Contents

Title Page
Key to Symbols used & Bibliography 4
Preface 5
 
Third Move Alternatives
1 3.Bc4?! & 3.Nc3 7
2 3.d4 21
3 6.Nxd7 55
3.Nxe5 d6
4 4th Move Alternatives 75
4.Nf3 Nxe4
5 5.Bd3 & 5.c4 102
6 5.d3 123
7 5.Qe2 139
Old Main Line
8 5.d4 154
9 9.Qc2 169
10 9.Nc3 186
Modern Main Line
11 5.Nc3 200
12 11.Kb1 & 11.Rhe1 221
Avoiding the Petroff
13 Centre Game 240
14 King’s Gambit 264
15 Vienna Game 287
16 Bishop’s Opening 301
 
Appendix: Konguvel Ponnuswamy – Swapnil Dhopade 317
Variation Index 320

1
 Playing the Petroff
By
Swapnil Dhopade

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

First edition 2020 by Quality Chess UK Ltd

Copyright © 2020 Swapnil Dhopade

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 978-1-78483-105-9

Hardcover ISBN 978-1-78483-106-6

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: info@qualitychess.co.uk
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

Edited by Andrew Greet
Proofreading by Colin McNab and John Shaw

2
Cover design by Kallia Kleisarchaki and Adamsondesign.com
Front cover photo by Kallia Kleisarchaki
Author photo on back cover by Eesha Karavade

3
 Key to symbols used
² White is slightly better
³ Black is slightly better
± White is better
µ Black is better
+– White has a decisive advantage
–+ Black has a decisive advantage
= equality
© with compensation
„ with counterplay
ƒ 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

Bibliography
I mainly relied on the following sources:
 
Cohen: A Vigorous Chess Opening Repertoire for Black, New in Chess 2013
Khalifman: Opening for White According to Anand 1.e4 (Book 1), Chess Stars 2003
Ntirlis: Playing 1.e4 e5 – A Classical Repertoire, Quality Chess 2016
Obodchuk: The Four Knights Game, New in Chess 2011
Sakaev: The Petroff: an Expert Repertoire for Black, Chess Stars 2011
Shaw: Playing 1.e4 – Caro-Kann, 1...e5 and Minor Lines, Quality Chess 2016
 
Electronic Resources
Mega Database
Correspondence Database
The Week in Chess

4
 Preface

Welcome, dear readers, to my first opening book – on the Petroff Defence. The idea for this book
first occurred during the 2018 Olympiad in Batumi, Georgia, an event which marked a turning point
in my career. Having become a grandmaster in 2016, I had been coaching talented youngsters
(including GM Raunak Sadhwani, who became a GM at just 13 years of age!) for quite a while and
preparing opening files for them. In 2018, I was given the opportunity to work as part of a team of
seconds, headed by GM Jacob Aagaard, helping the Indian Women’s team at the Olympiad.
Performing this work at a world-class competition required me not only to prepare opening ideas at a
more intricate level, but also to supplement my ChessBase analysis with brief but lucid comments at
critical moments, to help the players retain the most important information.
 
Sometime towards the end of the event, encouraged by the positive reception my opening preparation
had received from the players, I proposed to Jacob the idea that I might write a book for Quality
Chess on the Caro-Kann, my lifelong favourite response to 1.e4. Jacob replied that it would be great
if I could join the Quality Chess team and write an opening book... but a future Caro-Kann book had
already been planned with another author. To my great surprise, Jacob then suggested that I write a
book on the Petroff Defence! My instant reaction was of disbelief: how could I write a book on an
opening that I never played in my entire chess career? Jacob replied, “You are a researcher. If you
research this opening well, I am sure you can present a top-class repertoire for Black in the Petroff
Defence.”
 
These words, coupled with the success of my opening preparation in the Olympiad, gave me
confidence that I could face this new challenge. Jacob also mentioned that it had been many years
since an in-depth book had been published on the Petroff; moreover, Quality Chess had never
published such a book, so it was a great opportunity to fill this gap for readers. All this gave me
added motivation to write the best Petroff book I possibly could. In addition to the Petroff itself, we
also decided to include a slightly less detailed section on 1.e4 e5 lines where White avoids 2.Nf3, to
give the reader all the essential information needed to meet 1.e4 with 1...e5.
 

The Writing Process

I started, as usual, from scratch! The much-needed inspiration was provided by Fabiano Caruana,
who employed the Petroff Defence in his World Championship match against Carlsen. The world
number two was already a renowned Petroff expert, so there was no doubt that Carlsen and his team
of seconds would make the Petroff a focal point of their preparation. In such a situation, many
players would adopt a new opening for the match in order to surprise the opponent and avoid
preparation, but Caruana stayed true to the Petroff, believing in his home analysis and the inherent

5
soundness of the opening itself. Carlsen only played 1.e4 in two of the twelve classical games of the
match, and both times Caruana drew quite comfortably with the Petroff. I was especially attracted by
Game 11, where Caruana embraced opposite-sides castling and was clearly ready for a fight. I
analysed the whole system in great detail (you can find it in Chapters 11 and 12) and was amazed at
the dynamic potential this opening has to offer. In the game, Carlsen avoided Black’s attacking ideas
and went for an early queen exchange, which enabled the challenger to make a relatively easy draw.
In a way, this change in the character of the play reflects the instructive value of the Petroff: aside
from being a top-tier opening, it enables you to switch from sharp tactics and brutal attacks in one
game to dry positions and equal endgames in the next.
 

Repertoire Choices
Even though the Petroff has a solid reputation and this book is expected to give a bulletproof
repertoire, I have not shied away from recommending complicated lines when the position demands
such an approach – there are many such instances where Black has excellent chances to gain the
initiative or an outright advantage.
 
Compared to a lot of opening books, the volume and complexity of analysis is not too daunting. With
that being said, there are certain critical lines where the play becomes rather sharp, perhaps involving
sacrificial shenanigans leading to an eventual perpetual check. Such variations are not too frequent;
and when they had to be included, I have added as many comments as possible along the slippery
paths, which should ease the process of memorizing the most important points. I assure you that if
you study this book well, you will have every reason to be confident of getting a comfortable
position out of the opening against 1.e4.
 
I have analysed the recommended repertoire using the strongest cloud engines possible, but they
were as much of a hurdle as a boon for me, as “0.00” seldom tells the full story. There were several
lines which I analysed through to a 0.00 evaluation, but which I rejected as unsuitable due to my own
feeling that the positions would be slightly uncomfortable for Black in a practical game. Retaining
one’s sense of how pleasant a position would be in a practical game, in spite of what the engine may
say, is a major challenge for all authors and players nowadays. At all times, I have endeavoured only
to recommend lines that are relatively easy to understand and play in a practical game.
 
Having done so much work preparing this repertoire, I could not resist giving it a try in a tournament
game, and you can find my win over IM Konguvel in the appendix on page 317. The process of
researching and writing this book lasted from October 2018 (when the Olympiad ended) up until the
time of publication, with a few late updates even taking place during the editing/proofreading stage. I
am sure you will enjoy the benefits of my painstaking research, and face 1.e4 with a smile and
confidence in your future games!
 
GM Swapnil Dhopade
Amravati, India

6
April 2020

7
A) 3.Bc4?! 8
B) 3.Nc3 Bb4!? 4.Nxe5 0-0 5.Be2 Re8 6.Nd3 Bxc3 7.dxc3 Nxe4 8.0-0 d6 10
B1) 9.f3 11
B2) 9.Re1 Nd7 13
B21) 10.Bf1 13
B22) 10.Bf3 14
B3) 9.Nf4 Nd7 16
B31) 10.f3 17
B32) 10.c4 18

 
1.e4 e5 2.Nf3 Nf6
We begin our study of the Petroff by considering two sidelines: the dodgy
A) 3.Bc4?! and the more sensible B) 3.Nc3.
 
3.d4 will be analysed in Chapters 2 and 3, and the main line of 3.Nxe5 will follow in Chapters 4-12.
 
3.d3 has been played in thousands of games but after 3...Nc6 Black has relatively easy play, most
likely in a reversed Philidor after ...d5, although a more cautious set-up with ...Bc5 and ...d6 could

8
also be considered, depending on how White plays.
 
3.Qe2, 3.Bd3 and 3.Bb5 all appear on the database, but you shouldn’t need an opening book to play
against such moves.
 

A) 3.Bc4?!
 
This move is generally connected with a dubious gambit.
 
3...Nxe4!
3...Nc6?! has been a frequent choice, but why transpose to a Two Knights Defence when you can
simply capture a central pawn and fight for the advantage?
 
4.Nc3
This is White’s usual scheme, aiming for fast development.
 
Black has an easy time after:
4.0-0?! d5 5.Qe2
5.Re1? is crushed by: 5...Nxf2!–+
5.Bb3 Nc6 6.Re1 Bc5!–+ also gives Black a winning advantage.
5...dxc4 6.Qxe4 Nc6! 7.Nxe5N
Other moves also leave White clearly worse. The text seems most natural but Black stays on
top with the precise continuation:

7...Qd4!
Forcing a queen exchange.

9
 7...Qe7! 8.Re1 Nxe5 leads to the same thing.
8.Re1 Qxe4 9.Rxe4 Nxe5 10.Rxe5† Be6µ
White’s queenside is undeveloped and Black has the bishop pair.
 
4.d3
With this move, White regains the pawn while accepting a positional disadvantage.
4...Nc5
Another good option is to trade off the bishop: 4...Nd6 5.Nxe5 Nxc4 6.dxc4 d6 7.Nd3 Be7 8.0-
0 0-0 9.h3 Nc6 and Black was comfortable in Ozolins – Yusupov, Puhajarve 2011.
5.Nxe5
5.Bg5?! Be7 6.Bxe7 Qxe7 7.0-0 d6 8.d4 exd4 9.Re1 Ne6 10.Nxd4 Nc6! 11.Nf3 0-0µ left
White with no real compensation for the pawn in Williamson – Hall, corr. 2014.
5...d5 6.Bb3 Bd6 7.d4 Nxb3 8.axb3 0-0 9.0-0

9...Bxe5 10.dxe5 d4!³

Black’s bishop is superior to its counterpart, which gave him a positional advantage in Keuter –
Noble, corr. 2017.
 
4...Nxc3
4...Nc6?! lets White off the hook by transposing to a well-known variation of the Two Knights.
After 5.Nxe4 (but not 5.Bxf7†? Kxf7 6.Nxe4 d5 7.Neg5† Kg8 with a horrible position for White)
5...d5 6.Bd3 dxe4 7.Bxe4 Bd6 8.0-0 0-0 the position is about equal.
 
5.dxc3
Now Black can afford to be greedy with:
 

10
5...f6!
This not only protects the e5-pawn but also prevents any Ng5 tricks.
5...d6?? and 5...Nc6?? both allow 6.Ng5+– with a disaster on f7.
 
If for any reason you prefer not to keep the extra pawn then 5...c6 6.Nxe5 d5 shuts out the bishop,
with easy equality. For instance, 7.Bd3 Bd6 8.0-0 0-0 9.Re1 Qc7= and Black had a fine position in
Samsonkin – Plotkin, Sault Ste. Marie 2017.

6.0-0 Nc6
6...c6?? 7.Nxe5 fxe5 8.Qh5†+– would be a disaster!
 
7.Nh4
This is the most direct attacking try.
 
A slow set-up such as 7.Qe2 g6 8.b4 Bg7 9.Bb3 Qe7µ leaves Black in no danger and he can slowly
regroup his pieces. For example, he may consider transferring his knight via d8 to f7, followed by
...0-0.
 
7.Re1 does not achieve much. 7...d6 8.a4 was tried in a correspondence game, but 8...Ne7 9.Bb3 a5!?
stops White’s queenside expansion (not that it was all that scary anyway). Having given up a pawn,
White’s problem is that he has no convenient way of opening lines towards the enemy king; instead
he just has to sit and wait while Black regroups his pieces slowly. 10.Be3 Qd7 11.Nd2 Qf5 12.f3
Bd7 13.Bc4 h5 14.Qe2 Qg6 15.Bb5 Nc6µ White had no real compensation for the pawn in Pessoa –
Badolati, corr. 2009.
 
7...g6 8.f4
White tries to open lines but Black has more than one good way of dealing with the attack.
11
8...f5!
8...Qe7 has also proved to be a good option but I prefer the text move: Black hits the knight and
prepares to close the centre with ...e4.
 
9.Nf3
9.Nxf5? is met by 9...d5! 10.Bxd5 Bxf5 11.fxe5 Bc5† 12.Kh1 Ne7–+ forcing a queen exchange,
after which Black is easily winning.
 
9...e4 10.Ng5
10.Qd5? Qe7 11.Ng5 h6–+ is even worse for White.
 
10...Bc5† 11.Kh1 Qf6

12
12.Qd5!
This is White’s best chance to make a game of it, although he is still clearly worse.
 
12.Bf7† Ke7 13.Bd5 d6 14.Re1 created a threat of Bxe4 in Schlechter – Marco, Berlin 1897, but the
attack could have been convincingly refuted by:

14...h6!N White must either retreat his knight, when he is simply a pawn down for nothing, or go
down with guns blazing after 15.Bxe4 hxg5 16.Bxc6† Be6, when the terrible threat of ...Rxh2†
forces White to give up material.
 
12...d6N 13.Nxh7! Qe7 14.Ng5 Nd8µ

13
 White’s accurate play has enabled him to equalize the material balance. However, his queen and
bishop will soon be pushed back by ...c6 and ...d5, and Black’s superior pawn structure will give him
a clear advantage.
 
 

B) 3.Nc3 Bb4!?
 
Black gets ready to castle quickly while temporarily sacrificing the e5-pawn. He can easily regain it
in the near future by means of ...Bxc3 followed by ...Nxe4. The text move is a characteristic choice
for Petroff players, as Black continues to delay ...Nc6.
 
Having said that, transposing to the Four Knights Game with 3...Nc6 is a completely sound and
reliable way of playing for Black. If this already forms part of your repertoire and you are happy with
the resulting positions, then you may wish to skip the rest of this chapter.
 
4.Nxe5 0-0 5.Be2
5.Nd3 Bxc3 6.dxc3 Nxe4 will surely transpose.
 
Trying to hang on to the extra pawn, for instance with 5.d3?!, only leads to trouble for White after
5...d5.
 
It is worth briefly mentioning that 5.Bc4 prevents Black’s intended ...Re8, but 5...Qe7 does the same
job perfectly adequately. After the knight retreats from e5, ...Bxc3 and ...Qxe4† will lead to an equal
game.
 
5...Re8 6.Nd3
6.Nf3 has been tried by a few strong players but 6...Bxc3 7.dxc3 Nxe4 8.0-0 d6= leads to
essentially the same type of position as the main line, but White’s options are limited by the position
of the knight on f3 instead of d3.

14
6...Bxc3
Black has to part with his bishop pair in order to regain the pawn. Even though White possesses the
two bishops, their scope is limited by the slightly damaged state of White’s queenside pawn structure.
 
7.dxc3
7.bxc3?! just makes it harder for White develop his dark-squared bishop.
 
7...Nxe4 8.0-0 d6
8...d5 is the other main line and Sakaev prefers it in The Petroff: an Expert Repertoire for Black.
However, I do not share his concerns after the text move, and it is positionally more desirable to post
this pawn on a dark square to complement Black’s remaining bishop.
 
In the near future, Black will almost certainly develop the b8-knight to d7, from where it can go to
f6, f8 or c5 depending on circumstances. Black could bring his bishop to f5 at some point, but it may
be better placed on b7 in some variations.
After several natural moves, we have reached the point where White must take a serious decision. He
has played B1) 9.f3 in some games, but his two most popular ways to continue are B2) 9.Re1,
followed by Bf1 or Bf3, and B3) 9.Nf4.

B1) 9.f3

15
 White kicks the knight from the centre in order to pin it on the next move. Sakaev awards this
move an exclamation mark when explaining why he prefers 8...d5 over 8...d6.
 
9...Nf6 10.Bg5
White hopes to create some discomfort in Black’s camp.
 
10...h6
Sakaev gives some examples after 10...Nbd7 followed by ...b6. I think it is slightly more accurate
to delay ...b6, unless there is a special reason to play it, such as in the 12.c4 line in the notes below.
 
11.Bh4 Nbd7
The knight should almost always be developed to d7 in these lines, as it has a variety of future
prospects from this square. Black’s only real worry in this position is the pin on the f6-knight, but in
the lines below we will see exactly how he should deal with it.
 
12.Nf2
The knight intends to hop to e4 or g4, putting pressure on f6.
 
12.c4 has only been played in one game, Ostermeyer – Schulte, Germany 1989, but it is a typical
move which White often plays at some point in these positions. White gains a bit of space on the
queenside, and also prepares c4-c5 in some lines. Hence, it should be usually met by:

16
12...b6!N Aside from preventing c4-c5, this is a desirable move anyway for Black, who should
generally arrange his queenside pawns on dark squares. 13.Re1 Bb7 14.a4 a5= Having set up his
ideal pawn structure on the queenside, Black is now ready to play ...Nf8 followed by ...Ng6,
breaking the pin.
 
Another possibility is:
12.Re1 Nf8
The knight is once again heading to g6 to break the pin.
13.Qd2!?
A game continued: 13.Bf1 Rxe1 14.Qxe1 Bf5 (but not 14...Ng6?! immediately, since 15.Bxf6
Qxf6 16.Qe8†± is annoying for Black) 15.c4 Ng6 16.Bf2 b6= Again we see this typical
reaction to the pawn on c4. White’s bishop pair was balanced out by Black’s better pawn
structure on the queenside in Dempster – Bronts, corr. 2010.
13...Ng6 14.Bf2 Bf5N 15.Bf1
After developing the pieces well, it is time for Black to arrange his queenside pawns in an ideal
formation with:

17
15...b6!
This also limits the activity of White’s dark-squared bishop. My illustrative line continues:
16.h4!?
Trying to be aggressive on the kingside.
16...Qd7 17.g4 Bxd3 18.cxd3 h5!
It is important to prevent h4-h5.
19.g5 Nd5„
White’s kingside play has come to a halt. Even though Black has traded off both his bishops for
enemy knights, his knights will have chances to exploit the weaknesses created by White’s recent
pawn advances.

18
12...Nc5!?N
Black has tried a few other reasonable moves, one example being 12...Qe7 as played in Nisipeanu
– Fridman, Eforie Nord 2009, but the text move is my preference.
12...g5?! 13.Bg3 Nh5 would be a good idea to exchange White’s strong dark-squared bishop, if it
were not for 14.f4! Nxg3 15.hxg3ƒ when White will gain a strong initiative on the kingside.
 
13.Re1
13.Ng4 is comfortably met by 13...Bxg4 14.fxg4 Nce4 followed by ...Qd7 and Black is fine.
 
Now that White’s rook has moved away from the f-file, Black can safely play:
 
13...g5 14.Bg3 Nh5=
With the dark-squared bishop about to be exchanged, Black is comfortable.

B2) 9.Re1

9...Nd7
As usual, Black develops his knight to its best square. White has tried several options but, since his
last move implied that the e-file will figure in his plans, we will concentrate on the two most popular
options of B21) 10.Bf1 and B22) 10.Bf3.

B21) 10.Bf1
 

19
This is the less challenging option. White makes no attempt to seize the initiative and prefers to play
slowly.
 
10...b6
Developing the bishop to b7 is a simple and effective way for Black to play.
 
11.f3
Another game continued:
11.Be3 Bb7 12.f3 Ng5
12...Nef6 is fine as well.
13.Qd2 h6 14.Qf2
My recommended improvement is:

14...Qf6!?N
Transferring the queen to g6 while also connecting the rooks.
14...Ne5 was played in Zeng Chongsheng – Bai Jinshi, Xinghua 2016, when White could have
obtained a slight advantage with: 15.Nxe5!N 15...Rxe5 (or 15...dxe5 16.a4! Qf6 [16...a5
17.Bb5 is annoying for Black] 17.a5 Ne6 18.b4² with pressure on the queenside) 16.Qg3 Re6
17.Bd3 Qe7 18.Bf2 Re8 19.h4 Rxe1† 20.Rxe1 Ne6² With the queens still on the board, Black
is under slight pressure.
15.Bd4 Qg6 16.Qg3
16.Nf4?! is well met by 16...Qf5 17.Qg3 Ne5 when White has to be careful not to become
worse.
16...Ne6!=
Black should be fine after a queen exchange.
11...Nec5

20
 Black prepares to transfer his knight to e6, where it will be well placed.
 
12.Be3 Ne6 13.Nb4 Bb7 14.a4 a5 15.Nd5
Now it is important to challenge White’s powerful knight immediately with:

15...Nf6!N
15...Rc8?! led to a passive position for Black after: 16.Qd2 Qh4 17.c4 Ne5 18.Bf2 Qh5 19.Ra3!±
A nice transfer of the rook! White was clearly on top in Jensen – Groffen, corr. 2018.
 
