2) enables Black to gain a tempo on the bishop with 8...dxc4 9.Bxc4, leading to easy equality. For instance, 9...Nbd7 10.0-0 c5 11.Qe2 Ne4 12.Nxe4 Bxe4 13.Bxe7 Qxe7= and White had nothing in Kamsky – Karjakin, Tashkent 2012. D41) 8.Rc1 White delays any movement of his light-squared bishop and aims to develop pressure against the opponent’s central pawns. 8...c5

9.dxc5 After 9.cxd5 Nxd5 10.Bxe7 Qxe7 11.Nxd5 Bxd5! 12.dxc5 bxc5 Black’s broken queenside structure is no real problem, and 13.Bc4 Bc6 14.Qe2 Nd7= gave him plenty of counterplay along the b-file in M. Gurevich – Aronian, Saint Vincent 2005. 9...bxc5 10.cxd5 10.Be2, as was played in Le Quang – Matamoros Franco, Las Vegas 2014, can be well met by:

198

10...dxc4!N Leading to a thematic structure where Black is fine. 11.0-0 h6 12.Bh4 (in the event of 12.Qxd8 Rxd8 13.Bh4 Nc6 14.Bxc4 g5! 15.Bg3 Nh5 Black gets the advantage of the bishop pair, so White faces a struggle to avoid becoming worse) 12...Nd5 13.Bxe7 Qxe7 14.Bxc4

14...Rd8 15.Qe2 Nb6 16.Bb5 N8d7 17.Rfd1 Rac8 18.Ba6 Nf6= Black is fine; the pawns on c5 and b2 are equally weak. 10...Nxd5!?N 10...exd5 is also playable but the text move fits perfectly with some of the other lines we have seen, where Black adopts the same pawn structure and has no problems.

199

11.Qb3!? 11.Bxe7 Qxe7 transposes to some existing games. A good illustrative example continued 12.Be2 (12.Nxd5 Bxd5 takes us back to the M. Gurevich – Aronian game quoted earlier) 12...Rc8 13.0-0 Nd7 14.Qa4 c4 15.Rfd1 N7b6 when Black’s active piece play fully compensated for the pawn weaknesses in D. Kovalev – Zubov, Mukachevo 2013. 11...Nb6 12.Bxe7 Qxe7 13.Be2 N8d7 14.0-0

14...c4! Not the only decent move, but it’s the cleanest way to demonstrate that Black is not worse.

200

15.Bxc4 Bxf3 16.gxf3 Nxc4 17.Qxc4 Ne5 18.Qe4 Qg5† 19.Kh1 Qh5

White has nothing better than repeating moves with: 20.Kg2 Qg5† 21.Kh1= D42) 8.Qc2

Like the previous variation, White is making a constructive move while avoiding any movement of the light-squared bishop, which would lose a tempo to ...dxc4. 8...Nbd7!? 201

I also considered 8...c5 9.dxc5 bxc5 10.cxd5 Nxd5 (10...exd5 is also possible) 11.Bxe7 Qxe7 and, although Black has a playable position with the thematically broken queenside structure, I prefer the text move, which sets up an interesting pawn sacrifice. 9.cxd5 The most consistent. I also examined 9.Rd1 c5! when White may try to exploit the pressure along the d-file, but Black’s lead in development always provides enough counterplay. The critical line continues: 10.dxc5N 10...Nxc5 11.Bxf6 Bxf6 12.cxd5 exd5 13.Bc4

13...Rc8! 14.Nxd5 b5! 15.Nxf6† (15.Ba2?! Na4 16.Qf5 Bxb2 17.0-0 Rc5³) 15...Qxf6

202

16.Bxb5 Bxf3 17.gxf3 Qxf3 18.Rg1 Qh3 With White’s king stuck in the centre, Black has plenty of counterplay.

9...Nxd5 9...exd5 is playable but I prefer to keep the centre more fluid. 10.Bxe7 Qxe7! 10...Nxe7 is pretty solid but I prefer to fight for the initiative with a pawn sacrifice, which is justified by Black’s lead in development. 11.Nxd5 Bxd5 12.Qxc7 Against the toothless 12.Bd3, the correct path was shown as long ago as 1940! 12...Bxf3 13.gxf3 Nf6 14.Rc1 c5 15.dxc5 bxc5

203

16.Qxc5 (16.0-0 Qb7 17.f4 Rab8=) 16...Qb7 17.Qe5 Qxf3 18.Rg1 Rac8 Black was at least equal in Sämisch – Normann, Bad Elster 1940.

12...Rac8! The problem with 12...Rfc8 is shown in the following line: 13.Qf4 Rc2 (13...Nf6 14.Bd3 Bc4 15.Bxc4 Rxc4 16.0-0 left Black with minimal compensation in Riazantsev – B. Socko, Riadh [blitz] 2017) This position occurred in Alvarado Rodriguez – Liberadzki, Warsaw 2012, when 14.e4!N would have favoured White. The difference between this and our main line becomes clear after 14...Bb7 15.Bd3! when 15...Rxb2?? is refuted by 16.Qc7!, attacking the b7-bishop while also threatening Qc3, trapping the rook.

204

13.Qxa7 Another natural continuation is: 13.Qf4 Rc2 14.e4 Ba8! In contrast to the line given in the last note, Black slides the bishop away from a possible Qc7 attack, which means that the threat to the b2-pawn is real. 15.Rb1

15...f5! 16.e5 Rd8 17.Bd3 Rc7 18.0-0 Nf8 Black had full compensation in Jarabinsky – Terekhov, corr. 2011. The d4-pawn is weak and most of White’s pieces are rather passive. 13...Ra8 14.Qc7 Rfc8 Amusingly, we have almost the same position as in the 12...Rfc8 line, except that here the a7-pawn has been removed. As you have probably guessed, this is no bad thing, as the open a-file produces additional tactical opportunities.

205

15.Qf4 Rc2 16.Rb1 16.Bd3?? is hit with a sledgehammer: 16...Qb4†! 17.Kf1 (or 17.axb4 Rxa1#) 17...Qxb2 and White can resign. 16.e4? also enables Black to exploit the open a-file in a similar fashion: 16...Qb4†! 17.Nd2 Qxd4 and although White’s situation is not quite as devastating as in the line above, Black is still winning. 16...Ba2 17.Rd1 Rxb2 18.Bd3 Rxa3 19.0-0 Bd5 White has caught up with development while Black has regained both of the sacrificed pawns, so the chances are balanced.

206

20.e4 Rxd3 21.Rxd3 Bc4= The players soon agreed a draw in Dzenis – Packroff, corr. 2013. Conclusion The Petrosian System arises after 4.a3 Bb7 5.Nc3 (or 4.Nc3 Bb7 5.a3). This chapter has dealt with an assortment of sidelines after our classical reply of 5...d5. I recommend meeting 6.Qa4† with 6...Qd7, and 6.Bf4 with 6...Bd6. In both cases, Black offers to exchange the piece White has just moved, in order to speed up his own development. Next we analysed 6.Qc2, when 6...dxc4! 7.e4 c5 is the principled reply, leading to interesting play after the recent invention of 8.Bf4!? as well as the more traditional 8.d5 exd5 9.exd5. In the latter case, it is worth remembering that a timely ...h6 is a good way to limit White’s activity by preventing Bg5. Finally we analysed 6.Bg5 Be7. It is hard to give much general advice here, as many options and ensuing pawn structures are possible. It is useful to remember that, like in variation A, 7.Qa4† should be met by 7...Qd7!. Another thematic structure occurs after 7.Bxf6 Bxf6 8.cxd5 exd5, when Black gets an active game by preparing a quick ...c5. Finally, 7.e3 0-0 leads to something resembling a major variation of a Queen’s Gambit Declined, with White having spent a tempo on a2-a3, which considerably simplifies Black’s task of equalizing.

207

A) 7.e4?! 104 B) 7.Qa4† 106 C) 7.Nxd5 Qxd5! 110 C1) 8.g3 110 C2) 8.e3 112

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.a3 Bb7 5.Nc3 d5 6.cxd5 This is White’s most popular choice by a considerable margin. 6...Nxd5 As usual, we avoid the more rigid ...exd5 pawn structure where possible. 208

This chapter will deal with some minor options: A) 7.e4?!, B) 7.Qa4† and C) 7.Nxd5. 7.Bd2 and 7.e3 are analysed in the next chapter, while the big main line of 7.Qc2 can be found in Chapter 8. 7.Bg5 Be7 leads back to variation D1 of the previous chapter. 7.g3 The fianchetto set-up is playable but it contains no real venom here, as Black is well placed to contest the light squares. 7...Nxc3 8.bxc3 Be7 9.Bg2 0-0 10.0-0 c5 11.Qc2 This seems like White’s best try. After 11.Be3 Nd7 12.Qd3 Qc8 13.a4 Nf6 Black had a comfortable game in Al-Zendani – Le Quang, Guangzhou 2010.

11...cxd4 11...Nd7?! allows a thematic trick: 12.Ng5! Bxg5 13.Bxb7 Rb8 14.Bg2² White had a small edge with virtually no risk in Vi. Kovalev – Avdeenko, Tomsk 2008. 12.cxd4 12.Ng5? d3! 13.Qxd3 Qxd3 14.exd3 Bxg2 15.Kxg2 Rd8µ leaves White in a depressing endgame due to his pawn weaknesses. 12...Nc6 13.Rd1 Rc8 14.Qd3

209

This occurred in Loureiro – Sunye Neto, Sao Jose de Rio Preto 1995. The simplest continuation is: 14...Bf6N= Black has no problems. A) 7.e4?! Nxc3 8.bxc3 Bxe4

This pawn sac has been tried by several strong players but White’s compensation is questionable. 9.Ne5 Qh4! This move may appear artificial but it serves an important purpose in limiting White’s activity on the kingside – especially with regard to the queen. 210

9...c6? cannot be recommended. True, after 10.Qe2 Bg6 11.h4 Qd5 Black went on to win in Piket – Korchnoi, Roquebrune 1992, although White certainly has compensation at this stage. However, 10.Qh5! is a finesse which puts Black’s 9th move out of business. The point is revealed after 10...Qc7N (10...g6 11.Qe2 Bf5 12.g4 Qd5 13.gxf5! Qxh1 14.Nxf7!+– was devastating in Hart – B. Watson, Auckland 2010) 11.Qe2 Bg6 12.h4± when Black is in trouble. 10.g3

10...Qd8! The queen is not really wasting time, as White’s development has been disrupted and he now has to spend another tempo safeguarding his rook. 10...Qf6 is less accurate; after 11.Bb5† c6 12.f3! Bd5 13.Be2 b5 14.a4 White has promising play for the pawn. 11.Rg1 11.Bb5†? c6 12.f3 cxb5 13.fxe4 Nd7µ was poor for White in A. Mikhalevski – A. Sokolov, Biel 1992. 11.Qa4†?! is not much better. Play continues 11...c6 12.f3 Bd5 13.c4 and now a serious improvement is:

211

13...b5!N (in the game Black tried to get too clever with 13...Be4?, after which 14.Be3! left White with a strong initiative for the pawn in Kopasov – M. Kaufmann, email 2003) 14.cxb5 Bd6µ Black has returned the extra material to reach an excellent position where White suffers from serious weaknesses. We have been following the game Aoiz Linares – Veingold, Barcelona 1992. Black has a few good options but the most promising seems to be:

11...Bd5!N 12.Bb5† Black would be happy to provoke 12.c4, when 12...Bb7 leaves the light-squared bishop restricted, thus limiting White’s active possibilities. Play might continue 13.Qa4† c6 14.Be3 Be7 15.Bg2 0-0µ when, aside from being a pawn down, White’s king faces an uncertain future.

212

12...c6 13.Bd3 b5! Securing the future of the excellent bishop on d5. 14.a4 a6 15.axb5 cxb5 My analysis continues:

16.c4 bxc4 17.Qa4† Nd7 18.Bxc4 Bxc4 19.Nxc4 Be7 20.Ne5 f6 21.Nc6 Qc7³ The game goes on, but White is clearly struggling to find compensation for the pawn. B) 7.Qa4†

We have, in the previous chapter, already seen a few incarnations of this queen check, which White 213

hopes will disrupt our development and provoke a concession of some kind. This approach was employed by Garry Kasparov back in 1983, and has been used by many other GMs. 7...Nd7 I favour this natural developing move. 7...Qd7 8.Qc2 Nxc3 9.bxc3 leaves Black with a less harmonious set-up compared with variation B of Chapter 8. 7...c6 is playable and has scored well for Black; nevertheless, I regard this move as a slight concession. 8.Nxd5 8.Ne5?! Nxc3 9.bxc3 Bd6 gives White absolutely nothing, for instance:

10.Nxd7 (10.Nc6?? could have led to disaster for White in Petronic – Ostojic, Belgrade 1989, if Black had only found 10...Qh4!N with the deadly threat of ...Nc5. White has no real choice but to retreat with 11.Nb4, when 11...c5–+ leaves him hopelessly uncoordinated and behind in development.) 10...Qxd7 11.Qxd7† Kxd7 12.f3 f5 13.e3 c5 Black had the more pleasant game in Dzagnidze – A. Muzychuk, Khanty-Mansiysk 2014. 8...Bxd5 8...exd5 was Korchnoi’s choice, which brought him a draw against Kasparov in Game 3 of their 1983 Candidates match. However, blocking the diagonal of the b7-bishop is an indisputable achievement for White, so most strong players have preferred the bishop recapture.

214

9.Qc2 This is the most ambitious try; White wants to establish a strong pawn centre. 9.Bg5 This move is popular yet harmless. Azmaiparashvili has played it three times; strangely, his opening play became worse each time, as shown in the examples below. 9...Be7 10.Bxe7 Qxe7 11.Rc1?! This is the most interesting move to analyse, as long as it’s from Black’s side of the board! White is playing with fire, attacking a pawn at the expense of his development. 11.Ne5 is safer, when 11...a6 12.Qxd7† Qxd7 13.Nxd7 Kxd7 14.f3 f5 gave Black comfortable equality in Azmaiparashvili – Grischuk, Moscow (rapid) 2002, which was the first of the aforementioned games of the Georgian GM.

215

11...0-0! Naturally we can sacrifice the c-pawn. 12.Rxc7 Obviously this is the critical move to consider. It’s extremely risky though: White not only expends another tempo, but also opens the c-file which Black can now use to invade. I checked two other ideas: a) 12.e3 is safer but Black is at least equal after: 12...c5 13.Bb5 Nf6 14.dxc5 bxc5 15.0-0 Rab8 16.Qa5 This was Buhmann – Cvitan, Pula 2003, and now 16...Rb6!N³ would have caused White some problems, with ...Rfb8 or perhaps ...Bxf3 followed by ...Qb7 coming soon. b) 12.e4? was Azmaiparashvili’s bizarre attempt to improve, but it can be refuted by: 12...Bxe4 13.Rxc7

216

13...Rfc8!!N (13...Nc5!? 14.dxc5 Qxc7 15.Qxe4 Qxc5 gave Black a good position with rook against two minor pieces in Azmaiparashvili – Macieja, Ermioni Argolidas 2006, but the text move is even better) 14.Rxd7 Qf6 Threatening to win the rook with ...Bc6, while also setting up attacking ideas against White’s king. 15.Qd1 Qf4 16.Bc4 (16.Nd2 Bc6 traps the rook) 16...Bxf3 17.Qxf3 Qc1† 18.Ke2 Qxc4† 19.Ke3 Rf8µ White will be hard pressed to keep his position together with his king placed in such a way.

12...Nc5! Conveniently escaping the pin while activating the knight. 13.Rxe7 Nxa4 14.Kd2 14.b3 Bxb3 15.Nd2 Ba2 16.e4 Rfc8 was horrible for White in Bonin – Adorjan, New York 1986. 14...Rfc8 15.Ng5 Bb3 Black had a dangerous initiative in Azmaiparashvili – Kramnik, Cap D’Agde (rapid) 2003.

217

9...Be7 The other natural continuation 9...c5 10.e4 Bb7 11.Bf4! offers White some more activity. 10.e4 10.Bf4N 10...Rc8 11.e4 Bb7 transposes to the main line. 10...Bb7 11.Bf4 11.Bb5 has achieved a plus score for White but 11...0-0 12.Bc6 Nc5! is a nice resource which enables Black to unblock the c-pawn at once. 13.Bxb7 (after 13.dxc5N 13...Bxc6 14.cxb6 Qd6 15.bxc7 Rac8 16.0-0 Qxc7 17.Nd4 Bb7 Black’s excellent bishops provide full compensation for the pawn) 13...Nxb7 14.0-0 c5 15.Rd1 Qc7 16.dxc5 Here I found a slight improvement over Mchedlishvili – Werle, Emsdetten 2010:

218

16...Qxc5N 17.Qe2 Rfd8= Black has no problems.

11...Rc8 11...c5N is possible although 12.dxc5 gives Black something to think about, since 12...Nxc5 13.Bb5† forces the king to move. The text move is a simpler solution. 12.Rd1 Other continuations don’t bother Black either, for instance: 12.Bb5N 12...0-0 13.Bc6 Bxc6 14.Qxc6 Nb8 15.Qa4 Qd7 16.Qxd7 Nxd7 17.Ke2 c5 is equal. 12.Bc4N 12...Nf6 13.Bb5† c6 14.Bd3 c5 15.Qe2 0-0 16.dxc5 Rxc5 17.0-0 Qa8= also gives Black no 219

problems. 12...0-0 13.Bd3 13.Bb5!? c6 14.Be2 occurred in Sanikidze – Matlakov, Gjakova 2016. My new idea is 14...Nf6N 15.00 c5! when any problems along the d-file are illusory, for instance:

16.dxc5 Rxc5! 17.Qa4 Qa8 18.e5 Bc6 19.Qb3 Nd5 Black has a comfortable game and the e5-pawn might become weak in the long run.

13...c5 14.d5 White relies on the power of the passed pawn. Keeping the c-file closed is also desirable for him.

220

14...c4! Fortunately, the rook still has some value on c8. 15.Be2 exd5 16.exd5 Bf6 17.0-0N White should not keep his king in the centre any longer. The over-aggressive 17.h4 Re8 18.Ng5 Nf8³ led White nowhere in Kincs – Amstadt, Zalakarosi 2008.

17...Re8 18.Be3 a6 19.a4 h6 White’s pieces are rather ineffective and don’t support the passer, so Black is at least not worse. C) 7.Nxd5 Qxd5! White’s opening play would be fully justified after 7...Bxd5 8.Qc2, when the position is the same as variation B except that Black’s knight is on b8 instead of d7. Play may continue 8...Be7 9.e4 Bb7 10.Bf4² and White is doing well.

221

The text move is clearly best: the queen is active yet not vulnerable to attack, and Black has good control over the central light squares. We will consider C1) 8.g3 and C2) 8.e3. 8.Bf4 has been played several times but it proves harmless after 8...Bd6, for instance: 9.Bxd6 Qxd6 10.e3 0-0 11.Bd3 Nd7= Nutiu – Parligras, Baile Tusnad 1999. C1) 8.g3

This has been quite a popular choice; apparently the centralized queen is an inviting target. However, White comes under pressure in the centre after the following strong reply. 222

8...Nc6! Another attractive option is: 8...c5!? 9.Be3 9.Bg2?! Nc6 10.Qa4 was seen in Mietner – Wegener, Recklinghausen 1999, when Black should have played 10...0-0-0!N 11.dxc5 b5!, punishing White for keeping the king in the centre for too long. For instance: 12.Qc2 Nd4 13.Qd3 Nb3 14.Rb1 Qxd3 15.exd3 Rxd3µ 9...Nc6 10.dxc5 Qxd1† 11.Rxd1 Bxc5 12.Bxc5 bxc5

13.Rc1 13.Bg2 Nd4 14.Nxd4 Bxg2 15.Rg1 cxd4 16.Rxg2 0-0-0 is equal. 13...Ke7 14.Rxc5 Rac8 Black’s development advantage fully compensated for the pawn in Salvatore – Wassilieff, corr. 2007.

223

9.Be3 This is White’s only way to maintain the balance. 9.Bg2?! This is the move White would like to play, but the following complications favour Black. 9...Nxd4! 10.Nh4 Qa5† 11.b4 Bxb4† 12.axb4 Qxa1 13.Bxb7 Rd8 14.Kf1 This position was reached in S. Ivanov – Anastasian, St Petersburg 1994, and a couple of subsequent games. Black can obtain a clear advantage with the following improvement:

14...Qc3!N 15.Qd2 15.Bf4 0-0µ 15...Qxd2 16.Bxd2 Nb3 17.Bc6† Ke7 18.Bc3 Rd1† 19.Kg2 Rxh1 20.Kxh1 Rd8µ 224

Black has excellent winning chances due to his extra queenside pawns, which will soon become passers. 9...0-0-0 10.Bg2 e5 11.dxe5

11...Nxe5N This simple innovation is a safe equalizer, which sees Black regain the pawn in a comfortable situation. The more complicated alternative is: 11...Qxd1†!? 12.Rxd1 Rxd1† 13.Kxd1 Nxe5 14.Bh3†! (after 14.Rg1 Nxf3 15.Bxf3 Bxf3 16.exf3 Be7³ White suffers from an inferior pawn structure) 14...Kb8 15.Nxe5 Bxh1 16.Nxf7 Rg8 17.f3 The position is unclear and holds mutual chances, and a draw ensued in Uberos Fernandez – I. Jones, corr. 2016. 12.Qc2 In the event of 12.Qxd5 Rxd5 (but not 12...Bxd5? 13.Bh3† Nd7 14.0-0-0ƒ) 13.0-0 Nxf3† 14.Bxf3 Rd8= Black has nothing to worry about. I also checked 12.Qa4 Kb8 13.0-0 Nxf3† 14.Bxf3 Qd7 15.Qc2 (15.Qxd7 Rxd7=) 15...Bxf3 16.exf3 Qd3 17.Qa4 Qd7= when the activity of White’s pieces can be neutralized, while Black’s pawn structure is preferable in the long run. 12...Qa5† 13.Kf1 13.Bd2 Nxf3† 14.Bxf3 Qc5 is also level.

225

13...Ng4 14.Bd4 Bxf3 15.exf3 Rxd4 16.fxg4 Qc5= Black has no problems and the opposite-coloured bishops make a draw a likely outcome. C2) 8.e3

This normal move was played in several game, including Fedorovtsev – Smirnov, St Petersburg 2005. Surprisingly, I can offer a normal developing move as a novelty. 8...Be7N 9.Bd3 c5 10.e4 Qd6 11.0-0 0-0 11...cxd4 is also good enough, for instance: 12.e5 Qd7 13.Nxd4 0-0 14.Be3 Nc6=

226

12.e5 12.dxc5 can be met by 12...bxc5!?, in the style of some other lines we have seen, such as D41 from the previous chapter. Here too, after 13.Be3 Nc6= Black’s control over the d4-square and future play along the b-file compensates for the structural drawbacks. 12...Qd7 13.Qe2 Nc6 14.dxc5 bxc5! By now you should be expecting this choice of recapture! 15.Rd1 Qc7 16.Qe4 g6 Black is not worse at all. Before completing development, White has to move his queen to avoid a nasty discovered attack.

227

17.Qf4 Rfd8 18.Be3 Rd5 The vulnerability of the e5-pawn prevents White from developing any attack on the kingside. 19.Be4 Rxd1† 20.Rxd1 Rd8

21.Rc1 Qd7 22.h4 Other pawn moves on the kingside should be met in the same way. 22...Nd4„ White can exchange the knight with either his knight or his bishop; either way, Black gets a passed pawn and a full share of the chances. 228

Conclusion This short chapter has dealt with a few sidelines after 6.cxd5 Nxd5, beginning with the pawn sacrifice 7.e4?!. There is no doubt that White’s gambit is objectively unsound, so all you have to do is remember some key lines and remain vigilant at the board. 7.Qa4† is not too challenging although please remember that, unlike the previous chapter where we blocked this check with ...Qd7, here we should prefer 7...Nd7 and recapture on d5 with the bishop. Finally we considered 7.Nxd5 Qxd5! when Black’s queen is rather well placed in the centre, as evidenced by the lines after 8.g3 Nc6! when White must play accurately to maintain the balance. 8.e3 is safer but this is clearly not an opening variation which will cause Queen’s Indian players to lose sleep.

229

A) 7.Bd2 Nf6 115 A1) 8.Bg5 116 A2) 8.Qc2 117 B) 7.e3 Be7 118 B1) 8.Qc2 119 B2) 8.Bb5† c6 121 B21) 9.Nxd5 122 B22) 9.Bd3 c5! 124 B221) 10.0-0 124 B222) 10.Nxd5 125 B223) 10.e4 Nxc3 11.bxc3 Nc6 12.0-0 cxd4 13.cxd4 0-0 14.Be3 Rc8 15.Qe2 Na5 127 B2231) 16.Rfd1 129 B2232) 16.Rfe1 131

230

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.a3 Bb7 5.Nc3 d5 6.cxd5 Nxd5 Having analysed a few sidelines in the previous chapter, we will move on to the more important options of A) 7.Bd2 and B) 7.e3. 7.Qc2 is the most popular move of all, and will receive dedicated coverage in the next chapter. A) 7.Bd2

This modest-looking move was first seen in 1983. Since then, it was successfully employed by many strong players including Tal, Romanishin, Tukmakov and more. The main idea behind it is to prepare e2e4 while anticipating ...Nxc3, at which point White’s bishop will take up an active post on c3. 7...Nf6 There are a few other decent options, 7...Be7 and 7...Nd7 being the most popular. I like the text move though; Black avoids the ...Nxc3 exchange and takes control over the e4-square, while preparing an 231

effective ...c5 break. Beliavsky & Mikhalchishin do not mention this possibility in their 2008 book, even though it had been tested by players such as Anand and Kramnik prior to that. White’s two most important options are A1) 8.Bg5 and A2) 8.Qc2. 8.Bf4 is harmless: 8...Bd6 9.Bg3 (White also gets no advantage after 9.Bxd6N 9...cxd6 10.e3 0-0 11.Bd3 a6 12.0-0 b5 13.e4 Nc6=) 9...0-0 10.e3 Nbd7 11.Rc1 Bxg3 12.hxg3 c5= Black had nothing to worry about in Gelfand – Leko, Moscow (blitz) 2009. 8.Rc1!? This option is not without venom, but Black should be fine after a few careful moves. 8...Be7 8...c5 is well met by 9.e4! a6 10.e5 with active play for White, as in Ding Liren – Hou Yifan, Shenzhen (rapid) 2011. 9.Qa4†!? 9.Bf4 0-0 10.e3 was seen in A. Petrosian – Lerner, Riga 1985. My suggestion would be 10...Nbd7N 11.Be2 c5 12.0-0 Nh5 13.Be5 cxd4 14.Bxd4 Nhf6= and Black is fine. The text move is more critical. In Farago – Schneider Zinner, Lodi 2008, Black went for safety with 9...c6, but he could have solved his opening problems more convincingly with:

9...Nbd7!N 10.Ne5 c5 11.dxc5 Bxc5 12.Rd1 The threat is Bg5, so... 12...h6! 13.Bf4 a6 14.Nxd7 Nxd7 15.Ne4 b5 16.Qc2 Bxe4 17.Qxe4 0-0 18.g3 Nf6 Black has escaped from the nasty pin, and now it is White who needs to be careful due to his lag in development. A1) 8.Bg5

232

A common continuation. 8...Be7 9.Qa4†!? 9.e3 proved to be harmless after 9...0-0 10.Be2 Nbd7 11.0-0 h6 12.Bf4 Bd6= in S. Ivanov – Roiz, Dagomys 2009.

9...Qd7!N I was surprised to discover that this natural move has yet to receive a practical test. The arising position resembles variation D2 of Chapter 5, where we meet 6.Bg5 Be7 7.Qa4† with 7...Qd7!; see page 89 for a reminder. The strategic idea behind White’s previous move is clearly seen in the following game: 9...Nbd7 10.Ne5 0-0 11.Nc6 Bxc6 12.Qxc6 Nd5 13.Bd2 Nxc3 14.Bxc3² White was comfortably better due to the bishop pair in Dreev – Sakaev, Dos Hermanas 2003. 10.Qc2 0-0 11.e4 11.Rd1 may invite interesting complications after: 11...h6 12.Bh4 Qc8 13.e4 c5 14.d5

233

14...exd5! 15.e5! Re8! 16.exf6 Bxf6† 17.Be2 Bxh4 18.Nxh4 d4 The arising unbalanced position is not worse for Black, to say the least. 11...h6 12.Bh4 c5

13.dxc5 bxc5 14.Rd1 Qc7 15.Bg3 Qc8 16.Bb5 Nc6= Black’s control over the d4-square fully compensates for the weakened queenside structure. A2) 8.Qc2

234

This is the most ambitious choice; White is insisting on the early e2-e4 advance. 8...c5 9.Bf4 9.0-0-0?! proved to be too optimistic after: 9...a6 10.Bg5 (10.dxc5 Nbd7! 11.cxb6 Rc8 12.Be1 Qxb6ƒ) 10...cxd4 11.Nxd4 Nbd7 12.e4 Qc7³ Dubov – Leko, Moscow 2014. Harmless is: 9.dxc5 Bxc5 10.b4 (also after 10.Bf4 0-0 11.e3 Nbd7 12.Rd1 Qe7 13.Bb5 Rfd8 14.Bc7 Rdc8 15.Bg3 Nf8 Black had no problems in Moreno – Rivas Romero, corr. 2009) 10...Be7 11.e4 Nc6 12.Bf4 0-0 13.Rd1 Qc8

14.Bd3N (after 14.e5? Nxb4! 15.axb4 Ne4 White was crushed in Jobava – Kramnik, Dortmund 2006) 14...Nh5 15.Be3 Qb8 16.0-0 Nf4 With comfortable play for Black. 235

9...Nbd7 Another solid way of handling the position is: 9...cxd4!? 10.Nb5N This seems best. (The only game from this position saw 10.Rd1 a6 11.Nxd4 Nbd7 12.Nf3 Rc8 13.Qd2 b5³ and White had some problems in Dreev – Anand, Hyderabad 2002.) 10...Nd5 11.Qa4 Nc6 12.Nbxd4 Qd7 13.Bg3 a6=

10.Nb5!?N This is the most aggressive move, and invites interesting complications. 10.dxc5 has been played but it supports Black’s development. 10...Bxc5 11.e4 Rc8 12.Rd1 a6 13.Qd3 b5 14.Be2 This occurred in Stepanov – Farias, corr. 2005, and now I suggest:

14...h6!N 15.0-0 0-0 In the arising position with a symmetrical pawn structure, Black benefits from 236

having the more harmonious piece arrangement. 10...Nd5 11.Bg5 After 11.Nd6† Bxd6 12.Bxd6 cxd4 13.Rd1 Rc8 14.Qa4 Rc6 15.Bg3 0-0 Black’s active piece play is at least as important as White’s bishop pair. 11...Qb8 12.e4 N5f6

13.Qc1! Preparing Bf4 is White’s best and perhaps only good idea, but Black has enough resources to deal with it. 13...Nxe4 14.Bf4 Bd6 15.Nxd6† Nxd6 The pin looks awkward but White is unable to inflict any harm with it. 16.Bb5 Qc7 17.0-0 0-0 18.dxc5 bxc5 19.Rd1 Bd5= White can regain the pawn, but in no way is he fighting for an advantage. B) 7.e3

237

This is a popular move and has been tested extensively at the highest level. 7...Be7 This is the most popular of several playable moves. We will consider the rare B1) 8.Qc2, followed by the much more popular B2) 8.Bb5†. 8.Nxd5 Qxd5 9.Bd3 c5 transposes to variation B222 on page 125. 8.Bd3 is slightly illogical as, compared with variation B22 on page 124, White has neglected to provoke the ...c6 move. Nevertheless, the simplest solution for our purposes is 8...c5, transposing to our main line with one less move played. 8.Bd2 is a pretty harmless version of variation A, as White has given up on the idea of playing e2-e4 in one move. Play may continue 8...0-0 9.Bd3 c5 10.dxc5 (also after 10.Qc2 Nf6 11.dxc5 Bxc5 12.0-0 Nbd7 13.Rfd1 Qe7 Black was fine in A. Petrosian – Shneider, Telavi 1982) and now Black just needs one accurate move:

238

10...Nf6! (10...Bxc5 is less precise, and 11.Nxd5N 11...Qxd5 12.Qc2 Qh5 13.Be4 offers White a small edge.) 11.Qe2 Bxc5 12.e4 Ng4 13.0-0 Nc6= Black had no problems in Schussler – Moisieev, Naleczow 1979. B1) 8.Qc2

This seems like the most interesting way for White to deviate from the main line. 8...Nxc3 9.bxc3 Nd7 10.Bd3 c5 11.0-0 The present position has been reached via a few different move orders, including one game with Dreev on White’s side, so we should take it seriously. The situation resembles variation B1 of the next chapter 239

on page 145, and the two lines can easily converge. Here White will lose a tempo with e3-e4, but his queen will be able to move directly from c2 to e2, whereas in the next chapter it has to take a slower route. 11.Bxh7? This weak move was seen in Kuligowski – Josteinsson, Reykjavik 1982. Black should have responded with:

11...f5!N 12.Bg6† Kf8 Sometimes White can manage with such a misplaced bishop, but here his other pieces are not well placed, so Black will obtain both a material and a positional advantage. 13.Qa2 The desperate 13.h4 Bxh4–+ does not help, since Black is also threatening 14...Bxf2†. 13...Bd5 14.Qe2 cxd4 15.cxd4 Or 15.exd4 Qc7 16.Bd2 Rh6! 17.Bxh6 Qxc3† is another nice winning line.

240

15...Kg8! The last preparatory move, vacating f8 for the knight. 15...Rh6? would be a mistake due to 16.e4. 16.h4 Bxf3 17.gxf3 After 17.Qxf3 Nf8 18.Bxf5 exf5 19.Qxf5 Rxh4 White has only two pawns for the piece, and Black should be able to convert his material advantage into a full point. 17...Nf8 18.Rg1 Bxh4 19.f4

19...Rh6 20.Bh5 Bxf2†! 21.Kxf2 Qh4† 22.Kf1 Qxh5–+ White avoided losing a piece but Black was able to pick up an extra pawn, and he also benefits from having the superior minor piece.

241

11...Qc7 11...Rc8 led to an eventual success for Black in Dreev – Anand, Madras (5) 1991, but Anand went for ...Qc7 a few moves later anyway. I prefer to make the queen move now, having a particular follow-up in mind. 12.Qe2 0-0 13.e4 This position was reached in Dokhoian – Magerramov, Klaipeda 1988, and several other games. There are a few decent options but I find the following idea especially appealing:

13...cxd4!?N 13...Rfd8 14.Bb2 Rac8 is also perfectly playable and would lead to a long positional struggle. 242

14.cxd4 Interestingly, the database contains some games where this position arose with White to move, but only because Black played imprecisely in the opening. I found an interesting way to exploit Black’s development advantage. 14...e5! Exploiting the fact that White’s bishop is still on c1 instead of b2. 15.d5 The most ambitious reply. 15.Bb2 is more modest, and after 15...exd4 16.Nxd4 the activity of White’s minor pieces should be neutralized by means of:

16...Bd6! 17.h3 (17.Nf5?! Bxh2† 18.Kh1 Be5³ does not offer White much for a pawn; 17.g3 Nc5 18.Nb5 Qe7 19.Nxd6 Qxd6=) 17...Be5 18.Nb5 (18.Nf5 Nc5=) 18...Qd8 19.Rad1 Bxb2 20.Qxb2 Qe7 21.Rfe1 Nc5= 15...Nc5 16.Rd1 I also considered: 16.Bb2 Bd6 17.Rac1 (The knight transfer to the queenside will not bother Black: 17.Nd2 Rad8 18.Nc4? Ba6! and White is in trouble.) 17...Qe7„ The d5-pawn is blocked and it is not easy to find a constructive plan for White. Moreover, White’s a-pawn may prove weak in the long run. Black also does well after: 16.a4 Nxd3 17.Qxd3

243

17...f5! 18.Rd1 fxe4 19.Qxe4 Rad8 20.Bb2 Bf6 21.Bxe5 Bxe5 22.Nxe5 Qc5= 16...f5!

17.Bc4 In the event of 17.exf5 Nxd3 18.Qxd3 Rad8 Black will regain the pawn with an excellent position due to the bishop pair. 17...Kh8 18.exf5 Rxf5 19.Be3 Raf8„ Black has good prospects along the f-file.

244

B2) 8.Bb5†

This is the main line by far. White is provoking the ...c6 advance, so that Black will temporarily lose control over the long diagonal. 8...c6 It is worth analysing B21) 9.Nxd5 before tackling the main line of B22) 9.Bd3. 9.Ba4?! looks awkward. 9...Nxc3 10.bxc3 was seen in Novak – Buse, corr. 2009, when Black missed a nice idea:

10...Ba6!N 11.Bc2 0-0 12.e4 c5 White has nothing better than exchanging bishops, and after 13.Bd3 Bxd3 14.Qxd3 Nc6 15.0-0 Rc8 Black has excellent prospects.

245

9.Bc4?! also makes little sense, as the bishop makes an easy target for Black’s pieces. 9...0-0 10.0-0 c5 11.Nxd5 exd5 12.Ba2?! (12.Be2N is preferable, but even then after 12...c4! 13.b3 b5 14.a4 a6 Black’s play seems easier) In Korotkjevich – Kabatianski, Germany 2006, Black missed a chance to seize the advantage:

12...c4!N 13.Bd2 Nd7 14.Bb1 b5³ Black is making great progress on the queenside, whereas White’s setup is rather passive. B21) 9.Nxd5 Qxd5 9...exd5 is another possible way of handling the position. After 10.Bd3 0-0 11.0-0 Nd7 12.b4 the position is hardly dangerous for Black, but the arising pawn structure with a passive bishop on b7 will not be to everyone’s taste.

246

10.Bd3 0-0 11.Qc2 Qh5! This more accurate than 11...h6, as played in Miles – Portisch, Plovdiv 1983. 12.Ne5 This move is intended to show that Black’s queen is awkwardly located. However, moving the same piece twice while already behind in development is a clear concession as well. 12.0-0N is safer but 12...c5 13.Be4 Bxe4 14.Qxe4 Nd7= reaches a harmless position for Black. The extravagant 12.g4!?N 12...Qh3 13.Ng1 might backfire after13...Qxg4! (if a draw is an acceptable result, it is also possible to play 13...Qh4, when White has nothing better than repeating with 14.Nf3) 14.Bxh7† Kh8 15.Be4 Nd7 The arising position favours Black as his pieces are better developed. 12.Rg1!? is another move with an awkward feel to it. 12...Ba6 13.g4 Bxd3 14.Qxd3 Qd5 15.e4 Qd6 16.Be3 This position has been reached in two correspondence games. In my opinion, it makes sense to open up the position in order to highlight the drawbacks of White’s set-up, so I recommend:

247

16...c5!N 17.0-0-0?! (17.Rd1 is the lesser evil, but after 17...Rd8 18.Qc2 Qc6 White’s king is stuck in the centre and he must take care not to become worse.) The text move appears most consistent but after 17...Rc8 18.d5 e5!ƒ Black’s queenside attack practically plays itself, while his own king is absolutely safe on the kingside.

12...c5 13.Be4 The seemingly aggressive 13.h4 is no more than a drawing attempt. Black can keep the game going with: 13...f5!? (13...Bxg2 is the drawing line: 14.Rh2 Bb7 15.Be2 Qf5 16.Bd3 Qh5 17.Be2=)

248

14.Qb3 (14.Bc4?! runs into 14...Kh8! 15.Bxe6 Bf6ƒ when White’s king is under serious pressure) 14...Bd5 15.Bc4 Rd8 16.Bxd5 Rxd5 17.Nd3 Qf7 18.Nf4 Rd7 19.Qxe6 cxd4 With approximate equality. 13...Bd5

14.0-0 14.Bxd5 exd5 15.0-0 gives Black a choice between 15...Rd8, which transposes to the main line below, and the equally good 15...Rc8!?. 14...Rd8 Covering the important d5-spot and getting ready to develop the knight. 249

15.Bxd5 exd5

16.Bd2N 16.h3?! is too passive, and 16...Bd6 17.Nf3 Nd7 18.Bd2 Nf6 19.Rac1 c4 gave Black the upper hand due to his excellent prospects on the queenside in Hlavacek – Dob, corr. 2010. 16...Bd6 17.Nf3 Nc6 18.dxc5 bxc5 The activity of Black’s pieces fully compensates for any potential vulnerability of the hanging pawns. B22) 9.Bd3

250

White’s ambitions are mostly connected with seizing the centre and developing a kingside attack. Therefore the bishop is perfectly located on the b1-h7 diagonal, and the text move has been overwhelmingly the most popular choice. 9...c5! I was surprised to discover that this natural move is only Black’s fourth-most-popular choice. There are three noteworthy replies: B221) 10.0-0, B222) 10.Nxd5 and B223) 10.e4. 10.Bb5† proves ineffective after 10...Bc6, for instance: 11.Qa4 Qd7 12.Nxd5 exd5 13.Bxc6 Qxc6 14.Qxc6† Nxc6

15.dxc5 bxc5 16.Bd2 Kd7³ Black had the better piece play in Swayams – Vaibhav, Paris 2017. B221) 10.0-0

251

Surprisingly, after this natural move White already has to be careful to maintain the balance. 10...cxd4 11.exd4 Nd7 Black has successfully completed his development and the isolated d-pawn could prove a significant weakness. 12.Qe2 Another attempt to develop a kingside attack is 12.Re1 0-0 13.Bc2, intending to put the queen on d3.

13...Nxc3! A typical approach. The placement of the bishop on c2 makes it harder to protect the pawn on c3, so this is a perfect moment to change the pawn structure. 14.bxc3 Qc7 15.Qd3 g6 16.Bh6 Rfc8³ Black had won the opening battle in Lyell – Juhasz, Zalakaros 2010. 252

12...0-0 13.Ne4 Piece exchanges are usually unfavourable for White, as they reduce his attacking potential. Indeed, after 13.Nxd5 Bxd5 14.Bf4 Nf6 15.Ne5 Bb7³ White was doomed to passive defence in Gligoric – Milovanovic, Pozarevac 1995. 13...Qc7! I like this natural approach. Black connects his rooks and intends to eliminate one of the enemy bishops with ...Nf4. 14.Neg5 h6

253

15.Nxe6?! This piece sacrifice doesn’t work, but White’s position was already worse. 15...fxe6 16.Qxe6† Rf7 17.Bg6 Raf8µ Despite the approximate material balance, Black was much better due to his domination over the light squares in Galyas – Wang Shuai, Budapest 2001. B222) 10.Nxd5 Qxd5 As usual, we avoid recapturing with a pawn on d5 where it would block the bishop. 11.e4

11...Qd6! 11...Qd7 has occurred in a few games but the text move seems more precise. It is useful to prevent Bf4, and perhaps the knight may go to d7 instead of c6 in some lines. 12.dxc5 Hitting the queen offers White nothing, either before or after castling: 12.e5 Qd7 (12...Qc6!? 13.Qe2 0-0 14.Bg5 Qc7 looks equally good) 13.dxc5 bxc5 14.0-0 0-0 15.Qc2 h6 16.Rd1 Qc7 17.Be3 Nd7 Black has excellent play due to the weak pawn on e5. Similarly, after 12.0-0 0-0 13.e5 (13.dxc5 bxc5 transposes to the note to White’s 13th move in the main line below) 13...Qd7 (13...Qc6!? is possible here too) Black’s central pressure and powerful light-squared bishop start to tell, for instance: 254

14.Qe2 Nc6 15.dxc5 Rfd8 16.Rd1 Ba6! Exploiting White’s lack of development and inviting interesting complications. 17.Bxh7† Kf8! 18.Rxd7 Bxe2 19.Rxd8† Rxd8 20.Bc2 Bxc5 White has to struggle for equality. 12...bxc5 13.Bb5† I also considered: 13.0-0N 13...0-0 14.Qe2 (14.Be3 Rd8 15.Qb3 Qb6=) 14...Rd8 15.Rd1

15...Nc6! (15...Ba6 can be met by 16.Bg5! when White avoids losing material and obtains equal chances: 16...Bxd3 17.Bxe7 Qxe7 18.Rxd3 Rxd3 19.Qxd3 Nc6 20.Rc1 Rd8 21.Qe3 Nd4=) 16.Be3 e5 17.Rac1 Nd4 Black has excellent play.

255

13...Nc6 14.Qxd6N It looks logical for White to swap queens in the hope of exerting pressure against the c-pawn in the endgame. 14.Be3 Qc7 15.Qa4 0-0 16.0-0 Rfd8 17.Rfd1 Ne5= gave Black no problems in Koelewijn – Gierden, corr. 2014. 14...Bxd6 15.0-0 15.Be3 a6 16.Ba4 0-0 17.Rc1 Nd4 18.Nxd4 cxd4 19.Bxd4 Bxe4= is comfortable for Black.

15...0-0! 15...Ke7 is slightly less precise in view of 16.Be3 Nd4 17.Nxd4 cxd4 18.Bxd4 Bxe4 19.f3 Bd5 20.Ba6 with a tiny advantage for White. 16.Rd1 Rfd8 17.Be3 a6 18.Ba4 e5!= Black’s control over the d4-spot fully compensates for his structural drawbacks. B223) 10.e4

256

I consider this White’s most ambitious and challenging move. 10...Nxc3 11.bxc3 Nc6 12.0-0 This position has been reached via some different move orders. A complex battle lies ahead, with White generally looking towards the kingside and Black aiming for counterplay along the c-file and on the queenside in general. 12...cxd4 The immediate 12...0-0 allows 13.d5 Na5 14.c4, which I would prefer to avoid, even though 14...b5! 15.dxe6 bxc4 16.exf7† Rxf7 17.Bc2 Qxd1 18.Rxd1 Bf6 19.Rb1 Re8 reached an acceptable position for Black in Danielian – Parligras, Chalkis 2010. 13.cxd4 0-0 14.Be3 This is the best way of protecting the d-pawn. Still, there is a drawback behind it: now the c3-spot is weak, so Black’s rook will have nice prospects along the c-file. 14.d5? would be premature, as 14...exd5 15.exd5 Na5µ simply wins a pawn. The main alternative is: 14.Bb2?! This has been tried by some strong players but it is a mistake, as will soon become clear. 14...Bf6! The pressure on the d4-pawn is so strong that White can hardly avoid the positionally undesirable e4-e5 advance. If the bishop was on e3 this would not be so bad, as the bishop could still play a useful role on the kingside, but on b2 it will be poorly placed.

257

15.e5 15.Qa4 Rc8 16.Rad1 is strongly met by 16...Qe8!, and after 17.Ba6? Nxd4 18.Qxe8 Nxf3† 19.gxf3 Rcxe8 20.Bxb7 Bxb2µ White faced a depressing ending a pawn down in L. Szabo – Polugaevsky, Budapest 1967. I also examined 15.Bc2 Rc8 16.Qd3 g6 17.Rfd1 Na5³ when White was doomed to a passive defence in Bergmann – Wambach, Germany 1993. 15...Be7 16.Qe2 Qd5! 16...Na5 17.Nd2 is okay for White, so Black brings his queen to an active position first, taking advantage of a tactical nuance.

17.Rad1 17.Be4? is tactically refuted by 17...Nxd4!. True, White can avoid going a pawn down with 258

18.Bxh7† Kxh7 19.Bxd4, but after 19...Rac8µ Black was dominating in Ryskin – Cherepkov, Minsk 1982. 17...Na5 18.Rfe1 Rac8³ Black had an excellent game in Krensing – Krivoborodov, Pardubice 2008. 14...Rc8 15.Qe2 Na5 This move opens lines for the b7-bishop and c8-rook, while preparing a future occupation of the c4square.

In this complicated position, White must make an important decision regarding the placement of his rooks. We will analyse B2231) 16.Rfd1 and B2232) 16.Rfe1 in detail. 16.a4 is rather slow, and 16...Nc6! highlights the weakening of the b4-square. Following 17.Ba6 Qd7 Black was at least equal in Gheorghiu – Karpov, Malta (ol) 1980. 16.Rad1!? is a risky but playable pawn sacrifice, which was tried in Grooten – Vanheste, Eindhoven 1982. I think Black should have accepted the offer with:

259

16...Bxa3!N Black is in good theoretical shape after this move, although he should not underestimate his opponent’s attacking chances. We will analyse a) 17.d5 and b) 17.Ng5!?. a) 17.d5 This move appears tempting but it doesn’t quite work, although Black needs to be precise. 17...exd5 18.exd5 18.e5 is a thematic attempt to develop the initiative, but after 18...Be7 19.Nd4 Nc6 White does not have enough compensation for two pawns. 18...g6! This cold-blooded move puts a cold shower on White’s attacking plans.

19.Qa2 260

19.d6 Bxd6 20.Bh6 Bxf3 21.Qxf3 Qh4µ does not work for White. 19...Bb4! Black must be ready to cover the long diagonal. The timid 19...Bd6? runs into 20.Bh6 Re8 21.Bb5! Re4 22.Qb2 with a powerful initiative. 20.Bh6 Bxd5 21.Qa4 Bxf3 22.gxf3 Qh4 23.Bxf8 Kxf8³ With two pawns for the exchange plus the better pawn structure, Black is clearly on top. b) 17.Ng5!? This crude attacking move may be White’s only way to maintain the balance. 17...Be7! This is the safe approach. 17...h6 is playable but 18.Nh3 gives White definite attacking chances. 18.Nxh7! 18.Qg4 Rc3! leaves White without a convincing follow-up, for instance: 19.d5 (19.e5? Rxd3 20.Rxd3 Ba6–+) 19...e5 20.Qh5 Bxg5 21.Bxg5 f6 22.Bc1 Nc4³ 18...Kxh7 19.Qh5† Kg8 20.e5 f5 21.exf6 Rxf6

White has to search for a draw with: 22.Qh7† Kf7 23.Qh5†= Black can either go back with his king immediately, or play for a win with 23...g6!? 24.Qh7† Ke8 25.d5! Bxd5 26.Bg5! Qd6!. In the latter case, the machine gives 0.00 but any result would still be possible in a practical game. B2231) 16.Rfd1

261

With this move, White supports a future d4-d5 push but abandons all hope of achieving the ideal combination of rooks on d1 and e1. 16...Rc3! This strong move should offer Black excellent counterplay. 17.a4 White has also tried: 17.d5 exd5 18.exd5 18.e5? proved to be a mistake after 18...Nb3 19.Ra2 Nc5µ in Barbeau – Sambuev, Montreal 2016. In Kalantarian – Anastasian, Yerevan 1994, Black did not find the best way to neutralize White’s central play.

262

18...Bf6!N 19.Rab1 19.d6?! runs into 19...Bxf3 20.gxf3 Rc6³ when the d6-pawn will fall. 19...Bxd5 20.Be4 Bxe4 21.Rxd8 Rxd8„ Black has full material compensation for the queen, his pieces are active and he has no serious weaknesses, so his chances are certainly not worse. The text move is White’s best try, but it offers Black an important tempo for consolidating.

17...Qa8! 18.Rac1 I checked a couple of other options:

263

18.Nd2?! is too passive, and 18...Rd8 19.Rac1 Rxc1 20.Rxc1 Nc6³ left Black with an excellent position in Khotenashvili – Tsatsalashvili, Tbilisi 2010. 18.d5 This has been the most popular choice but Black is well placed to deal with it. 18...exd5 19.exd5 19.e5?! Nb3! 20.Rab1 Nc5 21.Bc2 Ne6³ leaves White without enough play for the pawn. 19...Nb3

20.Rab1 Bxd5 21.Nd4 Nc5 22.Bxh7† Kxh7 23.Qh5† Kg8 24.Nf5 We have been following a correspondence game in which the tactics resulted in a quick draw. A possible way to keep the game going is:

264

24...Rxe3!?N The game concluded peacefully after: 24...Bf6 25.Rxd5 g6 26.Nh6† Kg7 27.Nf5† ½–½ Halldorsson – Timson, corr. 2005. 25.fxe3 Bf6 26.Rxd5 g6 27.Nh6† Kg7 28.Nf5† gxf5 29.Qxf5 Qc8„ Black is not worse and the game goes on. Let’s return to the main line, where we have been following Fineboim – Schindler, corr. 2009. It seems to me that Black’s most convincing route to a safe position is:

18...Rxd3!N 19.Rxd3 19.Qxd3 Bxe4 20.Qe2 Bd6 offers Black superb compensation for the exchange. 19...Ba6 20.d5 20.Qc2 Bxd3 21.Qxd3 Rc8= gives Black easy play. 20...Bxd3 21.Qxd3 Rd8 22.Rd1 22.Rc7 is met by 22...Bd6 when the rook has to retreat.

265

22...Nc6! The knight joins in the fight against the passer. 23.Qc4 exd5 24.Rxd5 24.exd5? Nb4µ picks up the pawn. 24...h6„ Black is not worse at all. B2232) 16.Rfe1

266

This is the most flexible way for White to handle the position. The rook on e1 helps to make the d4-d5 break more effective, while the d1-square is still available for the other rook. 16...h6! This prophylactic move cuts out a number of White’s tactical resources by depriving his pieces of the important g5-square. The most common continuation is 16...Rc3, when White usually moves his a-pawn out of danger. However, I found a new and much more dangerous idea: 17.Rad1!N 17...Rxa3 18.d5ƒ 16...Kh8 was once used to draw against the young Garry Kasparov, but the following example illustrates White’s attacking potential: 17.Rad1 Qc7 18.Ng5! Bxg5 19.Bxg5 White was clearly better in Lputian – Giorgadze, Pavlodar 1982. 17.a4 The somewhat mysterious 17.Reb1?! Qd6 18.Bd2 Nc6 19.Bc3 Rfd8 20.Rd1 Na5 21.Bb4 Qc7³ led White nowhere in Nogler – Buettner, email 2013. The more aggressive 17.Ne5N should be met by 17...Nc6!, and after 18.Nxc6 Rxc6 the exchange of knights makes Black’s position safer. For instance:

19.Qg4 (19.Red1 is also well met by 19...Bg5!, and if 20.f4 Be7 Black is doing well) 19...Bg5 20.Rad1 Bxe3 21.Rxe3 Qg5 White’s attack comes to an end, so Black is fine. Finally, I considered an interesting new idea involving a pawn sacrifice by White: 17.Rad1!?N 17...Bxa3 18.Ne5 Black is well prepared for the standard break in the centre: 18.d5?! exd5 19.exd5 Bb4 20.Rf1 Re8 267

21.Bb5 Re7³ White does not have enough for the missing pawn.

18...Bd6 19.f4 The natural 19.Ng4 is well met by 19...f5! when the obvious sacrifice doesn’t work: 20.Nxh6† gxh6 21.Qh5 Qe8! 22.Qxh6 Rf7 White has nothing better than liquidating into a bad endgame. 23.exf5 Rh7 24.Qxe6† Qxe6 25.fxe6 Rg7 26.g3 Bd5µ White’s passers are going nowhere. 19...Nc6 The knight joins the defence with good effect. Already a pawn down, White has to go ‘all in’. 20.Ng4 In the event of 20.Bb1 Nb4 21.d5?! exd5 22.exd5 Rc7 23.Bf2 Qc8 24.Qe4 f5µ White’s attack comes to an end. 20...Nb4 The best defensive approach is to counterattack! 21.Bb1 f5! As usual, this pawn advance limits White’s attacking abilities, even though it apparently makes the king more exposed. 22.d5! White has to play dynamically, so we have ‘fire on the board’! 22...fxg4 23.dxe6 Qc7 24.e5 Bc5 25.Rd7 Bxe3† 26.Qxe3 Qc6 27.Qd2

268

What a spectacular position! At the moment Black is a full piece up, but White’s rook on the 7th rank enables him to achieve a draw after: 27...Qc5† 28.Kh1 Bd5 29.f5 Qc3 30.Qxc3 Rxc3 31.f6 Bxe6 32.Rxg7† Kh8 33.Rh7†= With perpetual check. 17...Nc6 We have already seen some examples of this theme: once a3-a4 is played, the b4-square becomes an inviting outpost for Black’s minor pieces, especially the knight.

18.Rad1N I find this the most logical continuation. Now White’s pieces are located perfectly to support the d4-d5 push, so Black should be careful. 269

18.Ba6? Bxa6 19.Qxa6 Nb4 20.Qxa7 Nc2 was losing for White in Pearce – Silva Filho, corr. 2011. 18.Rac1 This move is not the most challenging, as a rook exchange along the c-file will remove a potential attacker and bring Black a step closer to a promising endgame. 18...Nb4 19.Bb1 Ba6 20.Qd1 20.Qd2N is hardly an improvement: 20...Rxc1 21.Rxc1 Qd7 22.Ne5 Qxa4 23.h4 (23.Bxh6? doesn’t work: 23...gxh6 24.Qxh6 Rc8 25.Rxc8† Bxc8 26.Qh5 Qe8µ) 23...Rc8 24.Rxc8† Bxc8 25.Qc1 Qe8 26.Qc7 White’s active pieces enable him to regain the pawn, but Black has nothing to worry about. 26...Bxh4 27.Qxa7 Bf6=

20...Rxc1N This is my natural new idea. 20...Bb7 21.Qb3 Rxc1 22.Rxc1 Qb8 was fine for Black in Rodriguez Rey – Martins, corr. 2015, but I don’t see any need for the ‘silent draw offer’ at move 20. 21.Qxc1 Nd3 22.Bxd3 Bxd3 Black’s bishops are strong, so White has to force matters to avoid being worse.

270

23.Bxh6 gxh6 24.Qxh6 Bf6 25.Re3 Bb1 26.Ng5 Bxg5 27.Rg3 Bxe4 The bishop returns to the defence just in time. 28.Rxg5† Bg6 29.Rxg6† fxg6 30.Qxg6†= 18...Nb4 19.Bb1 Qc7 Now White is at a crossroads.

20.d5!? This is the sharpest and most critical continuation. I checked two other logical ideas: 20.Ne5 Nc2! 21.Rf1 21.Rc1? blunders a pawn: 21...Nxd4!µ 271

21...Bg5! The engine likes 21...Nxe3 22.fxe3, but opening up the f-file looks scary to my human eye. Instead I wish to provoke the f2-f4 advance in order to exchange White’s dark-squared bishop without giving him an open f-file and strengthening his centre in the process.

22.f4 22.Bxg5 hxg5 23.Nf3 Qc4! allows Black to liquidate into a better endgame. 22...Nxe3 23.Qxe3 Be7 24.f5 Qc3! Active counterplay is the best defensive approach.

25.Qf2 25.Qe2 Bf6 26.fxe6 fxe6 27.Ng6 Bxd4† 28.Kh1 Rxf1† 29.Rxf1 Re8 30.Ba2 Qc6„ 25...Ba6 26.Rfe1 Rcd8 27.Ba2 Kh7„ 272

Black has a strong pair of bishops and pressure against the centre, giving him enough play to offset White’s attacking ideas. The other natural continuation is: 20.Rc1 Preventing Black’s knight and queen from infiltrating along the c-file. The drawback is that White has to put his attacking plans on hold.

20...Qd6 20...Qd7?! is riskier in view of 21.Ne5! Qxa4 22.Qg4ƒ with good attacking chances. 21.Ne5 Nc6 22.Nc4 22.Nxc6 Bxc6 23.Qd3 e5„ seems okay for Black. 22.Ng4!? h5 23.Nh6† is an enterprising idea, but my analysis indicates that White gets no more than a draw: 23...gxh6 24.Bxh6 Nxd4 25.Qxh5 e5 26.Ba2 Ne2† 27.Kh1 Nf4 28.Bxf4 exf4 29.Rxc8 Bxc8 30.e5 Qd2 31.Qg6† Kh8 32.Qh6† Kg8= 22...Qd7 23.Red1

273

23...Na5! It is important to unblock the c-file at once. 24.Nxa5 bxa5 With an outpost on b4 and pressure against the a4-pawn, Black has enough activity to offset the damaged pawn structure. For example: 25.d5 Rxc1 26.Rxc1 exd5 27.Qd3 Rd8 28.e5 g6 29.Bxh6 Rc8„ 20...exd5 21.Nd4! The knight is heading for f5, so White’s attacking potential should not be underestimated. 21.exd5? is too simplistic, and after 21...Rfe8 22.Rc1 Bc5 Black wins a pawn.

274

21...dxe4 22.Nf5 22.Qg4 is well met by 22...h5!. Black returns one of the extra pawns in order to win a vital tempo for consolidating. 23.Qxh5 Rfd8

24.Qg4 (24.Nf5 Bf6 25.Bg5 Qe5³) 24...Bf6 White’s attack has come to an end, and it’s time for him to look for a way to equalize by winning back the sacrificed pawn. 22...Rfd8 Black prepares to exchange a pair of rooks, while also vacating the f8-square.

23.Qg4 White’s other option is: 23.Bxh6 Bf8 24.Qg4 Rxd1 25.Rxd1 Qe5 26.Bf4 Qf6

275

27.Nh6† Kh7 28.Bxe4† Bxe4 29.Qxc8 Qxf4 30.Qxf8 Nd3 31.Qxf7 Qxh6 In the resultant endgame, Black has the better practical chances as he can use his active pieces to create threats against White’s king. 23...g6 24.Nxh6† 24.Nxe7† Qxe7 25.Bxh6 Rxd1 26.Rxd1 Rd8 looks fine for Black. 24...Kf8

25.Nf5 Rxd1 26.Rxd1 Rd8 Black may appear slightly weak on the dark squares, but White does not have any serious threats. 276

Exchanging rooks helps the defence, especially since White is slightly weak on the back rank. 27.Bh6† Kg8 28.Nxe7† Qxe7

29.Rxd8† Qxd8 30.h3 Nd3 The powerful bishops provide White with just enough compensation for a pawn, but he is not realistically fighting for an advantage.

Conclusion This chapter has dealt with two significant lines after 6.cxd5 Nxd5. The first of them, 7.Bd2, has the idea to recapture with the bishop in the event of an exchange on c3, so 7...Nf6 feels like a logical reply. White’s two main options of 8.Bg5 and 8.Qc2 may both lead to lively play, but Black generally gets a full share of the chances with a timely ...c5 combined with active development. The other big topic was 7.e3. This move may appear timid but White generally intends to follow up with a later e3-e4 after developing his bishop to d3. Our main line continues 7...Be7 8.Bb5† c6 9.Bd3 c5! 10.e4 Nxc3 11.bxc3 when the pawn structure resembles a Grünfeld but with our bishop on e7 instead of g7. Black’s task will be to create enough counterplay on the queenside to prevent White from building an effective attack on the other wing. The c-file is an obvious source of activity, while another useful pointer to remember is to meet a3-a4 (which is often played to avoid leaving the pawn hanging on a3) by retreating the knight from a5 to c6, intending to occupy the newly weakened b4-square.

277

A) 8.Qxc3 h6! 139 A1) 9.b4 139 A2) 9.Bf4 140 A3) 9.e3 142 A4) 9.g3 143 B) 8.bxc3 c5 9.e4 Nd7 145 B1) 10.Bd3 145 B2) 10.Bf4 Be7 149 B21) 11.dxc5?! 150 B22) 11.d5 151 B23) 11.Bd3 Qc8! 153 B231) 12.Qa4 154 B232) 12.Qb1 155

278

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.a3 Bb7 5.Nc3 d5 6.cxd5 Nxd5 7.Qc2 This is White’s most popular choice. By comparison with 7.e3, White avoids blocking the dark-squared bishop and gives himself the option of seizing the centre with e2-e4. On the other hand, White may have to spend time moving his queen again, as the c-file is likely to be opened at some point. 7...Nxc3 I recommend this, the most popular of several playable moves. Black removes the central tension and will choose a plan of development according to how White recaptures. We will analyse A) 8.Qxc3 followed by B) 8.bxc3, the latter move being the more popular by a ratio of more than three to one. A) 8.Qxc3

This continuation was first seen in 1982, but it gained popularity from the late 80s and through the 90s, helped by the efforts of grandmasters like Bareev and especially Dreev. The somewhat unusual location of White’s queen has its merits: the pressure along the c-file might cause problems, and if Black plays 279

...c5 then the queen may be useful on the long diagonal. However, as tournament practice has proven, losing control over the e4-square is a concession. 8...h6! This prophylactic move enables Black to comfortably complete his development. After both 8...Be7 9.Bf4 and 8...Bd6 9.Bg5 Qc8 10.e3 0-0 11.Rc1 Black is liable to experience problems with the vulnerable c7-pawn. White has four main ideas: A1) 9.b4, A2) 9.Bf4, A3) 9.e3 and A4) 9.g3. A1) 9.b4

9...Nd7 9...Be7 is more common, and the only move mentioned by Beliavsky & Mikhalchishin, but text seems more active. 10.e3 Bd6 I find this to be the most flexible set-up. Black takes control over e5 and leaves the e7-spot vacant for the queen. 11.Bb2 0-0 12.Be2

280

12...a5!?N This is my new idea. The pressure on the b4-pawn is provoking White to release the tension. 12...Qf6!? is another attractive way of handling the position. A game continued: 13.0-0 Qg6 14.Rfd1?! Removing the rook from the kingside is a risky decision. (14.Rac1N 14...f5 15.Nh4 Qf7 16.Qc2= looks preferable.) 14...f5 15.Rac1 Nf6ƒ Black had great attacking prospects in Mandrikin – Predke, Taganrog 2013. 13.0-0 Qe7 14.b5 14.d5 Nf6 15.dxe6 fxe6! offers Black excellent play along the f-file.