16.c4 Nc5
Before exchanging on d5, Black removes his knight from the would-be attack.
 
17.Qd2
17.Bg5 is of no concern due to 17...Rxe1 18.Qxe1 Bxd5! 19.cxd5 h6 when White is forced to take
on f6, otherwise he will lose the d5-pawn. 20.Bxf6 Qxf6 The position is equal, but in a practical
game Black would have some chances to press for more.
 
17...Nxd5 18.cxd5 Re5!
Black has a good game, as shown by the following brief lines.

21
19.Bc4
19.Bxc5 Rxe1 20.Rxe1 dxc5=
 
19...Ba6!
It may seem counterintuitive to exchange the ‘bad’ bishop on c4, but Black’s bishop was equally
ineffective on b7.
 
20.Bxa6 Nxa6=
Black has no problems.

B22) 10.Bf3

22
10...Ndf6
This is the correct reply.
 
10...Nec5?!
This move was played in Mamedov – Bu Xiangzhi, Riyadh (rapid) 2017, but White could have
seized the initiative with the precise reply:
11.Nxc5!N 11...Nxc5 12.Bg5!
Provoking a concession of some sort.
12...f6
12...Qd7 clumsily blocks the c8-bishop. After 13.Be3!± White’s bishops are well placed and it
will take time for Black to bring his queenside pieces into the game.
13.Be3 c6
Covering the d5-square and the b7-pawn. Now White can provoke further weaknesses with:
14.Bh5! g6 15.Be2 Bf5 16.c4±
Black is under unpleasant pressure, due to his weak pawn structure and White’s powerful bishops.
 
11.c4
This has been White’s usual choice. I checked two other ideas:
 
11.Be3
Here my preference would be:

23
11...h6!N
Black frees the h7-square for his bishop in case it is needed.
11...Bf5?! was played in Slobodjan – Klein, Germany 1992, when 12.g4!N 12...Bg6 13.Nf4² is
the line which inspired the novelty above.
However, 11...d5!? seems like another fully playable approach, as seen in Liwak – Murdzia,
Lubniewice 2002.
12.c4 Bf5=
With a firm grip on the e4-square, Black is doing fine.
 
11.Bxe4
White opted for this simplifying move in Liu Chang – Zhao Yuanhe, Wuxi 2016. I suggest
meeting it with:
11...Rxe4N 12.Bg5 Rxe1† 13.Qxe1

24
13...Bd7!
The active 13...Bf5 is less accurate due to 14.Nf4! and Black has to waste time dealing with the
threat of Nd5.
Now Black is ready to play ...h6 with a comfortable position. For example:
14.f3 h6 15.Bh4 g5 16.Bf2 Qf8=
Black is fine.
 
11...h6!
An important move! Just as in the 11.Be3 line in the note above, Black frees the
h7-square before playing ...Bf5.
 
12.b3!
White’s best idea is to develop the bishop to b2, from where it will exert pressure on Black’s
kingside, especially the g7-point.
 
12...Bf5 13.Bb2 c6!
This move not only limits the scope of the f3-bishop but also prepares for a ...d5 break before
White can develop significant threats on the kingside. The critical continuation is:
 
14.h4!N
14.Qc1 is less challenging. 14...Ng5 15.Rxe8† Nxe8 16.Qf4 Bxd3 17.cxd3 Nxf3† 18.Qxf3 Qg5
Black has equalized comfortably and the players soon agreed a draw:

25
19.Qe3 Nc7 20.Bc3 ½–½ Solak – Jakovljevic, Herceg Novi 2008.
 
The text move is a better try for White, as ...Ng5 is now impossible. Black should instead break open
the centre with:

14...d5!
There is no time to lose!
14...Qd7?! is too slow and 15.Nf4!± gives White great play on the kingside, supported by the strong
bishop on b2.
 
15.cxd5 Nxd5 16.c4 Ndf6„

26
 The position is roughly balanced, as Black’s stronghold on the e4-square compensates for White’s
bishop pair.
 

B3) 9.Nf4

9...Nd7
As usual, we develop our knight to its best square. For some reason, in The Four Knights Game,
Obodchuk relegates this move to a brief note in order to focus on a main illustrative game in which
the completely unnecessary 9...Nf6?! was played.
 
White’s most popular options are B31) 10.f3 and B32) 10.c4.
 
Obodchuk gives 10.Re1 followed by a game reference where Black played weakly and soon fell into
a losing position. His position should be sound enough though; out of a few reasonable moves,
10...b6 seems a good choice which is consistent with some of the other lines where we develop the
bishop to b7 early in the game. 11.c4 Bb7 12.f3 occurred in Gu Tianlu – Lei Tingjie, China 2016, and
now 12...Nec5N would have given Black a perfectly decent position.
 
Another idea is:
10.Be3
Black has tried various knight moves from here, but for some reason no one has opted for the
straightforward plan of developing the bishop with:
10...b6N
Just as in variation B32 below, this fianchetto works well.
11.c4
The aggressive 11.Qd5?! achieves nothing after 11...Rb8 12.Bc4 Ng5, followed by moves like
...Ne5 and ...Bb7, when White’s pieces will be pushed back.
27
11...Bb7 12.Re1 Qf6!
The queen is well placed here.

13.Qd4
Or 13.Qc1 Ne5 14.a4 a5 15.f3 Nc5= and Black is fine.
13...Qxd4 14.Bxd4 Nef6=
Preventing Nd5. Black has a good position and is ready to double his rooks on the e-file next.

B31) 10.f3
 
White pushes the knight away before embarking on a further plan.
 
10...Nef6 11.c4 Qe7
Putting pressure on the e2-bishop and threatening ...g5.
 
12.Rf2
12.Re1N 12...Qe5! activates the queen to good effect. From its central post it hits the
b2-pawn, thus preventing the development of the c1-bishop. Therefore 13.Nd3 is logical, when play
continues: 13...Qd4† 14.Nf2

28
14...Qc5! Black’s queen is more active than its counterpart, so he avoids the exchange on this
occasion. 15.b3 b6 16.Bb2 Bb7 17.Bd4 Qf5= Black is absolutely fine.
 
12...Nc5
12...Qe5 is a reasonable option here too.

13.b3 Qe5
Activating the queen while preventing Bb2 for the moment.
 
14.Rb1 Bd7
We are following a high-level game in which Black decided to post his bishop on c6.

29
 14...b6 looks like a good alternative: the bishop goes to b7 next, with essentially the same type of
position, but with Black’s bishop on a slightly safer square and with his queenside pawns in the most
desirable formation.
 
15.Bf1 Bc6 16.Qd2
White supports the knight on f4, thus preparing Bb2. The time has come for us to improve Black’s
play:

16...Ne6N
The tempting 16...Nfe4? 17.fxe4 Nxe4 can be refuted by: 18.Qe1!N (rather than 18.Qe3? Nc3!
when Black went on to win in Artemiev – Firouzja, St Petersburg 2018) 18...Nxf2 19.Qxf2 Qe1
20.Qxe1 Rxe1 21.Kf2±
 
17.Nd3
17.Nxe6 Rxe6 18.Bb2 Qh5 is about equal, with Black intending ...Rae8 next.
 
17...Qh5 18.Bb2
18.Nb4 Bd7 19.Bb2 a5 20.Nd3 Bc6 transposes.
 
18...a5!
Covering the b4-square and arranging the pawns in an ideal formation.
 
19.Re1 b6 20.Rfe2 Qg6=
Black is doing okay, a natural plan being ...Nf8 followed by exchanges along the e-file.

 
30
 B32) 10.c4

Here Black has tried all sorts of d7-knight jumps, but I think the simplest approach is to develop
the bishop to b7 and avoid committing the knight just yet.
 
10...b6 11.Bf3
11.Be3N transposes to 10.Be3 b6N 11.c4 in the note on the previous page.
 
11.a4!?N 11...a5 12.Ra3! is an interesting attempt to bring the queen’s rook into the game.
Nevertheless, Black is fine after 12...Bb7 13.Rh3 Ndc5 14.b3 Qf6!= when, with Bb2 prevented, we
are completely safe on the kingside.
 
Another good example continued:
11.f3 Nef6 12.Rf2
12.b3 was played in Hong Xing – Xiang Zeyu, Shijiazhuang 2016. I sugges12...Qe7!?N,
preventing Bb2 for the moment due to the fork on e3, winning the knight on f4. 13.Rf2 (after
13.Qd4 Qe5= Black is happy with the queen exchange) 13...Qe5 14.Rb1 Bb7= The position is
similar to variation B31 above and Black is doing fine.

31
12...Bb7 13.Bf1 Ne5
Black wants to transfer the knight to g6, offering the trade of knights.
14.Nd5
14.b3N 14...Ng6 15.Nxg6 hxg6 16.Bb2 is met by 16...Nh5!= with the idea of ...Qh4, when
Black is fine.
14...Nxd5 15.cxd5 Qf6
Black posts his queen on a good square and is ready to double his rooks on the e-file.

16.f4?!
This weakens the e4-square for no good reason.
16.Be3N is better, when play may continue 16...Re7 17.Qd2 Qg6 18.a4 a5= followed by ...Ba6
to exchange the light-squared bishops. We saw the same idea earlier in the chapter in variation
32
 B21.
16...Nd7 17.c4
In Fomin – Bao Qilin, St Petersburg 2016, Black could have obtained some initiative with:
17...Nc5N
Targeting the weak e4-square.
 
11...Bb7 12.Nd5
12.Re1N 12...f5! cements the active knight on e4. 13.Nd5 is a logical move to cut off its support.
My analysis continues 13...c6 14.Bxe4 Rxe4 15.Rxe4 fxe4 16.Nc3 and now instead of defending the
e4-pawn, Black can counterattack the c4-pawn with:

16...Ne5! 17.Nxe4 (17.b3?! is met by the strong 17...Qf6! when Black may seize the initiative)
17...Nxc4 The position is balanced and the opposite-coloured bishops give the game a somewhat
drawish character.
 
Here I think Black should immediately kick the knight away from d5 with:

33
12...c6!N
Even though this weakens the d6-pawn, Black can use the time gained to prepare for the ...d6-d5
break at the right moment.
 
12...Rb8?! was played in Nitin – Tran, Penang 2015, but it looks like a waste of time to me.
 
13.Nf4 Qc7 14.b3
In the event of 14.Re1 Nef6 White has forced the active knight to retreat; on the other hand, Black
gets the option of ...Ne5, which was not possible earlier as the e4-knight would have been left
hanging. For instance: 15.Be3 Ne5 16.Be2 Rad8 All of Black’s pieces have been mobilized and the
position is equal.
 
14...d5!
Initiating a tactical sequence.
 
15.cxd5 cxd5 16.Nxd5!
If White refuses to capture this pawn, Black will play ...Rac8 and target the weakness on c2.
 
16...Qe5 17.Bxe4 Qxa1
Black could also consider 17...Qxe4!? 18.c4 Ne5© with interesting compensation.
 
18.Nxb6!
18.Nc7 is less accurate: 18...Rxe4 19.Nxa8 (19.Qxd7? Rf8–+ simply leaves White an exchange
down) 19...Nf8 20.Nc7 (20.Qd8?! Qxa2 21.Nc7 h6³ and White has to play precisely to avoid

34
becoming significantly worse) 20...Rd4 21.Qe2 Qxa2 The position is close to equal but Black has a
slight pull.
 
18...Nxb6!?
This would be the ambitious choice.
 
After 18...Bxe4 19.Nxa8 Rxa8 20.Qxd7 Qxa2= Black will pick up another pawn on the queenside,
leading to a drawish position with level material and opposite-coloured bishops.
 
19.Bxb7 Rad8÷
White will lose one of the queenside pawns sooner or later, but he will have some compensation
due to his bishop pair.
 

Conclusion
 
This chapter has dealt with a couple of sidelines that can occur after the opening moves 1.e4 e5
2.Nf3 Nf6. Black should be happy to see the first of them: 3.Bc4?! is well met by the simple
3...Nxe4!, when White faces an unenviable choice: if he wants to play with equal material he will
have to allow a quick ...d5, when Black gets an easy game by targeting the bishop; therefore most
players continue in gambit style with 4.Nc3, but then 4...Nxc3 5.dxc3 f6! gives Black a clear
advantage in all variations.
 
3.Nc3 is a more serious option, when 3...Nc6 would lead to the tried-and-tested Four Knights Game.
Instead, 3...Bb4!? is a decent choice for Petroff purists. The usual continuation is 4.Nxe5 0-0 5.Be2
Re8 6.Nd3 Bxc3 7.dxc3 Nxe4 8.0-0 d6, leading to a complex middlegame where White’s bishop pair
is definitely worth something, but Black has a slightly better pawn structure and a variety of good
squares for his pieces. The b8-knight invariably goes to d7, after which it can choose between many
possible paths, depending on how White plays.
The c8-bishop is often well placed on b7, although sometimes it goes to f5. In general, Black does
well to play solidly and patiently, but we saw some interesting lines where a more dynamic approach
with ...c6 and ...d5 was called for; the reader would do well to study those lines especially closely.

35
A) 4.dxe5 d5 5.Nbd2 Nxd2 22
A1) 6.Bxd2 Be7 23
A11) 7.c4 24
A12) 7.Bd3 26
A2) 6.Qxd2 27
B) 4.Bd3 d5 31
B1) 5.dxe5 Nc5 31
B11) 6.0-0 Be7 7.Nc3 c6 31
B111) 8.Nd4 32
B112) 8.Re1 33
B12) 6.Be2 Be7 7.0-0 0-0 35
B121) 8.Nc3 35
B122) 8.Be3 37
B2) 5.Nxe5 Nd7 40
B21) 6.Nc3 41
B22) 6.Qe2 43

36
 B23) 6.0-0 Nxe5 7.dxe5 Nc5 46
B231) 8.Nc3 47
B232) 8.Be2 50
B233) 8.Be3 51

 
1.e4 e5 2.Nf3 Nf6 3.d4
This move has been played less frequently than 3.Nxe5 (and indeed 3.Nc3) but it is still a major
option which has been employed by numerous world champions and other top players.
 
3...Nxe4
White has two serious options: A) 4.dxe5 and B) 4.Bd3. The latter is more popular, and our
coverage of it will take up the next chapter as well.
 
4.Nxe5 gives Black a choice: 4...d5 5.Bd3 transposes to variation B2 of this chapter, and 4...d6 5.Nf3
(5.Nxf7?? is refuted by 5...Qe7!–+) 5...d5 transposes to Chapter 6.
 
A) 4.dxe5

4...d5
4...Bc5?! 5.Bc4! is dangerous. Even if Black can draw with precise defence, there is no point in
taking such a risky road when we can instead get a sound position by simple means.
 
5.Nbd2

37
 5.Bd3 transposes to variation B1.
 
5.exd6 is completely harmless as it just allows Black to develop his bishop with tempo. 5...Bxd6
6.Bd3 Nc5 7.Be2 0-0 8.0-0 was seen in Mrkonjic – A. Saric, Sibenik 2005, and here I like:

8...Bf5N Black brings his bishop to an active square and is ready to bring the remaining pieces
quickly into the game. A sample line could be 9.Bg5 Qc8 10.Nc3 Re8 11.Nb5 Bf8 12.Nbd4 Be4=
and Black is doing well.
 
I also considered:
5.Be3 Bc5
5...Be7 was also fine in Jobava – Grischuk, Khanty-Mansiysk 2009, but trading bishops works
well.
6.Bxc5 Nxc5

38
7.Nc3
White’s best approach may be to play 7.c4!?N, just aiming to simplify the position. Play may
continue 7...Be6 8.cxd5 Qxd5 9.Nc3 Qxd1† 10.Rxd1 Bg4 11.Be2 Nc6 (or 11...Nbd7) when
Black is fine and White should take care of his e5-pawn.
7...c6 8.Qd2 0-0 9.0-0-0 Re8
Black already stands better: the e5-pawn is a target, and Black’s queenside attack will soon
gather pace.
10.Re1 Nbd7 11.h4 b5!
Gaining space and intending to drive White’s knight to an even worse square.
11...f6?! was the wrong idea in Bartel – Galkin, Khanty-Mansiysk 2007, even though Black’s
position was still objectively at least equal at this stage. The e5-pawn is a weakness and two of
White’s pieces are tied up defending it, so Black should not go out of his way to exchange it.
12.Bd3
This position was reached in Overmark – Zesiger, corr. 2016. Black has a few tempting
continuations but my favourite is:

39
12...b4!N 13.Nd1
13.Ne2 Nxd3† 14.Qxd3 Nxe5 15.Nxe5 Rxe5µ simply wins a pawn.
13...Nxd3† 14.cxd3 d4!µ
Preventing White from playing d3-d4. The knight on d1 is terribly passive, and if White grabs the
b4-pawn then ...c5 gives Black the makings of a tremendous initiative on the queenside.
 
5...Nxd2
In A Vigorous Chess Opening Repertoire for Black Cohen prefers 5...Nc5, finding it “less
cooperative”, but I believe the text move to be a good choice. Sakaev covers both options, displaying
no obvious preference.
 
Both A1) 6.Bxd2 and A2) 6.Qxd2 deserve attention. The latter move has been more popular,
although Sakaev does not mention it.
 

A1) 6.Bxd2 Be7

 
Now we have another split, with A11) 7.c4 and A12) 7.Bd3 both having been tested at a high level.
 
7.Be2?! just gives White a passive version of variation A12. 7...0-0 8.0-0 c5 9.Bf4 Nc6 10.Qd2 has
been played a few times, and here I think it’s best to develop actively with:

40
10...Bf5! Taking full advantage of the absence of a bishop on d3. (10...Be6 is a normal developing
move, and after 11.c3 Qd7 12.h3 Rad8 13.Rfd1 h6= Black was fine in Nepomniachtchi –
Mamedyarov, Moscow 2018, but I think we should be more ambitious.) 11.Rad1 Qb6!³ Black was
slightly better due to her active piece placement and control over the centre in Fatalibekova –
Ruchieva, Elista 1998. Assuming White deals with the threat to the b2-pawn, a good plan for Black
over the next few moves will be to centralize the rooks with ...Rad8 and ...Rfe8.

A11) 7.c4

41
 This has been played by Nepomniachtchi and a few other strong GMs, although most of the games
took place several years ago. White wants to clarify the situation in the centre by immediately
challenging it. Often he will think about moving the queen to c2 or b3 to prepare long castling.
Sakaev also mentions the text move as being problematic for Black, no doubt influenced by his game
against Nepomniachtchi as mentioned in the next note, but I can recommend a safe reply:
 
7...c5!
7...d4 is objectively not bad but I would prefer not to make such a committal move so early.
 
7...0-0 8.Qc2 c6 led to an eventual draw in Nepomniachtchi – Sakaev, Dagomys 2009, but Black’s
last move is rather passive and White could have obtained some initiative in the early middlegame.
 
The problem with 7...Nc6 is that White can force Black to commit himself in the centre with: 8.Qb3!
(rather than 8.Qc2 as played in Robson – Ivanchuk, Tromso 2013) 8...d4 (or 8...dxc4 9.Bxc4 0-0
10.Bc3², preventing the exchange of knights with ...Nd4, with some pressure for White due to his
active pieces) 9.Bd3² Arnold – Azevedo, corr. 2014. Compared to the 8.Qb3 line in the notes to our
main line below, Black’s pawn is still on c7, which makes his life harder.
 
8.Qc2N
8.cxd5 Qxd5 is likely to transpose to our main line after 9.Qc2 Be6.
 
An important alternative is:
8.Qb3
This forces Black to close the centre, but it also costs a tempo for White as his queen will have
to drop back to c2 sooner or later.
8...d4 9.Bd3 Nc6

42
10.0-0N
10.0-0-0? is a bad decision as the king will be vulnerable on the queenside. In Buchmann –
Schulz, email 2001, Black should have played: 10...Rb8!N Intending ...Be6 and, after suitable
preparations, ...b5. The following sequence looks logical: 11.Qc2 Be6 12.h4 h6 Covering the
g5-square before resuming play on the queenside. 13.Rde1 a6 14.Be4 Preventing ...b5, but not
for long. 14...Qd7 15.g4 And now 15...b5! 16.cxb5 Na7!µ (but not 16...axb5? 17.Bxc6 Qxc6
18.Nxd4=) gives Black a powerful initiative on the queenside.
10...Qc7 11.Rfe1 h6
Covering the g5-square and preparing ...Be6. Black may also think of expanding on the
kingside with ...g5 in the near future, as shown below.
12.h3
12.Be4 can be met by 12...Be6 13.Qd3 g5! 14.h3 Rg8³ with good prospects on the kingside.
12...Be6 13.a3 g5 14.Qa4
Intending to break with b2-b4.

14...Qd7!
Threatening ...Nxe5.
15.Qc2 0-0-0 16.b4 Rdg8ƒ
The dangerous ...g4 break is on the cards.
 
8...Be6!
I prefer this to 8...Nc6 9.cxd5 Qxd5 10.Bc4 Bf5! 11.Bxd5 Bxc2 when the ending is close to equal,
but White may get some chances due to the soon-to-be-doubled pawns on the c-file.
 
9.cxd5

43
 9.0-0-0?! is again too risky for White: after 9...Nc6 10.Kb1 Qd7 11.cxd5 Bxd5 12.Bc3 Qe6³
Black gets fine play on the queenside.
 
I also considered 9.Bd3 Nc6 10.0-0 dxc4 11.Bxc4 Bxc4 12.Qxc4 Qd7= followed by ...Qe6, when
Black is doing fine.
 
9...Qxd5 10.Bd3 Nc6 11.Be4
Inviting simplifications.
 
11.0-0 can be met by: 11...0-0-0!? 12.Be4 Qc4 13.Qb1 White keeps the queens on in order to create
some chances against the black king, but a good answer is:

13...Qa6! Moving the queen away from the exposed position on c4. 14.Be3 Kb8 15.Rc1 Nd4=
Black’s pieces are well placed and he is about to get a passed d-pawn.

44
11...Qc4 12.Qxc4 Bxc4 13.b3 Ba6 14.Bxc6†
This seems like the only ambitious try. White spoils Black’s structure at the first opportunity, but it
is of no great concern.
 
14...bxc6 15.0-0-0 c4=
Black’s pair of bishops easily compensates for the slight weakness of his pawn structure.
 

A12) 7.Bd3

7...c5 8.c3 Nc6

45
 Black continues in typical fashion, playing these two natural moves before deciding how to
develop his light-squared bishop.
 
9.0-0
It is important to realize that White is not only looking towards the kingside, but also preparing to
break with b2-b4 in the short term.
 
9...Bg4!
Black develops the bishop with tempo, due to the threat to the e5-pawn.
 
9...Be6 can be met with 10.b4! cxb4 11.cxb4 Bxb4 12.Bxb4 Nxb4 13.Nd4 Nxd3 14.Qxd3 0-0 15.f4
when White has some pressure on the kingside, although it is worth mentioning that Black can hold
his own after:

15...g6 16.f5 gxf5 17.Nxf5 Qg5!?N (17...Kh8 also proved to be fine after 18.Rab1 Qg5 19.Rxb7
Rfb8 20.Rb3 Rxb3 21.axb3 Rc8= in Ragger – Gelfand, Graz [4] 2019) 18.Rf3 Kh8 19.Raf1 Rac8=
 
10.Re1 Qd7
Black continues to prevent the b2-b4 break, the point of the text move being to stop any Qa4†
tricks.
 
11.h3 Bh5
11...Bf5?! allowed White to carry out his plan after 12.Bxf5 Qxf5 13.b4! 0-0 14.bxc5 Bxc5 15.Be3
Bxe3 16.Rxe3 Rad8 17.Qa4² when Black was under slight pressure in Anton Guijarro – Nisipeanu,
Batumi (ol) 2018.
 

46
12.Bf4 Qe6!
12...0-0?! allows White to move away from the pin with: 13.Qc2! Bg6 (13...Bxf3? doesn’t work
here due to 14.Bxh7† Kh8 15.Bf5!± Be4? 16.Qxe4!+–) 14.Bxg6 hxg6 15.Rad1 Rad8 16.Bg5²
 
The text is an excellent waiting move. With the bishop away from d2, Black need not fear the b2-b4
break. I believe that White’s best try is:

13.Be2!N
By supporting the knight in this way, White enables the queen to move to a better square.
 
13.a3 has the idea to play for the thematic b2-b4 break, but it takes some time and it is not so
effective for White now: 13...0-0 14.b4 (14.Qc2?! Bxf3 15.Bxh7†? Kh8 16.Bf5 fails here due to
16...Be4! when White doesn’t have the Qxe4 trick, as Black’s queen is supported on e6) 14...h6
15.Bg3 b6 Black is doing fine. White went for the ending with 16.Nd4?! Bxd1 17.Nxe6 fxe6
18.Raxd1, perhaps hoping that the bishop pair would count for something, but the strong 18...c4!³
enabled Black to seize the initiative in Grischuk – Caruana, Berlin 2018.
 
13...0-0 14.Qd2 h6
In positional terms, Black is doing fine, but we should check what happens if White tries to
complicate matters.
 
15.Nd4!? cxd4 16.Bxh5 dxc3 17.bxc3
It is important for Black to play:

47
17...d4!
Eliminating the d-pawn before it becomes weak.
 
18.Bg4 dxc3 19.Qxc3 Bb4! 20.Qg3 Qg6 21.Red1 Rad8=
Black is doing fine.

A2) 6.Qxd2
 
This may seem an odd choice but it has its benefits; for instance, the queen may go to f4 to support
White’s kingside play. Another point is that a subsequent c2-c4 will immediately threaten the d5-
pawn, unlike in variation A11 when the bishop on d2 blocked White’s queen along the d-file.
 
6...c5!
6...Be7 has been more popular but I found the text move to be an effective antidote against White’s
set-up with the queen on d2. From here, Black’s next few moves tend to be quite easy: he wants to
play ...Nc6, ...h6 and ...Be6, after which he can seriously consider ...g5.
 
7.c4
Let’s see what happens if White opts for simple development: 7.Bd3 White places his bishop on its
ideal square and prepares to castle. 7...Nc6 8.0-0 h6 Before developing his bishop to e6, it is useful to
cover the g5-square. 9.Re1 Be6 10.c3

48
10...Qd7 11.a3 g5!³ With ...0-0-0 coming up next, Black had promising play on the kingside in Audi
– Nikcevic, Djenovici 2018.
 
Fabiano Caruana is one of the top ‘role models’ that Petroff players should look to emulate. A game
of his last year continued:
7.c3 Nc6 8.Qf4?!
Hari transfers the queen to the kingside a bit too early. As a general rule, such a manoeuvre is
ineffective when Black’s bishop is still on f8.
8...h6 9.Bd3 Be6 10.0-0
Now Caruana seized the initiative.
10...g5! 11.Qe3
The strongest follow-up would have been:

49
11...g4!N
Improving on 11...Qc7, which gave Black somewhat better chances in Harikrishna – Caruana,
Saint Louis (rapid) 2019. The text move is more energetic.
12.Nd2
12.Ne1 Qb6!µ prepares ...0-0-0 followed by ...c4 at the right moment, leading to a much better
endgame for Black. (12...c4 13.Bc2 d4³ is slightly less accurate, but still rather promising.)
12...c4! 13.Bc2 Qc7 14.f4 0-0-0µ
White’s pieces are misplaced and Black is ready to break in the centre with ...d4.
 
7...d4
In variation A11 I was reluctant to commit to a rigid central formation too soon. However, the
presence of the queen rather than the bishop on d2 inhibits White’s development, which makes this a
more appealing version for Black. From this point, a useful guideline is that Black should generally
aim to exchange the light-squared bishops.
 
8.Bd3 Nc6 9.0-0
Inferior is:
9.Qf4?!
I repeat: this move is not so good when the bishop has not moved from f8, as a subsequent Qg3
will not hit the g7-pawn.
9...h6 10.Qg3

50
10...Qc7!
Tying the enemy knight to the defence of the e5-pawn.
10...Be6?! was well met by 11.Nd2! and White was ready to advance his f-pawn in Howell – A.
Saric, Rijeka 2010.
11.h4
Preventing ...g5.
11...Be6
This natural developing move incidentally threatens ...Nb4.
12.a3 0-0-0 13.h5

13...Qd7!N

51
 The idea of this novelty is to exchange the light-squared bishops with ...Bf5, which is the main
strategic aim for Black in this structure.
13...Kb8 was a bit less purposeful in Dedina – Simeonov, corr. 2018.
14.0-0 Bf5 15.Ne1 Rg8³
Intending to break with ...g6 at a suitable moment. Black has better chances on the kingside than
White has on the queenside.

9...Be7 10.a3
This is always a useful move; White eliminates any ideas involving ...Nb4, and also prepares a b2-
b4 break at the right moment.
 
10.Qf4
Now that Black’s bishop has moved from f8, transferring the queen to g3 is a more serious
option. However, the problem is that Black will simply castle on the queenside and ignore the
attack on g7!
10...Be6 11.Qg3
Cohen reaches this position in a note when discussing his reasons for avoiding 5...Nxd2, and
evaluates it as better for White. As you can see from the notes that follow, I have reached a
different conclusion.
11...Qd7!
11...Qc7?! proved to be inaccurate after 12.Ng5! Bxg5 13.Bxg5² when the bishop pair was
significant in Almasi – Socko, Khanty-Mansiysk 2007.
12.Bd2
Grabbing the ‘free’ pawn with 12.Qxg7? runs into 12...0-0-0 13.Qh6 Nb4!³ when Black
regains the material with an obvious initiative.
12.a3!?N is a possible try, when my analysis continues: 12...0-0-0 13.b4 h5! Gaining space on
the kingside and threatening ...h4. 14.Bg5 f6! A typical break in this structure, opening lines on

52
 the kingside and in the centre. 15.Bd2 Bf5 16.Bxf5 Qxf5= White must be careful not to become
worse.
The text can be met by:

12...0-0-0!?N
The most energetic, although 12...h6 was also pretty good for Black in Wohl – Csonka,
Oberwart 2006.
13.Ng5
After 13.Qxg7? h6!–+ the threat of ...Rdg8 is too dangerous.
13...Bxg5 14.Bxg5
14.Qxg5 can be met by 14...h6 15.Qg3 h5! 16.h4 and now the typical exchanging move
16...Bf5³, when I believe Black is slightly better due to the strong protected passed pawn on d4.
Now Black has the strong:

53
14...f6! 15.exf6 gxf6 16.Bf4
16.Bxf6 Rhg8 17.Qh4 Rdf8³ gives Black excellent play for the pawn.
16...Bf5!³
Thematically exchanging bishops, after which the protected passed d-pawn gives Black slightly
better prospects.
 
10...Bg4
Black develops quickly and intends to trade the light-squared bishops in one way or another.
 
11.Re1
Another logical try is: 11.Qf4N 11...Bh5! The bishop is heading for g6. 12.Qg3 Bg6 13.Bxg6
fxg6! This is generally the correct recapture in this structure. (13...hxg6? 14.Re1±, with the idea to
break with e5-e6 at the right moment, is dangerous for Black) 14.Ng5 Bxg5 15.Bxg5 Qc8 Black
wants to establish a blockade on the light squares, so it feels logical for White to try:

54
16.e6 This leads to an almost forced sequence of moves: 16...h6 17.Bf4 0-0 18.Rae1 Rf6 19.e7 Qd7
20.Bd6 Re8! The position proves to be equal, for instance: 21.Bxc5 b6 22.Bb4 a5 23.Bd2 Rxe7=
 
Here I suggest:

11...0-0!N
The tempting 11...Bxf3 12.gxf3 was not so easy for Black in Kayser – Farias, corr. 2013.
 
12.Be4
12.Qf4 is again met by 12...Bh5= with the idea of ...Bg6.
 

55
12...Qd7 13.Qd3 g6 14.h3 Bf5!=
The potential doubled f-pawns will be practically impossible for White to attack, so Black is doing
fine.

B) 4.Bd3 d5

We will analyse B1) 5.dxe5 followed by the main line of B2) 5.Nxe5.
 
B1) 5.dxe5
 
The same central structure will be reached later in the chapter in variation B23. The difference here is
that White’s knight remains on f3, rather than being traded on e5 for its counterpart on b8. The
present version is inferior for White, as it will be harder for him to threaten f4-f5, which is his most
promising plan in the dxe5 structure.
 
5...Nc5
We will consider White’s two main options: B11) 6.0-0 and B12) 6.Be2.
6.Nc3 c6 7.Nd4 is mentioned by Cohen as a reason to prefer 5...Be7 over 5...Nc5 on the last move,
but I am not concerned by it. 7...g6!? is one good answer (which Cohen mentions himself, without
further comment), and 7...Nxd3† 8.Qxd3 Na6 also looks pleasant enough for Black.
 

B11) 6.0-0 Be7

 

56
It is also possible to exchange the bishop immediately with 6...Nxd3, after which 7.Qxd3 Be7 8.Nc3
c6 is likely to transpose to one of the main lines below.

7.Nc3
7.Be2 is possible, with a transposition to variation B12.
 
7...c6
7...0-0? loses a pawn to 8.Nxd5+– so Black supports it before castling.
 
We will study B111) 8.Nd4 and B112) 8.Re1, after first checking some minor alternatives:
 
8.Bf4 0-0 9.Bg3 proves to be a dubious manoeuvre after: 9...Nxd3 10.Qxd3

57
10...Na6! A typical way of developing the b8-knight. 11.Nd4 Nb4!?N (11...f5 is also decent: 12.exf6
Bxf6 13.a3N [13.Rae1?? was easily refuted by 13...Nb4 14.Qd2 Bxd4 15.Qxd4 Nxc2–+ in
Amonatov – Mosadeghpour, Ahvaz 2018] 13...Nc5 14.Qd2 Qb6³) 12.Qe2 c5 13.Ndb5 Bf5³ Black’s
pieces are well placed while White’s lack harmony.
 
8.Be2
Saving the bishop from getting exchanged is not such a good idea at this point, as White’s play
is too slow.
8...Nbd7!
8...0-0 transposes to variation B121 below. Black is doing absolutely fine in this case, but the
text move is more precise, as the pressure on the e5-pawn prevents White’s thematic plan of
Nd4 followed by f2-f4.
9.Re1
I also considered 9.Bf4 Ne6!?N (9...f6 was fine too in Dragun – Ziaziulkina, Rewal 2011)
10.Bg3 0-0 when the knight on e6 is ideally placed and Black can look to gain space on the
queenside next. For example, 11.h3 a5 12.a3 Ndc5³ followed by ...b5, with fine play for Black.
9...0-0 10.Bf1 f6
A draw was agreed here in Papp – Paschall, Budapest 2010, but Black is firmly in the driving
seat. For example:

58
11.Bf4
11.Nxd5!? is an interesting attempt to complicate the game, but after 11...cxd5 12.Qxd5† Kh8³
I don’t rate White’s compensation.
11...fxe5 12.Nxe5! Ne6!
12...Rxf4?! 13.Nxd5!÷ leads to a mess.
13.Bg3 Nxe5 14.Rxe5 Bf6³
Black has the better chances due to his well-placed pieces.
 

B111) 8.Nd4

This is White’s most popular continuation, preparing f2-f4.

59
 
8...Nxd3
There is no particular reason to delay this exchange any longer.
 
With that being said, 8...0-0 is also quite playable, and if 9.Be2 we transpose to variation B121
below.
 
9.Qxd3 0-0 10.f4
White would love to gain additional space with f4-f5, but we can nip this idea in the bud.
 
10.Re1!? transposes to variation B112 below.
 
10.Nf5 can be simply met by: 10...Bxf5 11.Qxf5 Qd7 12.Qd3 Na6³ The knight is heading towards
c5 or b4 and Black had nice control of the light squares in Berndt – Ernst, Germany 2009.
 
10...f6!?
Breaking up White’s pawn structure before it expands any more on the kingside.
10...f5 was fine for Black in Navara – Kramnik, Prague 2008, but I like the idea of opening the
position for the two bishops.
 
11.Bd2
11.exf6? Bxf6µ gives Black a dream position.
 
11.e6 has to be met by 11...f5!, cutting off the support of the e6-pawn. After 12.Nxf5 Bc5† Black
soon regains the pawn with a fine position, for instance: 13.Be3 White should exchange a pair of
bishops. (13.Kh1?! Bxe6 14.Ng3 Nd7µ gave Black the bishop pair plus superior central control in
Goerens – Faseyitan, Lucerne [ol] 1982) 13...Bxe3† 14.Nxe3 Bxe6³ Black’s stable pawn structure in
the centre gave him a slight edge in Slawinski – Mamonovas, corr. 2008.
 
11...Na6 12.a3 Nc5 13.Qg3
Now Black has an important manoeuvre:

60
13...Qb6 14.Kh1 Qa6!
An excellent move, threatening ...fxe5.
 
15.Kg1
15.b4? would be a desirable move but 15...Qc4!µ is a convincing answer.
15.Rfe1 fxe5 16.fxe5 Qc4 17.Be3 Ne6³ gives Black a fine position, similar to the main line below.
 
15...Bd7 16.Be3 fxe5 17.fxe5 Ne6³
With the bishop pair and a firm blockade on the e6-square, Black had slightly better chances in
Morrow – Morley, corr. 2016.
 

B112) 8.Re1

61
 I think this is the best way for White to proceed.
 
8...0-0 9.Nd4
9.Bf1?! Bg4³ leaves White having to deal with an unpleasant pin.
 
9...Nxd3 10.Qxd3 Na6
The knight heads for the e6-square via c5 or c7.
 
11.a3
I checked a few other ideas:
 
11.Nf5? is bad due to 11...Nb4! 12.Nxe7† Qxe7 13.Qd1 Bf5µ and Black exerts strong pressure,
Hoffmann – Wuchterl, Germany 1992.
 
11.Bf4 is met by 11...Nc5 12.Qg3 Bh4, preventing any Bh6 ideas. 13.Qe3 Ne6 14.Nxe6 Bxe6
15.Ne2 Be7 With a solid game and the bishop pair, only Black could hope to be better in Georgiadis
– Tsiamis, Kallithea 2009.
 
11.Qg3
White threatens Bh6, but an excellent reply is:
11...Bh4 12.Qf3 Qd7!
Black’s main idea is to kick the d4-knight with ...c5, as the knight no longer has the f5-square
available. Black can also think of exchanging queens with ...Qg4, as shown after:

62
13.Bf4
13.h3?! is well met by 13...c5 14.Nb3 Nb4³ followed by ...Qf5, with pressure against the
queenside.
13...Qg4!
13...c5? doesn’t work due to 14.e6!±.
14.Qxg4 Bxg4³
Black is slightly better due to his bishop pair and good central control, Tiits – Koesebay, corr. 2009.
 
11...Nc5 12.Qg3
12.Qf3 a5! secures the position of the c5-knight. 13.b3 was seen in Svana – Husek, Slovakia 2018,
when Black could have seized a clear advantage with:
 

63
13...f6!N 14.e6 White needs to keep the position closed, otherwise Black’s bishops will become too
powerful. 14...Bd6 15.Bf4 Re8µ White is about to lose the e-pawn.
 
12...Bh4 13.Qe3
13.Qf3 was played in Sitnikov – Kozlitin, Prokopevsk 2012, when 13...f6!N would have been
strong, for instance: 14.e6 f5 Taking control of the e4-square for the knight. 15.Bf4 Ne4 16.g3 Bf6
17.Nxf5 Bxe6 18.Nxe4 Bxf5 19.Nxf6† Qxf6³ Black is slightly better in this opposite-coloured-
bishop position because the light squares around the white king are weak.

13...Ne6 14.Nce2 Nxd4 15.Nxd4 Re8 16.Bd2 Bd7=

64
 With the bishop pair and a solid centre, Black was comfortably equal in Schonberger – Krebs,
email 2012.
 
 

B12) 6.Be2

This is the most popular move. White prevents the exchange of his strong bishop.
 
6...Be7 7.0-0 0-0
We will analyse B121) 8.Nc3 and B122) 8.Be3.
8.c4 gives Black a comfortable game after exchanging queens: 8...dxc4 9.Bxc4 Qxd1 10.Rxd1 Be6
11.Bxe6 In Kosteniuk – Ju Wenjun, Huaian (blitz) 2016, Black recaptured with the f-pawn and
eventually won, but objectively it is best to keep the pawn structure intact with 11...Nxe6N, when
Black is doing fine in the endgame.

B121) 8.Nc3 c6

65
 White has played this way in several games, but the problem is that his knight is rather passive on
c3 and he has lost the option of playing c2-c4.
 
9.Nd4
Unblocking the f-pawn is logical.
 