281

14...Rac8 Overprotecting the c7-pawn and keeping the ...c6 break in mind. White’s queen is evidently not well placed on c3. 15.Rac1 15.d5?! e5 16.e4 f5³ makes it difficult to protect the d5-pawn. 15...f5! I like this positional concept: Black takes full control over the light squares, while the weakening of e5 is not so important. 16.Nd2 16.Ne5 Nxe5 17.dxe5 Bc5³ leaves White with no real compensation for his inferior structure.

16...Nf6 17.Nc4 f4 Black’s pieces are active and White’s king is in some danger. A2) 9.Bf4 Before playing e2-e3, White develops his bishop outside the pawn chain. 9...Bd6

282

10.Ne5 This is White’s most ambitious attempt, but he risks falling behind in development. 10.Bxd6 cxd6 gives Black easy equality; you don’t have to know any more than this. 10.Bg3 may be best, although it doesn’t offer White any advantage: 10...0-0 11.e3 Nd7 12.Bb5 Bxg3 13.hxg3 c5 14.Rd1 In Ftacnik – Farago, Altensteig 1987, Black could have neutralized the mild pressure along the d-file by means of:

14...Rc8N 15.dxc5 Rxc5 16.Qb4 Rd5= 10...Nd7 11.e3 283

11.f3 0-0 12.Rd1 Qe7 13.e4 Rfd8 14.Be2 Rac8 15.Nd3 Bxf4 16.Nxf4 c5 was comfortable for Black in Najer – Estremera Panos, Linares 2001. 11...0-0

12.Nc6!?N I was surprised to discover that this natural move has yet to be tested. After 12.Bb5 Nxe5 13.dxe5 Be7 14.Rd1 Qc8 15.Rd7 Bc5 Black had the better pawn structure in Alekseev – Dmitrieva, corr. 2014. 12...Qf6! 12...Bxc6 is best avoided in view of 13.Bxd6 cxd6 14.Qxc6 Qe7 15.Ba6 when White has a small but risk-free edge due to his control over the c-file. 13.Bg3 Rfe8! Black is preparing ...e5 in order to make use of his development advantage and unblock the c-pawn. A logical continuation would be:

284

14.Rc1 e5 15.Be2 exd4 16.Qxd4 Qxd4 17.Nxd4 Bxg3 18.hxg3 c5= The position is balanced, although plenty of play remains. A3) 9.e3

This has been White’s most popular choice but the prospects of the c1-bishop are not great now. Black has a few attractive options, my favourite being: 9...Bd6 10.Bb5† 10.b4N 10...Nd7 takes us back to variation A1.

285

10...c6

11.Ba4 White may also try: 11.Bd3 0-0 12.b3 I also checked 12.e4N 12...c5 13.dxc5 bxc5! 14.0-0 Nc6 15.Be3 e5 when I prefer Black, whose knight will reach the perfect d4-spot. 12...Nd7 13.Bb2 We have been following Daroczy – Pecotic, email 2005. I like the following natural way of handling the position:

13...Nf6N 14.0-0 14.e4 c5 15.e5 Nd5 16.Qd2 Be7³ is promising for Black. 286

14...c5 15.dxc5 Bxc5 16.Rfd1 Qe7 17.b4 Bd6= 11...0-0 12.0-0 This position was reached in Dreev – Anand, Madras (3) 1991, and a couple of subsequent games, all of which continued with 12...c5. Instead, I believe Black should have continued:

12...Nd7!N 13.e4 Presumably this is the move that Black was keen to prevent in the aforementioned games, but there is nothing to fear. 13...c5 14.e5 Bc7 15.Rd1 Qe7 16.b4 Bd5! White’s centre is under pressure. Play may continue:

287

17.bxc5 bxc5 18.Bxd7 Qxd7 19.dxc5 f5! Black has an excellent play for a pawn, due to his powerful bishops. A4) 9.g3

This hasn’t been the most popular move, but it is quite logical to fianchetto and aim for queenside pressure. 9...Bd6 10.Bg2 0-0 11.0-0 Nd7 12.b4 I considered a few other possibilities: 12.b3 is playable but rather timid compared to the main line. 12...Nf6 13.Bb2 Rc8 14.Rfd1 Be4 15.Nd2 288

Nd5 16.Qc4 Bxg2 17.Kxg2 This occurred in Borras Julian – Chomicki, corr. 2013, when Black should have played:

17...f5!N 18.e4 fxe4 19.Nxe4 Qe8 Black has a comfortable game due to his control over d5, and the light squares generally. 12.Rd1 Nf6 13.Ne1 White does not mind trading bishops, as he hopes to exploit the light-square holes on the queenside. 13...Bxg2 14.Kxg2 Qe7 15.Nd3 We have been following Marquez Abreu – Leemans, corr. 2014. My new idea is:

15...Rad8!?N It seems to me that Black should prepare to attack the enemy centre. 289

16.Qc6 16.b4? runs into 16...e5! and White is in trouble. 16.Bf4 Nd5 17.Bxd6 cxd6= is harmless for Black. 16...e5 17.dxe5 Bxe5 18.Bd2 Bd6 19.Re1 Qe4† 20.Qxe4 Nxe4= Black has nothing to worry about. The text move is the most ambitious try; White is seizing more space on the queenside and making it harder for Black to play ...c5. I suggest deviating from Akwei – Huzita, corr. 2014, with the following new idea:

12...Be4!?N 13.Rd1 Black’s strategy is well illustrated in the following line: 13.Bb2 Nf6 14.Rfd1 Nd5! 15.Qc4 f5!³ With full control over the light squares. 13...Nf6 14.Ne1 White wishes to carry out a thematic plan of exchanging bishops and building a strong centre with e2e4. However, it takes a long time. 14...Bxg2 15.Nxg2 Qd7 Connecting the rooks and taking control over the b5-square. Black has a harmonious set-up with good prospects for counterplay, as the following line illustrates.

290

16.Qf3 a5! Just in time, before e2-e4 comes. 17.bxa5 17.e4?! Qc6!³ leads nowhere for White. 17...Rxa5 18.Bxh6 Rf5 19.Bf4 g5

20.e4 White avoids losing a piece, but gets no advantage.

291

20...Nxe4 21.Qxe4 gxf4 22.gxf4 White is a pawn up, but almost all of his isolated pawns are weak. Black only needs to focus on securing his king. 22...Rd8 23.Rd3 White’s main plan is exchanging the powerful rook on f5, so that the knight will not be tied to the defence of the f4-pawn.

23...Qe7 The queen is heading for f6, where it will be closer to the king and put more pressure on White’s weaknesses. 24.Rad1 Qf6 25.Rf3 Kf8 26.Rdd3 Rh5 Black is at least equal; an important point is that liquidating into an endgame is likely to favour him. White’s extra pawn is not contributing much and it is hard to find constructive ideas for him. B) 8.bxc3

292

This recapture is “probably the only way to fight for an opening advantage” according to Beliavsky & Mikhalchishin. White intends e2-e4 next, with a strong centre and attacking chances on the kingside. 8...c5 8...Be7 9.e4 0-0 is somewhat more popular and also offers Black a solid position. However, I prefer to challenge the opponent’s pawn centre immediately. 9.e4 Nd7 White has two natural plans of development: B1) 10.Bd3 and B2) 10.Bf4. B1) 10.Bd3

293

10...Qc7! This move causes White some coordination problems. The point is that ...cxd4 is a positional threat, and White cannot put his queen on the ideal e2-square due to the potential check on c3. 11.Qb1 This is the most common and logical choice. The queen protects the rook in the corner and supports the light-squared bishop, offering hopes of a future attack along the b1-h7 diagonal. 11.0-0?! cannot be recommended for White: 11...cxd4 12.cxd4 Qxc2 13.Bxc2 Be7 14.Bb2 0-0 15.Rfd1 Rac8 16.Bd3 Rfd8³ Black had the easier play in Zoler – Gerzhoy, Ramat Aviv 2004. 11.Bb2 has been tested at GM level but is harmless: 11...cxd4 12.cxd4 Qxc2 13.Bxc2 Ba6! Out of a few decent options, this is my favourite. 14.Rc1 A draw was agreed here in Christiansen – Miles, Portoroz 1985. A logical continuation could be:

294

14...Be7 15.Kd2 0-0 16.Bc3 Bb7 17.Bb4 Bxb4† 18.axb4 a5= Black has nothing to worry about. 11.Qd2 is well met by 11...Nf6, provoking White to fix his central structure. 12.e5 (12.Bb5†N 12...Bc6 13.Qe2 Bxb5 14.Qxb5† Qd7 15.Qd3 Be7=) 12...Nd5 Black is fine, for instance:

13.0-0 (13.Bb5†?! only helped Black after 13...Bc6 14.Bxc6† Qxc6 15.c4 Ne7³ in Balta – R. Ward, corr. 2016) 13...cxd4 14.cxd4 Be7 15.Ng5 h6 16.Ne4 0-0 17.Qe2 f5 18.exf6 Nxf6 The respective isolated pawns are equally weak, so Black has no problems. 11.Qa2 This has been tested by some strong players and has achieved a better statistical score than the main line. White may or may not recentralize his queen once he has castled. 295

11...Be7 It would be a mistake to try 11...cxd4?! 12.cxd4 Qc3† 13.Ke2². White’s king is perfectly safe and the open c-file will cause Black problems. 12.0-0 0-0 13.a4 13.Qe2 has been the most popular choice but it transposes to variation B1 of the previous chapter – see page 120 for the continuation beginning with 13...cxd4N. 13.Bb2 allows Black to highlight the poor location of White’s queen by means of: 13...b5! (13...Rc8 was equal in Dreev – Greenfeld, Fügen 2006) 14.a4 (14.Bxb5? Bxe4 is poor for White) 14...a6 15.Qb1 Rfb8 Black had the more harmonious set-up and the better chances in Capuano – Moyseenko, corr. 2011. I recommend meeting the text move with a novelty:

296

13...Rfd8!N With the following idea in mind: 14.a5 bxa5! 15.Qxa5 15.Bd2 seems preferable although 15...a4 16.Qxa4 a5 offers Black fine counterplay thanks to the passer on the a-file. 15...Qxa5 16.Rxa5 cxd4 17.cxd4 Nc5! 18.dxc5 Rxd3³ White has a difficult endgame ahead.

11...Be7 Another common way of handling the position is 11...g6 in the spirit of the Grünfeld, but I would rather not invite 12.h4!?. Beliavsky & Mikhalchishin opine that the insertion of this move and 12...h6 favours Black, but I believe White has more attacking chances here than in our main line. 12.0-0 0-0 We have a typical scenario for the Petrosian System: White’s pawn centre looks nice but Black’s pieces are better coordinated, and might put strong pressure on it. 13.Be3 Many different moves have been tried here. I will present my analysis of three other options, which should provide you with enough ideas to meet any of White’s plans. 13.e5 has been tried by a few strong players but I see no reason for White to commit to a rigid pawn formation at this stage. A good example continued: 13...h6 14.Be3 Rfd8 15.Nd2 Rac8 16.Ne4 In Naumkin – Czebe, Balatonlelle 2002, Black missed a strong idea:

297

16...f5!N 17.exf6 Nxf6 18.Nxf6† Bxf6³ White’s attacking chances are gone, and his centre is under strong pressure. 13.Re1 Rfd8 14.a4 (14.Bb2 Rac8 15.Re3?! was the overambitious continuation of Dankert – Lau, Germany 1985, when Black could have seized the advantage by means of: 15...e5!N 16.d5 c4 17.Bc2 f5! 18.exf5 Bxd5³) 14...h6 15.Bd2 This was Jovanovic – Ermolaev, corr. 2009. Here I suggest:

15...Rac8!N Simply developing the last piece, and intending to meet 16.a5 with 16...bxa5! 17.Qb5 cxd4 18.cxd4 a4! to exploit the loose bishop on d3, as seen after: 19.Qxa4 Nc5 20.Qc2 Nxd3 21.Qxd3 Bc5³ 13.a4 Rfd8 14.h3 As usual, 14.a5?! bxa5! works well for Black, for instance: 15.Bd2 h6 16.Qa2 (16.Qb5 Qb6 298

17.Qxa5 Qxa5 18.Rxa5 cxd4 19.cxd4 Nc5!µ) 16...cxd4 17.cxd4 Bb4³ White cannot regain the pawn.

14...Nf8!? I like this thematic manoeuvre: the knight is heading for g6, where it protects the king and may jump actively to f4. 15.Be3 15.a5N is more reasonable here than on the previous move, but 15...bxa5 is still fine for Black. My analysis continues: 16.Bf4 Bd6 17.Bxd6 Rxd6 18.Qa2 Ng6 19.Rfe1 (19.Qxa5?! Qxa5 20.Rxa5 cxd4 21.cxd4 Nf4³) 19...Ne7 20.Qxa5 Qxa5 21.Rxa5 cxd4 22.cxd4 Nc6 23.Ra4 a5= The endgame holds no danger for Black. 15...Ng6

299

16.g3 After 16.Re1 Rac8 White runs out of constructive ideas, since 17.a5?! bxa5 18.Qa2 cxd4 19.cxd4 Bb4 is simply bad for him. 16...Rac8= Black had no worries in Parsons – Bellegotti, corr. 2015.

13...Rfd8 The rook is not just putting pressure on the d4-pawn, but also freeing the f8-square for the minor pieces and setting up a potential ...Nc5 trick, which we have already seen in a few of the notes above. 14.Nd2 This retreat is connected with some aggressive ideas: White may advance his f-pawn and/or play e4-e5 followed by Ne4. Another game continued: 14.a4 Rac8 15.Re1 h6 16.Ra2 (16.a5?! runs into 16...bxa5!³ as usual.) White is trying to activate his the rook along the second rank but it feels too artificial. Black has a harmonious setup and a few ways of developing counterplay, but I especially like the plan played in the game:

300

16...Nb8! The knight is heading for c6. 17.Rb2 cxd4 18.cxd4 Nc6³ White was suffering due to the strong pressure on his central pawns in Demetrio – Brodda, corr. 2007. 14...Rac8 15.e5 15.f4? is refuted by 15...cxd4 16.cxd4 Nf6 17.Kh1 Qc3 when White’s position is on the verge of collapse. We have been following the game Kononenko – Fedorchuk, Badalona 2005. Black should have played:

15...h6!N 16.Ne4 After 16.f4 f5! 17.Bc4 Qc6 18.Rf2 Nf8³ Black’s positional advantage is indisputable. 301

16...f5 17.exf6 Nxf6 18.Nxf6† Bxf6³ In purely structural terms, the pawns on d4 and e6 are equally weak, but Black’s pieces are much better located. B2) 10.Bf4

This is the most popular continuation, and the most accurate according to Beliavsky & Mikhalchishin. 10...Be7 The main idea behind White’s last move is illustrated in the following fashionable line: 10...cxd4 11.cxd4 Rc8 12.Qb3 The tactical justification is shown after 12...Bxe4 13.Ba6 Bd5 14.Qa4 Bc6 15.Qd1 Ra8?! (The lesser evil was 15...Be7 16.0-0 0-0 17.Bxc8 Qxc8 18.Ne5²) 16.Rc1 and Black was in trouble in Morozov – Cornell, corr. 2014. Here White faces an important choice. We will analyse the tricky but inaccurate B21) 11.dxc5?!, followed by B22) 11.d5 and finally the most popular B23) 11.Bd3. White has occasionally tried: 11.Bb5 a6 12.Bd3 This strikes me as a bit too sophisticated, as White is unlikely to benefit from the inclusion of ...a6. 12...cxd4 13.cxd4 Rc8

302

14.Qe2 14.Qb1 is well met by: 14...Rc3! 15.0-0 0-0 16.a4N (after 16.Rc1 Rxc1† 17.Bxc1 h6 18.Bf4 Nf6 19.a4 a5 White faced some problems in Martys – Do Amaral, email 2006) 16...Qa8 17.Re1 a5 18.Bd2 Rcc8= 14...b5 White’s only challenging idea is to strike at the queenside immediately with: 15.a4!?N 15.0-0 allows the consolidating 15...Nb6, and after 16.Rfc1 0-0 Black was doing well in Dean – Shpaer, email 1996. 15...Bb4† 16.Bd2 Bxd2† 17.Qxd2 0-0

18.axb5 axb5 19.0-0 Nf6 20.Rfe1 Ra8= 303

The pressure against e4 compensates for the weakness of Black’s passer. B21) 11.dxc5?!

This anti-positional move has been employed in three correspondence games, with White scoring two wins and a draw so far. White wants to exploit the fact that Black has not castled, but we have more than one good answer. 11...Bxc5 11...0-0!?N also seems to refute White’s approach. For instance, 12.cxb6 (or 12.Rd1 Qc8 13.cxb6 Nc5 14.Nd2 axb6µ) 12...Nc5 13.Nd2 Bg5! 14.Bxg5 Qxg5 with a powerful initiative. 12.Rd1 12.Qa4 was seen in Miralles – Caiafas, Lucerne 1985, when the natural 12...a6!N 13.Rd1 b5 14.Qc2 Qb6 would have secured a big positional advantage for Black. We have been following Riemer – Stolle, corr. 1983, and one later game. Black should have taken his queen off the d-file with:

304

12...Qc8!N 13.Bb5 Bc6 14.Qa4 This seems like the obvious try but it doesn’t work. 14.a4 is the lesser evil although 14...a6 15.Bxc6 Qxc6 16.0-0 0-0³ still leaves White suffering from a worse structure. 14...Bxb5 15.Qxb5 a6

16.Qd3 Bxa3 17.Nd2 After 17.0-0 Ra7 nothing can stop Black from castling.

305

17...Be7 18.Nc4 Qc6

19.Qxd7† 19.0-0?! Nf6 20.Nd6† Bxd6 21.Bxd6 Nxe4 22.Qf3 0-0-0 is winning for Black. 19...Qxd7 20.Rxd7 Kxd7 21.Nxb6† Kc6 22.Nxa8 Rxa8µ White was able to regain the pawn but the resulting endgame is unpleasant for him, since the passed apawn is a monster. B22) 11.d5

This advance must always be taken seriously, but Black has enough resources to deal with it here. 306

11...exd5 12.exd5 0-0 13.Rd1 13.c4 is harmless: 13...Bf6 14.Rd1 Re8† 15.Be2 Qe7

16.Be3N (16.Rd2? allowed the unpleasant 16...g5! 17.Bg3 g4µ in Miniboeck – Kiss, Budapest 1987) 16...Ne5 17.Nxe5 Bxe5 18.0-0 Bc8= The d5-pawn is safely blockaded, so Black has nothing to worry about. I also checked 13.Bd3N 13...g6 14.c4 which leads to similar play as the line above: 14...Bf6 15.Rd1 Re8† 16.Be2 (16.Be3 Ne5 17.Nxe5 Bxe5 18.0-0 Qd6 19.h3 Bc8=) 16...Qe7 17.Be3 Ne5 18.Nxe5 Bxe5 19.00 Bc8 20.Qd2 Bd7= Once again, Black has a firm blockade.

307

13...Re8 This is my recommended move, which leads to a safe, healthy position when followed up correctly. Black can take the game in a completely different direction with: 13...g5!? This comes with a health warning, but it may appeal to players with an appetite for risk. 14.Bg3 14.Be3 is less challenging, and after 14...g4 15.Ng1 f5 Black had excellent play in Radjabov – Kramnik, Linares 2003. 14...f5 15.Bd3 In the event of 15.Bc4N 15...Kh8 16.h4 g4 17.Ng5 Bxg5 18.hxg5 Qxg5 White has sufficient compensation for a pawn, but not more.

15...f4!N It is necessary to improve on 15...g4? 16.Ng1 f4 17.Bxh7† Kh8 18.Bf5, which was extremely dangerous for Black in Krasenkow – Schandorff, Esbjerg 2003. 16.Bxh7† 16.h4 g4 17.Ng5 Bd6„ is another messy position. 16...Kh8 17.Bf5 Bd6!÷ The last move improves on 17...Kg7?! 18.c4, which was given by Beliavsky & Mikhalchishin, after which Black’s position is too dangerous. Even though 13...g5 might be objectively playable, it would not be surprising if White had a big improvement somewhere, and the whole line seems unnecessarily risky. 14.Be2 c4! It is important to split White’s central pawns. After 14...Bf6?! 15.0-0 a6 16.c4² Black was doomed to a passive defence in Mamedyarov – Polgar, 308

Geneva (rapid) 2013. 15.0-0 Qc8!N The queen is heading for c5, where it will protect the c4-pawn and attack White’s a- and d-pawns, while also helping to connect the rooks. Beliavsky & Mikhalchishin do not consider this move, which is understandable as it was (and still is) untested – although it does transpose to an older correspondence game in which the queen went to c8 several moves earlier. 16.Rfe1 Bf8 17.d6 Qc5 18.Be3 Qc6

19.Qf5 19.Rd4 Ne5 20.Red1 Rad8³ 19...Nf6 20.Bd4 Re6÷/³ We have been following Thomsen – Koster, corr. 2006. The game ended in a draw, but it is White who has to play more precisely in order to maintain the balance. The d6-pawn is more of a weakness than a strength, and the pressure along the a8-h1 diagonal makes it hard for White to develop any activity. B23) 11.Bd3

309

This is the most popular move. White develops the bishop to its ideal square and overprotects the e4pawn, so that the queen can be relocated. 11...Qc8! 11...0-0 12.0-0 Rc8 13.Qe2 Nf6 14.Rfe1 has also been tested at a high level. However, in this case the presence of Black’s queen on her starting square renders Black’s set-up less harmonious. The text move is connected with an interesting positional concept of exchanging the light-squared bishops. White has two main replies: B231) 12.Qa4 and B232) 12.Qb1. As usual, it would be a mistake for White to allow an early queen trade: 12.0-0?! cxd4 13.cxd4 Qxc2 14.Bxc2 In Krasenkow – Gelfand, Bled (ol) 2002, the players agreed a draw here, but Black could have played on.

310

14...0-0N I prefer Black’s chances, for instance: 15.Bd3 Rac8 16.Rfc1 Nf6 17.Nd2 Rfd8³ The awkward-looking 12.Qb2 is unlikely to pose problems. 12...0-0 13.0-0 In Persson – Haugen, corr. 2003, Black’s simplest solution would have been: 13...Ba6N 14.Qe2 Bxd3 15.Qxd3 Rd8= B231) 12.Qa4

Preventing ...Ba6 is a serious option, but Black should be fine. 12...0-0 13.0-0 Rd8! 13...Nf6 has scored well but I consider the text move more accurate. Black is threatening to exchange 311

the important bishop on d3 by means of ...cxd4 followed by ...Nc5. 14.Qb5! 14.Rfd1N looks like a more natural way to protect the bishop but 14...Qc6! 15.Qxc6 Bxc6 reaches an unpleasant endgame for White, who will have to take care of some weak pawns. 14...Nb8 15.Qb1 Ba6 16.Rd1 Bxd3 17.Qxd3 Nd7

18.d5 White is playing in the most ambitious way possible. 18...exd5 19.exd5 c4 20.Qe2 Re8 21.Nd4 We have been following the high-level game Gelfand – Korchnoi, Tilburg 1992. My suggested improvement is:

312

21...Bf8!N 22.Qf3 Nc5 23.Nc6 Nd3„ Reaching a double-edged position where Black’s chances are no worse. B232) 12.Qb1

Finally we come to the most popular choice. 12...Ba6! This is the main idea behind Black’s previous move. The bishop exchange reduces White’s attacking abilities, and leads to a relatively peaceful middlegame with a healthy structure for Black.

313

13.0-0 0-0 14.Re1 Arguably the most natural move, supporting the e-pawn and adding force to a later d4-d5 push and subsequent opening of the e-file. Naturally I analysed some other moves as well: 14.Rd1 Bxd3 15.Qxd3 cxd4 16.cxd4 Qb7= was fine for Black in Yermolinsky – A. Ivanov, Philadelphia 2008. The immediate 14.d5 doesn’t offer White anything special: 14...Bxd3 15.Qxd3 c4 16.Qd2N (16.Qc2?! as played in Ghosh – Agdestein, Internet [rapid] 2017, should be met by 16...exd5N 17.exd5 Nc5 with some advantage for Black)

16...exd5 17.exd5 Re8 18.Rfe1 Qc5 19.Be3 Qd6= The passer has been neutralized, so Black is safe. 14.e5 is another ambitious move: White commits to a kingside attack but takes a positional risk by fixing his central structure. 14...Bxd3 15.Qxd3 Rd8 16.Qe4 (16.h4 h6 17.Rfd1 Qc6 18.Qe3 b5³ was excellent for Black in Zakharevich – Shipov, Elista 1994) 16...Nf8 17.h4 Qd7 18.Be3 Rac8 19.Qg4 cxd4 20.cxd4 This was Laxman – Bharambe, Mumbai 2016. Here I suggest going for safety on the kingside with:

314

20...f5N 21.exf6 Bxf6 22.Bg5 Bxg5 23.hxg5 Ng6 24.Rfe1 Rc6= Black has nothing to worry about. A more positional plan is: 14.a4 White removes the pawn from the sights of the e7-bishop and sets up a potential minority attack on the queenside. 14...Bxd3 15.Qxd3 Rd8 16.a5 This seems the most consistent. With that being said, there is nothing wrong with 16.Rfe1, which transposes to the main line below. 16...Nf6 17.Rfb1 Qb7 18.Bg5 cxd4 19.cxd4 After a series of natural moves, the players reached this position in Ingersol – Taboada, email 2003. My suggested improvement is:

315

19...h6N 20.Bxf6 20.a6 Qd7 21.Bh4 Rac8 22.Rc1 b5„ leaves Black with a passed pawn which might become dangerous in the long run. 20...Bxf6 21.axb6 axb6= The pawns on b6 and d4 are equally weak.

14...Bxd3 15.Qxd3 Rd8! This continuation seems most flexible. Releasing the tension in the centre seems somewhat premature: 15...cxd4 16.cxd4 Qb7 17.Rad1 h6 18.d5 White’s heavy pieces perfectly supported the powerful passed d-pawn in Erdos – Roser, Cappelle-laGrande 2008. 316

16.a4 In comparison to the note above, 16.Rad1?! now drops a pawn in view of 16...c4! 17.Qe2 Bxa3³.

16...cxd4 Another reasonable way of handling the position is: 16...a6!? 17.Rad1 cxd4 18.cxd4 Qc6„ 17.cxd4 Qb7 Black has reached a harmonious set-up, with his rooks connected and the queen safe from White’s pieces. 18.Rab1 The ambitious 18.a5 comes with a clear strategic risk, since Black can push: 18...b5! 19.a6 Qb6 20.Bd2 (I also examined 20.Be3 b4 21.Nd2 Rac8 22.Rec1 Rxc1† 23.Rxc1 Nb8! when White has nothing better than simplifying with: 24.d5 Qxa6 25.Qxa6 Nxa6 26.Ra1 Nc5 27.Rxa7 Bf8=)

317

20...b4 21.Reb1 h6 22.Rb3 Rab8 The strong b-pawn and vulnerability of the a6-pawn meant that White was under some pressure in Manannikova – Morokova, corr. 2013. White’s last move prepares a4-a5 without allowing ...b5, so Black’s most challenging answer seems to be:

18...a6! Maintaining the tension on the queenside. 18...Rac8 19.a5 h6 20.h3 looks slightly more pleasant for White. 19.Re2 318

The aggressive 19.h4 is well met by 19...Nf6! with pressure against the e4-pawn, tying White’s pieces to its defence. (19...Rac8 would be less precise in view of 20.Bg5 Bxg5 21.hxg5 when the open h-file might cause Black some problems) 20.Ne5 Rac8 White fails to create any threats on the kingside and his central pawns are rather vulnerable. 19...Rac8 20.h3 We have been following Koskela – Mrkvicka, corr. 2004. Now I really like the following original concept:

20...a5!N Black seems to be weakening the important b5-spot, but in return he gets a valuable outpost on b4, from where the bishop can neutralize any pressure along the b-file. Moreover, the a4-pawn is fixed as a potential target for the future. 21.Qb3 Qa6 22.Reb2 Bb4 Black’s pieces are getting more active, so White has to take care to maintain the balance. Conclusion This chapter concludes our coverage of the Petrosian System by dealing with the most popular 7.Qc2 variation. After the natural 7...Nxc3 we started by analysing 8.Qxc3, when 8...h6! is an important move to prevent the Bg5 pin. White has several playable continuations but Black should not be particularly worried by any of them. 8.bxc3 is more challenging. 8...c5 9.e4 Nd7 is our response, when 10.Bf4 Be7 is the main line. We looked at several options, the main ones being 11.d5 when Black has enough resources to deal with the 319

passer, and 11.Bd3 Qc8! with the plan of exchanging the light-squared bishops. A long battle lies ahead, but Black should be able to enter the middlegame on equal footing, with a safe and harmonious position.

320

A) 6.cxd5 exd5 7.Bg2 Bd6 8.0-0 0-0 160 A1) 9.Re1 162 A2) 9.Ne5 163 B) 6.Bg2 Be7 7.0-0 0-0 164 B1) 8.cxd5 165 B2) 8.Ne5 c5 167 B21) 9.Nb3 167 B22) 9.dxc5 169 B3) 8.b3 c5 171 B31) 9.dxc5 171 B32) 9.cxd5 173 B33) 9.Bb2 174 B34) 9.Ne5!? 176

321

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Ba6 Equally playable is 4...Bb7, but I chose to focus on the more dynamic text move. 5.Nbd2 This move was first seen in Pirc – Nimzowitsch, Bled 1931. Out of all White’s possible ways of protecting the c4-pawn, this one seems the most natural at first glance, as White develops the knight while also fighting for control over the key e4-square. However, it also suffers from some drawbacks: the knight can no longer be developed on the optimal c3-square, while the queen and dark-squared bishop are temporarily blocked, so Black has good prospects of fighting for the centre. 5...d5 Other moves are possible but I like this direct approach. Black’s energetic play is justified by the slightly passive placement of White’s knight. We will consider the immediate fixing of the central structure with A) 6.cxd5, followed by the more popular B) 6.Bg2. 6.b3 has no independent value; after 6...Be7 there is nothing better than 7.Bg2 0-0 8.0-0, transposing to variation B3 on page 171. A) 6.cxd5 exd5

322

This pawn exchange is double-edged. White has reached a potentially favourable pawn structure but Black can benefit from the semi-open e-file, while White’s knight would obviously be happier on c3 than on d2. 7.Bg2 A few games have continued: 7.Ne5 Bd6 This line does not have much independent value, since the following attempt to interrupt Black’s natural development proves ineffective: 8.Qa4†N White should prefer 8.Bg2 0-0 9.0-0 with a transposition to variation A2 below. The overoptimistic 8.b4? 0-0 9.Rb1 was played in Rajkovic – Benkovic, Subotica 2008. Black should have exploited his opponent’s poor development with: 9...c5!N 10.b5 Bc8 11.Bg2 Bf5 12.Rb2 a6µ

323

8...c6 9.Bg2 0-0 The e2-pawn is unprotected, so White has trouble castling. 10.Ndf3 Ne4 11.Nd3 Re8 12.0-0 Bb5 13.Qc2 c5 Black has excellent play in the centre. 7...Bd6 This is the most active way to develop the bishop, as it fights for control over e5 while leaving the efile open for a rook.

8.0-0 0-0 White’s two main options are A1) 9.Re1 and A2) 9.Ne5.

324

The quiet 9.b3 does not pose Black any problems, as the d2-knight is poorly placed. 9...Nbd7 10.Bb2 Re8 11.Re1 Qe7 12.e3 Ne4 13.Qc2 occurred in Chandra – Azarov, Dayton 2017, and now I like the following plan for Black:

13...Bb7N The bishop has done its job, so it drops back to reinforce the central light squares. 14.Rac1 f5! „ Black’s last move is justified by the fact that White is not in a good position to plonk his knight on e5. It is hard to suggest another constructive plan for him, while Black enjoys an active position with a strong knight on e4. A rare alternative is: 9.Nb1 Rerouting the knight to its best square makes a lot of sense but the manoeuvre costs valuable time. I only found one game from this position, which we will follow for a few moves. 9...Nbd7 10.Nc3 Re8 11.Re1 h6 12.Nh4 Bb7 13.Qb3 This was Cvek – Babula, Austria 2008. Here I found an interesting new idea:

325

13...Ne4!?N Black’s lead in development enables him to play aggressively, even sacrificing a pawn. 14.Nxd5 14.Nf5 Bf8 15.Nxd5 Ndc5! 16.dxc5 Bxd5 17.Qd3 Bxc5 18.Be3 c6= reaches a comfortable position for Black. 14...c5! This typical break works especially well here, as White’s pieces are uncoordinated. 15.Nf5 15.Be3?! is risky in view of 15...cxd4 16.Bxd4 Ndc5 17.Bxc5 Bxc5 18.Bxe4 Rxe4 19.Rad1 Qg5 when Black’s bishop pair looks scary. 15...cxd4 16.Nxd6 Nxd6 17.Qd3 Nf6 18.Nxf6† 18.Qxd4 Bxd5 19.Bxd5 Nf5–+ 18...Qxf6 19.Bxb7 Nxb7 With active pieces and chances to put pressure on the e2-pawn, Black is at least equal. A1) 9.Re1

326

This move should not be dangerous, as long as Black meets it safely with: 9...Bb7! The bishop no longer has anything to do on a6 so it bolsters the central light squares, without fearing any potential pinning of the d5-pawn along the h1-a8 diagonal. The careless 9...Re8?! runs into 10.e4!. For instance, 10...dxe4 11.Ng5 Nbd7 12.Ndxe4± and Black was in trouble in Engman – Aguettaz, Stockholm 2007. 10.Ne5 Nbd7 11.Ndc4 Be7

327

12.Bf4 12.Nxd7N 12...Qxd7 13.Ne5 Qe6 14.Bf4 c5 offers Black good play in the centre. 12...Re8 13.Nxd7 Qxd7 14.Rc1 c5 15.dxc5 We have been following Browne – Arnason, Reykjavik 1990, in which Black recaptured with the bishop. Instead I prefer:

15...bxc5N 16.Na5 Ba6 Black’s position is generally healthy, but it is necessary to check the following forcing line. 17.e4 Qb5! 18.e5 Qxa5 19.exf6 Bxf6 20.Bxd5

328

20...Rxe1† 21.Qxe1 Qxe1† 22.Rxe1 Rd8= Black’s active pieces fully compensate for White’s slightly preferable pawn structure. A2) 9.Ne5

This active move has been the choice of several strong GMs. 9...c5 This move is well timed, as White’s last move reduced his control over the d4-square, while the knight on e5 also becomes slightly unstable now. 10.Ndf3 10.Ndc4 is risky in view of: 10...dxc4! 11.Bxa8 cxd4 12.Nf3 (12.Nc6 Nxc6 13.Bxc6 Qc8 14.Bf3 Be5 15.Qa4 was seen in Schandorff – Fries Nielsen, Norresundby 1992, when 15...Nd7!N followed by ...Nc5 would have been excellent for Black) This position was seen in Ivanchuk – Leko, Monte Carlo (rapid) 2006. My suggestion is:

329

12...Bc5!N 13.Bf4 Re8 The powerful central pawns offer Black more than enough play for the small material sacrifice. 10...Re8

11.Bf4N 11.Nd3?! Nc6 12.Be3 c4 13.Nf4?! (13.Nde5 Rc8³) 13...b5 was clearly better for Black in Laxman – Ghosh, Bhubaneswar 2018. The text move seems like a more natural method of development. A logical continuation is:

330

11...h6 12.Rc1 Nbd7 13.Re1 Nf8 14.Nc6 Qd7 15.Bxd6 Qxd6 16.Nce5

16...Bb7 17.e3 Rac8„ Black has good prospects on the queenside. B) 6.Bg2

Keeping the tension is more popular. 6...Be7 In comparison to the previous variation, 6...Bd6 is less logical when the e-file is closed. This is well illustrated after 7.0-0 0-0 8.b3 Nbd7 9.e4! Nxe4 (9...dxe4 10.Ng5 also poses Black some problems) 331

10.Nxe4 dxe4 11.Ng5 Nf6 12.Nxe4 Nxe4 13.Bxe4 and White was better in Gyimesi – Gordon, Liverpool 2006. 7.0-0 White has tried several moves here but most of them transpose to later lines, as White has no good reason to delay castling. For instance, 7.Qc2 0-0 8.0-0 is covered under 8.Qc2 in the notes below; 7.cxd5 Nxd5 8.0-0 0-0 transposes to variation B1; and 7.Ne5 0-0 8.0-0 is variation B2. 7.Qa4† is a genuine alternative, attempting to disrupt Black’s smooth development, but White fails to achieve his objective after: 7...c6 8.Ne5 (8.cxd5 is also well met by 8...b5! 9.Qc2 cxd5 10.0-0 0-0 11.Ne5 Qb6 12.Nb3 Rc8 and Black was better in Petrov – Kurajica, Porec 1998) 8...b5! 9.cxb5 cxb5

10.Qd1 0-0 11.0-0 Qb6 12.Nb3 Nbd7 13.Bg5 Bd6 14.Bf4 b4 Black was fine in Lputian – Lautier, Palma de Mallorca 1989.

332

7...0-0 We have reached a significant branching point, where B1) 8.cxd5, B2) 8.Ne5 and B3) 8.b3 all deserve attention. 8.Qc2 This continuation seems slightly illogical as the White queen may be subject to harassment, either along the c-file or after ...Nc6 and a possible jump to b4 or d4. 8...c5 9.dxc5 9.Rd1 Nc6 10.dxc5 bxc5 11.a3 Rc8 12.e3 Qb6 was comfortable for Black in Kristiansen – Adamski, Polanica Zdroj 1975. 9...bxc5

333

10.Ne5 I also considered 10.e4 Nc6 11.exd5N (11.Rd1?! occurred in Zara – Nacht, Bucharest 1967, when 11...d4N 12.Ne1 e5 would have given Black the upper hand) 11...exd5 12.Re1 Re8 13.Qa4 Bb7 14.cxd5 Nxd5„ when the activity of Black’s minor pieces fully compensates for White’s preferable pawn structure. 10...Qc7 11.Ndf3 Bd6 12.cxd5 The natural-looking 12.Bf4N runs into 12...g5! 13.cxd5 gxf4 14.dxe6 fxe6 15.Ng5 Bxe5 16.Nxe6 Qd6 17.Nxf8 Nc6, reaching an unbalanced position where Black’s pieces are much more active. 12...exd5

13.Nd3 Re8 Black had no problems in Epishin – Gulko, Reggio Emilia 1991. B1) 8.cxd5 Nxd5 8...exd5!? 9.Ne5 c5 seems perfectly playable but I slightly prefer the knight recapture on this occasion.

334

9.Re1 This is White’s most ambitious option, unpinning the e2-pawn and preparing to seize space in the centre. 9.Ne5 c5 10.dxc5 Qc7 transposes to a line covered on pages 169-170 – see 10.cxd5 Nxd5 in the notes to variation B22. 9...c5 10.dxc5 I also examined 10.e4!?N 10...Nb4 11.d5, a thematic plan which should always be considered in such positions. The most accurate reply is:

11...Nd7! (I am less convinced by 11...exd5 12.exd5 Bb7 13.Nc4 when Black can take on d5 with any of 335

the three pieces, but White will obtain good compensation in each case) 12.dxe6 fxe6 13.Qb3 Qe8 14.Qxe6† Kh8ƒ White’s position is rendered difficult by his lack of harmony and the weakness of the d3-square. After the text move I found an important improvement.

10...bxc5!N Maintaining optimal coordination between our pieces is more important than any slight damage to the pawn structure. After 10...Bxc5 11.a3! Nc6 12.e4 Nf6 13.e5 Nd5 14.Ne4 the vulnerability of the bishop on c5 enabled White to develop some initiative in Kovac – Petrik, Slovakia 2011. 11.Ne5 Qc7 12.Ndc4 Rd8 13.Bd2 I also considered 13.Qa4 f6 14.Nd3 Qd7! when Black manages to interrupt the connection between the opponent’s pieces. For instance:

336

15.Qxd7 Nxd7 16.Na5 Rac8„ Black’s activity counts for at least as much as his extra pawn island.

13...Nd7 White’s pawn structure remains preferable but it is not so easy to maintain the blockading knight on c4, so Black’s isolated pawn may soon be able to advance. A logical continuation could be: 14.Qa4 Nxe5 15.Qxa6 Nxc4 16.Qxc4 Bf6 17.Rab1 Rab8 18.b3 Nb6 19.Qc2 c4 White’s bishop pair is countered by Black’s active piece play, so the position remains balanced. B2) 8.Ne5

337

This has been used by several strong GMs. 8...c5 Once again, this advance becomes more effective when White has placed his knight on e5 and delayed the development of his queenside pieces. White has two main replies: B21) 9.Nb3 and B22) 9.dxc5. 9.cxd5 is a reasonable move but it does not have much independent significance after 9...Nxd5. For instance: 10.Ndf3N (10.dxc5 has been the universal choice so far, after which 10...Qc7!? reaches a line covered on pages 169-170; see 10.cxd5 Nxd5 in the notes to variation B22) 10...cxd4 11.Nxd4 Bf6 12.Bf4 Bb7 13.Ng4 Bxd4 14.Qxd4 Nc6 Black is fine. B21) 9.Nb3 This somewhat surprising way of putting pressure on Black’s central pawns was employed by Hikaru Nakamura, among other strong players. 9...Bb7 10.dxc5 bxc5

338

11.Bg5 An important alternative is: 11.cxd5 exd5 11...Bxd5 has proved to be a reliable choice but I prefer the text move, since it keeps more dynamic potential in Black’s position. 12.Bg5

12...Na6! It is vital to complete development as soon as possible. It turns out that White’s knights are not well placed, and Black can look to mobilize his central pawns at a suitable moment. 12...Nbd7? 13.Nxd7 Qxd7 14.Bxf6 gxf6 left Black with too many weaknesses in Kazoks – Camilleri, corr. 1985. 339

12...Qc7?! would also be careless in view of 13.Bxf6! Bxf6 14.Ng4 Nd7 15.Nxf6† Nxf6 16.Rc1 when Black is heading for serious strategic troubles. 13.Qd2 13.Bxf6N 13...Bxf6 14.Ng4 gives White no advantage after 14...Rc8 15.Nxf6† Qxf6 16.Qd2 Qb6 followed by ...d4. The text move was played in a correspondence game, which is worth following for a few more moves. 13...Rc8 14.Rad1 Qc7 15.Bxf6 Bxf6 16.Ng4 Be7 17.Ne3

17...d4 17...Rfd8!?N also deserves attention. For instance, 18.Nxd5 Bxd5 19.Bxd5 c4 20.Nc1 Nb4 21.e4 g6 22.Rfe1 Qb6 and Black’s active pieces provide full compensation for the pawn. 18.Nd5 Bxd5 19.Bxd5 Nb4 The position remained balanced in Kund – Serban, corr. 2005.

340

11...Na6 12.Qd2 Rc8 13.Rfd1!?N This seems to me to be the most consistent choice. 13.cxd5 exd5 transposes to the Kund – Serban game in the previous note. 13.Rac1 was played at the elite level but 13...h6!N is a strong reply. (13...Qc7 14.Bxf6 Bxf6 15.Ng4 Be7 16.cxd5 exd5 was about equal in Nakamura – Karjakin, Wijk aan Zee 2013) Play may continue:

14.Bxf6 (14.Bf4 Re8 15.Rfd1 Bd6³) 14...Bxf6 15.Ng4 Bg5 16.f4 Be7 A complex position has arisen, where I prefer Black. The bishop pair is a useful asset, while the f2-f4 advance has permanently loosened White’s kingside. The text move intensifies the pressure on the d5-pawn, so Black has to do something about the pin. Fortunately, he can solve his problems with a pawn sacrifice. 341

13...Qc7! 14.Bxf6 Bxf6 15.Ng4 Be7 16.cxd5 exd5 17.Bxd5 Rfd8 18.e4 White has succeeded in winning a pawn, but Black’s pieces are better coordinated. My analysis continues: 18...Nb4 19.Ne3

19...Bg5! 20.Qe2 After 20.f4?! Bf6 21.Rac1 Bxd5 22.exd5 Qb6ƒ Black will regain the pawn while keeping all the merits of his position. 20...Bxe3 21.Bxb7 Qxb7 22.fxe3 Qxe4 342

Black is by no means worse. B22) 9.dxc5

9...Qc7!? This is an interesting way to pinpoint the drawback of White’s set-up – namely the fact that the knight on e5 cannot be conveniently protected. 9...bxc5 is perfectly playable; 10.b3 is the most logical reply, which transposes to variation B31 on page 171. 10.Nd3 I checked two alternatives: 10.Ndf3 bxc5 11.cxd5 11.b3?! is too slow. 11...Bb7 12.Bb2 occurred in Rost – Dutra Neto, email 2004, and now after 12...d4!N 13.e3 dxe3 14.fxe3 h6³ Black has clearly won the opening battle. 11...Nxd5 Once again, Black’s active pieces will make up for his slight structural deficiency. 12.Qc2 Nc6 13.Nxc6 Qxc6 14.Bg5 f6 15.Bd2 Qb6 16.Rfe1 This occurred in Carlsen – Karjakin, Sao Paulo/Bilbao 2012. I think the most accurate continuation is:

343

16...Rab8N 17.b3 Nb4 18.Qb2 Rfd8= The activity of Black’s pieces fully compensates for the split queenside pawns. White may also try: 10.cxd5 Nxd5

11.Bxd5!? 11.Ndf3 bxc5 transposes to the Carlsen – Karjakin game noted above. 11...exd5 12.Ndf3 Bxc5 13.Bf4 13.Qxd5 Bb7 14.Qd3 Nc6 15.Bf4 Nxe5 16.Bxe5 Qc6 17.Rad1 Rfe8 18.Bc3 a5 gave Black full compensation for the pawn in Kamanel del Corral – Peri, corr. 2013. 13...Qb7 14.Rb1 f6 15.Nd3 344

This was Babula – Markos, Banska Stiavnica 2012. Here I suggest:

15...Be7N 16.Bxb8 16.Nd4 Nc6 17.Nxc6 Qxc6 18.Rc1 Qe8 looks somewhat shaky for White, whose king is left without its main defender. 16...Raxb8 17.Qa4 Rbc8 18.Rfd1 Rfd8 The powerful bishops provide Black with good prospects.

10...bxc5 11.Qa4 Bb7 12.Nb3 We have been following Brousek – Le Duigou, email 2002. I like the following way of handling Black’s position:

345

12...Bc6N 13.Qa5 Qb6! Highlighting the awkward placement of the queen on a5. 14.Ne5 Bb7 15.Rd1 15.Bg5 is no improvement: 15...h6 16.Bxf6 Bxf6 17.Ng4 Be7 18.Rfd1 dxc4 19.Qxb6 axb6 20.Bxb7 Ra7= 15...Rc8 16.Bg5 Nc6 17.Nxc6 Bxc6 18.cxd5 Bxd5 Black has a comfortable game, for instance: 19.e4 Bb7

346

20.e5 Bxg2 21.Qxb6 axb6 22.Kxg2 Nd5= The position remains balanced. B3) 8.b3 This has been somewhat less popular than 8.Ne5 but one could argue that it is more logical to finish developing the queenside before taking action in the centre. The text move protects the c4-pawn and prepares to develop the bishop on b2. Later, White may look to prepare e2-e4.

8...c5 We will consider B31) 9.dxc5, B32) 9.cxd5, B33) 9.Bb2 and B34) 9.Ne5!?. As you might expect, 347

several transpositions are possible and I have tried to point out the most likely of them. B31) 9.dxc5 bxc5 10.Ne5 Bb7 White is keeping the tension along the long diagonal; but as we will see, he is unlikely to benefit from it.

11.Bb2 Nbd7 12.Nd3 12.e3 should not bother Black, as it is rather slow. After 12...Nxe5 13.Bxe5 Qb6 14.Qe2 a5 15.Rab1 Rfd8 16.Rfd1 Qa6= Black was fine in Andersson – Schreiber, corr. 2007. It is worth checking a slightly more forcing way for White to exchange knights: 12.Nxd7 Qxd7

348

13.cxd5 exd5 14.e4 d4! Quite a committal move, since it gives White an outpost on c4 and a mobile pawn majority on the kingside. On the flip side, the b2-bishop is blocked and Black has good prospects on the queenside. 14...Rad8 seems less precise in view of 15.exd5 Bxd5 16.Nc4, when White’s better queenside structure gave him a slight advantage in Kramnik – Anand, Frankfurt 1999. 15.Re1 Rfe8

16.Nc4 I also considered 16.e5N 16...Nd5 17.Ne4 but found 17...f5! 18.exf6 Nxf6 to be fine for Black. 16...Bf8 17.Qd3 Qe6 18.Rad1 a5„ In this complex position, Black was by no means worse in Krzyzanowski – Kukk, corr. 2002.

349

12...Qb6 13.e3 After 13.cxd5 exd5 14.Rc1 Rfe8 15.Rc2 Bf8 16.Re1 a5 The pressure on the b3-pawn yielded Black decent counterplay in Skembris – A.H. Williams, Thessaloniki (ol) 1984. 13...a5 14.Rb1 Bc6 15.Re1

15...Rfe8!N This innovation offers Black comfortable play. 15...a4 proved to be slightly premature after 16.cxd5 exd5 17.b4! cxb4 18.Ba3² in Murray – P. Gaprindashvili, corr. 2002. 350

16.cxd5 exd5 17.e4 17.Qc2 Bf8 18.Rbc1 Rac8„ is fine for Black.

17...dxe4 18.Nxe4 Nxe4 19.Bxe4 Bf8 20.Bxc6 Qxc6= Black’s control over the light squares fully compensates for his slight structural weaknesses. B32) 9.cxd5 This forces Black to make an important decision.

9...exd5! 351

9...Nxd5 10.Bb2 Nc6 11.dxc5 Bxc5 reaches a position with a symmetrical pawn structure, where 12.a3! offers White a slight plus. The text move is more challenging, as the d2-knight remains rather passive. 10.Ne5!? 10.Bb2 Nc6 leads to variation B33 on the next page. Despite being unusual, White’s last move is not without venom. In my opinion, Black should insist on placing the knight on c6 with: 10...Qc7! 10...cxd4 11.Bb2 Bc5 12.Ndf3 Re8 13.a3 Nbd7 14.Nd3 has been tested at a high level, but it seems to me that White manages to regain the pawn in a favourable situation. 11.Bb2

11...Nc6N Improving over 11...Nbd7, as was played in one game. 12.Nxc6 Another logical continuation is 12.Rc1 Rad8 13.Nxc6 (13.Ndf3 Ne4 14.Nxc6 Qxc6 15.Ne5 Qe6„) 13...Qxc6 14.dxc5 bxc5 15.Re1 Qb6 16.Qc2 Rfe8„ when Black’s hanging pawns are well supported by his harmoniously placed pieces. 12...Qxc6

352

13.dxc5 I also checked 13.Rc1 Qe6!? (13...Rad8 is a decent alternative which transposes to the previous note) 14.Re1 Ne4 15.Nxe4 dxe4 16.Qc2 f5 17.dxc5 Bxc5 when the powerful bishop on b2 is balanced out by Black’s space advantage. 13...bxc5 I found a correspondence game which reached this position via a different move order. Play continued: 14.Re1 Rad8 15.Qc2 Qe6 16.Rac1 In Schrancz – Kozlowicz, email 2003, Black went for the overly committal 16...d4?! and eventually lost. A simple improvement is:

353

16...Rfe8N Maintaining the tension. Play might continue: 17.e4 d4 18.Ba3 Rc8„ A long, strategic struggle lies ahead. B33) 9.Bb2

This has been the most common continuation, though it has scored rather poorly. 9...Nc6 10.cxd5 354

10.dxc5 bxc5 11.cxd5 exd5 12.Rc1 Re8 13.Re1 Rc8 transposes to the main line below. 10.Rc1 Nxd4 11.Nxd4 cxd4 12.Bxd4 Rc8 13.cxd5 occurred in Hacker – Hertweck, Muenster 1987. My natural improvement is:

13...Nxd5N 14.Rxc8 Qxc8 15.Qa1 Qc2 16.Nc4 Bxc4 17.bxc4 Nf6 Black has comfortable play. 10...exd5 11.Rc1 Re8

This is the kind of position where the drawbacks of White’s knight on d2 are obvious. Black’s pressure on the d4-pawn feels strong, while Black does not have to worry about the typical pressure on the d5pawn, which would be present if the knight was on c3.

355

12.Re1 Rc8 13.dxc5 13.Bh3 Rc7 14.dxc5 was seen in Bernasek – Navara, Czech Republic 2012. My improvement is:

14...bxc5!N Presumably Navara rejected this due to 15.Bxf6, but after 15...Bxf6! 16.Rxc5 Nb4 17.Rxc7 Qxc7 18.Qb1 Qa5 the powerful bishops provide Black with full compensation for the pawn. 13...bxc5 Again I favour the hanging-pawns structure, which preserves more dynamic potential in the position. 13...Bxc5 proved worse after 14.Bh3 Rc7 15.a3 Ne4 16.e3² in Sandler – Rogers, Adelaide 2004. 14.e4 This advance is absolutely thematic but it has one disadvantage, which becomes clear after Black’s next move. 14...Nb4! The most energetic response, aiming for the d3-square. 15.Ne5N This is White’s best. The ambitious 15.e5? was seen in Korobov – Iordachescu, Ohrid 2001. Here Black missed a chance to exploit the unfortunate placement of the d2-knight by means of:

356

15...Ng4!N 16.e6 f5!µ 15...dxe4 16.Nxe4 Qxd1 17.Rexd1 Nxe4 18.Bxe4 Nxa2 19.Ra1 Nb4 The activity of White’s pieces enables him to maintain the balance by regaining the sacrificed pawn, but he is not fighting for an advantage. B34) 9.Ne5!?

White temporarily gives up the d4-pawn in order to activate his knight. 9...cxd4 10.Ndf3 357

10.Bb2 is not particularly challenging. 10...Bb7 11.Bxd4 Nbd7 12.cxd5 (12.Ndf3 dxc4 13.Nxc4 Rc8 was equal in Weiss – Mutsuaki, email 1996) 12...Nxe5 13.Bxe5 This occurred in Rakic – Kalashyan, Herceg Novi 2008, when 13...Nxd5N 14.Nc4 b5 would have been fine for Black. The text move is slightly trickier. Both of White’s knights are heading towards c6, so Black should be careful. The best solution has already been demonstrated in some correspondence games. 10...Bc5 11.Nxd4 Nbd7! Having moved his bishop away from e7, Black will not be harmed by the appearance of a knight on c6. 12.Nec6 Qc7 13.cxd5 Nxd5

14.Bxd5!? This has been the universal choice in correspondence games, all six of which were subsequently drawn. White gives up the important light-squared bishop in order to disrupt Black’s coordination. I also considered 14.Bb2N 14...Bb7 15.e4 N5f6 16.b4 Bxc6 17.Nxc6 Qxc6 18.bxc5 bxc5 when White may be able to claim sufficient compensation for the pawn, but nothing more. 14...exd5 15.Bf4 Qc8 16.b4 Another game continued 16.Rc1 Re8 17.Nb4 (17.b4N 17...Bc4! 18.bxc5 bxc5 transposes to the main line below) 17...Nf6 18.Nxa6 Qxa6 when the activity of Black’s pieces fully compensated for the isolated d-pawn in Roesch – Wernikiewicz, corr. 2016.

358

16...Re8! The game is entering a tactical phase, where both sides must tread carefully and equality will be the outcome. 17.bxc5 17.Qd2 Bxd4 18.Nxd4 Bc4 was level in Satici – Dunlop, corr. 2014. The isolated pawn is not particularly vulnerable, while Black’s bishop is excellent and White has to keep an eye on the weak light squares around his king. 17...bxc5 18.Rc1! 18.Re1 cxd4 19.Nxd4= occurred in Cheek – Ivec, corr. 2016. The evaluation is much the same as in the above game; the missing b-pawns do not particularly favour one side or the other.

359

18...Bc4! The key move, after which Black has no particular problems. 19.Na5 cxd4 20.Nxc4 Another game continued 20.Qxd4 Nb6 21.Rfe1 Qe6= and the game went on for a while, but the final result was the same in Ham – Edwards, corr. 2015. 20...dxc4 21.Qxd4 Nb6 22.Be3 The passed pawn ensures that Black will always have enough counterplay. Indeed, a correspondence game in 2016 was agreed drawn here. In an earlier encounter, White ‘fought on’ for longer:

360

22...Qc6 23.Rfd1 Re4 24.Qd6 A draw was agreed here in Legemaat – Meiners, corr. 2014. Conclusion 5.Nbd2 is a serious option, which I recommend meeting with 5...d5. Although White sometimes exchanges in the centre immediately, the majority of games continue 6.Bg2 Be7 7.0-0 0-0, at which point many plans and move orders are possible. One of the main themes of the chapter is the fact that White’s knight is poorly placed on d2, whereas it would be much happier on c3. Nevertheless, White has several options and Black must be ready to handle a multitude of different pawn structures, especially after carrying out the ...c5 advance. Hanging pawns are a common occurrence, while sometimes it makes more sense to play with isolated a- and c-pawns, with active pieces to compensate for the weakened structure. The final variation with 8.b3 c5 9.Ne5!? leads to especially interesting play, but correspondence games have demonstrated that Black’s position is absolutely reliable.

361

A) 7.d5!? 180 B) 7.0-0 cxd4 8.Nxd4 Bxg2 9.Kxg2 Be7 10.Nc3 0-0 182 B1) 11.e4 183 B2) 11.Rd1 185 C) 7.dxc5 bxc5 8.0-0 Be7 9.Nc3 0-0 10.Rd1 Qb6 186 C1) 11.Rb1 188 C2) 11.Qb3 189 C3) 11.Qc2 191 C4) 11.Bf4 d6 192 C41) 12.Qb5 193 C42) 12.Rd2 195 C43) 12.Rab1 Nbd7 13.b4 Rfb8! 196 C431) 14.Be3!? 197 C432) 14.a3 200

362

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Ba6 5.Qa4 This method of protecting the pawn dates back to the game Grünfeld – Nimzowitsch, Breslau 1925, which also happens to be the first game on the database in which 4...Ba6 was played. Apart from defending the c4-pawn and eyeing the bishop on a6, the queen also pins the d7-pawn. However, in comparison to 5.Qb3, White is not supporting a possible d4-d5 advance, which makes the plan of ...c5 more effective. 5...Bb7 Black is not losing a tempo at all, since the queen will be slightly worse on a4 than on d1 when the ...c5 break comes. It seems less precise to start with 5...c5, as it yields White an interesting extra option: 6.d5!? (6.Bg2 Bb7 transposes to our main line) 6...exd5 7.cxd5 Bb7 8.e4 Qe7 9.Bd3 Nxd5 10.0-0 White had promising play for the pawn in Pedersen – Dolezal, corr. 2012.

6.Bg2 6.Nc3 has no independent value: 6...c5 7.dxc5 (after 7.d5? exd5 8.cxd5 Nxd5 White has no real 363

compensation for the pawn) 7...bxc5 8.Bf4 Be7 It’s obvious that White has nothing better than 9.Bg2 followed by 0-0, transposing to normal pathways. 6...c5 We will consider the rare but interesting pawn sacrifice A) 7.d5!?, followed by the more commonplace B) 7.0-0 and C) 7.dxc5. A rather passive alternative is: 7.e3 Be7 8.0-0 0-0 You don’t really need to know any specific details from here, but I have included a few illustrative lines to show how Black can handle the position when White plays with no ambition. 9.Nc3 After 9.Rd1 Qc8 10.Nc3 cxd4 11.exd4 d5 12.cxd5 Nxd5 13.Nxd5 Bxd5 White had no real compensation for the isolated pawn in Litvinov – Kachar, Saratov 2006. 9...Qc8! An excellent move. Black avoids the potential opposition of rook and queen along d-file and makes d4-d5 impossible, since the b7-bishop is now protected. 10.Bd2 We have been following Gjorgjieski – Kukov, Skopje 2011. My new way of handling the position is:

10...d6!N My other idea was 10...cxd4N 11.exd4 d5 but after 12.cxd5 Nxd5 13.Rac1 Black’s queen is clearly misplaced. 11.Rac1 Nbd7 12.Rfe1 Ne4 Black has excellent play due to his control over the light squares.

364

A) 7.d5!?

This pawn sacrifice has been a rare choice. The same idea has become extremely popular in the 5.Qc2 variation, so it is worth checking it here as well. Overall, the vulnerable location of the queen on a4 makes it easier for Black to solve his development problems, but it is important to recognize an important difference in the way Black should capture on d5. 7...exd5 8.cxd5 Bxd5! 8...Nxd5?! would be the correct choice with the queen on c2 but here it does not make much sense, as after: 9.0-0 Be7 10.Rd1

White’s queen cannot be attacked by the knights. Play may continue 10...Nc7 11.Nc3 0-0 12.Bf4 and White obtains excellent compensation for the pawn. 365

The text move is now known to be a serious mistake in the 5.Qc2 variation, but here it is fine as Black will gain a whole tempo after dropping the bishop back to c6. 9.0-0 I also checked: 9.Nc3 Bc6 10.Qc2N 10.Qb3 Be7 11.Ne5 (11.0-0 transposes to the main line below) 11...0-0 12.Nxc6 Nxc6 13.0-0 Rc8 14.e4 d6 left Black with no problems due to his control over the d4-square in A.S. Karlsson – Jonsson, Reykjavik 1993.

Funnily enough, the current position has been reached many times in the 5.Qc2 variation, but always with White to move instead of Black. Although that might sound like a cause for celebration, you should be aware that the practical results with White to play have been nothing short of a massacre in White’s favour. Obviously the extra tempo is a big help for Black, but White still undeniably has some compensation. Therefore it makes sense not to be too greedy and focus on fast development with: 10...d5 11.Ne5 Be7 White has a few ways of regaining his pawn, but none of them offer him any advantage. 12.Bg5 12.0-0 0-0 13.Rd1 Bb7 14.Qb3 Qe8 15.Nxd5 Nxd5 16.Bxd5 Nc6 17.Nd7 Nd4 18.Rxd4 cxd4 19.Nxf8 Bxd5 20.Qxd5 Bxf8 is another possible line resulting in equality. 12...0-0 13.Nxc6 Nxc6 14.Bxf6 Bxf6 15.Nxd5 Rc8

366

16.0-0 In the event of 16.0-0-0 Nb4 17.Nxb4 cxb4 18.Rxd8 Rxc2† 19.Kxc2 Rxd8= White’s slight structural edge is meaningless. 16...Nb4 17.Nxf6† Qxf6 18.Qb3 Rfd8 19.Rfd1 Nc6= The game continues, but Black is fine.

9...Be7 10.Nc3 Bc6 11.Qb3 0-0 12.Bf4 In the event of 12.Rd1N 12...d6 13.Bf4 Nh5 14.e3 Nxf4 15.exf4 Nd7 16.Rd2 Nf6 White has some compensation due to his pressure along the d-file, but it does not yield any more than equality. The text move was played in an all-GM encounter in 2017, which we will follow for a few more moves.

367

12...Na6 13.Rad1 Nc7

14.e4 The best and most ambitious choice. If 14.Nh4 Bxg2 15.Kxg2 Ne6 16.Be5 Ng4 White fails to create any serious threats. 14...Ne6 15.e5 I also considered 15.Ne5N 15...Nd4 16.Qc4 Bb7 when the strong knight offers Black decent counterplay. For instance:

17.Be3 Qb8 18.Nd3 b5 19.Nxb5 Nxb5 20.Qxb5 Bxe4 White should be able to draw but he is definitely not playing for a win. 368

15...Ne8 16.Be3 All this was played in Markos – Kelires, Heraklion 2017. The presence of the e5-pawn in Black’s camp makes it difficult to activate the knight and connect the rooks. Therefore, it makes sense to play:

16...d6!N The pin along the d-file is of no concern. For instance: 17.h4 Qc7 18.exd6 18.Nd5 Bxd5 19.Qxd5 Rd8³ 18...Nxd6

369

19.Nd5 Bxd5 20.Qxd5 Rad8 The onus is on White to demonstrate sufficient compensation for the pawn. B) 7.0-0 Even though this move has been employed by such great players as Korchnoi, Polugaevsky, Nikolic and others, Black’s task seems much easier here than in variation C with 7.dxc5. Tournament practice has clearly demonstrated that the exchange of light-squared bishops makes it easier for Black to develop his counterplay on the queenside.

7...cxd4 8.Nxd4 Bxg2 9.Kxg2 Be7 370

9...Qc8 is the most popular continuation but I don’t see any reason to commit the queen so soon, as it can head for b7 any time. In some situations Black may wish to put the queen on c7 instead of c8, in order to keep the option of a quick ...Rc8. 10.Nc3 There is hardly anything better than this natural developing move. 10.Rd1 0-0 11.Nc3 is merely another way of reaching variation B2 below. I also checked 10.Bf4 0-0 11.Rd1 Qc8 12.Nb5 (12.Nc3 Qb7† reaches variation B2) 12...Qc6† 13.f3 a6 14.N5c3 Qb7= as played in Amado – Franco Ocampos, Villa Ballester 1981. White’s eccentric knight manoeuvre has not particularly harmed his chances, but it has certainly not brought him any closer to an opening advantage. 10...0-0

White has tried a number of move orders from here; I will focus on the most natural options of B1) 11.e4 and B2) 11.Rd1. B1) 11.e4 Like in many Hedgehog positions, White is aiming to seize space. 11...Qc7 12.Bf4 After 12.Rd1?! Rc8! 13.b3 Qb7 14.f3 a6³ White suffered from the unfortunate placement of his queen in Brunk – Genov, Berlin 1993. Lines like this are the reason why I prefer to avoid the more common

371

move order with 9...Qc8. 12...Qb7 The queen harmoniously replaces the bishop and supports both ...b5 and ...d5, while also menacing White’s king.

13.f3 I also considered 13.Qc2N 13...Rc8 14.Qe2 as a natural attempt to regroup the queen, but it takes a lot of time. Play may continue: 14...Nc6 15.Rad1 Nxd4 16.Rxd4

16...Bc5 17.Rd3 Bb4 18.f3 Rxc4 19.Rxd7 Nxd7 20.Qxc4 Be7= White may have avoided the loss of a pawn, but Black is still comfortably equal.

372

13...a6 14.Qc2 14.Rfd1 transposes to variation B2 below. The disadvantages of leaving the queen on a4 for too long are illustrated in the following case: 14.Rad1 d6 15.Nc2 Rd8 16.Rf2 Nc6 17.Rfd2 Ne8 18.Be3

18...Rdc8! Black starts to put pressure on the c4-pawn. 19.Na3 (19.Bf2 Nf6³) 19...Nf6 20.Bf4 Ne5³ White had some problems in Mukhtarov – Gashimov, Baku 2001. 14.e5 is playable although Black can liberate himself quite easily with: 14...Nh5 15.Be3

15...f6! 16.exf6 Nxf6 17.Rad1 Rc8 18.Rfe1 Nc6 19.Nxc6 Rxc6 With mutual chances, Baikov – Shariyazdanov, Elista 2001. 373

14...d6 15.Be3 The threat of ...e5 forces White to waste a tempo, which doesn’t bode well for him, despite the relatively closed character of the position. 15...Nbd7 16.Rac1 Here I found a slight improvement over the existing games.

16...Rfc8!N In my opinion, it makes sense to leave the queen’s rook on its initial spot for a while. After 16...Rac8 17.Qe2 Rc7 18.b3 White had the traditional space advantage in Voiculescu – Rost, corr. 2010. 17.Qe2 d5! This typical break enables Black to activate his pieces and equalize cleanly. 18.exd5 exd5 19.cxd5 Nxd5 20.Nxd5 Qxd5 21.Rxc8† Rxc8 22.Qxa6

374

22...Ra8 22...Rc4 is also good enough. For instance, 23.Rd1 Ne5 24.Rd2 Rxd4 25.Bxd4 Nxf3 26.Qc8† Bd8 and perpetual check is unavoidable. 23.Qb5 Qxb5 24.Nxb5 Rxa2= The extra pawn is gone, so White fails to claim any advantage. B2) 11.Rd1

Black’s plan remains pretty similar after this move.

375

11...Qc7 12.Bf4 12.Ndb5 Qc6† 13.f3 a6 14.Nd6 Qc7 15.Nce4 Nxe4 16.Nxe4 occurred in Bykhovsky – Goldin, Rishon LeZion 1997. A natural improvement is:

16...Rc8N 17.b3 f5 18.Bf4 e5 19.Bg5 fxe4 20.Bxe7 exf3† 21.exf3 d5 With excellent play for Black. 12...Qb7† 13.f3 After 13.Nf3 Rc8 14.Kg1 a6 White already has to be careful to maintain the balance. For instance:

15.Rac1 Nc6 16.Bg5 (16.Bd6?! Bxd6 17.Rxd6 has occurred in a few games but for some reason no one continued with 17...b5!N when Black takes over the initiative) 16...Rc7 Black was fine in Languidey – El Debs, Curitiba 2016.

376

13...a6! It is vital to take control over b5. After the inaccurate 13...Rc8?! 14.Ndb5 Qc6 15.Rac1 Black was doomed to passive defence in Euwe – Alekhine, Netherlands 1938. 14.e4 14.Rac1 was seen in some high-level games, but after 14...Rc8 White has nothing better than 15.e4, transposing to our main line. Surprisingly, the most common 14.Rd2 may land White in difficulties: 14...Rc8 15.Rad1 d6 16.Nc2

377

16...b5! 17.cxb5 d5 18.Nd4 Nbd7ƒ Black’s initiative was worth more than the missing pawn in Tugsavul – Mer, corr. 2011.

14...Rc8 15.Rac1 Once again 15.Rd2?! does little to improve White’s position, and after 15...d6 16.Nde2 Ne8 17.Rc1 Nd7³ Black had the easier play in Chetverik – Tunik, Sochi 2005. The ambitious 15.e5 leads nowhere after 15...Nh5 16.Be3 f6 17.exf6 Nxf6 when White’s king was somewhat exposed in V. Ivanov – Yunusov, Russia 1991. 15...d6 16.Nc2 Ne8 17.Ne1 Nc6 18.Nd3 White is manoeuvring sensibly but his last move relieved the pressure along the d-file, allowing Black to improve his passive knight.

378

18...Nf6! 19.Be3 Nd7 20.Qc2 b5 Black succeeded in launching his queenside counterplay in Galanov – Badolati, corr. 2006. C) 7.dxc5 bxc5 7...Bxc5 is a solid alternative but I find it more attractive to change the pawn structure by capturing towards the centre, while taking control over the d4-square.

8.0-0 White can hardly benefit from delaying castling. For instance, 8.Nc3 Be7 9.Bf4 0-0 10.Rd1 Qb6 and by now there is clearly nothing better than 11.0-0, transposing to 10.Bf4 Qb6 11.Rad1 in the next note 379

below. 8...Be7 9.Nc3 0-0 10.Rd1 This has been the most popular choice by far. 10.Qb3 has little independent value after 10...Qb6. For instance: 11.Qxb6 axb6 12.Bf4 (12.Rd1 Nc6 transposes to 12.Qxb6N 12...axb6 in the notes to variation C2, on page 189) 12...Nc6 13.Ne5N 13...Rfd8 14.Rfd1 Na5= and Black is fine. The other main option is: 10.Bf4 Qb6 This has occurred in quite a lot of games, most of which have transposed to one of our main lines after a subsequent Rfd1. Here I will just mention a few independent possibilities. 11.Rad1 11.Qb5 d6 12.Qxb6 axb6 13.Rfd1 Rd8 transposes to variation C41; see page 194 for the continuation, with 14.Nb5 the main line. 11...d6 12.Rfe1 12.Rd2 Ne4!N 13.Nxe4 Bxe4 is similar to variation C42 and is likely to transpose after a subsequent Rfd1.

This position was reached in Salov – Milos, Las Vegas 1999. White’s last move was clearly intended to prepare an advance of the e-pawn. Therefore it makes sense for Black to play: 12...Rd8N Overprotecting the d6-pawn in anticipation of the following moves. 13.e4 Nc6 Black has excellent prospects due to his control over the d4-spot.

380

10...Qb6 This is the best place for the queen, where it is relatively safe and can protect the d-pawn for the foreseeable future (once it arrives on d6). We will analyse four main options: C1) 11.Rb1, C2) 11.Qb3, C3) 11.Qc2 and the most popular C4) 11.Bf4. 11.Qb5 has been played several times but I consider it completely harmless. A good example continued: 11...Qc7 (11...Rd8 is another solid continuation but I would prefer to keep more tension in the position) 12.Qb3 (12.Bf4 d6 13.Rd2 a6 14.Qb3 leads to the same position) 12...a6 13.Bf4 d6 14.Rd2 Nc6 15.Qd1 Rad8 16.Bg5 h6 17.Bxf6 Bxf6 Black had excellent prospects in Priehoda – Carlsson, Czech Republic 2018. C1) 11.Rb1 Rd8 White’s last move supports a possible advance of the b-pawn, so it makes sense to avoid ...d6 for the moment. 12.Bf4

381

12...h6 This prophylactic move restricts the mobility of White’s minor pieces. 12...d6? 13.b4! offers White a dangerous initiative. However, 12...Nc6 seems like a playable alternative. 13.Bd6 Bxd6 14.Rxd6 Qc7 Black will neutralize the temporary activity of White’s rooks: 15.Rbd1 (I also considered 15.Nb5N 15...Qb8 16.Rd2 d6 17.Ne1 a6 18.Nc3 Ne5=) 15...a6 16.Qc2 Rab8 17.R6d2 h6= Black had no particular difficulties at this stage in Nikolic – A. Ivanov, Acqui Terme 2015. 13.Qb5 The main alternative is: 13.a3 a5! It is important to understand that giving away the b5-outpost does not lead to any serious problems. 14.Qb5

382

14...Qa7! This retreat looks somewhat awkward, but the queen will only be temporarily misplaced. 14...Qxb5 seems less attractive. In Kramnik – Bacrot, Dortmund 2009, White recaptured with the cpawn and got nowhere, but the simple 15.Nxb5N improves. For instance, 15...Be4 16.Bc7 Bxb1 17.Bxd8 Bxd8 18.Rxb1 and White has a small edge due to his more active bishop. 15.Bd6 After 15.Qb3?! d6³ White was running short of constructive ideas in Gavrilov – Myshkin, Moscow 2012. 15...Bxd6 16.Rxd6 Bc6 17.Qb3 Qc7 18.Rbd1 Ra6 As often happens in this line, White will soon have to cede control over the d6-square.

19.R6d2 383

19.R1d2 Ne8 20.R6d3 Rb6= is fine for Black. 19...Ne4 20.Nxe4 Bxe4 21.Qe3 d5 Black had no problems in Lloyd – Pilipchuk, corr. 2012. This position was reached in Kramnik – Carlsen, Bilbao 2010, and a number of subsequent games. I propose the following new idea:

13...d6!N 14.b4 This is clearly the only serious attempt to put pressure on Black. If White does not play this move, he will find it hard to justify the placement of the rook on b1. 14...cxb4 15.Qxb4 15.Rxb4?! would be a mistake in view of 15...Nbd7! 16.Be3 Qxb5 17.Rxb5 Bc6³ when White has no real compensation for his structural defects. 15...Qxb4 16.Rxb4 Bc6 17.Nd4 Bxg2 18.Kxg2 Unlike the note above, White has sufficient piece activity to make up for Black’s preferable pawn structure. 18...Rc8 Eyeing the weak pawn on c4 while covering the c6-square. 19.Rb7 19.Ndb5 e5 20.Bd2 Nbd7 21.Ra4 a6 22.Na3 Nc5 is pleasant enough for Black. 19...Nbd7 20.Ncb5 Rxc4 21.Bxd6 Bxd6 22.Nxd6 Ra4= 384

Black has no problems. C2) 11.Qb3

Compared with the 11.Qb5 line noted on page 187, here White offers the queen trade while leaving the b5-square vacant for the knight. 11...Nc6 11...Rd8 has been a more common choice but I see no reason to leave the knight undeveloped for any longer. 12.Bf4 12.Ne5N is a natural try but it gives White no advantage after: 12...Rfd8 13.Qxb6 (13.Bf4 Na5=) 13...axb6 14.Nxc6 Bxc6 15.Bxc6 dxc6 16.Rxd8† Bxd8 17.b3 Nd7= I also examined 12.Qxb6N 12...axb6 13.Ne1 Na5 14.Bxb7 Nxb7 and now an important line is:

385

15.e4 This is a natural attempt to secure a favourable pawn structure but Black has a good answer: 15...Na5! 16.e5 Ng4 17.Rxd7 Bd8 18.Nd3 Nxc4 19.f4 f6 The activity of the knights offers Black excellent counterplay.

12...Na5! Provoking the queen exchange and preparing to advance in the centre. 13.Qxb6 axb6 14.b3 d5 15.Ne5 15.Na4? Nxc4 was horrible for White in Gunina – Javakhishvili, Kutansi 2015. I also considered 15.cxd5N 15...Nxd5 16.Nxd5 Bxd5 17.Ne5 Bxg2 18.Kxg2 Rfd8 when Black is absolutely fine. 386

15...Rfd8

16.Rac1 After 16.Na4 Ra6 Black had nothing to worry about in L.B. Hansen – Stefansson, Reykjavik 1995. 16...Kf8 17.Bf3 17.e3N is safer, when 17...dxc4 18.Bxb7 Nxb7 19.Nxc4 Rxd1† 20.Rxd1 b5 21.Nxb5 Rxa2= offers Black an easy route to equality. This position was reached in Neiksans – Romanov, Norway 2015. Black has several playable moves but the following option stands out as the most promising and ambitious:

387

17...d4!N 18.Bxb7 After 18.Nb5 Bxf3 19.Nxf3 Nc6 Black has an edge due to his space advantage. 18...Nxb7 19.Na4 Bd6

20.Nc6 After 20.Nxb6 Rxa2 White is under pressure due to Black’s strong centre and active rook. 20...Bxf4 21.Nxd8 Bxc1 22.Nxb7 Bh6 Black is at least equal in the resulting endgame.

388

C3) 11.Qc2

This move indicates that White wishes to do battle in the middlegame. 11...d6 12.b3 This set-up is intended to limit Black’s counterattacking abilities on the queenside. Later, however, White might attempt to exert pressure along the a1-h8 diagonal, perhaps even advancing his kingside pawns. 12.Bf4?! does not combine well with White’s last move. He will be too slow to put pressure on the d6pawn and the bishop may prove to be misplaced, as the following game demonstrates: 12...Rd8 13.Rd2 Nc6 14.Rad1

389

14...Nd4! This knight jump is only possible due to the unfortunate location of White’s dark-squared bishop. 15.Nxd4 cxd4 16.c5 dxc5 17.Na4 Qb5 18.Bxb7 Qxb7 19.Nxc5 Qb5³ White’s opening strategy had clearly failed in Das – Kunte, Calicut 2003. 12...Nc6 13.Bb2 Nb4! The move disrupts the coordination between White’s pieces. Black does not fear losing time after a2a3, as this would render White’s queenside structure unstable.

14.Qb1 14.Qd2N may appear more natural as it keeps the rooks connected, but the drawback is that White loses control over the e4-square. Play may continue 14...Rfd8 15.e3 Ne4! 16.Qe2 Nxc3 17.Bxc3 Be4 when Black’s play seems easier, as ...a5-a4 is coming soon. 14...Rfd8 15.e3 The more aggressive 15.e4N leaves a hole on d4 and hardly poses any problems. For instance, 15...Nd7 16.Rd2 a6 17.Qf1 Qc7 18.Rad1 Nc6„ and Black has good prospects on the dark squares. 15...Bc6 16.e4 Qb7 17.Re1 a5 18.Nd2 We have been following an old game, Kan – Smyslov, Moscow 1952, with both sides playing natural and strong moves up to this point. Here Black can improve with:

390

18...Nd7!N Smyslov moved the knight to g4 instead but the text move is more accurate. The knight is not only eyeing the e5-square, but also b6, where it will support a possible ...a4 break. Another benefit is that the f6-square is now vacant for the bishop. 19.a3 White cannot put up with the knight on b4 indefinitely. 19...Na6 20.f4 20.Nd5!? is a nice idea which we should always watch out for, but Black can safely ignore it with 20...Bf8, after which 21.Bc3 Nc7 22.Ne3 Be7 23.f4 Bf6„ leads to mutual chances.

391

20...Nb6! The positional threat of ...a4 provokes White into fixing his queenside pawn structure. 21.a4 Bf6 22.Nb5 Bxb2 23.Qxb2 Nb4 24.Re3 d5„ In this complex position, Black’s chances are by no means worse. C4) 11.Bf4

This is the most popular and active continuation. 11...d6 392

It is worth making this move as Black’s position becomes more flexible, and the pressure against the d6-pawn will be manageable. The greedy 11...Qxb2?? backfires after 12.Rab1 Qxc3 13.Rxb7+– and the queen will soon be trapped. 11...Rd8 is a popular alternative but its drawbacks were illustrated in the following encounter: 12.Rd2 d6 13.Rad1 h6 14.Qb5 Ne8 15.g4! Nd7 16.Qxb6 Nxb6 17.b3² Black was doomed to a passive defence in Roiz – Leko, Saratov 2011. I believe White’s three most important options are C41) 12.Qb5, C42) 12.Rd2 and C43) 12.Rab1. Another quite popular, yet harmless, continuation is: 12.Qb3 Rd8 13.Rd2 13.Qxb6 axb6 transposes to variation C41 below. The text move prepares to double the rooks, but the d6-pawn is easily protected. Moreover, White’s queenside pawns will become more vulnerable after the rook moves away from a1, which explains Black’s next move. 13...Qxb3! 14.axb3 Nc6 The b3-pawn is fixed as a clear weakness.

15.Rad1 Ne8 Now White’s play comes to an end, whereas the weakness of the b3-pawn soon starts to tell. 16.Ne1 A later game continued 16.Bg5 f6 17.Bf4 Na5 18.Rd3 g5 19.Be3 g4 20.Nd2 Bxg2 21.Kxg2 Nc6 and White was doomed to a passive defence in Yarmolyuk – Sacerdotali, corr. 2011. 16...Rdb8 17.Nd3 Nd4 White was in trouble in Olafsson – Lautier, Antwerp 1998.

393

C41) 12.Qb5 Rd8 I see no problem allowing the queen exchange, which can also arise via the 12.Qb3 move order as noted above. With that being said, keeping the queens on with 12...Qc7!? is a good alternative which has scored well for Black.

13.Qxb6 White should not delay this move, as evidenced by 13.Rd2?! Qc7! when White lacks any constructive plan, whereas Black’s knight is heading towards d4: 14.Rad1 a6 15.Qa4 h6 16.Ne1 Nc6³ Khademalsharieh – Short, Guilan 2016. 13...axb6

394

14.Nb5 This is White’s most aggressive choice. Other moves are not too challenging, for instance: 14.Rd2 is well met by 14...Ne4! 15.Nxe4 Bxe4 16.Rad1 d5 17.cxd5 exd5 18.a3 Nc6 and Black had excellent play in Korchut – Daanen, corr. 2011. 14.b3 Nc6 15.Rd2 was tried in Fridman – Neiksans, Liepaja (rapid) 2016, when Black could have solved his space problems by means of:

15...d5!N 16.Ne5 (16.cxd5 Nxd5 17.Nxd5 exd5 18.Rad1 b5 and Black is at least equal) 16...Nxe5 17.Bxe5 Ng4 18.Bc7 Rdc8

395

19.Bf4 (19.Bxb6?? Bf6! wins material) 19...g5 20.Be3 Nxe3 21.fxe3 Bf6 Despite White’s ugly pawn structure, the position is equal due to the pressure on d5. 22.Rc1 Rd8 23.cxd5 Bxc3 24.Rxc3 exd5= 14...Ne8! Any movement of Black’s central pawns would definitely suit White, but the text move keeps things rock-solid. 15.a3 White fared no better in either of the following examples: 15.h4 h6 16.Rd2 Nc6 17.Ne1

396

17...g5! 18.hxg5 hxg5 19.Be3 Na5 20.Bxb7 Nxb7³ Black was starting to take over in Purnama – S. Zhigalko, Istanbul (ol) 2012. 15.a4 Nc6 16.b3 Na5 17.Nd2 Bxg2 18.Kxg2 occurred in Lajthajm – Bogosavljevic, Belgrade 2016. A simple and strong continuation would have been:

18...d5!N 19.cxd5 exd5 Black has no problems whatsoever.

15...h6 The immediate 15...Nc6 is also perfectly playable. 16.Rac1 Nc6 17.Ne1 Na5 Black had a comfortable position in Vallejo Pons – Anand, Linares 2005. 397

C42) 12.Rd2

Unlike the 11...Rd8 line noted earlier, White’s simple plan of doubling rooks does not work particularly well here. Even though this position has been reached in dozens of grandmaster games, the following strong reply is practically unknown. 12...Ne4! The exchange of knights helps Black to solve any problems connected with a slight lack of space. 12...Nc6!? is another fully satisfactory response which has scored well for Black, but I like my recommended move even more. 13.Nxe4 Bxe4 14.Rad1 e5 15.Bg5 f6 16.Be3

398

White has provoked Black into weakening some light squares and, given enough time, he could establish a positional advantage. Unfortunately for him, the e3-bishop is misplaced and Black has excellent prospects for counterplay. 16...Nc6 17.Ne1 White swaps bishops in the hope of exposing the weak light squares in Black’s camp, but White’s lack of harmony prevents him from benefiting from it. 17...Bxg2 18.Nxg2 f5 19.f3 Qc7 20.Bf2 Up to now, Black had done everything right in Wessman – Bergstrom, Stockholm 1987. Here it would have been best to prevent White’s knight from moving towards d5.

399

20...f4!N 21.Qc2 Rf6 22.Qe4 Raf8 White finds himself in a passive position with no real counterplay. C43) 12.Rab1

This can be regarded as the main line. Since Black’s last move blocked the bishop on e7, it is quite logical for White to prepare the b2-b4 break. 12...Nbd7 13.b4 This move seems most consistent, and it is no surprise that it has been played in almost all games from this position. An interesting alternative would be 13.e4!?N, utilizing the loose knight on d7 to gain space in the centre. My suggestion is 13...Rad8 14.Qc2 and now it is important for Black to regroup in the following way:

400

14...Nb8! The knight is heading towards d4, as it often does in these positions, especially after e2-e4 has been played. 15.e5 dxe5 16.Nxe5 Bxg2 17.Kxg2

17...Nfd7! Swapping off White’s active knight. (17...Nc6 18.Qb3 Nd4 19.Qxb6 axb6 20.Nb5!² puts Black under some pressure.) 18.Nxd7 Nxd7 19.h4 Nb8= Black has nothing to worry about.

401

13...Rfb8! An important consolidating move. Now almost all of Black’s pieces are located on the queenside, where the action is about to start. Interestingly, this move was not played until 2011, but since then it has been tested in more than thirty games. Previously 13...cxb4 was thought to be a good reply, based on the small tactical finesse of 14.Rxb4 Bc6! when Black is fine. However, it was later discovered that 14.Qxb4! is a significant improvement, when Black has some problems along the b-file. We have reached an important position, where C431) 14.Be3!? and C432) 14.a3 should be investigated. I would evaluate the two moves as being of roughly equal value, although the latter has been a lot more popular. C431) 14.Be3!? Bc6! This provocative move is the only effective way to stabilize the situation on the queenside.

402

15.Qc2 15.bxc5N 15...Bxa4 16.cxb6 axb6= leads nowhere special for White. I also checked 15.b5N 15...Bb7 16.Qc2 a6 (16...Qc7 transposes to the main line but there is no need to move the queen against this particular move order) 17.a4 axb5 18.axb5 Qc7 when Black has a solid position with good prospects along the a-file. 15...Qc7 16.b5 This ambitious advance is White’s only challenging idea. After the harmless 16.bxc5 Rxb1 17.Rxb1 Nxc5 18.Bxc5 dxc5= a draw was soon agreed in Lukyanenko – Warzecha, corr. 2016. 16...Bb7 17.a4

403

White’s pawn majority on the queenside appears promising but it is actually hard to make further progress with it, whereas Black can prepare an eventual ...d5 break. 17...h6 18.Bd2 I also considered 18.Bf4N 18...Nb6 19.Qb3 a5 20.bxa6 Bxa6 21.Nd2 Ra7 22.h3 Nfd7 23.Nb5 Bxb5 24.axb5 Ra4„ when Black has excellent prospects along the open a-file. 18...Nb6 This position has been reached in two correspondence games. In each of them White protected the c4pawn with his queen, but the choice of square actually makes a big difference to the way Black should react. 19.Qa2 The earlier game continued: 19.Qd3

404

19...a5! The queenside is safely blocked and the protected passed pawn on b5 will only become dangerous after many piece exchanges. At present, Black benefits from a central pawn majority and healthy prospects in the middlegame. 20.e4 White has to be active in the centre. 20...Nfd7 21.Bf4 Rd8 22.Qe2 22.Bxd6?! Qxd6 23.Qxd6 Bxd6 24.Rxd6 Nxc4 25.Rdd1 Ndb6 is more comfortable for Black. We have been following Torrijos Alhambra – Figueredo Company, corr. 2012. My suggested improvement is:

22...Nf8!N 405

Intending to regroup with ...Ng6. 23.e5 This is the most challenging idea; White is trying to break through to the weak c6-spot. The quieter 23.Rbc1 Ng6 24.Be3 Bf6„ leads to a long strategic struggle where Black’s chances are not worse. 23...Ng6 This leads to a series of mostly forcing piece exchanges.

24.exd6 Nxf4 25.gxf4 25.dxc7 Nxe2† 26.Nxe2 Rdc8 27.Ne5 Bxg2 28.Kxg2 Rxc7 also leads to a double-edged endgame. 25...Bxd6 26.Ne5 Bxe5 27.Qxe5 Qxe5 28.fxe5 Bxg2 29.Kxg2 Nxc4

30.Ne4 Nb6 31.Nxc5 Rac8„ 406

The weakness of the a4- and e5-pawns offers Black sufficient counterplay. The text move was played in a later game which was eventually won by White. However, at this point Black could have highlighted the drawback of the queen’s location by means of:

19...a6!N The game continued 19...a5 20.Re1 Rd8 21.e4 Nfd7 22.Nd1! Bf6 23.Nb2 Nf8 24.Nd3 when White had a nice position and eventually prevailed in Garau – Tassone, corr. 2015. The knight manoeuvre to d3 shows why it was useful not to have the queen on that square in the closed structure after ...a5. 20.a5 Presumably this move is the reason why Black avoided this line in the aforementioned game, but I like the look of the following positional sacrifice. 20...axb5! 21.axb6 21.Nxb5 Qd8 22.Be1 Ra6 23.Ne5 Bxg2 24.Kxg2 Nbd7 25.Nxd7 Qxd7 also leads to double-edged play. 21...Qd8 22.Qb2 b4

407

23.Nb5 Qxb6 24.Bf4 White can change the character of play with 24.Bxh6!? gxh6 25.Nxd6, but 25...Bxf3 26.Bxf3 Ra7 27.Nb5 Rd7 is fine for Black. 24...Rd8 We have reached an unusual situation where White’s material advantage is of little consequence, as his pieces are heavily restricted by Black’s strong pawns. 25.Ra1 After 25.Ne1 Bxg2 26.Nxg2 Rd7 27.Qc2 Qc6 White’s position looks rather passive. Black might start pushing the pawns at any moment. 25...Rxa1 26.Qxa1 26.Rxa1? e5 27.Bd2 e4µ is poor for White. 26...e5 27.Bc1 e4 28.Nh4 d5 29.Nf5 Bf8© Black has excellent play. C432) 14.a3 This leads to a different character of game from the previous line, with more fluid play on the queenside.

408

14...a6 This is an important prophylactic move, securing the c7-square for the queen. 15.Qa5!? This has been White’s main try at correspondence level, so it’s only a matter of time before somebody tries it over the board. (In fact, the same idea has already been tested in the near-identical position with pawns on h3 and h6.) I checked two other moves: After 15.Qb3 Qc7 16.b5 Ne4! 17.a4 Nxc3 18.Qxc3 axb5 19.axb5 Nb6 Black had the more pleasant position due to his control over the a-file in Moiseenko – Leko, Saratov 2011. 15.Qc2 cxb4!? I prefer this rare move over 15...Qc7, which has featured in a couple of high-level games. 16.axb4 Qc7 As we will see, Black’s slightly weak pawns on a6 and d6 are balanced out by the enemy targets on b4 and c4.

409

17.Qb3 In the event of 17.c5N 17...e5 18.Be3 dxc5 19.bxc5 Black can safely capture on c5 with either his knight or his bishop; in both cases, White should be able to regain the pawn with equality, but nothing more. 17.Qa2N 17...Nh5 18.Be3 Rc8 19.Rdc1 Nhf6= is also fine for Black. 17...h6 18.Rbc1 This was White’s choice in the correspondence game. Once again, 18.c5N 18...e5 19.cxd6 Bxd6= leads nowhere for White. After 18.Rdc1N 18...Bf8 the vulnerability of White’s queenside pawns offers Black good chances. A possible continuation is 19.c5 e5 20.cxd6 Qxd6 21.Be3 Bc6 when the position is essentially equal, but White has to be slightly more careful due to the weakness of the b-pawn.

410

18...g5!? With White’s rooks lined up in this particular fashion, 18...Bf8? would be a big mistake due to 19.c5!. On the other hand, 18...Nh5N 19.Be3 Nhf6= is a sensible alternative to drive the bishop away. 19.Be3 Rc8 20.Na4 Rab8 21.b5 Qa5 22.Ra1 Qc7 23.Rac1 ½–½ Dullemond – Kuhne, corr. 2014. Returning to the main line, we see an interesting concept: White is attempting to provoke a weakening of his own pawn structure in order to make inroads on the queenside and exploit the power of the g2-bishop.

15...Ne4! Black ignores the provocation and invites favourable simplifications. The timid 15...Qxa5?! 16.bxa5 Bxf3 (16...Ne8 17.Ne5! also favours White) 17.Bxf3 Rxb1 18.Rxb1 Rb8 19.Rb7 leads to an unpleasant endgame, where the weak a6-pawn would be a constant worry for Black. 16.Qxb6 Nxb6 17.Nxe4 Bxe4 18.Rbc1 18.bxc5N is tactically viable, but hardly an improvement over the existing games. The forced line continues: 18...Bxb1 19.cxd6 Bxd6 20.Bxd6 Nxc4

411

21.Ne5 Nxe5 22.Bxe5 Ba2 23.Bxb8 Rxb8= Reaching a completely equal endgame. 18...cxb4 18...Rd8 is also reasonable and has led to draws in all five of the correspondence games in which it was played, but the text move seems more straightforward. 19.axb4 Rc8 20.Bxd6 Bxd6 21.Rxd6 Nxc4

22.Rd4 Bxf3 23.Bxf3 Ne5 24.Rxc8† Rxc8 25.Bb7 Rb8 26.Bxa6 Nc6 Black easily drew in Cattani – Rost, corr. 2012, and a subsequent correspondence game.

412

Conclusion 5.Qa4 may not be the most theoretically critical move but it remains quite a popular choice. After the standard response of 5...Bb7 6.Bg2 c5 we looked at three main options. 7.d5!? can be compared with the much more popular pawn sacrifice with the queen on c2, as discussed in Chapters 12 and 13. This version is less dangerous overall but, after 7...exd5 8.cxd5, it’s important for Black to take on d5 with the bishop rather than the knight, and to be willing to return the extra pawn to reach a safe position. Next we looked at 7.0-0 cxd4 8.Nxd4 Bxg2 9.Kxg2, which is playable but not dangerous at all. The exchange of light-squared bishops gives Black a pleasant version of a Hedgehog, with chances to fight for the initiative if he can successfully arrange ...b5. 7.dxc5 bxc5 is the main line of the chapter, when the asymmetrical pawn structure makes for a strategically complex middlegame. Play generally continues 8.0-0 Be7 9.Nc3 0-0 10.Rd1 Qb6, after which many moves are possible, but 11.Bf4 d6 12.Rab1 is most critical. Black’s most accurate response is 12...Nbd7 13.b4 Rfb8! when White has a couple of options. The ensuing positions are challenging for both sides and a close study of the relevant correspondence games and my accompanying analysis should prove invaluable in understanding the complexities.

413

A) 6.Bg2 204 B) 6.Bf4 Be7! 7.Nc3 d5! 8.cxd5 Nxd5 205 B1) 9.Nxd5N 206 B2) 9.Bg2 207 C) 6.Bg5 Be7 7.Nc3 0-0N 8.e4 c5! 209 C1) 9.Rd1 210 C2) 9.d5 212 D) 6.Nc3 Be7! 7.e4 d5 213 D1) 8.exd5 214 D2) 8.cxd5 Bxf1 9.Kxf1 cxd5 10.e5 Ne4!N 11.Kg2 Nc6 0-0 13.h4 Rc8 215 D21) 14.Ne1!? 217 D22) 14.Be3 218

414

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Ba6 5.Qb3 This move was first seen in 1981, but mostly gained popularity in the early 2000s, since when it has been employed by many leading players. Unlike the other methods of defending the c4-pawn, White keeps a watchful eye on the d5-square. On the flip side, Black can look to develop counterplay using the slightly exposed position of the white queen. 5...c6!? The point of this slightly unusual move is to play ...d5 and recapture with the c-pawn if White exchanges. I find it a logical plan, as the c4-pawn can no longer be supported by a pawn. 5...Nc6 is most popular move by far, and it is perfectly playable. The only slight drawback of this move is that certain lines have been played and analysed all the way through to a drawn endgame, which is theoretically acceptable but not necessarily an ideal scenario for a practical player. The immediate 5...d5 is also possible, when 6.cxd5 Qxd5 leads to a different type of struggle. Against our chosen move, White has four sensible options: A) 6.Bg2, B) 6.Bf4, C) 6.Bg5 and D) 6.Nc3. A) 6.Bg2 Out of White’s four options, this is the only one which is completely harmless, as it makes no attempt to stop Black from carrying out his plan. 6...d5 7.cxd5 7.Nbd2 Be7 8.0-0 Nbd7 reaches a version of the Closed Catalan where White’s queen is clearly misplaced; it would be happier on c2. 9.Re1 (9.Qa4N may be a better try, though after 9...Bxc4! 10.Nxc4 b5 11.Qc2 bxc4 12.b3 cxb3 13.Qxc6 bxa2 White still has to justify the pawn sacrifice) 9...0-0 10.e4

415

10...Nxe4! 11.Nxe4 Bxc4 12.Qc2 dxe4 13.Qxc4 exf3 14.Bxf3 Rc8³ White had less than adequate compensation for the pawn in Kadimova – Stefanova, Kusadasi 2006. 7...cxd5

8.Nc3 Be7 9.Bf4 9.Ne5 0-0 10.0-0 Bb7 11.Rd1 Nc6 12.Bf4 Na5 13.Qc2 Rc8= was also fine for Black in Anand – Sasikiran, Bastia 2011. 9...0-0 10.Rc1 Here I found a promising new idea. 416

10...Nc6!N Offering a pawn sacrifice. The modest 10...Bb7 11.0-0 Nc6 led to equality in Ivanchuk – Leko, Monte Carlo (rapid) 2003. 11.Nxd5 This is the critical try, although Black gets superb compensation. After 11.0-0 Na5 12.Qa4 Qe8 13.Qxe8 Rfxe8³ we reach a queenless middlegame with a symmetrical pawn structure, where Black’s pieces are placed better. 11...Qxd5 12.Qxd5 Nxd5 13.Rxc6

417

13...Bb5 14.Rc1 Rac8ƒ White has serious problems due to his lack of harmony and lagging development. B) 6.Bf4

This move carries some venom and should be handled carefully. 6...Be7! 6...d5 seems inferior, though not for the reason you might think! 7.cxd5 cxd5

418

8.e3! (Surprisingly, 8.Bxb8N yields no advantage: 8...Bc4 9.Qa4† b5 10.Qd1 Rxb8 11.b3?! [11.Nbd2 is preferable, though Black is still not worse] 11...Bb4† 12.Nbd2 Ne4ƒ Black has an excellent position.) 8...Bxf1 9.Kxf1² Riazantsev – Ionov, Krasnoyarsk 2003. 7.Nc3 d5! Now is the right time for this move. 7...0-0 has been more popular but I believe it is less precise in view of: 8.e4 d5 9.cxd5 cxd5 10.exd5 Nxd5 11.Nxd5 Qxd5 12.Qxd5 exd5 13.Bxb8 Bxf1 14.Rxf1 Raxb8 15.Kd2² White had a risk-free endgame edge in Karpov – Leko, Cap d’Agde (rapid) 2003.

8.cxd5 Nxd5 419

8...cxd5? does not work here in view of 9.Bxb8 Bc4 (9...Rxb8?? loses a whole piece after 10.Qa4†) 10.Qa4† b5 and now, unlike the similar line in the note to Black’s 6th move, 11.Nxb5 spoils Black’s plans. Recapturing with the knight not only avoids the tactical problem mentioned above, but also targets the bishop on f4. White’s most natural replies are B1) 9.Nxd5N and B2) 9.Bg2. White can hardly hope to benefit from delaying the development of his kingside pieces: 9.Rc1 Nxf4 10.gxf4 0-0 11.e3 (11.Bg2N transposes to variation B2 below) 11...Bxf1 12.Kxf1 This was Lerner – Yegiazarian, Moscow 2005, and now a natural improvement is:

12...Nd7N 13.Kg2 Qc8 14.Rhd1 Qb7 15.Qc2 Nf6= Black is ready for ...c5, either immediately or after developing one or both rooks. B1) 9.Nxd5N 9...cxd5!

420

As we will see, this positionally desirable recapture is tactically justified. 10.Bxb8 10.e3 is well met by 10...Qc8! 11.Bb5† Bxb5 12.Qxb5† Qc6 13.a4 a6 14.Qb3 0-0 15.0-0 Qb7 16.Rfc1 Nc6 17.Rc3 Rfc8 18.Rac1 Bd8= followed by exchanges on the c-file. 10...Bc4 11.Qa4† b5 Unlike the 8...cxd5? line mentioned in the earlier note, White is unable to capture the b-pawn since the c3-knight has been exchanged. However, we still need to see what happens if White traps the bishop. 12.Qc2 Rxb8 13.b3 Qa5† 14.Nd2 0-0 15.bxc4 bxc4 Black’s initiative full compensates for the missing piece, as the following lines demonstrate.

421

16.Bh3 I also considered 16.Bg2 c3 17.Nb3 Rxb3! 18.Qxb3 c2† 19.Kf1 Rc8 when Black will regain the material and be left with no problems. 16...Ba3 17.0-0 Rb2 18.Nxc4 dxc4 19.Qxc4 Rb4

20.Qc5 20.Qd3?! Bb2³ gives Black a slight initiative. 20...Qxc5 21.dxc5 Rc4= Black regains the pawn to reach a drawish endgame. 422

B2) 9.Bg2

I will take this natural developing move as the main line. 9...0-0 10.Rc1 Nxf4 11.gxf4 Qd6! This precise move enables Black to quickly connect the rooks. 12.e3 Nd7

13.Ne5 I also examined the logical alternative: 423

13.Qa4N White wants to castle, so he chases the bishop away. 13...Bb7 14.0-0 Qc7!? The main purpose of this prophylactic move is to guard the bishop on b7. I see no way for White to exploit the alignment of rook and queen on the c-file. I also analysed 14...Rfd8 15.Ne5 Nxe5 16.dxe5 Qc7 17.Nb5 Qc8 18.Rfd1 cxb5 19.Rxc8 bxa4 20.Rdxd8† Bxd8 21.Bxb7 Rxc8 22.Bxc8 when White will win a pawn, but the opposite-colouredbishop endgame is an easy draw.

15.Rfd1 15.Ne5 Nxe5 16.fxe5 c5„ is similar. 15...Rfd8 16.Ne5 16.a3 Rac8 17.b4 Qb8 is also about equal. 16...Nxe5 17.fxe5 c5

424

This standard break in the centre offers Black decent counterplay. For instance: 18.d5 Qxe5 19.dxe6 Qg5 20.exf7† Kxf7 21.Ne4 Qg6„ 13...Rac8 14.Qa4 We have been following Jirovsky – Obsivac, Czech Republic 2014. My new idea is:

14...Nb8!N Keeping the bishop in place, in order to stop White from castling. 15.a3 15.Rg1 f6 16.Nf3 Kh8 17.Bf1 Bxf1 18.Kxf1 Nd7„ leads to a complex strategic struggle with mutual 425

chances. 15...f6 16.Nc4 16.Nf3 c5„ also leads to double-edged play.

16...Qd8 17.Qb3 Kh8 18.0-0!? This seems the most ambitious and critical move, inviting the following transformation. 18...b5 19.Nd2 b4 20.axb4 Bxf1 21.Nxf1 White’s positional compensation for the exchange looks undeniable, but Black can activate his pieces by means of:

426

21...e5! 22.fxe5 fxe5 23.dxe5 Na6 Black has plenty of counterplay and his chances are not worse. C) 6.Bg5

Just like the previous option, this bishop development is connected with a concrete idea against an early ...d5. 6...Be7 Once again, we should develop this bishop before committing ourselves in the centre. 6...d5 has achieved a plus score but I prefer White’s chances after: 7.cxd5 cxd5 8.Bxf6! Qxf6 (8...gxf6 is safer but White is slightly better) 9.Nc3 Qd8 10.Ne5 Bd6 11.Qa4† Kf8 12.e4 White had a powerful initiative in Kulaots – Alekseev, Moscow 2004. 7.Nc3 7.Nbd2 seems too passive: 7...d5 8.Bg2 (after 8.cxd5 cxd5 9.e3 Qc8 10.Qa4† Qd7 11.Qxd7† Nfxd7 Black effortlessly equalized in Obukhov – Ionov, Krasnoyarsk 2003) 8...0-0 9.0-0 Nbd7 10.Rfe1 Bb7 11.Rac1 c5 Black had a comfortable position in Miton – David, Haguenau 2013. The ambitious 7.Bxf6 Bxf6 8.e4 may backfire, as Black’s dark-squared bishop could become powerful. 8...d6 9.Nc3 0-0 10.Rd1 This was seen in Chuchelov – Babula, Germany 2005. Black’s play can be improved with:

427

10...Nd7N 11.Bg2 d5! 12.cxd5 cxd5 13.e5 (Black also has no problems after 13.exd5 Re8! 14.Nb5 exd5† 15.Kf1 Nf8=) 13...Be7 14.Bf1 Qc8 15.Qa4 Bxf1 16.Kxf1 Qb7 17.Kg2 Rfc8 Black has excellent prospects on the queenside, whereas the lack of a light-squared bishop significantly reduces White’s attacking abilities. This position has been tested a few times at a high level. However, in my opinion the best continuation has not yet been played.

7...0-0N This is the most flexible move. On the flip side, it allows White to establish a strong pawn centre, but we have an effective way of countering it. 7...d5 has been more common but I am not satisfied with Black’s prospects after 8.cxd5 cxd5 9.e3!N. For 428

instance: 9...Bxf1 10.Kxf1 Nc6 11.Rc1 Rc8 12.Kg2 0-0 13.Qa4 Qd7 14.Ne5 Qb7 15.Bxf6 gxf6 16.Nd3 Black remains under pressure along the c-file. 8.e4 c5! 8...d5 fails to equalize after 9.Bxf6 Bxf6 10.cxd5 exd5 11.Bxa6 Nxa6 12.exd5² when Black does not have much compensation for the pawn. The text move is much more interesting. Usually in this pawn structure, d4-d5 would give White an easy space advantage, but here it would involve a pawn sacrifice due to unfortunate location of the bishop on g5.

I analysed two main options, which seem to be of roughly equal value: C1) 9.Rd1 and C2) 9.d5. 9.dxc5 bxc5 10.e5 proves to be harmless after: 10...Ne8 11.Bxe7 Qxe7 12.Bg2 Nc6 13.0-0 Rb8 14.Qa4 Bb7

429

15.Rad1 d6 16.exd6 Nxd6= Black’s knights are heading to d4. C1) 9.Rd1 Bb7 10.d5 This pawn sacrifice is the only challenging continuation. 10...Nxd5 11.cxd5 Bxg5 12.Bc4!? White simply develops and plays for a positional bind. I analysed two alternatives, each with its own merits and drawbacks. 12.h4 Be7 13.Bc4 enables Black to simplify matters with: 13...exd5 14.Bxd5 Bxd5 15.Qxd5 Nc6 16.e5 Re8 17.0-0 h6 18.Rfe1 Bf8

430

White has nothing better than regaining the pawn with 19.Qxd7, and after 19...Qxd7 20.Rxd7 Rad8 Black has no problems. 12.Bg2 This is also not too threatening after: 12...exd5 13.Nxd5 Nc6 14.e5

14...d6! 15.exd6!? This leads to double-edged play. 15.0-0 is safer, and after 15...dxe5 16.Nc3 Qe7 17.Nd5 Qd8= White has nothing better than repeating moves. 15...Re8† 16.Kf1 Bf6 17.Nc7 Rf8 431

18.h4 The greedy 18.Nxa8? Qxa8 19.Ne1 Rd8µ invites serious trouble, as White’s position desperately lacks harmony. 18...Rc8 19.Qa4 Rxc7 20.dxc7 Qxc7 Black has full compensation for the exchange.

12...e5! I prefer this over 12...exd5 13.Bxd5 Nc6 (unlike the analogous position in the 12.h4 line above, 13...Bxd5?? is not an option here as 14.Qxd5+– comes with a lethal fork) 14.e5 Be7 15.0-0 Qc8 16.Rfe1 when White has at least enough compensation for the pawn. 13.Nxg5 432

After 13.Nxe5 d6 14.Nf3 Bf6 15.0-0 Nd7 the arising version of the Benoni is comfortable for Black, whose dark-squared bishop is extremely powerful. 13...Qxg5 14.Nb5 Qd8 15.f4 d6 16.fxe5 dxe5 17.d6 White has to play energetically in order to justify the pawn sacrifice, but now the b7-bishop is unblocked.

17...Nd7 18.Nc7 Rb8 19.0-0 Nf6 In the arising complex position Black’s chances are not worse. White’s pieces are active but his king is slightly exposed, and he is still a pawn down. C2) 9.d5

433

This is the most straightforward reply to Black’s last move. 9...Nxd5 10.cxd5 Bxg5 11.Bxa6 Nxa6 Despite the closed character of the position, White has decent compensation due to his space advantage and Black’s poor coordination. 12.Rd1 In the event of 12.0-0 Nc7 13.Rad1 exd5 14.Nxd5 Be7 Black should be able to return the extra pawn to simplify the position and activate his pieces. For instance:

15.Rfe1 Nxd5 16.Qxd5 Qc7 17.Qxd7 Rac8 18.e5 Rfd8=

434

12...exd5 12...Nc7!? also deserves attention, for instance: 13.dxe6 (13.0-0 transposes to the previous note) 13...Nxe6 14.0-0 Be7 15.Qa4 d6 16.e5 Qe8 17.Qxe8 Rfxe8 18.exd6 Bf6 Black should be able to neutralize the passer. 13.Rxd5 13.Nxd5 is less challenging in view of 13...Re8 14.0-0 Bf6 15.Rfe1 d6 when White is running out of constructive ideas. 13...Be7 14.Qa4 After 14.0-0 Nb4 15.Rd2 Qc8 the activity of White’s pieces compensates for the missing pawn, but he cannot claim any advantage. Play might continue:

435

16.a3 c4 17.Qa4 Nd3 18.b3 (18.Nd5 Re8 19.b3 Bf8 20.bxc4 Nc5=) 18...Qc6 19.Qxc4 Qxc4 20.bxc4 Nc5=

14...Nb4 14...Nb8?! seems too passive, and 15.0-0 a6 16.e5 offers White good attacking chances with the queen poised to swing across the 4th rank. 15.Rxd7 Qe8 16.0-0 16.a3 is well met by: 16...Bf6! 17.0-0 Bxc3 18.axb4 Qxe4 19.Ng5 Qg4 20.bxc3 Qxg5= 16...c4! 436

As usual, the knight is heading to d3. An interesting role reversal is taking place, as now it is Black who offers a pawn sacrifice to improve the activity of his pieces. 17.Rxa7 17.Ne5 Bf6 18.Qxb4 Bxe5 19.Rfd1 Bxc3 20.Qxc3 Qxe4 gives Black easy equality. 17...Rxa7 18.Qxa7 Bc5 19.Qa4 Qe6 Black’s active piece play fully compensates for the missing pawn. D) 6.Nc3

Even though this is only White’s third choice in terms of popularity, it is arguably the most challenging option of all. 6...Be7! I was surprised to discover that this natural move has been tested only twice. 6...d5 has been more common but it may invite some troubles: 7.cxd5 cxd5 8.Bf4 Bc4 (8...Nbd7 9.Qa4 Bb7 10.Nb5 Bc6 11.Qb3 Bxb5 12.Qxb5² gave White a safe edge with the two bishops in Fressinet – Bauer, Villandry 2012) 9.Qc2 Nc6 This was Thejkumar – B. Kumaran, Dharamshala 2014, and now White could have extended his initiative by means of:

437

10.e4!N 10...Rc8 (10...Bxf1? 11.Nxd5! Rc8 12.Nxf6† gxf6 13.Kxf1+–) 11.Bxc4 dxc4 12.0-0ƒ White has a dangerous lead in development. 7.e4 7.Bf4 and 7.Bg5 transpose to variations B and C respectively. 7.d5 was not too threatening in Rajkovic – Chelushkina, Belgrade 2008. Simplest is 7...cxd5N 8.cxd5 and now Black has a couple of good options:

8...Nxd5!? (8...Bb7 leads to easy equality after something like 9.dxe6 dxe6 10.Bg2 0-0 11.0-0 Nbd7=) 9.Nxd5 exd5 10.Qxd5 Nc6„ Black has plenty of activity to make up for the isolated d-pawn.

438

7...d5 We will consider the simplifying D1) 8.exd5 followed by the more combative D2) 8.cxd5, with e4-e5 to follow. 8.e5 is a slightly inaccurate move order. After 8...Ne4 White has nothing better than 9.cxd5 Bxf1 10.Kxf1 when 10...cxd5 transposes to variation D2, but Black has the extra option of 10...Nxc3 11.Qxc3 Qxd5, reaching a comfortable position. D1) 8.exd5 cxd5

9.Qa4†!? 439

9.cxd5 Bxf1 10.Kxf1 Nxd5 11.Nxd5 Qxd5 12.Qxd5 exd5 is equal, with the knight coming to c6 next. The text move is a small finesse: White ‘loses’ a tempo in order to lure the knight to d7, where it will be slightly less active. He still gets no advantage though. 9...Qd7 10.Qxd7† Nbxd7 11.cxd5 Bxf1 12.Rxf1 Nxd5 13.Nxd5 exd5

14.Ke2 Nf8 15.Bd2 Ng6 16.Rac1 Bd6= White can hardly claim an advantage in the endgame with a symmetrical pawn structure. D2) 8.cxd5

440

This move, in conjunction with e4-e5, is the more ambitious continuation. 8...Bxf1 9.Kxf1 cxd5 10.e5 10.exd5 transposes to the note to White’s 9th move in variation D1 above. This important position was reached in Barlov – Pogorelov, Las Palmas 1994, when Black retreated the knight to d7. A more active and stronger move is:

10...Ne4!N As we will see, Black has enough dynamic resources to justify placing the knight on a seemingly unstable square. 11.Kg2 White will surely need to play this move at some point, so I will present it as my main line. 11.h4 Nc6 12.Kg2 transposes to 12.h4 in the notes to the main line below. 11.Qc2 sees White try to provoke an exchange of the active knight, but 11...Qc7 is a good answer, and after something like 12.Bd2 0-0 13.Kg2 Rc8 14.Rhc1 Qb7 Black is fine. The immediate attempt to trap the knight by means of 11.Ne2 Nc6 12.Ne1 seems ineffective after:

441

12...f6! 13.f3 Ng5 14.exf6 (14.Bxg5 fxg5 15.Kg2 0-0 16.Rc1 Qd7 is fine for Black) 14...Bxf6 15.Bxg5 Bxg5 16.f4 Bf6 17.Nf3 0-0= The pawns on d4 and e6 are equally weak. 11...Nc6

Several move orders are possible from here. I think White’s most challenging plan is to drop the c3knight back to e2 at some point, in order to set up the possibility of Ne1 followed by f2-f3 to harass the e4-knight. However, to make this plan effective, White will need to throw in h2-h4 as well, otherwise Ne1 will simply be met by ...f6 as in the note above. All this takes time, so White may be well advised to develop his bishop and stick a rook on the c-file somewhere along the way. Black, on the other hand, has a fairly simple plan of castling and ...Rc8, after which he will choose a plan to counter whatever White may be doing. The ...f6 break is likely to play a key role, although in certain situations Black may keep 442

the kingside closed and play on the opposite flank. 12.Ne2 White could also start with: 12.h4 0-0 13.Be3 13.Ne2 transposes to the main line below. I also checked 13.Bf4 Qd7 14.Rac1 Rac8, and now I doubt that White has anything better than 15.Ne2, with a transposition to the note on 14.Bf4 in our main line.

13...Rc8 14.Qd1 Na5 White is not close to trapping the knight here, so Black has no real need for ...f6. 15.Nxe4 dxe4 16.Ng5 Qd5 We have reached a complex position where Black has plenty of queenside counterplay to distract White from his kingside ambitions. 12...0-0 13.h4 This move is necessary if White wants to make trapping the knight a realistic threat. 13.Ne1 hardly poses Black any problems after 13...f6. For instance: 14.f3 Ng5 15.exf6 Bxf6 16.Bxg5 Bxg5 17.f4 Bf6 18.Nf3 Ne7=

443

13...Rc8 The stage is set for a complex middlegame. I analysed the ambitious and rather risky D21) 14.Ne1!?, trying to trap the knight, followed by the more steady plan of development with D22) 14.Be3. 14.Bf4 Qd7 15.Rac1 f6! sees Black challenge the opponent’s central pawns, in the spirit of the French Defence. 16.exf6 Bxf6

17.Qe3 Rce8 18.Rhe1 h6= Black has a pretty comfortable version of a French Tarrasch structure, due to the removal of the light-squared bishops. Black has decent control over the centre, and the vulnerability of the light squares around White’s king might become an issue in the long run.

444

D21) 14.Ne1!?

White postpones the development of his queenside pieces in order to trap the knight. We will respond by sacrificing it and blasting open the centre. 14...f6 15.f3 Nxg3! 16.Kxg3 fxe5 17.dxe5 Nxe5© Black has two pawns for the piece, while White’s pieces are uncoordinated and the king is weak. Here are a few illustrative lines: 18.f4 I also considered18.h5 Bg5 19.Be3 Nc4 20.f4

20...Nxe3 (20...Bf6!? is also good enough) 21.Qxe3 Bxf4† 22.Nxf4 Qg5† 23.Kh3 Qxf4 24.Qxf4 Rxf4 445

With three pawns against a knight, Black’s chances in the endgame are by no means worse. The text move is ambitious and critical, but it exposes the white king even more. Black can effectively exploit this by means of:

18...Rc4! 19.Nf3 19.fxe5? Bxh4† 20.Kh3 Bxe1 gives Black a crushing attack. 19...Nxf3 20.Qxf3 e5 Black has a dangerous initiative on the kingside. D22) 14.Be3

446

This seems like a safer/saner choice than the previous variation. White develops another piece and supports the central pawns, while keeping the Ne1 plan as an option for later. 14...Na5 15.Qd3 Nc4 16.Bc1 This retreat looks passive, but White is intending to chase the knights away soon. 16.Rab1 Qd7 17.Bg5 f6 18.exf6 Bxf6 offers Black decent counterplay along the f-file. 16...f6 16...Qd7!? also deserves attention, intending to meet 17.Ng5 with 17...f6! 18.Nxe4 dxe4 19.Qxe4 fxe5 20.dxe5 Qb5© with sufficient activity for the pawn. 17.Nf4 17.b3? runs into 17...fxe5 18.bxc4 Nxf2! 19.Kxf2 e4 and White’s position is ruined. 17...Qd7 18.Re1 Rc6! 18...fxe5?! would be somewhat premature, as after 19.dxe5 the e4-knight is hanging.

447

19.b3 19.exf6 gxf6 20.Nh2 is hardly better in view of 20...Rfc8 21.f3 e5!„ with good play in the centre. 19...Na3 20.exf6 gxf6 21.Bxa3 Bxa3

22.Rxe4! The knight was so strong that White could not have put up with it for much longer. The exchange sac is a decent try but it offers no advantage. 22...dxe4 23.Qxe4 Bc1!? The engine offers other some moves leading to equality, but the text seems simplest. Black restricts the 448

enemy rook while controlling some kingside squares with the bishop. 24.Nd3 Rfc8 White has nothing better than forcing a draw, the following line being the most straightforward way of doing so.

25.Rxc1 Rxc1 26.Nxc1 Rxc1 27.Qg4† Kf7 28.Qh5† Kg7 29.Qg4†= Conclusion 5.Qb3 is a former sideline which has become relatively mainstream over the past decade or two. I find 5...c6!? to be an interesting response which has been unfairly neglected up to now. Then 6.Bg2 d5 is absolutely harmless for Black, but White’s other three moves all deserve close attention. 6.Bf4 should be met by the precise 6...Be7!, waiting for 7.Nc3 before striking in the centre with 7...d5!. In the event of 8.cxd5, it is important to recapture with the knight, leading to a rich position with mutual chances. 6.Bg5 presents a different set of challenges. Once again Black should start with 6...Be7, but he should meet 7.Nc3 with 7...0-0N 8.e4 c5!, the point being that d4-d5 will be met by the thematic ...Nxd5 trick, exploiting the loose bishop on g5 to win a pawn. The story does not end there as White gets reasonable compensation, but Black should be able to maintain the balance. Finally we analysed 6.Nc3, when 6...Be7! is once again the way to go. After 7.e4 d5, the simplifications resulting from 8.exd5 cxd5 offer White no advantage. The more critical option is 8.cxd5 Bxf1 9.Kxf1 cxd5 10.e5, when 10...Ne4!N is an important innovation. My analysis indicates that the open c-file and the possibility of a timely ...f6 offer Black enough counterplay, and it would be interesting to see these 449

ideas tested in practice.

450

A) 6.Bg2 Nc6 221 A1) 7.0-0 222 A2) 7.dxc5 222 B) 6.d5 exd5 7.cxd5 Bb7 224 B1) 8.e4 Qe7 225 B11) 9.Bd3 225 B12) 9.Bg2 227 B2) 8.Bg2 Nxd5 229 B21) 9.Qb3 230 B22) 9.0-0 233

451

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Ba6 5.Qc2 This move dates back to 1930 but its merits were not fully understood until the mid-2000s, at which point it became extremely popular at high levels. White is not only protecting the c4-pawn but also taking control over e4, supporting a possible e2-e4 advance. The drawback is that White loses control over the d5-square, which explains Black’s next move. 5...c5 White has two main ways to meet the challenge to his central pawn. A) 6.Bg2 playable but not particularly threatening, while B) 6.d5 is the more dangerous move which has become the main line. 6.dxc5 Bxc5 7.Bg2 Nc6 transposes to variation A2 below. 6.e4?! This central advance is premature. 6...cxd4 7.e5 After 7.Nxd4 Bc5 8.Nb3 Bb4† 9.Bd2 Nc6 10.Bg2 Rc8 11.0-0 0-0³ White was suffering from slow development in Franco Raymundo – Gallego Eraso, Monzon 1987. The text move is more aggressive but the advanced e5-pawn will soon come under fire. 7...Bb7 8.Bg2 Ng4 9.0-0 This occurred in Kopasov – Manelyuk, Dagomys 2009, and now my new idea is:

452

9...Nc6!N 10.h3 10.Bf4 is hardly better in view of 10...d6 11.Qa4 dxe5 12.Nxe5 Ngxe5 13.Bxe5 Qd7 14.Bxd4 Nxd4 15.Qxd7† Kxd7 16.Bxb7 Rd8³ with a positional advantage for Black. 10...Ncxe5 11.Nxe5 Bxg2 12.hxg4 Bxf1 13.Kxf1 Rc8 14.Qe2 d5³ In the resulting unbalanced position I prefer Black’s chances, due to his powerful central pawns. A) 6.Bg2

6...Nc6 The d4-pawn is attacked, although White isn’t obliged to protect it immediately. Only two moves have independent value: A1) 7.0-0 and A2) 7.dxc5. 453

It is not too late for White to play more forcefully with 7.d5, when 7...exd5 8.cxd5 Nxd5 9.0-0 Bb7 reaches the critical main line as discussed in variation B22 and the next chapter. 7.Qa4 Bb7 is also likely to transpose elsewhere: 8.d5 (8.dxc5 Bxc5 is covered under 8.Qa4 Bb7 in the notes to variation A2 below) 8...exd5 9.cxd5 Nxd5 10.0-0 Be7 11.Rd1 See variation D of the next chapter on page 250. A1) 7.0-0 cxd4 8.Rd1

8...Bc5 Even though White will eventually regain the pawn, it will cost him a few tempos. 9.a3 Bb7 The light-squared bishop has done its job on a6, so it drops back to the main diagonal. 10.b4 Be7

454

11.Bb2 11.Nxd4?! Nxd4 12.Rxd4 Bxg2 13.Kxg2 enables Black to establish a favourable pawn structure by means of 13...a5! 14.b5 Qc7 15.Nd2 Rc8³, putting pressure on the backward c4-pawn. 11...0-0 12.Nxd4 Nxd4 13.Rxd4 Bxg2 14.Kxg2 Qc7 14...a5N 15.b5 Rc8 16.e4 Qc7= is also good enough.

15.Nd2 e5 16.Rh4? This awkward move only leads to trouble. More to the point was 16.Rd3 Qc6† 17.Kg1 Qe6 with mutual chances.

455

16...h6 17.Qf5 d6 Black was clearly better in Terzic – Farago, Passau 1994. A2) 7.dxc5 Bxc5

White’s last move seems rather toothless and it has not been tested at GM level. Indeed, Black easily manages to complete development and exert pressure along the c-file. 8.0-0 8.Qa4 Bb7 9.0-0 0-0 10.Nc3 Be7 11.Bf4 (11.Rd1?! transposes to a game, when 11...Na5 gave White problems protecting the c4-pawn in Blanco Ronquillo – Pulvett, Caracas 2014) 11...Na5 12.Nb5

456

12...a6 (12...d5!?N= would also be fine) 13.Nd6 Bc6 14.Qd1 Ne8 15.Nxe8 Rxe8 In this equal position a draw was agreed in Velikov – Barczay, Zalaegerszeg 1979, although there is nothing to stop either side from fighting. 8...0-0 9.a3 d5 10.cxd5 This position arose in a bunch of games, and each time Black recaptured with the knight. Although there is nothing wrong with that, I think it would be more interesting to change the pawn structure with:

10...exd5!?N 11.Rd1 11.b4?! Bd4 12.Bb2 Rc8 offers great piece play for Black. The lack of development in White’s camp would also tell after 11.Nc3 d4 12.Ne4 d3! 13.Nxf6† Qxf6 14.exd3 Rac8 when Black has full compensation for the pawn. 11...Ne4 12.Nc3

457

12...Nxf2!? This move invites interesting complications. 12...Nxc3 13.bxc3 is a more straightforward continuation, leading to approximate equality. 13.Rxd5 Ng4† 14.Rxc5 bxc5 15.Ng5 g6 16.Bxc6 White has picked up a second piece for the rook, but it’s not such a big deal. 16...Qd4† 17.e3 Qf6! 18.Bf3 Ne5 The vulnerability of White’s king offers Black decent compensation for the slight material deficit. B) 6.d5

458

This is the most energetic and consistent continuation, which implies a pawn sacrifice. 6...exd5 7.cxd5 Bb7 The d5-pawn is doomed but White has a few ways of sacrificing it. The main options are B1) 8.e4 and B2) 8.Bg2. 8.Nc3? This is clearly inferior. 8...Nxd5 9.Bg2 No better is: 9.Qe4† Qe7 10.Qxd5 (10.Ne5 is well met by 10...f5! 11.Qxf5 Nxc3 12.Qh5† g6 and White was lost in Ismailov – Ivanilov, Maralsay 2010) 10...Bxd5 11.Nxd5 Qd8 12.Bf4 d6–+ Despite the impressive activity of his minor pieces, White still only had two of them for the queen in Kantsler – Kiselev, Belgorod 1990. 9...Nxc3 10.Qxc3 d5 11.0-0 In Gallana – Lantini, Arvier 2008, a natural and strong continuation would have been: 11...Nd7N 12.Rd1 Nf6 13.Nh4 Qd7³ White does not have much for the pawn. 8.Bg5?! This was played in a rapid game by Topalov who, aside from being a former World Champion, is also an expert in the 5.Qc2 variation. Despite his success in the game, White’s unusual last move has some drawbacks, which can best be exploited by:

459

8...Be7!N 8...Bxd5 has been played in all six games so far, but this gives White more chance of justifying his last move. For instance, 9.Nc3 Be6 was seen in Topalov – S. Zhigalko, Kiev (rapid) 2013, and now 10.Ne5!N 10...Nc6 11.Nxc6 dxc6 12.Bg2 Rc8 13.Rd1 would have offered White interesting compensation for the pawn. 9.Nc3 9.Bxf6 is the only way to protect the central pawn, but 9...Bxf6 10.e4 0-0 11.Nc3 d6 leaves Black with a favourable version of a Modern Benoni due to the powerful, unopposed dark-squared bishop. 9.d6 is hardly an improvement; after 9...Bxd6 10.Bg2 Nc6 11.Nc3 Be7 12.Rd1 h6 13.Bxf6 Bxf6 14.0-0 0-0 White has a hard time proving any compensation. 9...Nxd5 10.Qe4

460

10...f5! 11.Qe5 Even worse is 11.Qxf5?! Nxc3 12.bxc3 Bxg5 13.Nxg5 Qe7 14.f3 Rf8 15.Qg4 h6 16.Ne4 Bxe4 17.fxe4 Nc6 when White’s position is ruined. 11...0-0 12.Nxd5 Bxg5 13.Nxg5 Qxg5 14.Bg2 Nc6 15.Qc7 Rab8 16.0-0 Qd8 17.Qd6 Re8³ Black keeps his extra pawn while the activity of White’s pieces is mostly neutralized. B1) 8.e4

This is the most challenging of White’s alternatives to the main line. 8...Qe7 Of course, Black should not allow his opponent to complete development and maintain his powerful pawn centre. On the flip side, Black will spend a few extra tempos moving his queen, allowing White to build a lead in development. White has two significant options: B11) 9.Bd3 and B12) 9.Bg2. After 9.Nc3?! Nxd5 10.Nxd5 Bxd5 the exchange of knights helps Black. 11.Bd3 was seen in Gyger – Schnur, email 2006, and here I suggest:

461

11...Nc6N 12.0-0 Be6 13.Bd2 Qd8 14.Bc3 Nb4 15.Bxb4 cxb4 16.Nd4 Bh3 17.Rfd1 Bc5³ Black manages to solve his development problems and maintain the extra pawn. B11) 9.Bd3

I believe this to be slightly the weaker of the two main options. Still, it has been tested at a high level and leads to complicated play. 9...Nxd5 10.a3 The main drawback of White’s last move is that he has to spend a tempo guarding b4.