I also checked:
9.Be3 Nbd7
Unless White solves the problem of the c3-knight, Black will always be slightly better.
10.Re1
10.Rb1 with the idea of b2-b4 can be met by: 10...a5 11.a3 Ne6 12.Re1 Qc7 The e5-pawn is
now a problem. 13.Bd4 Nxd4 14.Qxd4 f6! 15.exf6 Bxf6µ Black’s dark-squared bishop was too
strong in Handke – Moen, Oslo 2009.
10...Re8 11.Qd4
Black proceeded with 11...Nf8 in Matinian – Melkumyan, Sitges 2017, but it was better to exert
pressure on the e5-pawn with:

66
11...Bf8!N
The queen goes to c7 next, and White will have a hard time defending the weak pawn. For
example:
12.Bf1 Qc7 13.Qf4 b5!µ
With dominating control over the queenside and the centre. (However, please note that 13...Nxe5??
loses to 14.Nxe5 Qxe5 15.Bxc5+–.)
After the text move, I believe Black should execute the following thematic central break without
delay:
 
9...f6! 10.Bf4
After 10.exf6 Bxf6 11.Be3 Ne6 12.Nxe6 Bxe6 13.Bg4 Bxg4 14.Qxg4 Qd7³ the bishop on f6 was
strong while the knight on c3 remained misplaced in Dembo – Mamedjarova, Khanty-Mansiysk (ol)
2010.
 
10.f4
Sakaev covers this variation in a note and claims that Black must play accurately to equalize,
but I am of a different opinion.
10...fxe5 11.fxe5 Ne6
Sakaev only mentions 11...Nbd7 in his analysis, but Black has better moves available.
11...Qc7!?N 12.Bf4 Ne6 13.Nxe6 Bxe6 also looks quite promising.

67
12.Be3
12.Nf5?! Bc5† 13.Kh1 Nd7 14.Qe1 Nxe5 15.Qg3 occurred in Hruby – Pokora, Czech
Republic 2004, when 15...Ng6!N–+ would have left White without the slightest compensation
for the lost pawn.
12.Rxf8† Bxf8³ 13.Nf3? Nd7 14.Bd3 was seen in Van den Heever – Mikheev, Khanty-
Mansiysk (ol) 2010, when Black should have continued with 14...g6!N followed by ...Bg7 and
...Qc7, when the e5-pawn will soon fall.
12...Rxf1†
12...Nxd4 followed by ...Be6 and ...Nd7 also gives Black slightly better chances.
13.Qxf1?!
13.Bxf1 is preferable although 13...Nxd4 14.Bxd4 Nd7³ still favours Black.
Now in Tarleva – Glidzhain, Anapa 2011, Black missed a simple way to increase his advantage:

68
13...Bc5!N 14.Qf2 Bxd4 15.Bxd4 Nxd4 16.Qxd4 Qb6µ
Black will get a clearly better endgame due to the weakness of the e5-pawn.
 
Now in Guliyev – Nezar, Nancy 2008, it would have been best for Black to develop the b8-knight
with gain of tempo:

10...Nbd7!N 11.exf6
11.e6 is met by 11...Ne5 when the e6-pawn is weak, and if 12.Bg4 Nxg4 13.Qxg4 f5! 14.Nxf5
Nxe6³ Black keeps the upper hand.
 
11...Nxf6 12.Be5 Bd6³

69
 Black’s pieces are well placed while the knight on c3 remains a problem for White.
 

B122) 8.Be3

This is the main move.

 
8...Nc6
Intending ...Bg4, putting pressure on the e5-pawn.
 
8...c6 is a reliable alternative although, compared to the previous variation, White has not yet
committed his knight to c3. Still, after 9.c4 dxc4 10.Bxc4 Nbd7 the position is balanced.
 
9.Nc3
9.c4? would be a clear mistake here due to 9...dxc4 10.Bxc4 Bg4µ when the e5-pawn comes under
pressure.
 
9...Be6 10.a3
White’s idea is to kick the c5-knight with b2-b4. There are other games featuring strong GMs,
showing how the position can be interpreted in different ways:
10.Qc1 prepares Rd1 to put pressure on the d5-pawn, so 10...Qc8 is a natural reply, mirroring
White’s intentions. After the further 11.Rd1 Rd8 12.Bg5 h6 13.Bxe7 Nxe7 14.Nd4 (Sakaev mentions
14.Qe3 as slightly favourable for White but I see no problems for Black after either 14...b6 or
14...Nd7 followed by ...Nf8) 14...Bg4= Black was fine in Almasi – Landa, Reggio Emilia 2008. A
good plan from here will be to move the knight away from c5 and try to get in ...c5, gaining space in
the centre.
 

70
10.Bb5
This aims at ruining Black’s pawn structure on the queenside but that is not so much of an
issue; in fact, I like the idea of forcing White’s hand with:
10...a6!? 11.Bxc6 bxc6 12.Nd4 Bd7
Black has the simple plan of ...f6, opening the position for his two bishops.

13.Qd2
13.Nb3 occurred in Wastney – Dive, Wellington 2013, when the straightforward 13...Nxb3N
14.axb3 f6 15.exf6 Rxf6 would have reached a balanced position where Black’s activity and
two bishops are no less important than his weakened queenside structure.
13...f6 14.f4 fxe5 15.fxe5 a5
It is useful to control the b4-square.
16.b3 Qe8!
A nice manoeuvre of the queen to g6.

71
17.Bg5 Rxf1† 18.Rxf1 Bf8
Of course Black should not give up his bishop pair.
19.Qf2 h6 20.Be3 Qg6=
The position remained more or less equal in Schreiner – Thelen, email 2011. However, in a
practical game I would slightly prefer Black due to the potential power of the two bishops, plus the
fact that the e5-pawn could become a target.
 
I also considered White’s other possible move to the b5-square:
10.Nb5
White aims to solve the problem of his passive knight by transferring it to d4.

10...Nd7

72
 Black relocates a knight as well, hitting the e5-pawn.
11.Bf4 a6 12.Nbd4 Nxd4 13.Nxd4
 
13.Qxd4 can be met by the thematic 13...f6, and after 14.Rae1 c6 15.Qd2 fxe5 16.Nxe5 Nc5=
Black was fine in Telepnev – Osipov, corr. 2009.
13...c5 14.Nxe6 fxe6
This time it is White who has the bishop pair, but his dark-squared bishop is restricted by the
pawn on e5 and Black has good chances to gain space on the queenside.
15.Bg3 c4!
It is important to get this move in before White can play c2-c4.
16.Bg4 Nc5 17.b4 cxb3 18.axb3
18.cxb3 a5 secures the knight on c5, and after 19.Qd4 Kh8 20.Rad1 Ra6! Black had a reliable
position in Zelcic – A. Saric, Bosnjaci 2010.
18...a5 19.f3
White intends to improve his dark-squared bishop by transferring it to the g1-a7 diagonal.
Therefore it is best to get rid of his bishop pair with:

19...Bh4! 20.Qd4!
This is White’s best attempt to cause problems.
20...Bxg3 21.Qxc5 Bxe5 22.Bxe6† Kh8 23.Rae1 Bd6 24.Qxd5 Qc7 25.h3
White has won a pawn but Black has good compensation due to the weak dark squares around
the white king, which are especially noticeable with opposite-coloured bishops. I recommend:

73
25...Rad8N
25...Bg3 led to a subsequent draw in Shirov – Gelfand, Kallithea 2008, but the text move seems
safest to me. Black threatens to win the queen with a bishop check, as well as the simple
capture of the c2-pawn.
26.Qc4 Qc5†! 27.Qxc5
27.Kh1 Qg5!© is annoying for White.
27...Bxc5† 28.Kh1 Rd2=
White has an extra pawn but no advantage, due to Black’s activity and the drawing tendencies of
the opposite-coloured bishops in the event that all the rooks are exchanged.
 
10...a6
It is worth investing a tempo to prevent any ideas involving Bb5 or Nb5. Another point is that
White will not be able to drive the knight away from c6 with b4-b5.
 
11.Qe1!?
This move strengthens the b2-b4 push: White anticipates the reply of ...Ne4 and so he supports the
c3-knight in advance. Another point of the text move is to bring a rook to d1.
 
White has also tried the immediate 11.b4 Ne4 12.Qd3 (12.Qe1? is strongly met by 12...Nxb4!) when
the best reply is:

74
12...Bf5!N Maintaining the tension and creating some threats. 13.Rfd1 (White can sacrifice the
exchange with 13.Nxd5!? Ng3 14.Qc4 Nxf1 15.Rxf1 but after 15...Qd7 16.Rd1 Qe6 White’s
compensation is sufficient only for equality) 13...Qc8! Supporting the f5-bishop and threatening to
take on f2. 14.Nd4 Nxd4 15.Qxd4 Nxc3 16.Qxc3 c6= With a balanced position.
 
This position was reached in Bologan – Huzman, Budva 2009, plus some other games involving
strong GMs. I found a promising new idea:

11...Nd7!?N
Immediately hitting the e5-pawn.
 

75
12.Rd1
 
12.Bf4!? enables Black to win a pawn, although White can get just enough compensation for it.
The critical line continues: 12...g5! 13.Bg3 g4 14.Nd2 Ncxe5

15.f4! gxf3 16.Nxf3 White gets some compensation for the pawn due to Black’s slightly exposed
kingside16...Nxf3† 17.Rxf3 Bf6 18.Bd3 c5 Although the position is dynamically balanced, White
will have to play precisely to prove his compensation. A good plan for Black could be ...Re8 and
...Nf8, bringing the knight to the defence of the king.
 
12...Ndxe5 13.Nxe5 Nxe5
Now it’s important for White to play:

76
14.f4!
Kicking the knight and threatening f4-f5.
 
The tactical shot 14.Nxd5? doesn’t work due to: 14...Bxd5 15.c4 (or 15.f4 Qd6!µ) 15...Nxc4!
16.Bxc4 Bxc4 17.Rxd8 Raxd8–+ Black will emerge with two rooks and a pawn for the queen.
 
14...Ng4 15.Bxg4 Bxg4 16.Rxd5 Qc8=
White has some space on the kingside but Black has the bishop pair, and the position remains
balanced.
It is worth mentioning that White does not have time to trap the light-squared bishop with 17.f5?
due to 17...c6, breaking the support of the f5-pawn.

B2) 5.Nxe5

77
 This can be considered the main line of the 3.d4 variation, and our coverage of it will span the
remainder of this chapter and all of the next one.
 
5...Nd7
Black immediately challenges the knight on e5. The main line is 6.Nxd7, and it will be examined
in the next chapter.
 
Before then, we will consider three significant alternatives: B21) 6.Nc3, B22) 6.Qe2 and B23) 6.0-0.
 
6.Nxf7 leads straight to a draw after: 6...Kxf7 (6...Qe7?! is inferior in view of 7.Qe2 Qxf7 [or the
even worse 7...Kxf7? 8.Qh5† Ke6 9.Bxe4! dxe4 10.d5† Kd6 11.Bf4†+–] 8.f3 Ndf6 9.Nd2ƒ and
White has some initiative) 7.Qh5† Ke6

78
8.Qe2 (8.Qg4† leads to the same result) 8...Kf7 9.Qh5† Ke6 10.Qe2 Kf7 11.Qh5† Ke6 ½–½ Zubov
– Rakhmanov, Dubai 2018.

B21) 6.Nc3
 
White offers the exchange of both pairs of knights.
 
6...Nxc3 7.bxc3 Nxe5 8.dxe5
White’s pawn structure has been compromised but, thanks to the advanced pawn on e5, he hopes to
create some attacking chances on the kingside.

79
8...Bc5
Several other moves have been tried but I like this one the most.
 
9.0-0
The alternative is to prevent ...0-0 by means of:
9.Qh5 Be6
With kingside castling not possible, Black prepares to send his king in the opposite direction.
10.0-0
10.Rb1?! has been played in a few games but it is not so effective, as Black does not need to
save the b7-pawn after all: 10...Qd7! 11.Bg5 (11.Rxb7?! is a risky sacrifice for White: 11...Bb6
12.Qg5 Qc6 13.Qxg7 0-0-0 14.Rxb6 axb6 15.0-0 Rdg8ƒ followed by ...d4 with good play for
Black) 11...h6 12.0-0 Rg8! 13.Bd2 (13.Bh4?? Bg4–+) 13...0-0-0³ Black’s position was
preferable in Movsesian – Haba, Czech Republic 2004. He can start playing on the kingside
soon, and White’s weak queenside pawns can always be a factor in an eventual endgame.
10...Qd7 11.a4!?N
This seems the most logical choice, gaining space on the queenside even before Black castles.
11.Bg5 was played in Movsesian – Navara, Prague 2005, when Black could have played
11...h6!N 12.Qh4 Rg8 13.Bd2 g5! 14.Qxh6 Be7ƒ followed by ...0-0-0, with a strong initiative
on the kingside.
11.Qh4 h6 covers the g5-square and intends ...Qe7 next. White responded with 12.Qg3 in
Turov – Weglarz, Decin 1998, and now 12...0-0-0N would have given Black good prospects.
11...0-0-0 12.a5 a6
Black is doing fine. My illustrative line continues:
13.Ba3 Bxa3 14.Rxa3 c5 15.Rb1

80
15...c4 16.Bf1 Qc7 17.Ra2 g6=
Black’s superior pawn structure more than makes up for the shortcomings of his slightly inferior
bishop.
 
9...0-0 10.Qh5
White induces weaknesses on the kingside, but they can be covered by our bishop.
 
10...g6 11.Qh6
I found a useful improvement:

11...Be7!N

81
 The bishop drops back to cover the weak dark squares.
 
Black was able to hold after 11...f5?! in Huschenbeth – Bogner, Le Castella 2018, but I think it’s
better to wait for White to play f2-f4 before playing ...f5.
11...Re8 gave White decent chances on the kingside after 12.Bg5 Be7 13.Bxe7 Qxe7 14.f4 Qf8
15.Qh4 Qe7 16.Qg3² in Lorenzo de la Riva – Laxman, Barcelona 2012.
 
12.Be3 c6
This is a useful way to strengthen Black’s position while waiting to see what White will do.
 
13.f4
Gaining space on the kingside seems like the most logical way to proceed.
 
13.Rae1 can be met by 13...Re8 intending ...Bf8, and after 14.Qf4 Be6 15.Qg3 b5= Black is fine.
 
13...f5!
This is a typical reaction against f2-f4, which is well worth remembering. Black is doing fine, for
instance:

14.exf6 Bxf6 15.f5?!

Opening lines on the kingside seems natural, but Black gets the upper hand after:
 
15...Bxf5 16.Bxf5 gxf5 17.Rxf5 Bxc3³
Finally one of those weak queenside pawns has fallen, and White does not have enough firepower
for a successful attack.
 

82
 

B22) 6.Qe2

This has been played by many strong GMs but it has become a rarity in recent years, as it has
become clear that Black obtains plenty of play for the pawn which is about to be sacrificed.
 
6...Nxe5 7.Bxe4
7.dxe5?! makes no sense because 7...Nc5 gives Black a better version of variation B23 below, as
White’s queen is misplaced on e2.
 
7...dxe4 8.Qxe4
8.dxe5?! was well met by 8...Qd4!³ in Torralba Brosa – Morais Carreras, Montcada 2017, the point
being to prevent Bf4, as the b2-pawn hangs.
 
8...Be6 9.Qxe5
The alternative is even less attractive for White:
9.dxe5?! Bd5 10.Qg4 h5 11.Qh3 Qe7
Cohen focuses on 11...Be6 but I find the text move more attractive.
If the e5-pawn falls, Black will benefit from having two bishops in an open position. Therefore
White usually defends it with:
12.f4
12.0-0 simply hands Black a slight plus after 12...Qxe5 13.Nc3 Be6 14.Qh4 Be7 15.Qa4† c6³
when the bishops were strong in Vaclav – Hatzl, corr. 2009.
The strongest way for Black to continue is:

83
12...Qe6!
Exchanging queens is beneficial for Black, as White will come under pressure on the kingside.
13.Qxe6† fxe6 14.Kf2
Black also gets a good initiative against White’s other tries:
14.0-0 appears logical, but after 14...Bc5† 15.Kh1 0-0-0 16.Nd2 Rhf8 17.c4 Bc6 18.Nb3 Bd4
19.Nxd4 Rxd4 20.Be3 Rxc4³ Black had regained the pawn and his bishop was clearly stronger
than its counterpart in Haessel – Kamberi, Chicago 1997.
14.Nc3!? sees White return the extra pawn in order to focus on development. However, after
14...Bxg2 15.Rg1 Bc6 16.Be3 Rd8 17.Ne2 Kf7 18.Nd4 Be4³ Black continued to exert some
pressure with his bishop pair in Kotyurgin – Golubenko, corr. 2014.

14...g5!
84
 Opening up lines against the white king.
15.Nc3 Bc5† 16.Be3 Bxe3† 17.Kxe3 gxf4† 18.Kxf4 Bxg2 19.Rhg1
The most accurate continuation is:

19...Rg8! 20.Rad1 Ke7 21.Ke3 Raf8

Black kept some initiative with no risk in Maggiora – Llorente Alvarez, email 2012.

9...Qd7 10.0-0
An important alternative is:
10.Be3
Maintaining the option of long castling. I suggest meeting it with:
10...Bd6!?

85
 10...0-0-0?! is inaccurate in view of 11.Qa5! Kb8 12.0-0 when White had good chances to
consolidate his extra pawn in Fleischanderl – Bueno, corr. 2002.
10...Bb4†!? seems like a reliable alternative though.
11.Qxg7
11.Qa5 has been played in a bunch of games but it is ineffective here due to 11...Qc6!N,
attacking both g2 and c2. Play may continue 12.0-0 (or 12.Nc3 Qxg2 13.0-0-0 0-0³ when
Black’s king is safe and he has better prospects due to the bishop pair) 12...Qxc2 13.Qb5† c6
14.Qxb7 Rb8ƒ followed by ...0-0, when Black is extremely active.
11.Qg5!?N can be simply met by 11...0-0 12.0-0 Rae8© when Black’s active bishops offers him
lasting compensation for the pawn, and he has good chances to create play on the kingside.
11...0-0-0

12.Nc3
It is high time for White to think about his king safety and development, so it is logical to get
the knight out and prepare long castling.
12...Bb4 13.0-0-0 Rhg8 14.Qe5
The players agreed a draw here in De Magalhaes – Badolati, corr. 2009. Black has full
compensation for the two pawns and can continue in a few different ways, the safest being:
14...Bxc3 15.bxc3 Qc6 16.Qc5 Qxc5 17.dxc5 Rxg2=
With an equal endgame.
 
10...0-0-0©
Black has promising compensation in the spirit of the Marshall Attack, except that he has castled
on the queenside, which opens up additional attacking possibilities.

86
11.Be3 Bb4!
A subtle move, intending to provoke an undesirable pawn advance from White. Another purpose of
the text move is to prepare ...f6, as the white queen will not have the a5-square available.
 
12.a3
12.Nc3?! f6 13.Qg3 Bxc3! 14.bxc3 reaches a situation where White’s extra pawn is of little value
and Black is well placed to start an attack, as shown after:

14...h5! 15.h4 g5! 16.f3 Rdg8µ Karpov is arguably the greatest of all time in positions with opposite-
coloured bishops, but even he was unable to solve White’s problems in Karpov – Larsen, Tilburg
1980.

87
 
12.a4 intends to exchange queens with Qb5, but 12...a6 stops it, and after the further 13.c3 Bd6
14.Qh5 Rde8³ Black was doing well in Da Riva Alonso – Cook, corr. 2016. A natural plan is to
expand on the kingside with ...f5-f4, with a good initiative.
 
12.c3
This is an obvious candidate, challenging the bishop.
12...Bd6!
12...f6 has been more popular but there is no need for it here.
13.Qh5
13.Qa5 is thematic but 13...Bd5! is a strong reply: the bishops on d5 and d6 are going to create
a lot of problems for White. For example: 14.Nd2 A logical move, but it is met by a nice
tactical shot. (14.Qxa7? Qg4 15.f3 Qh5 16.h3 Rhe8!–+ and White will be crushed on the
kingside; 14.h3 may be White’s best but 14...Kb8© offers Black fine compensation for the
pawn) 14...Bxg2! 15.Kxg2 Qg4† 16.Kh1 Qh3 17.f4 Qxe3³ White’s exposed king remained a
source of problems in Michalczak – Gawehns, Germany 1995.

13...f5!
A typical way to gain space on the kingside.
14.f4 Rhe8 15.Qe2 Bf7 16.Nd2 Qe6 17.Rae1
A draw was agreed here in Martynov – Podvoysky, corr. 2012. The obvious continuation would
be:
17...Qxa2=
With a balanced position.

88
12...f6 13.Qe4 Bf5 14.Qf3 Bg4 15.Qg3?!
I think White should settle for 15.Qe4 Bf5= with a repetition. The text move keeps the game going
but to White’s detriment.
 