462

10...Nc6 11.0-0 Nc7 If Black can safely establish a knight on d4, it will greatly reduce White’s attacking abilities. 12.Nc3 Ne6 13.Be3 Qd8 14.Rad1 White’s set-up looks rather aggressive, so Black has to play accurately. My new idea is:

14...g6!N Black is not only preparing a fianchetto, but also neutralizing the potential pressure along the b1-h7 diagonal. 14...Be7 led to an eventual draw in Mamedyarov – Gelfand, Wijk aan Zee 2006, but I find it less convincing as it can be met by a spectacular exchange sacrifice: 15.e5!?N 15...Ncd4 16.Bxd4 Bxf3 (16...cxd4? 17.Be4 is bad for Black) 17.Be3 Bxd1 18.Rxd1 Despite his significant material advantage, I don’t trust Black’s position. The king is stuck in the centre, while White’s bishops are extremely powerful. 15.Bc4 This is the most logical choice, activating the bishop and unblocking the d-file. 15.Nb5?! proves ineffective after 15...a6!, since 16.Qc3? Rg8–+ leaves the knight without a safe square. 15...Bg7 16.Nb5 16.Qd2 Ned4 17.Nxd4 Nxd4 18.Bxd4 cxd4 19.Nb5 0-0 also favours Black, for instance:

463

20.Rfe1 d5 21.Bxd5 Bxd5 22.exd5 a6 23.Nxd4 Qxd5 White has regained the pawn but Black is still somewhat better due to his powerful bishop. 16...0-0 17.Nd6 The knight has reached its perfect destination but it shouldn’t bother Black.

17...Qf6! This active move reminds us that there are targets in White’s camp as well. 18.Qe2 Rab8 19.Rd2 Qe7 Black is ready to occupy the key d4-spot at any moment. That would force major simplifications, so 464

claiming adequate compensation for the pawn is no easy task for White. B12) 9.Bg2

This method of development seems most natural. 9...Qxe4† Liquidating into a pawn-up endgame is undeniably an achievement for Black, although White will have some compensation. The greedy 9...Bxd5? 10.0-0 Bxe4 11.Qd1 d5 12.Re1 leaves White with a powerful initiative. 10.Qxe4† Nxe4 11.0-0 Bxd5 12.Re1 Despite the simplifications, the significance of White’s lead in development should not be underestimated.

465

12...f5! It is important to overprotect the knight and vacate the f7-square for the king. Now Black is ready to break the pin. The other natural move 12...Be7 runs into: 13.Nh4! Nf6 14.Nf5 Nc6 15.Nxg7† Kd8 16.Nf5 Bxg2 17.Kxg2 Nd5 (17...d5 was better, although 18.Nxe7 Nxe7 19.Bg5 Nfg8 20.b4! cxb4 21.a3 offers White promising play for the two pawns) 18.Rd1 Ndb4 19.Nc3 White had a powerful initiative for the pawn in Altrichter – Spindelboeck, corr. 2006. 13.Nc3 The impulsive 13.g4? allows Black to complete development and maintain all the merits of his set-up: 13...Be7 14.gxf5 Nc6 15.Ng5 Nf6–+ White’s position is ruined. 13.Nh4?! is also dubious in view of 13...Kf7 14.Nxf5 Nf6µ and Black is a healthy pawn up.

466

13...Bc6 14.Nh4 It seems natural for White to win back one of the pawns; on the other hand, Black will have time to catch up with development. White would not benefit from liquidating the opponent’s light-squared bishop after: 14.Ne5 Be7 15.Nxc6 Nxc6 16.Nxe4 fxe4 17.Bxe4 Rc8 18.Bf4

18...Kd8! The king is out of danger, and Black’s pieces start to cooperate well. In particular, the knight is heading for d4. 19.Rad1 g5 20.Bc1 h6 21.b3 Nd4 Black has fully consolidated, so White will have to struggle for a draw. A more viable alternative is: 467

14.Bf4!? Be7 15.Rad1 White mobilizes the last of his forces.

15...0-0N 15...Bf6 proved playable in Tanis – Laan, corr. 2015, but I find it more logical to remove the king from the danger zone as soon as possible. 16.Nd5 16.Ne5 Bf6 17.Nxc6 Nxc6 18.Nxe4 fxe4 19.Rxd7 transposes to the line below, with one less move played. 16...Bxd5 17.Rxd5 Nc6! Finally connecting the rooks. Giving a pawn back is a low price for that!

18.Nd2 468

White also achieves nothing with 18.Rxd7 Bf6 19.Nd2 Rad8 20.Rb7 Rf7, when all of Black’s pieces are active. 18...Bf6 19.Nxe4 fxe4 20.Rxd7 Equally good is 20.Bxe4 Rf7 21.Rd2 Kh8 when White has nothing better than regaining the pawn: 22.Rxd7 Rxd7 23.Bxc6 Rdd8 24.Bxa8 Rxa8= Black is fine. 20...Rad8 21.Rc7 Nb4 22.Bxe4 Nd5= Black manages to exchange one of his opponent’s bishops, so his problems are solved.

14...Kf7 15.Bf4 15.Nxf5?! is weaker in view of: 15...Nxc3 16.bxc3 g6 17.Nh6† Bxh6 18.Bxh6 Bxg2 19.Kxg2 Nc6 20.Rad1 d5! 21.Rxd5 Rad8 22.Rxd8 Rxd8³ White has restored material equality, but now he is suffering from structural drawbacks. 15...Nxc3 16.bxc3 Bxg2 17.Kxg2 Nc6

469

18.Rad1 Once again, 18.Nxf5?! would be premature: 18...g6 19.Ne3 Bg7 20.Rad1 Rhd8 21.Rd3

21...d5! 22.Bg5 This intermediate move enables White to grab the pawn, but after 22...Re8 23.Rxd5 h6³ his minor pieces are located awkwardly. 18...Rd8 19.Bc7 In the event of 19.Nxf5 d5 20.Rd3 g6 Black has nothing to worry about; the most White can hope only for is to regain the pawn and equalize. 19...Re8 20.Rxd7† Re7 470

21.Rdxe7† A more ambitious choice is 21.Rexe7† Bxe7 22.Bf4, leaving the rook on the 7th, but this does not bother Black either: 22...Kf6 23.Rc7 Na5 24.Rxa7 Re8 25.Nf3 Ke6 With the active king within reach of the weak queenside pawns, only Black can be better. 21...Bxe7 22.Nxf5 Re8 The endgame was equal in Lysyj – Debashis, Moscow 2018. B2) 8.Bg2

471

8...Nxd5 Let me remind you not to copy the approach from variation A of Chapter 10 with 8...Bxd5?, since 9.Nc3 Bc6 does not gain a tempo against the queen on a4. White plays 10.e4 and builds up a poisonous initiative, as evidenced by a practical score in excess of 80%. After the text move we will start by looking at B21) 9.Qb3, a respectable sideline, before introducing the main topic of B22) 9.0-0, coverage of which will span the end of this chapter plus the next one. B21) 9.Qb3 In comparison to the main line, White creates immediate threats along the diagonals, so Black has to retreat his knight at once.

9...Nf6 10.Ne5 This is the most consistent move and the only one to have been played thus far, even though White is moving the same piece twice and delaying his own development. I also considered: 10.Nc3N 10...Nc6 11.Bf4 (after 11.0-0 Be7 12.e4 0-0 13.e5 Ne8 14.Rd1 Nc7 15.Bf4 Ne6 16.Nd5 Re8³ White fails to claim enough compensation for the pawn) 11...Be7 12.Rd1 0-0 13.0-0 Na5 14.Qc2 d5 15.Be5 g6 16.e3 The pressure on the d5-pawn is quite strong, but Black can afford to give it back. 16...Qd7 17.Ne1 Qe6 18.Bxf6 Bxf6 19.Nxd5 Rad8³ White has restored material equality but Black’s pieces are better coordinated. 10...d5 11.Nc3 Be7

472

12.Qa4† White has also tried: 12.0-0 0-0 13.Rd1 Nc6 14.Nxd5 Nxd5 15.Bxd5 Qe8!

16.Bf4 (in the event of 16.Nd7N 16...Nd4 17.Rxd4 cxd4 18.Nxf8 Bxd5 19.Qxd5 Bxf8 20.Bd2 Rd8 21.Qf3 Qe6= Black’s pressure along the e-file fully compensates for the isolated pawn) 16...Rd8 17.e4 Bf6 18.Nxc6 Bxc6 Black had excellent play in Brunsteins – Standaert, corr. 2011. 12...Nbd7 Black should be willing to give up the bishop pair to retain castling rights. The inferior 12...Kf8?! 13.Bf4 Qe8 14.Qxe8† Kxe8 was played in Hawkins – Wells, England 2014. Had

473

White played 15.0-0-0!N 15...Na6 16.Nc4 Rd8 17.Ne3, he would have been able to regain the pawn in a favourable situation.

13.Nc6 Bxc6 14.Qxc6 Rc8 15.Qa4 15.Qb7 is hardly more promising. After 15...0-0 16.Nxd5 Nxd5 17.Qxd5 Bf6 18.0-0 Ne5 19.Qb7 Rc7 20.Qa6 Qc8 21.Qxc8 Rcxc8, the activity of all his pieces meant Black did not experience any problems in Fedorovsky – Ghosh, Riga 2017. 15...d4 16.Nb5 0-0 17.0-0 Ne5

18.Qxa7 Another natural continuation is: 474

18.Nxa7N 18...Ra8! 19.Bxa8 Qxa8 This looks extremely dangerous for White, as his knight is stuck and many of his other pieces are undeveloped. However, his material advantage should not be underestimated, so both sides need to be accurate.

20.f4! This is the only way to liberate the knight. 20...Ng6 21.b3 h5! White’s king is rather exposed, so he should try to simplify matters. 22.Nc6 Qxa4 22...Qb7!? 23.Qb5 Bd6 also comes to consideration. 23.bxa4

475

23...Rc8 24.Ne5 Nxe5 25.fxe5 Nd5 White’s rooks lack open files and he has several weak pawns, so Black’s chances are by no means worse. 18...Nc6 19.Qb7 Na5 20.Qa7 The queen is awkwardly located, but this is the best White can do. After 20.Qf3?! Qd7 21.a4 c4 22.Bd2 Nb3 23.Rad1 Qe6³ Black takes over the initiative.

20...Re8!?N If a draw is an acceptable result, Black can repeat the position with 20...Nc6= as occurred in Kramnik – Leko, Dortmund 2008. 21.Bf4 Bf8 22.e3 Another unbalanced position arises after: 22.Bc7 Qd7 23.Qxb6 Nc4 24.Qb7

476

24...Re7! 25.Rac1 Rxc7 26.Nxc7 Nd6 27.Qb6 Qxc7 28.Qxc7 Rxc7 Black has a comfortable game, since White’s rooks are not active enough. 22...Nd5 23.Bxd5 Qxd5 24.Nc7 Rxc7 25.Qxc7 White must avoid 25.Bxc7? Nc4 26.Qa4 Re7 when Black is winning.

25...g5! The bishop is trapped, but the complications are only just starting. 26.Qxb6 gxf4 27.Qxa5 fxe3 Black’s pawns look scary and White’s kingside is about to be opened up. 477

28.fxe3

28...Bh6! 29.h4 Rxe3 30.Qc7 Kg7 The strong passed d-pawn and vulnerability of White’s king provide Black with full compensation for the exchange. B22) 9.0-0

9...Nc6 Another common moves order is 9...Be7 10.Rd1 Nc6, but it yields White some extra possibilities in 10.Qe4 and 10.Ne5. 478

After the text move, 10.Rd1 is by far White’s most popular choice, and we will analyse it carefully in the next chapter. Before then, we will consider some minor options. 10.Qe4† This rare move has its merits and drawbacks. True, our knight is forced to retreat to e7, leaving the dark-squared bishop blocked for a while. On the flip side, White is moving his queen again and again, so his own development is seriously delayed. 10.a3 has been played a few times; after 10...Be7 White has nothing better than 11.Rd1, transposing to variation B of the next chapter. 10.Nh4?! is premature. 10...Nc7 11.Qe4† Qe7 12.Qxe7† Bxe7 13.Nf5 Bf8 14.Nc3 occurred in Ding Liren – Ghaem Maghami, Subic Bay 2009, when Black should have played:

14...g6N 15.Bf4 0-0-0 16.Nd6† Bxd6 17.Bxd6 Rhe8µ Despite the bishop pair, White has no real compensation for the pawn. 10...Nde7 11.Rd1 11.Nc3 is likely to transpose within a few moves, for instance: 11...d5 12.Qa4 Qd7 13.Bf4 (13.Rfd1 leads straight to our main line, and was the actual move order of the Mueller – Zill game quoted later) 13...d4

479

By now, White really should play 14.Rfd1, with a transposition to our main line after 14...Rd8. Instead after 14.Ne5?! Nxe5 15.Qxd7† Kxd7 16.Bxb7 Rb8 17.Bg2 N7g6 18.Bxe5 Nxe5 19.Rfd1 Bd6 White will struggle to maintain the balance. 11...d5 12.Qa4 Qd7 13.Nc3 Rd8! 13...d4 was played in Kozul – S. Zhigalko, Rogaska Slatina 2011, when White missed a good opportunity to seize the initiative: 14.b4!N 14....Nxb4 15.Qxd7† Kxd7 16.e3 Ned5 17.Nxd5 Nxd5 18.exd4 Despite the simplifications, Black’s king feels exposed on the open board.

14.Bf4 14.e4?N 14...d4 15.Nd5 Ng6 leaves White with no compensation for the pawn.

480

I also examined an untested but quite natural idea: 14.Bg5N 14...f6 15.Bf4 Provoking ...f6 might offer White some additional tactical possibilities, but Black is still doing fine:

15...d4 15...g5!? invites complications: 16.Ne4! Ng8! 17.Nfxg5 fxg5 18.Nxg5 Nd4 19.Rxd4 h6 20.Ne6 cxd4 21.Nc7† Kf7 22.Qxa7 The computer calls this equal but, for human players, it’s clearly a wild position where anything could happen. 16.e3 Ne5 17.Qxd7† Rxd7 18.Bxe5 dxc3 19.Bxc3 Nf5= The position has stabilized and the chances are equal. 14...d4

481

15.e3N White urgently needs to open files if he is to exploit the lack of development in Black’s camp. 15.Ng5? proved to be too optimistic after: 15...h6 16.Nb5 Nd5 17.Bxd5 hxg5

18.Nc7† Ke7 19.Bxg5† f6 20.Bf4 Qh3–+ Mueller Toepler – Zill, corr. 2008. 15...Ng6 16.Qb5 After 16.exd4 Nxd4 17.Nb5 Qxb5 18.Rxd4 Qxa4 19.Rxa4 Rd7 Black is out of danger. The text move also enables White to regain the pawn, but Black still faces no problems. My analysis continues:

482

16...Be7 17.exd4 Nxf4 18.gxf4 Nxd4 19.Nxd4 Bxg2 20.Kxg2 cxd4

21.Rxd4 Qxb5 22.Rxd8† Kxd8 23.Nxb5 a6= The resulting endgame is equal. Conclusion 5.Qc2 has risen from a mere sideline to become one of the most topical main lines in the Queen’s Indian. After the natural 5...c5 we checked a few lines after the quiet 6.Bg2 Nc6, but the really interesting battleground occurs after 6.d5 exd5 7.cxd5 Bb7, when Black wins a pawn but White gets full compensation due to his lead in development and active pieces. Despite being relatively uncommon, 8.e4 Qe7 leads to interesting positions after both 9.Bd3 and 9.Bg2. Black is theoretically fine, but it is definitely worth knowing a few theoretical details such as 14...g6!N in the former line and 12...f5! in the latter. The main line continues 8.Bg2 Nxd5, when 9.Qb3 is an interesting deviation, which we answer with 9...Nf6 intending ...d5. If White hurries to regain the sacrificed pawn, the position will simplify and Black will be fine. White’s more ambitious approach is to bring his knight to c6 to trade it for the lightsquared bishop, but this gives Black time to establish a powerful pawn wedge with ...d5-d4, after which he will have plenty of counterplay. Finally, we looked at some rare lines after 9.0-0, with the more important main lines to follow in the next chapter.

483

A) 11.Nc3 237 B) 11.a3 Nc7 12.Nc3 0-0 13.Bf4 Ne6 14.e3 Nxf4 238 B1) 15.exf4 239 B2) 15.gxf4 241 C) 11.Qf5 Nf6 12.e4 g6 13.Qf4 0-0 242 C1) 14.Nc3 243 C2) 14.e5 Nh5 245 C21) 15.Qg4 245 C22) 15.Qc4 247 D) 11.Qa4 Nf6 250 D1) 12.Nc3 252 D2) 12.e4 253 D3) 12.Nh4 0-0 259 D31) 13.Nc3 259 D32) 13.Nf5 260

484

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Ba6 5.Qc2 c5 6.d5 exd5 7.cxd5 Bb7 8.Bg2 Nxd5 9.0-0 Nc6 10.Rd1 This is the big main line of the 5.Qc2 variation. 10...Be7 This move is clearly best, as Black defends the knight indirectly while developing another piece and preparing to castle. This brings us to an important tabiya. Even though Black has developed all his minor pieces, he is still a long way from consolidating his extra pawn. The pressure along the d-file is annoying, and White has decent attacking prospects in the event that Black castles on the kingside; in particular, his king’s knight often goes to f5. We will analyse A) 11.Nc3, B) 11.a3, C) 11.Qf5 and D) 11.Qa4, the last two of which account for the great majority of games. The tactical point of Black’s last move is obvious: 11.Rxd5? Nb4 12.Qc3 Nxd5 13.Qxg7 Bf6 14.Qg4 Qe7 and Black should convert his material advantage into a full point. A) 11.Nc3

485

This move invites simplifications which ease Black’s task. 11...Nxc3 12.bxc3 d6 13.Ne5 13.Bf4 0-0 14.Nd2 Qc7 15.Nc4 leads to the same thing. 13...Qc7 14.Nc4 0-0

15.Bf4 Rad8 16.Ne3 The pressure along the d-file offers White some compensation but Black should not be in any danger, as the following examples illustrate.

486

16...Ne5! 16...Rfe8 17.Nd5 Qc8 18.Nxe7† Nxe7 19.Bxb7 Qxb7 20.Rxd6 h6 was also equal in Class – Tulyakov, corr. 2011. The text move forces White to work a bit harder to regain his pawn. 17.Nd5 Bxd5 18.Rxd5 Ng6 19.Be4 Bf6

20.Bxg6 hxg6 21.Rad1 Be5= A draw soon followed in Lauer – Poetz, corr. 2013. Even so, in a practical game White would still have to find a few good moves to ensure equality. B) 11.a3

487

Covering the b4-square is a natural way of renewing the threat to the knight. Compared to variations C and D below, White is playing for longer-term compensation based mainly on pressure along the d-file, and does not intend to advance his e-pawn too quickly, at least after our next move. 11...Nc7 Another common continuation is 11...Nf6, in which case advancing the e-pawn suddenly makes a lot more sense: 12.e4 0-0 13.e5 Ne8 14.Nc3 Nc7 15.Be3© White has promising compensation and has scored well from here. 12.Nc3 12.e4?! makes little sense without the knight on f6 as a target. 12...Ne6 13.Be3 0-0 14.Nc3 occurred in Raghunandan – Tuomainen, Barbera del Valles 2016, when Black should have opted for the simple and strong:

14...d6!N 15.e5 Qb8! 16.exd6 Bxd6³ White does not have much for the pawn. In some games White started with 12.Bf4, but it doesn’t have much independent value: 12...Ne6 13.e3 (13.Nc3 0-0 14.e3 is the same) 13...0-0 14.Nc3 This transposes to our main line below. 12...0-0 13.Bf4 This is White’s clear first choice. Other moves are easier to handle, for instance: The aggressive 13.Ne5?!N only invites 13...Qc8! 14.Nxc6 (14.Nxd7? Rd8 is even worse for White) 14...Bxc6³ when the exchanges favour Black. Just as on the previous move, it no sense to weaken the d4-square with: 13.e4?! Ne6 14.Be3 Bf6 15.Rd2

488

15...Ne5! 16.Nxe5 Bxe5 17.Ne2 Ba6 White had a hard time finding compensation in S. Singh – Sangma, Tirupati 2012. 13.Qf5?! resembles variation C, but combining this queen move with a2-a3 seems illogical. I found a natural improvement over a few existing games:

13...d6!N 14.Bf4 Ne6 15.Nd5 g6 16.Qh3 Re8 17.Nxe7† Qxe7 18.Bxd6 Qf6µ White has regained the pawn but he now suffers from a distinct lack of harmony. 13.h4 was played in Sutkus – Lucki, corr. 2009. Here I would favour 13...d5!?N, more or less forcing White to regain the pawn as follows: 14.Ng5 Bxg5 15.Bxg5 Qd7

489

16.Nxd5 Nxd5 17.Rxd5 (17.Bxd5 Nd4 18.Bxb7 Qxb7 19.Qd3 h6= is also fine for us) 17...Qe6 18.Qf5 Qxf5 19.Rxf5 Rae8 Black’s active piece play fully compensates for White’s bishop pair.

13...Ne6 14.e3 I also examined 14.Rd2N 14...Ncd4 15.Nxd4 Nxd4 16.Qd1 Bxg2 17.Kxg2 Qc8 18.e3 when White manages to regain the pawn, but 18...Ne6 19.Rxd7 Rd8= offers Black comfortable play. 14...Nxf4 Now White must make an important choice between B1) 15.exf4 and B2) 15.gxf4. The two moves are equally good but each leads to a distinctive structure with its own characteristics and nuances.

490

B1) 15.exf4

This recapture offers White a super-solid kingside structure and active prospects in the centre. The drawbacks are the weakening of the d4-square and the absence of any attacking prospects along the gfile. A further point is that Black may be able to simplify by giving up the d-pawn in some scenarios, with chances to reach a superior endgame due to his ability to create a passed pawn on the queenside. 15...d6 16.Rd2 Re8 17.Rad1 h6 Black covers the g5-square, thus preparing to retreat the bishop and start some play along e-file. White’s pressure along d-file is sufficient to maintain equality, but nothing more. 18.Nb5!?N White failed to achieve anything with either of the following moves: The overoptimistic 18.Qf5?! Qc8 19.Qh5 was played in Le Quang – Sasikiran, Loo 2013. My new idea is:

491

19...Na5!N 20.Nd5 Bf8³ White does not have much for the pawn. 18.Ne5 Nd4! is also fine for Black after: 19.Rxd4 Bxg2 20.R4d2 Bb7 21.Qb3 (21.Qf5N 21...Rf8 22.Nc4 Bf3= is also not dangerous) 21...Rf8 22.Ng6 Re8 23.Ne5 Rf8= White found nothing better than accepting the draw by repetition in Jakovenko – Karjakin, Kazan 2014. The text move is a natural choice to attack d6. For some reason, it has not yet been tried in practice.

18...Qb8 19.Qf5! 19.Nxd6?! is premature in view of 19...Bxd6 20.Rxd6 Nd4 21.Nxd4 Qxd6 22.Bxb7 cxd4 23.Bxa8 Rxa8³ when White must fight for a draw. 19...a6 20.Nxd6 Bxd6 21.Rxd6 Nd4 492

22.R1xd4! cxd4 23.Rd7 Rf8 The activity of White’s pieces compensates for the exchange, but he is no more than equal. 24.Ne5 I also checked 24.Rxb7 Qxb7 25.Ng5 g6 26.Qh3 Qxg2†! 27.Kxg2 hxg5 when Black is out of danger and White must play for a draw with 28.f5!= to avoid becoming worse. 24...Bxg2 25.Nxf7 Ra7 26.Nxh6† Kh8 27.Nf7† Kg8= B2) 15.gxf4 This seems like a more ambitious choice. White keeps control of the d4-square and opens the g-file, which may yield some attacking chances in the long run. Nevertheless, the text move also renders White’s king more exposed, as highlighted in our main line below. 15...d6 16.Rd2 This position was reached in Lenic – Navara, Reykjavik 2015, and a few other games. I like the following new plan:

493

16...Qe8!N The queen is moving towards the kingside, highlighting a drawback of White’s decision to recapture with the g-pawn. 17.Rad1 After 17.Nd5 Rd8 White has nothing better than 18.Rad1, transposing the main line below. 17...Rd8 18.Nd5 This seems best. In the event of 18.Ng5 Bxg5 19.fxg5 Na5 20.Bxb7 (after 20.Nd5?! b5!³ Black’s extra pawn starts to tell) 20...Nxb7 21.Qf5 Qe6 White will have to play precisely to claim full compensation for the pawn.

494

18...Kh8 18...f5!? is also possible, but there is no need to hurry – it is useful to remove the king from the a2-g8 diagonal in advance. 19.b4 A harmless alternative is: 19.Nxe7 Qxe7 20.Ng5 f5 21.Qa4

21...h6! 22.Bxc6 Bxc6 23.Qxc6 hxg5 24.Rxd6 gxf4 White will have to be careful to hold a draw. I also considered 19.Qf5 Bc8 20.Qe4 (20.Qh5?! g6 21.Qh6 f6³ leaves White’s queen misplaced) 20...f5 21.Qa4 Bd7 when White can hardly hope for an advantage.

495

19...f5! As you may have guessed, this space-gaining move was one of the main ideas behind putting the queen on e8, as well as the king on h8. 20.Ne1 Bf6 21.Qa4 Rf7 22.Qb5 Ba8 It is hard for White to improve his position, and he can only hope that Black will not find a favourable way to open the kingside. C) 11.Qf5

Compared with the other main queen move to a4, here the queen moves more directly towards Black’s 496

king. On the flip side, it can be easily attacked by Black’s pieces. 11...Nf6 12.e4 It hardly makes sense to postpone this thematic advance. For instance: 12.Nc3 d6 13.Bf4 13.e4?! Qd7 14.Qf4 0-0 15.e5 Ne8 16.Be3 Rd8 leaves White without much play for the pawn. 13...0-0 14.Rd2 g6 15.Qc2 d5

16.Rad1N After 16.Ne5 Nd4 17.Qa4 Qe8 18.Qd1 Kg7 19.e3 Ne6 White had some problems in Gatineau – Esipenko, Mamaia 2017. The text move is a natural attempt to improve. 16...Qc8! 17.e3 17.Nxd5? runs into a simple refutation: 17...Nxd5 18.Rxd5 Nb4–+ 17...Rd8 18.Ne5 Qe6= White should be able to regain the pawn, but Black has no problems. 12...g6 This move is quite committal but Black has make it in order to castle. The immediate 12...0-0? invites serious trouble after 13.e5 g6 14.Qh3. 12...d6 has also been tested at a high level, but after 13.e5 Qd7 14.Qxd7† Nxd7 15.exd6 Bf6 16.Re1† Kf8 17.Nc3 White’s powerful passer gives him the easier play, to say the least. 13.Qf4 This is the only adequate retreat.

497

13.Qh3? would be a poor choice in view of: 13...Nxe4 14.Bh6 d5 15.Nc3 Nxc3 16.bxc3 Qc8µ 13...0-0 13...Nh5?! has been tested in some correspondence games but I don’t like it so much, since Black’s king remains stuck in the centre. After 14.Qh6 Bf8 15.Qe3 Bg7 the castling problem is solved, but 16.e5! leaves the knight short of squares. I tried to improve Black’s play with 16...0-0N 17.g4 Re8 18.gxh5 Nxe5 19.Nxe5 Bxg2 20.Kxg2 Rxe5 21.Qf3 when the position is still complex, but White is clearly better. White has tried two moves here: C1) 14.Nc3 is a natural developer but C2) 14.e5 is more testing. C1) 14.Nc3

This offers Black more breathing room in the centre. 14...d6 15.b3 White cannot afford slow play, such as 15.Be3? Re8 16.Nb5 Nh5 17.Qh6 Bf8 when Black was already winning in Jayakumar – Gareev, Los Angeles 2012. 15.e5? also fails to impress. In Xiu Deshun – Liu Guanchu, Beijing 2017, the natural 15...dxe5N 16.Nxe5 Qc8 17.Qf3 Na5µ would have left White without adequate compensation. 15...Nh5 Now White requires a novelty to make sense of his position.

498

16.Qh6N The passive 16.Qe3?! is not really in the spirit of the position. 16...Bf6 17.Bb2 was seen in Scepanovic – Perunovic, Belgrade 2007, when Black should have played:

17...Re8N 18.Rab1 Ne5 19.Nxe5 Bxe5 20.f4 Bh8 21.e5 Bxg2 22.Kxg2 Qb8!³ Highlighting the drawback of the white queen’s location. 16.Qd2?! Bf6 17.Bb2 was hardly any better in Swapnil – Bok, Bangkok 2016. My new idea is:

499

17...Ng7!N 18.Nd5 Bxb2 19.Qxb2 f6 20.b4 Rb8³ White fails to claim adequate compensation. 16...Bf6 Covering the long diagonal.

17.Ng5 This is obviously the right choice, as liquidating the dark-squared bishop is a clear achievement for White. The prosaic 17.Bb2?! Nd4³ is clearly worse for White.

500

17...Bxg5 18.Bxg5 f6 19.Bc1! White should reroute the bishop to the long diagonal. 19.Be3?! Qd7 20.Nd5 Rae8 21.f3 Ne5 22.g4 Ng7µ would be worse for him.

19...Nd4 20.Bb2 Ng7 21.Nd5 Nge6 22.Rd3 Bxd5 23.exd5 Ng7

24.Re1 Ngf5 25.Qd2 Re8= The strength of the knight on d4 means that White has nothing better than exchanging his dark-squared bishop, after which Black will not be in any danger. C2) 14.e5 501

This is the most consistent and strongest choice, which offers White decent attacking prospects. 14...Nh5 White’s two most important options are C21) 15.Qg4 and C22) 15.Qc4. 15.Qh6? This tempting yet substandard move has been played in a few high-level games. Black has more than one decent reply but the most convincing is:

15...Qc7! 15...Nd4 16.Nxd4 Bxg2 17.Kxg2 cxd4 18.Rxd4 Qc7 19.Nc3 Qxe5 20.Rxd7 Qe6 was an easy route to equality in Bacrot – Karjakin, Dubai (rapid) 2014, but Black can play for more. 502

16.Bg5 16.Nc3 Nxe5 17.Nxe5 Bxg2 18.Kxg2 Qxe5 19.Rxd7 Rfd8–+ is useless for White. 16.Qe3 d6 17.exd6 Bxd6 18.Nc3 Rad8 19.Nd5 Qc8 20.Qc3 Ne7µ also leaves him with nothing for the pawn. 16...Nxe5! The tactics work out perfectly for Black. 17.Nxe5 Qxe5 18.Bxe7 Bxg2 19.Bxf8

19...Bc6! White will have to return his extra material to avoid being killed on the light squares. 20.Nc3 Rxf8 21.Qe3 Qxe3 22.fxe3 Re8µ Black converted his endgame advantage in Schulz – Ryvova, corr. 2008. C21) 15.Qg4

503

This move was played by the AlphaZero program against Stockfish in the well-publicized 2017 match. 15...Qb8! This is the most reliable move, attacking the e5-pawn while getting the queen to safety. 15...Re8 seems less natural, and 16.Nc3 Qb8 17.Nd5 Bf8 18.Bf4 gave White a powerful initiative, which was later converted into an impressive victory in AlphaZero – Stockfish, engine game 2017. 15...d5 16.exd6 Bxd6 17.Bg5! Qb8 18.Qd7 also gave White promising play for the pawn in Zhukova – Kursova, Mardin 2011. 16.Nc3 In a couple of games White went for a different move order: 16.Bh6?! Re8 17.Nc3 This is less accurate in view of:

504

17...Nxe5! 18.Nxe5 Bxg2 19.Nxf7? 19.Nxg6N is objectively the lesser evil, even though 19...hxg6 20.Kxg2 Qb7† 21.Qf3 Qxf3† 22.Kxf3 Rad8 reaches an endgame where it will not be easy for White to claim sufficient compensation for the pawn. 19...Nf6!µ 19...Bc6!N is the more accurate engine move which leads to a near-decisive advantage if followed up precisely, but there are some weird details in the variations. The text move is a more human choice. 20.Qh4 20.Qc4 b5! is also clearly better for Black, the main idea being 21.Nxb5 d5 of course.

20...Bc6 21.Bf4 Nh5! 505

Black was clearly better and went on to win a good game in Y. Vovk – Melkumyan, Lvov 2007. The text move is more natural, and leads to interesting complications.

16...Nxe5! 17.Nxe5 Bxg2 18.Nxd7 Qb7 19.Nxf8 Bh1 19...Nf6!? is another move leading to sharp complications and eventual equality after: 20.Qh4 Bh1 21.f3 Qxf3 22.Rd2 c4 23.Nd7 Re8 24.Rf2 Bc5 25.Be3 Bxe3 26.Nxf6† Kf8

27.Qh6†!? (several correspondence games have ended in perpetual after 27.Nd7† Kg8 28.Nf6†=) 27...Bxh6 28.Rxf3 Bxf3 29.Nxe8 Kxe8„ We have reached an endgame with mutual chances.

506

20.Nd5 I also checked 20.Ne4!?N 20...Nf6! (not 20...Bxe4? 21.Rd7 Qc6 22.Nxg6 hxg6 23.Rxe7 Bh1 24.Qh3±) 21.Nxf6† Bxf6 22.Qh3 Rxf8 23.Bh6 Re8© when Black’s control over the light squares yields full compensation. 20...Bxd5 21.Qd7 Nf6 22.Qxb7 Bxb7 23.Nd7 Nd5 24.Ne5 Bf6© Black was subsequently outplayed in Kachanov – Camps, corr. 2015, but at this stage he had decent compensation for the exchange, with an extra pawn and a strong pair of bishops. C22) 15.Qc4

507

15...d5! I like the concept behind this move: since White’s queenside pieces are not developed, it makes sense to open the position, even if it means returning the extra pawn. 16.exd6 So far, everyone has responded this way, fighting for the initiative rather than taking back the material. 16.Rxd5N is indeed toothless. 16...Na5! is a simple tactic leading to simplifications, and after 17.Rxd8 Nxc4 18.Rxa8 Rxa8 19.Nc3 Rd8 20.b3 Na5= Black is comfortable. 16...Bxd6 The pin along the d-file may appear dangerous, but White’s undeveloped pieces are in no position to attack the bishop. 17.Nc3 This has been played in almost all the games to have reached this position so far. White can bring his bishop into action first: 17.Bh6 Re8 18.Nc3 Na5 19.Qg4 Bc6 20.Ng5 Nf6 21.Qh4 The natural play of both sides led to an exciting position in Rodshtein – Volke, Kallithea 2008. I think Black’s safest continuation would have been:

21...Qe7N 22.Nd5 Bxd5 23.Bxd5 Nxd5 24.Rxd5 The exchange of knights has made the h7-pawn vulnerable, but Black is in time to consolidate with: 24...Be5! 25.Re1 f6

508

26.Rexe5! There is no way back, so White has to extend his initiative with a further sacrifice. 26.Nf3?! Nc6µ leaves White a pawn down with no real compensation. 26...fxe5 27.b4 Rad8!? If a draw is an acceptable result, Black can play 27...cxb4 28.Rxa5 bxa5 29.Qc4† Kh8 30.Nf7† Kg8= when White must take a perpetual. 28.Rxd8 Rxd8 29.bxa5

29...Rd1† 30.Kg2 Rd4 31.Qh3 Qb7† Black’s chances in this unbalanced position are by no means worse.

509

17...Na5 18.Qd3 18.Qa4?! is inaccurate. 18...Bc6 19.Qc2 Qc7 20.Bh6 occurred in Postny – Stefansson, Stockholm 2016, when Black should have continued:

20...Rfe8!N 21.Nd5 Qb7 22.g4 Bxd5 23.gxh5 Bxf3 24.Qc3 Be5 25.Qxf3 Qxf3 26.Bxf3 Rad8µ White finds himself in a difficult endgame with no compensation for the pawn. 18...Be7! This is clearly best; the bishop is covering the exposed dark squares around the king. 18...Bc7 is less precise in view of 19.Qc2 when White had promising play for the pawn in Gelfand – Leko, Jermuk 2009, and several other games. 510

19.Qe2 Qe8 20.Bh6 Ng7 21.Nd5 I also checked 21.Rd3N 21...Bf6 22.Re3, but found that 22...Qc8! 23.Ne4 Bxe4 24.Rxe4 Nc6 25.Re1 Qc7³ enables Black to consolidate while keeping his extra pawn. The text move is most logical, as it virtually forces Black to give up the bishop pair.

21...Bxd5 22.Rxd5 Bf6 23.Qd2 White’s initiative is mainly based on the activity of his heavy pieces, but this is a temporary factor. 23...Rd8 24.Re1 Rxd5 25.Qxd5 Qd8 26.Ng5 White has to press against f7. 26.Ne5?! is less precise in view of 26...Qxd5 27.Bxd5 Bxe5 28.Rxe5 Rd8 when, despite his powerful bishops, White had to struggle for a draw in Bromberger – Volokitin, Warsaw 2015.

511

26...Nc6! Centralizing with great effect! Alternatively, 26...Qxd5 27.Bxd5 Bxg5 28.Bxg5 Re8 29.Rd1 Ne6 30.Bf6© leads to an endgame where White is pretty safe and Black will have to take care against the mighty bishops. 27.Ne4 Nd4 28.Qb7 This is White’s latest try. Instead, 28.Qd7 Ndf5 29.Nxf6† Qxf6 30.Bf4 was recommended by Hera & Tuncer in their well-researched guide to this gambit variation. White’s active bishops offer full compensation for the pawn, but Black can stay safe with:

30...Ne6!N (30...Qxb2 led to an eventual draw in Joppich – Vesely, corr. 2013, but the text move is less

512

risky for a practical player) 31.Be5 Qe7 32.Qxe7 Nxe7 33.Bf6 Nf5 34.Bd5 Nfd4= Black has an extra pawn and there is not much White can do to improve his position, other than winning back a pawn by swapping one or both of his bishops for the black knights, which will lead to a drawn ending. 28...Qe7 29.Bd2! White is creatively maintaining the tension. In case you hadn’t guessed, we are following a pair of correspondence games. 29...Ngf5 Black has consolidated and will keep his extra pawn, but White’s strong bishops are enough to maintain the balance against the knights. 30.Nxf6† The more recent game continued 30.Bc3 Qxb7 31.Nxf6† Kg7 32.Nh5† Kh6 33.Bxb7 Kxh5 34.b4 Rb8 35.Bg2 Kh6 and a draw was soon agreed in Claridge – Salzmann, corr. 2016. 30...Qxf6 31.Qxa7 Nb5! This tactical resource enables Black to restore his one-pawn advantage.

32.Qd7 Qxb2 33.Bd5 Nc7!? 33...Qd4N 34.Re4 Qa1† 35.Re1 Qd4= was a quicker route to a draw. 34.Bb3 Qd4 35.Qxc7 Qxd2

513

36.Rd1 Qb2 37.Rd7 Qf6 38.Bd5 Nd4= We have been following the well-played game Walczak – Horvat, corr. 2014. An impasse has been reached, as White can hardly improve his pieces, while Black cannot abandon the defence of the f7-pawn. Thus, a draw was soon agreed. D) 11.Qa4

This way of handling the position is conceptually similar to variation C, as White renews the threat to the d5-knight while preparing to deploy the queen on the kingside. But chess is a concrete game, and naturally there are some differences in how the subsequent play unfolds.

514

11...Nf6 Once again, this is clearly the best retreat; the knight supports the ...d5 advance and takes control of some important central squares. On the flip side, Black is inviting White to push his e- or even g-pawn to attack it. White has three main options: D1) 12.Nc3, D2) 12.e4 and D3) 12.Nh4. In one game, White tried: 12.g4?! This opening experiment was successful in Inarkiev – Ganguly, Berlin (blitz) 2015, but it is not really correct. Had it been a longer time control, Black might have shown the drawback of White’s last move by means of:

12...h6N 13.h4 a6! It turns out that White’s queen doesn’t have many good squares available. 14.g5 hxg5 15.hxg5 b5 16.Qc2 I also checked: 16.Qf4 Nh5 17.Qg4 d6 18.Nc3 Qc8 19.Qh4 White’s queen is poorly located, but this is the only way to avoid exchanging queens. 19...Nd4 20.e4 Nxf3† 21.Bxf3 Rh7 22.Be3 g6³ We have reached a double-edged position where White’s king is no safer that Black’s, while he is still a pawn down.

515

16...Nb4 17.Qc3 Nh5! This move invites favourable complications. Of course, 17...Nbd5 18.Qd3 Nb4= is fine if Black is happy with a draw. 18.Ne5 This is the only adequate response; the d-pawn is a clear target for White’s pieces. 18...Bxg2 19.Kxg2 19.Rxd7? Qc8 20.Kxg2 Nc6 leads to a decisive material gain for Black. 19...Bxg5 20.Qf3 After 20.Rxd7? Qf6 21.Rxf7 Qh6–+ White’s king is absolutely helpless.

20...Bf6! 21.Nxd7 Once again, 21.Rxd7?! does not work. After 21...Qc8 White does not have time to complete his 516

development in view of: 22.Nc3 Nc6! 23.Nxf7 Kxd7 24.Nxh8 Qxh8µ 21...Ra7! The opposition of rook and queen along the d-file does not bother Black, since his pieces are better mobilized. 22.Nxf6† Qxf6 23.Qe4† Qe7 24.Nc3 Nf6 25.Qf5 Qe6 Black has successfully consolidated, so White has to struggle for a draw. D1) 12.Nc3

Despite being a seemingly natural developing move, this has been a pretty rare choice. It has scored quite well though. 12...0-0 13.g4 13.Nh4 is covered under the move order 12.Nh4 0-0 13.Nc3 – see variation D31 on page 259. An inferior alternative is: 13.Bg5? h6 14.Qh4?! The lesser evil was 14.Bxf6 Bxf6µ, but in that case White clearly does not have much for the pawn. This dubious piece sac was tried in Tugui – Papin, Cappelle-la-Grande 2013. Black responded with 14...Nd4!?, which brought him an advantage and a subsequent victory. However, he could have refuted the sacrifice outright with:

517

14...hxg5N 15.Nxg5 Nh5! The only defence, but good enough. 16.Nce4 After 16.Qxh5 Bxg5 17.Be4 g6 18.Bxg6 fxg6 19.Qxg6† Kh8 there is no perpetual so Black is winning, for instance: 20.Ne4 Qe7 21.Nxg5 Nd4–+ 16...Bxg5 17.Nxg5 Nf6 18.Rd6 Re8

19.Rad1 Rxe2 20.Bf3 Re5 21.Bxc6 21.Rxf6 Rxg5–+ 21...Bxc6 22.Rxc6 dxc6 23.Rxd8† Rxd8–+ With two active rooks against the queen, plus an extra pawn, Black is winning.

518

This position was first tested in Shirov – Aronian, Elista (2.2) 2007, and has been repeated in several games. It seems to me that the most effective way of neutralizing White’s activity has not yet been shown:

13...Rb8!N Having analysed a few options, I believe the main cause of potential danger to be the hanging bishop on b7. Therefore it makes sense to protect it immediately. Here is one line I analysed which illustrates the above point. 13...h6 was played in Inarkiev – Kovalenko, Berlin (blitz) 2015, and now 14.h4!N would have been a good response. For instance, 14...d6 15.g5 hxg5 16.hxg5 Nh5 17.Qh4 g6 18.Ne5! and White has promising play for a pawn. The last move exploits the loose bishop, and prepares Ng4 to target the dark squares on the kingside, perhaps supported by the other knight coming to e4 or d5. 14.g5 Nh5 15.Qg4 g6

519

16.Qxd7 This is the safer way of regaining the pawn. 16.Rxd7 is well met by 16...Bc8!, leading to the liquidation of White’s best pieces. Play continues 17.Rxd8 Bxg4 18.Rxb8 Rxb8 when White must play accurately in order to neutralize Black’s activity and maintain the balance. 16...Nd4 17.Qxd8 Rfxd8

18.Nxd4 cxd4 19.Nb5 Bxg2 20.Kxg2 Rd5 In the arising endgame, Black does not experience any problems. 520

D2) 12.e4

Hitting the recently returned black knight with e4-e5 is a natural reaction from White. 12...0-0 13.e5 I am not sure why White would postpone this move, but he has occasionally done so: 13.Nc3 Qb8! This precise move enables Black to take control of e5 while avoiding any pin along the d-file. White’s best reaction is:

14.Nd5N 521

14.Bf4?! is too slow, and after 14...d6 15.Rd2 Ne5 16.Nh4 g6 White did not have much for the pawn in Clemens – Debashis, Hoogeveen 2016. The text move is more consistent; White is trying to exploit the temporary lack of harmony in Black’s camp. 14...d6 15.Bg5 In the event of 15.Nxe7† Nxe7 16.Nh4 Qe8! 17.Qc2 Ng6³ White’s initiative comes to an end. 15...Qd8 16.Nxe7† Qxe7 17.Nh4 Nd4

18.Rxd4! cxd4 19.Bxf6! 19.Nf5 Qe6 20.Qxd4 Rad8³ does not give White enough. 19...Qxf6 19...gxf6 20.Nf5 offers White definite compensation for the exchange. 20.e5 dxe5 21.Bxb7 Rae8 We have reached an interesting, roughly balanced position with rook against two minor pieces, where the powerful central pawns should offer Black plenty of counterplay.

522

13...Ne8 14.Nc3 Nc7 It makes sense to activate the knight and connect the heavy pieces as soon as it possible. Now White is at a crossroads; he has a space advantage and decent attacking prospects due to powerful e5-pawn, but he still has to come up with a plan of attack. 15.Be3 This natural developer is the best and most popular choice. White diminishes Black’s control of d4 and prepares to double the rooks, leading to strong pressure along the d-file. 15.Qg4 Ne6 16.Ne4 This aggressive set-up was tested in Bhambure – Godbole, Mumbai 2014. I think it should be met with the following thematic countermeasure:

523

16...f5! 17.exf6 Bxf6 18.Nfg5 After 18.Nd6 Ba6 Black’s minor pieces are extremely active, so White’s compensation is questionable. 18.Bg5 Bxg5 19.Nfxg5 Ncd4 20.Nxe6 Nxe6³ is also good for Black. 18...Bxg5 19.Nxg5 Nxg5 20.Bxg5

20...Ne5! 21.Bxd8 Nxg4 22.Bxb7 Raxd8³ White remains a pawn down and will have to grovel for a draw. White tried advancing the h-pawn in one game but it doesn’t seem effective: 15.h4 Ne6 16.h5 f5! As usual, this move enables Black to release the tension on the kingside and activate his rook. 524

17.Nd5 We have been following Jankovic – Palac, Sibenik 2008. A natural improvement is:

17...Kh8N 18.Be3 After 18.h6 g5! 19.Nxe7 Qxe7 20.Nxg5 Nxg5 21.Bxg5 Qxg5 22.Rxd7 Qxh6 23.Rxb7 Nd4µ the powerful knight gives Black the upper hand. 18...Qe8 19.Nxe7 19.h6 g5 20.Nxe7 Qxe7 21.Nxg5 Nxg5 22.Bxg5 transposes to the note above. 19...Qxe7 White fails to claim adequate compensation for the pawn. For instance: 20.Qh4 Rad8³ Finally, White has tried the mysterious plan of provoking the knight to come to e6 with: 15.Bf4 Ne6 I examined a new try for White:

525

16.Rd2!?N 16.Be3 was seen in Sargissian – Beliavsky, Paks 2008, and one other game. Presumably White’s idea in luring the knight to e6 was to enable his knight to go to d5, but I find it hard to believe that this is worth losing a tempo for. Afte16...f5!N 17.Nd5 Qe8 18.Rd2 Rd8 19.Rad1 Kh8 Black reaches a comfortable position while maintaining his extra pawn. 16...Na5! 17.Nd5 If 17.Rxd7 Qe8³ the pin is annoying for White. Also after 17.Qxd7 Qxd7 18.Rxd7 Rfe8 19.Nd5 Bf8 White has equalized the material balance, but Black’s chances are preferable due to his superior piece coordination. The text move is the best way; White should maintain the tension. Nevertheless, Black is doing well after:

526

17...Bc6 18.Qc2 b5! 19.b3 Rc8 Black has a safe position with excellent prospects on the queenside.

15...f5! I like this thematic continuation. If White avoids capturing en passant, the mobility of his queen and knights will be severely limited, whereas Black’s queen will have some prospects along the e8-h5 diagonal. 16.exf6 Since White’s pieces are currently better coordinated, it makes sense to open the position while avoiding the problems noted above. On the other hand, swapping off the e5-pawn gives Black useful breathing room. Alternatives are less challenging, for instance: 16.Qc4† seems rather pointless, and after 16...Ne6 17.Nd5 Qe8 18.a3 Rc8 Black was doing well in Groth – Penzler, corr. 2009. 16.Rd2 seems too slow; after 16...Qe8 17.Rad1 Rd8 18.Ng5 Kh8³ White failed to achieve anything on the kingside while remaining a pawn down in Sciarretta – Langer, corr. 2011. It is also worth checking: 16.e6?! Nxe6 17.Ne5 Even though White’s concept is not quite correct, it is worth seeing how Black should navigate the complications.

527

17...Nxe5 18.Bxb7 In Potkin – Volokitin, Aix-les-Bains 2011, the players agreed a draw here, but Black could have obtained some advantage as follows: 18...Bf6! 19.f4 19.Bxa8? Qxa8 gives Black too much initiative, for instance: 20.Nd5 Nf3† 21.Kg2 Nfd4–+ Black is also doing great after: 19.Bf4 Rb8 20.Bd5 Qe7 21.Re1 d6 22.Bd2 g5 23.f4 gxf4 24.gxf4 Kh8 25.fxe5 dxe5 Black has three pawns for the piece and his king is much safer.

19...Ng4 20.Rxd7 Qe8 21.Bxa8 Qxa8 The vulnerability of White’s monarch will lead to a loss of material. 22.Qb3 c4! 23.Qxc4

528

23...Qg2†! 24.Kxg2 Nxe3† 25.Kh1 Nxc4³ The active minor pieces are clearly stronger than the rook, so Black can press in the endgame. Finally, 16.Nd5!?N is an interesting move which, for some reason, nobody has tried yet. After 16...Nxd5 17.Rxd5 White will double his rooks as fast as possible. After 17...Qe8 18.Rd2 Rd8 19.Rad1 it seems to me that Black should not be too greedy and instead simplify the position with:

19...Na5! 20.Rxd7 Bc6 21.Rxd8 Bxa4 22.Rxe8 Rxe8 23.b3 Bc6= With a balanced endgame. 16...Bxf6

529

17.Rd6 The presence of the rook in Black’s camp is rather annoying, but he can deal with it. 17...Qe7 18.Rad1 Rad8 19.Ne4 Ne8 20.Nxf6† I also checked 20.Bf4!?N 20...Nxd6 21.Bxd6 Qe6 22.Bxf8 Rxf8 23.Nd6 when White appears to be doing well, since moving the bishop on b7 would allow a nasty tactic with Ng5. However, Black has a tactical solution of his own:

23...Nd4! 24.Nxb7 Qe2 Black’s active pieces offer plenty of play for the material. For instance: 25.Nxd4 (25.Nd2 b5 26.Bf1 Qxd2 27.Rxd2 bxa4=) 25...Bxd4 26.Qb3† c4 27.Bd5† Kh8 28.Bxc4 Bxf2† 29.Kg2 Qe4† 30.Kh3 Qxb7 This should be a draw, but White has to be more careful.

530

20...Nxf6 Black no longer has the bishop pair; on the positive side, the knight guards the d7-pawn, freeing up the rook for other activities.

21.Bf4 Rde8 Now all three of Black’s heavy pieces are well placed on open files. 22.a3! White is creatively maintaining the tension. In particular, a timely b2-b4 could prove annoying for Black. Once again though, there are correspondence games which demonstrate that the position remains balanced. 22...Qe4 23.Qxe4 Rxe4 24.Bg5 Also after 24.b4N 24...cxb4 25.Nd2 Re7 26.axb4 Ba6 Black does not experience any problems. 24...Bc8 25.Nh4

531

25...Rd4 I also like the untested 25...Rfe8!?N 26.Bxe4 Nxe4, whereby Black sacrifices the exchange to change the course of the game. Play may continue:

27.R6d5 d6 28.Re1 Nd4 29.f3 Nf6 30.Rxe8† Nxe8= The rook is trapped, so the material balance will be restored with an equal endgame. 26.Nf5 Rxd1† 27.Rxd1 d5

532

28.Bxf6 Rxf6 29.Bxd5† Kf8 30.Ne3 Nd4= A draw was soon agreed in Pasko – Boskovic, corr. 2012. D3) 12.Nh4

This move has been the most popular choice of all. The appearance of White’s knight on f5 would significantly increase White’s attacking potential. In addition, the long diagonal is unblocked, so the d5square may be used by White’s pieces. 12...0-0 We will consider D31) 13.Nc3 followed by the main line of D32) 13.Nf5.

533

D31) 13.Nc3 This rare move seems a little inconsistent. Still, it has been employed at the top level and must be taken seriously. 13...g6 Black should take the opportunity to cover the f5-square.

14.Bg5 I also considered 14.Bf4N, which is playable but slightly less accurate. A good reply is: 14...a6 (Black could also transpose to the main line with 14...Rc8, but the text move seems slightly more useful) 15.Bd6 Re8 16.Qf4 White maintains some pressure on the d-pawn, but Black can free his position by giving it up. 16...Nh5! 17.Bxe7 Nxe7 18.Qd6 Bxg2 19.Nxg2 Nf5 20.Qxd7 Nd4 Black is fine. 14...Rb8 15.Bf4 Rc8 16.Bh6 It looks odd to move the bishop for a third time, but my guess is that White wanted to lure the rook to c8 in order to prevent Black from moving his queen to that square. 16.Rd2N can be met by 16...a6! 17.Qd1 b5 when the queenside expansion offers Black excellent counterplay. 16...Re8 17.Qf4 We have been following the top-level game Aronian – Nakamura, Moscow 2016. White’s last move prepared Nf5 to eliminate the dark-squared bishop. Black has a number of playable moves but my favourite is:

534

17...d5!N I like this concept: Black gives back his extra pawn in order to free his position and activate his pieces. 18.Bxd5! 18.Nxd5 is worse in view of 18...Nh5! 19.Nxe7† Qxe7 20.Qg5 Qxg5 21.Bxg5 Rxe2 22.Rd7 Ba8 when White does not have enough for the pawn. 18...Nxd5 19.Nxd5 Liquidating the powerful light-squared bishop is an achievement, but the opposition of Black’s queen and the enemy rook along the d-file is a cause for concern. Fortunately, the following centralization of the knight comes just in time:

535

19...Nd4! 20.Rxd4 Bxd5 21.Rd2 Bxh4 22.gxh4 Re4 Another important move, giving Black time to break the pin. 23.Qf3 Rd4 Both sides have weaknesses and the position is balanced. A logical continuation is: 24.Rxd4 cxd4 25.Qf4 f6 26.Rd1 Qe7 27.Rxd4 Bxa2= A draw seems likely. D32) 13.Nf5

536

This is the most active and consistent move, and the more popular choice by far. 13...d5 Once again, we make this move with a view to returning the extra pawn in a favourable situation. 14.Nc3 Nd4! 15.Nxd4 cxd4 The troublesome f5-knight has been exchanged, and the appearance of either the rook or the queen on d4 will offer Black an important tempo for activating his pieces.

16.Rxd4 Less challenging is: 16.Qxd4 Bc5 17.Qh4 (17.Qa4 transposes to 17.Rd1 in the note to the main line below) 17...Ne4 18.Qxd8 Raxd8 19.Bxe4 dxe4 20.Be3 This endgame arose in Radjabov – Harikrishna, Shamkir 2016. In the game, Black exchanged on e3 and drew easily. If Black wishes to play more ambitiously, he could consider:

537

20...Be7!?N 21.Nb5 a6 22.Nd4 Bf6 Black has a comfortable position and can try to make the bishop pair count. 16...Bc5 17.Rd3 This seems like slightly the more challenging of the sensible rook moves. We must also consider the alternative: 17.Rd1 Qc8! The b7-bishop is protected, so Black’s set-up is harmonious now. However, it is connected with a pawn sacrifice. 18.Nxd5 Accepting the challenge. Trying to establish a blockade on d4 and play against the isolated pawn seems too passive: 18.e3N 18...Rd8 19.Ne2 Ne4 20.Nd4 h5„ Black has good prospects on the kingside. 18...Nxd5 19.Bxd5 Bxd5 20.Rxd5

538

20...Qe6 Black activates his queen with a double attack, so White still has no time to complete development. 21.Qd1 21.e4 Rfe8 22.Bf4 Qxe4 23.Qxe4 Rxe4 24.Rad1 Rae8 25.Kf1 f6 reached an equal endgame in Leko – Karjakin, Wijk aan Zee 2010. 21...Qf6! Attacking f2, while eyeing b2 and preparing to challenge for the d-file.

22.e3 The most challenging – White insists on maintaining his material advantage. 22.Bf4 Qxb2 23.Be5 Qb4 24.Qd3 Rfe8 25.Rc1 Qa3 was equal in Fridman – S. Zhigalko, Porticcio 2018. 539

22...Rad8 23.Rxd8 Rxd8 24.Qe2 h6 25.Kg2 a5= White was unable to develop his bishop without losing his extra pawn in Narayanan – Debashis, Patna 2017.

17...Qe8 It is useful to get the queen off the d-file while offering a favourable exchange. 18.Qh4 18.Be3 Qxa4 (18...Bxe3 19.Rxe3 Qxa4 20.Nxa4 is the same thing) 19.Nxa4 Bxe3 20.Rxe3 d4 21.Rd3 Bxg2 22.Kxg2 Rfe8= soon led to a draw in Can – Kovalenko, Iasi 2015, and a few other games. 18...Ba6! 18...Ne4 led to a draw in a previous high-level game but I find the text move and subsequent follow-up more convincing. Black continues to activate his pieces with tempo.

540

19.Rd1 d4! 20.Bxa8 Black is doing fine after: 20.Ne4N 20...Nxe4 21.Bxe4 h5! 22.Bxa8 Qxa8!? (an equally valid option is 22...Qxe2, when White should force a perpetual: 23.Bh6! Rxa8 24.Bxg7 Kxg7 25.Qg5† Kf8 26.Qh6†=)

23.b4 Bxe2 24.Rd2 Re8 25.Bb2 Re4 26.Qh3 Bxb4 27.Rxd4 Bc5 Black’s bishop pair and the light-square targets on the kingside provide Black with full compensation for the exchange.

541

20...Qxa8!?N The game continued 20...dxc3 21.Bf3 Bxe2 22.Re1 Bxf3 23.Rxe8 Rxe8 24.Be3 cxb2 25.Rb1 Ba3 26.Qa4 Be4 27.Qxa3 Bxb1 28.Qxb2 Be4 and Black achieved an effortless draw in Wojtaszek – Leko, Reykjavik 2015. The text move is an interesting way of keeping the game more complex and fighting for the initiative. 21.Rxd4 Re8! Black should not be too focused on regaining material: 21...Bxd4? 22.Qxd4 Re8 23.Be3± 22.Be3 If White tries to exchange his rook for another bishop by means of 22.Rc4, Black should maintain the tension: 22...Qc6! 23.Rxc5 bxc5 24.Qa4 Qxa4 25.Nxa4 Rxe2 26.Be3 Black’s queenside pawns are weak, so White still hopes for some kind of endgame advantage, but Black can sustain his initiative with:

542

26...Ng4! 27.Rd1 h6 28.Nxc5 Bc4 29.b3 Bb5 30.a4 Bc6 The activity of Black’s pieces means that he is not worse.

22...Qc6! The threats along the a8-h1 diagonal are dangerous, so White has to be careful. 23.Qf4 The timid 23.Rad1?? leads to disaster after 23...Bb7, since f2-f3 would leave the bishop hanging. 23...Bb7 24.f3 Nh5 25.Qg5

543

25...Nxg3! 26.hxg3 Also after 26.Qxg3 Rxe3 27.Rad1 Re8= Black is doing well. For the moment Black is a rook down, but White cannot maintain his material advantage after: 26...h6 27.Qf5 27.Qf4?! g5 28.Qf5 Rxe3 is only dangerous for White.

27...Rxe3 28.Rad1 Re8= The complications have subsided and the resulting position is approximately equal.

544

Conclusion This chapter has dealt with the important tabiya after 10.Rd1 Be7, which is the main line of the topical pawn sacrifice associated with the 5.Qc2 variation. We started by checking the harmless 11.Nc3, followed by 11.a3 which is more interesting – especially since the subsequent choice between 15.exf4 and 15.gxf4 has such an impact on the type of position that ensues. The majority of this chapter was devoted to the coverage of two queen moves, 11.Qf5 being the first. After the standard reply of 11...Nf6, play usually continues 12.e4 g6 13.Qf4 0-0 14.e5 Nh5, when the queen may move to either g4 or c4. In both cases, play becomes quite concrete, with analysis and correspondence games demonstrating that Black is okay in all variations. We then looked at 11.Qa4 Nf6, when the game may develop in various ways. Once again, the concrete nature of the game makes it hard to offer many general principles, but I would mention the ...f5 advance as an important theme in lines where White drives our knight back with e4-e5, plus the general concept of returning the extra pawn in order to free the black position. In the main line of 12.Nh4 0-0 13.Nf5, we even saw Black sacrificing the exchange, obtaining full compensation thanks to his superior activity.

545

A) 7.Bc3 266 B) 7.Nc3 d5 8.cxd5 exd5 267 B1) 9.Qc2 268 B2) 9.Bh3!? 269 C) 7.Bg2 0-0 8.0-0 d5 271 C1) 9.Bc3 272 C2) 9.Nc3 c6! 273 C21) 10.Rc1 274 C22) 10.Bf4 275 C3) 9.Ne5 Bb7 10.Nc3 Nbd7 276 C31) 11.Bf4 278 C32) 11.cxd5 281

546

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Ba6 5.b3 This continuation is by far White’s most popular choice. Indeed, compared to other options, none of White’s pieces is tied to the protection of the c4-pawn, and the c3-square is vacant for the knight. However, there are also drawbacks behind it: White’s last move failed to develop a piece, and it slightly weakened the c3-square.

5...Bb4† Out of several playable moves, I favour this one, which happens to be the most popular. 6.Bd2 This is the only sensible reply. 6.Nbd2? allows 6...Bc3! 7.Rb1 Bb7³ when White’s natural development is interrupted and he risks inviting serious troubles.

547

For example, I was surprised to discover that 8.Bb2 Ne4! 9.Rg1 Qf6! has even occurred at GM level; White’s position here is already lost! 6...Be7 The point of this standard manoeuvre is that White’s bishop is worse on d2 than on c1, since it blocks the d-file and will require an additional tempo to relocate to b2. We have reached the first branching point of the chapter, where we should consider A) 7.Bc3 and B) 7.Nc3, before analysing the normal C) 7.Bg2. 7.Qc2 is met by 7...d5 when White will almost certainly transpose to normal positions by playing Bg2 in the near future. A) 7.Bc3

548

This move prepares Nbd2, which is a standard regrouping plan, but White usually carries it out a bit later, after first developing his kingside. 7...d5 8.Nbd2 8.Bg2 0-0 9.0-0 is covered later in variation C1. 8.cxd5 exd5 will almost certainly lead to normal lines after a subsequent Bg2. 8...0-0 9.Ne5?! With this, White pushes the creative opening play a step too far. He should not neglect his development at this early stage of the game. 9.Bg2 is advisable, with a likely transposition to variation C1 after 9...Nbd7 10.0-0.

549

9...dxc4! 10.bxc4 10.Bg2 Nd5µ is obviously poor for White. 10...Bb7 11.Nef3 This isn’t a move White wants to make, but he hardly has anything better. 11...Nbd7 12.Bg2 Rc8 13.e3 c5 Black already had the upper hand in Bianshi – Bartsch, corr. 2010. B) 7.Nc3

550

This is a more serious alternative. Given the chance, White would like to carry out a quick e2-e4 advance. Despite the earlier g2-g3, the light-squared bishop may remain on the f1-a6 diagonal. 7...d5 The venom of White’s set-up was perfectly illustrated in the following top-level game: 7...c6 8.e4 d5 9.Qc2 dxe4 10.Nxe4 Bb7 11.Neg5!? c5 12.d5! exd5 13.cxd5 h6 14.Nxf7! Kxf7 15.0-0-0 White had full compensation for a piece in Topalov – Anand, Sofia 2005, and went on to win in great style. 7...0-0 is a respectable alternative, when 8.e4 d5 9.cxd5 Bxf1 10.Kxf1 exd5 11.e5 Ne4 has been tested in hundreds of games, with generally fine results for Black. However, since the ...d5 structure is at the heart of our repertoire in the main lines, it makes sense to aim for similar play (and a possible transposition) here. 8.cxd5 8.Bg2 0-0 9.0-0 leads straight to variation C2. 8.Rc1N has not yet been played here, but it is worth mentioning that 8...c6! is most accurate, preparing to recapture on d5 with the c-pawn. (8...0-0 is perfectly playable, and is likely to transpose to a later variation in our repertoire.) We will follow a game which arrived here via a different move order: 9.cxd5 cxd5 10.Bg2 0-0 11.Ne5?! This loses time. (11.0-0 is better, and transposes to 10.cxd5 cxd5 11.Rc1, as given in the notes to variation C2 on page 273.) 11...Nfd7 12.Nf3 Bb7 13.0-0 Nc6³ In the arising position with a symmetrical pawn structure, Black had the better-placed pieces in Lovkov – Grigoriants, St Petersburg 2006. 8...exd5

From this position, the great majority of games have continued with 9.Bg2, leading to normal paths. 551

For instance, 9...0-0 10.0-0 and we have reached the next chapter, where 10...Re8! will be our choice. However, there are a few independent possibilities, so we will consider B1) 9.Qc2 and B2) 9.Bh3!?. 9.Rc1 0-0 10.Bh3!? is covered under the 9.Bh3!? move order below. B1) 9.Qc2 0-0 10.Bg5!? 10.Bh3 Re8 11.0-0 c5 transposes to 11.Qc2N 11...c5 in the notes to variation B2 below.