15...Bd6 16.f4
White clings to his extra pawn but creates new weaknesses in his camp.
 
16...Bf5 17.Qf2 Rhe8
It is obvious that Black has more than enough compensation for the pawn. The immediate threat is
...Bxc2.
 
18.Re1 Qb5 19.b3 Bg6!
Making room for the queen to come to f5.
 
20.Nc3 Qf5 21.Re2 Re7 22.b4 Rde8µ
Black is clearly better due to the powerful pressure along the e-file and on the light squares, Vesely
– Cruzado Duenas, corr. 2012.
 

B23) 6.0-0

89
 Black should meet this move by trading off the active knight.
 
6...Nxe5 7.dxe5 Nc5
Just as in the earlier variation B1, we retreat the knight and prepare to exchange the strong d3-
bishop.
 
White’s most important responses are B231) 8.Nc3, B232) 8.Be2 and B233) 8.Be3.
 
White can also go for a direct expansion on the kingside with 8.f4!?, but Black can counter it with:
8...Nxd3 9.Qxd3 (9.cxd3 Bf5N= followed by ...h5 gives Black good control on the light squares, as
10.g4? Bd7 leaves White too exposed) 9...g6! Covering the f5-square.

90
10.Nc3?! (10.Be3 is preferable, when 10...c6 transposes to variation B233 below) 10...Bc5†!
Exploiting the lack of a bishop on e3 to develop with tempo. 11.Kh1 c6³ We have transposed to the
Papp – Pap game, mentioned under the 9.f4?! line in the notes to variation B231 below.

B231) 8.Nc3

The idea behind this move is usually to transfer the knight to e2 at some point.
 
8...c6
Adding extra protection to the d5-pawn.
 
9.Ne2
Two other moves deserve close attention:
 
9.f4?!
This move is pretty thematic and has been the most popular choice, but it is inaccurate.
9...Nxd3!
9...g6 is also playable although White has scored well after 10.Ne2.
The text move is more accurate. The point is that Black should wait to see how White
recaptures on d3 in order to determine whether or not ...g6 is needed.
10.cxd3
10.Qxd3 is well met by 10...Bc5† 11.Kh1 (11.Be3?? loses a piece after 11...Bxe3† 12.Qxe3
d4–+) 11...g6 12.Ne2 Bf5 13.Qc3 Qb6³ and Black had firm control over all the key squares in
Papp – Pap, Szombathely 2010.

91
10...Bf5!
It is obviously desirable to block the further advance of White’s f-pawn, but the real trick is to
appreciate that Black does not need to fear g2-g4.
11.Ne2
11.g4? leads to a chronic weakening of the kingside, and after 11...Bd7 12.f5 Qh4! White is in
trouble. Black will play ...h5 next and (after the kingside opens up) can castle long whenever he
needs to.
11...Qb6† 12.d4 h5!
Advancing the pawn to h4 takes away the g3-square from White’s knight and also gains space
on the kingside.

13.a3N

92
 13.Ng3?! Bg4µ followed by ...h4 is great for Black.
13.Bd2?! was simply met by 13...Qxb2 14.Rf3 Qc2 15.Qxc2 Bxc2µ and Black was a
comfortable pawn up in Trani – Sukhanitskij, corr. 2012.
13...h4 14.h3 Be7³
Black has a pair of well-placed bishops; the light-squared one is especially strong.
 
9.Qf3!?
Transferring the queen to the kingside is an interesting plan: the great Geller played this way
twice in the 1980s, and a few other strong GMs have played it more recently.
9...Be7
Black was successful with 9...Qh4 in Smirin – Smith, Helsingor 2015, but there is no real need
to prevent White’s manoeuvre.
10.Qg3
I recommend an important novelty:

10...0-0!N
10...g5?! led to an interesting game in Sameeh – Kobese, Tanta 1998, but there is no need for
such antics.
The text move is better, as it turns out that White’s ‘threat’ is not really dangerous.
11.Bh6
11.f4? is met by the typical break 11...f6!µ when Black has a great game.
11...Ne6 12.Bf5
This is the whole point behind the queen transfer to g3, but Black can easily defuse the attack.
12...Bh4!
It is important to note that the immediate 12...Bg5? does not work due to 13.Bxg5 Qxg5
14.Bxe6 Qxg3 15.Bxf7†!± when White wins a valuable pawn.
13.Qg4

93
 Only now should Black play:

13...Bg5! 14.Bxg5 Qxg5 15.Qxg5

15.Bxe6?! Qxg4 16.Bxg4 Bxg4³ reaches an endgame where only Black can realistically have
winning chances. White’s far-advanced pawn on e5 is slightly vulnerable, and Black can break
with ...f6 at the right moment.
15...Nxg5 16.Bd3
It seems logical for White to preserve his ‘superior’ bishop, but Black is still doing fine.
16...Bd7=
Black has no problems and will break with ...f6 at the right moment.
 
9...Be7 10.f4
10.Be3 has been the choice of some correspondence players. Black can simply go for: 10...0-0
11.Ng3 (11.c3 creates an escape square on c2 for the bishop, so of course Black exchanged it with
11...Nxd3 12.Qxd3, and now the thematic break 12...f6! 13.exf6 Rxf6 14.Ng3 Rf7 15.Rfe1 Bd7= left
him with the bishop pair and good prospects in Sharpell – Boettcher, corr. 2009) 11...Nxd3! It is
important not to delay this exchange any longer, otherwise White may consider putting the bishop on
f5. 12.Qxd3 Now the simplest approach would be:

94
12...f6!N Again we see this thematic break. (12...Re8 was also not bad in Henri – Ponomarev, corr.
2013) 13.exf6 Bxf6 14.Bd4 Qd6= With ...Bd7 coming soon, followed by mobilizing the queen’s
rook, Black is doing fine.

10...f6!
Immediately executing the desired central break before White gets a chance to advance further on
the kingside with f4-f5.
 
11.Be3
This was Mamedyarov’s choice.

95
11.Nd4?! 0-0 12.Bf5 was played in Khader – Sarsam, Beirut 2000, when Black could have seized the
initiative with:

12...fxe5!N 13.fxe5 Bxf5 14.Nxf5 Ne4! The knight is excellently placed here. 15.Qg4 Bc5† 16.Kh1
g6µ Black’s pieces are more active than their counterparts, especially his knight.
 
11...Nxd3 12.Qxd3 0-0 13.c4!
White tries to break up Black’s pawn centre as well. I recommend a novelty:

13...dxc4N
This seems simplest, although 13...fxe5 also worked out fine in Mamedyarov – Kramnik, Wijk aan
Zee 2008.

96
 
14.Qxc4† Qd5! 15.Qxd5† cxd5
The endgame is balanced. It is worth adding one important detail:
 
16.Nc3 Be6 17.f5
This may appear unpleasant, but Black has a nice resource.
 
17...d4! 18.Bxd4 fxe5 19.fxe6 exd4 20.Nd5 Bd6=
One way or another, the respective central pawns are both likely to fall, resulting in complete
equality.

B232) 8.Be2

Preserving the bishop has been White’s most popular choice.

 
8...c6
This is always a useful move in this structure.
 
9.c4!?
Challenging Black’s central structure in this way is ambitious, but it gives Black a chance to seize
the initiative.
 
One of the points of Black’s previous move is revealed afte9.f4?, when 9...Qb6!µ is unpleasant for
White.
 

97
9.Be3 reaches a position covered under 9.Be2 in the notes to variation B233 below.
9.c3 is rather slow, and after something like 9...Be7N 10.Be3 0-0 11.Nd2 Qc7 12.f4 f6= Black is
doing fine.
 
9.Nd2 is playable but not too threatening: 9...Be7 10.Nb3 (10.Nf3 Qc7 11.c4 dxc4 12.Bxc4 Bg4=
followed by ...Rd8 and/or ...0-0 is good for Black, Haraldsson – Nierobisz, corr. 2005) 10...0-0

11.Be3 Nxb3 12.axb3 Qc7 13.f4 f6= Following this thematic break, challenging the e5-pawn, Black
had a fine position in Zelcic – Kovacs, Oberwart 2012.

9...d4! 10.Nd2
10.b4N is well met by 10...d3!³.

98
 
10...d3 11.Bf3 Bf5!N
 
11...Be6 was the continuation of Saxena – Coyne, corr. 2008, but I prefer to utilize the bishop to
overprotect the d-pawn, while keeping the e6-square free for the knight. The critical continuation is:
 
12.b4 Ne6 13.c5
And now it’s important to play:

13...Qd4! 14.Rb1
14.Qa4!? is another idea, when 14...Bxc5 15.Bxc6† bxc6 16.Qxc6† Ke7 17.bxc5 Rhd8 18.Nf3
Qd5³ keeps some initiative for Black due to the strong pawn on d3.
 
14...Be7
Black is doing fine. A possible continuation is:
 
15.Bxc6†!? bxc6 16.Qf3 0-0 17.Qxf5 Rfd8©
White has won a pawn but Black’s active pieces and the passer on d3 more than make up for it.

B233) 8.Be3 c6
 
Once again, we start by playing this always-useful move, keeping the other pieces flexible.
 
9.f4!

99
 This is the critical continuation. White does not spend time safeguarding the light-squared bishop,
but rather prefers to gain space on the kingside with f4-f5.
 
A slower, relatively harmless alternative is:
9.Be2 Be7 10.f4
10.c3 transposes to the note on 9.c3 in variation B232 above.
10.c4 is a typical break in this structure, but after 10...dxc4 11.Qxd8† Kxd8 12.Bxc4 Be6
13.Bxe6 Nxe6 14.Nc3 Kc7 Black held comfortably in Adams – Yu Yangyi, Shenzhen 2017.
10...0-0

11.c4
After 11.Nc3?! f6! 12.exf6 Rxf6³ White’s knight was clearly misplaced in Zelcic – A. Saric,
Opatija 2012.
11...dxc4 12.Bxc4 Qxd1 13.Rxd1 Bg4=
Black was fine in Lebedev – Capuano, corr. 2003.

100
9...Nxd3
Black preferred 9...g6 and obtained a reasonable position in Jones – Smirin, Batumi (ol) 2018, but I
think it’s best to eliminate the bishop immediately.
 
10.Qxd3
10.cxd3?! can be met by 10...Bf5!, blocking White’s f-pawn and intending ...h5, gaining complete
control of the light squares on the kingside. My illustrative line continues: 11.g4? (White should
settle for 11.Qe2 h5 although Black is obviously comfortable) 11...Bd7 12.f5 This is how White
would like to play, but his position becomes unglued after:

101
12...Qh4! 13.Qf3 h5! 14.g5 Be7µ White is unable to keep his kingside pawn formation intact, and
Black can evacuate his king to the queenside whenever he needs to.
 
10...g6!
Controlling the f5-square.
 
11.Nc3
The knight is heading for e2, from where it can go to g3 or d4 to help control the f5-square.
11.Nd2 can be met by 11...c5!?, gaining space in the centre and covering the d4-square. Then 12.c4
d4 13.Ne4!? is an interesting attempt to complicate things, but Black is safe enough after:

13...Bf5! 14.Bd2 Be7 15.Qe2 Qd7 Black proved to be fine after 16.Nd6† Bxd6 17.exd6† Qe6
18.Qe5 Qxe5 19.fxe5 Kd7= with an equal endgame, Kolcak – Copar, corr. 2008.
 
Now I prefer to play on the light squares with:

102
11...Bf5!?N
11...Be7 reaches a position that has been tested in many grandmaster games. Black’s position
seems quite reliable; see, for instance, Adams – Sandipan, Gibraltar 2018.
 
12.Qd2 h5 13.Qf2!
I also considered:
13.Ne2 h4!
Covering the g3-square and introducing the possibility of ...h3 at some point.
14.Nd4
White aims for the f4-f5 break.
14.h3 stops Black’s h-pawn from going any further but gives Black time to control another
important square with 14...c5! 15.c3 Qd7= when he is doing fine.

103
14...Be4 15.f5 gxf5 16.Nxf5 Qd7
Now Black’s king is ready to escape to the queenside.
17.Qf2
17.Nd6† Bxd6 18.exd6 Rg8 19.Rf2 h3³ is good for Black, who will take on d6 next.
17...0-0-0!
Black should not worry about the a7-pawn, as he will get a strong attack in return for it:

18.Bxa7
18.Nxh4 is well met by: 18...Rg8! Threatening ...Qg4. 19.Nf3 (19.Qxf7? Be7!–+) 19...d4!
20.Bf4 Bc5µ With too many problems for White.
18...Rh5 19.Nxh4 Be7 20.Nf3 Rg8µ
The attack is too strong.
104
13...b6
Preventing Bxa7 and improving Black’s control over the dark squares in general, especially in
connection with Black’s next move.
 
14.Ne2 c5!
Preparing to gain additional space in the centre with ...d4.
 
14...h4?! 15.Nd4² would be annoying.
 
15.Ng3 Be6 16.f5 gxf5 17.Nxf5
White has achieved the desired pawn break but Black has sufficient counter-chances.

105
17...d4 18.Bd2 h4 19.Rae1 Qd5!
Centralizing the queen and preparing ...0-0-0 and/or ...h3.
 
20.Nxh4
20.Bg5 h3 21.gxh3 Rxh3ƒ looks promising for Black.
 
20...Rg8
Black has good compensation for the pawn, for example:

21.g3 0-0-0 22.Ng2!

A good manoeuvre, but it does not solve all of White’s problems.
106
 
22...Qb7 23.Nf4 Bc4 24.Nd3 Be6©
White has to be careful of the weak light squares near his king.
 

Conclusion
 
This has been the first of two chapters dedicated to 3.d4 against the Petroff. 3...Nxe4 is our reply,
when 4.dxe5 d5 5.Nbd2 was the first major option we considered. After 5...Nxd2, both 6.Bxd2 Be7
and 6.Qxd2 c5! turn out fine for Black, according to my analysis.
 
4.Bd3 is more popular, when 4...d5 gives White another choice. 5.dxe5 Nc5 reaches the same
thematic structure as the previous variation, but with the possibility of Black’s knight exchanging
itself for the d3-bishop, rather than the enemy knight. White may or may not safeguard the bishop by
retreating it to e2, but in general Black will look to stay solid in the centre and carry out the ...f6
break, which should offer him plenty of activity when timed correctly.
 
Finally we looked at 5.Nxe5 Nd7, which can lead to a variety of scenarios. 6.Nc3 sees White
weaken his pawn structure in pursuit of the initiative, but Black can neutralize it with careful play
and hope to exploit White’s pawn weaknesses in a future endgame. Next we considered 6.Qe2,
which wins a pawn, but has mostly fallen out of favour since it became clear how strong Black’s
compensation is. Finally, 6.0-0 Nxe5 7.dxe5 Nc5 reaches the same thematic structure as variations A
and B1, but with an extra pair of knights having been traded. Once again, the position presents
challenges for both sidesrutor.i, but a timely ...f6 break usually assures Black of good chances.

107
A) 8.Qh5 Qf6! 57
A1) 9.Qxd5 57
A2) 9.Nc3 58
B) 8.Nc3!? Nxc3 9.bxc3 0-0 10.Qh5 f5 61
B1) 11.Bg5 62
B2) 11.Rb1 63
B3) 11.Re1 c6 12.Bg5 Qc7 64
B31) 13.c4 65
B32) 13.Re3!? 66
C) 8.c4 c6 9.cxd5 cxd5 67
C1) 10.Qh5 0-0 11.Qxd5 Bc6 12.Qh5 g6 68
C11) 13.Qh6 68
C12) 13.Qh3 68
C2) 10.Nc3 Nxc3 11.bxc3 0-0 12.Qh5 g6! 13.Qxd5 Qc7 14.Bh6!? Rfd8 71
C21) 15.Qg5 72
C22) 15.Qf3 73
108
 
1.e4 e5 2.Nf3 Nf6 3.d4 Nxe4 4.Bd3 d5 5.Nxe5 Nd7 6.Nxd7
This is the most popular move by a wide margin. White agrees to the knight trade but wants to
keep his pawn on d4, rather than allow the dxe5 structure from variation B23 of the previous chapter.
 
6...Bxd7 7.0-0
Castling is by far the most natural and popular move. Other options exist, but do not require a great
deal of attention:
 
7.Qh5 Nf6 gives White nothing.
 
7.Nd2 is less than ambitious, and 7...Nxd2 8.Bxd2 Qf6!? (or 8...Qe7†) is fine for Black.
 
7.Qf3 gives Black a few playable options, the cleverest being:

7...Bb4†!? (7...Bd6?! does not really work due to 8.Bxe4 Qe7 9.Nc3) 8.c3 Bd6 Black is fine thanks
to the pawn on c3, as 9.Bxe4?! is well met by 9...Qe7 and if 10.Nd2 f5 things are really becoming
unpleasant for White.
 
7.Nc3
If White is going to play this move, he usually does so after castling.
7...Nxc3 8.bxc3 Bd6 9.Qh5!?
9.0-0 transposes to variation B below.
The text move is an independent try which has scored quite well for White. Now ...0-0 is
obviously not possible, but Black can arrange long castling easily enough.
9...Qe7† 10.Be3

109
 Black can solve his opening problems with just one accurate move:

10...Bc6!
This is better than defending the pawn with 10...c6, as whenever White breaks with c3-c4, it
will activate the bishop on c6.
10...Be6 has been the most popular choice but 11.Bg5 prevents Black from castling and leaves
him with some mini-problems, which he has not been too successful at solving in practice.
11.0-0 g6 12.Qh3
12.Qh6 can be met by 12...0-0-0 with good prospects for Black.
12...0-0 13.c4 dxc4 14.Bxc4 Rad8=
Black was doing fine in Nunn – Fridman, London (blitz) 2017.

110
7...Bd6
Black prepares ...0-0, so White must act immediately in order to create some chances. Here White
has three main options: A) 8.Qh5, B) 8.Nc3!? and C) 8.c4
 

A) 8.Qh5 Qf6!
 
An important move. Black responds to the attack on d5 by attacking the d4-pawn in return, and also
prepares ...0-0-0 if needed.
 
8...c6?! is too slow. 9.Bxe4 dxe4 10.Nd2² With the queen on h5, Black does not have options like
...f5 or ...Bf5 in order to defend the e4-pawn.
 
8...Nf6 is a popular alternative but I find the text move more appealing.
 
We will analyse A1) 9.Qxd5 and A2) 9.Nc3.
 
9.Be3 threatens f2-f3 as well as Qxd5; the best way to parry both threats is: 9...Bf5! (9...0-0-0?! was
seen in Ivanchuk – Giri, Tbilisi 2017, when the simple 10.Qxd5N 10...Bc6 11.Qh5± would have left
Black without enough compensation for the pawn)

10.f3 g6 11.Qh6 Bf8 12.Qf4 Bd6 13.Qh6 Bf8 Neither player had a good way to avoid the repetition,
so a draw was agreed in Guseinov – Fridman, Tallinn 2019.

A1) 9.Qxd5

111
 Eating the pawn is a critical move to consider, but Black gets a lot of activity for it.
 
9...Bc6 10.Qc4?!
Trying to save the extra pawn is the most ambitious approach, but excessively risky for White.
 
10.Bb5?! is also bad due to 10...0-0-0! 11.Qh5 (11.Bxc6?? simply loses to 11...Bxh2† 12.Kxh2 Rxd5
13.Bxd5 Qd6† 14.Kg1 Qxd5–+ as in Burrows – Ferguson, Sunningdale 2007) 11...Bxb5 12.Qxb5
Qxd4³ when White’s undeveloped queenside gave Black at least a slight edge in Rysbayeva – Otazo
Sanchez, Dresden (ol) 2008.
 
White’s best and safest approach is to give the pawn back with 10.Qh5N 10...Qxd4 11.Nd2 0-0
12.Nxe4 Bxe4 13.Be3 Qd5= with a level position.
 
10...0-0-0
The pair of bishops on d6 and c6 pointing at the kingside is not good news for White! Here are
some sample variations.
 
11.Be3
11.Nc3? Bxh2† 12.Kxh2 Rxd4–+ and White’s king will be executed.
11.c3? is strongly met by: 11...Bd5! 12.Qa4 Nxf2! 13.Rxf2 Qh4 14.h3 Qg3–+
 
11.Nd2? is also bad in view of 11...Bxh2†! 12.Kxh2 Rxd4 13.Qb3 Nxd2 14.Bxd2 Bxg2! and Black
wins.
 
11.h3 Ng5‚ and there are numerous sacrifices in the air on the kingside.
 

112
I found a useful novelty:

11...Rhe8!N
11...Bf4 was fine for Black and a draw was soon agreed in Callow – Garcia Rojas, email 2003, but
we can play more ambitiously.
 
Bringing the rook to the e-file creates chances of an exchange sacrifice on e3, as shown in the
illustrative line below.
 
12.Nc3 Qh4 13.h3 Nxc3 14.Qxc3 Rxe3! 15.fxe3 Qg3
Black has a powerful attack, for example:
 
16.e4 Qh2† 17.Kf2 f5!µ
The white king is going to suffer.
 

A2) 9.Nc3

113
 This is the usual move. Black should respond by eating the central pawn.
 
9...Qxd4 10.Be3
The alternative is:
10.Nxd5 Bc6 11.Ne3
After 11.Nc3?! Nxc3 12.bxc3 Qxc3 13.Bg5 h6 14.Rae1† Kf8³ White did not have enough
compensation in Kurnosov – Guliyev, Bastia 2010.
11...g6 12.Qh3
12.Qh6 allows 12...0-0-0 with good prospects for Black. 13.Nc4 was seen in Rodriguez Vila –
Umetsubo, Florianopolis 2015, when 13...Bc5!N 14.Be3 Qd5³ would have posed White some
problems.
This position was reached in Deep Fritz 10 – Kramnik, Bonn (4) 2006. I suggest the simple
continuation:

114
12...0-0N
Black is fine, for example:
13.Nc4 Rfe8 14.Be3 Qf6 15.Nxd6 Nxd6 16.b3 Be4=
Getting rid of White’s bishop pair.
 
10...Qe5 11.Qxe5† Bxe5 12.Nxd5
12.Nxe4 dxe4 13.Bxe4 0-0-0 should be equal: 14.c3 (14.Bxa7? would be too greedy! 14...b6 15.a4
[15.c3 Ba4! covers the d1-square and prepares for ...Rd2; Black is winning after 16.b3 Bb5 17.Rfc1
Rd6–+ intending ...Bc6 followed by ...Kb7] 15...Bxb2 16.Ra2 Bc3µ White is about to lose material)
14...Rhe8 15.Bd5 Be6 16.Bxe6† Rxe6= With a completely level endgame, Filippov – Egin, Tashkent
2007.
 
12...Nf6 13.Rae1 Nxd5 14.Bd4
14.Bd2 f6 15.f4 can simply be met by 15...Nxf4 16.Bxf4 0-0, and after the further exchanges
17.Bxe5 fxe5 18.Bc4† Kh8 19.Rxe5 Rxf1† 20.Kxf1 Re8 21.Rxe8† Bxe8= the ending was a clear
draw in Hovhannisyan – Jablonicky, Riga 2018.

115
14...0-0-0!
14...f6?! proves to be inaccurate: 15.Bxe5 0-0-0 16.Bg3² Nb4 (or 16...g6 17.a3² and White keeps
the bishop pair, which makes life uncomfortable for Black in this structure, Lu Shanglei – Bai Jinshi,
China 2018) 17.Re7 Nxd3 18.cxd3 g5 19.Rc1 c6 20.f3² Even though Black held the draw fairly
easily in Lu Shanglei – Zhou Weiqi, China 2018, the bishop on g3 and rook on e7 are annoying, and
White could have made his opponent suffer if he had handled the next few moves more precisely.
 
15.Bxe5
15.Rxe5 Nf4!N is precise, as Black gets ready to meet Rg5. (15...Nb4 16.Rg5 Nxd3 17.cxd3 left
Black under some pressure in Palac – A. Saric, Pozega 2018) 16.Re7 (16.Rg5?! Rhg8!³ turns out well
for Black here) 16...Nxd3 17.cxd3 Bc6 18.Bc3 Rd7 19.Rfe1 f6= The endgame is equal.
 
15...Nb4
An important move: Black’s main aim is to eliminate the bishop pair as soon as possible.
 
16.Bc4!?
This is the most ambitious choice, which has been tested in a lot of correspondence games.
 
16.Be4?! would save the bishop pair but hang the a2-pawn, enabling Black to fight for the advantage
with: 16...f6! (16...Bb5?? backfires after 17.Bxg7 Rhg8 18.Bf6 Rde8 19.Bf5†+– when White wins a
pawn while saving the exchange) 17.Bc3 Nxa2 18.Ba5 b6 19.Ra1 bxa5 20.Rxa2 Rde8!³ Attacking
the bishop and planning ...Re5 next.
 
16.Bxg7 Nxd3 17.cxd3 Rhg8

116
18.Be5 (18.Bf6N should be met by 18...Bc6!, threatening to take on d3 as well as g2, with an equal
position) 18...Bc6 19.f3 (19.Bg3 Rxd3 20.Re7 Rd7= was also clearly equal in Anderskewitz – Owen,
corr. 2017) 19...Rxd3 20.Bg3 Re8= The endgame with rook and opposite-coloured bishops should
end in a draw, Shankland – Yu Yangyi, Danzhou 2018.

16...Nxc2
16...f6?! would be imprecise due to 17.Bc3 Nxc2 18.Re7² when Black’s knight c2 is out of play
and White exerts pressure on the 7th rank.
 
17.Re2

117
 17.Rc1 is met in the same way as the main line: 17...f6! 18.Bxf6 (18.Bxc7 Kxc7 19.Rxc2 Bc6=
was also fine for Black in Lane – Zelesco, Auckland 2014) 18...gxf6 19.Rxc2 Rhe8 Another equal
endgame arose in Zelcic – A. Saric, Zagreb 2008.
 
17...f6!
An important intermediate move.
 
18.Bc3 Bf5 19.g4
I also considered:
19.Re7 Rd7 20.Rxd7 Kxd7 21.g4 Bg6
Black cannot take the pawn with 21...Bxg4? due to 22.Bd3+– and the knight is trapped.
22.h4
Now Black’s best idea is to prepare the return of the knight to d4 with:
22...c5 23.Rd1†
23.h5 Be4= is also fine for Black.
23...Ke7
23...Nd4 24.Bxd4 cxd4 25.h5 Be8 26.Rxd4† Kc7= is also satisfactory.
24.h5 Be8 25.Bd5 b6 26.Be4 Nd4 27.Bxd4 cxd4 28.Rxd4

28...g6!
This endgame is completely equal and White soon gave up trying for anything more.
29.hxg6 hxg6 30.Rc4 Kd8
½–½ Del Lungo – Cantelli, corr. 2018.

118
19...Bg6 20.f4!
This is White’s only serious try for an advantage, but we have another correspondence game
showing how to neutralize it.
 
20...Nd4 21.Re7 b5!
An important move, forcing White to go for an opposite-coloured-bishop endgame.
 
22.Bxd4
22.f5? turns out badly for White after 22...bxc4 23.fxg6 hxg6³ when taking on g7 will allow
...Ne2†.
 
22...bxc4 23.Bxa7 Rhe8
The endgame holds no danger for Black, for instance:
 
24.Bc5
24.Rxg7 can be met by 24...Bd3 25.Rd1 Re2„ when Black’s rook on the second rank gives him
good counterplay for the pawn, especially with opposite-coloured bishops.
 
24...Rd2 25.f5 Rxe7 26.Bxe7 Bf7 27.Rf2 c3 28.bxc3 Rxf2 29.Kxf2 Bxa2=
The draw was inevitable in Borzenko – Espinoza, corr. 2009.
 

B) 8.Nc3!?

119
 White voluntarily allows his pawns to be doubled on the c-file. On the plus side, White quickly
eliminates the active knight on e4 and hopes to seize the initiative on the kingside by following up
with Qh5.
 
8...Nxc3 9.bxc3 0-0 10.Qh5
10.Rb1 b6 should transpose to variation B2 after 11.Qh5 f5.
 
The text move is the real reason why 8.Nc3 is a credible try for an advantage. The double threat
against h7 and d5 forces Black’s reply:
 
10...f5
This weakens Black’s position, especially the e5-square. Now one of White’s main ideas will be to
exchange the dark-squared bishops and try to create some pressure down the e-file.
 
I analysed three main options: B1) 11.Bg5, B2) 11.Rb1 and B3) 11.Re1.
A quick exchange of the dark-squared bishops by means of 11.Qf3 c6 12.Bf4 is not so effective
due to 12...Qc7 13.Bxd6 Qxd6= when White does not have time to plant a rook on e5.
 
11.c4
This move is also pretty harmless.
11...dxc4 12.Bxc4† Kh8 13.Bg5
13.Rb1 b6 14.Re1 Qf6 15.Bb2 Rae8 16.Ba1 Re4!³ and Black seized the initiative in Kanarek –
Kuzubov, Minsk 2017.
13...Qe8 14.Qh4 Qg6 15.Be7
Black was ready to play ...Rae8, hence the text move. In Nataf – Bluvshtein, Montreal 2008,
Black exchanged on e7 and went on to hold, but I think it would be even simpler to play:

120
15...Rfe8N 16.Rfe1 Bxe7 17.Rxe7 Qd6 18.Rae1
18.Rf7?! is bad due to 18...Re4³.
18...Qb4!
Hitting e1 and c4. White has to settle for the perpetual after something like:
19.R1e3 Qxc4 20.Rxg7 Kxg7 21.Qg5†=
 

B1) 11.Bg5

121
 This been played in a lot of correspondence games but Black has comfortably held a draw in
almost all of them.
 
11...Qe8
Another possibility is: 11...Be8 12.Qh4 Qd7 13.Bf4 Bxf4 14.Qxf4 White has carried out his main
aim of exchanging the dark-squared bishops, but 14...Qd6 15.Qxd6 cxd6 16.Rab1 Bc6 17.Rfe1 g6=
reached a typical solid endgame with no problems for Black in Bobel – LeBled, corr. 2018.
 
12.Qf3 Qg6 13.Bf4
White has achieved his aim of exchanging the dark-squared bishops but has spent several tempos
doing it.
 
13...Bxf4 14.Qxf4 Qc6!
Protecting the c7-pawn as well as hitting the c3-pawn.
14...Qd6 is a slightly passive way of defending in the endgame, but nevertheless completely sound.
15.Qxd6 cxd6 16.Rfb1 Bc6 17.a4 Rac8 18.Kf1 g6 19.Ke2 Rc7 20.Kd2 Kf7 Black held firm in
Hedlund – Bokar, corr. 2012. Nevertheless, the engine likes White and in a practical game he would
be able to play on for a while without any risk, so I would tend to prefer the cleaner path to equality
featured in our main line.

15.Rab1 a6
Covering the b5-square.
 
16.c4
16.Rfe1 is met by 16...Rae8 17.Rxe8 Rxe8 with no problems for Black.
 

122
16...dxc4 17.d5
17.Be2!? with the idea of Bf3 can be met by 17...Be6 18.Bf3 Bd5, and after 19.Bxd5† Qxd5
20.Qxc7 b5= the major-piece endgame was equal in Hedlund – Yefremov, corr. 2012.
 
17...Qa4!
Ensuring that the queens will leave the board.
 
18.Qxc4 Qxc4 19.Bxc4 b5 20.Bd3
The endgame is equal and the players did not bother playing any longer in Daubenfeld – Rubinas,
corr. 2018.

B2) 11.Rb1 b6

White can insert these moves before proceeding with one of the other normal continuations.
However, I think that the inclusion of these two moves is more likely to favour Black. A relevant
point, which is far from obvious right now, is that in the event of c3-c4, the pawn on b6 ensures that
White will not be threatening c4-c5.
 
12.Re1
12.Bg5 Qe8 13.Qf3 Qg6 14.Bf4 Bxf4 15.Qxf4 Qd6 16.Qxd6 cxd6 resulted in an equal endgame
in Korneev – Landa, Casablanca 2005. In the equivalent position after 11.Bg5, it was noted that
White would have some slight pressure in the endgame, although Black should definitely hold.
However, in the present position White is unable to play a2-a4 due to having moved the ‘wrong’ rook
to b1, so he has no advantage whatsoever.

123
 12.c4 can be met by the interesting reply: 12...Qf6!? (the normal 12...dxc4 is completely fine too)
13.Bb2 (or 13.cxd5?! Qxd4³ followed by ...Rae8, with good chances for Black to seize the initiative)
13...dxc4!?N (13...Be8 was also perfectly okay in Vaindl – Nejtek, corr. 2003) 14.Bxc4† Be6!
Making use of the queen on f6. 15.d5 is met by 15...Bf7= and Black is doing fine.
 
12...c6 13.Bg5 Qc7
Black continues exactly as in variation B3 below.
 
14.c4
14.Be7 is simply met by 14...Rfe8. A clash between two legendary players continued: 15.Bxd6
Qxd6 16.h3

This was Geller – Yusupov, Minsk 1979. Black has several playable moves but I think the simplest
approach is to avoid any c3-c4 ideas entirely with 16...b5!?N, when Black is completely fine.
 
Now Black can carry out a typical manoeuvre:
 
14...Be8! 15.Qh3
The alternative is:
15.Qf3 dxc4
15...Bxh2†?! 16.Kh1 Bd6 17.cxd5 gives White chances for some initiative.
16.Bxc4† Bf7 17.Bxf7† Rxf7 18.h3
18.h4 Qd7 19.h5 was well met by 19...Bf8= and Black was fine in Naiditsch – Fridman, Bastia
2013.
18...Qd7
Black covers the e6-square and threatens ...f5-f4.

124
 The hasty 18...f4?! would be unnecessarily risky. 19.Re6 h6 Now 20.Bh4 is assessed by the
engine as slightly better for White, while 20.Bxh6!?÷ is also playable. There is no reason for
Black to allow such possibilities.
19.Rbd1
Now the simplest move is:

19...Re8N
19...f4!? also turned out fine in Amaral – Lins, corr. 2012.
20.Rxe8† Qxe8 21.d5 cxd5 22.Qxd5 Bf8=
Followed by ...h6 and ...Kh7.

15...dxc4 16.Bxf5!?

125
 White aims to create some imbalance.
 
16.Bxc4† Bf7 17.Be6 Bxe6 18.Rxe6 Rae8 proved to be fine for Black in two games: 19.Rbe1
(19.Qb3 Rxe6 20.Qxe6† Rf7 was also equal in Shirov – Kramnik, Cazorla 1998) 19...Rxe6 20.Rxe6
Qd7 White could manage nothing better than a quick draw after 21.Qb3 Kh8 22.Qe3 Kg8 23.Qb3
½–½ in Wolf – Efremov, corr. 2010.
 
16...h6 17.Be6† Bf7 18.Bxh6 gxh6 19.Qxh6
White hopes to create some attack against the king, but Black can maintain the balance in a forcing
way:

19...Bxh2†! 20.Qxh2
Or 20.Kh1N 20...Bf4 21.Qg6† Kh8 22.Qf6† Kg8= and White has to settle for a perpetual.
 
20...Qxh2† 21.Kxh2 Bxe6 22.Rxe6 Rxf2 23.Rxc6 Rxc2=
Black should never have lost this endgame in Areshchenko – G. Szabo, Plovdiv 2012.
 

B3) 11.Re1

126
 This is the most popular move overall, and it has been tested extensively in high-level GM games.
 
11...c6
Defending the d5-pawn in advance, while making the c7-square available for the queen.
 
11...Qf6 is a logical move, but I think White gets a slight pull after 12.Qf3 c6 13.Bf4!² as in S.
Zhigalko – Fridman, St Petersburg 2018. The problem for Black is that he cannot force White to
exchange the dark-squared bishops immediately, so White can take his time to build up the pressure.
 
12.Bg5 Qc7
We will analyse B31) 13.c4 and B32) 13.Re3!?.
 

B31) 13.c4

127
 This has been the most popular choice but I consider it harmless.
 
13...Be8 14.Qh3
14.Qh4 does not change much, as Black can react in the same way. 14...dxc4 15.Bxc4† Bf7
16.Bd3 Now the simplest way to secure the f5-pawn is: 16...g6 This seems to weaken the dark
squares around Black’s king, but they can easily be defended, either by the dark-squared bishop or, in
the case of the g7-square, by the queen along the 7th rank. (16...Bg6 and 16...Qa5 are also playable)
17.Bf6 Rae8= Black was ready to exchange rooks and had no problems in Szymanski – Teeriaho,
corr. 2011.
 
14...dxc4 15.Bxc4†
15.Bxf5!? is the same idea that we saw in the Areshchenko – G. Szabo game in variation B2 above.
The only difference is that the moves Rb1 and ...b6 have not been played here, which does not alter
the evaluation at all. Thus, play once again continues 15...h6! 16.Be6† Bf7 17.Bxh6 gxh6 18.Qxh6
followed by the familiar liquidating combination:

128
18...Bxh2†! 19.Qxh2 Qxh2† 20.Kxh2 Bxe6 21.Rxe6 Rxf2 22.Re7 In this equal rook endgame, a
draw was agreed in Vachier-Lagrave – Gelfand, Khanty-Mansiysk 2009.
 
15...Bf7
Offering the exchange of bishops.
 
16.Be6
This has been the main try for White, but it leads to massive exchanges after:
 
16...Bxe6 17.Rxe6 Rae8! 18.Rae1
Taking the f5-pawn does not change much: 18.Rxe8 Rxe8 19.Qxf5 Bxh2† 20.Kf1 Bd6 and the
position remained balanced in Galytskyi – Halliwell, corr. 2018.
 
18...Rxe6 19.Rxe6
Black just needs one more accurate move to solve the last of his opening problems:

129
19...Qb6!
Exploiting the weakness of the first rank while also hitting the d4-pawn.
 
20.Qb3 Qxb3=
With the queens disappearing from the board, the ensuing endgame is equal and many
correspondence games have been drawn from here, for instance Lebled – R. Moll, corr. 2012.
 

B32) 13.Re3!?

130
 This rare move gained more followers after Nepo used it successfully last year. White focuses on
doubling along the e-file rather than breaking with c3-c4 immediately.
 
13...g6 14.Qh4
14.Qh6 does not change much. 14...Rae8 15.Rae1 Rxe3 16.Rxe3 occurred in Firat – Keler, Konya
2019, and now 16...b5!N= is a good move, for the same reasons as in the main line below.
 
14...Rae8 15.Rae1 Rxe3 16.Rxe3
It is important to stop the c3-c4 break with:

16...b5!N
Black is fine after this prophylactic move, and can trade rooks with ...Re8 next.
 
In the first two games Black erred with the natural-looking 16...Re8?!, allowing 17.c4!² (improving
on 17.Be7, as played in Fercec – Doric, Pula 2014, which was the only previous game where 13.Re3
was played) when Black’s position becomes uncomfortable and the weaknesses in his camp begin to
tell. Nepo exploited them cleanly and won impressively in Nepomniachtchi – Yu Yangyi, Astana
2019.
 
16...f4?! does not really help Black. 17.Re1 Re8 was seen in Idani – Kuzubov, Linares 2019, when
18.Rxe8†N 18...Bxe8 19.Bh6² would have maintained some pressure.
 
16...Rf7!? 17.c4 b6 was not so bad, and Black went on to hold in Korneev – Kardashevskiy, Sochi
2019. Nevertheless, life becomes simpler for Black if he prevents c3-c4 altogether.
 
17.Be7

131
 Exchanging the bishops seems like the most logical attempt to exploit the weak dark squares.
However, with so few pieces remaining on the board, Black can hold his own.
 
17...Re8 18.Qf6
18.Bxd6 Qxd6 19.Rxe8† Bxe8= is even more straightforward, as White cannot infiltrate with his
queen.
 
18...Rxe7 19.Rxe7 Bxe7 20.Qxe7 a5!?
20...Qc8= is also fine; I just want to illustrate that Black even has a tempo to spare.
 
21.h4 Qc8=
The queen comes to e8 in time to deal with any threats on the kingside. Black has a bad bishop but
White’s bishop is also limited by Black’s wall of pawns, and White’s queen is unable to accomplish
anything on the dark squares by herself.
 

C) 8.c4

This has been White’s most popular choice by a wide margin.

 
8...c6
The rare 8...0-0!? has scored well but I am happy to recommend the traditional response.
 
9.cxd5
9.Nc3 0-0 10.cxd5 Nxc3 11.bxc3 cxd5 transposes to variation C2 below.

132
 
Trying to win a pawn with 9.Re1 can be rebutted in more than one way. 9...0-0!? is the more
ambitious choice. (9...Bxh2† should lead to a draw after 10.Kxh2 Qh4† 11.Kg1 Qxf2† etc.; instead,
White preferred 10.Kf1? in Pfaffel – Sniesko, Frohnleiten 1999, and a couple of other games, when
the simple 10...0-0Nµ would have left White in a rotten position) 10.Bxe4 dxe4 This position has
occurred a few times via transposition.