White’s play over the past few moves may appear somewhat artificial, but this scheme has been tried by such strong players as Eljanov, Piket, Avrukh and more. In most cases, White intends to play e2-e3 to exchange the light-squared bishops, and he does not mind castling artificially by means of Kxf1 and Kg2. After careful consideration, I came up with a new idea for Black. 10...c5!?N 11.e3 This is consistent with White’s earlier play. In the event of 11.Bg2?! Nc6 12.Rd1 h6 13.Be3 Rc8 14.0-0 Ne4µ the lack of harmony in White’s camp starts to tell. 11...Bb7 11...Bxf1 12.Kxf1 Nc6 is perfectly playable, but I see no reason to facilitate White’s development. 12.Bg2 Nc6

552

13.0-0 13.dxc5 runs into 13...d4! 14.Rd1 bxc5 15.0-0 Ba6 16.Rfe1 Nb4 17.Qb1 Bd3µ and White cannot avoid a loss of material. 13...cxd4 14.Nxd4 Nxd4 15.exd4 h6 16.Bf4 Rc8= Black has no problems. B2) 9.Bh3!?

There are some merits behind this way of development. In particular, Black will not be able to put a rook on c8. On the other hand, losing control over the long diagonal is a concession. 553

9...0-0 10.0-0 In the event of 10.Rc1N Black can play 10...Re8, when 11.0-0 transposes to the main line below, or he could consider 10...c5!? 11.0-0 Nc6, which has been tested successfully in a couple of high-level games. 10...Re8 This will be our choice in the more popular position with the bishop on g2, and it works well here too. 11.Rc1 The most natural, but I also checked two other moves: 11.Re1 c5 12.Bg5 Nc6 13.dxc5 13.Rc1 leads straight to our main line below. In Genov – Mladenov, Sofia 2010, Black recaptured with the bishop. My analysis indicates that a better option is:

13...bxc5!N 14.Rc1 Nb4! As we will see, Black’s piece activity fully compensates for the vulnerability of the hanging pawns. 15.a3 15.Qd2 Ne4 16.Bxe7 Qxe7 17.Nxe4 dxe4 18.Nh4 e3 19.Qxe3 Qxe3 20.fxe3 Nxa2 21.Rxc5 Rxe3 results in equality. 15...d4 16.Na4

554

16...Na2! 17.Rc2 Nc3 18.Nxc3 dxc3 19.Ne5 Qxd1 20.Rxd1 Rad8„ Black has excellent play along the open d-file. 11.Qc2N This move is untested but is likely to transpose to a high-level game after: 11...c5 12.Be3 Nc6 We have reached the game in question, and will follow it briefly. 13.Rfd1 h6 14.Rac1 At this point I have found an important improvement:

14...Bf8!N 14...c4? was played in Ivanchuk – Lautier, Moscow (5.5) 2001, when 15.bxc4!N 15...Bxc4 16.Ne5! 555

would have been close to winning for White. The text move completes Black’s harmonious development, and his activity along the e-file is at least as strong as White’s pressure against the d5-pawn. A brief illustrative line is: 15.Qb2 Rxe3! 16.fxe3 Qe8 17.Qd2 Rd8 With ...Ne4 coming soon, Black has a dangerous initiative which more than compensates for the exchange.

11...c5 12.Re1 I also checked the more forcing 12.Bg5N 12...Nc6 13.dxc5 (13.Re1 transposes to the main line) but found that after 13...bxc5 14.Bxf6 Bxf6 15.Qxd5 Nd4! White achieves nothing, for instance:

16.Nxd4 cxd4 17.Qc6 Qb6 18.Nd5 Qxc6 19.Rxc6 Bxe2=

556

12...Nc6 13.Bg5 Nxd4 14.Nxd4 cxd4 15.Qxd4

15...h6 16.Bf4?! White should have settled for something like 16.Bxf6N 16...Bxf6 17.Qxd5 Bxc3 18.Qxd8 Raxd8 19.Rxc3 Rd2= or 16.Qa4!?N 16...hxg5 17.Qxa6 Bc5, although in both cases Black has no problems. 16...Bc5 17.Qa4 In Polak – Nayhebaver, Bojnice 2016, Black missed a powerful idea:

17...Bb7!N 18.b4 Bxf2†! 19.Kxf2 g5 20.Bd2 d4‚ With a venomous attack.

557

C) 7.Bg2

This is the main line of course. 7...0-0 If you are new to the Queen’s Indian you may assume that castling is an automatic decision, but in fact 7...c6 has been historically far more popular. The reasoning behind the latter move is that Black wants to play ...d5 and be ready to recapture with the c-pawn in the event that White exchanges on d5. Although that approach is solid and respectable, I am recommending what I consider a more interesting way of playing, involving ...d5 without a preliminary ...c6. The prevailing wisdom used to be that the ensuing ...exd5 structure would generally lead, after a subsequent ...c5, to the thematic hanging-pawns structure, which was believed to be positionally riskier for Black. However, modern practice and analytical tools indicate that Black is theoretically fine. Moreover, from a practical perspective Black tends to obtain plenty of activity and prospects for counterplay. By the way, the immediate 7...d5 is also possible and will almost certainly transpose, but I like the idea of castling and keeping White guessing about our intentions. 8.0-0 Other options require no special analysis: moves such as 8.Bc3 and 8.Nc3 are met by 8...d5, when White hardly has anything better than castling in the near future. 8.Ne5 d5 9.Nc3 (9.0-0 leads straight to variation C3) should be met by 9...c6! by analogy with variation C2 below. Again, White can hardly benefit from delaying castling for much longer. 8...d5 558

We have reached an important branching point. The big main line arises after 9.cxd5 exd5, and this will receive special coverage in the next chapter beginning on page 285. In the rest of this chapter we will analyse White’s various attempts to avoid the cxd5 exchange, or at least delay it until such a time that it does not simply transpose to the next chapter. The three main options are C1) 9.Bc3, C2) 9.Nc3 and C3) 9.Ne5. 9.Qc2 is easily met by: 9...c5! Since the e2-pawn is unprotected, this break is even more effective than usual. 10.dxc5 bxc5 11.Nc3 Nc6

12.Rad1 (12.Rfd1 occurred in Gebigke – Figura, Berlin 2009, when the same plan of 12...Rc8N 13.Bf4 Qa5„ would have been promising for Black) 12...Rc8 13.Bf4 Qa5 Black had excellent piece play in Wang Hao – Harikrishna, Danzhou 2016. 559

C1) 9.Bc3 White makes room for the knight to go to d2, where it will protect the c4-pawn – a standard method of regrouping, especially in the 7...c6 variation. In the present position though, Black is poised to carry out the ...c5 break without wasting a tempo on ...c6, so his counterplay comes with much more force. 9...Nbd7 10.Nbd2 In the event of 10.Ne5 Bb7 Black is ready for ...c5, with easy equality.

10...Rc8 10...c6?! makes little sense here, but it is worth mentioning that this move would transpose to a welltested line of the 7...c6 system, where Black has achieved solid results. Thus, it is obvious that in our version, not having to waste time on that pawn move, Black has excellent prospects. 10...c5 is a sensible move which has achieved good results, but somehow I prefer to bring the rook to a better position before launching our counterplay. 11.Re1 11.Rc1 occurred in Gauglitz – Gruenberg, Fürstenwalde 1981, when 11...Ba3!?N 12.Rc2 c5 13.Nb1 cxd4 14.Nxd4 Bc5= would have given Black easy play. 11...c5 12.Rc1 12.cxd5?! only helps to activate Black’s pieces, and 12...Nxd5 13.Bb2 cxd4 14.Bxd4 Bb4!³ was becoming unpleasant for White in Liu – Bitoon, Manila 2007.

560

12...cxd4 13.Bxd4 13.Nxd4N looks more natural to me, but a good reply is: 13...Ba3 14.Rc2 e5 15.Nb5 Bxb5 16.cxb5 Qe7 17.Nb1 Bc5„ Black’s strong pawn centre is at least as important as White’s bishop pair.

13...Ba3 14.Rc2 Qe7 15.cxd5 Rxc2 16.Qxc2 Rc8 17.Qb1 Nxd5 Black’s control over the c-file plus the weak c3-square caused White definite problems in Bouaziz – Farago, Lugano 1985. C2) 9.Nc3

9...c6! This seemingly modest move is perfectly timed! The resulting position has been reached many more 561

times via the 7...c6 move order, but here we should be happy to bring about the transposition, as the early Nc3 is well known to be ineffective against that scheme. White’s main problem is that he cannot comfortably protect the c4-pawn. It seems to me that White has two main attempts to justify his opening play: C21) 10.Rc1 and C22) 10.Bf4. The most common continuation has been 10.cxd5 cxd5, but it is toothless in the extreme, as the bishop on g2 does not have much of a role in the game. I don’t see the point in further detailed analysis; many moves are possible and you can simply play the middlegame with confidence, having effortlessly equalized from the opening. One example continued 11.Rc1 Nbd7 12.a4 Bb7 13.Bf4 a6 and Black had no problems in Christiansen – Adorjan, Linares 1985. 10.Qc2?!

The database shows a plus score for White after this move, but it seems to me that he is simply making a dubious pawn sacrifice. 10...dxc4 11.bxc4 11.Rfd1 Nbd7 12.e4 cxb3 13.axb3 Bb7 14.Bf4 a5 15.Na4 occurred in Liska – Pulpan, Kouty nad Desnou 2017. A natural improvement is 15...Qc8N 16.Rac1 c5 when Black succeeds in liberating the b7-bishop, leaving White with no real compensation. (Incidentally, Pulpan faced a similar scheme a few rounds later in the same tournament: the game and suggested improvement are mentioned below.) 11...Bxc4

562

12.Rad1 The attempt to regain the pawn by means of 12.Ne5 Ba6 13.Nxc6 Nxc6 14.Bxc6 runs into 14...Rc8 15.Qa4 Bb4! and Black takes over the initiative. 12...Nbd7 White had less than adequate compensation for the pawn in Naumkin – Estremera Panos, Saint Vincent 2000. C21) 10.Rc1

10...Nbd7! Unlike the note above, it’s not a good idea to grab the pawn here: 10...dxc4 11.bxc4 Bxc4 12.Ne5 Ba6 563

13.Nxc6 Nxc6 14.Bxc6² Machan – Pulpan, Kouty nad Desnou 2017. 11.Bf4 Rc8 Once again, the greedy 11...dxc4?! is not to be recommended: 12.bxc4 Bxc4 13.Nd2 Ba6 14.Bxc6 Rc8 15.Qa4ƒ Sygulski – Grabowski, Porabka 1986. 12.Qd3 12.Nd2?! restricts the mobility of White’s bishop, so after 12...Nh5 13.Be3 f5! Black was already better in Gofshtein – Tiviakov, Leiden 2010. Even though the queen may appear awkwardly placed on the same diagonal as the a6-bishop, the text move seems to be White’s best way of protecting the c4-pawn, as the queen also supports a future e2-e4 break. I found a new way for Black to combat this plan.

12...Ba3!?N Nudging the rook to a slightly worse square. 13.Rc2 Bb4!? By eyeing the knight on c3, Black prevents the e2-e4 advance. 13...c5 looks like a fully viable alternative. 14.Rd1 14.a3 Bxc3 15.Qxc3 c5 also offers plenty of counterplay. 14...Re8 15.a3 Bxc3 16.Qxc3 c5 564

Black’s active piece play fully compensates for White’s pair of bishops. C22) 10.Bf4

This is similar to the previous line, but here we see White trying to save time by doing without the Rc1 move. Once again, Black should ignore the c4-pawn for the time being and focus on completing development. 10...Nbd7 11.Qd3 11.Nd2 is not a stupid move although retreating the knight has its drawbacks. A good reply is: 11...Nh5 (11...Rc8 12.e4 c5 could also be investigated) 12.Be3 (12.e4? Nxf4 13.gxf4 Bb4 14.Qf3 Qh4µ left White with serious weaknesses in Benjamin – Farago, Philadelphia 1984) 12...Rc8 13.a4 f5!? Black had a pleasant version of a Stonewall structure in Li Chao – Wojtaszek, Huaian (blitz) 2016. 11...Rc8 12.e4 This is the natural and ambitious move, but after opening the centre the unfortunate location of White’s queen will start to tell. 12.Rac1 takes us back to variation C21 above. 12.Rfd1 c5 gives Black easy play, for instance: 13.Nb5 Bxb5 14.cxb5 cxd4

565

15.Rac1 (After 15.Nxd4? e5 16.Nc6 Rxc6 17.bxc6 Nc5 18.Qf5 exf4 Black was much better in Vyzmanavin – Illijin, Bern 1993) 15...Nc5 16.Qb1 Nce4 Black had excellent prospects in Chomicki – Wernikiewicz, email 2005.

12...dxc4!? This has only been played once but I find it the most interesting option. 13.bxc4 b5 14.cxb5 This move leads to complications which are not unfavourable to Black. I also examined: 14.Qc2N 14...bxc4 15.Rab1 Qa5 16.Rfd1 Rfd8 We have reached a complex position where White may be able to claim sufficient compensation for the pawn, but certainly not enough for an 566

advantage. 14...cxb5

15.e5! 15.Nxb5 is convincingly met by 15...Nc5! 16.Qb1 Ncxe4 when White suffers from a lack of harmony. 15...Qa5! 16.exf6 Rxc3 17.Qxc3! Qxc3 18.fxe7 The dangerous e7-pawn and the imminent activity of White’s rooks along the open c-file should not be underestimated. 18...Re8 White can maintain the balance with the following natural novelty. 19.Rac1N After 19.Bd6? Qc6 20.Bb4 Bb7 Black was winning in Miton – Nikolov, Plovdiv 2008. 19...Qa3 20.Rc7 Rxe7 21.Rxa7 Nf6 White can and should force a draw by repetition:

567

22.Bc1 Qd6 23.Bf4 Qa3= C3) 9.Ne5

This way of maintaining the tension is the most effective. White’s knight advance comes with great effect, as the pressure along the h1-a8 diagonal is significant now. 9...Bb7 9...c6 10.Bc3 Nfd7 11.Nxd7 Nxd7 12.Nd2 is a huge main line which usually arises via the 7...c6 move order. I would rather avoid this, as there is a massive amount of theory Black needs to know, and White takes little risk if he knows what he is doing. 568

10.Nc3 This is the most natural move, and the most challenging. 10.cxd5 was played in Shomoev – Predojevic, Moscow 2009. There is nothing wrong with 10...exd5, as played in the game, since White hardly has anything better then 11.Nc3, when 11...Re8 leads straight to variation G of the next chapter. However, 10...Bxd5!?N looks like a good way to exploit White’s unusual move order. Play may continue 11.Bc3 Bxg2 12.Kxg2 Qd5† 13.f3 Nbd7 14.e4 Qb7 and Black has no problems. 10.Qc2 has been played a few times, but I doubt that White can achieve anything by continuing to delay the development of the b1-knight. 10...c5 11.dxc5 This was Spalir – Bartos, Pardubice 2015, when 11...Bxc5N looks fine for Black, for instance:

12.cxd5 (I also considered 12.Nc3 Nbd7 13.Nd3 Bd4! 14.Rad1 dxc4 15.Bxb7 cxd3 16.Qxd3 Nc5 17.Qf3 Nxb7 18.Qxb7 Qc8= and Black has no problems) 12...Bxd5 13.e4 Bb7 14.Nc3 Nbd7 Black has excellent play; the weakness of the d4-spot might cause White problems in the long run. 10.Bc3 clears the d2-square for the knight. This works well in the lines with an early ...c6 but it is rather slow here, as Black has saved a tempo which can be used for counterattacking purposes. Out of a few attractive options, I favour the flexible 10...Qc8!?, protecting the bishop on b7 and vacating d8 for the rook. A good example continued:

569

11.Nd2 dxc4 12.Ndxc4 Bxg2 13.Kxg2 Qb7† 14.Kg1 (14.f3N is hardly an improvement in view of 14...Nd5 15.Bd2 Rd8 16.Qc2 Na6 and ...c5 is coming.) 14...Na6 15.Rc1 Rac8 16.Qd3 In Montecatine Rios – Kudrin, Marchena 1990, the most logical continuation would have been:

16...Rfd8N 17.Rfd1 c5 Black has excellent play. 10...Nbd7 White is at a crossroads. If he wishes to maintain the central tension, then C31) 11.Bf4 is his most important option. The alternative approach is to clarify the central structure with C32) 11.cxd5. 11.Nxd7?! is a clear concession, and after 11...Qxd7 12.Bc1 Rfd8 13.cxd5 Nxd5 Black was already better in Kelecevic – Tukmakov, Winterthur 2002.

570

11.Rc1 is reasonable, but after 11...c5 White has nothing better than transposing to variation C31 with 12.Bf4. I also checked 12.Bg5N but found it to be toothless after: 12...cxd4 13.Nxd7 Qxd7 14.Qxd4 Rfd8 15.cxd5 Nxd5 16.Bxe7 Qxe7= Finally, a rare but sensible alternative is: 11.e3!? Finding a good home for the queen is a challenge for White in these positions, so vacating the e2square is quite logical. However, it doesn’t pose Black any problems. 11...c5 12.Qe2 Qc7 13.f4 Rfd8 14.cxd5 Nxd5 15.Ne4? More to the point was 15.Nb5N 15...Qb8 16.Rac1 a6 17.Nc3, maintaining the balance.

15...f5! With this excellent move, Black takes control over the light squares. Now the mobility of White’s minor pieces is severely restricted. 16.Nf2 cxd4 17.exd4 Nxe5 18.Rac1 Qd7 19.fxe5 Ba3 Black enjoyed a clear positional advantage in Sek – Bocharov, Vladivostok 2016. C31) 11.Bf4

571

Now all of White’s minor pieces are actively placed, exerting pressure on the centre and queenside. 11...c5 This counterattacking move should not be delayed. 12.Rc1!? Maintaining the tension is the most complex continuation. Obviously we should consider both of the possible pawn exchanges as well: 12.dxc5 Nxc5 13.cxd5 exd5 The appearance of an isolated d-pawn is not such a big achievement for White, as he does not have much control over the key d4-square. 14.Rc1 Worth attention is 14.Nb5!?N, but also then after 14...Ne6 15.Be3 Re8 16.Nd4 Bd6 17.Nef3 Ba6 18.Rc1 Rc8„ Black’s pieces are very active. 14...Ne6 15.Nb5 In Kanep – Dubrovin, Tallinn 2006, Black should have continued:

572

15...Bc5!N 16.Nf3 Nxf4 17.gxf4 Ne4„ Black is doing fine in this double-edged position. 12.cxd5 The timing of this move enables Black to trade some pieces in the centre. 12...Nxe5 13.Bxe5 After 13.dxe5?! Nxd5 14.Qc2 Nb4 15.Qb1 Bxg2 16.Kxg2 f5 Black enjoyed the better pawn structure in Davidov – De Souza, corr. 2007. 13...Nxd5 14.Nxd5 Bxd5 15.e4 Bb7

16.Qg4 Another natural try is 16.d5, as seen in Hamid – Arkell, Plymouth 1992. My suggested 573

improvement is 16...Bd6!N 17.Bb2 exd5 18.exd5 Qd7 19.Re1 Rae8= when Black is fine; the d5pawn is blockaded, and it may become vulnerable in the future. 16...f6 17.Qxe6† Kh8 18.Bf4 cxd4 19.Rfc1 We have been following Yusupov – A. Sokolov, Riga (8) 1986. I think Black should have neutralized his opponent’s activity along the c-file by means of:

19...Rc8N 20.Rxc8 Bxc8 21.Qc4 Bg4 22.h3 Bh5„ The strong passed d-pawn offers Black a full share of the chances. 12...cxd4 13.Qxd4 Bc5 14.Qd1!? 14.Qd2 is the more obvious retreat, when 14...Nxe5 15.Bxe5 Ng4 16.Qf4 occurred in Soreghy – Krueger, corr. 2007. Here Black missed a nice idea:

574

16...f5!N I like the aggressive spirit of this concept. It is fully justified by the lack of coordination among White’s pieces. 17.Bc7 This is the only way to avoid a serious loss of material. Nevertheless, the bishop remains quite vulnerable. 17...Qf6 18.cxd5 e5 19.Qd2 f4© Black’s kingside initiative is worth at least as much as the sacrificed pawn. The text move was played in Kuehne – E. Popov, corr. 2011. My suggestion for Black is:

14...Nh5!?N 15.Nd3 Naturally, I also considered: 15.Nxd7 Qxd7 16.Be5 (the greedy 16.cxd5?! Nxf4 17.gxf4 Rad8 offers Black a dangerous initiative for the pawn) 16...d4 17.Bxb7 Qxb7 18.Bxd4 Rad8 19.e3 Nf6

The pin along the d-file leaves White with no choice but to move his queen away, and after 20.Qe2 Bxd4 21.exd4 Rxd4= Black regains the pawn with a balanced game. 575

15...Ba3! This is an important detail, as it turns out that each possible rook move comes with a certain drawback. 16.Rb1 16.Rc2 allows 16...dxc4! 17.Bxb7 cxd3 18.Qxd3 Nxf4 19.gxf4 Rb8= when Black is fine. 16...Qc8! The point of Black’s play is that the pin along the c-file gives him time to consolidate. My main line continues:

17.Bc1 Bxc1 18.Rxc1 dxc4 19.Ne4 Qb8 20.Rxc4 Bd5 21.Rc1 Nhf6= 576

Black has no problems. C32) 11.cxd5 exd5

The position is similar to the 9.cxd5 exd5 main lines, which we analyse in the next chapter. Compared with variation G on page 300, the only difference is that here Black has played ...Nbd7 instead of ...Re8. Against several of White’s possible continuations, Black can simply play ...Re8 and transpose to one of the lines in the next chapter. Naturally, we will focus on any independent possibilities in this section. 12.Nd3 If White wishes to avoid transposing to the next chapter, then this is his best option. 12.Qc2 c5! 13.e3 can hardly threaten Black. 13...Re8N looks like a good reply, although there was also nothing wrong with 13...Rc8 in Hoem – Salomonsen, Bergen 2005. 12.Bg5 should be met by 12...Re8, reaching variation G2 of the next chapter. I would like to mention in passing that 12...h6?! should be avoided in view of 13.Bf4! when the extra ...h6 move weakens Black’s kingside. For instance: 13...Re8 14.Rc1 Nf8 15.Bh3 Ba3 16.Rc2 a6 17.Nb1 Bd6 18.Nc6 Bxc6 19.Rxc6² Black was under pressure in Fridman – Chuchelov, Belgium 2008. 12.Bf4 Re8 both lead to the next chapter; see variations G3 and G5 respectively, on pages 303 and 306. Finally, 12.f4?! c5 13.e3 was overambitious by White in Chatterjee – Chumfwa, Khanty-Mansiysk (ol) 2010. Black should have fixed the pawn structure by means of:

577

13...cxd4!N 14.exd4 Ne4 15.Be1 Ndf6³ With full control over the e4-outpost, Black has the better prospects.

12...a5! White’s last move introduced the positional threat of b3-b4, so it is important for Black to respond accordingly. 12...c5?! proved to be a mistake after 13.Be3! Ne4 14.Rc1 and the pressure on the d5-pawn was too strong in Kadlec – Vavrla, Czech Republic 1999. The merits of White’s flexible set-up were well illustrated in the following encounter: 12...Re8 13.b4! Now Black’s activity in the centre is severely limited. 13...c6 14.Qb3 Nf8 15.Bg5 Ne6 16.Bxf6 Bxf6 578

17.e3² Black was doomed to passive defence in Aseev – Golovchenko, St Petersburg 2003. 13.Rc1N This move has not been tested, but it’s clearly a logical choice and will shortly transpose to a high-level game. I checked three other continuations: 13.Nf4N 13...c6 14.Re1 Re8 15.Qc2 Rc8 16.Rad1 Bb4 leads to a complex positional struggle, where Black’s chances are not worse. 13.a3N This new move immediately transposes to an older game. 13...Ne4 Once again, Black fights against the squeezing b3-b4 move. 14.Be1 14.Qc2N is well met by: 14...Bf6! 15.Nf4 (the inaccurate 15.e3?! runs into 15...c5 16.dxc5 Rc8!ƒ) 15...Bxd4 16.Rad1 Nxd2 17.Rxd2 Bxc3 18.Qxc3 Nf6 19.e4 c6 White can regain the pawn, but he will not end up with more than equality.

14...Ndf6!N This is the best way to strengthen Black’s position. 14...c5?! is risky. 15.Nf4 Ndf6 16.dxc5 Nxc3 17.Bxc3 bxc5 18.Bxf6 Bxf6 occurred in Horvitz – Browne, Chicago 1988, and now 19.Rb1!N± would have enabled White to pick up the d5-pawn for no compensation. 15.e3 15.b4 axb4 16.axb4 Qd7„ Black is poised to take over the a-file. 15...c5 Black has a pleasant position, and is practically a tempo up on the Horvitz – Browne game. 579

A somewhat more recent correspondence game continued: 13.Qc2 Re8

14.Rad1 I also considered: 14.a3 c5! This strike against the d4-pawn is well timed; the d5-pawn is not too vulnerable and White’s queenside has been slightly loosened by his last move. 15.dxc5 bxc5 16.Nf4 Nb6 17.Rfd1 Qc8„ 14...Rc8 15.Nf4 c6 16.Rfe1 White prepares for aggressive actions in the centre. 16...Bf8 17.Bc1 h6 18.e4 All this occurred in Rydholm – Nyvlt, corr. 2009. It seems to me that Black’s most natural reaction would be:

580

18...dxe4N 19.Nxe4 Nxe4 20.Rxe4 Rxe4 21.Bxe4 Nf6 22.Bg2 Nd5= The pressure along the long diagonal is neutralized, so Black has nothing to worry about.

13...Re8 With this move, we transpose to the aforementioned high-level game, which we will follow for a few more moves. 14.Bf4 c6 By overprotecting the d5-pawn, Black enables the knight on f6 to move. 15.Rc2 Nf8 16.Bg5 It’s not easy to suggest a better option than exchanging the problematic bishop. 16...Ne6 17.Bxf6 Bxf6 18.e3 Here I have a new idea to improve on Gelfand – Karjakin, Beijing 2013.

581

18...Be7!?N The idea is to take control over the dark squares on the queenside and make it more difficult for White to advance the b-pawn. 19.Qc1 19.Na4 Bd6 20.Qd2 is hardly an improvement; after 20...Rc8 White runs out of constructive ideas. 19...Bd6 20.Na4 Rc8 21.Rd1

21...g6 22.Qb2 h5!? Intending kingside action with ...h4. A complex struggle lies ahead, but Black’s chances are by no 582

means worse. Conclusion This chapter introduced the important topic of 4.g3 Ba6 5.b3, when 5...Bb4† 6.Bd2 Be7 is our choice. For a good few decades, the plan of ...c6 followed by ...d5 was widely regarded as Black’s most reliable follow-up, but more games and analysis have convinced me that Black gets a good game by playing a quick ...d5 without the preliminary ...c6. We started by analysing some sidelines, of which 7.Nc3 d5 8.cxd5 exd5 9.Bh3!? is the most interesting. Black’s chances are not worse, but it is important to be aware of the pros and cons of the bishop on h3 and how the ensuing play will differ from the standard lines with the bishop on g2. We then moved on to the more popular 7.Bg2 0-0 8.0-0 d5, when another branching occurs. It is important to remember that after 9.Nc3 we should revert to a more traditional set-up with 9...c6!, when the positioning of the knight on c3 makes it harder for White to defend the c4-pawn. True, against both 10.Rc1 and 10.Bf4 Black should refrain from capturing on c4 too quickly; but the fact remains that White will have to make some kind of concession to safeguard the pawn at some point. In the last part of the chapter we considered the active option of 9.Ne5, when 9...Bb7 10.Nc3 Nbd7 should be played. Transpositions to the next chapter are possible, but we focused on independent possibilities after both 11.Bf4 and 11.cxd5. Both options lead to lively positions with chances for both sides, where Black’s prospects are not worse.

583

A) 11.Bf4 287 B) 11.a3 288 C) 11.Rb1 290 D) 11.Rc1 292 E) 11.Re1 293 F) 11.Qc2 c5!? 295 F1) 12.dxc5 296 F2) 12.Rfd1 299 G) 11.Ne5 Bb7 300 G1) 12.Bc1 300 G2) 12.Bg5 302 G3) 12.Bf4 303 G4) 12.Qc2 304 G5) 12.Rc1 Nbd7 306 G51) 13.Nb5 307

584

G52) 13.Rc2 309 G53) 13.Bf4 310 1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Ba6 5.b3 Bb4† 6.Bd2 Be7 7.Bg2 0-0 8.0-0 d5 9.cxd5 This is the most popular choice. White releases the central tension in a position where Black does best to recapture with the e-pawn. 9...exd5 9...Nxd5 is playable and fairly solid, but Black’s counterattacking abilities are limited and his practical results have not been so good. The text move brings us to an important position. For any readers who are fearful of hanging pawns, let me reassure you that, in those lines where we follow up with ...c5, Black’s active piece play will offer plenty of counterplay.

10.Nc3 This is the most natural and common move: White develops his last minor piece and improves his central control. The resulting position has been reached in over a thousand games on my database, taking into account transpositions involving an earlier Nc3. 10.Ne5 Re8 hardly leaves White with anything better than 11.Nc3, which transposes to variation G. The only serious independent alternative is: 10.Bc3 As usual, this move is aimed at overprotecting the central dark squares. The disadvantages are obvious: White’s bishop moves to a presently blocked diagonal, and his knight will be placed less actively in the event of Nbd2. 585

10...Re8

11.Bb2 11.Qc2?! was a less accurate choice in Burmakin – S. Zhigalko, St Petersburg 2013, when Black missed an interesting opportunity: 11...c5!N 12.dxc5 (after 12.Rd1 Nc6 13.e3 Rc8ƒ White’s lack of harmony starts to tell) 12...Nc6! 13.cxb6 Qxb6 With a powerful initiative for the pawn. 11.Ne5 Bd6 has also been played, when 12.Nd2N is best. (After the slow 12.Nd3?! Nbd7 13.Nd2 Ne4³ Black had the better game in Laruelle – Latzke, Cannes 1997.) 12...Nbd7 13.Ndf3 c5 14.Bb2 Qe7 15.Rc1 cxd4 16.Nxd7 Qxd7 17.Nxd4 Black has plenty of activity to compensate for the isolated pawn after 17...Be5 or even 17...h5!?. 11...Nbd7 12.a3 12.Nc3N is well met by 12...Bb4! 13.Re1 Ne4 14.Qc2 Rc8 intending ...c5, with excellent counterplay. 12...Ne4 13.Nbd2 This was seen in Ponomarev – T. Schmidt, corr. 2004. I recommend the following way of regrouping the pieces:

586

13...Bd6N 14.Re1 After 14.Nxe4 dxe4 15.Nd2 Nf6 I prefer Black, who has more space and good attacking prospects on the kingside. 14...Qe7 15.e3 f5 16.b4 c6 17.Rc1 Bb7 Black is at least equal. The e4-knight is annoying for White, whereas the e5-square is well guarded by Black’s pieces. 10...Re8! 10...Bb7 has been significantly more popular but there is no need to hurry with the bishop retreat, as the pressure along the a6-f1 diagonal can still bother White in some lines.

The present position has been known since the 1980s, but it has risen to prominence in recent years as 587

White has struggled to find any advantage. Most notably, Sergey Karjakin played it four times in the 2016 Candidates tournament, achieving a 100% success rate (meaning he drew all four games as Black). We are at an important crossroads, where White has a multitude of plans at his disposal. Black’s counterplay is based on attacking the d4-pawn by means of a timely ...c5 break. This usually requires some preparation, such as transferring the b8-knight to e6, but we will discuss this and other nuances as we work our way through the different variations. We will consider A) 11.Bf4, B) 11.a3, C) 11.Rb1, D) 11.Rc1, E) 11.Re1, F) 11.Qc2 and G) 11.Ne5. 11.Bg5 Nbd7 12.Ne5 Bb7 leads to variation G2 on page 302. A) 11.Bf4 This has only rarely been seen, but it’s a sensible and thematic move. 11...Nbd7 Manoeuvring the knight to e6 is a thematic plan, and the presence of the bishop on f4 makes it all the more appealing. 12.Qc2N This novelty immediately transposes to a Gelfand game, so it should be taken seriously. Other logical moves tend to transpose elsewhere, for example: 12.Ne5 Bb7 leads to variation G3 on page 303. 12.Rc1 Nf8 13.Ne5 Bb7 is variation G53 on page 310. 12...Nf8 13.Rfd1 Ne6 14.Be5 Bb7 15.a3 After a series of natural moves, Black found a nice idea.

588

15...Ne4! 16.Nxe4!?N This seems like a logical attempt to improve White’s play. The game continued: 16.b4 Nxc3 17.Qxc3 c6 18.Rab1 a5! Opening the a-file at the perfect moment. 19.Qc2 (19.b5?! cxb5 20.Rxb5 Ba6µ) 19...axb4 20.axb4 Ba6³ Black had an excellent position in Gelfand – Martirosyan, Moscow 2016. 16...dxe4 17.Nd2 f6 18.d5! The critical move, without which White’s play would make no sense. 18...Qxd5 19.Bxe4 The arising complications might appear promising for White, but we have an excellent rejoinder.

589

19...Qc5! 20.Qxc5 It would be risky to grab a pawn with 20.Bxh7† Kh8 21.Bc3 Ng5 22.Bf5 Bd6, when the vulnerability of White’s king starts to tell. 20...Nxc5 21.Bxb7 Nxb7 22.Bxc7 Bc5! Black liquidates into a comfortable endgame. For instance:

23.e4 Rac8 24.Bf4 g5 25.Be3 Bxe3 26.fxe3 Nc5= Any benefit of White’s token extra pawn is counterbalanced by his inferior structure. B) 11.a3 590

White not only covers the b4-square, but also intends to expand on the queenside by pushing the bpawn. 11...Nbd7! The thematic 11...c5 does not completely equalize after: 12.dxc5 bxc5 13.Ne5 Bb7 14.Rc1! (improving on 14.Bf4 as played in Caruana – Karjakin, Moscow 2016) 14...Na6 (I also examined: 14...Nbd7N 15.Nxd7 Qxd7 16.Bg5 d4 17.Bxb7 Qxb7 18.Bxf6 Bxf6 19.Na4²) 15.Bg5!? (15.Nd3 was also reasonable in Cheparinov – Navara, Minsk 2017) 15...Rb8 (15...h6 16.Bxf6 Bxf6 17.Nd3² puts the hanging pawns under strong pressure) Black went on to hold a draw in Shuler – Palmateer, corr. 2015. In general his position is playable after 11...c5, but I prefer to avoid these positions where White can exert slight pressure without much risk. 12.b4 The other ambitious try is: 12.Ne5 Bb7 13.b4 Bf8 14.Nd3 (White can maintain the centralized knight, but only at the cost of making positional concessions: 14.f4 c5 15.e3 cxd4 16.exd4 Rc8„)

591

14...Ne4 15.Rc1 a5! Well timed, since White cannot play b4-b5 due to the loose pawn on a3. 16.Bf4 c6 Black had excellent play in Sychev – Neiksans, Minsk 2017. The text move is most consistent with the previous a2-a3. Here I found a natural improvement for Black:

12...c6!N This move prevents b4-b5 and overprotects the d5-pawn, thus making it possible for the f6-knight to move if necessary. 12...Bb7 is too passive, and after 13.Qb3 c6 14.Rfe1 a5 15.bxa5 Rxa5 16.e4 dxe4 17.Ng5 White had a lot of activity in Cheparinov – Debashis, Doha 2016.

592

12...Bc4?! is premature in view of 13.Bf4 c6 14.Nd2, chasing the bishop away. 13.Ne5 This seems most challenging. The appearance of White’s knight on this square usually makes the position tenser, and this is no exception. After 13.Re1 Bc4 14.Qc2 b5 15.e4 Nxe4 16.Nxe4 dxe4 17.Rxe4 Bd5= Black has full control over the light squares. 13...Nxe5 14.dxe5 Ng4 15.f4 Bc4

16.h3 Also after 16.Re1 a5! 17.h3 Nh6 18.bxa5 Rxa5 Black’s counterplay on the queenside makes his position rather attractive. 16...Nh6 17.g4 a5! The poor knight may be stuck on h6 for a while, but the other pieces have targets on the queenside. 18.Rb1 axb4 19.axb4

593

19...f5 20.g5 Nf7„ Black has blocked the kingside and begun the process of improving his knight, leading to at least equal chances. C) 11.Rb1

Just as in the previous line, White is preparing to advance his b-pawn. Karjakin faced this move twice in the 2016 Candidates tournament. 11...Nbd7! Once again, Black should develop his last minor piece before taking concrete action. The tournament 594

winner opted for this move the second time around. We will follow his game against Nakamura, who must surely have considered the text move in his preparation. The earlier game instead saw 11...c5?! 12.dxc5 bxc5 13.Ne5 Bb7 14.Bf4 Bf8 15.Rb2 and Black was under definite pressure in Topalov – Karjakin, Moscow 2016. 12.b4 White has seized some space but his strategy is not without drawbacks. 12...Bc4! Black takes control over some light squares on the queenside now. Unlike the previous variation, it is not necessary to prepare the bishop move with ...c6. 13.Bf4 The other natural move is 13.Qc2!?N, which should be energetically met by: 13...c5! (less convincing is 13...b5 14.Bf4 Bd6 15.Bxd6 cxd6 16.Nd2! Bxe2 17.Nxe2 Rxe2 18.Qd3 and Black’s pawns are rather weak) 14.bxc5 bxc5 15.Be3 Rc8 16.Bh3 cxd4 17.Nxd4 Bc5

The activity of Black’s pieces fully compensates for the isolated d-pawn. An illustrative line is: 18.Rfd1 Rxe3! This positional sacrifice highlights the dynamic potential of Black’s set-up. 19.fxe3 Qe8 20.Qd2 Rd8© With full compensation for the exchange.

595

13...Ne4! This strong move invites complications which are not unfavourable to Black. The justification comes from the vulnerability of the a2-pawn, which explains why we could not play the same way in the 11.a3 line above. 13...c6?! is inaccurate in view of 14.Nd2ƒ. 14.Nxe4 The modest 14.Qc2N 14...Nxc3 15.Qxc3 a5 16.a3 axb4 17.axb4 c6 offers White no advantage, to say the least. 14...dxe4 15.Nd2 15.Ne5?! Nxe5 16.dxe5 Qxd1 17.Rfxd1 Bxe2 18.Rd7 Bd3 19.Rc1 c5³ leaves White without full compensation for the pawn. 15...Bxa2 16.Bxe4 16.Rb2?! Bd5 17.Nxe4 Bxb4!³ works out well for Black. 16...Bxb1 17.Qxb1 The arising position may appear promising for White, as his bishops are powerful and two of Black’s pieces are under attack. However, things turn out fine after:

596

17...Nf6! 18.Bxa8 Qxa8 Regaining the exchange is no great achievement. 19.e4 19.Nf3 Bf8 20.Ne5 was hardly an improvement in Michalik – Navara, Prague (blitz) 2016. My suggestion is:

20...c5N 21.bxc5 bxc5 22.dxc5 Bxc5 Black has no problems. 19...Rd8 20.Be3 597

20...a5!?N An interesting new idea, although I must stress there is nothing at all wrong with 20...Ng4, which led to an eventual draw in Nakamura – Karjakin, Moscow 2016. 21.bxa5 bxa5 22.Qc2 a4 23.d5 a3 24.Ra1 Qb7„ The passed a-pawn compensates for White’s strong centre, and the position remains dynamically balanced. D) 11.Rc1

This is quite a popular move, although a lot of the time it transposes to a later variation after a knight 598

hop to e5. 11...Nbd7 12.Rc2 The more common 12.Ne5 Bb7 leads straight to variation G5. 12.Bf4 does not have much independent value after: 12...Nf8 (12...c6!? is another attractive option. 13.Ne1 occurred in Blagojevic – A. Kovacevic, Cetinje 2013, when 13...Nf8N 14.Rc2 Rc8 would have offered Black excellent play.) 13.Ne5 Bb7 We have transposed to variation G53 on page 310. 12...Nf8 13.Ne5

This position was reached in Bluebaum – Meskovs, Germany 2017. Rather than retreating the bishop to b7, as occurred in the game, Black can simply allow the knight foray to c6: 13...Ne6!N 14.Bc1 14.Nc6 Qd7 15.Nxe7† Qxe7 16.Be3 Rad8 looks entirely comfortable for Black. 14...Rc8 15.Nc6 15.Bb2 c5 16.dxc5 d4 17.Na4 bxc5„ gives us plenty of activity. 15...Qd7 16.Nxe7† Qxe7! Exchanging the dark-squared bishop certainly counts as an achievement for White, but it has cost him several tempos. Black is fully mobilized and ready to begin active operations in the centre. Incidentally, his last move offers a pawn sacrifice which I think White does best to decline.

599

17.Bb2 17.Nxd5 Nxd5 18.Bxd5 Rcd8 19.Bxe6 Qxe6 looks shaky for White, whose king is exposed. 17...c5 18.Nxd5 Nxd5 19.Bxd5 Nxd4 20.Bxd4 cxd4 Black has no problems. E) 11.Re1

This modest-looking move makes no attempt to refute Black’s set-up, but it has been employed by some strong players including Ding Liren, so we should consider it seriously.

600

11...Nbd7! I prefer this over the much more common 11...Bb7, which allowed White to successfully regroup his pieces by means of 12.Bc1! Na6 13.Bb2 c5 14.Rc1 when the d5-pawn was under pressure in Ding Liren – Harikrishna, Tsaghkadzor 2015. 12.Ne5 In comparison to the note above, now 12.Bc1N is well met by 12...Bb4! 13.Bb2 Ne4 14.Rc1 Ndf6 with excellent counterplay. For example: 15.Ne5

15...Nxf2!? This is by no means a necessity, but it’s a convincing route to a draw. (15...c5 16.e3 Rc8„ is a sensible way to keep more tension in the position.) 16.Kxf2 Rxe5 17.dxe5 Ng4† 18.Kg1 Bc5† 19.e3 Nxe3 20.Qf3 Ng4† 21.Kh1 Nf2† With a draw by perpetual. 12.e4!?N This has not yet been tested, but it’s a natural follow-up to White’s last move. 12...dxe4 12...Nxe4 gives White the useful option of 13.Nxd5.

601

13.Ng5 13.Nxe4 Nxe4 14.Rxe4 Bb7 15.Re1 Nf6 16.Rc1 h6 17.Bf4 c5= is entirely harmless. I also considered 13.Ne5 Bb7 14.Qc2 c5 15.Nxe4 Nxe5 16.dxe5 Nxe4 17.Bxe4 Bxe4 18.Rxe4 Qd5= and Black is fine. 13...Rc8 14.Ngxe4 14.Bh3 Bb7 15.Ngxe4 Rb8= also leads nowhere for White. 14...Nxe4 15.Nxe4 Nf6 The position remains about equal. 12...Bb7 The weakness of the d5-pawn and c6-square necessitated this retreat. However, with the knight already on e5, White is in no position to carry out Ding Liren’s regrouping.

602

13.Rc1!?N This untested move seems like the most natural choice to me. Others are less challenging: 13.f4 is positionally risky, and after 13...c5 14.e3 Nf8 15.Ne2 Ne4 Black was already making good use of the e4-outpost in S. Farago – B. Szabo, Budapest 2012. 13.Qc2 c5 was fine for Black in Agzamov – Kuzmin, Tashkent 1980. 13.Nxd7 releases the tension prematurely, and after 13...Qxd7 14.Rc1 h6 15.Bf4 Rac8 Black had comfortable play in Elarbi – Supatashvili, Al Ain 2013. As usual, the placement of White’s bishop on d2 instead of b2 hampers his coordination. 13...c6 I recommend this solid approach, although other moves are certainly playable. Despite the novelty on the last move, we have now transposed to a correspondence game. It is worth pointing out that 13...Ba3 can be met by 14.Rc2, since the greedy 14...Nxe5 15.dxe5 Rxe5?! invites trouble: 16.Nb5 Bc5 17.b4ƒ

603

14.e4!?N This seems like the most obvious move to consider. The game proceeded more slowly with: 14.Nd3 Nf8 15.Bg5 Ba3 16.Rc2 Ne6 17.Bxf6 Qxf6 18.e3 At this point my preferred continuation would be:

18...Qe7!?N The game continued with 18...Nc7 and Black eventually prevailed in Shabaev – Mauro, corr. 2014. However, the text move feels like a more logical choice for a human. 19.Qd2 a5 Black’s idea is to take control over the queenside dark squares and limit White’s activity there. 604

20.Na4 Qd8 21.Rd1 Rc8 Overprotecting the c6-pawn, so that the light-squared bishop can be placed on a6. Black is doing fine. The text move appears active and logical, but Black solves all his problems after: 14...Ba3! 15.Rc2 Nxe5 16.dxe5 Rxe5 17.Bf4 Re8

18.exd5 Rxe1† 19.Qxe1 cxd5 20.Bg5 The pin along the long diagonal enables White to regain the pawn with equal chances, but nothing more. F) 11.Qc2

605

This queen move vacates the d1-square for the rooks. Later, after the queen’s rook has moved to c1 or d1, the queen may relocate to b1 (or possibly b2) to support the b3-b4 advance. Occasionally the queen might even go to the more exotic f5-square. Rather than wait to see which of these plans White chooses, I recommend immediate counterplay with: 11...c5!? 11...Nbd7 is a reasonable move which has scored well, but I prefer to carry out the pawn break immediately. One reason is that I want to develop the knight to c6 rather than d7, as the queen on c2 may be hit by a timely ...Nb4 or ...Nd4. White’s two main responses are F1) 12.dxc5 and F2) 12.Rfd1. 12.Rad1 seems less logical, as it is not clear what the other rook will achieve on the kingside. After the further 12...Nc6 13.Be3 (13.dxc5 is well met by 13...d4! 14.Ng5 Nb4 15.Qb1 h6 16.Nge4 bxc5³) 13...Rc8 14.Qb2 h6 15.Rfe1 Bf8 Black had the more harmonious position in Potkin – Navara, Jerusalem 2015. F1) 12.dxc5 bxc5

606

13.Ne5 Another concrete approach to the position is: 13.Ng5 g6 White has provoked a slight weakening of our kingside. However, the lack of harmony in White’s camp makes it easier for Black to develop counterplay in the centre. 14.Be3!? After 14.Rad1 Nc6 15.Rfe1 Rc8 16.Qb1 d4 17.Na4 Nd5 White was doomed to passive defence in Raykin – Dothan, corr. 2003.

14...Nc6 15.Qd2 15.Rad1N is well met by 15...Rc8! when the critical line continues: 16.Nxd5 Nd4! 17.Nxe7† Rxe7 18.Bxd4 cxd4 19.Qb2 Rxe2 20.Qxd4 Qxd4 21.Rxd4 Rxa2 Black is in no way worse. 607

15...h6 16.Nh3 White gets no advantage after: 16.Nxd5N 16...hxg5 17.Nxf6† Bxf6 18.Bxc6 Qxd2 19.Bxd2 Rxe2 20.Bxa8 Bxa1 21.Rxa1 Rxd2= 16...Rc8! 17.Bxh6 d4 18.Ne4

This complicated position was reached in Lauer – Schakel, corr. 2013. I believe Black’s most promising continuation would have been: 18...d3!N 19.exd3 Bxd3 20.Rfe1 Bxe4 21.Bxe4 Nxe4 22.Rxe4

22...Nd4! 23.Qd3 Qd7 24.Kg2 Bf8 The misplaced knight on h3 and the weak squares around White’s king offer Black excellent play for a pawn.

608

13...Bd6! This seems like the most challenging option, as driving away the strong knight should improve Black’s counterattacking abilities. 13...Bb7 transposes to variation G4 on page 304. Although Black’s position is quite acceptable there, I believe that the text move is more accurate against the particular move order White has chosen here. 14.Nc4 This logical retreat has been White’s choice in all four of the games so far. White intends to eliminate one of our bishops. The ambitious 14.f4 is positionally risky, and after 14...Bb7 15.Rad1 Nc6„ Black is at least equal. 14.Nd3!?N keeps the tension but loses some time, and after 14...Nc6 Black’s active piece play fully compensates for any pressure on the hanging pawns. For instance:

609

15.Bg5 (15.Qd1 Rc8 16.Bg5 Be7 17.Rc1 Nb4!„) 15...Nd4 16.Qd1 Qa5 17.Bd2 Qc7 18.Nf4 Rad8„ Black is doing well. Finally, we should also consider the thematic knight sac: 14.Nxf7!?N 14...Kxf7 15.Bg5 After 15.Nxd5 Nbd7 16.Rfd1 Rb8 17.Bc3 Kg8 Black manages to consolidate and maintain his material advantage. 15...Nbd7 16.Bxd5† Kf8 17.Ne4 After the materialistic 17.Bxa8?! Qxa8 18.Rad1 Be5³ Black takes over the initiative due to his powerful minor pieces, especially the light-squared bishop. 17...Be5 18.Rad1 Rc8

610

19.Nxf6 In the event of 19.Bxf6 Nxf6 20.Bb7 Qb6 21.Bxc8 Bxc8 22.Qxc5† Qxc5 23.Nxc5 Bh3³ White’s rook will be caught, so Black will emerge with a piece for three pawns. The pawns are a long way back, so Black has some winning chances. 19...Nxf6 20.Bb7 Qb6 21.Bxc8 Bxc8 22.Bxf6 Bxf6 23.Qxh7 Qc6„ Black’s superb bishops make up for White’s slight material advantage.

14...Nc6 15.Nxd6 Qxd6 Although White’s bishop pair is an asset, Black’s remaining minor pieces are perfectly located. In particular, the queen’s knight is heading for d4. 16.Rfe1 16.Bf4 is hardly an improvement in view of: 16...Qe6 17.Rfe1 (17.Qd2 d4 18.Na4 Ne4„) 17...Rad8 18.Rac1 Nd4 19.Qb2 Qg4 20.f3 Qh5 Having provoked a weakening of White’s kingside, Black had excellent prospects in R.B. Ramesh – Marti Pericot, corr. 2016. After the text move, I found a natural improvement over Zubov – Kryvoruchko, Lvov 2014.

611

16...Nd4N 17.Qd1 Ne6! 18.e4 After 18.Rc1 Rac8„ Black’s pieces are located more harmoniously, so the hanging pawns have excellent support. 18...dxe4 19.Nxe4 Nxe4 20.Bxe4

20...Rac8 21.Be3 Nd4 22.Bg2 Rcd8„ White’s preferable pawn structure is fully counterbalanced by Black’s activity. F2) 12.Rfd1

612

This is more of a patient, positional approach than the previous line. 12...Nc6 13.Be3 13.Bg5 has been tried but the exchange on f6 suits Black just fine: 13...h6 14.Bxf6 Bxf6 15.e3 cxd4 16.Nxd4 Nxd4 17.exd4 Rc8 Black was comfortably equal in Bu Xiangzhi – Matlakov, Doha 2016. 13...Ng4! This aggressive move is fully justified, and White already needs to improve over the existing games.

14.dxc5!N 14.Qf5?! has been played in all seven games so far, but it proves to be weaker after: 14...Nxe3 15.fxe3 613

15...g6! 16.Qxd5 Nb4 17.Qe4 Bg5 18.Ne5 cxd4 19.exd4 Rc8ƒ White came under heavy pressure in Torrijos Alhambra – Hernaez Fernandez, corr. 2009. 14...Nxe3 15.fxe3 Nb4 16.Qb2 Bf6! 16...Bxc5 may look tempting but 17.a3 Bxe3† 18.Kh1 Nc6 19.Nxd5 turns out in White’s favour.

17.a3 Nc6 18.Qc2 bxc5 19.Rac1 Ne7 In the arising complex position, Black’s chances are by no means worse. G) 11.Ne5

614

Finally we come to the most popular and ambitious continuation. The knight takes up an imposing position in the centre, while opening the diagonal for the powerful bishop on g2. 11...Bb7 The bishop has done its job on the a6-f1 diagonal, and now it drops back to bolster the d5-pawn and guard the c6-square. Now that White’s knight has stopped controlling the d4-spot, the ...c5 break is more likely to be effective. We will analyse G1) 12.Bc1, G2) 12.Bg5, G3) 12.Bf4, G4) 12.Qc2 and G5) 12.Rc1. 12.a3 should be met by 12...Nbd7N, leading back to page 289 – see 12.Ne5 Bb7 in the notes to variation B. (Instead, 12...c5 13.dxc5 bxc5 transposes to the note to Black’s 11th move in variation B, where my assessment is that Black is not quite equalizing.) G1) 12.Bc1 This somewhat slow method of handling the position is not without merits. The bishop is heading to b2, after which White’s set-up will be quite harmonious. In particular, it will make the thematic ...c5 break less effective.

615

12...Nbd7 13.Bb2 Bf8 14.f4 This seems most consistent, as the presence of White’s knight in the centre is annoying for Black. On the flip side, weakening the e4-square is an obvious concession. 14.Rc1!? This move was tried by the great Mikhail Botvinnik in a World Championship match, so it deserves to be mentioned. 14...Nxe5 15.dxe5 Rxe5 16.Nb5 Re7 17.Bxf6 gxf6 The weakening of Black’s kingside should offer White enough compensation for the pawn, but nothing more.

18.e4 616

A more recent game continued 18.Qd3?! c6 19.Nd4 Qd7 20.Nf5 Re5 21.e4 Rae8 and Black was clearly better in Ladva – Neiksans, Viljandi 2016. 18...dxe4 19.Qg4†?! An overoptimistic decision. White should have settled for 19.Nxc7N 19...Qxd1 20.Rfxd1 Rb8 21.Nd5 Bxd5 22.Rxd5, maintaining the balance.

19...Bg7 20.Rfd1 Qf8 21.Nd4 Bc8 22.Qh4 f5 23.Nc6 Re8 White’s attack has come to an end, and Black won convincingly in Botvinnik – Bronstein, Moscow (11) 1951.

14...c5 15.e3 cxd4 16.exd4 We have been following the high-level game Anton Guijarro – Harikrishna, Gibraltar 2016. 617

My new idea is:

16...a6!?N Intending to seize more space on the queenside and secure the comfortable b6-spot for the knight. 17.a4 In the event of 17.Qd3 b5 18.Rae1 b4 19.Na4 a5„ Black does not experience any difficulties. The aggressive 17.g4 is well met by: 17...Bb4! 18.g5 (18.Qd3 a5 19.Qg3 Ne4 20.Nxe4 dxe4„) 18...Bxc3 19.Bxc3 Ne4

20.Bb2 f6 21.gxf6 Ndxf6„ The powerful knight on e4 offers Black promising play.

618

17...Rc8 18.Rc1 Bb4

19.Qd3 Nf8 20.Rc2 Ne6„ The d4-pawn is under pressure, and the f6-knight might jump to e4 at any moment. G2) 12.Bg5

12...Nbd7 This position might also arise after 11.Bg5 Nbd7 12.Ne5 Bb7, as noted earlier on page 287. 13.Rc1 The aggressive 13.f4?! leaves a hole on e4 and might backfire: 13...c5 14.Kh1 Rc8! (14...Ne4?! was 619

only about equal in Sargissian – S. Zhigalko, Dubai [rapid] 2014) 15.Bh3 (I also considered 15.Rc1 cxd4 16.Qxd4 Ba3!ƒ when White’s set-up seems really shaky)

15...cxd4 16.Nb5 d3! 17.Qxd3 Nc5 18.Qd1 d4† 19.Nf3 Nce4! 20.Bxc8 Qxc8 21.Kg1 Qh3 Black had a crushing attack for the exchange in Ugrinovsky – Novikov, corr. 2016. 13...Ba3! 13...c6!? 14.f4 c5„ is another decent option but I like the idea of exploiting White’s vulnerable dark squares on the queenside.

14.Rc2 After 14.Nxd7?! Qxd7 15.Bxf6 Bxc1 16.Qxc1 gxf6 White had insufficient compensation for the

620

exchange in Sargissian – Tomashevsky, Huaian (rapid) 2016. 14...Nxe5 15.dxe5 15.Nb5!?N is an interesting tactical resource but it’s hardly a significant improvement. My analysis continues: 15...Bd6 16.Nxd6 cxd6 17.Bxf6 Qxf6 18.Rc7 Ba6 19.dxe5 dxe5 20.Bxd5 Rad8

21.Rxa7 Bxe2 22.Qxe2 (22.Bxf7†?? should obviously be avoided: 22...Qxf7 23.Qxe2 Qxa7–+) 22...Rxd5= 15...Rxe5 16.Bf4 Re7 17.Nb5 Bc5

18.Nxc7?! 621

Objectively, White’s best continuation was 18.b4N 18...Bxb4 19.Nxc7 Rc8 20.Nxd5 Bxd5 21.Bxd5 Rd7 22.Rxc8 Qxc8 23.e4 Nxe4 24.Qf3. In that case, he would be a pawn down but his powerful bishops would offer him good drawing chances. The text move was played in Michalik – Sadzikowski, Czech Republic 2017. Black went on to lose the game, but only after missing the following great opportunity: 18...g5!N 19.Nxa8 gxf4µ Black’s minor pieces are much stronger than White’s rook. G3) 12.Bf4

This choice of square for the bishop is usually connected with attacking the c7-pawn. 12...Nbd7 13.Nb5!? I am surprised that this direct attacking move has not yet been tried in a high-level game. White’s idea is to meet 13...a6 with 14.Nxf7, as occurred in one correspondence game. The much more common 13.Rc1 transposes to variation G53 on page 310. 13...Nxe5!?N This move provokes complications which I am happy to enter. 14.Bxe5 14.dxe5 Nh5 15.Bc1 g6 16.Bb2 c6 17.Nd4 c5 18.Nc2 Qd7 offers Black excellent chances due to his better pawn structure.

622

14...Ng4! 15.Bf4 In the event of 15.Nxc7 Nxe5 16.Nxe8 Nd7 17.Nxg7 Kxg7 White is equal at best. His rooks have little purpose, whereas Black’s minor pieces can easily find good squares.

15...Qd7 16.Nxc7 g5 17.Nxe8 gxf4 We have reached an original and unbalanced position, offering interesting play for both sides. In my opinion, Black’s chances are not worse. G4) 12.Qc2

This natural move vacates the d1-square for one of the rooks.

623

12...c5!? 12...Nbd7 is a sensible developer which has been used by some top players. Having analysed both options, I prefer to challenge White’s pawn centre immediately. 13.dxc5 bxc5 14.Rad1 14.Rfd1N would hardly be an improvement. For instance, 14...Qc8 15.Na4 Nbd7 16.Nxd7 Nxd7 17.Rac1 d4„ and Black’s pressure along the e-file compensates for the vulnerability of the c5-pawn. 14...Qc8 15.e4 The attempt to apply piece pressure on Black’s central pawns proved ineffective in the following game: 15.Bg5 Na6 16.Na4 h6 17.Bxf6 Bxf6 18.Nd3

18...c4! 19.bxc4 dxc4 20.Bxb7 Qxb7 White needed to play precisely in order to maintain the balance in Wang Hao – Sasikiran, Guangzhou 2010.

624

15...Bf8 16.f4 I will take this as the main line, as Aronian scored an excellent win with it. A previous correspondence game took a different direction: 16.Bf4 d4 17.Nd5!? This natural move is connected with a piece sacrifice. I also examined 17.Na4N 17...Nh5 18.Nd3 Nxf4 19.Nxf4 Nd7„ when the bishop pair offers Black decent counterplay. Among other plans, he may look to challenge the knight on a4 and force through ...a5-a4 in order to weaken White’s queenside structure. 17...Nxd5 18.exd5

18...g5! 19.Nf3! This is White’s best attempt to justify his material sacrifice. 625

After 19.Nxf7N 19...gxf4 20.Ng5 Qd7 21.Bh3 Qg7 22.Be6† Rxe6 23.Nxe6 Qf6 24.Nxc5 Bxd5 25.Ne4 Qe5 26.Rfe1 Nc6 Black manages to complete development and maintain the powerful passer on d4, so White is in some danger. 19...gxf4 20.Ng5 f5

Unlike the note above, the f7-pawn prevented Black’s queen from participating in the defence. Black therefore has to make this additional pawn move, enabling White to increase the momentum of his attack. He still has no advantage though, as the game demonstrates. 21.Rde1 21.Bh3N is another logical attempt to develop the attack. Play continues 21...Re5! 22.Rfe1 Rxd5 23.Re6 Nc6 24.Bxf5 d3! 25.Qc4 when White will recoup material but Black improves his pieces: 25...Nd4 26.Re5 Qc6 27.Bxh7† Kg7 28.Be4 Rad8 29.Bxd5 Rxd5 30.Rxd5 Qxd5 31.Qxd5 Bxd5 32.Rxd3 Be7„ Black’s minor pieces are well coordinated, so his chances are by no means worse. 21...c4! This excellent tactical resource enables Black to secure a great spot for his knight. 22.bxc4 In the event of 22.Rxe8 Qxe8 23.Qxf5 Qg6 24.Qxf4 Bd6 25.Qc1 h6 26.Ne4 Bf8 27.bxc4 (27.Qxc4? Ba6µ) 27...Nd7„ White’s attack comes to an end, while his passers are blocked.

626

22...Na6 23.Re6 Rxe6 24.Nxe6 Nc5 By now, Black was no longer in danger, and White had to play accurately to hold a draw in Kuta – Alexa, corr. 2013. Our main line also sees White in an aggressive mood, as he is willing to expose his own king in order to maintain the powerful knight on e5. My suggested improvement is:

16...dxe4!N 16...d4 is riskier in view of 17.Nd5 Nxd5 18.exd5 when White’s lead in development starts to tell. Let me share the following game, which beautifully illustrates White’s attacking potential: 18...f6? (18...Nd7 19.Bh3 Re7 was called for, but it looks ugly to the human eye) 19.Rde1! fxe5 20.fxe5 Nd7 21.e6 Nf6 22.Rxf6! gxf6 23.Qf5 White had a crushing attack in Aronian – Navara, Saint Louis 2017. 627

17.Nxe4 Nxe4 18.Bxe4 Bxe4 19.Qxe4 Nd7

Black is in time to connect the rooks and solve the last of his problems. I will include an illustrative line. 20.Qc4 Nxe5 21.fxe5 Qe6 22.Qxe6 Rxe6 23.Bc3 c4! White has no advantage, for instance:

24.bxc4 24.b4 Ra6! 25.Rd7 Ra3= offers Black enough counterplay.

628

24...Rc8 25.Rd7 Rxc4 26.Bb2 Be7= The draw is almost certain. G5) 12.Rc1

This natural developing move has been the most popular choice of all. 12...Nbd7 We reply with an obvious developer of our own. We have a final branching, with three main options for White: the direct G51) 13.Nb5, the patient G52) 13.Rc2 and the most popular G53) 13.Bf4. 13.Bg5 takes us back to variation G2. 13.Nd3 a5 transposes to variation C32 of the previous chapter; see page 283 for the continuation beginning with 14.Bf4. G51) 13.Nb5 c5 14.Bf4

629

The activity of White’s pieces may appear annoying, but there is a good solution: 14...Nf8! As usual, the knight is heading for e6 in order to challenge White’s centre and disturb the bishop on f4. 15.dxc5 15.Bh3!? This move takes away the last available squares from Black’s queen. White is inviting interesting complications, which have been explored in correspondence games: 15...a6!