11.Rxe4 is the obvious continuation, but Black obtains excellent compensation after 11...Bf5 12.Re1
Qh4 13.g3 Qh3© when White’s queenside is undeveloped and the light squares around her king are
weak, Agaeva – Mamedjarova, Baku 2001.
 
9...cxd5
White has two serious tries: C1) 10.Qh5 and C2) 10.Nc3.
 

C1) 10.Qh5
 
This can be met by a simple, thematic pawn sacrifice.
 
10...0-0 11.Qxd5 Bc6 12.Qh5
12.Qf5 g6 13.Qh3 transposes to variation C12 below.

133
12...g6
We have a further split, with C11) 13.Qh6 and C12) 13.Qh3 both requiring attention.
 

C11) 13.Qh6 Re8

 
Black brings his rook to the open e-file while preparing ...Bf8.
 
14.Be3
White defends the d4-pawn in anticipation of Black’s next move.
 
14.Nc3 is playable but completely unambitious: 14...Bf8 15.Qf4 Qxd4 (15...Bd6 16.Qh6 Bf8=
forces a draw if Black wants it) 16.Bxe4 Bxe4 17.Be3 Qe5= With further exchanges on the way,
leading to an equal endgame, Tomczak – Nurkiewicz, Warsaw 2018.
 
14...Bf8 15.Qh3?!
15.Qf4 enables Black to force a repetition with 15...Bd6 16.Qh6 Bf8=; and if you wish to keep the
game going, then 16...Nf6!? could be considered, intending ...Nd5 next. As uninspiring as this is for
White, he should settle for it, as the text move enables Black to fight for the advantage.

134
15...h5!
This method of gaining space on the kingside is a typical idea in such positions. Black now
threatens to trap the queen with ...Bd7, and there is not much that White can do about it.
 
16.Nc3 Bd7 17.Qf3 Bg4 18.Qf4 g5!
Black opted to force a quick draw with 18...Bd6 19.Qh6 Bf8 in Fleck – Treppner, Germany 1998,
but he can play for more.
 
19.Qxe4 Rxe4 20.Nxe4 f6³
Black had slightly better chances in this materially unbalanced position in Wurschner –
Fleischanderl, email 2006.
 

C12) 13.Qh3 Ng5!

 
Black continues to harass the queen and the knight prepares to come back to e6, putting pressure on
the d4-pawn.
 
14.Qg4
This move looks logical and has been the most popular choice, but it offers Black good chances to
seize the initiative.
A safer, though unambitious, approach is:
14.Bxg5 Qxg5 15.Nc3 Rad8
Black centralizes his rook and put pressure on the d4-pawn.

135
16.Rad1
16.Ne4?! was an inaccuracy in Townsend – Jimenez Alonso, email 2000, when 16...Qe7!N
17.Rfe1 Bb4 18.Nc3 Qf6³ would have enabled Black to regain the pawn while keeping some
initiative.
16.Be4 was played in Meli – Panov, email 2010, when 16...Qd2!N would have attacked both d4
and b2, thus winning back the pawn with an equal position.
16...Rfe8 17.Bc4 b5
White can and should steer the game towards a draw with:
18.Bxf7†
A draw was agreed here in J. Buecker – Zapf, email 2007, due to:
18...Kxf7 19.d5!
19.Qxh7†? Kf6 20.d5 Re7 leaves White with no real attack or compensation for the piece.
19...Bd7 20.Qxh7† Kf8
20...Kf6!? could be a risky winning attempt. 21.f3! Qh5! 22.Ne4† Rxe4 23.fxe4† Kg5
24.Qxh5† gxh5 results in a double-edged endgame where any result is possible: White has a
material advantage, with rook and three pawns versus two minor pieces, but his pawns are
blockaded and Black’s bishops are superb. Nevertheless, I would say Black runs a slightly
bigger risk of losing than White.
21.Qh8† Kf7 22.Qh7†=
 
14.Qh6 Ne6 15.Be3
15.Nc3 Nxd4 16.Bg5 Be7 17.Bxe7 Qxe7= was simply equal in Aldrete Lobo – Cornejo, email
2007.
The text can be met by a thematic trick, which we will also see in the main line.

136
15...Bf4! 16.Qh3?
White should settle for 16.Bxf4N 16...Qxd4= when Black regains the piece with an equal
position.
16...h5!
16...Qf6 was pleasant for Black in Sermek – Timoscenko, Ljubljana 2000, but the text move is
more incisive. The threat of ...Ng5 is unpleasant for White, and he has nothing better than
giving up a piece for some pawns with:
17.Bxg6 Ng5 18.Qxh5

18...Qd5! 19.Bh7† Kg7! 20.Qg4 f5! 21.Bxf5 Qxf5³

Reyes – Etchechoury, Argentina 2000. After a forced sequence, we are about to enter an endgame
where White temporarily has four pawns for a piece. However, one of those pawns is about to fall,
137
and the others are a long way back. Black’s pieces are more active, so overall we can evaluate his
winning chances as quite reasonable.

14...Ne6
The threat is ...h5, picking up the d4-pawn.
 
15.Bh6
Another way of safeguarding the central pawn would be: 15.Be3 h5 16.Qh3 White’s queen must
remain on the kingside, otherwise ...Qh4 would be too much to handle. Now in Lautier – Gelfand,
Las Vegas 1999, Black could have obtained the upper hand with:

138
16...Nf4!N 17.Bxf4 Bxf4³ Followed by picking up the d4-pawn, with the bishop pair in an open
position.
 
The text move prevents ...h5, but Black has another strong option.
 
15...Bf4! 16.Bxf4
Grabbing the exchange with 16.Bxf8? is refuted by: 16...h5 17.Qd1 (17.Qh3? loses the queen to
17...Ng5 18.Qh4 Nf3†–+)

17...Bxg2! 18.Kxg2 Qg5† 19.Kh3 (19.Kh1 Qh4–+ wins) and now Black has the strong 19...Bc7!
20.f4 Nxf4† 21.Rxf4 Qxf4–+ and Black has a winning attack.
 
16...Qxd4
Black regains the piece while keeping a slight edge. For example:
 
17.Nc3 h5 18.Qh4
18.Qg3 Nxf4 19.Rfd1 Qf6 20.Bf1 Rad8³ left White under some pressure on the kingside in Skvira
– Blitsko, corr. 2011,
 
18...Qxf4 19.Qxf4 Nxf4 20.Be4 Bxe4 21.Nxe4 Ne2† 22.Kh1 Rac8³
White had to play precisely to hold a draw in Souza – De Magalhaes, corr. 2009.
 

C2) 10.Nc3

139
10...Nxc3 11.bxc3 0-0 12.Qh5 g6!
Once again, Black sacrifices the d5-pawn temporarily.
12...f5?! can be compared to variation B, as analysed earlier in the chapter. The major drawback for
Black compared to those positions is that here White has already eliminated his doubled c-pawns,
which gives him better chances.
 
13.Qxd5 Qc7 14.Bh6
This has been the most popular choice.
 
14.h3 enabled Black to regain the pawn with ease after 14...Be6 15.Qf3 Qxc3= in Timman – Van der
Sterren, Gouda 1997.
 
14.h4 also allows 14...Be6 15.Qf3 Qxc3 and Black is fine, for instance: 16.Bh6 Rfe8 17.Qf6 Bf8
18.Bxf8 Rxf8 19.Be4 Rad8 20.Rfd1 Bd5= Guseinov – Safarli, Porticcio 2017.
 
14.g3!?
This move has been tried by several strong GMs and has scored well for White, but Black
should be fine with accurate play.
14...Be6 15.Qf3 Qxc3 16.Rb1 Qxd4 17.Be4
17.Bb2 is met by 17...Bd5! when Black has no problems whatsoever.

140
17...Qc4! 18.Bb2
18.Qf6 is met by 18...Be7!=.
18...f6!
18...Be7? 19.Rfc1 Qxa2 20.Rc7 Bd8 21.Rxb7 gave White a decisive initiative in Negi – So,
New Delhi 2011.
19.Rfd1
19.Bxb7 Rad8= (or 19...Rab8N=) also gives Black no problems, Guseinov – Grigorov, Plovdiv
2012.
19...Be5 20.Bxb7 Rab8=
This position has been reached in three correspondence games, all of which were drawn.
 
A final idea is:
14.c4!? Bxh2† 15.Kh1
White saves his c-pawn at the expense of his h-pawn, which gives the game a different
character from the other possibilities. The position should still be equal though.
15...Bc6 16.Qg5 Bd6 17.d5
17.Bb2 Rae8 produces a positional threat of ...Be4, which White should prevent in some way.
18.Rae1 (18.c5?! Be7³; 18.d5N 18...Be5 19.Qd2 Bd7 20.Rfe1 b6=) 18...Qd8 19.Qxd8 Rxd8
The endgame was level in Jakovenko – Gelfand, Dagomys 2009.
Now it is important to utilize a small tactical finesse:
17...Rae8!
17...Bd7? allows 18.Bb2 Rae8 19.Rfe1± when the bishop on b2 is a monster.
The text move avoids this by enabling us to play ...Be5 in time. For example:
18.Bb2
18.dxc6? Re5–+ is the justification of Black’s last move.
18...Be5 19.Bxe5 Rxe5 20.Qg3 Ba4 21.Rae1 Rfe8=

141
 Black is fine.

14...Rfd8
Black should not be in too much of a hurry to regain the pawn, as 14...Bxh2†?? 15.Kh1 Rfe8
16.f4! destroys all harmony between his pieces: 16...Bg3 (or 16...Re3 17.Qg5 Qd8 18.Rf3+– and the
bishop is lost) 17.Qg5+– Threatening both the bishop and mate with Qf6.
 
I analysed two main options: C21) 15.Qg5 and C22) 15.Qf3.
 
15.h3 Be6= enabled Black to take on c3 next in D. Fischer – Abramov, corr. 2008.
 
15.h4 is also met by 15...Be6 when Black should have no problems, although it is worth adding a few
more moves: 16.Qf3 (16.Qg5 Qxc3 17.Rfd1 Bf8 18.Bxf8 Kxf8= gave Black no real problems in
Bologan – Inarkiev, Poikovsky 2009) 16...Qxc3 17.Bg5
 

142
17...Qxd4!? A promising exchange sacrifice! 18.Bxd8 Rxd8 19.Be4 h5!© Threatening ...Bg4. Black
had powerful compensation in Lobanov – Gaehwiler, Skopje 2019.
 
15.Rab1 can be safely met by15...Bxh2† 16.Kh1 Bd6

17.Rfe1 (17.Rxb7?? loses after 17...Bc6 18.Rxc7 Bxd5–+ and the rook is trapped; 17.Qxb7?! is also
not great in view of 17...Qxc3³ when both the d3-bishop andd4-pawn are attacked) 17...Bf8= Black
is doing fine, Rogos – Velilla Velasco, corr. 2016.
 

C21) 15.Qg5
143
 
This has been tried by several strong players but without much success for White.
 
15...Qxc3
It is better to capture this pawn than the one on h2, as now the d4-pawn becomes isolated.
 
16.Rfd1
16.Qf6 transposes to variation C22 below.
 
16...Bf8 17.Bxf8
17.Rac1 is met by 17...Qa3 18.Bxf8 Qxf8 when the position remains balanced, Pheby – Binder,
corr. 2012.
 
17...Rxf8 18.Rac1 Qa3 19.Be4
The seemingly logical 19.Rc7?! is rebutted by 19...Ba4! 20.Rd2 Bc6 21.h4 Qd6³ and Black seized
the initiative in Souleidis – Grabarczyk, Lüneburg 2016.
 
White can try to create some play on the kingside with 19.h4 but it is also met by 19...Ba4!, messing
up the coordination in White’s camp. Best play continues 20.Qa5! b6 21.Qa6 Rad8! with equality, as
in Daroczy – Walther, corr. 2016.

19...Ba4!
Again we see this typical idea, disturbing White’s coordination.
 
20.Rd3 Qb2 21.h4 Rae8!
Pushing the bishop away from the ideal e4-square before occupying the c-file.

144
 
22.Bf3
22.f3 can be met by 22...Bc6 23.d5 f5! 24.dxc6 fxe4= and Black maintains the balance.
 
22...Rc8„
Black had sufficient counterplay in Schramm – Repp, corr. 2015.
 

C22) 15.Qf3

15...Qxc3 16.Qf6 Bf8 17.Bxf8 Rxf8

This position has been tested in a lot of games. Material is equal and so is the position; if anything,
White has to be more careful as the d4-pawn could easily become a target. Here are some examples.
 
18.Be4
White quickly became worse after 18.Qf3?! b6 19.d5 Rac8 20.h3 Rfe8³ when Black controlled
both the e- and c-files in Schneider – Semrl, corr. 2009.
 
18...Bc6 19.Rac1
19.Rfc1 soon leads to a repetition: 19...Qb2 20.Rcb1 Qc3 21.Rc1 ½–½ So – Gelfand, Amsterdam
2010.
 

145
19...Qa5 20.Rc5
20.Bxc6 bxc6= led to a drawn endgame in Ninov – Delchev, Bankia 2011.
 
20...Qd8 21.Qxd8 Rfxd8
With a completely equal endgame, Boskovic – Mannermaa, corr. 2012.
 

Conclusion
 
This chapter concludes our coverage of the 3.d4 Nxe4 4.Bd3 d5 5.Nxe5 Nd7 variation, by analysing
the main line of 6.Nxd7 Bxd7 7.0-0 Bd6, when White has three main options. We started with 8.Qh5
Qf6!, when 9.Qxd5 takes a pawn but offers Black plentiful compensation, while 9.Nc3 generally
results in simplifications to an endgame where Black is fine, although there are some subtleties
which are worth remembering.
 
We then moved on to 8.Nc3!? Nxc3 9.bxc3 0-0 10.Qh5 f5, when White hopes that the weakening of
his queenside structure will be outweighed by Black’s dark-square weaknesses resulting from the
...f5 advance. My analysis shows that Black is fine in all variations, although again there are some
theoretical nuances which are worth knowing about.
 
Finally we analysed the most popular 8.c4 c6 9.cxd5 cxd5, leading to a final split. 10.Qh5 0-0
11.Qxd5 Bc6 12.Qh5 g6 is another version of the thematic sacrifice of the d5-pawn, when Black gets
great compensation by combining threats against the d4-pawn and the white queen, as well as the
castled king in some lines. 10.Nc3 is safer, when 10...Nxc3 11.bxc3 0-0 12.Qh5 g6! 13.Qxd5 Qc7
sees Black regain the sacrificed pawn with a satisfactory position. Here too, there are certain lines
which should be memorized, but the details are not too complicated.

146
A) 4.Nxf7?! Kxf7 76
A1) 5.d4 c5! 76
A11) 6.dxc5 77
A12) 6.Nc3 78
A2) 5.Nc3!? 79
B) 4.Nd3 Nxe4 5.Qe2 Qe7 82
B1) 6.b3 83
B2) 6.Nc3 84
B3) 6.Nf4 Nc6 86
B31) 7.c3 87
B32) 7.Nd5 88
C) 4.Nc4 Nxe4 90
C1) 5.d4 90
C2) 5.Nc3 92
C3) 5.Qe2! Qe7 6.Ne3 Be6 94
C31) 7.Nc3 95

147
 C32) 7.c4!? Nc6!? 8.Nc3 Nf6 9.Ncd5 Nd4! 10.Qd3!? Bxd5 11.cxd5N Qe5 96
C321) 12.f4!? 98
C322) 12.g3 99

 
1.e4 e5 2.Nf3 Nf6 3.Nxe5
This move is the gateway to the main lines of the Petroff.
 
3...d6
Now 4.Nf3 is the usual move, and we will analyse it in Chapters 5-12.
The present chapter will deal with three significant sidelines: A) 4.Nxf7?!, B) 4.Nd3 and C)
4.Nc4.
 

A) 4.Nxf7?!
 
This is the Cochrane Gambit. Objectively it is dubious, but it may turn out to be a dangerous weapon
if Black doesn’t know how to handle it.
 
4...Kxf7
In return for the knight, White has picked up two pawns and lured the black king to a slightly more
exposed position. There are no immediate mating threats, so White should instead aim to gradually
gain space with moves like Nc3, d2-d4 and f2-f4. If White can achieve this set-up it will be difficult
for Black to find good posts for his pieces, as White’s pawns will control so many important squares.
Therefore a key principle of our opening strategy will be to prevent White from achieving his desired
central formation with pawns on d4, e4 and f4.
 
We will examine A1) 5.d4 and A2) 5.Nc3!?.
 
The other natural move has long been known to be inferior:
5.Bc4†?! d5! 6.Bb3!?
The text move is a bit more challenging than 6.exd5 Bd6 7.0-0 Rf8 when Black was clearly
better in Reinderman – Bosboom, Groningen 1995. Even though White has three pawns for the
piece, his doubled d-pawns are not so useful, and Black will enjoy full harmony after ...Kg8.
Now it is in Black’s best interest to provoke the capture on d5:

148
6...c5!N
Threatening ...c4.
6...Bd6 was not so convincing in Forsberg Conde – Xanco Compains, Barcelona 2011.
7.exd5
7.d3 will be met by 7...Bg4 8.f3 Be6µ and Black is clearly on top.
The pawn tension in the centre is now cleared and the bishop on b3 is blocked, so Black is free
to develop smoothly.
7...Qe8†!µ
This intermediate move prevents 0-0, and Black will follow up with ...Bd6 with an obvious
advantage.
 

A1) 5.d4

149
 This is White’s most popular try but we have an effective answer to it.
 
5...c5!
It is important to challenge the central pawn before White can establish the optimal set-up
described earlier.
 
Please also note that 5...Nxe4? should be avoided due to 6.Qh5† g6 7.Qd5†± and White picks up the
knight.
 
We will analyse A11) 6.dxc5 and A12) 6.Nc3.
 
Just as on the previous move, it is worse for White to play:
6.Bc4†?! d5 7.exd5 Bd6
Black has more than one good option, with 7...b5!? the engine favourite, but the text move is
simple and good.
8.0-0
8.dxc5?! Re8† 9.Be3 (9.Kf1 Bxc5–+) was played in Nagley – Glew, email 2000, when
9...Bg4!N would have developed the bishop with tempo, and after 10.Qd3 Bxc5µ Black has
clearly won the opening battle.

150
8...Nbd7N
8...cxd4 was decent in Mestrovic – J. Howell, Cannes 1995, but the text move seems slightly
more accurate. Black maintains the tension and guards the c5-pawn with the knight, ensuring
that the bishop will be able to stay on the blockading d6-square in the event of dxc5.
9.Nc3 Re8µ
With ...Kg8 coming next, White’s three pawns are clearly no match for Black’s extra piece.
 

A11) 6.dxc5 d5!

 
An important resource: Black prepares to recapture on c5 with the bishop, while closing the a2-g8
diagonal at the same time.
 
7.e5
7.exd5? Bxc5–+ was dismal for White in Wakeham – Niewold, corr. 2013, so the text move is his
only challenging idea.
 
7...Ng4 8.Bc4!?
This is an interesting try but it falls short after precise play by Black.
 
8.f4? weakens the kingside too much, and 8...Bxc5 was already winning for Black in Salem Elier –
Heldt, Willingen 2016.
 
8.h3 Nxe5 9.Qh5† Ng6 10.Be3 Be6µ (intending ...Qh4) left Black in control in Lemke – Welti, corr.
2017.

151
 
Finally, 8.Be2N is met by 8...Nxe5 9.b4 Nbc6µ and White does not have sufficient compensation.

8...Bxc5! 9.Qxd5†
9.Bxd5†? has not been played; Black simply replies 9...Ke8–+ when there is no convenient way to
defend the f2-point.
 
9...Qxd5 10.Bxd5† Be6! 11.Bxb7 Bxf2† 12.Ke2 Nd7 13.Bxa8

13...Bc4†!
Black has the advantage if he takes on a8 immediately, but it is even better to drive White’s king to
a worse position.

152
 
14.Kd1 Rxa8µ
Black’s precise play yielded a near-decisive advantage out of the opening in Spitz – Copar, corr.
2011.
 

A12) 6.Nc3

6...cxd4 7.Bc4†
I also checked: 7.Qxd4 Nc6 8.Qe3 (8.Bc4† d5! transposes to the note to White’s 8th move below)
8...Be6!N Preventing Bc4† and supporting ...d5. (the immediate 8...d5 was less accurate in
Kozhevnikov – Demidov, Novosibirsk 2002) 9.Be2 d5 10.exd5 Nxd5 11.Qf3† Kg8–+ Black should
be able to finish consolidating and win with his extra piece.
 