16.Nxf7! 630

Only this spectacular continuation might justify White’s previous play. 16...Kxf7 17.Bc7 The queen is trapped, but Black will pick up three minor pieces. 17...axb5 18.Bxd8 Rexd8 19.Rc2

19...Ne6 This is the most ambitious move, provoking a minor-piece exchange which favours Black in the long run. 19...b4!? is a decent alternative: 20.f3 (20.Qc1!?N 20...Kg8 21.dxc5 bxc5 22.Rxc5 Bxc5 23.Qxc5 is a reasonable attempt to change the course of the game, though after 23...Bc8 24.Bxc8 Rdxc8 25.Qxb4 Rxa2 it is White who will have to be more careful to hold the endgame) 20...Kg8 21.e3 Bd6 This was Koskela – Nickel, corr. 2013. Despite his marginal material disadvantage, Black’s chances seem preferable; the minor pieces are active, while White’s heavy pieces have no access to open files. 20.Bxe6† White is also struggling after 20.e3 b4³, as evidenced by Black’s 3/3 score in correspondence games from this position. 20...Kxe6 21.f3 Kf7

631

This occurred in Jandek – Siikaluoma, corr. 2011, and several subsequent correspondence games. In the arising unbalanced situation Black’s chances seem preferable, since White’s heavy pieces have no open files, whereas the pressure along the a-file is rather strong. This assessment is confirmed by the fact that Black has achieved a huge score in the correspondence games (+4 =2 –0 at the time of writing). 15...bxc5

16.Nc4 The harmless 16.Nd3 Ne6 17.Be5 Nd7 18.Bb2 was played in Javakhishvili – Khukhashvili, Tbilisi 2002. Black responded by deploying the knight to b6, but I consider it somewhat misplaced there. A significant improvement is:

632

18...Qb8!N Intending to advance the d-pawn. 19.e3 Otherwise ...d4 comes anyway, with excellent prospects for Black. 19...d4!³ Highlighting the loose knight on b5, and leaving White with some problems to solve. The text move is a thematic idea in the hanging-pawns structure: when the b7-bishop is unprotected, White can put his knight on c4 and create various tactical motifs. 16...Ne6 17.Be5 In the event of 17.Nbd6N 17...Nxf4 18.Nxb7 Qc7 19.gxf4 Qxb7„ White is unlikely to benefit from exchanging the light-squared bishop, as his kingside pawn chain has been damaged. 17...Ng4! 18.Bb2 Ba6! With accurate play, Black forces the active enemy pieces to retreat. 19.Nca3 Bf6 20.Bxf6 Nxf6 White was suffering from lack of harmony in Pacan Milej – Kantor, Rymanow Zdroj 2017. G52) 13.Rc2

633

This move may appear toothless but it is actually quite purposeful. The dark-squared bishop can be rerouted to b2, and later White may double his rooks. 13...c6 13...c5 proved to be premature after 14.Nd3! cxd4 15.Nb5 Nc5 16.Nxd4² in Miranda – Wajnberg, corr. 2008. 14.e4 This active move seems like the most natural reaction after Black has blocked the view of the b7bishop. I also examined: 14.Bc1N 14...Nxe5 15.dxe5 Nd7

634

16.f4 (16.e4 is well met by 16...d4! 17.Qxd4 Bc5 with better chances for Black) 16...f6 17.e4 d4 18.exf6 Nxf6 Black is by no means worse. Another natural continuation is: 14.Bf4 Nf8

15.Bc1!? Dropping the bishop out of harm’s way while claiming that Black’s last move was not helpful, as it took the pressure off the knight on e5. I also examined 15.e4 Ne6 16.Be3 (inserting 16.exd5 cxd5 makes no difference) 16...Rc8 (16...Bb4N looks like an equally valid move order) 17.exd5 cxd5 18.Qd3 Bb4 and Black had no problems in Macarruda – Mulligan, corr. 2016. 15...Bd6 16.f4 Ne6 17.e3 This was Schmidt – Ottesen, corr. 2016. I find the most logical continuation to be:

635

17...c5N 18.Nb5 Bf8 19.dxc5 Bxc5„ 14...dxe4 15.Nxe4 Nxe4 16.Bxe4 Nxe5 17.dxe5 It looks as though White will obtain some pressure thanks to his extra central pawn and strong bishop on e4, but the next move solves all of Black’s problems.

17...Bd6! 18.Bxc6 18.f4?? runs into 18...Bxe5! 19.fxe5 Qd4†, and White is almost lost. 18...Bxc6 19.Rxc6 Bxe5= As played in Folk Gilsanz – Morrow, corr. 2014, and several subsequent games, all of which were 636

drawn. G53) 13.Bf4

Taking aim at c7 is a natural plan but Black has a good answer. 13...Nf8! The knight’s transfer to e6 is especially effective with the bishop on f4 acting as a target. 14.Bg5 This has been the most popular choice, although the position is open to a number of interpretations. Here are five other options, arranged in approximately ascending order of importance: a) The pseudo-active 14.g4?! is well met by standard counterattacking measures: 14...Ne6 15.Bg3 c5 16.dxc5 Bxc5³ White’s set-up was already rather shaky in Damljanovic – A. Kovacevic, Palic 2014. b) 14.Nd3 Ne6 15.Be5 has been tried in three games. In each of them, Black based his strategy on hunting down the bishop by moving his knight to either d7 or g4. Although that approach seems valid, I would like to propose a different idea:

637

15...Ne4!?N 16.Nxe4 dxe4 17.Nf4 Nxf4 18.Bxf4 Bd6 19.Bxd6 Qxd6= White can hardly hope for an advantage with his bishop on g2 so restricted. c) 14.Qd3 Ne6 15.Rfd1 c6 16.Be3 This has occurred in a few high-level games. Black’s best continuation is:

16...Rc8! 16...Nd7 was also playable in Gelfand – Leko, Tromso (ol) 2014, but I see no special reason to exchange pieces. 17.Nf3 17.Bh3 Ba3 18.Rc2 c5³ is pleasant for Black. Also after 17.Rc2 Bd6 White must worry about the ...c5 advance, so I prefer Black. 638

White found nothing better than retreating his knight voluntarily in B. Kogan – Ottesen, corr. 2015, which implies that Black’s opening has been a success. Here I suggest: 17...Bd6N Intending ...Qe7, when I prefer Black’s chances. d) 14.Bh3!? This somewhat artificial move may invite interesting complications. 14...c5

15.e3 15.Nb5!? is a challenging though ultimately risky move, which has already been covered on page 307 – see 15.Bh3!? in the notes to variation G51. 15.Qd3 failed to impress after 15...Bd6 16.Nb5 Bb8 17.e3 Ne6 when Black was already better in Fages – Sacerdotali, corr. 2010. The text move is safest, but it is hardly a serious threat to us. 15...Ne6 16.Bxe6 fxe6 17.Qe2 Rc8 Black had nothing to worry about in Nucci – Ovaskainen, corr. 2013. e) Finally, a high-level game continued: 14.Nb5 Ne6 15.Nc6

639

Usually in these positions, the appearance of White’s knight on c6 is a sign that something has gone wrong for Black. However, in this specific situation it gives White nothing special. His pieces are not sufficiently well placed to support the knight, so the best that he can do is swap it for the bishop on e7. Meanwhile Black is poised to swap off the bishop on f4, damaging White’s pawn structure in the process. 15...Qd7 16.Nxe7† Rxe7 17.a4 I also considered 17.Nc3N 17...Nxf4 18.gxf4 c6 19.e3 Bc8 20.Kh1 Qd6 when Black is by no means worse. White’s damaged pawn structure on the kingside might count against him in the long run.

17...c6 18.Nc3 Nxf4 19.gxf4 a5 White ran out of constructive ideas and went on to lose in Gelfand – Karjakin, Astana (rapid) 2012. 640

Returning to the main line, in anticipation of ...Ne6, White intends to trade on f6 and cement the position of his knight with f2-f4.

14...Ne6 15.Bxf6 Bxf6 16.e3 c5 17.f4 White continues with his plan. A less aggressive continuation is 17.Re1 Rc8 18.Ng4 Be7 19.dxc5 Rxc5 when the activity of Black’s pieces fully compensates for the isolated pawn, as in Gerhards – Jensen, corr. 2008.

17...Rc8 White’s set-up looks active but it is tough for him to make much progress; the d4-pawn is under strong pressure. 641

18.Qd2 18.dxc5 Bxe5 19.fxe5 was tried in Eldridge – Raessler, corr. 2009. I see nothing wrong with the most natural reply:

19...Rxc5N 20.b4 Rc7 21.Nxd5 Rd7 22.e4 Ng5 Black will regain the pawn, so White will have to take care to maintain the balance, as his pawn structure is worse. 18...cxd4 19.exd4 Chances are level, as demonstrated by both engine analysis and correspondence praxis. One such game continued:

642

19...Rc7 But not 19...Be7? 20.Nxd5! when Black lost a pawn and had to suffer for a long time to make a draw in Kazakovskiy – T.D.V. Nguyen, Marianske Lazne 2017. 20.b4 g6 21.Nb5 Rxc1 22.Rxc1 a6 23.Nc3 Qd6 The position remained equal in Oezmen – Schakel, corr. 2014. Conclusion This chapter has completed our repertoire after 5.b3 Bb4† 6.Bd2 Be7, by analysing the main line which occurs after 7.Bg2 0-0 8.0-0 d5 9.cxd5 exd5 10.Nc3. In this popular position we favour the flexible 10...Re8!, which has been championed by Karjakin. White then has several playable moves, each with their own nuances. For example, 11.a3 and 11.Rb1 both prepare to advance the b-pawn, but it is important to understand the slight difference between them, namely the fact that Black can post his bishop on c4 at a certain moment in the latter line, based on the fact that the enemy pawn will be hanging on a2. Other moves such as 11.Rc1, 11.Re1 and 11.Qc2 all contain their own subtleties as well. We ended by analysing the most popular option of 11.Ne5, when 11...Bb7 is necessary to strengthen Black’s position in the centre. This will generally be followed by ...Nbd7 to challenge the strong knight. As usual, Black’s subsequent play will depend on what White is doing, but in general we should keep in mind a well-timed ...c5 break as a source of counterplay, as well as the ...Nf8-e6 manoeuvre, especially in those lines where White develops his bishop to f4.

643

A) 3.a3 315 B) 3.e3 316 C) 3.Bg5 318 D) 3.g3 Bb4† 4.Nd2 c5 321 D1) 5.dxc5 322 D2) 5.a3 Bxd2† 6.Qxd2 cxd4 7.Nf3 d5 326 D21) 8.Bg2 327 D22) 8.cxd5 329

644

1.d4 Nf6 2.c4 e6 White’s two most common moves in this position are 3.Nc3 and 3.Nf3, inviting the Nimzo-Indian and Queen’s Indian. In the final four chapters of the book we will discuss his other options. In this chapter we will briefly look at the sidelines A) 3.a3, B) 3.e3 and C) 3.Bg5, before beginning our coverage of the Catalan after D) 3.g3. 3.f3? is a stubborn attempt to arrange e2-e4, but 3...d5 is a convenient reply. Then 4.Nc3 invites a transposition to a popular variation of the Nimzo-Indian, but 4...c5! is a stronger reply which enables Black to attack the centre much faster. After the further 5.e3 Nc6 6.cxd5 Nxd5 7.Bb5 Bd6 Black was much better in Mietkiewicz – Pacan Milej, Suwalki 2015. A) 3.a3

This radical way of avoiding the pin is occasionally tried by strong players, but is mostly aimed at avoiding theory. Compared to the more respectable Petrosian System, Black has more opportunities to attack the centre, making the loss of a tempo more relevant. 3...d5 3...c5!? is another decent choice, but I guess 4.d5 exd5 5.cxd5 will not be to everyone’s taste, even though Black is practically a tempo up compared to a normal Modern Benoni. 4.Nc3 I checked two other options. After 4.Nf3 the most challenging reply seems to be: 4...dxc4! 5.e3 b5 6.a4 This is the only adequate response, but it’s a clear admission that a2-a3 was premature. 6...b4 7.Bxc4 This position usually arises from a Queen’s Gambit Accepted, but with White to move instead of Black!

645

7...Bb7 8.0-0 Nbd7 9.Qe2 (9.a5 c5 10.Qe2 Rc8 11.b3 Be7 12.Rd1 0-0 13.Bb2 Qc7 was also fine for Black in Morais Pinto – Loureiro, Sao Paulo 2001) 9...c5 10.Rd1 Be7 11.Nbd2 0-0 Black had a pleasant game in Rossetto – Maderna, Mar del Plata 1947. 4.cxd5 exd5 5.Nc3 c6 leads to a Carlsbad structure where a2-a3 is not always a useful move. In the Queen’s Gambit Declined, one of Black’s biggest challenges is to develop his light-squared bishop actively, but here he can accomplish it easily as he is virtually a tempo up. For instance:

6.Bg5 (6.Qc2 g6 7.Bg5 Be7 8.e3 Bf5 9.Bd3 Bxd3 10.Qxd3 0-0 11.Nge2 Nbd7 led to balanced play in Szabolcsi – Lobzhanidze, Vaujany 2010) 6...Bf5 7.e3 Nbd7 8.Bd3 Bxd3 9.Qxd3 Be7 10.Nf3 0-0 11.0-0 Ne4 12.Bxe7 Qxe7 13.Rfc1 f5 Black had excellent prospects on the kingside in Damm – Boensch, Bad Woerishofen 1992. 646

4...c6!? Black has a few ways to reach a pleasant version of some normal opening where a2-a3 has been played prematurely, but I find the text move the most attractive. Having already wasted a tempo on the queenside, White cannot afford to sacrifice the c4-pawn, so he should either release the tension or play e2-e3, blocking the dark-squared bishop. 5.e3 5.cxd5 exd5 is covered via 4.cxd5. 5...Bd6 6.Nf3 0-0 7.b4 This is the only way to make sense of the early a2-a3. 7...Nbd7 8.Bb2 Qe7 9.Qc2 I also considered 9.c5 Bc7 10.Be2 e5 11.0-0 e4 12.Nd2 Re8 13.b5 Nf8 when Black has serious attacking potential on the kingside. 9...e5 9...a6!? is a decent alternative but Black is ahead in development, so it makes sense to open the position at once. 10.cxd5 cxd5 11.dxe5 Nxe5 12.Nb5 White has to take control of d4.

647

12...Nxf3† 13.gxf3 Be5 14.Bxe5 Qxe5 15.Nd4 Re8 16.Be2 Bh3 Black had excellent play in Agrest – Serper, Oberwart 1994. B) 3.e3 b6 As usual, when dealing with such an unassuming third move, you could go for any sensible option such as 3...d5 or 3...c5, taking into account your own repertoire and preferences. However, for our purposes the text move seems like a convenient choice for a Queen’s Indian player. 4.Nc3 Bb7 5.Nge2!? This seems like White’s most interesting option. 5.Nf3 takes us back to variation D of Chapter 2. As usual, 5.f3 is well met by 5...d5, preventing White from seizing the centre. After 6.cxd5 exd5 7.Bd3 Bd6 8.Nge2 0-0 Black is doing well, for instance:

648

9.0-0 (9.e4?! only invites trouble, since White’s pieces are not developed: 9...dxe4 10.fxe4 Nc6! 11.Bg5 [11.e5 Bxe5 12.dxe5 Nxe5 13.Bc2 Qxd1† 14.Bxd1 Bxg2–+] 11...h6 12.Bh4 Be7 13.Bb1 Ng4ƒ) 9...Re8 10.a3 c5 Black had the more comfortable game in Manukyan – Matevosyan, Yerevan 2016.

5...d5 I was surprised to discover that White’s last move is not without venom. For instance, after 5...Be7 6.d5 0-0 7.e4 White has great prospects due to his space advantage. 6.cxd5 exd5 7.g3 Bd6 8.Bg2 0-0 9.0-0 Nbd7 10.Nf4 It is time to improve Black’s play.

649

10...c6N I don’t like 10...Bxf4, as in Gareev – Feng, Las Vegas 2014, in view of 11.exf4N 11...Re8 12.Qc2 Ne4 13.Re1 Ndf6 14.Be3² when Black is under some positional pressure. 11.f3 The quiet 11.b3 Re8 12.Bb2 Ne4 does not pose Black any problems. White’s ambitions are mainly connected with the e3-e4 advance, so the text move is critical. As always, Black should not ignore this positional threat.

11...c5! 650

This typical break invites interesting complications. 12.Nfxd5 Nxd5 13.Nxd5 Bxd5 14.dxc5

14...Bc4! 15.cxd6 Bxf1 16.Bxf1 Qf6 17.e4 Rfe8 We reach an unbalanced position where Black’s chances are preferable due to his better development; White’s pieces are unable to support the central pawns. C) 3.Bg5

This is the most natural of the sidelines we have seen so far. Indeed, what can be wrong with developing the bishop and pinning the opponent’s knight? However, it turns out to be rather committal, 651

since Black can exploit the vulnerable location of bishop. Still, it has been employed by some strong grandmasters, including Avrukh, Ivanisevic and Bauer, and was even used by Alekhine in a simul game! 3...h6 It makes sense to clarify White’s intentions now, and force the bishop off the c1-h6 diagonal, at least if it wants to maintain the pin. One of the ideas behind White’s last move is to meet 3...Bb4† with 4.Nd2, which is okay for Black, but quite distinct from the Nimzo-Indian. 4.Bh4 An inferior option is: 4.Bxf6 Qxf6 5.Nc3 Bb4 Black has an excellent version of the Nimzo-Indian Leningrad System. 6.Rc1 After 6.Nf3 Bxc3† 7.bxc3 d6 8.e4 e5³ White lacks any compensation for his damaged structure.

6...c5 7.e3 After 7.d5?! exd5 8.cxd5 0-0 9.Nf3 d6 White has no time to take action in the centre, whereas Black’s plan is obvious. 10.e3 Bg4 11.Be2 Nd7 12.0-0 Bxf3 13.Bxf3 Bxc3 14.bxc3 b5 Black was clearly better, due to his fast play on the queenside in Padevsky – Cobo Arteaga, Lugano (ol) 1968. 7...0-0 8.a3 cxd4 9.exd4 Bxc3† 10.Rxc3 Nc6 11.Nf3 d5 Black was at least equal in Zamikhovsky – Boleslavsky, Kiev 1958.

652

4...c5! I consider this the most challenging reaction to White’s unusual plan of development. 5.d5 5.e3 is rather timid. 5...cxd4 6.exd4 (6.Bxf6, as in Efimov – Zaitsev, Nareto 1999, can be well met by 6...Qxf6N 7.exd4 Nc6 8.Nf3 g5!µ when the absence of White’s dark-squared bishop already starts to tell) The desertion of White’s bishop from the queenside can be exploited by means of:

6...Qb6! 7.Qb3 Qxd4 8.Nf3 Qc5 White did not have much play for the central pawn in Van Ginkel – Rotstein, Latschach 2010. After 5.Nf3 Black has a few good possibilities, but I prefer 5...Qb6! to attack the vulnerable pawn on b2. 6.Qb3 Nc6 7.e3 occurred in Schleicher – Puschendorf, Schwarzburg 2005, when I think Black should 653

have played:

7...g5!N 8.Bg3 Qxb3 9.axb3 Ne4 The bishop will soon be exchanged, leaving Black free to exert strong pressure along the h8-a1 diagonal. For instance, 10.Nc3 Nxg3 11.hxg3 g4 12.Ne5 cxd4 13.exd4 Nxd4 14.0-0-0 Nxb3† 15.Kc2 Bg7 and Black is clearly better.

5...exd5 6.cxd5 d6 7.Nc3 Be7 The opposition of the bishops makes it difficult for White to complete development, as there are various tactics involving ...Nxd5 and/or ...Nxe4 if a pawn goes there too. 8.e4 8.Nf3 0-0 9.e3 transposes to 8.e3 below.

654

8.e3 looks like a concession, as White’s expansion in the centre is slowed down. 8...0-0 9.Nf3 a6 10.a3 (10.a4 Nxd5! 11.Qxd5 Bxh4 12.Nxh4 Qxh4 13.Qxd6 Qb4 14.Qd2 Nc6µ left White suffering from a lack of development in Turgut – D. Gurevich, Chicago 1997) 10...b5 11.Bd3 This occurred in Ivanisevic – Brkic, Legnica 2013. My new idea is:

11...Nxd5!N 12.Nxd5 Bxh4 13.Nxh4 Qxh4³ White has inadequate play for the pawn.

8...0-0 Black could not take the e4-pawn due to the check on a4, but after castling the threat is real. 9.Bg3 9.Nf3? simply drops the pawn: 9...Nxe4 10.Nxe4 Bxh4–+

655

The fact that the text move is almost the only way to avoid losing material move illustrates the lack of harmony in White’s camp. Here I found a way to improve Black’s play from J. Jones – Buttell, Cardiff 2009. 9...Re8N 10.Be2 10.Bd3 Nbd7 11.Nge2 Nh5 does not look attractive for White either.

10...Nbd7 11.Qc2 Bf8 12.Nf3 White can hardly postpone developing the knight any longer. The aggressive 12.f4? b5 13.Bxb5 Nxe4 14.Nxe4 Qa5† leads to a quick disaster for White.

656

12...Nh5 13.Bh4 g5 14.Bg3 Nxg3 15.hxg3 Bg7 With full control over the dark squares, Black has excellent prospects. D) 3.g3

With this move, White rules out the Queen’s Indian and invites Black to play ...d5, leading to a Catalan. 3...Bb4† This move is a convenient choice for us, as it invites a transposition to a Nimzo-Indian while also giving Black nice options against White’s other ways of blocking the check. 4.Nd2 4.Nc3 transposes to Chapter 7 of Grandmaster Repertoire – The Nimzo-Indian. 4.Bd2 is the main line, which will be analysed in the final three chapters. The text move will be analysed in the remainder of this chapter. This seemingly modest method of development was first seen in 1948, but the main merits of it were illustrated in the 1980s and 90s, when it was employed by Gelfand, Bareev and other Catalan experts. To this day, it remains a frequent guest at the top level. The main point of it is to claim that the bishop is not well placed on b4, as a subsequent a2a3 will force Black to either retreat it or swap it for the knight. 4...c5 The awkward location of the knight on d2 makes this standard break more effective. In particular, White cannot respond with d4-d5.

657

4...d5 5.Bg2 0-0 6.Ngf3 is another major variation, where White tries to exploit the drawbacks of the bishop’s placement on b4. Here are a few illustrative lines: 6...dxc4 7.Qc2 a5 (7...b5?! 8.a4 c6? 9.axb5 cxb5 runs into the thematic refutation 10.Ng5! h6 11.h4 and Black was already lost in Karner – Taggatz, Austria 2010) 8.a3 Bxd2† 9.Bxd2 b5 10.a4 c6 11.e4 White had interesting play for the pawn due to his strong centre in Tomashevsky – Ding Liren, Palma de Mallorca 2017.

We will analyse D1) 5.dxc5 and D2) 5.a3. A harmless alternative is: 5.Ngf3 cxd4 5...Nc6?! needlessly loses the bishop pair after 6.a3, so we shouldn’t consider it seriously. 6.Nxd4 6.a3 gives Black a pleasant choice between 6...Bxd2† 7.Qxd2 d5, which transposes to variation D2, and 6...Be7 7.Nxd4 Nc6, which also looks perfectly playable. 6...Nc6

658

7.Nc2 This is the most ambitious way of handling the position. 7.Nxc6 dxc6 8.Bg2 e5 9.Qc2 0-0= gave Black no problems in Pieper Emden – Rechmann, Germany 1995. 7.a3N 7...Be7 8.Nxc6 dxc6 9.Bg2 e5= is also pleasant for Black. 7...Be7 8.Bg2 0-0 9.0-0 d5 Black’s knights are better placed than their counterparts, so there is no reason to refrain from opening the position. 10.cxd5 Nxd5 10...exd5!? seems equally good, and after 11.Nb3 Ne4 12.Be3 Bf6 13.Nbd4 Re8„ Black’s active piece play fully compensated for any structural drawbacks in Plat – Akesson, Pardubice 2017.

659

11.e4 11.Nc4 Qc7 12.e4 Nb6 13.Qe2 Nxc4 14.Qxc4 Bd7 15.Be3 occurred in Remon – Vilela de Acuna, Bayamo 1980, when 15...e5!N 16.Rac1 Be6 17.Qa4 Rfd8³ would have enabled Black to claim the upper hand. 11...Ndb4 Black had good prospects along the d-file in Arca – Alpaslan, Antalya 2017. D1) 5.dxc5

This move has been tried by such strong players as Nakamura, Nisipeanu, Chernin and more. However, it looks like something of a concession, as now Black’s strong bishop will not be exchanged, whereas White still lacks any superiority in the centre. 5...Bxc5 6.Bg2 0-0 7.Ngf3 Nc6 Another attractive option might be 7...b6!? 8.0-0 Bb7 9.b3 Be7 10.Bb2 d6 when Black has a decent version of a Hedgehog, whereas White’s set-up looks rather unambitious. 8.0-0 d5 9.Qc2 With this move, White eyes the bishop on c5 and vacates the d1-square for the rook. White has a wide choice of possibilities, but I found nothing which fully justifies the poor location of the knight on d2. Here are some examples: 9.a3 a5 does not change a great deal. For example: 10.b3 d4

660

11.b4 (11.Ne1 e5 12.Nd3, as was played in Szuper – Zielinski, Dobczyce 2010, should be naturally met by: 12...Bd6N 13.Bb2 Bf5³ White suffers from lack of space.) 11...axb4 12.Nb3 Be7 13.Nfxd4 Nxd4 14.Nxd4 e5 15.Nc2 Be6 Black had comfortable play in Chernin – I. Sokolov, Austria 1995. 9.Nb3 Bb6 10.Bf4?! has been played, presumably with the idea of maintaining the central tension in order to restrict the movements of the bishop on c8. A strong reply is:

10...dxc4 11.Nbd2 c3! 12.bxc3 Nd5 White’s queenside structure has been damaged, and the same is about to happen on the kingside as well. 13.Nc4 Nxf4 14.gxf4 Qf6 15.Nxb6 (15.e3 is hardly an improvement in view of 15...Bc7! followed by ...e5) 15...axb6 16.Qd2 Ra4 17.e3 e5 White’s set-up looked pretty dicey due to the kingside weaknesses in Kurajica – Zelcic, Sibenik 2012.

661

9.cxd5 exd5 gives Black a comfortable IQP position, for instance: 10.Nb3 Bb6 11.Nbd4 Re8

Compared to the main lines of the Tarrasch Defence, the isolated pawn is not really weak, since White’s pieces are located passively and cannot attack it. We will follow a nice example: 12.b3 Bg4 13.Bb2 Ne4 14.Rc1 Rc8 15.Rc2 Qd6 16.a3 h5! Black’s pieces are much more active, so he can afford to start aggressive actions on the kingside. 17.b4 h4

In Rakhmanov – Shimanov, Sochi 2017, White should have invited simplifications with 18.Nxh4N 18...Nxd4 19.Rxc8 Nxe2† 20.Qxe2 Bxc8, but even then I prefer Black, who continues to exert strong pressure on the f2-pawn. A final option is: 9.e3 b6 I like this natural approach, intending to solve Black’s only problem: the passive light-squared 662

bishop. After 9...Qe7 10.cxd5 exd5 11.Nb3 Bb6 12.a4 a6 13.Nbd4 Bg4 14.Qb3 Ba7 15.Bd2 White had a tiny advantage in Nakamura – Eljanov, Moscow 2010. 10.a3 10.cxd5 Nxd5 11.Nb3 Ba6 12.Re1 Ndb4 gave Black the more active play in Pakosta – Mozny, Czech Republic 1993. 10...a5 11.b3 Ba6 12.Bb2 We have been following Nogueiras Santiago – Seirawan, Rotterdam 1989. A new but highly natural way of playing is:

12...Qe7N 13.Qc2 Rfd8 14.Rfc1 Rac8 Black has nothing to worry about.

663

9...d4!? There are other decent possibilities but I favour this aggressive approach. The space advantage is on Black’s side now. 10.Nb3 Be7 11.Bf4N Stopping ...e5 seems crucial – so I was surprised to discover that the text move has not yet been played. 11.Rd1?! e5 12.e3 sees White attacking the d4-pawn. However, his set-up lacks harmony, and Black can demonstrate this by means of:

12...a5!N (improving over 12...Bg4, when 13.exd4 e4 14.Ng5 Bxd1 15.Qxd1 gave White full compensation for the exchange in Topalov – Carlsen, Nice [rapid] 2008) 13.exd4 a4 14.Nbd2 exd4 15.Ne4 Nb4 16.Nxf6† Bxf6 17.Qd2 Qe7ƒ The strong passed pawn makes White’s position difficult.

664

11...Ng4! Of course, Black insists on playing ...e5. 12.h3 An equally good alternative is: 12.Bh3 e5 13.Bxg4 exf4 14.Bxc8 Rxc8

15.Rfd1 Bf6! (but not 15...fxg3? 16.hxg3 Bf6 17.Nbxd4 Nxd4 18.Nxd4 Bxd4 19.e3 when White wins a pawn for nothing) 16.gxf4 Qc7 17.Qf5 Rcd8 18.Rac1 Rfe8 19.Rd2 Re6 Black has enough play for the pawn, thanks to his pressure along the e-file and the vulnerability of White’s king. 12...e5 13.hxg4 exf4 14.gxf4 Bxg4 665

Visually Black’s position seems excellent, but we should not forget about the vulnerability of the d4pawn.

15.Ne5! This is the most challenging idea. 15.Rad1 Qc7 16.Nbxd4 Bxf3 17.Bxf3 Qxf4= is fine for Black. 15...Nxe5 16.fxe5

16...a5! Black should activate his rook in the most effective way possible.

666

17.Rad1 In the event of 17.Qe4 Qc8 18.Nxd4 Ra6 19.Qxb7 Rg6 Black is fine and White should look for a drawing line. 17...Qc8 18.Nxd4 18.Rxd4?! Bf5 19.Be4 Bxe4 20.Qxe4 a4 21.Nc1 Ra6 puts White in serious danger.

18...Ra6! With the rook heading for h6, Black’s compensation is undeniable. For instance: 19.Be4 Rh6 20.Bf5 Bxf5 21.Nxf5 Rg6† 22.Kh2 Bc5 23.e4 f6„ The position is objectively equal, though White has to be slightly more careful due to his vulnerable king. D2) 5.a3

667

This is the more challenging move, forcing Black to give up the bishop pair. 5...Bxd2† 6.Qxd2 6.Bxd2? cxd4 7.Bg2 Nc6 8.Nf3 e5µ leaves White with no compensation for the pawn. 6...cxd4 7.Nf3 This is the most logical choice, avoiding losing time with the queen. 7.Qxd4 Nc6 White has tried several queen moves but Black retains at least equal chances in all lines. For instance: 8.Qd3 a) 8.Qh4? was seen in Galopoulos – Theodoulidis, Achaia 2013. Now Black could have exploited the poor location of the queen by means of: 8...e5!N 9.Bg2 (or 9.Bg5 Qa5† 10.Bd2 Qb6µ) 9...Nd4 10.Rb1 d5 11.Bg5 (11.cxd5 Bf5 12.e4 Nxe4 13.Qxd8† Rxd8–+) 11...Nf5 12.Bxf6 gxf6 13.Qh5 dxc4 Black is already close to winning. b) 8.Qd6?! was another overambitious choice in Bragin – Khokhlov, Tomsk 2006, when 8...Qa5†!N would have been strong: 9.Qd2 (9.Bd2 Qb6 10.Bg2 Qxb2 11.Rd1 Qd4³; or 9.b4 Ne4 10.Qxe6† dxe6 11.bxa5 Nxa5³ and White’s queenside structure is ruined) 9...Ne4! 10.Qxa5 Nxa5 11.Bg2 Nf6 12.Nf3 Nxc4 White does not have much for the pawn. c) 8.Qd1 has been tried by three grandmasters. I think the best way of exploiting White’s lack of development is: 8...Na5!?N 9.c5 (9.b3? Qb6 10.Rb1 Nxc4 picks up the pawn) 9...b6 10.Bg2 Bb7 The exchange of light-squared bishops definitely favours Black. 11.Bxb7 Nxb7 12.cxb6 axb6 13.Nf3 0-0 14.0-0 d5 15.Be3 Re8 16.Rc1 (16.Bd4?! Nd7³) 16...e5 With a strong central pawn majority, Black does not face any problems. 8...b6 668

As usual, Black has a clear plan of attacking the weak c4-pawn. 9.Nf3 9.Bg2 Bb7 10.Nf3 is the same.

9...Bb7 10.Bg2 Rc8N Surprisingly, this natural developing move is a novelty. The tempting 10...Ne5 was tested in a few high-level games, but for some reason no one has tried 11.Nxe5!N 11...Bxg2 12.Rg1 Bb7 (12...Be4 13.Qd4 d6 14.Ng4!ƒ) 13.Bg5 Qc7 14.Bxf6 gxf6 15.Ng4, which offers White interesting play. 11.0-0 Also after 11.Bg5 h6 12.Bxf6 Qxf6 13.Rd1 0-0= Black has nothing to worry about.

11...Na5 12.b3 d5 13.Bd2 669

13.cxd5 Bxd5 14.b4 Nc4 looks somewhat unpleasant for White, since the presence of the strong knight in his camp could prove annoying. 13...0-0 14.Bxa5 bxa5 The doubled a-pawns are only superficially vulnerable. In reality, the b3-pawn is no less of a weakness. 15.Rfc1 dxc4 16.Rxc4 Qb6=

7...d5 White has two main options: D21) 8.Bg2 and D22) 8.cxd5. 8.Qxd4N 8...Nc6 resembles the previous note; here too, White can hardly hope for an opening advantage after losing so much time with his queen. For instance: 9.Qh4 0-0 10.Bg2 dxc4 11.0-0 e5 12.Qxc4 (12.Bg5 h6 13.Bxf6 Qxf6 14.Qxc4 Be6=) 12...Be6 13.Qh4 h6 Black is fine. 8.Nxd4 should be met by 8...dxc4N when White has nothing better than 9.Bg2, transposing to the note to move 9 in variation D21 below. D21) 8.Bg2 White is playing in the spirit of the Catalan, temporarily giving up a pawn in order to exert pressure along the h1-a8 diagonal.

670

8...dxc4 9.Qxd4 9.Nxd4 has not been played and indeed it seems less challenging. For instance, 9...0-0 10.0-0 e5! 11.Nb5 Nc6 12.Qxd8 Rxd8 13.Be3 Bg4 with excellent piece play for Black. 9...Qxd4 10.Nxd4 0-0 11.Be3 This has been played in two high-level games. White develops the last of his minor pieces and prepares to bring a rook to the c-file. I also examined: 11.0-0N 11...Rd8 12.Be3 (12.Nb5 Nc6 13.Be3 Nd5 is absolutely fine for Black) 12...Nd5 13.Rfd1

13...Nc6! Black has no need to fear the ensuing damage to his pawn structure. 14.Nxc6 bxc6 15.Bc5 Ba6

671

16.Rd2 Rdb8 17.Rc1 Rb3 Black’s extra pawn and pressure along the b-file fully compensate for his structural weaknesses and White’s bishop pair. Black also doesn’t face any problems after: 11.Bf4N 11...Nd5 12.Rc1 Bd7 (an interesting alternative is 12...Nd7!? 13.Bxd5 exd5 14.Nb5 Nc5 15.Nc7 Be6 16.Nxa8 Rxa8 with full compensation for the exchange, since White’s rooks have no access to open files)

13.Rxc4 Nc6 14.Nxc6 (after 14.0-0?! Nb6 15.Bxc6 Nxc4 16.Bxb7 Rae8³ White has to struggle for a draw; and 14.Nb3 Rac8 15.0-0 Nce7= is harmless) 14...Bxc6 15.0-0 Bb5 16.Rc2 Rac8 17.Rfc1 Rxc2 18.Rxc2 Nxf4 19.gxf4 b6= With a drawish endgame.

11...Nd5 12.Rc1 White is regaining the pawn, but paying a price in terms of the bishop pair and his kingside structure. 672

12...Nxe3 13.fxe3 Nd7! Since there is no way to neutralize the powerful bishop, Black is going to do the next best thing and remove all his pieces from the long diagonal!

14.Rxc4 Nf6 15.0-0 15.Rc7 e5 16.Nb5 Rd8 17.0-0 transposes to the main line below. White intends to double his rooks on the c-file. In Wojtaszek – Karjakin, Shamkir 2018, a good continuation would have been:

15...e5!N 16.Nb5 Rd8 17.Rc7 a6 18.Nc3 Rb8 19.Re7 After 19.Rd1 Rxd1† 20.Nxd1 e4 21.Nc3 Bf5 the g2-bishop is blocked, so Black has comfortable play. 673

19...b5 20.Rxe5 Be6 21.Rc5 b4 22.axb4 Rxb4= White’s extra pawn is balanced by the fact that his e-pawns are doubled and isolated. D22) 8.cxd5

This has been White’s most popular choice. I like the following reply: 8...e5! 8...Qxd5 has been much more common, but 9.Qxd4 Nc6 10.Qxd5 Nxd5 offers White a tiny advantage due to the bishop pair. The text move invites interesting complications. Black is trying to exploit the lack of harmony in opponent’s camp. 9.Bg2!?N This new move deserves close attention. 9.Nxe5 This is the only move to have been tested so far. My analysis continues: 9...Qxd5 10.Nf3 Nc6 11.Bg2 0-0 12.0-0 This interesting position arose in Juhasz – Kunze, Germany 2004. My new idea is:

674

12...Bf5N The bishop is heading for e4, where it will neutralize its counterpart on g2. 13.b4 13.Nh4 Be4 14.f3 Bg6 15.Nxg6 hxg6 16.f4 Qe6 gives Black excellent play along the e-file. 13...Be4 14.Bb2 a5 15.Rfd1 Rfd8 The activity of Black’s pieces fully compensates for White’s bishop pair. I also examined the following ambitious try: 9.Qg5!?N 9...0-0 10.Qxe5 White manages to grab the central pawn but now the e-file is open, so Black’s lead in development starts to tell:

675

10...d3! 11.Bg5 dxe2 12.Bxe2 Nbd7 13.Qd6 Also after 13.Qd4 Re8 14.d6 h6 15.Be3 Nb6 16.0-0 Bg4 17.Rfe1 Nbd5= White is unable to protect the passer, so Black has no problems. 13...Re8 14.Nd4

14...h6 Black manages both to regain the pawn and to liquidate the opponent’s bishop pair. 15.Bxf6 Nxf6 16.Qxd8 Rxd8 17.0-0 Rxd5 18.Rfd1 Bh3 19.Bf3 Rd6 20.Ne2 Rad8=

9...Nbd7 9...Qxd5?! runs into an improved version of the plan from the note above: 10.Qg5! 0-0 11.Qxe5 and White is better.

676

10.0-0 Now 10.Qg5? is pointless in view of 10...h6! intending 11.Qxg7?? Rh7 and the queen is trapped. 10...0-0 The closed character of the position prevents White from taking advantage of his bishop pair. 11.Qc2 After 11.e3 Re8! 12.exd4 e4 13.Ne5 Nxe5 14.dxe5 Rxe5 White loses a pawn for inadequate compensation.

11...b6! The materialistic 11...Re8 can be met by 12.e4! dxe3 13.Bxe3 Nxd5 14.Rfe1 when White gets interesting play for the pawn, due to his superior piece coordination. 12.e4 a5! After 12...dxe3 13.Bxe3 Bb7 14.d6 Rc8 15.Qe2 the d6-pawn might cause Black problems. The text move keeps things more closed while securing the a6-square for the bishop. 13.Bg5 13.Re1 Ba6 14.Bd2 Rc8„ also gives Black decent prospects.

677

13...Ba6 14.Rfc1 Re8 A complex strategic struggle lies ahead, where Black’s chances are not worse. Conclusion We started this chapter by considering some rare options. 3.a3 is a radical way of preventing the NimzoIndian, after which 3...d5 gives Black a comfortable version of one or another variant of the Queen’s Gambit or Semi-Slav. 3.e3 is not so bad, but White loses some flexibility and 3...b6 is a convenient reply for Queen’s Indian practitioners. Then we looked at 3.Bg5, which appears active but runs a serious risk of leaving the queenside dark squares exposed. 3.g3 is the big topic for the remainder of the book. We meet it with 3...Bb4†, and if 4.Nd2 c5 Black successfully challenges the enemy pawn centre. Then 5.dxc5 is pretty harmless but 5.a3 Bxd2† 6.Qxd2 cxd4 7.Nf3 is more interesting. We respond with 7...d5, when 8.Bg2 resembles an Open Catalan while 8.cxd5 e5! leads to dynamic play. The ensuing positions are often lively and challenging for both sides, but I believe Black is fine everywhere.

678

A) 8.Na3!? 335 B) 8.b3 336 C) 8.Nc3 338 D) 8.Bf4 343 E) 8.Qb3 b6 9.Nc3 Ba6 10.cxd5 cxd5 11.Ne5 Bb7 347 E1) 12.Rac1 352 E2) 12.Rfc1 354

679

1.d4 Nf6 2.c4 e6 3.g3 Bb4† 4.Bd2 This is the much more common way of blocking the check. 4...Be7 This apparent tempo loss is a typical response to White’s last move. The idea is the same as in the popular 4.g3 Ba6 5.b3 Bb4† variation of the Queen’s Indian: the bishop on d2 obstructs White’s control of the d-file and White is denied the easy option of a queenside fianchetto. 5.Bg2 Starting with 5.Nf3 has no independent value. For instance, 5...d5 6.Qc2 c6 and White hardly has anything better than transposing to normal paths with 7.Bg2. 5...d5 6.Nf3 This is the characteristic Catalan set-up, and is almost always played. Both 6.Qb3 c6 7.Nf3 and 6.Qc2 0-0 7.Nf3 c6 transpose to lines which will be examined later. 6.Nc3 c6 also has little independent significance: 7.Nf3 leads to a version of the normal set-up where the knight has gone to c3 early, while 7.cxd5 exd5 transposes to the note below. The only independent alternative worth mentioning is: 6.cxd5 exd5

680

White’s kingside fianchetto is not particularly scary in the Carlsbad central structure. Still, it is worth considering how to approach the manoeuvring game which lies ahead. 7.Nc3 c6 8.e3 0-0 9.Nge2 A similar position was covered in Chapter 12 of Grandmaster Repertoire – The Nimzo-Indian Defence. The fact that White has gained a tempo with Bd2 here helps him slightly, but is no great achievement; Black’s task of equalizing is still much simpler here than in the main line of the Catalan. 9...Re8 10.0-0

10...Bf8N I find this new move to be the most flexible choice. Black unblocks the e-file and avoids committing the queen’s knight for the time being. This is useful because, as a rule of thumb, Black 681

should generally meet f2-f3 with ...c5, before White gets a chance to build a big pawn centre with e3-e4. Should f2-f3 and ...c5 occur, the knight will be better on c6 than d7. You don’t need to know any more than this, but I will include some lines for illustrative purposes. 11.Qc2 After 11.f3 c5 Black’s scheme works to perfection. For instance, 12.Nf4 Nc6 13.Be1 Rxe3 14.dxc5 Re5 15.Bf2 d4 with excellent play for Black. 11...g6 Since 12...Bf5 is now an annoying positional threat, White’s next move is practically forced.

12.f3 c5 13.Rad1 Nc6 14.dxc5 14.Bc1 cxd4 15.Nxd4 Bd7 16.Qf2 Nxd4 17.exd4 Bc6= is harmless. 14...Bxc5 15.Kh1 The opposition of rook and queen along the d-file is annoying for Black, so it makes sense to accept the challenge. 15...Bxe3 16.Bxe3 Rxe3 17.g4 White’s pieces are better developed, so he should be able to regain the pawn.

682

17...Re8! 18.g5 Nh5 19.Nxd5 The most ambitious. In the event of 19.Rxd5 Qc7 20.Rd2 Be6 21.f4 Rad8 the vulnerability of White’s king might tell in the long run. 19...Qxg5 20.Nc7 Bf5 21.Qc4 Rxe2 22.Qxe2 Nf4 23.Qf2 Rc8

24.Qg3 Qxg3 25.hxg3 Rxc7 26.gxf4 Nb4© We reach an unbalanced endgame where, in view of White’s poor pawn structure, Black’s chances are by no means worse.

683

6...0-0 7.0-0 White can, of course, start with some other move like 7.Qc2 or 7.Qb3, but after 7...c6 he will almost always follow up by castling, thus transposing to one of our main lines. 7...c6 We have reached a major branching point. By far the most popular move is 8.Qc2, and we will analyse it in detail in the next two chapters. Before then, we will consider several significant sidelines: A) 8.Na3!?, B) 8.b3, C) 8.Nc3, D) 8.Bf4 and E) 8.Qb3. 8.Bg5 does not have much independent significance: 8...Nbd7 9.Nbd2 (9.Qc2 leads straight to variation A in the next chapter) 9...h6 10.Bxf6 Nxf6 White has nothing better than 11.Qc2, transposing to variation A of the next chapter. A) 8.Na3!?

684

This move appears strange but it was successfully employed by Levon Aronian and has since been tried by other strong players. The main merits of this continuation are that it protects the c4-pawn while keeping the c-file unobstructed. 8...Ne4!? I like this active response. The venom of White’s set-up was well illustrated in the following game: 8...b6 9.Rc1 Bb7 10.Ne5 Nfd7 (10...Nbd7 11.cxd5 cxd5 12.Nc6²) 11.Nd3! a5 12.cxd5 exd5 13.e4 dxe4 14.Bxe4 Nf6 15.Bg2 Bxa3 16.bxa3² Black was doomed to passive defence in Martirosyan – Sevian, Saint Louis 2017. 9.Bf4 Nd7 10.Qc2 f5 Black has a good version of a Stonewall, since White’s knight has developed so far away from the centre. 11.Rad1 g5 12.Bc1 This position occurred in Aronian – Radjabov, Geneva 2017. A good way for Black to complete development would have been:

685

12...a5!?N But not 12...b6?? 13.cxd5 cxd5 14.Qc6 winning. 13.Ne1 Transferring the knight to d3 is a thematic plan. The ambitious 13.Ne5 Nxe5 14.dxe5 b6 15.g4 Ba6 16.b3 Qd7 leads to a tense position, where Black’s pieces are placed more harmoniously. 13...Nd6 14.b3 Bf6 15.Nb1 The knight hardly has a better route into the game. 15...b5! The value of ...a5 becomes clear: if White swaps pawns and moves his queen to c6, the reply ...Ra6 will expel the invader. 16.Nd2 Ba6„ Black has a harmonious position with plenty of active prospects. B) 8.b3

686

It is slightly unusual to see this move so early but, on the whole, it is pretty thematic for these positions and may easily transpose to one of the lines in the next two chapters after a subsequent Qc2. 8...b6 9.Nc3 This is the most ambitious choice. With the c4-pawn protected, White is preparing the e2-e4 advance. 9.Qc2 Nbd7 10.Rd1 transposes to variation D of the next chapter. The other logical try is: 9.Bc3 White opts for a double fianchetto and intends to prepare e2-e4. This plan takes time though, so Black will be able to complete development and carry out the liberating ...c5 break. 9...Bb7 10.Nbd2 Nbd7 11.Re1 c5 12.Ne5 The knight enters the centre and unblocks the powerful light-squared bishop. Other options are less ambitious, for instance: 12.cxd5 Nxd5 13.Bb2 Rc8 14.Rc1 cxd4 15.Nxd4 Rxc1 16.Qxc1 Qa8 17.Nc4 N5f6= was fine for Black in Inarkiev – Khairullin, Sochi 2007. 12.Bb2 Rc8 13.Rc1 dxc4 14.Nxc4 b5 (14...Rc7!?N also deserves attention, intending ...Qa8 with a harmonious set-up) 15.Nce5 Nxe5 16.Nxe5 Bxg2 17.Kxg2 Qd5† 18.f3 Rfd8 19.e4 Qb7 was absolutely equal in Yuffa – Dubov, Sochi 2017. 12...cxd4 13.Bxd4 Nxe5 14.Bxe5 Qd7 15.e4 This position was reached in Yuffa – Gelfand, Moscow 2017. There are few equally good options, but I prefer:

687

15...dxc4!?N 16.Nxc4 Rfd8 17.Qe2 Qd3 Black has excellent play along the d-file.

9...Ba6 10.Bf4 Nbd7 White has tried a wide choice of moves from here. 11.Qd3!? The self-pinning move makes some sense, as the queen protects the c4-pawn while preparing e2-e4. 11.Qc2 Rc8 12.e4 dxc4 13.Rfd1 transposes to variation D32 of Chapter 19. 11.Nd2?! is well met by: 11...Nh5 12.Be3 Rc8 (12...f5!? is also logical) 13.a4 f5 14.cxd5 exd5 15.Nf3 This occurred in Li Chao – Wojtaszek, Huaian 2016. Black could have continued his expansion on the 688

kingside with:

15...g5!N Black has good prospects, for instance: 16.a5 bxa5 17.Ne5 Nxe5 18.dxe5 Ng7³ 11.Rc1 Rc8 12.Qd3 c5 13.Nb5 has occurred in a few games. I like the following natural way of handling the position:

13...dxc4N 14.bxc4 Bb7! The bishop has done its job on a6, so now it drops back to a more useful diagonal. 15.Nxa7 (15.Rfd1 cxd4 16.Qxd4 Rc5=) 15...Ra8 16.Nb5 Rxa2

689

17.Ra1 (17.Qb3 Qa8 18.Nc7 Qa4=) 17...Qa8! Black conveniently activates his queen with good chances. For instance: 18.Rxa2 Qxa2 19.Bd6 Be4 20.Qc3 Bxd6 21.Nxd6 Bc6=

11...c5 12.a4 This is the most ambitious. The appearance of White’s knight on b5 certainly has the potential to cause problems. 12.Rac1 Rc8 transposes to the 11.Rc1 line noted above. 12.Nb5 can be neutralized by 12...Bb7 13.Bc7 Qc8, when White has nothing better than repeating with: 14.Bd6 Qd8=

690

12...dxc4!N I find it important to open the long diagonal. The timid 12...cxd4?! yielded White a powerful initiative after 13.Nxd4 Rc8 14.Ndb5 in Polugaevsky – Timman, Breda 1979. 13.bxc4 cxd4 14.Nxd4 Nc5 Black gains an important tempo which enables him to consolidate. 15.Qd1 Bb7

691

16.Nc6 Bxc6 17.Bxc6 Rc8 18.Bf3 Qxd1 19.Rfxd1 Nfd7= White’s bishop pair and active pieces provide enough compensation for his damaged structure, but nothing more. C) 8.Nc3

This move is quite committal, since it makes it tougher for White to regain the pawn after Black’s reply. 8...dxc4 8...Nbd7 is another common continuation but I see no reason to allow White to protect the pawn. 9.a4 I also considered: 9.Bg5 Nd5 10.Bxe7 Qxe7 11.Ne5 (11.a4N 11...a5 12.Ne5 Nb6 13.b3 cxb3 14.Rb1 Nd5 15.Qxb3 Nd7 16.Ne4 Rb8³) 11...b5

692

12.a4!?N (after 12.Ne4?! f6 13.Nf3 Nd7µ White had no compensation for the pawn in Fehr – Tochtrop, Dortmund 2005) 12...Nxc3 13.bxc3 Bb7 14.Qb1 a6 15.axb5 cxb5 16.Bxb7 Qxb7 17.Nxc4 White manages to regain the pawn, but his pawn structure is inferior, so after 17...Qc7 18.Ne5 Nd7 Black’s chances are preferable.

9...Nbd7 A solid alternative is: 9...c5!? 10.dxc5 10.Bg5 would prove toothless after 10...h6 11.Bxf6 Bxf6 12.dxc5 Nd7 13.Nd2 Nxc5 14.Nxc4 Bd7 with excellent play. 10...Nc6 11.Qc2N White needs to improve on 11.Bf4?! Bxc5 12.Nd2 e5 13.Bg5 Be6 14.Bxf6 gxf6 when Black 693

enjoyed the bishop pair as well as an extra pawn in Miton – Gajewski, Lublin 2008. Also harmless is 11.Bg5 Bxc5 12.Nd2 as in Li Chao – Vitiugov, St Petersburg 2012. At this point I suggest 12...h6N 13.Bxf6 Qxf6 14.Nce4 Qe7 15.Nxc5 Qxc5 16.Qc2 Nd4 17.Qxc4 Qxc4 18.Nxc4 Nxe2† 19.Kh1 Nd4 when White’s compensation definitely doesn’t yield more than equality. This new way of handling the position seems like White’s best idea. Now d1 is vacant for the rooks, while the queen gets ready to capture on c4.

11...Bxc5 12.Rad1 12.Ne4 Be7 13.Qxc4 is hardly better in view of: 13...Qd5! 14.Nxf6† Bxf6 15.Qxd5 exd5= 12...Be7 13.Bf4 Qa5 14.Nd2 Nd4 15.Qb1 Nd5 16.Qe4 Nxf4 17.gxf4 Nc6 18.Nxc4 Qb4 Black’s bishop pair is balanced by White’s active pieces, and the chances are equal. 10.e4 White builds a powerful pawn centre and restricts the mobility of Black’s pieces. 10.Qc2?! a5 11.e4 offers Black an improved version of the main line after:

694

11...e5! Further analysis of this position can be found on page 365, when we arrive here via the 8.Qc2 move order. 10.a5 This slower way of handling the position enables Black to attack the centre with: 10...c5! 11.e3 Another natural continuation, 11.Bf4N, is well met by: 11...Nd5 12.Qd2 cxd4 13.Qxd4 Nxf4 14.Qxf4 Rb8 15.Rfd1 b5 16.axb6 Qxb6= This occurred in Sumets – Mchedlishvili, Fujairah 2012. I recommend the following new way of handling the position:

11...Rb8!N Black consistently removes his pieces from the long diagonal. 695

12.Qa4 After 12.Ne5 Nxe5 13.dxe5 Nd5 14.Qe2 b5 15.axb6 axb6 16.Qxc4 Nxc3 17.Qxc3 Bb7³ Black has the better pawn structure. 12...cxd4 13.exd4 a6 14.Qxc4 b5 15.axb6 Rxb6 16.Na4 Rb5 17.Rfc1 Bb7 Black has excellent prospects against White’s weak pawns.

10...e5!N This counter-sacrifice is aimed at activating the minor pieces. Although it’s a novelty here, the same idea is commonplace in the similar position with the bishop on c1 instead of d2, and indeed we soon transpose to several such games. In the event of 10...b6 11.Qe2 Ba6 12.Rfd1 Black manages to maintain the extra pawn, but his position is too passive for my taste. 11.dxe5 11.Nxe5? would be a serious mistake: 11...Nxe5 12.dxe5 Ng4 13.Bf4 g5µ 11...Ng4 12.Bf4 The dynamic potential of Black’s set-up is well illustrated in the following line: 12.e6 Nde5 13.Nxe5 Nxe5 14.exf7† Rxf7 15.Be3 Qxd1 16.Raxd1 Bg4„ After the text move we transpose to a large number of games in which Black’s bishop went directly from f8 to e7 and White’s from c1 to f4.

696

12...Qa5 13.e6 The following moves are less challenging: 13.Qd4?! Rd8 14.Qxc4 Ndxe5 15.Nxe5 Nxe5 16.Qb3 Ng6 17.Be3 Be6 18.Qc2 Ne5µ gave Black a clear positional superiority in Wojtaszek – Koneru, Lausanne 2006. 13.Qd2 Bb4 14.e6 fxe6 15.Bd6 Bxd6 16.Qxd6 Nde5 17.Nxe5 Nxe5 was better for Black in Hausner – Kramnik, Germany 1992. 13.Qe2 Ndxe5 14.Nxe5 Nxe5 doesn’t look attractive for White either, even though he can regain the pawn:

697

15.Nd5 cxd5 16.Bxe5 dxe4 17.Bxe4 Be6 18.Bxb7 Rad8 White was suffering from a lack of harmony and a weak b2-pawn in Mesquita – De Souza, corr. 2006. The text move is critical. White gives up the pawn in a favourable situation, damaging Black’s kingside pawn chain.

13...fxe6 14.Qe2 14.e5 has been tried several times but is not dangerous. After 14...Ndxe5 15.Nxe5 Nxe5 16.Qh5 Bd6 White has nothing better than a draw by repetition, for instance:

17.Ne4 (or 17.Rfe1 Rf5 18.Qe8† Rf8=) 17...Rf5 18.Qe8† Rf8 19.Qh5 ½–½ Topalov – Kasparov, Sarajevo 2000.

698

14.Bh3!? This was a recent invention, seen in Sarana – Dreev, Yaroslavl 2018. A natural improvement over Black’s play is: 14...Nge5!N 15.Bxe6† Kh8

16.Nxe5 White also fails to achieve much after: 16.Nd2 Qb4! 17.Rb1 Nd3 18.Na2 Qc5 19.b4 Qd4 20.Be3 Qf6 21.Bxc4 N7e5 Black has sufficient compensation for the pawn, to say the least. 16...Nxe5 17.Bxc8 Raxc8 18.Bxe5 18.Nd5 cxd5 19.Bxe5 dxe4 20.Qd4 Bf6 21.Bxf6 Rxf6 22.Qxe4 Qb4 offers Black comfortable equality. 18...Qxe5 19.Qe2 Rcd8

699

20.Rad1 20.Qxc4?! Rd2 gives Black too much activity. 20...b5 21.axb5 cxb5 22.Nd5 Bc5 23.Kg2 a5 24.f4 Qe8 We reach an unbalanced position where Black’s chances are by no means worse; the queenside pawns may be supported by the powerful bishop.

14...Nde5 15.Nxe5 Nxe5 16.Bxe5 Qxe5 17.Kh1! This prophylactic move is definitely the best. White’s chances are mainly connected with advancing the kingside pawns. 17.Qxc4?! is premature. 17...Qc5 18.Qb3 Kh8 19.Rac1 was seen in Hafner – Macs, corr. 2012, when Black should have played:

19...Qh5!N 20.f4 e5 21.f5 Bc5† 22.Kh1 b6 23.Qc2 a5³ Black is better due to his powerful bishops. 700

17...Qc5 18.e5 In return for the sacrificed pawn, White has managed to seize a lot of space and block the light-squared bishop for a while.

18...Bd7 19.f4 Rad8 Black completes the development of his queenside pieces. 20.Ne4 20.Rfc1 Be8 21.Ne4 leads to the same thing. 20...Qb4 21.Rfc1 Be8! 21...b5?! 22.axb5 cxb5 23.Rxa7 Bc6 24.Rc7 offers White a strong initiative due to his control of the 7th rank. The text move improves Black’s harmony, and losing the extra pawn isn’t such a big deal. 22.Rxc4 Qb3 23.Rc3 Also after 23.Bh3 Bf7 24.a5 a6= Black is doing well. 23...Qb4 24.a5 I also considered: 24.Qc2N 24...Bg6 25.Rb3

701

25...Rd2! 26.Rxb4 Rxc2 27.Rxb7 Rxg2 28.Kxg2 Bxe4† 29.Kh3 Bc5 The powerful bishops put White in a difficult situation.

24...Bg6 25.Rc4 The immediate 25.a6N also allows the rook’s centralization: 25...Rd4 26.Nd6 Bxd6 27.exd6 Rxd6 28.axb7 Qxb7 29.Rca3 Qb5= 25...Qb5 26.a6 Rd4 27.b3 bxa6 28.Bf3 c5= Black had no problems and a drawn endgame soon ensued in Rydholm – Loinjak, corr. 2014. D) 8.Bf4 702

This move was first seen in 1985. White avoids spending a tempo moving his queen and gets ready to place a rook on c1. 8...dxc4 This seems like the most challenging reaction. If Black develops routinely, the c-file might become a significant factor, for instance: 8...b6 9.Nc3 Bb7 10.Rc1 Nh5 (10...Nbd7 11.cxd5 cxd5 12.Nb5 offers White slight but lasting pressure along the c-file) 11.Be5 Nd7 12.cxd5 exd5 (12...cxd5 13.Nb5 Nxe5 14.dxe5 leaves the knight on h5 poorly placed) 13.e4! f6? (the lesser evil was 13...Nhf6 14.exd5 Nxe5 15.dxe5 Nxd5 16.Nxd5 cxd5 17.Nd4²) 14.exd5 cxd5 15.Bc7! Qxc7 16.Nh4 Black was crushed in Gelfand – Aronian, Monte Carlo (blindfold) 2007. 9.Ne5 9.Nc3?! b5 leaves White with no real compensation for the pawn, unless he tries 10.Bxb8 Rxb8 11.Ne5 Bb7 12.Nxc6 Bxc6 13.Bxc6 b4µ, in which case he has only traded a material disadvantage for a positional one. 9.a4 Nd5 10.Bd2 seems pointless, as it doesn’t prevent Black’s plan: 10...b5 11.axb5 cxb5 12.Ba5 Qe8 13.e4 Nf6 White did not have much for the pawn in Heinke – Eglseder, email 2008 The text move enables White to regain the pawn, but his bishop will soon be exchanged.

703

9...Nd5 10.Nxc4 Nxf4 10...Nd7?! needlessly allows 11.Bd2². 11.gxf4 Nd7 In this interesting position, both sides have their merits. White has a strong centre and a definite space advantage, while Black’s position lacks any weaknesses and his bishop pair might prove significant in the long run. His main challenge is developing the light-squared bishop. We will aim to solve that issue by arranging a fianchetto over the next few moves.

12.e3 White will have to protect the f4-pawn at some point, so starting with this move seems most flexible. 704

12.Nbd2 Qc7 13.e3 transposes to 13.Nbd2 in the note to the main line below. 12.Nc3 gives Black a few options: a) the simplest choice, for our purposes, is 12...Qc7 13.e3 leading to our main line. b) A secondary idea is 12...Nb6!? 13.Ne5 f6 14.Nf3 Nd5 15.e3, as played in Bruzon – Karjakin, Tsaghkadzor 2015, as well as some correspondence games. I suggest a new concept:

15...Nxc3!?N 16.bxc3 Rb8 with a typical double-edged middlegame ahead. 12...Qc7 13.Nc3 13.Nbd2 diminishes White’s central control, and 13...b6 14.Rc1 Bb7 15.Nf3 Rad8 16.Qa4 c5 was equal in Scheidig – Guzy, corr. 2006. 13.Qc2N hardly has any independent value, and only offers Black extra options. 13...b6 (13...b5!? 14.Ne5 Nxe5 15.fxe5 Bb7 16.Nd2 c5 17.Bxb7 Qxb7 18.dxc5 Qd5„ is another tempting idea) 14.Ne5 Nxe5 15.dxe5 Bb7 Black doesn’t experience any problems, for instance: 16.Nc3 a6 17.Rfd1 Rfd8 18.Rac1 c5= 13...b6 14.Rc1 The opposition of rook and queen may seem annoying for Black, but White cannot exploit it at this stage. I also examined 14.a3N, intending to seize more space on the queenside. Black should respond with 14...a5 15.Rc1 Ba6! 16.Ne4 c5, achieving excellent play. The attempt to stop ...c5 with 14.Ne5 Bb7 15.Qf3 was seen in Yuffa – Inarkiev, Chelyabinsk 2016. In 705

my opinion, Black has no reason to leave the powerful knight on e5, and should have played:

15...Nxe5N 16.fxe5 Rab8 Intending ...c5. 17.Qg3 Kh8! A useful prophylactic move. 18.Rad1 (18.f4 is well met by the thematic 18...f5!) 18...Qd7 19.Kh1

19...f5! This thematic move limits any further advance of White’s central pawns and reminds him of his somewhat exposed king. 20.f4 Rg8 21.e4 Rbf8 A complex struggle lies ahead, with Black’s chances being no worse. 14...Bb7 15.Qc2 I checked two other queen moves: 15.Qe2 has its drawbacks, as can be seen after: 15...Rac8!N (15...Rad8 proved less precise after 16.Rfd1 706

c5 17.d5 exd5 18.Bxd5 Nf6 19.Bf3 in Short – Agdestein, Helsingor 2017) 16.Rfd1

16...Ba6! 17.b3 Rfd8 18.Qf3 Nf6 19.Ne5 c5 20.dxc5 Qxc5 21.Rxd8† Bxd8 22.Bf1 Bxf1 23.Kxf1 Qa3= 15.Qf3 should be met by: 15...Rab8!N (it makes no sense to leave the c-pawn pinned, as occurred after 15...Rac8 16.Rfd1 Rfd8 17.a3 Ba8 18.Qg3 Nf6 19.Ne5 and White was better in Y. Vovk – Kravtsiv, Lvov 2017) 16.Qg3 (16.Qh3 Rfd8 17.Rfd1 b5 18.Ne5 Nxe5 19.fxe5 c5=)

16...f5! 17.Ne5 (the careless 17.Rfd1?! invites 17...Rf6! with serious threats on the kingside) 17...Nxe5 18.fxe5 Qd7 19.Rfd1 Kh8 Black has interesting prospects on the kingside, based on the ...g5 advance. 15...Rac8 Black can also consider 15...Rad8!?, when 16.Rfd1 transposes to a line analysed later – see page 390, 707

where 16...c5 is nice for Black. 16.Rfd1 Rfd8 After a series of mostly natural developing moves, both sides will have to make more complex decisions in the next phase.

17.a3 17.f5?! is premature in view of 17...e5 18.d5 cxd5 19.Nxd5 Bxd5 20.Rxd5 Nf6 and Black was slightly better in Fey – Mass, Oberhof 1999. 17...Nf8! Most games from this position have continued 17...a5, but I see no reason to weaken the queenside pawn chain. Instead, the knight heads to g6 in order to secure Black’s kingside and threaten that of the opponent. 18.b4 This seems most consistent. In the event of 18.Ne5N 18...Ng6 19.Qe2 Nxe5 20.fxe5 c5 Black successfully activates his pieces:

708

21.Bxb7 Qxb7 22.d5 exd5 23.Qf3 Qc6 24.Rxd5 Rxd5 25.Nxd5 Qe6= The aggressive 18.f5 has been played a few times. I like the following reaction: 18...c5 19.d5

19...Ba6! It is important to drive the strong knight away from the centre. 20.Nd2 (I checked 20.Bf1N but found it to be toothless after: 20...b5! 21.d6 Bxd6 22.Rxd6 Rxd6 23.Nxb5 Bxb5 24.Nxd6 Qxd6 25.Bxb5 Qe5 26.fxe6 Nxe6=) 20...exf5 21.Qxf5 Ng6 22.Qh3

709

22...Rb8! All this happened in Mislin – Conde Poderoso, corr. 2017. Black’s last move vacated the c8square for the bishop. Black’s bishop pair and safer king fully compensate for White’s superiority in the centre. 18...Ng6 19.Be4 f5 20.Bh1 c5 21.Nb5 The ambitious 21.d5!?N leads to interesting complications:

21...cxb4 22.d6 Qxc4 23.dxe7 Nxe7 24.Bxb7 bxc3 25.Bxc8 Rxd1† 26.Rxd1 Qxc8 Black is out of danger thanks to the strong passed pawn, which will be securely protected by the knight coming to d5. 21...Qb8 22.Bxb7 Qxb7

710

23.dxc5 White has managed to open the d-file and gain access to d6, but Black’s counterplay comes in time. 23...Qf3! 24.Rxd8† Rxd8 One way or another, the game is heading for perpetual check, so a draw was agreed in Glorstad – Teeriaho, corr. 2011. E) 8.Qb3

Compared to the more popular 8.Qc2 variation, this move enables White to follow up with Nc3 without losing a pawn. On the other hand, the queen blocks White’s b-pawn and does not help to control 711

e4. In the 1980s, White’s set-up was successfully employed by Bareev, Khalifman and others, but nowadays it is a rare guest at the top level. With that being said, Fabiano Caruana used it twice in the 2018 Candidates tournament, and other strong GMs have also tested it from time to time. 8...b6 8...Nbd7 is a common alternative but I find the text move more flexible. In the event that White meets the imminent ...Ba6 with cxd5, Black may benefit from the option of placing the knight on c6. 9.Nc3 This is the most natural and consistent move. White has tried several other ideas: I was surprised to discover that 9.Rc1?! has been played by a few GMs, and that nobody opted for the critical reply: 9...dxc4!N 10.Qxc4 Ba6 11.Qa4 Bxe2 12.Nc3 Bxf3 13.Bxf3 Qd7³ White does not have full compensation for the pawn. A slightly better version of the above idea is: 9.cxd5 cxd5 10.Rc1 White is first to occupy the open file but the majority of his pieces are rather passive, so it shouldn’t bother Black.

10...Bb7 11.a4!? Quite an interesting concept: White intends to push the pawn all the way to a6, securing the b5square for the knight along with a lot of space on the queenside. 11.Bb4 proved harmless after 11...Nc6 12.Bxe7 Qxe7 13.Nc3 Rac8 14.e3 Na5 15.Qd1 Nc4 16.b3 Nd6= in Ki. Georgiev – Ladva, Puhajarve 2013. I also considered: 11.Ne5 Nfd7! Black wants to get rid of the annoying knight, having in mind that the b8-knight will be perfectly placed on c6. 12.Nd3 (12.Nxd7 Qxd7 13.Bb4 Nc6 14.Bxe7 Qxe7 712

was equal in Rakotomaharo – Zhu Chen, Tromso [ol] 2014) 12...Nc6 13.e3 Rc8 14.Na3 Nf6 15.Be1 Ne4 16.Rc2 Nd6 17.Rac1 This position occurred in Ki. Georgiev – Pikula, Veliko Gradiste 2018. After 17...Qd7N 18.Qa4 Rfd8 Black reaches a harmonious set-up, so White has nothing better than inviting simplifications: 19.Nb4 Nxb4 20.Qxd7 Rxd7 21.Bxb4 Rxc2 22.Rxc2 a6=

11...Ne4 The following game illustrates the potential potency of White’s concept: 11...Nc6 12.a5 Rc8 13.a6 Ba8 14.Na3 Ne4 15.Be1 Qd7 16.e3 Bxa3?! 17.Qxa3 Rc7 18.Bf1 White was clearly better in Nikolic – Panchenko, Chelyabinsk 1989. 12.Be1 After 12.Ne5 Nxd2 13.Nxd2 a5! the drawback of the earlier a2-a4 is obvious: the b4-outpost belongs to Black’s pieces. 14.e3 Bd6 15.Nd3 Nc6= 12...Nd7N This natural developing move improves over 12...Nd6 13.Ne5 Nd7 14.Nc6 Bxc6 15.Rxc6 when White had a small but stable edge in Bareev – Chernin, Pula 1988. 13.a5 13.Nc3 is again well met by 13...a5! 14.Nb5 Rc8 15.Qd1 Ba6 with no problems for Black.

713

13...Ba6! 14.e3 Qb8 15.Nc3 bxa5! Releasing the tension at a good time. 16.Qxb8 Rfxb8 17.Rxa5 Bd3= Black has no problems. A somewhat similar plan to the line above is: 9.Ne5 Ba6 10.cxd5 10.Bc3 was recently tried in Fedoseev – Gelfand, Poikovsky 2018. My preferred plan is 10...Qc7N to cover c6, so that the b8-knight can be developed. 11.Nd2 Nbd7 12.Qa4 (12.Nxd7 Qxd7 13.Rfd1 Rfd8 14.Rac1 Rac8 15.e3 c5= is harmless for us) 12...Bb7 13.e4 dxc4 14.Ndxc4 b5 15.Qa5 White has managed to seize the centre while avoiding a loss of material, but after 15...Qc8! 16.Nxd7 Nxd7 17.Ne3 Bd8 18.Qa3 a5 Black has excellent counterplay due to the poor location of the queen on a3. 10...cxd5

714

11.Rc1!? 11.Nc3 transposes to our main line on page 351. The text move is an interesting pawn sacrifice which deserves attention. 11...Bxe2 12.Nc3 Ba6 13.a4N The weaker 13.Bg5?! Bb7 14.Nb5 was seen in Chabanon – J. Horvath, Paris 2003. Black should have played 14...Ne8!N 15.Bxe7 Qxe7 when White’s compensation for the pawn is questionable.

13...Bb7 14.Nb5 Ne8 15.Rc2 White’s activity on the queenside might become dangerous, so it makes sense to simplify matters: 15...a6! 16.Nc7! Nxc7 17.Qxb6 Nb5! 18.Qxd8 Rxd8 19.axb5 axb5 Black is fine.

715

White has also tried a few bishop moves, beginning with the harmless 9.Bb4. Black has several decent way of handling the position, the most challenging being 9...c5 10.dxc5 Na6!, exploiting the poor placement of White’s bishop.

11.cxd5 Nxd5 12.Ba3 Nxc5 13.Qc2 Bb7 14.Rd1 We have been following Azmaiparashvili – Gelashvili, Kallithea 2003. My new idea is 14...Rc8N 15.Nc3 Qc7 16.Rac1 Qb8, connecting the rooks and obtaining a flexible set-up with comfortable equality. 9.Bg5 Ba6 10.Nbd2 Protecting the c4-pawn in this way is a concession of sorts, as White’s knight does not put much pressure on Black’s centre.

10...Nbd7 11.Rfe1 Bb7! 716

11...c5?! is premature. 12.cxd5 exd5 occurred in Tukmakov – A. Ivanov, New York 1990, when 13.Ne5!N 13...Bb7 14.Nxd7 Qxd7 15.dxc5 bxc5 16.Rad1² would have put the hanging pawns under serious pressure. 12.e4!?N After 12.Rac1 c5 Black had excellent play in Miton – David, Haguenau 2013. The text move is more consistent.

12...dxe4 13.Nxe4 Nxe4 14.Bxe7 Qxe7 15.Rxe4 c5 16.d5 Qd6 17.Rd1 17.Nh4 proves ineffective after 17...g6!, covering the f5-square. 18.Re3 exd5 19.Rd1 Nf6 20.cxd5 Rfe8³ The knight is clearly misplaced. 17...exd5 18.cxd5 Nf6 19.Ree1 Rfe8= The d-pawn is firmly blockaded and might even become weak in the long run. A final alternative line is: 9.Bf4 Ba6 10.Nbd2 10.cxd5 cxd5 11.Nc3 Nc6 12.Rfc1 Na5 13.Qd1 Rc8 14.Ne5 Bb4 was equal in Cebalo – Nogueiras Santiago, Reggio Emilia 1985. 10...Nbd7 11.Rfd1 After 11.Rfe1?! c5 12.Qa4 Bb7 the poor location of White’s queen enabled Black to seize the initiative in Chloupek – Kholmov, Mlada Boleslav 1995.

717

11...Rc8!N For some reason, this natural move has not been played yet. 11...c5 is somewhat premature as, unlike the 11.Rfe1?! line above, the rook on d1 is well placed for any opening of the centre. After 12.Rac1 Qe8 13.Qa4 Bb7 14.Ne5² Black was under some pressure in Aronian – Beliavsky, Warsaw 2005. 12.Rac1 Nh5 13.Be3 Now the bishop is blocking the e-pawn, so e2-e4 is impossible. 13...h6 Inviting White to show his cards. 14.Ne5 Nhf6 15.Qa4 Nb8 16.Bf4 b5 17.cxb5 cxb5 18.Qb3 Qb6 Black has comfortable play.

718

9...Ba6 10.cxd5 cxd5 11.Ne5 This is the most aggressive continuation. The knight occupies the centre and takes control over the c6spot, which is especially important when Black’s bishop has gone to a6. 11.Rfc1 This move has scored well but it does not prevent Black from completing development with:

11...Nc6! 12.Nxd5 Other moves are toothless, for instance: 12.Bf4 Rc8 13.Qa4 (after 13.a4?! Ne4 14.Nb5 Qd7 15.Ne5 Nxe5 16.Bxe5 Nd2 17.Qd3 Nc4³ White’s queenside set-up was rather dicey in Van Wely – A. Ivanov, Tripoli 2004) 13...Bb7 14.Ne5 Nxe5 15.Bxe5 a6 Black had comfortable play in Shomoev – Smirnov, Kirov 2012. 12...Qxd5 13.Qxd5 Nxd5 14.Rxc6 Bxe2 15.Ne5 Bb5 16.Rc2 We have been following L.B. Hansen – Muir, Gothenburg 2005. A natural improvement for Black is:

719

16...Rfc8N 17.Rac1 Rxc2 18.Rxc2 Kf8= Black is fine, as the following transformation does nothing for White: 19.Bxd5?! exd5 20.Rc7 Ke8³ White cannot make use of the 7th rank and has no compensation for the enemy bishop pair.

11...Bb7 Moving the bishop twice looks like a concession, but provoking the earlier cxd5 exchange has made Black’s position easier to handle, so the loss of tempo was not in vain! It is worth mentioning that the careless 11...Nfd7? leads to a quick disaster after 12.Nxd5! exd5 13.Nxf7 Rxf7 14.Bxd5, as occurred in a couple of games.

720

We will analyse E1) 12.Rac1 and E2) 12.Rfc1. There may not appear to be a big difference between them, but in fact the choice of rook has a significant bearing on how Black should continue. 12.Bf4 is less challenging, and 12...Nc6 13.Nxc6 Bxc6 14.Rfc1 Qd7 15.a4 Rfc8 was fine for Black in Kallai – Magerramov, Baku 1986. E1) 12.Rac1

12...Nbd7 Unlike variation E2 below, this quiet developing move enables Black to solve his problems. 12...Nfd7?! is too risky in view of 13.Nxd5! exd5 14.Bxd5 Bxd5 15.Qxd5 Nxe5 16.Qxa8 Qxd4 17.Bc3 Qd6 18.Qxa7 and White was clearly better in Ray – Anderson, corr. 2010. I also considered 12...Nc6 13.Nxd5 Nxd4 14.Nxe7† Qxe7 15.Qc4 Bxg2 16.Kxg2 Rfd8 17.f3 when White obtained slight but long-lasting pressure in Tyulenko – Grebenshchikov, corr. 2011. 13.Nd3 I also checked 13.Nb5N 13...a6! 14.Nxd7 Nxd7 15.Nc3 b5= when Black has no worries. The text move keeps more tension in the position but White’s rooks are less harmonious than they would have been on a1 and c1. In particular, there is no plan of advancing the a-pawn. We have been following Perelman – Del Cuadro, Buenos Aires 2004. In the game Black proceeded with 13...Ne4 followed by ...Nd6, but I find it more logical to reroute the other knight with:

721

13...Nb8!?N 14.Nb5!? The most aggressive. 14.e3 Nc6 15.Rfd1 Qd7 16.Qa4 Rfc8= is harmless. 14...Ne4 15.Bf4 15.Bb4 Nc6 16.Bxe7 Qxe7 17.Bxe4 dxe4 18.Ne5 Rfc8= is safe enough for Black. 15...Nc6 16.Ne5 Na5 17.Qd3 a6

18.Nc7!? 722

I also considered: 18.Nc3 f6 19.Ng4! (the more obvious 19.Nf3?! Rc8 20.Nd2 f5³ leaves White in a passive position) 19...Nxc3 20.Rxc3 Rc8

21.Rfc1 (21.Rxc8 Bxc8 22.Rc1 Nc4 23.b3 Nd6=) 21...Rxc3 (also possible is 21...Rc4 22.Ne3 Rxc3 23.Rxc3 Bd6 24.Bxd6 Qxd6=) 22.Rxc3 Nc6= White’s pressure along the c-file is neutralized. The text move invites interesting complications after: 18...Rc8! 19.Nxa6 The tempting 19.Nxf7?! doesn’t work: 19...Kxf7 20.Bh3 Ng5 21.Bg4 Nc6 22.h4 Rxc7 23.Bxc7 Qxc7 24.hxg5 g6³

19...f6 723

White has won a pawn, but the poor location of his knight on a6 offers Black sufficient counterplay. 20.Rxc8 Bxc8 21.Ng4 Nc6 22.Nb8! Avoiding 22.Rc1 Nxd4 23.Qxd4 Bxa6 24.Bc7 Bc5 25.Qxc5 Nxc5 26.Bxd8 Rxd8³ when White will suffer in the endgame. 22...Nxb8 23.Bxb8 Bd7 24.Bf4

24...e5 25.Be3 Bxg4 26.f3 exd4 27.Qxd4 Bc5 28.Qd3 Bh5 29.fxe4 dxe4 30.Qb3† Bf7= The Catalan bishop has been blocked, so Black has comfortable play. E2) 12.Rfc1

724

This move illustrates White’s intention to play mostly on the queenside, where his major pieces are located. Moreover, the f1-square is now vacant for the light-squared bishop. 12...Nfd7! My analysis indicates that this standard reply is more effective with White’s second rook on a1 rather than c1. 12...Nbd7 13.Nd3 Nb8 was played in Alburt – Tal, Taxco 1985, when 14.e3N 14...Nc6 15.Qa4 Rc8 16.b4 would have left Black under slight but long-lasting positional pressure. 12...Nc6 was tested more recently at the top level13.Nxd5 Nxd4 14.Nxe7† Qxe7 15.Qc4 Bxg2 16.Qxd4 Rfd8 17.Qf4 White had a slight edge and almost managed to convert it into a full point in Caruana – Ding Liren, Berlin 2018. 13.Nxd5!? This sharp continuation is the best attempt to highlight the drawbacks of leaving the b8-knight on its initial spot. Quieter methods do not seem to bother Black. For instance, 13.Nd3 Nc6 14.e3 Rc8 15.a4 was seen in Huang Qian – Guo Qi, Xinghua 2018, when the following regrouping plan would have been logical:

725

15...Qc7!N 16.Qd1 Qb8 17.Be1 Rfd8 18.Nb5 a6 19.Nc3 h6 With roughly equal chances. 13...exd5 Of course, not 13...Nxe5? 14.Nxe7† Qxe7 15.Bb4+– when Black loses the exchange. 14.Bxd5 Bxd5 15.Qxd5 The point of White’s small combination is that the rook on a8 is trapped.

15...Nxe5 16.Qxa8 Qxd4 17.Bc3 The analogous position with White’s rook on f1 instead of a1 has already been evaluated as bad for Black. In the present version, White’s king is more exposed, so Black should continue: 726

17...Qc5! I also examined 17...Qd6N 18.Qxa7, which leads to an unbalanced position offering interesting play for both sides. With that being said, White’s chances are slightly preferable after accurate play.

For instance: 18...Ng4 19.Qa4 Qe6 20.f3! (20.Bd4 Qh6 21.h4 Bxh4 gives Black enough of an attack to force a draw at least) 20...Bc5† 21.Bd4 b5 22.Qd1 Ne3 23.Qd3²

18.Qxa7!?N Taking another pawn seems the most principled choice. 18.Qe4 f6 19.Bxe5 Qxe5 20.Qxe5 fxe5 21.Rc7 Bc5 was seen in Sloth – Morgado, corr. 1995, at which point a draw was agreed.

727

Indeed, after 22.e3 Rf7 23.Rc8† Rf8 White has to repeat moves, as the endgame after 24.Rxf8†?! Kxf8 25.Kg2 Ke7 offers winning chances only to Black. My other idea was 18.b4N 18...Qb5 19.Qe4, trying to exploit the temporary lack of harmony in Black’ camp. However, a good answer is:

19...Ng6 20.a4 Qc6! Black is just in time to consolidate and start attacking White’s queenside pawns. For instance, 21.Qxc6 Nxc6 22.b5 Nd8 The knight heads towards a solid outpost on c5. 23.Rd1 Ne6 24.Rd7 Ra8 25.a5 Bc5 Black has nothing to worry about.

728

18...Ng4! Now we see the difference! With all of White’s pieces on the far side of the board, White’s king is rather vulnerable. 19.Be1 19.Rf1 is equally good, with 19...Nc6 20.Qd7 Qh5 21.h4 Qg6 giving Black sufficient play on the kingside. An illustrative line is:

22.Rad1 h5 23.Rd5 Ne3 24.fxe3 Qxg3† 25.Kh1 Qxh4† 26.Kg1 Qg3†= With perpetual check.

729

19...Qh5 20.h4 Bc5 21.Qb7 21.Qa4 Ne5 22.Qe4 leads to the same.

21...Ne5 22.Qe4 Nbd7 White’s potential passer on the queenside might be dangerous, so Black has to play energetically. 23.Rc2 Nf6 24.Qf4 h6! Covering g5, thus preventing White from forcing the queens off.

25.b4 Nd5 25...Bd6!? 26.a4 Bb8 also offers Black enough counterplay. 730

26.Qe4 Nf6 27.Qb7 Bd4 28.Rd1

28...Qg6! The double attack is very effective. 29.Rcd2 Qxg3† 30.Qg2 Qxg2† 31.Kxg2 Bc3= The resulting endgame is drawish. Conclusion This chapter introduced the popular sequence of 4.Bd2 Be7 5.Bg2 d5 6.Nf3 0-0 7.0-0 c6, at which point we considered several interesting alternatives to the most common 8.Qc2. 8.Na3!? and 8.b3 are playable moves which aim to exert positional pressure in one way or another. Against the former, converting to a Stonewall set-up works well against White’s estranged knight. Against the latter, simple development followed by ...c5 offers Black equal play. 8.Nc3 is an ambitious choice, which has the drawback of allowing 8...dxc4. The main line continues 9.a4 Nbd7 10.e4 e5!, when Black temporarily returns his extra pawn in order to free his pieces. The ensuing positions can become complex and double-edged, but Black is in good theoretical shape. 8.Bf4 is another interesting choice, when 8...dxc4 9.Ne5 Nd5 10.Nxc4 Nxf4 11.gxf4 reaches a thematic, double-edged structure. Black should continue 11...Nd7 intending a queenside fianchetto, hoping to make use of the two bishops in the long run. The space-gaining ...f5 is worth remembering as an important motif in many lines.

731

Finally, we analysed 8.Qb3, which has both pluses and minuses compared with the more common development of the queen to c2. After 8...b6 we considered several options, the most critical being 9.Nc3 Ba6 10.cxd5 cxd5 11.Ne5 Bb7. At this point the play becomes rather concrete, with 12.Rac1 and 12.Rfc1 each requiring a precise reply with knowledge of the ensuing nuances.

732

A) 9.Bg5 359 B) 9.Rc1 361 C) 9.a4!? a5! 363 C1) 10.Na3 365 C2) 10.Rc1 366 D) 9.Rd1 b6 10.b3 Ba6 11.a4 c5 369 D1) 12.cxd5 371 D2) 12.a5 373 D3) 12.Na3 375

733

1.d4 Nf6 2.c4 e6 3.g3 Bb4† 4.Bd2 Be7 5.Bg2 d5 6.Nf3 0-0 7.0-0 c6 8.Qc2 As I mentioned in the previous chapter, this is White’s most popular move by a wide margin. The queen takes up a flexible position and guards c4, while making room for the rook to go to d1 or perhaps c1. 8...Nbd7 8...b6 is another popular move which often leads to the same main lines. Each move order has some minor pros and cons, but I eventually settled on the knight development. We have reached an important branching point, where White has several options. In this chapter we will analyse A) 9.Bg5, B) 9.Rc1, C) 9.a4!? and D) 9.Rd1. 9.Bf4 is the most popular move of all, and it will be analysed separately in the next and final chapter. 9.Nc3?! resembles variation C of the previous chapter, but this is a worse version for White, as there is no need to have the queen on c2 if the c4-pawn is being sacrificed anyway. Black gets an excellent game with: 9...dxc4 10.e4 (10.a4 a5 transposes to 10.Nc3?! dxc4 in the notes to variation C; see page 365, where it is shown that 11.e4 e5! gives Black the upper hand) 10...b5³ The game goes on, but White does not have enough play to justify the pawn sacrifice. 9.b3 has been played in lots of games but does not require special attention, as the ensuing play will either resemble or transpose directly to lines examined elsewhere. For instance: 9...b6 10.Bc3 (the most common 10.Rd1 leads straight to variation D) 10...Bb7 11.Nbd2 c5= This can be compared with the 9.Bc3 line in the notes to variation B of the previous chapter on page 336. The placement of White’s queen on c2 does him no favours here, as the rook will come to c8 soon. Black has an easy game; no further variations are needed. A) 9.Bg5

734

This move generally signifies that White does not mind ceding the bishop pair in order to carry out e2e4 as soon as possible. 9...h6 10.Bxf6 White switched plans with 10.Bf4 b6 11.Nbd2 in Banov – Spassov, Sofia 2011, when 11...Ba6N would have offered Black a good game. For instance:

12.Rfd1 (White achieves nothing with 12.e4 Nxe4 13.Nxe4 dxe4 14.Ne5 Nxe5 15.dxe5 Qc7 16.Qxe4 c5=) 12...Rc8 13.Rac1 g5! 14.Be3 Ng4 15.Qb3 c5„ Exploiting the poor location of White’s darksquared bishop. 10...Nxf6 11.Nbd2 735

White is ready for e2-e4, which will enable him to seize the centre and activate the Catalan bishop. On the flip side, the position opens, so Black’s bishops gain more potential.

11...b6 12.e4 12.Ne5 is inconsistent and after 12...Bb7 13.Rfd1 Rc8 Black is ready for ...c5. Therefore White prevents it, but only invites further troubles. 14.c5 bxc5 15.dxc5 Qc7 16.f4 a5 17.Nb3 This position occurred in Zeng Chongsheng – Wei Yei, Wuxi 2016. I suggest the following plan:

17...Rcd8!N Black would like to mobilize his central pawns, so he prepares ...Nd7 without losing the a5pawn. 18.Nd4 Nd7 19.Nxd7 Rxd7 20.e3 Re8³ 12...dxe4 12...dxc4!?N 13.Nxc4 Bb7 14.Rfd1 Qc7 also offers Black comfortable play. 736

13.Nxe4 Bb7 14.Rfd1 Qc7 15.Nxf6† The careless 15.Rd2?! c5 16.d5 exd5 17.Nxf6† Bxf6 18.cxd5 Qd6µ was clearly worse for White in Michalet – Jacobs, Cappelle-la-Grande 1989. 15...Bxf6 16.c5 Preventing ...c5 is White’s only challenging plan. 16...Rfd8 17.Rac1 We have been following Yanchenko – Tsoi, St Petersburg 2016. I recommend the following fresh way of handling the position:

17...bxc5!?N The game continuation of 17...Rd7 can be well met by 18.Ne5!N 18...Bxe5 19.dxe5 Rxd1† 20.Rxd1 Qxe5 21.b4 and White has more than enough compensation for the pawn. 18.dxc5 In the event of 18.Qxc5 Rd5 19.Qc2 Rad8 the d4-pawn is even more vulnerable than Black’s c-pawn. 18...Rxd1† 19.Rxd1 Rd8 The powerful unopposed dark-squared bishop fully compensates for Black’s structural drawbacks. B) 9.Rc1

737

Quite a tricky move: White is preparing for the queenside fianchetto, but Black has a more flexible way to improve his position. 9...a5! Avoiding 9...b6, when 10.a4! transposes to an unfavourable line mentioned on page 364; see 9...b6 10.Rc1! in the notes to variation C. The text move is an excellent reply. Black prevents any queenside expansion with b2-b4 and invites White to show his cards. 10.Bg5!? 10.a4 transposes to variation C2, so it only remains to check a few other bishop moves: 10.Bf4 can be met by 10...Nh5 11.Be3 Nhf6 when the bishop is clearly not well placed on e3, so White has nothing better than 12.Bf4 or 12.Bg5. The passive 10.Be1 is well met by 10...b5 (10...b6!? 11.cxd5 cxd5= is also good enough) 11.c5 Ne4 and Black is fine, for instance: 12.Nfd2?! (12.b4N looks better, when 12...axb4 13.Bxb4 e5 14.a4 maintains the balance) 12...f5 13.f4 g5! Black had some initiative in Urban – Macieja, Warsaw 2008.

738

10...h6 11.Bf4 Losing a tempo in this way is part of an interesting plan, which was demonstrated by Boris Gelfand in 2018. Giving up the dark-squared bishop is a clear concession: 11.Bxf6 Bxf6 12.Nbd2 This was Starostits – Volodin, Riga 2017, when 12...b5!N would have been logical. Play might continue: 13.cxd5 (after 13.c5?! e5³ Black’s unopposed dark-squared bishop starts to tell) 13...cxd5

14.e3 (14.e4 dxe4 15.Nxe4 Bb7 is comfortable for Black, as is 14.Qc7 Nb6 15.Ne5 Ba6=) 14...Ba6 15.Qc7 (15.Bf1 Rc8 16.Qb3 Qb6=) 15...a4 16.Bf1 Rc8 17.Qd6 Nb6= 11...Nh5 12.e3!? 739

White’s concept involves provoking Black to expose his king; thus, the reason for inducing ...h6 becomes clear.

12...g5! Accepting the challenge. 12...Nxf4?! 13.exf4 b5 14.c5 leaves Black in a passive position, where ...h6 has clearly loosened his kingside. White is free to develop play along the e-file and on the kingside, while benefiting from his considerable space advantage. 13.Be5 Ng7 13...f6?! 14.cxd5! fxe5 15.dxe6 Nb6 16.Qg6† Ng7 17.Nxg5 hxg5 18.Be4 offers White a dangerous initiative, with a guaranteed draw by perpetual if he wants it. 14.g4 I also examined: 14.Bxg7N 14...Kxg7 15.e4 15.Nc3 f5 is pleasant for Black, who controls the dark squares and may consider further expansion on the kingside, whereas White’s intentions are unclear. The text move is the only real attempt to exploit the weaknesses created by ...g5, but the danger for White is that Black’s dark-squared bishop becomes more powerful.

740

15...dxc4 16.Qxc4 e5! 17.d5 17.dxe5? g4 18.Ne1 Nxe5µ is bad for White. Also after 17.Nxe5 Nxe5 18.dxe5 Be6 19.Qe2 Qd4µ Black’s bishops reign supreme. 17...g4 18.Ne1 If 18.Nfd2 cxd5 19.exd5 f5 20.Nc3 e4³ the knight will reach e5 soon, after which Black will dominate. 18...cxd5 19.Qxd5 19.exd5?! f5 20.Rd1 Bd6 21.Nc3 Nc5 22.Nb5 b6³ is nice for Black. 19...Bg5 20.Rd1 Black’s only issue is developing the light-squared bishop, so... 20...Nb6!

741

21.Qb3 21.Qxe5†? Bf6 22.Rxd8 Bxe5 23.Rxf8 Bxb2µ sees Black regain the material with a great position. 21...a4 22.Qc2 Qf6 23.Nc3 Be6 Black succeeds in developing his bishop, obtaining an excellent position.

14...Nxe5 15.Nxe5 After 15.dxe5 Bd7 16.Nc3 f5! 17.exf6 Rxf6 Black has fine prospects along the f-file. 15...Bd6 16.Nd2 f6 17.Nd3 17.Ng6?! looks too risky in view of 17...Rf7 18.h3 Bd7 19.f4 Be8 when the knight on g6 looks shaky.

17...f5 742

17...Bd7!?N also deserves attention, after which 18.e4 dxe4 19.Nxe4 Bc7 20.d5 exd5 21.cxd5 Bxg4 leads to sharp play. 18.h3 h5 19.Ne5 After 19.Nf3 fxg4 20.hxg4 hxg4 21.Nfe5 Bd7 22.Qe2 Be8 23.Qxg4 Qf6„ White’s control over the e5-spot does not yield any advantage. We have been following Gelfand – Meier, Jerusalem 2018. A natural improvement for Black is:

19...Bd7N 20.Ndf3 Be8 I think Black’s pieces are better placed to carry out active operations on the kingside. C) 9.a4!?

743

The idea behind this somewhat unusual move is to make it more difficult for Black to develop his lightsquared bishop. 9...a5! I prefer this over 9...b6 10.Rc1! when the following examples show the venom behind pushing the apawn: 10...Bb7 (10...Ba6 11.b3 c5 12.a5 Bb7 13.a6 Bc6 occurred in Voit – Gunina, Moscow 2018, when 14.cxd5!N 14...Nxd5 15.e4 N5f6 16.Nc3ƒ would have offered White promising play in the centre) 11.a5 c5 12.a6 Bc6 13.cxd5 Bxd5 14.Nc3 Bc6 This was Narayanan – Nakar, Sandavagur 2018, and now the natural 15.e4N 15...cxd4 16.Nxd4 Ne5 17.Nxc6 Nxc6 18.Rd1± would have posed Black serious problems.

White has two main ways of meeting our last move: C1) 10.Na3 and C2) 10.Rc1. 744

10.Bf4 b6 is not dangerous, for instance: 11.cxd5 Nxd5 12.Nc3 (After 12.Qxc6N 12...Ba6 13.Qc2 Rc8 14.Nc3 N7f6 15.Rfc1 Bb4 16.Bd2 Bxe2 Black regains the pawn with a good position.) This was Llanes Hurtado – Bauer, Haguenau 2013, when Black should have played:

12...Nxf4N 13.gxf4 Bb7 14.Ne5 Rc8 15.Rfd1 Nf6 16.e3 Nd5 17.Rac1 Nb4 18.Qb3 Bd6 Black has a solid position, with two bishops and control over the b4-outpost. 10.Rd1 can be compared with variation D but it makes little sense, as White has lost flexibility on the queenside. 10...b6 11.b3 Ba6 12.Bc3 (The overly aggressive 12.Bf4 Rc8 13.Ne5?! c5 14.dxc5 took place in Kozul – Saric, Stari Mikanovci 2009, when the natural 14...Bxc5N 15.Nxd7 Qxd7 16.Nc3 Qe7ƒ would have left Black with full control over the dark squares.) This position occurred in Kyhos – Ertl, corr. 2010. My natural new idea is:

745

12...Rc8N 13.Nbd2 c5 14.Rac1 cxd4 15.Nxd4 Ba3 16.Ra1 Qe7 Black is not worse at all. A somewhat dubious pawn sacrifice is: 10.Nc3?! dxc4 11.e4 I also examined: 11.Nb1N 11...b5!? Not the only playable move, but I find it the most attractive. 12.axb5 cxb5 13.Ne5 Ra6 14.Nc6 Rxc6 15.Bxc6 Qb6 16.Bg2 a4 17.Nc3 Bb7 In this unbalanced position I prefer Black; White’s rooks have no access to open files, whereas Black’s queenside majority is rather dangerous. Establishing a strong pawn centre is White’s most natural idea, but his pieces are not developed enough to support such ambitious play.

11...e5! 746

This excellent reply enables Black to activate his pieces and establish an outpost on d3. The same idea was seen in variation C of Chapter 17. 12.dxe5 12.Nxe5?! Nxe5 13.dxe5 Ng4 14.Bf4 g5 15.Rad1 Qb6 16.Bc1 Nxe5µ is even worse for White. 12.d5 Re8 13.Rad1 Bc5³ 12...Ng4 13.Bf4N This move is the most challenging – at least White manages to provoke Black into creating a weakness of some kind. 13.e6 fxe6 14.h3 Nge5 15.Nd4 Nc5 16.Be3 Qb6 was clearly worse for White in Kelires – Kuljasevic, Rio 2018. 13...g5

14.e6! Nc5! 14...fxe6 15.Bd2 Nge5 16.Nxe5 Nxe5 17.Rad1 offers White definite compensation for the pawn, though even then Black should be fine. 15.exf7† Rxf7 16.Bd2 Qd3 17.Qxd3 Nxd3 Black has excellent play. C1) 10.Na3

747

This move looks somewhat artificial, but it enables White to maintain the tension on the queenside while connecting the rooks. The drawback is obvious: it diminishes White’s control of e4, so Black has time for regrouping the pieces. 10...Bd6 11.Ne1 The other natural continuation is 11.Rfd1N 11...Qe7 12.Rac1, but in that case Black can comfortably complete development by means of 12...b6! 13.cxd5 cxd5 14.Nb5 Ba6 15.Qc6 Ne4, reaching an excellent position. The text move has been employed twice by Boris Gelfand, one of the world’s foremost Catalan experts. My new idea for Black is:

748

11...Re8!N Since the queenside fianchetto is hard to carry out, it is crucial for Black to play ...e5 at some point. This preparatory move seems the most precise. Instead, after 11...Qe7 12.Nd3 e5 13.cxd5 Nxd5 14.Bxd5 cxd5 15.Nb5² White managed to liquidate the strong dark-squared bishop in Gelfand – Tomashevsky, Moscow 2017. 12.Nd3 e5 13.cxd5 e4! This intermediate move is vital, since 13...Nxd5? 14.Nc4± is simply bad. 14.Nf4 An equally good alternative is: 14.Nc5 cxd5 15.Nb5 Bb8 16.f3 (16.Bg5 h6 17.Nxd7 hxg5 18.Nxb8 Rxb8=) 16...h6 17.fxe4 Nxc5 18.dxc5 Nxe4„ Black has good counterplay along the e-file. 14...cxd5 15.Nb5 Bb8 In comparison to the Gelfand – Tomashevsky game, the c7-square is covered and the c8-bishop is well protected. 16.Rac1 h6 A tough strategic battle lies ahead, where Black’s chances are by no means worse. For instance:

17.Nc7 Bxc7 18.Qxc7 Qxc7 19.Rxc7 g5! 20.Nh3 Nb8³ Despite having the bishop pair, White is under some positional pressure as the g2-bishop is locked out of the game. C2) 10.Rc1 749

This move also discourages the queenside fianchetto, but keeps the b1-knight more flexible. 10...Ne4 The knight occupies a strong central outpost in order to interrupt the coordination between White’s pieces. Moreover, Black might consider developing some play on the kingside with ...f5 and possibly ...g5. The venom of White’s set-up was illustrated in the following game: 10...b6 11.cxd5 cxd5 12.Nc3 Ba6 13.Nb5 Rc8 14.Qb3 Black was under permanent pressure and eventually lost in Wang Yue – Kravtsiv, Riyadh 2017. 11.Be1 This modest-looking move was successfully employed by World Champion Magnus Carlsen. I checked three other ideas: 11.Bf4N makes little sense, as 11...g5 12.Be3 f5 offers Black rich play on the kingside. 11.Be3 This occurred in Aronian – Wei Yi, Wijk aan Zee 2017. I suggest the following natural plan:

750

11...b6!?N This works better here than on the previous move. 12.cxd5 12.Nc3 Nxc3 13.bxc3 Ba6 14.cxd5 cxd5= gives us no problems. 12.Ne5 Nxe5 13.dxe5 Nc5 is also comfortable for Black, who has neutralized the pressure along the c-file while gaining stability for the knight. 12...cxd5 13.Na3 13.Ne5 is hardly more promising in view of: 13...Nxe5 14.Bxe4 Ng4! 15.Bxh7† Kh8 White is a pawn up but his lack of harmony starts to tell. For instance, 16.Bd2 (16.Bf4? is nicely refuted by 16...e5! 17.dxe5 Bc5 18.Rf1 g5 19.Bc1 f5! 20.exf6 Qxf6–+ with decisive threats) 16...Bf6 17.h3 Nxf2 18.Kxf2 g6 and Black regains the material in a favourable situation.

751

13...Bxa3! In such a closed position with a fixed pawn structure, Black will not miss the bishop too much. 14.Rxa3 Ba6 15.Qd1 Rc8 16.Raa1 Qe7 Black has no problems. 11.Ne1 Finally, against this move I suggest another novelty: 11...Nxd2N Liquidating the bishop secures Black full control over b4. Instead, after 11...Nd6 12.Na3 b6 13.cxd5 cxd5 14.Qb3 Ba6 15.Nb5 Rc8 16.Bf4 Bxb5 17.axb5² Black was suffering from the weak c6-square in Aronian – Anand, Stavanger 2017. 12.Nxd2 Bf6 13.Nef3 In the event of 13.e3 e5 14.dxe5 Nxe5 15.cxd5 cxd5 16.Ndf3 Nc6, Black’s bishop pair, especially the powerful dark-squared unit, fully compensates for the isolated pawn.

13...Nb8! I like this manoeuvre, which highlights the positional drawback of the a2-a4 plan. The position is closed, so Black can afford to leave some of his pieces undeveloped for a while. My other idea was 13...b6 14.e4 Bb7 15.e5 Be7 16.cxd5 exd5 with a sharp, double-edged game. This seems playable, but I prefer the text move. 14.e4 14.Ne5 Na6 15.Ndf3 Nb4 16.Qc3 b6 is comfortable for Black. 14...Na6 15.e5 Be7 16.c5 If White maintains the tension with 16.Qc3, Black can play 16...dxc4! 17.Nxc4 Nb4 18.Nfd2 b6 19.Ne4 Ra7 20.Ned6 Ba6 when the d4-pawn is a target. The text move is more ambitious; White’s space advantage is undeniable. 16...b6 Of course, Black hurries to attack the opponent’s pawn chain. 752

17.cxb6 Also after 17.Nb3 bxc5 18.dxc5 Qc7 19.Qc3 Nb4 20.Re1 Ba6 21.Nfd4 Bc4 Black’s chances are by no means worse; White’s queenside pawns are weak. 17...Nb4 18.Qd1 Qxb6 19.Nb3 Ba6 20.Rc3 Rab8 Black has sufficient counterplay along the b-file.

11...f5 12.Nbd2 The more aggressive 12.Nc3 Bd6 13.Nd2, as in Deac – David, Calimanesti Caciulata 2017, leaves the c4-pawn unprotected, so it makes sense to play: 13...Nxd2N 14.Bxd2 dxc4

15.e3 (15.e4?! seems strange when White’s pieces are concentrated on the queenside, as Black is better prepared for a central confrontation: 15...e5! 16.Ne2 fxe4 17.Qxc4† Kh8 18.Bxe4 exd4 19.Nxd4 Ne5³) 15...e5 16.Ne2 Kh8 17.Qxc4 Nb6 18.Qb3 e4 Black has excellent prospects in this reversed French 753

structure with a passive bishop on g2. 12...Bd6 13.e3 White’s unassuming set-up is mainly designed to inhibit the development of the c8-bishop. Beyond that, it is not clear what his plan should be. Therefore, I see nothing wrong with expanding on the kingside with:

13...g5!?N In the following top-level game Black was determined to carry out the fianchetto, but failed to solve all his problems: 13...Ra7 14.Qd1 b6 15.cxd5 cxd5 16.Nb1 Ba6 (better was 16...Rc7!N 17.Na3 Rxc1 18.Qxc1 Ba6 19.Nb5 Bxb5 20.axb5 Qc8 when Black gets close to full equality) 17.Na3 Qa8 18.Nb5 Bxb5 19.axb5 White had a small but long-lasting advantage in Carlsen – Eljanov, Stavanger 2016. 14.Qd1 Qf6 14...g4!? also looks interesting, for instance: 15.Nxe4 dxe4 16.Nd2 h5„ 15.Rc2 Rf7 16.Rac1 Qg7 17.Nb3 Bc7 White’s position looks somewhat passive, and he cannot claim any advantage. D) 9.Rd1

754

This is a popular and thematic option. For the moment the d-file is closed but, considering that the ...c5 break is a big part of Black’s plans, the logic behind putting the rook here is obvious. 9...b6 Black proceeds with his usual plan of development. 10.b3 10.Bf4 is the most popular continuation; we will cover this position under the 9.Bf4 b6 10.Rd1 move order; see variation D of the next chapter on page 391. 10.a4 has scored quite well but 10...a5= seems harmless for Black. Unlike the analogous position which could result from variation C2, which we avoided in view of a quick cxd5, here the white rook stands on d1 instead of c1, so Black has nothing to fear from the opening of the c-file. Another unimpressive idea is: 10.Ne5 Compared to variation D of the next chapter, where the knight moves to e5 after the bishop has gone to f4, here the d-file is blocked and White’s bishop is less active. All this means that Black can confidently trade off the active knight with: 10...Nxe5 11.dxe5

755

11...Ng4! 11...Nd7 is reasonable but Black can play more ambitiously. 12.cxd5 cxd5 13.e4?! This is the move White wants to play, but it does not work. 13.Bc3 is better, but after 13...Bc5 14.e3 Qc7, as in Windvogel – Ivana Maria, Gauteng 2011, White still faces the prospect of losing his e5-pawn for uncertain compensation. 13...d4 14.Bf4

14...g5! 15.Qe2 Nxf2 16.Qxf2 gxf4 17.gxf4 In Lenic – Mchedlishvili, Germany 2013, Black’s most accurate continuation would have been: 17...Ba6Nµ White has to contend with numerous unpleasant ideas including ...d3, threatening ...Bc5 and ...Rc8-c2, 756

and ...Kg8 followed by ...Rg8.

10...Ba6 10...Bb7 is a popular alternative but I prefer the text move. It also makes a convenient choice for our repertoire, as there are certain move orders where White starts with some other move and then meets ...Ba6 with b2-b3. 11.a4 This is an important option, which received a lot of attention after being recommended by Boris Avrukh in the first ever Grandmaster Repertoire book. Avrukh recommended a different plan in the updated Volume 1A in 2015 (see the next chapter for more about this), but the text move is still a reasonable choice which continues to be used by strong players to this day. White’s other main option is 11.Bf4, which converts to variation D3 of the next chapter on page 400. 11...c5 Several other moves have been tested at high levels. However, Black’s set-up is already rather harmonious, so I see no reason to delay this key pawn break. White may respond with D1) 12.cxd5, D2) 12.a5 or D3) 12.Na3. 12.Nc3?! is a dubious pawn sac which Black should have no hesitation in accepting: 12...dxc4 13.bxc4 Bxc4 14.Ne5 Nxe5 15.dxe5 Nd5 16.Be1 Qc7

757

17.Ne4 (after 17.Nxd5?! Bxd5 18.Bxd5 exd5 19.Rxd5 a6µ Black’s queenside passers will be too powerful) 17...Ba6 18.Nd6 Rad8 White did not have enough compensation in Gelfand – Lautier, Biel 2001. D1) 12.cxd5 Nxd5

With this exchange, White is trying to exert pressure along the h1-a8 diagonal. On the other hand, Black’s light-squared bishop gets more freedom and his rook will be active on the c-file. 13.Nc3 The ambitious 13.e4N runs into 13...Nb4! 14.Bxb4 cxb4 15.Nbd2 Rc8 16.Qb2 Nf6 17.Ne5 Bb7, when 758

the vulnerability of the c3-square causes White definite problems. 13...Rc8 14.Qb2 After 14.Nxd5N 14...exd5 15.Bf4 cxd4 16.Qa2 d3 17.exd3 Nc5 18.Ne5 Ne6 Black has nothing to worry about. 14...Bb7

15.Nxd5 I also considered: 15.dxc5 This occurred in Svidler – Karjakin, Astana (rapid) 2012. In the game Black recaptured on c5 with a decent position, but he could have caused more problems with: 15...Bf6!N Attempting to exploit the poor coordination among White’s queenside pieces. My analysis continues: 16.e4!? 16.b4 bxc5 17.b5 blocks the c-file but 17...c4 18.e4 Nxc3 19.Bxc3 Qc7 20.Ne1 Rfd8„ gives Black plenty of activity. The text move invites interesting complications after:

759

16...Nb4! 17.Bg5! This is the only way for White to avoid material loses. 17...Bxg5 18.Nxg5 Qe7 19.cxb6 Ne5 20.Nh3 axb6 The activity of Black’s pieces offers him full compensation for the pawn. 15...Bxd5

16.Bc3 Bf6 17.Ne5 I also examined 17.Rac1, as played in Armani – Petruzzelli, corr. 2016, when 17...h6!N is a useful waiting move. (The game continued with the natural-looking 17...Qe7, when 18.b4! enabled White to generate some pressure.) Play might continue:

760

18.Qb1 White insists on the e2-e4 advance, but invites further simplifications: 18...Bxf3 19.Bxf3 cxd4 20.Bxd4 Bxd4 21.Rxd4 Qe7 Black’s knight is not really worse than the bishop, so the position is equal. 17...Bxg2 18.Kxg2 Qe7 Black moves his queen to safety and prepares to bring his last piece into play.

19.Nxd7 White also achieves nothing after: 19.Ng4 Bxd4 20.Bxd4 cxd4 21.Rxd4 h5 22.Rad1 Rfd8 23.Ne3 Nc5= 19...Qxd7 20.dxc5 Qb7† 21.f3 Bxc3 22.Qxc3 Rxc5 23.Qe3 h6 761

Black fully equalized in Rattinger – Lafarga Santorroman, corr. 2010. D2) 12.a5

White is trying to exert some pressure along the a-file, but delaying the development of the b1-knight diminishes his control over the centre. 12...dxc4 13.bxc4 Qc8! This multi-purpose move avoids the dangerous opposition of queen and rook along the d-file, while securing the b7-square for the bishop’s retreat. 14.Nc3!? This move was introduced in 2015 and has not yet been repeated, but I consider it the most ambitious continuation. White tries to put pressure on Black with the help of a temporary pawn sacrifice. 14.Ne5 only invites simplifications: 14...Bb7 15.Nxd7 Qxd7 16.Bxb7 Qxb7 Black was fine in Pham – Nguyen, Da Lat 2014. The most popular continuation has been: 14.Bf4 cxd4 15.Nxd4 Bb7 Black neutralizes the pressure along the h1-a8 diagonal while maintaining a healthy pawn structure. 16.Bxb7 After 16.e4 bxa5 17.Rxa5 a6 18.Ra2 Bb4 19.Rb2 a5 Black had comfortable play thanks to the b4outpost and passed a-pawn in Pfiffner – Wegelin, corr. 2010. 16...Qxb7 17.Nc3 This position occurred in Tomashevsky – Karjakin, Baku 2014. In this and a few other games Black opted to exchange on a5, but I prefer: 762

17...Nc5N 18.Rab1 18.e4?! loosens White’s position, making 18...bxa5! 19.Rxa5 a6³ more appealing. 18.Nb3 Qc6 19.axb6 axb6= is fine for Black. 18...Rfd8 19.e4 h6= Black has comfortable play.

14...cxd4!N This is the right moment to release the central tension. White’s idea was justified in the following example: 14...Bxc4 15.Bg5! b5 (15...h6 16.Bxf6 Nxf6 17.Ne5ƒ) 16.Ne5 cxd4 17.Nc6 Bc5 18.Bxf6 Nxf6 19.Nxb5 Bxb5 20.Qxc5 Bxc6 21.Bxc6 White regained the pawn and kept some positional pressure in Swiercz – Karjakin, Doha 2015.

763

15.Nxd4 Bb7 16.Bxb7 Qxb7 17.a6!? This is the most ambitious try, fixing the a7-pawn as a long-term target. If it ever falls, the a6-pawn will be close to promoting. 17.Bf4 transposes to the 14.Bf4 line noted above, where it was shown that 17...Nc5N offers Black a good game.

17...Qc7 18.Ncb5 Securing the b5-square for the knight is another benefit of the a5-a6 move. 18...Qc5 19.Bc1 The bishop is heading for a3. A more solid continuation is 19.Qb3 Qh5 20.Qf3 (or 20.Kg2 Rfc8 21.h3 Nc5=), but after 20...Qxf3 21.Nxf3 Rfc8 22.Rac1 Nb8= Black has nothing to worry about. 19...Qh5 20.Nc6 20.Ba3 can be met by 20...Bxa3 21.Rxa3 Qc5= when Black has enough pressure along the c-file to stop White from making progress. After the text move it looks as though White has achieved his goal of breaking through to a7. However, the desertion of the knights and other pieces from the kingside makes it possible for Black to develop significant counterplay there. 20...Bc5! 21.Kg2! A critical position has been reached.

764

21...Ng4! Black must bring more pieces into the attack. The tempting 21...Bxf2? is strongly met by 22.Qd3! Bc5 23.Qf3! when White’s pieces dominate over the board. 22.h4 This is the only way to parry the threats. 22...Nde5 23.Nxe5 Qxe5 24.Bb2

24...Qe3! 765

This powerful tactical resource enables Black to exploit White’s shaky kingside to win a pawn. 25.fxe3 Nxe3† 26.Kf3 Nxc2 27.Ra4 Nb4 28.Rd7 Nc6 We have reached an unbalanced endgame where White clearly has compensation for the pawn, but Black is doing fine. D3) 12.Na3

This was Avrukh’s recommendation and has been the most popular choice by far. White develops the last of his minor pieces and preserves the central tension. 12...Bb7 This bishop has nothing left to do on a6, so it returns towards the centre. 13.Qb2 This is the usual move. The queen leaves the c-file in anticipation of a rook arriving on c8. It also defends the knight on a3, thus enabling the a1-rook to move. Finally, it supports a possible b3-b4 advance. Releasing the tension simplifies Black’s task: 13.cxd5 Bxd5 14.Nc4 Rc8 15.Qb2 cxd4 16.Nxd4 Bxg2 17.Kxg2 Nc5 Black was easily equal in Ionov – Kunte, Hyderabad 2000. It is also worth mentioning: 13.Bc3 This has been tried in a few games. My new idea is:

766

13...Qc8!?N Avoiding the unpleasant opposition along the d-file. Black was successful with 13...Rc8 in Buhmann – Grischuk, Mainz (rapid) 2010, but I find this move less accurate in view of 14.Ne5!N 14...cxd4 15.Bxd4 Nxe5 16.Bxe5, when the d5-pawn comes under strong pressure. 14.dxc5 14.Bb2 is hardly an improvement: 14...dxc4 15.Nxc4 Be4 16.Qc3 Rd8= 14.Ne5 Rd8 15.Bb2 cxd4 16.Bxd4 Nxe5 17.Bxe5 h6= is also harmless. 14...bxc5 In this specific situation, when White’s queenside structure has been weakened by a2-a4, I prefer this recapture. 14...Nxc5 15.Ng5 g6 also looks acceptable, but is less dynamic. 15.e3 Rb8 16.cxd5 exd5 The pressure on the b3-pawn fully compensates for any slight vulnerability of the hanging pawns. For instance: 17.Bb2 Re8 18.Rac1 Bf8 19.Ne1 Qc6 20.Nd3 Qb6„ Black is not worse at all.

767

13...Ne4 The knight occupies a strong outpost and vacates the f6-square for the bishop, so White’s queen begins to feel less comfortable than before. 14.Be1 14.Bf4?! occurred in Novikov – Jakovenko, Moscow 2007, when Black missed a good opportunity to seize the initiative:

14...g5!N 15.Bc1 (15.Be3?! f5 looks even worse for White, with ...f4 coming next) 15...f5 16.cxd5 Bxd5³ Black is better, due to his full control over the central light squares.

768

14...Bf6 15.e3 Supporting the centre in this way has been the usual continuation. A natural alternative is: 15.Rac1 Here I found a new way of handling the position:

15...Nd6!?N Black intends to exchange the enemy knight when it arrives on b5 or c4, which should make his setup much more comfortable. 16.e3 16.Nb5 Nxb5 17.axb5 a6 gives Black easy equality, for instance: 18.bxa6 Rxa6 19.Bc3 Qa8 20.Ra1 Rxa1 21.Rxa1 Qb8= In the event of 16.cxd5 Bxd5 17.Bc3 Ne4 White has nothing better than retreating the bishop: 18.Be1 Nd6= 16...dxc4 17.Nxc4 17.bxc4 Qc7 18.Nb5 Nxb5 19.axb5 a6 20.Bc3 axb5 21.cxb5 Ra4 also offers Black equal play.

769

17...Nxc4 18.bxc4 18.Rxc4?! cxd4 forces White to settle for an unfavourable pawn structure, since 19.Nxd4? Bxg2 20.Kxg2 Ne5 wins material for Black, on account of the dual threats of ...Nxc4 and ...Qd5†. 18...Be4 19.Qe2 19.a5 Qc7 20.axb6 axb6= is fine for Black.

19...Re8 20.d5 This is the only way to fight for the advantage. However, White’s strategy carries certain risks. 20...exd5 21.cxd5 Qe7 White’s central pawns are not currently mobile, so he has to exchange the light-squared bishops if he wishes to get anywhere.

770

22.Nd2 Bxg2 23.Kxg2 g6 24.Nc4 a6 25.Rb1 Rab8 Black has sufficient counterplay on the queenside. Let’s return to the main line, which was recommended by Avrukh in his famous Grandmaster Repertoire 1 – 1.d4 Volume One.

15...Qb8! Avrukh did not consider this move in his book, but subsequent games have demonstrated that it is Black’s most reliable continuation. The queen moves away from the glare of the enemy rook, while making room for the f8-rook to take up a better spot. As we will see, the blocking of the rook on a8 is only temporary.

771

15...cxd4 is premature, and 16.Nxd4 Qb8 17.Rac1 gave White the better chances in Piket – Kramnik, Monaco (rapid) 1997, as Avrukh noted in the aforementioned book. 15...Qe7 has similar ideas to the text move, and has been tried in one game. Interestingly, this move gave Avrukh the chance to demonstrate his understanding of these positions with 16.Ne5!, after which White seized the initiative in Avrukh – Lerner, Israel 2009. The obvious advantage of our main line compared to this is that Black takes control of the e5-square and prevents the knight incursion. 16.Rac1 The is the most natural and popular move. 16.cxd5 Bxd5 17.Nc4 was tried in Horak – Keskowski, email 2003, and here I recommend the following improvement:

17...Qb7!N 18.Nfe5 cxd4 19.exd4 Nxe5 20.dxe5 Be7 Black has excellent prospects due to his control over the light squares. White also fails to achieve much after: 16.b4N 16...cxb4 17.Bxb4 (17.Qxb4 Rc8 18.Rac1 a5 19.Qb2 Be7 leads to the same type of position) 17...Rc8 18.Rac1

772

18...a5! Chasing the bishop away while fixing the a4-pawn as a potential weakness. 19.Be1 Be7 20.cxd5 Bxd5 21.Nd2 Qb7 Black has good play on the light squares, and the a4-pawn might become weak in the long run. True, the b6-pawn is also theoretically weak, but I don’t see how White can exert meaningful pressure against it.

16...Rc8 17.Qb1 This seems most consistent; White unpins the d4-pawn and keeps an eye on the e4-knight. I also considered: 17.Qe2 h6 18.Nd2 White hardly has anything better than exchanging the powerful knight on e4. (18.Nb5 cxd4 19.Nfxd4 dxc4 20.Rxc4 Rxc4 21.Qxc4 Nec5=) 18...Nxd2 19.Bxd2 dxc4 20.Bxb7 Qxb7 21.Nxc4 Rd8 Black has nothing to worry about. 773

17...Nd6! 17...cxd4?! would be premature in view of 18.Nxd4 Ndc5 19.cxd5 exd5 20.Bh3 when Black was doomed to a passive defence with an isolated pawn in Mikhalevski – Huzman, Beersheba 2014. The text move is better: Black clears the long diagonal for the b7-bishop and puts more pressure on the c4-pawn.

18.Nb5 Nxb5 19.axb5 a6! The b5-pawn was cramping Black’s queenside, so it makes sense to get rid of it. 20.cxd5 Bxd5 21.e4 Bb7

774

22.e5 Be7 23.bxa6 Rxa6 24.dxc5 Nxc5= We have been following Muck – Dudyev, corr. 2011. Black has solved his development problems and is not worse at all: White has a slight space advantage but Black is not too cramped, and the e5-pawn may become weak if White is not careful. Conclusion 8.Qc2 is a logical choice to guard the c4-pawn while vacating d1 for the rook, keeping White’s position flexible and harmonious; no wonder it is the main line. After the normal developing move 8...Nbd7 we considered a few sidelines, of which 9.Rc1 and 9.a4!? are both rather interesting. In both cases, Black should refrain from an immediate queenside fianchetto and instead play 9...a5! when his chances are fine, as long as he keeps a flexible attitude and is ready to modify his plans according to how White positions his pieces. The main line of the chapter was 9.Rd1 b6 10.b3 Ba6, after which the original Avrukh recommendation of 11.a4 deserves close attention. The logical 11...c5 is our reply, when 12.cxd5, 12.a5 and 12.Na3 all lead to interesting play, but my analysis shows that Black has a full share of the chances against each of them.

775

A) 10.Nbd2 382 B) 10.cxd5 384 C) 10.Nc3 385 D) 10.Rd1 Ba6 391 D1) 11.Nbd2 391 D2) 11.cxd5 cxd5 12.Ne5 Nxe5! 13.dxe5 Ng4! 392 D21) 14.Bf3 393 D22) 14.h3 394 D3) 11.b3 Rc8 12.Nc3 dxc4 400 D31) 13.Nd2 401 D32) 13.e4 403 D33) 13.bxc4!? 405 D4) 11.Ne5 Rc8! 407 D41) 12.cxd5 408 D42) 12.Nc3 Qe8!? 411 D421) 13.e4 412 776

D422) 13.Rac1N 413 1.d4 Nf6 2.c4 e6 3.g3 Bb4† 4.Bd2 Be7 5.Bg2 d5 6.Nf3 0-0 7.0-0 c6 8.Qc2 Nbd7 9.Bf4 This is White’s most popular move, and the recommendation in Avrukh’s updated Grandmaster Repertoire 1A – The Catalan. 9...b6

After this typical move, White has several options. We will analyse A) 10.Nbd2, B) 10.cxd5, C) 10.Nc3 and finally the most popular D) 10.Rd1, after first considering a few minor alternatives: 10.Rc1 does not have much independent value, since after 10...Bb7 11.Nc3 (or 11.cxd5 cxd5, transposing immediately to variation B) 11...Rc8 White has nothing better than 12.cxd5 cxd5, transposing to a position covered on page 384 – see 12.Nc3 Rc8 in the notes to variation B. It makes no real sense to invite simplifications with: 10.Ne5?! Nxe5 11.Bxe5 (11.dxe5 Nd7 12.h4 Ba6 13.Nd2 Rc8 14.Rad1 b5 15.cxb5 cxb5 16.Qb1 b4 left Black with the better pawn structure in I. Vovk – Lugovoi, Tallinn 2011) 11...Ba6 12.Nd2 Ng4 13.Bf4 In Viela – Silva, Sao Joao da Madeira 2013, Black could have seized the initiative with:

777

13...g5!N 14.Be3 f5 15.Qa4 Bb7³ The lack of harmony in White’s camp is obvious. The following option usually just transposes to one of our main lines: 10.b3 Ba6 11.Nc3 The most common 11.Rd1 leads straight to variation D3 on page 400. 11.Nbd2?! cannot be recommended for White. 11...Nh5 12.Be3 A sad necessity. 12...Rc8 13.Rfd1 f5³ White’s set-up lacks harmony.

11...Rc8 12.e4 12.Rfd1 is variation D3. 12...dxc4 White has nothing better than: 778

13.Rfd1 Reaching variation D32 on page 403. Finally, the thematic plan of advancing the a-pawn seems ineffective here: 10.a4 Ba6 11.b3 11.Nbd2?! Rc8 12.a5 bxa5 13.c5?! Bxe2 14.Rfe1 Bb5 left White without adequate compensation for the two pawns in Hodgson – Luther, Hastings 1996. 11...c5 Since Black’s pieces are better developed, I see no reason to delay this important break. Another interesting possibility is 11...Rc8!? 12.a5 b5 13.c5 b4, after which 14.Ra4 Ne4! 15.Rxb4 g5 16.Bc1 f5 offers Black rich play for the pawn. 12.Na3 This position occurred in Latorre – Butti, Asuncion 2013. I suggest:

12...Bb7N Taking control over the central squares e4 and d5. 13.dxc5 bxc5!? 13...Nxc5 also looks decent. 14.cxd5 Nxd5 15.Bd2 Nb4 The pawns on b3 and c5 are equally weak, so the position is balanced. A) 10.Nbd2 This knight development often makes the bishop less comfortable on f4, and the present case is no exception. 10...Nh5 779

10...Bb7 is another common response, but why should we allow e2-e4? 11.cxd5!? This is the most interesting continuation, inviting complications. Quieter play yields White nothing: 11.Be3 Bb7 12.Rfd1 12.Rad1 leaves the other rook out of play, so Black gets a fine game with 12...Rc8 13.Ne5 Nhf6 followed by queenside action. For instance, 14.Ndf3 c5 15.dxc5 Nxc5 16.Qb1 Qc7 17.cxd5 Bxd5 and Black had the more harmonious position in Oliva Castaneda – Merario Alarcon, Linares 2017.

12...f5! The Stonewall treatment works especially well here, as White does not have great control over e5 and the e3-bishop is a target for ...f4. 13.Nb3 13.Ne5 f4 14.gxf4 Nxf4 15.Bxf4 Rxf4 16.e3 Rf8 17.Ndf3 Nxe5 18.Nxe5 Bd6 is also more than okay for Black. 13...Rc8 14.c5 We have been following Moorthy – Kaufman, Bethesda 2016. This would have been a good moment for Black to activate his light-squared bishop by means of:

780

14...Ba6!N 15.a4 Bc4 With excellent prospects.

11...Nxf4 The inferior 11...cxd5?! allows 12.Bc7!N 12...Qe8 13.Rfc1² when White has an easy edge. 12.Qxc6 12.dxc6 occurred in Postny – Boruchovsky, Gjakova 2016. In the game Black took on e2, which is not a bad move, but I prefer 12...Nxg2!?N 13.cxd7 Qxd7 14.Kxg2 Ba6 15.Ne4 Bb7, when Black’s strong bishops and active prospects for the rest of his pieces offer him at least enough compensation for the pawn.

781

12...Nxe2† 13.Kh1 Rb8 14.dxe6

14...Nxd4! This nice tactical resource enables Black to secure the c5-square for his knight and activate all his pieces. 15.Nxd4 Nc5 16.exf7† Rxf7 17.Qd5 White has no choice, as 17.N2b3? Bb7 18.Qb5 Ba6–+ sees Black win the exchange. 17...Bb7

18.Qxd8† Rxd8 19.Bxb7 Rxd4 782

Black’s piece activity provided full compensation for the pawn in Sethuraman – Meier, Tromso (ol) 2014. B) 10.cxd5 cxd5

White hopes to utilize the open c-file, combined with his slight edge in piece mobility, to create concrete problems for Black. 11.Rc1 11.Nc3 Bb7 12.Rac1 (12.Rfc1 transposes to 12.Nc3 in the notes to the main line below) 12...Rc8 13.Qb3 a6 14.a4 Rc4 15.Na2 Rxc1 16.Rxc1 Qa8 17.Ne5 Rc8 gave Black no problems in Ortega Ruiz – Sorokin, Linares 2000. White also achieves nothing with: 11.Bc7 Qe8 12.Nc3 Bb7 13.Nb5

783

13...Ne4! Black has already equalized, and the careless 14.Qb3? runs into 14...Ndc5! 15.dxc5 Nxc5 16.Qb4 a5 and Black won a pawn in Bauer – Ivanchuk, Cap d’Agde 2012. 11...Bb7

12.Qb3 Another way of handling the position is: 12.Nc3 Rc8 13.Qd1 (or 13.Ne5?! Nxe5 14.Bxe5 Ne4! and White was already under pressure in Duda – Ding Liren, Liaocheng 2018) This position was reached in Andersson – Bareev, Ubeda 1997, and a few other games. I recommend the following new way of handling the position:

784

13...Nb8!?N Transferring the knight to c6 makes Black’s life easier. 14.Ne5 (After 14.Rc2 Nc6 15.a3 Na5 the b3-square is available for the knight, so White fails to double his rooks.) 14...Nc6 15.Nxc6 Rxc6 16.e3 Qd7= Black has nothing to worry about. 12...Rc8 13.Rxc8 Qxc8 14.Nc3

14...Bc6! This precise move makes it possible for Black to regroup his pieces and neutralize any pressure along the c-file. 14...Nh5 invites White to exchange the dark-squared bishops: 15.Bg5 Bxg5 16.Nxg5 Qb8 17.e3 Nhf6 18.Bf1 In this position with a symmetrical pawn structure, White’s pieces are somewhat more active. For 785

instance, 18...h6 19.Nf3 Ne4 20.a4 Rc8 21.a5 and Black was under slight but long-lasting pressure in Karjakin – Wei Yi, Wijk aan Zee 2018. 15.Rc1 Qb7 16.Nb5 Bxb5 17.Qxb5 Rc8 18.Rxc8† Qxc8 19.Ne1 Ne4 20.f3 Nd6 21.Qd3 Qc6= In the arising closed position, White’s bishop pair had no more than symbolic value in Muck – Mooij, corr. 2010. C) 10.Nc3

This natural developing move has the usual drawback: the c4-pawn is left unprotected. 10...dxc4 This reaction seems most challenging. White will be able to regain the pawn, but it will involve a positional concession. 11.Nd2 White should not delay regaining the pawn. For instance: 11.Rfd1? Nd5 12.Nd2 Ba6µ gave White no compensation in Jaracz – Miton, Warsaw 2006. 11.e4?! Bb7 12.Rad1 b5³ also favoured Black in Bogosavljevic – Miton, Vrsac 2008. 11.Ne5?!N 11...Nxe5 12.dxe5 Nd5 13.Ne4 Ba6³ also gives White insufficient play for the pawn. 11...Nd5 12.Nxc4 12.Nxd5?! cxd5 13.Nxc4 Ba6³ gave Black the more pleasant game in Aquino – Streit, corr. 1987. 786

12...Nxf4 13.gxf4 Bb7 We have reached a thematic Catalan structure, which also occurred in variation D of Chapter 17. Both sides have their issues. White’s important dark-squared bishop is gone, and his kingside structure has been compromised. On the flip side, White has better control over the e5-square and Black still suffers from the characteristic Catalan pressure along the h1-a8 diagonal.

14.Rfd1 This seems most natural. The opposition of the rook and the enemy queen always favours White, and the c1-square is left vacant for the other rook. Here are some examples involving slightly different setups: 14.Rad1 seems like an odd choice of rook. 14...Qc7 15.e3 Rad8 16.a3 (the careless 16.Qe4? Ba6 17.b3 b5 18.Ne5 Nxe5 19.fxe5 b4µ led to material losses for White in Grigoryan – Bluebaum, KhantyMansiysk 2015) 16...c5

787

17.Bxb7 (17.d5N meets with the typical response of 17...exd5 18.Nxd5 Bxd5 19.Rxd5 b5=; we will see more of this idea in the main line and accompanying notes below) 17...Qxb7 18.dxc5 Nxc5 19.b4 Nd7 Black was comfortably equal in Ghaem Maghami – Chandra, Philadelphia 2015. The aggressive 14.f5 exf5 15.Qxf5 is rather risky, since it helps Black to develop counterplay on the kingside. For instance:

15...g6 16.Qc2 (16.Qh3 Nf6 17.Rad1 Qc7 18.f4 Rad8 19.Ne5 Bd6„ also leads to a tense game) 16...Nf6 17.Rad1 Both sides had played quite naturally in Mann – Nagel, Bad Wiessee 2013. My new idea is:

788

17...b5!?N 18.Ne5 Rc8 19.a3 Bd6 20.f4 Qb6 It is hard to stop the ...c5 break, which will put White’s king in danger. 14.e3 Qc7 15.a4 15.Rfd1 transposes to the main line below. 15...a6! It’s important to cover the b5-square, so that the ...c5 break will be effective. 15...c5?! would be slightly premature in view of 16.d5 exd5 17.Nxd5 Bxd5 18.Bxd5 Rad8 19.Rfd1² when White takes full control over the light squares.

16.Rfd1N 16.a5 b5 17.Ne5, as in Saduakassova – Guo, Ashkhabad 2017, can be met by 17...Nxe5N 18.fxe5 789

c5 19.Bxb7 Qxb7 20.dxc5 Bxc5 when White starts to experience problems with his weak pawns. 16...Rfd8 17.Rac1 b5! The time has come to activate the queenside pieces.

18.Nd2 After 18.Ne5 Nxe5 19.fxe5 c5 20.axb5 Bxg2 21.Kxg2 cxd4 22.exd4 axb5„ the exchange of lightsquared bishops has left White’s king rather exposed. 18...c5 19.Nce4 Rac8 20.axb5 axb5 21.dxc5 Nxc5 22.Nxc5 Bxg2 23.Kxg2 Qxc5= The resulting endgame is level.

14...Qc7 15.e3 15.f5 was played in Gretarsson – Klein, Wijk aan Zee 2013, but I find it a strange choice after moving the f1-rook away from the kingside. I think Black should ignore it with 15...Rad8!?N 16.fxe6 fxe6, 790

obtaining excellent play along the f-file. Another interesting line is: 15.Ne5 Nxe5 16.fxe5 Rad8 17.f4?! An instructive mistake. 17.Rac1 occurred in Matzig – Hinz, email 2013, when I prefer: 17...Qb8!?N (The game saw 17...Qc8, which feels rather artificial to me.) 18.e3 c5 19.Bxb7 Qxb7 20.Qe4 Qxe4 21.Nxe4 cxd4 22.Rxd4 Rxd4 23.exd4 Rd8 24.Rc7 Kf8= Black has enough play along the d-file. The text move appears tempting, and was tested in Rodhstein – Khismatullin, Moscow 2011, but I consider it too risky. A strong reply is:

17...Kh8!N White is ready to meet 17...c5 with 18.d5, but this is not Black’s only idea! The text move prepares to open the g-file and there is not much that White can do to stop it. For instance: 18.Rac1 18.Ne4 f5 19.exf6 gxf6ƒ gives Black good prospects along the g-file. 18...f5! Fixing the f4-pawn so that when ...g5 is played, White will not be able to meet it with f4-f5. 19.e3 Rg8ƒ White faces a difficult future on the kingside.

791

15...Rad8 Hilton & Ippolito reach this position via an earlier transposition in Wojo’s Weapons, but do not consider the text move. The rook leaves the long diagonal in advance, so everything is ready for the thematic ...c5 break. 15...Rfd8 is also playable, when 16.Rac1 Rac8 transposes to variation D of Chapter 17. However, considering that ...c5 will usually be met by d4-d5, Black does not really need a rook on the c-file. 16.Rac1 Several other moves have been tried by strong players. The general rule is that Black will either play ...c5 on the next turn, or make some other move to exploit a particular drawback of whichever move White may choose. For instance: 16.a3 c5 17.d5 exd5 18.Nxd5 Bxd5 19.Rxd5 b5 20.Ne5 Nxe5 21.Rxe5 Rd7= did not make life difficult for Black in Kourousis – Rogozenko, Czech Republic 2012. 16.a4 Bb4 17.Na2 (17.Rac1 Bxc3 18.Qxc3 c5 was pleasant for Black in Kasimdzhanov – Matlakov, Dubai 2014; Black was also fine after 17.Ne4 c5 18.Ne5 Nxe5 19.dxe5 Bxe4 20.Bxe4 g6= in Banusz – Erdos, Sarajevo 2013) 17...Be7 Now the ...c5 advance is back on the menu. The only real attempt to fight for the advantage is 18.b4, as was played in Vakhidov – Grachev, London 2013. Here I like the following new idea:

792

18...g5!?N 19.fxg5 (19.f5 c5! 20.Bxb7 Qxb7 21.dxc5 bxc5 looks shaky for White, whose king is rather exposed) 19...Bxg5 20.Nc3 Kh8 21.Ne4 Be7 Black has excellent prospects on the kingside, as most of White’s pieces are located far away on the opposite flank. The drawback of 16.Qe2 was well illustrated in the following encounter: 16...Ba6 (Another interesting idea is: 16...b5!?N 17.Ne5 Nxe5 18.fxe5 a6 19.a4 This seems the only effective way to discourage ...c5, but it comes anyway! 19...c5! 20.axb5 Bxg2 21.Kxg2 axb5 22.Nxb5 Qc6† 23.Kg1 Rb8 24.Nc3 cxd4 25.exd4 Rfd8 White’s weak structure and exposed king offer Black full compensation for the pawn.) 17.Rac1 Nf6 18.Ne4 c5 19.Nxf6†?! (19.Qe1!N= was preferable) 19...Bxf6 20.d5 exd5 21.Bxd5 Rd7 White was under pressure in Giri – F. Berkes, Germany 2010. 16.Rab1 c5 17.d5 exd5 18.Bxd5 Nf6 19.Bxb7 Qxb7 20.a4 g6 (20...a6N is also fine, and can be compared with our main line. The presence of White’s rook on b1 instead of c1 does not really change anything.) 21.Rxd8 Rxd8 22.Rd1 Kg7 23.Qb3 This position was reached in So – Efimenko, Wijk aan Zee 2011. Black can choose between a few solid continuations but I prefer:

793

23...Ne4!?N 24.Rxd8 Bxd8 25.Nxe4 Qxe4 26.Nd6 Qe6 Liquidating into an equal endgame. For instance: 27.Qxe6 fxe6 28.Kg2 a6 29.Kf3 Kf8 30.Ke4 Ke7 31.Nc4 Bc7= Finally, I also analysed: 16.Rd2 c5 17.d5 exd5 18.Bxd5 18.Nxd5 Bxd5 19.Rxd5 b5= gives Black no problems.

18...Nf6 19.Rad1 White has nothing better, but now major simplifications are on the cards. 19...Nxd5 20.Nxd5 Bxd5 21.Rxd5 Rxd5 22.Rxd5 Rd8 23.Qd3 Rxd5 24.Qxd5 g6 25.a4 Kg7 26.b3 This happened in Matlakov – Shimanov, Khanty-Mansiysk 2013. White is trying to prove the superiority of his knight over the bishop, but this proves illusory after:

794

26...Qd8!N 27.Qb7 After 27.Qxd8 Bxd8 the endgame is close to equal but more dangerous for White, as Black has chances to create a passer on the queenside, and his bishop may attack the kingside pawns if White’s king moves too far away. 27...Qd1† 28.Kg2 Bh4 Black activates his pieces and seizes the initiative.

16...c5 17.d5 This is certainly the most consistent and ambitious choice. White achieved nothing with 17.Nb5 Qb8 18.Bxb7 Qxb7 19.dxc5 Bxc5 in Alf – Hoffmann, email 2011.

795

17...exd5 18.Bxd5 The unambitious 18.Nxd5 Bxd5 19.Rxd5 b5 20.Ne5 Nxe5 21.Rxe5 a6 has occurred a few times; Black obviously has nothing to worry about. 18...Nf6 19.Bxb7 Qxb7

20.a4 After 20.Rxd8 Rxd8 21.Rd1 g6 22.Ne5 it makes sense to liquidate the strong centralized knight by means of 22...Nd7!. A good example continued:

23.Qa4 Nxe5 24.Rxd8† Bxd8 25.Qe8† Kg7 26.fxe5 Bg5 27.f4 Bh4 28.e4 Qa6 Black had sufficient counterplay in Zawadski – Saitou, email 2013.

796

20...a6 21.Rxd8 The immediate 21.Qb3 runs into 21...Rb8= when ...b5 is coming. 21...Rxd8 22.Qb3 White is focused on stopping ...b5, but the queen’s desertion of the kingside starts to become a factor. In the event of 22.Rd1 Black can even temporarily leave the d-file with: 22...Rb8!? (22...g6 23.Rxd8† Bxd8 24.Qd3 Qd7 is also equal) 23.Ne5 b5 24.axb5 axb5= 22...Ng4!? 22...Rb8 23.Ne5 Bf8= was possible too, but the more active approach is fully justified. 23.e4?! 23.h3N was better, when 23...Qf3 24.hxg4 Qxg4† 25.Kf1 Qh3† 26.Kg1 Qg4†= forces a draw. We have been following the high-level game Grischuk – Kramnik, Moscow 2010. White’s risky last move could have been punished with the following accurate continuation:

23...Qc7!N 24.Ne2 24.Nd5 is also bad in view of 24...Rxd5! 25.exd5 Qxf4µ. Another unfortunate outcome for White could occur after: 24.e5?! Qc6 25.h3

797

25...Nxf2! 26.Kxf2 Bh4† 27.Ke3 Qe6 28.Rf1 Rd4! White’s position collapses. 24...Qc6 25.h3 25.f3 b5! 26.axb5 axb5µ is also nasty for White. 25...Qxe4 26.hxg4 Qxe2 27.Ne5 Rf8 White’s king is weak, so he will have to struggle hard for a draw. D) 10.Rd1

This is the most popular choice, and Avrukh’s recommendation. 798

10...Ba6 Another common continuation is 10...Bb7, but 11.Ne5 offers White chances for an edge without much risk. We have reached a significant branching point, where D1) 11.Nbd2 is not too challenging, but D2) 11.cxd5, D3) 11.b3 and D4) 11.Ne5 all demand serious attention. D1) 11.Nbd2

This method of protecting the c4-pawn has certain drawbacks: the knight is placed rather passively, while the dark-squared bishop has fewer retreat squares available. 11...Rc8 12.Rac1 12.e4?! Nxe4 13.Nxe4 dxe4 14.Ne5 is an overoptimistic attempt by White to regain the pawn in a favourable situation, but a strong reply is: 14...g5! 15.Be3 f5 White remains a pawn down after all! 16.Qa4 Nb8 17.Qb3 Bf6 Black’s chances were clearly better in Grachev – Motylev, Baku 2015. 12...Nh5 13.Be3 13.Ne5 has scored well for White, which is surprising, as giving up the bishop is a clear concession. 13...Nxe5 14.dxe5 (14.Bxe5 f6 15.Bf4 occurred in Drazic – Palac, Saint Vincent 2003, when 15...Nxf4N 16.gxf4 Bb4³ would have been promising for Black) 14...Nxf4 15.gxf4 Bb4 16.b3 Ba3 17.Rb1 Qe7 Black was somewhat better due to his bishop pair in Kersten – Palac, Dresden 2003.

799

13...Bd6! I like this way of handling the position. Now the f4-square is covered and e7 is vacant for the queen. 13...Nhf6 is the most popular choice, but why should we give White the option of repeating moves? 14.Bg5N This seems like a logical attempt to improve on the passive 14.Bf1?!, after which 14...Bb7 15.cxd5 cxd5 16.Qb1 h6 17.Rxc8 Qxc8 18.Rc1 Qb8 was slightly better for Black in Przedmojski – Wojtaszek, Zgierz 2018. 14...f6 15.Be3 g5! This move is not really weakening, since White’s pieces have no access to the kingside. At the same time, Black’s expansion on the kingside looks worrying for the first player. Play might continue:

800

16.h4 g4 17.Ne1 f5 18.Bg5 Be7 19.Bxe7 Qxe7 20.e3 c5 Black has excellent play along the c-file. D2) 11.cxd5 cxd5 Usually this exchange does not bother Black when played at such an early stage. However, White has a concrete idea in mind. 12.Ne5 The knight is heading for c6, in order to force Black to give up one of his bishops. And if Black exchanges on e5, the play becomes rather sharp, as we will see. 12.Nc3 is rather toothless, and 12...b5! is a good reply. For instance: 13.Ne5 (13.a3 Qb6 14.Ne5 Nxe5 15.Bxe5 Bb7 16.Qd3 Bc6 was comfortable for Black in Atalik – Beliavsky, Belgrade 1998) 13...Rc8 14.a3 Bb7 15.Qd3 a6 Black has comfortably equalized, and the impulsive 16.e4?! Nb6 17.exd5 Nfxd5 only gave Black a structural advantage in Nass – Venkatesh, Arinsal 2011. 12...Nxe5! Black should accept the challenge. Instead, after 12...Rc8 13.Nc6 Nh5 14.Bc1 Nb8 15.Nxe7† Qxe7 16.Qd2 Nc6 17.b3 Rfd8 18.Ba3 Qd7 19.Nc3 White had a small but long-term edge thanks to the bishop pair in Gelfand – Tomashevsky, Moscow 2016. 13.dxe5

801

13...Ng4! After 13...Nd7 14.e4 Black is virtually forced to give up a pawn with 14...Rc8 (or 14...d4 immediately) 15.Nc3 d4 16.Rxd4 when he does not have much compensation. Only the brave decision to leap forward with the knight can fully justify Black’s previous play. White may attack it with either D21) 14.Bf3 or D22) 14.h3. D21) 14.Bf3 f5

15.exf6 This is the most practical decision, even though it liberates the knight. 802

After 15.Qa4N 15...g5! 16.Bd2 Bc8 the e5-pawn is in danger, so White is forced to give up an important bishop: 17.Bxg4 Bd7 18.Qc2 fxg4 19.Be3 Rc8 20.Nc3 Be8 Black has excellent play, due to the powerful light-squared bishop. 15...Nxf6 16.Qa4 16.e4N is an ambitious attempt, but it is risky for White to open the centre while behind in development. A good reply is:

16...Qe8! 17.exd5 Nxd5 18.Bxd5 exd5 19.Nc3 Rc8 20.Rxd5 Bf6 Black has great play for the pawn. 16...Bb7 17.Nd2 An equally good alternative is 17.Nc3 Bc5 18.Rac1 Qe7 19.Be5 as played in Gelfand – Karjakin, Yurmala 2015. At this point I like the following ambitious idea:

803

19...Nd7N 20.Bd4 Qf7 21.Rf1 e5 22.Be3 Kh8 Black has a promising position with a mobile pawn centre.

17...Bd6!N It makes sense to exchange the bishops and diminish White’s control of e5, so that Black’s central pawns can advance at the right moment. 17...Qe8 is too passive for my liking, and after 18.Qxe8 Rfxe8 19.Nc4 Black was slightly worse in Sorokin – Fakhrutdinov, Yekaterinburg 2018. 18.Bxd6 Qxd6 19.Nc4 Qc7 20.Ne3

804

20...Qe5 21.Qa3 Rf7 22.Rac1 h5„ Black has an active game, with good prospects on the kingside. D22) 14.h3

14...Rc8 Black develops dynamically and forces White to make a decision. 15.Nc3 This is the most natural and popular continuation.

805

15.Qd2 This move also invites interesting complications. 15...g5! 16.hxg4 gxf4 17.gxf4 Rc4! White is a pawn up but the last precise move gives Black a few targets for attack. In particular, he intends to penetrate into White’s camp along the c-file, and also to attack the kingside pawns along the 4th rank.

18.Nc3!?N This new move seems rather interesting. 18.b3? Bb4 19.Qe3 Rc2µ is bad for White. In the event of 18.g5N 18...Qc7 19.Nc3 Bb4 20.e3 Bxc3 21.bxc3 Rxc3= Black manages to regain the pawn in a comfortable situation. The slower 18.a3 has been played in a couple of correspondence games, in which 18...f6! has proven a reliable reply. Now both kings are getting exposed, but Black’s pieces are much more active. 19.Nc3 fxe5 20.Nxd5! This nice tactical shot enables White to maintain the balance, but nothing more. 20...Rd4 21.Nxe7† Qxe7 22.Qe3 Qd6 23.Rxd4 exd4 24.Qe4 Rxf4 By now White has nothing better than tacitly offering a draw with 25.Qa8† Rf8 26.Qe4= as occurred in Bochev – Azevedo Pessoa, corr. 2012. (In the game Black rejected the immediate repetition and tried playing on, but had to agree a draw a few moves later anyway.) 18...Bb4 19.Bf3 f6! 19...Qc7 seems less convincing in view of 20.e3 Bxc3 21.bxc3 Rxc3 22.Kg2 and White is out of risk.

806

20.f5!? Also after 20.exf6 Qxf6 21.e3 d4 22.Ne4 Qh4 Black has rich play on the kingside. The text move is the most ambitious attempt; White is naturally trying to split the opponent’s central pawns. 20...Bxc3 21.bxc3 fxe5 22.fxe6 Bb7

Black has enough play against White’s exposed king. For instance: 23.e7 Qxe7 24.Bxd5† Kh8 25.f3 Qh4 26.Bxb7 Qg3† 27.Kh1 Qh3†= With perpetual check. 15.Qa4!?N This move has never been played but it looks quite challenging; both of Black’s minor pieces are 807

under attack now.

15...Bxe2 16.Re1 16.hxg4 Rc4 17.Qb3 Rb4 18.Qc2 Bxd1 19.Qxd1 Rxb2 20.Nc3 Qc7 offers Black excellent play along the c-file. 16...b5 17.Qa6 In the event of 17.Qb3 Bc4 18.Qf3 Nxf2! 19.Qxf2 Bc5 20.Be3 d4 21.Bf4 g5 22.Bd2 f6 Black’s initiative fully compensates for the material deficit, to say the least.

17...Nxf2 18.Kxf2 After 18.Rxe2 Nd3 19.Nd2 Nb4 20.Qxb5 Nc2 21.Nb3 Nxa1 22.Nxa1 Bg5 White is equal at best, as his king is rather exposed and his remaining bishop will be restricted by Black’s strong pawn chain. 808

18...Bc4 19.Kg1 b4 20.Qa4 Bc5† 21.Kh2 Bf2

22.Rc1 22.Nd2 Bxe1 23.Rxe1 a5 leads to a complicated position, where Black’s chances are not worse. 22.Rd1 g5 23.Bd2 Be2 24.Bxb4 Bxd1 25.Qxd1 Qb6 26.Qg4 Qe3 27.Nd2 Rc2 28.Bxf8 Kxf8 looks rather dicey for White, whose pieces are almost paralyzed. 22...Bd4 23.Rc2 g5 24.Bc1 Bxe5 25.Qxb4 Qf6 Black’s piece activity yields excellent counterplay.

15...Nxf2! Inviting White’s king to go for a walk.

809

16.Kxf2 Bc5† 17.Be3 17.Ke1?! is too risky in view of 17...f6! when Black’s major pieces will soon be ready to harass the king. 18.Qa4 fxe5 19.Bxe5

19...Bf2†! (19...Bc4 20.Qxc4?? Bf2† led to a quick win for Black in Inarkiev – Yu Yangyi, Doha [blitz] 2016, but 20.Ne4!N would have left the outcome unclear) 20.Kd2 Rf5 21.Rf1 (21.Qxa6 Rxe5ƒ also poses White serious practical problems)

21...d4! 22.Ne4 Be3† 23.Ke1 Rxe5 24.Qxa6 Qc7 25.Qd3 Rc5 26.Rd1 Rc1 27.Rf3

810

27...Rxd1† 28.Qxd1 Qc1 29.Nf2 Bd2† 30.Kf1 Qxd1† 31.Nxd1 Rc1µ Black finally regained the piece to reach a clearly better endgame in Calkins – Caron, corr. 2016.

17...Bxe3† 18.Kxe3 b5! This is the most effective way to activate the queen and keep White’s king under pressure. My other idea was 18...f6!?N, but I rejected it on account of 19.Kf2 b5 20.Qd2 b4 21.Na4 Qe8 22.b3 fxe5† 23.Kg1 Qg6 24.Qe3 Rc2 25.Re1 when Black does not have full compensation for the piece. 19.Qd3 19.Qd2 Qb6† 20.Qd4 is the same.

811

19...Qb6† 20.Qd4 Qb8 At the moment, Black only has a single pawn for the piece. On the other hand, the e5-pawn might fall at any moment, and White will have to spend precious time securing his king. 21.Kf2 This seems more natural; White’s king has to leave the danger zone. 21.Qb4 can be met by 21...Qb6† when White has nothing better than repeating the position. Instead 22.Kd2?! would be playing with fire: 22...Rc4 23.Qd6 Rc6 24.Qb4 Rfc8‚ I also considered: 21.b3N 21...f6 22.exf6 Rxf6 23.Bf3 (also after 23.Rac1 Qxg3† 24.Bf3 b4 25.Nxd5 exd5 26.Qxd5† Rf7 27.Rxc8† Bxc8= Black is out of risk) 23...b4 24.Na4

812

24...Bxe2!! 25.Kxe2 (25.Bxe2? is refuted by 25...Qxg3† 26.Kd2 Qg5† 27.Qe3 Rc2† 28.Kd3 Qg6† 29.Kd4 Rf2–+) 25...Qxg3 26.Rf1 Rc2† 27.Kd1 Rh2 White’s position lacks harmony and he will not be able to maintain his material advantage. For instance: 28.Qe3 Qxh3 29.Qd3 Rxf3 30.Rxf3 Rh1† 31.Ke2 Qh2† 32.Rf2 Qh5† 33.Qf3 Qxf3† 34.Kxf3 Rxa1 Black has four pawns for the piece and definitely should not face any problems, although White should be able to hold for a draw.

21...b4 22.Nb1 A sad necessity – White’s development is interrupted again. 22.Na4?! runs into 22...Qb5, after which 23.b3? Qxe2† 24.Kg1 Rc2–+ leads to a quick disaster. 813

22...f6! Black is not too concerned about regaining material – it is more important to involve the rooks in the attack. 23.Qe3 I also checked: 23.exf6 Rxf6† 24.Kg1 White’s king reaches a relatively safe spot, but now White’s pawns start to fall. (24.Bf3? Rc4 25.Qe3 Re4–+) 24...Bxe2 25.Re1 Qxg3 26.Nd2 Ba6 27.Rac1 Rcf8

28.Nf1 Bxf1 29.Rxf1 Rxf1† 30.Rxf1 Rxf1† 31.Kxf1 a5= Black’s three pawns are definitely worth a piece.

814

23...fxe5† 24.Kg1 e4! The mobility of White’s pieces will be heavily restricted by Black’s central pawns. 25.Nd2 Rc2 26.Rab1 Qb6! 27.Qxb6 axb6 Despite the material deficit, Black does not mind exchanging queens. Indeed, without the support of the queen, White will have a hard time protecting his weak pawns.

28.Bf1 In the event of 28.Nb3 Bxe2 29.Rdc1 Rxc1† 30.Rxc1 e5 31.Nd2 Bd3 32.Rc7 Rf6 Black’s passers are becoming powerful. 28...e3 29.Nf3 Bxe2 30.Bxe2 Rxe2 31.Nd4 Ref2 32.Re1 After 32.Rf1 Rd2 33.Rxf8† Kxf8 34.Nxe6† Ke7 35.Ng5 e2 36.Re1 d4 37.Kf2 d3 White is certainly not playing for a win.

815

32...e5!N 32...Rd2 33.Nb3 Rc2 was less convincing in Dominguez Perez – Karjakin, Saint Louis 2018. Had White played 34.Rxe3!N 34...Rff2 35.Nd4 Rg2† 36.Kf1 Rcf2† 37.Ke1 Rh2 38.Nf3 Black would not have been able to claim sufficient compensation for the piece. 33.Ne6 Rc8 34.Rxe3 White has finally liquidated the dangerous e3-pawn, but now both of the black rooks enter into his camp. 34...Rcc2 35.Rxe5

816

35...Rg2† 36.Kf1 Rh2= After accurate play by both sides, a draw is the logical outcome. D3) 11.b3

As mentioned earlier, this position can arise via other move orders such as 9.Rd1 b6 10.b3 Ba6 11.Bf4. 11...Rc8 12.Nc3 In addition to this most popular move, I checked a couple of rare alternatives: 12.Nbd2?! is too passive and Black easily gets a good game: 12...c5 13.e4 (even worse is 13.dxc5?! Bxc5 14.Qb2 Qe7µ as in Istratescu – Bossard, Paris 2014) 13...cxd4 14.exd5 Nxd5 15.Nxd4 Nxf4 16.gxf4 Qc7³ Black enjoyed both the better pawn structure and the bishop pair in Grabuzova – Kovalevskaya, Dagomys 2008. 12.a4!? This move is more interesting; White is playing in the spirit of the more popular a2-a4 line which was analysed in variation D of the previous chapter. 12...Re8!? I like this prophylactic move, which invites White to show his cards. 12...c5 is a logical alternative, when 13.Na3 cxd4 14.Nb5 offers White interesting play for the pawn. 13.a5 13.Nc3?! sees White mixing up his plans, and 13...dxc4 14.bxc4 Bxc4 15.Ne5 Nxe5 16.dxe5 Nd5 17.Ne4 Ba6 18.h4 h6³ leaves him with insufficient play for the pawn. 13.Na3 transposes to one correspondence game but it does not make much sense to me. Black could have exploited the lack of harmony in his opponent’s camp by means of 13...Bxa3!?N 14.Rxa3 c5 817

15.a5 b5, with excellent play along the c-file. 13...b5 14.c5 b4 Compared to the main line of the previous chapter, where White goes for a4-a5 with his bishop still on d2, here the desertion of White’s bishop from the queenside enables Black to activate his lightsquared bishop much more easily. 15.Re1 Ne4 16.Ne5 Nxe5 17.Bxe5 f6 18.Bf4 We have been following Heffalump – Hent, Internet (freestyle-rapid) 2006. My new idea is:

18...Ng5!?N 19.Nd2 White also achieves no advantage with 19.Bxg5 fxg5 20.Nd2 Bf6 21.Nf3 Qc7„. 19...Nf7 20.Bh3 f5 Black has excellent prospects on the kingside. Returning to our main line, we reach a popular theoretical position which can also arise via a 4.g3 Queen’s Indian. 12...dxc4 Black has a wide choice of logical continuations, including 12...h6, 12...Nh5, 12...Qe8 or even 12...Re8!?. However, accepting the pawn sacrifice seems most challenging. White has three main replies: D31) 13.Nd2, D32) 13.e4 and D33) 13.bxc4!?. 13.Ne5 is harmless due to 13...Nxe5 14.dxe5 Nd5, since the pin along the d-file is not dangerous for Black. For instance:

818

15.Ne4 (15.e4? Nb4 16.Qe2 Nd3 is rotten for White) 15...c3!? Securing the knight on d5. (15...b5!? also seems perfectly playable.) 16.Nxc3 Qc7 17.Nxd5 cxd5 18.Qxc7 Rxc7= Black had no problems in Konopka – Stohl, Czech Republic 2000. D31) 13.Nd2

This move blocks the d-file and seems to render White’s set-up less harmonious. On the other hand, the powerful light-squared bishop gets a clear view of the long diagonal. 13...b5 White’s idea is illustrated after 13...cxb3? 14.axb3± when the open a-file causes Black serious 819

problems. 14.bxc4 bxc4 15.Qa4 White has also tried: 15.e4 Qa5 16.Bf1 The most consistent move, trying to regain the pawn. 16.Na4?! occurred in Rustemov – Mitkov, Kallithea 2002, when 16...g5!N 17.Be3 Ng4 18.Nxc4 Bxc4 19.Qxc4 Nxe3 20.fxe3 c5 would have left White in a strategically difficult situation. 16...Bb4 17.Bxc4 Bxc3 18.Bxa6 We have been following the game Oganian – Zenzera, Batumi 2014. I see no reason for Black to deviate from the following natural continuation: 18...Qxa6N 19.Qxc3 c5 With excellent play.

15...Bb5! A nice tactical nuance, which forces White to find a novelty to maintain the balance. 16.Qxa7!N 16.Nxb5?! is well met by 16...Nb6! 17.Qxa7 cxb5 18.Bb7 c3! 19.Bxc8 Nxc8 20.Qa6 cxd2³ as in Bologan – Doettling, France 2008. The text move is more critical. White’s queen is in danger, but Black’s pieces cannot attack it yet. 16...g5! The idea behind this move is to lure the bishop away from controlling the b8-square. 16...Nb6 17.a4 Bb4 18.axb5 Bxc3 19.Ra6 Nfd5 leads to interesting play but I find the text move more 820

convincing.

17.Bxg5 Accepting the challenge. 17.a4!? gxf4 18.axb5 cxb5 19.Nxb5 c3 leads to double-edged play, where Black’s chances are not worse. 17...Ra8 18.Nxb5!? 18.Qb7 Rb8 19.Qa7 Ra8= is a repetition but we should also see what happens if White gives up his queen for some pieces. 18...Rxa7 19.Nxa7

821

19...Nb8! It is necessary to cover c6. 20.Nxc6 20.Bf4 Bd6 is fine for Black. 20...Nxc6 21.Bxc6 Qxd4 The combination of the passed a-pawn and the powerful light-squared bishop could prove dangerous, so we will extend the analysis to show how Black should deal with it.

22.Ne4 Qe5 23.Bxf6 Bxf6 24.Nxf6† Qxf6 25.a4 c3 26.Rdc1 Rc8 27.Bf3 c2 822

The c-pawn provides enough counterplay, so Black is out of danger. D32) 13.e4

13...b5! The tempting 13...Bb4 is well met by 14.e5! Bxc3 15.exf6 Bxa1 16.fxg7 Re8 17.Rxa1ƒ and White obtains a powerful initiative for the exchange. 14.bxc4 bxc4 15.Rab1 White takes control over the open file and invites Black to show his cards. 15.Nd2 transposes to 15.e4 in the notes to variation D31 above. The ambitious 15.Ne5 was played in Polak – Zpevak, Czech Republic 2012. Here it would have made sense to avoid the annoying opposition along the d-file by means of:

823

15...Qa5!N 16.Bd2 Qa3 17.Nxd7 Nxd7 18.Rab1 Rfd8 Black has no problems at all. White also achieves nothing with: 15.Qa4 Bb5 16.Nxb5 16.Qxa7?! occurred in Karthikeyan – Ramakrishna, Bhiwani 2010, when Black missed a strong idea: 16...Ra8!N 17.Qb7 Qa5 18.Nxb5 cxb5 19.Rdb1 b4ƒ The passed c-pawn is dangerous, especially since White’s queen is stuck in Black’s camp. 16...cxb5 17.Qxb5 Nxe4 18.Rac1 This position occurred in few games, but for some reason nobody played the natural continuation:

18...c3N 19.Ne5 Nxe5 20.dxe5 Nd2 The c3-passer is annoying, so White has to struggle for equality. 824

Finally, an aggressive try for White is: 15.d5!? cxd5 16.exd5 This was tested in Psakhis – Huzman, Tel Aviv 1994, plus some subsequent outings. So far, nobody found the best antidote for Black:

16...Qa5!N Instead, 16...exd5 17.Nxd5 Nxd5 18.Rxd5 Bb7 19.Rd2 offers White interesting compensation for the pawn due to the pin along the d-file. 17.Ng5 In the event of 17.dxe6 fxe6 18.Ng5 Nc5 19.Bh3 Nd3 20.Bxe6† Kh8 21.Bxc8 Bxc8 Black obtains more than enough compensation for the exchange. 17...Nc5 18.d6 White’s main hopes are connected with the powerful passed pawn. Instead, 18.dxe6 fxe6 transposes to the note above. 18...Bd8 19.d7

825

19...Rc7! 20.Nce4 20.Bxc7 Bxc7 21.Nce4 Nd3 is the same thing. 20...Nd3 21.Bxc7 Bxc7 22.Nxf6† gxf6 23.Ne4 Be5 24.Rab1 Rd8 The powerful knight, supported by the bishop pair, offers Black excellent play. 15...Qa5 16.h3 This useful prophylactic move, which secures the e3-square for the bishop, has been the choice of three GMs, all of whom won with White. 16...h6 Black adopts a similar waiting strategy. Now it is hard to suggest another obvious move to strengthen White’s position.

826

17.Re1!? In the event of 17.Bd2 Qc7 White has nothing better than searching for a repetition with 18.Bf4. If he plays more ambitiously with 18.Na4, then 18...Bb5 19.Re1 a5 20.Bf1 Qa7! 21.Bxc4 Qa6 22.Nb2 a4 puts White’s central pawns in serious danger. 17.g4?! proved to be overoptimistic in the following game: 17...Rfe8 18.Bd2 (18.g5 hxg5 19.Bxg5 e5!µ) 18...Qc7 19.Qc1 Rb8 20.Rxb8 Qxb8

21.g5 hxg5 22.Bxg5 c5 23.e5 Nd5 Despite his eventual defeat, at this stage Black was clearly better in Lputian – Mchedlishvili, Istanbul 2003. 17...Rfd8 18.Bd2 Qc7 19.Qa4 827

As usual, this move enables White to regain the pawn, but the c4-pawn will become more powerful. 19...Bb5 20.Nxb5 cxb5 21.Rxb5 c3„ All this happened in Ftacnik – Mazur, Slovakia 2016. In this case too, Black eventually lost the game, but at this stage he was at least equal thanks to the strong passed pawn. D33) 13.bxc4!?

For some reason, this move has not been tested much at high levels. Nevertheless, it seems the most principled, as it enables White to restore material equality by force. 13...Bxc4 14.Nd2 As often happens in this line, the light-squared bishop has no ideal retreat square. 14.e4?! Bb4 15.e5 Nd5 16.Ng5 g6 leaves White without adequate compensation for the pawn. 14...Ba6 I also examined 14...Bb5, but after 15.Qb3 Black has nothing better than retreating the bishop to a6 anyway, since 15...Nd5?! 16.Bxd5 exd5 17.Nxb5 cxb5 18.Qxb5 Nf6 19.a4² only helps White.

828

15.Qa4 Bb7 16.Qxa7 Ba8 White has regained the sacrificed pawn while maintaining the merits of his set-up, such as extra central space and active prospects for his pieces. On the flip side, the poor location of the queen on a7 yields Black some tactical possibilities; in particular, the thematic ...c5 break will gain in effectiveness. 17.e4 The quiet 17.Rac1 can be met by: 17...c5 18.dxc5 (an equally good alternative is 18.Nde4N 18...Nxe4 19.Nxe4 Bd5 20.dxc5 Ra8 21.Qc7 Nxc5=) 18...Bxg2 19.Kxg2 Ra8 20.Qb7 Nxc5

21.Qf3 (21.Qc6N 21...Ra6 22.Nc4 Qa8= does not change anything) 21...Ra6 22.Nc4 Qa8= Black was fine in Ulko – Kharitonov, St Petersburg 1999. Seizing the centre is White’s only real attempt to claim an edge. I propose a new plan for Black: 829

17...h6!N Not for the first time, this move is rather useful. Apart from avoiding back-rank problems later in the game, Black also makes it possible to throw in a ...g5 advance in order to force the bishop to give up control of the b8-square, thus putting White’s queen in more danger. I am not so keen on the previously played 17...Nh5 (or 17...Bb4N 18.Rac1 Nh5 19.Be3) 18.Be3 Bb4 19.Rac1 b5 because of the following improvement:

20.a3!N (20.d5 Bxc3 21.Rxc3 cxd5 was fine for Black in Baryshpolets – Kruglyakov, Kiev 2008) 20...Rc7 21.Qxc7 Qxc7 22.axb4 In this unbalanced position I prefer White. 18.h4 830

This way White secures the bishop on f4 for a while, but weakens the g4-square. Transferring the knight to d6 seems ineffective: 18.Nc4 b5 19.Nd6 Rb8

20.e5 White’s pieces were somewhat unstable, so this move seems best. 20...Bxd6 21.exd6 b4 22.Na4 Rb7 23.Qa6 Rb5! It turns out that the queen is still in danger! In particular, ...Ra5 is a deadly threat.

24.Nc5 Nxc5 25.dxc5 Rxc5 26.d7 Ra5 27.Qc4 Nxd7 Liquidating the strong passer is a big achievement for Black. 28.Rac1 c5³ 831

White will have to struggle for a draw. I also checked: 18.Rac1?! This careless move allows Black to strike out immediately with: 18...g5! 19.Be3 b5! Covering a4, so the queen will remain stuck on a7 for a while. 20.e5 Ng4

21.Nde4 In the event of 21.Nb3?! Nxe3 22.fxe3 Bb4 White’s position seems shaky. For instance, 23.a3 (23.Ne4? drops the queen: 23...Rc7 24.Qa6 Bb7 25.Qa7 Bc8 26.Qa8 Nb6 27.Qb8 Rb7–+) 23...Rc7 24.Qxc7 Qxc7 25.axb4 c5! Black happily gives up a pawn to improve the prospects of all his pieces. 26.Bxa8 Rxa8 27.bxc5 Qc6µ 21...Nxe3 22.fxe3 Rc7 23.Qa5 Rb7 24.Qxd8 Rxd8³ Black liquidates into a promising endgame, with ...c5 coming soon. 18...Bb4 19.Rac1 c5! 20.Bd6 20.Qa4 Qe7 21.dxc5 Bxc5³ puts White’s kingside under pressure; the weakness of the g4-square is relevant here. 20...Re8 21.Qa4

832

21...Ba5 21...Bxc3!? 22.Rxc3 b5! is an interesting alternative: 23.Qxb5 cxd4 24.Rxc8 Qxc8 25.e5 Bxg2 26.Kxg2 Nd5 27.Ne4 Qc2 28.Rxd4 Rd8© gives Black decent compensation for a pawn. 22.dxc5 Bxc3 23.Rxc3 Nxc5 24.Bxc5 Rxc5 25.Rxc5 bxc5 26.Nc4 Qb8 27.e5 Bxg2 28.Kxg2 Nd5 29.Nd6 Rd8= Black’s knight on d5 is just as strong as White’s on d6. D4) 11.Ne5

This popular continuation is Avrukh’s recommendation in Grandmaster Repertoire 1A, so we 833

obviously have to take it seriously. One idea of this move is to prepare e2-e4, since the g2-bishop is unblocked now. Another point is to try and get an improved version of variation D2 above, as the addition of the pawns on c4 and c6 reduces Black’s active possibilities in certain lines. 11...Rc8! Unlike the earlier variation D2, here Black should avoid 11...Nxe5 12.dxe5 Ng4 (12...Nd7 13.cxd5 cxd5 14.e4 has been mentioned on page 393 – see the 13...Nd7 line in the notes to variation D2), when Avrukh points out the strong improvement of 13.Bf3!N which yields White some advantage. The lack of an open c-file clearly inhibits Black’s activity here. After the text move, White usually either releases the central tension with D41) 12.cxd5 or maintains it with D42) 12.Nc3, the latter move implying a pawn sacrifice. As usual, 12.Nd2?! leaves the dark-squared bishop short of squares. 12...Nxe5 13.Bxe5 (13.dxe5 is hardly better in view of 13...Nd7 14.h4 h6 15.h5 b5! when Black seizes the initiative) 13...Ng4

14.e4 (14.Bf4N 14...g5! 15.Be5 Nxe5 16.dxe5 Qc7³ is problematic for White, as defending the loose pawn with f2-f4 would be too weakening) 14...Nxe5 15.dxe5 d4 16.Qa4 Bb7 17.Qxa7 Qc7 18.Qa4 c5³ Black had more than enough play for the pawn in Ibrayev – Halkias, Dresden (ol) 2008. White also gets nowhere after: 12.Qa4 Nb8 Of course, not 12...Bxc4? 13.Nxc6±. Now the poor location of White’s queen may start to tell. For instance:

834

13.Nd2 The tempting 13.Nc3? is tactically refuted by 13...b5! 14.Qc2 (or 14.cxb5 cxb5 15.Nxb5 Qb6 16.Nc3 Qxb2–+ and White’s position is ruined) 14...bxc4 15.b3 cxb3 16.axb3 c5 and Black was simply a pawn up in Hamitevici – Kuljasevic, Paleochora 2012. 13...b5 14.cxb5 cxb5 15.Qb3 Qb6 16.Qe3 Nc6 17.Nb3 b4 Black had the more comfortable play in Tukmakov – Korotylev, Geneva 2001. D41) 12.cxd5

This was recommended by Avrukh and it demands an accurate response.

835

12...Nxd5! 12...cxd5 transposes to the 12...Rc8 line mentioned on page 393, in the notes to variation D2, where it was noted that 13.Nc6 offers White a safe position with a small edge thanks to the two bishops. 13.Nc3 This is definitely the right choice. Of course we should also check what happens if White greedily grabs the c6-pawn: 13.Nxc6?! Nb4! 14.Qa4 Nxc6 15.Bxc6 Bxe2 16.Rc1 After 16.Re1 b5 17.Qxa7 Rxc6 18.Rxe2 Nb6 19.Na3 Nd5 Black had more than enough play for the pawn in H. Graf – Witzschel, email 2013. 16...Nf6 17.Nc3 Bd3 Black has no problems with development and the isolated d-pawn might become a significant weakness in the long run. For instance: 18.Rd1 Bc2 19.Qxc2 Rxc6 20.Qb3 This was Leutwyler – Tan, Helsingor 2017. My suggestion for Black is:

20...Nd5!?N 21.Nxd5 Qxd5 22.Qxd5 exd5 23.Rac1 Rfc8 It should be a draw, but Black has a tiny advantage due to his better bishop and the target on d4. 13...Nxe5 14.Bxe5 14.dxe5 is strategically risky, and after 14...Qc7 15.Ne4 Rfd8 Black may soon be ready to mobilize his queenside pawn majority. For instance: 16.Ng5 g6 17.h4

836

We have been following Halkias – Nakar, Paleochora 2016, and now the natural 17...Nxf4N 18.gxf4 c5 19.Be4 Bb7³ would have put White in a strategically difficult situation.

14...f6! This modern discovery solves all of Black’s problems in this line. In his 2015 book Grandmaster Repertoire 1A, Avrukh only considered 14...Nxc3, which had been played in all three of the games at the time, and showed that after 15.bxc3 White has decent chances to generate some pressure. The text move was introduced as an improvement for Black late in 2015, and has since proved a reliable choice in many games. 15.Bf4 Nxf4 16.gxf4

837

16...f5! An excellent move. Black’s pawn structure may contain some weaknesses but White is not well placed to exploit them (for instance, the knight is a long way from e5) and there are certain dynamic factors in Black’s favour, such as the bishop pair and the vulnerability of the white king. 17.Qa4 The above-mentioned merits of Black’s set-up were well illustrated in the following encounter: 17.e3 Rc7 18.Qa4 Qc8 19.Rac1 Kh8 20.Rd2 Rg8 21.Ne2 g5! Black took over the initiative in Banikas – Fressinet, Baku (ol) 2016. The quiet 17.Rac1, as in Erdos – Lupelescu, Romania 2015, can be met by 17...Rf6!N 18.e3 Rg6, with great play on the kingside. The text move is the most energetic choice. Black has to retreat the bishop, so the a7-pawn must fall.

838

17...Bb7 18.Qxa7 Qc7 19.Rac1!? This is a risky attempt to justify White’s pawn grab. The alternative is 19.Nb5 Qd7 with the following options:

a) Safest, and probably best, is 20.Nc3 Qc7= with a repetition. b) 20.Qxb6?? loses material after 20...Bd8 21.Qa7 Ra8–+ when the queen must retreat to c5, allowing Black to eat the knight on b5. c) 20.Rac1 is risky in view of 20...Rf7! 21.Nc3 b5, when the poor white queen will be out of play for a long time. 839

d) 20.Na3 is an attempt to preserve White’s extra pawn, but Black will have plenty of compensation. 20...b5 21.Qa5 occurred in Andersen – Thybo, Denmark 2018. Black has several playable continuations but 21...Kh8!N looks especially appealing, with ideas of attacking along the g-file. 19...Bh4! Reminding White that his king is not so safe. Black must avoid 19...Ra8? 20.Nb5± of course. 20.Nb5 Bxf2†! 21.Kh1 After 21.Kxf2 Qxf4† 22.Bf3 cxb5 23.Rxc8 Qxh2† 24.Kf1 Bxc8³ White’s king is too exposed. 21...Qd7 22.Qxb6?! 22.Nd6!N 22...Qxd6 23.Qxb7 Qxf4 24.Qxb6 would have maintained the balance. This position occurred in Georgiadis – Bok, Biel 2016. Black could have obtained some advantage with:

22...Ba8!N With the knight on b5 hanging and ...Be3 threatened, White is under some pressure. Here is an illustrative line: 23.Na7 Rb8 24.Qc5 Rxb2 25.Nxc6 Be3 26.Rc2 26.Ne7†?! Kf7µ hardly helps White.

840

26...Rxc2 27.Qxc2 Bxf4³ White suffers from a weak kingside and an excessive number of pawn islands. D42) 12.Nc3

This position has been debated in more than eighty games. I was surprised to discover that my preferred choice for Black has only been played in one of them! 12...Qe8!? Despite its rarity, this is quite a thematic idea. Black avoids the unpleasant opposition along the d-file, making it more feasible to exchange on e5. This idea is not considered by either Bologan in The Powerful 841

Catalan, or Hilton & Ippolito in Wojo’s Weapons. Most games have continued 12...Bxc4 13.Nxc4 dxc4. Although this seems perfectly playable, White’s light-squared bishop becomes rather powerful, so he is assured of decent compensation for the pawn. There is only one game after the text move, so what follows is almost entirely original analysis. The two most critical moves are D421) 13.e4 and D422) 13.Rac1N. Other moves do not offer White much: 13.cxd5N 13...cxd5 14.Rac1 Nh5 15.Be3 Nb8 (if a draw is an acceptable result, there is nothing wrong with 15...Nhf6=) 16.Qb3 Nc6 The knight is heading for a5 and c4, so White had better exchange it. 17.Nxc6 Rxc6 Black has no problems. 13.Qa4N 13...Nb8

14.cxd5 14.e4?! only invites trouble for White after 14...b5! 15.cxb5 cxb5 16.Qb3 b4. The slow 14.a3 can be met by 14...Nh5! (but not 14...dxc4? 15.Nxc6 Rxc6 16.Bxb8±) 15.cxd5 Nxf4 16.gxf4 cxd5 17.Qxe8 Rfxe8 18.f5 Nc6 19.Nxc6 Rxc6 20.fxe6 fxe6 21.Nxd5 Rc2 and Black is out of danger. 14...cxd5 15.Qxe8 Nxe8! This precise recapture makes Black’s position more harmonious. The knight is heading for d6 and later c4. Instead, 15...Rfxe8 16.a4 offers White a tiny edge.

842

16.e4 16.a4?! is now pointless in view of 16...g5 17.Be3 f6 18.Nd3 Nd6³. 16...dxe4 17.Nxe4 Nc7 18.Nc3 Bb4 19.Rac1 Bxc3 20.Rxc3 Nd5 Black has no problems. D421) 13.e4

This was the continuation of Carron – Simon, Lenzerheide 2010. My improvement is: 13...dxc4N 14.Nxd7 Qxd7 15.d5 White’s pieces are currently more active, so opening the centre is the most natural way to develop his 843

initiative. 15...cxd5 16.exd5

16...Bc5! However, it turns out that Black can also benefit from the liquidation of the centre, as now there are more good squares available for his minor pieces! 17.dxe6 Qxe6 18.Re1 Qd7 19.Rad1 We have a funny situation where Black’s queen is almost trapped in the middle of the board, but he has a nice resource available.

844

19...Rce8! The only move, but good enough. 20.Be5 Qe6 21.Re2 White is still targeting the queen, but Black has some attacking ideas of his own. 21...Bxf2†! 22.Kh1 Bc5

23.Bd6 Qxd6 24.Rxd6 Bxd6 Black is not worse at all, as he has full material compensation for the queen and his king is safer. D422) 13.Rac1N

845

This seems most challenging. White increases the pressure along the c-file and tries to punish Black for restricting the f8-rook. 13...Nh5 14.Be3 The other retreat 14.Bd2 is well met by 14...c5!, for instance: 15.Qa4 (or 15.Nxd7 Qxd7 16.dxc5 Bxc4 17.cxb6 axb6 18.Be3 Nf6 19.Bxb6 Bxa2 20.Bd4 Qb7 when Black has fully consolidated while maintaining his central pawn majority) 15...Nxe5 16.dxe5 Bxc4

17.Qxa7 (White must avoid 17.Qxe8? Rfxe8 18.b3 Ba6 19.g4 d4 20.Nb1 Bxe2µ) Black can and probably should force a draw with: 17...Ra8 18.Qb7 Rb8 19.Qc7 Rc8 20.Qxb6 Rb8= 14...Nxe5! 15.dxe5 Bxc4! 846

Offering a promising piece sacrifice.

16.g4 The knight is trapped, but White’s king is becoming more exposed. 16...f5! 17.gxh5 f4 Black’s pieces are active and his attack will quickly gather steam. Now we can appreciate the hidden value of the queen on e8! 18.Bd2 Qxh5 Black already has two pawns for the piece, and a third may fall at any moment. 19.Kh1 19.Bf3 Qxe5 20.Kh1 Bc5 21.Rf1 Rf6ƒ is also promising for Black. 19...Bc5! The immediate 19...Qxe5 allows White to secure the king with 20.Rg1 Bh4 21.Rcf1, although even here the evaluation is unclear. 20.Be1 Rf5 Black has excellent play for the piece. I will include a few more lines showing some tactical possibilities.

847

21.b3 Ba6 22.b4 Rg5! The greedy 22...Bxb4?! sees Black getting distracted from the important target, and 23.Qa4 Ba5 24.Bf3 leaves him suffering from poor coordination between the pieces. 23.f3 23.bxc5?? is refuted by 23...Qg4 with checkmate to follow. 23...Rxg2!? Not the only decent move, but the most straightforward for theoretical purposes. 24.Kxg2 Qg5†

848

25.Kh3 25.Bg3? Be3 is winning for Black. 25...Qh5† 26.Kg2 Qg5†= With perpetual check. Conclusion This chapter has concluded our Catalan coverage by analysing the popular 8.Qc2 Nbd7 9.Bf4, which can be considered the ultimate main line against our chosen set-up. We meet it with the standard developing plan of 9...b6, when the path divides. The options of 10.Nbd2 10.cxd5 and 10.Nc3 all have a right to exist; at the same time, they all have certain drawbacks and can be neutralized with accurate play. 10.Rd1 is the most thematic Catalan continuation, anticipating the future opening of the d-file. 10...Ba6 is our reply, when there are three quite challenging continuations. 11.cxd5 cxd5 12.Ne5 has the idea to penetrate to c6, but 12...Nxe5! 13.dxe5 Ng4! is a principled reply, whereby Black prepares to sacrifice the knight on f2 if the situation demands it. 11.b3 Rc8 12.Nc3 is another popular position, which can arise via several different move orders. 12...dxc4 is our reply, when White has three main options, all leading to interesting play, but with Black keeping a full share of the chances, often relying on dynamic resources such as meeting Qa4 with ...Bb5, returning the extra pawn in order to make the passed c4-pawn more powerful.

849

Finally, 11.Ne5 is a principled option but 11...Rc8! is a good reply. In the event of 12.cxd5, it is important for Black to recapture with the knight, keeping in mind a specific antidote to Avrukh’s recommended line. We ended the chapter by checking 12.Nc3, at which poin12...Qe8!? is a strong move which has the added bonus of considerable surprise value. The piece sacrifice in variation D422 is an important detail to justify Black’s opening play, and the plan of ...f5 and ...Qxh5 makes an especially nice impression.

850

Variation Index Chapter 1 1.d4 Nf6 2.c4 e6 3.Nf3 b6 A) 4.Bg5 Bb7 8 A1) 5.Nc3 h6 6.Bh4 Be7 7.Qc2 c5 8 A11) 8.Rd1 10 A12) 8.dxc5 12 A2) 5.e3 h6 6.Bh4 Be7 7.Nc3 0-0 13 A21) 8.Be2 14 A22) 8.Bd3 15 B) 4.Bf4 Bb4†!? 16 B1) 5.Nbd2 17 B2) 5.Nfd2 Nh5! 19 B21) 6.a3!? 19 B22) 6.Bg3 20 Chapter 2 1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.e3 Bb7 A) 5.a3 24 B) 5.Be2 26 C) 5.Nbd2 c5 29 C1) 6.b3 30 C2) 6.Bd3 31 D) 5.Nc3 d5 32 D1) 6.a3 32 D2) 6.cxd5 exd5 7.Bb5† c6 35

851

D21) 8.Ba4!? 35 D22) 8.Bd3 36 Chapter 3 1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.e3 Bb7 5.Bd3 d5 A) 6.cxd5 40 B) 6.0-0 Bd6 41 B1) 7.b4!? 41 B2) 7.Nc3 0-0 43 B21) 8.Qe2 43 B22) 8.cxd5 exd5 44 B221) 9.Ne5 44 B222) 9.Qc2 45 B3) 7.b3 0-0 8.Bb2 Nbd7 47 B31) 9.Nbd2 Ne4 48 B311) 10.Rc1 48 B312) 10.Qc2 49 B32) 9.Nc3 Re8!? 50 B321) 10.cxd5 51 B322) 10.Rc1 53 Chapter 4 1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.Nc3 Bb7 A) 5.Qd3 57 B) 5.Bf4 Bb4! 58 B1) 6.e3 58 B2) 6.Qb3 60 C) 5.Qc2 c5 61 C1) 6.e4 62

852

C2) 6.dxc5 bxc5 65 C21) 7.Bg5 65 C22) 7.Bf4 67 D) 5.g3 Bb4 68 D1) 6.Bd2 68 D2) 6.Bg2 70 Chapter 5 1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.a3 Bb7 5.Nc3 d5 A) 6.Qa4† 75 B) 6.Bf4 77 C) 6.Qc2 dxc4! 7.e4 c5 79 C1) 8.Bf4!? 82 C2) 8.d5 83 D) 6.Bg5 Be7 87 D1) 7.cxd5 88 D2) 7.Qa4† Qd7! 89 D21) 8.Qxd7† 89 D22) 8.Qc2 91 D3) 7.Bxf6 Bxf6 8.cxd5 exd5 93 D31) 9.Qb3 94 D32) 9.Qa4† 95 D33) 9.g3 96 D4) 7.e3 0-0 98 D41) 8.Rc1 98 D42) 8.Qc2 100 Chapter 6 1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.a3 Bb7 5.Nc3 d5 6.cxd5 Nxd5

853

A) 7.e4?! 104 B) 7.Qa4† 106 C) 7.Nxd5 Qxd5! 110 C1) 8.g3 110 C2) 8.e3 112 Chapter 7 1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.a3 Bb7 5.Nc3 d5 6.cxd5 Nxd5 A) 7.Bd2 Nf6 115 A1) 8.Bg5 116 A2) 8.Qc2 117 B) 7.e3 Be7 118 B1) 8.Qc2 119 B2) 8.Bb5† c6 121 B21) 9.Nxd5 122 B22) 9.Bd3 c5! 124 B221) 10.0-0 124 B222) 10.Nxd5 125 B223) 10.e4 Nxc3 11.bxc3 Nc6 12.0-0 cxd4 13.cxd4 0-0 14.Be3 Rc8 15.Qe2 Na5 127 B2231) 16.Rfd1 129 B2232) 16.Rfe1 131 Chapter 8 1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.a3 Bb7 5.Nc3 d5 6.cxd5 Nxd5 7.Qc2 Nxc3 A) 8.Qxc3 h6! 139 A1) 9.b4 139 A2) 9.Bf4 140 A3) 9.e3 142 A4) 9.g3 143

854

B) 8.bxc3 c5 9.e4 Nd7 145 B1) 10.Bd3 145 B2) 10.Bf4 Be7 149 B21) 11.dxc5?! 150 B22) 11.d5 151 B23) 11.Bd3 Qc8! 153 B231) 12.Qa4 154 B232) 12.Qb1 155 Chapter 9 1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Ba6 5.Nbd2 d5 A) 6.cxd5 exd5 7.Bg2 Bd6 8.0-0 0-0 160 A1) 9.Re1 162 A2) 9.Ne5 163 B) 6.Bg2 Be7 7.0-0 0-0 164 B1) 8.cxd5 165 B2) 8.Ne5 c5 167 B21) 9.Nb3 167 B22) 9.dxc5 169 B3) 8.b3 c5 171 B31) 9.dxc5 171 B32) 9.cxd5 173 B33) 9.Bb2 174 B34) 9.Ne5!? 176 Chapter 10 1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Ba6 5.Qa4 Bb7 6.Bg2 c5 A) 7.d5!? 180 B) 7.0-0 cxd4 8.Nxd4 Bxg2 9.Kxg2 Be7 10.Nc3 0-0 182

855

B1) 11.e4 183 B2) 11.Rd1 185 C) 7.dxc5 bxc5 8.0-0 Be7 9.Nc3 0-0 10.Rd1 Qb6 186 C1) 11.Rb1 188 C2) 11.Qb3 189 C3) 11.Qc2 191 C4) 11.Bf4 d6 192 C41) 12.Qb5 193 C42) 12.Rd2 195 C43) 12.Rab1 Nbd7 13.b4 Rfb8! 196 C431) 14.Be3!? 197 C432) 14.a3 200 Chapter 11 1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Ba6 5.Qb3 c6!? A) 6.Bg2 204 B) 6.Bf4 Be7! 7.Nc3 d5! 8.cxd5 Nxd5 205 B1) 9.Nxd5N 206 B2) 9.Bg2 207 C) 6.Bg5 Be7 7.Nc3 0-0N 8.e4 c5! 209 C1) 9.Rd1 210 C2) 9.d5 212 D) 6.Nc3 Be7! 7.e4 d5 213 D1) 8.exd5 214 D2) 8.cxd5 Bxf1 9.Kxf1 cxd5 10.e5 Ne4!N 11.Kg2 Nc6 0-0 13.h4 Rc8 215 D21) 14.Ne1!? 217 D22) 14.Be3 218 Chapter 12

856

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Ba6 5.Qc2 c5 A) 6.Bg2 Nc6 221 A1) 7.0-0 222 A2) 7.dxc5 222 B) 6.d5 exd5 7.cxd5 Bb7 224 B1) 8.e4 Qe7 225 B11) 9.Bd3 225 B12) 9.Bg2 227 B2) 8.Bg2 Nxd5 229 B21) 9.Qb3 230 B22) 9.0-0 233 Chapter 13 1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Ba6 5.Qc2 c5 6.d5 exd5 7.cxd5 Bb7 8.Bg2 Nxd5 9.0-0 Nc6 10.Rd1 Be7 A) 11.Nc3 237 B) 11.a3 Nc7 12.Nc3 0-0 13.Bf4 Ne6 14.e3 Nxf4 238 B1) 15.exf4 239 B2) 15.gxf4 241 C) 11.Qf5 Nf6 12.e4 g6 13.Qf4 0-0 242 C1) 14.Nc3 243 C2) 14.e5 Nh5 245 C21) 15.Qg4 245 C22) 15.Qc4 247 D) 11.Qa4 Nf6 250 D1) 12.Nc3 252 D2) 12.e4 253 D3) 12.Nh4 0-0 259 D31) 13.Nc3 259 857

D32) 13.Nf5 260 Chapter 14 1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Ba6 5.b3 Bb4† 6.Bd2 Be7 A) 7.Bc3 266 B) 7.Nc3 d5 8.cxd5 exd5 267 B1) 9.Qc2 268 B2) 9.Bh3!? 269 C) 7.Bg2 0-0 8.0-0 d5 271 C1) 9.Bc3 272 C2) 9.Nc3 c6! 273 C21) 10.Rc1 274 C22) 10.Bf4 275 C3) 9.Ne5 Bb7 10.Nc3 Nbd7 276 C31) 11.Bf4 278 C32) 11.cxd5 281 Chapter 15 1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Ba6 5.b3 Bb4† 6.Bd2 Be7 7.Bg2 0-0 8.0-0 d5 9.cxd5 exd5 10.Nc3 Re8! A) 11.Bf4 287 B) 11.a3 288 C) 11.Rb1 290 D) 11.Rc1 292 E) 11.Re1 293 F) 11.Qc2 c5!? 295 F1) 12.dxc5 296 F2) 12.Rfd1 299 G) 11.Ne5 Bb7 300 G1) 12.Bc1 300 858

G2) 12.Bg5 302 G3) 12.Bf4 303 G4) 12.Qc2 304 G5) 12.Rc1 Nbd7 306 G51) 13.Nb5 307 G52) 13.Rc2 309 G53) 13.Bf4 310 Chapter 16 1.d4 Nf6 2.c4 e6 A) 3.a3 315 B) 3.e3 316 C) 3.Bg5 318 D) 3.g3 Bb4† 4.Nd2 c5 321 D1) 5.dxc5 322 D2) 5.a3 Bxd2† 6.Qxd2 cxd4 7.Nf3 d5 326 D21) 8.Bg2 327 D22) 8.cxd5 329 Chapter 17 1.d4 Nf6 2.c4 e6 3.g3 Bb4† 4.Bd2 Be7 5.Bg2 d5 6.Nf3 0-0 7.0-0 c6 A) 8.Na3!? 335 B) 8.b3 336 C) 8.Nc3 338 D) 8.Bf4 343 E) 8.Qb3 b6 9.Nc3 Ba6 10.cxd5 cxd5 11.Ne5 Bb7 347 E1) 12.Rac1 352 E2) 12.Rfc1 354 Chapter 18 859

1.d4 Nf6 2.c4 e6 3.g3 Bb4† 4.Bd2 Be7 5.Bg2 d5 6.Nf3 0-0 7.0-0 c6 8.Qc2 Nbd7 A) 9.Bg5 359 B) 9.Rc1 361 C) 9.a4!? a5! 363 C1) 10.Na3 365 C2) 10.Rc1 366 D) 9.Rd1 b6 10.b3 Ba6 11.a4 c5 369 D1) 12.cxd5 371 D2) 12.a5 373 D3) 12.Na3 375 Chapter 19 1.d4 Nf6 2.c4 e6 3.g3 Bb4† 4.Bd2 Be7 5.Bg2 d5 6.Nf3 0-0 7.0-0 c6 8.Qc2 Nbd7 9.Bf4 b6 A) 10.Nbd2 382 B) 10.cxd5 384 C) 10.Nc3 385 D) 10.Rd1 Ba6 391 D1) 11.Nbd2 391 D2) 11.cxd5 cxd5 12.Ne5 Nxe5! 13.dxe5 Ng4! 392 D21) 14.Bf3 393 D22) 14.h3 394 D3) 11.b3 Rc8 12.Nc3 dxc4 400 D31) 13.Nd2 401 D32) 13.e4 403 D33) 13.bxc4!? 405 D4) 11.Ne5 Rc8! 407 D41) 12.cxd5 408 D42) 12.Nc3 Qe8!? 411 D421) 13.e4 412 860

D422) 13.Rac1N 413

861

Table of Contents Title page Key To Symbols Preface Bibliography Chapter 1 - 4.Bg5 and 4.Bf4 Chapter 2 - 4.e3 Introduction Chapter 3 - 4.e3 and 5.Bd3 Chapter 4 - 4.Nc3 Chapter 5 - 4.Bg5 and 4.Bf4 Chapter 6 - 4.e3 Introduction Chapter 7 - 4.e3 and 5.Bd3 Chapter 8 - 4.Nc3 Chapter 9 - 5.Nbd2 Chapter 10 - 5.Qa4 Chapter 11 - 5.Qb3 Chapter 12 - 5.Qc2 Sidelines Chapter 13 - 5.Qc2 and 10.Rd1 Chapter 14 - 5.b3 Introduction Chapter 15 - 9.cxd5 Chapter 16 - Sidelines Chapter 17 - 4.Bd2 Chapter 18 - 8.Qc2 Chapter 19 - 9.Bf4 Variation Index A) A1) A11) A12) A2) A21) A22) B)

3 5 6 7 8 41 74 110 145 208 230 278 321 362 414 451 484 546 584 644 679 733 776 851 851 851 851 851 851 851 851 851 862

B1) B2) B21) B22) 5.Be2 26 C) C1) C2) D) D1) D2) D21) D22) 6.cxd5 40 6.0-0 Bd6 41 7.b4!? 41 7.Nc3 0-0 43 8.Qe2 43

851 851 851 851 851 851 851 851 851 851 851 851 851 851 851 851 851 851

863