7...d5!
After 7...Be6?! 8.Bxe6† Kxe6 9.Qxd4© the awkward position of the king in the centre gave White
some compensation in Sulskis – Erenburg, Isle of Man 2005.
 
8.Nxd5
White also has a hard time after: 8.Qxd4 Nc6 9.Bxd5† Nxd5 10.Qxd5†

153
10...Qxd5N (this looks simpler to me than 10...Be6, as played in Barnett – Miller, Parsippany 2011)
11.Nxd5 (11.exd5? Nd4–+ wins immediately) 11...Nb4! 12.Nxb4 Bxb4†µ White’s three extra pawns
are not connected, and Black should eventually prevail thanks to his strong bishops.
 
8...Be6 9.Bg5
This is the best try, preventing Black from exchanging on d5.
9.0-0 occurred in Beveridge – Berna, corr. 2012, when Black could have played 9...Bxd5!N
10.exd5 Bd6µ, thematically blocking the d-pawn and intending ...Re8 followed by ...Kg8.
 
9.Qxd4N is a logical move to consider. White grabs the central pawn immediately but allows Black
to gain time after: 9...Nc6 10.Qd1 (10.Qe3 runs into 10...Bc5! 11.Qxc5 Nxe4 12.Qb5 Nd6–+ when
Black regains the piece with a won game)

154
10...Na5!µ After the bishop retreats Black will play ...Nxd5, with an obvious advantage.
 
9...Nc6 10.0-0 Be7
Black is doing well, as evidenced by the continuation of a correspondence game.

11.Nxe7
11.f4N is well met by 11...Bxd5 12.exd5 Na5!µ and the d5 pawn falls.
 
11...Qxe7 12.Bd3 h6 13.Bh4 g5 14.Bg3 Kg7µ
Black’s king is reasonably safe, and White was hard pressed to prove his compensation in Stanulis
– Dearlove, corr. 2002.

155
 

A2) 5.Nc3!?

This seems like the best move order for White.

 
5...c5
We should be ready to exchange on d4 as soon as White advances his pawn there. However, with
the knight already on c3, White can check on c4 without allowing the motif of ...d5 followed by
...Bd6, which we saw in so many of the lines above.
 
6.Bc4† Be6 7.Bxe6† Kxe6 8.d4!
White should open up the position in search of compensation.
 
8...Kf7
Black retreats the king from its exposed position.
 
8...cxd4?! transposes to the Sulskis – Erenburg game, referred to in the note to Black’s 7th move in
variation A12 above.
 
9.dxc5
Closing the centre with 9.d5?! can hardly be the best idea for White. The correct response is:

156
9...g6!N (rather than the passive 9...Be7?!, which gave White a chance to obtain good counterplay
against the king in Kaczmarek – Kuzmicz, Krakow 2016) 10.f4 Qe7 11.0-0 Nbd7 12.Re1 Bg7µ
Black is clearly in control.
 
9.0-0 should be met by 9...cxd4 10.Qxd4 and now 10...Qc8! is a nice detail to prevent a check on c4.
Best play seems to be:

11.Qa4 Intending to check on b3, but it is of no great concern. 11...Be7 12.Qb3† Qe6 13.Qxb7 Nbd7
White may have picked up a third pawn for the piece, but Black is now fully developed and ready for
action. 14.Qb3 Rhc8 15.Qxe6† Kxe6 16.Rd1 Rab8µ Black had all the chances in the endgame in
Alcala – Nichols, email 2011.

157
9...Nc6
9...Na6!? is another promising option which could be investigated.
 
10.Qe2!
This is the critical try, intending Qc4†.
 
10.cxd6? just helped Black to develop his pieces with 10...Bxd6 11.0-0 Re8µ in Hrabusa – Maros,
Slovakia 2009.
 
10.0-0?! is well met by 10...dxc5 11.Qe2 Qe8! (rather than 11...Qd7?, which would transpose to
11...dxc5?! as mentioned in the notes to the main line below) 12.Re1 Nd4 13.Qd3 Qe6µ and Black
was clearly better in Starke – Hudak, corr. 2009.
 
10...Qd7
Black prepares to insert his queen in response to Qc4†.
 
11.0-0
A notable game continued:
11.Be3
White prepares long castling but Black’s next move conveniently prevents it.
11...dxc5 12.f4
Sakaev points out a significant improvement for Black:
 

158
12...Nd4!N
The game continued: 12...Re8 13.e5 Ng4 14.Rd1 Qf5 15.0-0 h5!³ Establishing a firm grip on
the g4-square. White did not have quite enough compensation for the piece in Topalov –
Kramnik, Linares 1999. Nevertheless, Black should prefer the text move as it yields a larger
advantage.
13.Bxd4 cxd4 14.0-0-0 Rc8 15.e5
15.Kb1 Bc5 16.e5 Rhe8µ does not help White.
15...Rxc3! 16.bxc3 Ba3† 17.Kb1 Nd5 18.Ka1 Qc6
Sakaev stops here with the assessment that Black has an overwhelming advantage. This is
correct, but I will show a few more accurate moves:

19.e6† Ke7 20.Qd3 Ne3! 21.cxd4 Bb4!

159
 The knight cannot be captured due to the mating threats. White has nothing better than:
22.c4 Qxc4 23.Qxc4 Nxc4 24.Rc1 b5–+
Black should win the endgame with accurate technique.
 
Returning to our main line, White’s plan is to gain some initiative by advancing his f- and e-pawns,
but Black has more than one good way to continue.

11...Re8
This is the most straightforward route to an edge for Black, provided he follows it up correctly.
 
11...Qe6? allows White to show his idea: 12.f4! dxc5 13.f5+– with a winning initiative.
 
11...dxc5?! would also be a mistake due to 12.e5! Ng4 (12...Re8? is met by 13.Qc4†± when White
wins the piece back) 13.Ne4!© threatening e5-e6† or Ng5†. This gives White time to support the e5-
pawn with f2-f4, when Black’s king is under pressure and his forces are misplaced, all of which adds
up to strong compensation for White.
 
11...Rd8!?
This move is the favourite among correspondence players. The point is that sooner or later the
d-file will open up, when the rook will be usefully placed.
12.Bg5
12.f4? is no good here due to 12...d5! 13.e5 Bxc5†µ.

160
12...dxc5!
This is the safest continuation: Black gives back the piece to enter a comfortable endgame.
13.e5 Nd4 14.Qc4† Qe6 15.Qxe6† Nxe6 16.f4!
This trick enables White to equalize the material balance.
16...Nd4 17.exf6 gxf6 18.Bh4 Nxc2 19.Rad1 Be7

20.Nd5 b6 21.Bf2 Rhe8 22.g3 f5 23.Kg2 Rd7³

Only Black can be better here, as his knight is stronger than White’s bishop. That said, after
24.Nxe7 the players immediately agreed a draw in Taner – Bas, corr. 2012.
 
12.cxd6

161
 Black was ready to take on c5, so White had no real choice. Surprisingly, Black’s best move has
yet to be played:

12...Qxd6!N
12...Bxd6 13.f4© gave White decent play in Starostits – Seres, Rommelshausen 2002. The point of
taking with the queen is that White is prevented from expanding in the centre with f2-f4 and e4-e5.
 
13.Bg5
13.f4? can now be refuted by 13...Nxe4! 14.Nxe4 Qd4†.
 
13...Qe6³
Black intends ...Bb4 next, with a fine position.
 

B) 4.Nd3
 
This quirky-looking move is not even mentioned by Sakaev or Cohen in their respective repertoire
books on the Petroff. Indeed, it is only recently that it attracted the attention of many strong players:
before 2015 or so, there were virtually no GM games on the database with it. Since then, however, it
has been tested at the highest levels, with Carlsen using it in his 2018 World Championship match
against Caruana.
 
4...Nxe4 5.Qe2 Qe7
White has three main tries: B1) 6.b3, B2) 6.Nc3 and B3) 6.Nf4.
 

162
6.b4!?
This was played in Karklins – Ippolito, Philadelphia 2006. I suggest meeting it with:
6...Nc6N
For the moment we react in the same way as in variation B1 below, but on the next move a
different approach is needed.
7.Bb2 d5!
An important move, without which White would be able to create problems by kicking the
knight with b4-b5.
8.b5
8.a3 can be met by 8...Bf5 9.f3 Nd6 10.Qxe7† Nxe7³ when White faces a challenge in
relocating his pieces to optimal squares.

8...Nb4 9.Nxb4 Qxb4 10.Bc3 Qe7

Unpinning the knight.
11.Bb2 Be6
Black could repeat with 11...Qb4 but he has every reason to play for more than a draw.
12.Qe3 Qd7³
Black is ready to develop his kingside pieces, with slightly better chances due to his lead in
development.
 

B1) 6.b3

163
 This used to be the main move but it has proven to be completely harmless.
 
6...Nc6 7.Bb2 Bf5!
Preparing long castling.
 
8.Na3!?
This is the most interesting and ambitious continuation, avoiding the knight trade on c3 in order to
keep the game more complex. However, there is a risk that the knight on the rim will remain dim.
 
8.g4?! proves inaccurate after 8...Bd7 9.Bg2 0-0-0! when Black will regain the lost piece on e4 in the
event that White captures it. 10.0-0 Nc5µ In Glek – Ivanchuk, Pinsk 1986, White suffered from slow
development and a loose pawn on g4.
 
8.Nc3 is safe but harmless: 8...Nxc3 9.dxc3 (9.Bxc3N 9...0-0-0= is comfortable for Black) Now in
Karklins – Martinovsky, Chicago 1993, the simple 9...0-0-0N 10.0-0-0 g6 followed by ...Bh6 or
...Bg7 would have been completely fine.
 
8...0-0-0 9.0-0-0
9.Nc4? aims to reroute the knight to e3 but the problem is that it will not be able to stay there for
long: 9...d5! 10.Ne3 Bg6 11.0-0-0 d4 12.Nc4 b5! Kicking the knight back to the bad square that it
was trying to move away from. 13.Na3 a6µ The knight on a3 is misplaced and Black has a
significant space advantage.
 
9...d5!
Covering the c4-square and keeping the a3-knight out of the game.

164
 
10.g4 Bg6 11.f3 Nc5! 12.Nf4
12.Qxe7 Bxe7 13.Nf4 Ne6! essentially forces the exchange of knights, otherwise the black knight
on e6 would be too strong controlling the weak f4-square. 14.Nxe6 fxe6³ Followed by ...e5 with
better prospects for Black.
 
In Losev – Shevchenko, Moscow 1996, it would have been good for Black to proceed exactly as in
the note above:

12...Ne6!N
Challenging White’s knight and effectively forcing its exchange for the same reason as in the
previous note.
 
13.Nxe6 fxe6³
Followed by ...e5. White’s remaining knight is misplaced and his kingside pawn advances have left
some weaknesses.
 

B2) 6.Nc3

165
6...Nf6!?
6...Nxc3 7.dxc3 Qxe2† 8.Bxe2 Nc6 is quite playable as well.
 
7.Nf4
Threatening to jump to d5.
 
I also considered:
7.b3 Nc6 8.Bb2 Bf5
Now Black is ready to castle.
9.Nf4
Another possibility is: 9.0-0-0N 9...0-0-0 10.Qxe7 Bxe7 11.f3 White would like to play Re1
followed by Ne4. Also he has ideas like g2-g4 followed by Nf4. However, 11...d5! is a good
answer, covering the e4-square and intending ...d4. For instance, if 12.g4 Bg6 13.Nf4 then
13...d4 14.Nce2 d3„ offers Black good counterplay.

166
9...0-0-0
This is simple and good.
9...Nd4 may seem tempting but after 10.Qxe7† Bxe7 11.0-0-0 White has chances to seize the
initiative.
10.Qxe7 Nxe7 11.0-0-0 Ng6 12.Bd3 Bd7!?
White’s bishop is not so well placed on d3 so Black declines the exchange.
13.Nxg6 hxg6 14.Ne4 Nxe4 15.Bxe4 g5
The position was equal in So – Caruana, Saint Louis 2018.

7...Nc6!?
Black does not try to prevent White’s jump to d5 but rather creates his own to d4!

167
 
8.Nfd5 Qxe2† 9.Bxe2 Nxd5 10.Nxd5 Nd4
10...Kd7 was played in Uifelean – Chitescu, corr. 2011, but the text move is a rather thematic
response to the threat of Nxc7†. Now if White defends the c2-pawn, Black will play ...Ne6 followed
by ...c6 and ...d5.
 
11.Nxc7†
11.Bd1 Ne6 12.Bg4 is met by 12...Kd8= followed by ...c6, when Black is fine.
 
11.Bd3 Ne6 12.b3 c6 13.Ne3 d5= also gave Black no problems in Begliy – Duplenko, corr. 2010.
 
11...Kd8 12.Nxa8 Nxc2† 13.Kd1 Nxa1

14.d3
I checked two other ideas:
 
14.b3 Bf5 controls the c2-square for the knight. 15.d3 Be7 16.Bb2

168
16...Nxb3! 17.axb3 Kd7 18.Bxg7 Rxa8 The position is objectively equal but in a practical game I
would prefer Black due to his queenside pawn majority.
 
14.b4!? is similar to the line above, but this time Black will not be able to capture the b-pawn with
his knight. However, he has other resources: 14...Bf5 15.d3 Kd7 16.Bb2 Be7 (an interesting
alternative is 16...d5!? 17.b5 Bb4 18.Bxg7 Rxa8 19.Bxa1 a6!© when White’s pieces are
uncoordinated and Black will regain the pawn sooner or later) 17.Bxg7 Rxa8 18.Bxa1 a5 19.b5 a4©
Black has full compensation for the pawn, as White’s pieces are not in sync and his pawns on b5 and
a2 are targets.
 
14...Be6!
With the c2-square already in White’s firm control, there is no point in putting the bishop on f5, so
instead Black should target the a2-pawn.
 
15.Be3 Bxa2 16.Bxa7
Here I found a useful improvement.

169
16...d5!N
16...Be7 reached a curious situation of complete symmetry in Wei Yi – Xu Xiangyu, Daqing 2019.
Although this was not so bad for Black, I would prefer to develop the bishop actively on b4. Black is
doing fine, for example:
 
17.Bf3 Bb4 18.Bd4 Bb3† 19.Kc1 Ba4
First the bishop comes back, before the knight finally makes its escape from the corner.
 
20.Bxd5 Nb3† 21.Bxb3 Bxb3 22.Nb6 Re8©
With two strong bishops and the rook poised to invade on e2, Black has more than enough
compensation.
 

B3) 6.Nf4 Nc6

 
6...Nf6 used to be the main line, but the text move came into fashion when Caruana played it against
Carlsen in their World Championship match in 2018.
 
White has two main continuations: B31) 7.c3 and B32) 7.Nd5.
 
A third idea is:
7.Nc3N
This gives Black a choice.
7...Nd4!?
An interesting try to spice things up!

170
 If simple equality is more to your taste, then 7...Nxc3 8.dxc3 Qxe2† 9.Bxe2 Bf5= is completely
fine.
8.Qxe4 Qxe4† 9.Nxe4 Nxc2† 10.Kd1 Nxa1 11.b3
11.Bd3 can be met by 11...d5!÷ intending ...Bg4† and ...0-0-0, with a messy position.

11...c6 12.Bd3!
White should get his rook to e1 as soon as possible.
12.Bb2?! is inferior due to 12...d5 13.Ng3 Nxb3 14.axb3 f6³ when Black has better harmony
among his pieces than in the line below.
12...Be7
12...d5?! runs into 13.Ng5! followed by the annoying Re1†.
13.Bb2
13.Re1 0-0 14.Bb2 Nxb3 15.axb3 is the same thing.
13...Nxb3 14.axb3 0-0 15.Re1÷
Materially Black is doing well, but White’s pieces are active; with many pieces still on the board,
the position is dynamically balanced.
 

B31) 7.c3

171
 Nepomniachtchi has played this move a few times. White simply covers the d4 square, threatens
Nd5 and plans for a later d2-d4, building a strong centre and also in a way playing against the knight
on c6. However, the problem is that White’s plan is slow and Black can quickly complete
development without any problems.
 
7...Nf6
Preventing White’s knight jump looks simpler than 7...Nc5, which was Vidit’s choice against Nepo
at Wijk aan Zee 2019.
 
8.d4
This move seems most natural.
I also considered: 8.h4!? Bf5N (8...h5 was also reasonable in Nepomniachtchi – Wei Yi, Abidjan
[rapid] 2019) 9.Qxe7† Nxe7 10.f3 h5 Preventing g2-g4. 11.d4 g6= Followed by ...Bh6 and ...Kf8-
g7, when Black is doing fine.
 
8...g5!?
This is the most ambitious move, kicking the knight from its good post while gaining space on the
kingside.
8...Bd7N= followed by ...0-0-0 is a simple and safe alternative.

172
9.Qxe7†!?N
This seems a natural attempt to improve over 9.Nd3, when the knight is obviously not well placed,
although the move is not outright bad. 9...Rg8 (9...h6N is also decent) 10.Qxe7† (10.h4 h6=)
10...Nxe7 11.Nd2 This occurred in Obada – Zhuang, Mumbai 2019, when 11...Bd7N seems most
logical, to keep the e-file clear for the rooks after ...0-0-0. (The game continuation of 11...Be6 was
also quite playable though.)
 
9...Bxe7 10.Ne2
White intends Ng3 followed by Bd3 to conquer the f5-square, but it takes valuable time.
 
10...h6 11.Ng3 Be6 12.Bd3 0-0-0
Black is way ahead in development and can look to seize the initiative on the kingside with ...Rdg8
and ...h5-h4. Play may continue:
 
13.Nf5
Black should have no qualms about conceding the bishop pair with:

173
13...Rdg8! 14.Nxe7† Nxe7³
Black can exchange a pair of bishops with ...Bf5, and White is going to have a hard time on the
kingside even without queens.
 

B32) 7.Nd5

Just like in variation B2, we can meet this move by mirroring White’s threat.
 
7...Nd4! 8.Nxe7

174
 8.Qc4?? Nc3†! 9.Nxe7 Nxc2# is a pleasant fantasy line.
 
8...Nxe2 9.Nd5 Nd4 10.Bd3!?
This seems like White’s most interesting try, although it is not the only reasonable move.
10.Na3 is playable but harmless after: 10...Ne6 Intending ...c6 to kick the knight away, followed by
...d5. 11.f3 N4c5 12.d4 Nd7 13.c3 c6 14.Nf4 Nb6 15.Bd3 d5 16.Nc2 Bd6 17.Nxe6 Bxe6= Black
was doing fine in Carlsen – Caruana, London (6) 2018.
 
I also considered:
10.Nxc7†N 10...Kd8

11.Nxa8?!
This tempting move turns out to be a mistake.
11.Bd3! maintains a dynamic balance after: 11...Nxf2! 12.Kxf2 Kxc7 13.Nc3 (13.b3 Be6
14.Bb2 Nc6= followed by ...d5 and ...Bc5 is fine for Black; 13.c4 is met by 13...Be6,
threatening to open the position with ...d5, and after 14.Nc3 g6= followed by ...Bg7 Black is
doing well) 13...Be6 14.b3 d5 15.Bb2 Nc6= White still needs to arrange his pieces on optimal
squares.
11...Nxc2† 12.Ke2
12.Kd1?? loses to 12...Nxa1–+ when f2 hangs as well.
12...Nxa1 13.d3 Nc5
The position is symmetrical except for two details: Black’s knight is active on c5 compared to
its counterpart on b1, and his king is better placed on d8, covering the c7-square, than the king
on e2. In fact, White has nothing better than improving his king with:

175
14.Kd1 Be7 15.Nc3 Bf5
Intending ...Kd7.
16.Be3
16.Nd5 is met by 16...Nxd3 17.Be3 Nc5 18.b4 Be4³ and Black is at least slightly better; White
has to play precisely to avoid becoming seriously worse.
16...Kd7 17.d4 Rxa8 18.Bb5† Kd8 19.dxc5 Nc2³
White has to be careful not to end up in a clearly worse position.
 
10...Nc5 11.Nxc7†
This position was reached in Draskovic – Zlatin, Skopje 2019. I found an improvement, leading to
a more or less forced sequence of moves:

176
11...Kd7!N
Black placed his king on d8 in the game but the d7-square is more suitable here. The light-squared
bishop can get out via b7 and in some cases the rook may capture the knight on a8 if the opportunity
arises.
 
12.Nxa8 Nxd3† 13.cxd3 Nc2† 14.Kd1 Nxa1
White remains a pawn up, but of course the doubled d-pawns bring no immediate joy for him and
Black’s bishop pair must count for something.
 
15.b4 b5 16.Bb2 Bb7 17.Rg1
17.Bxa1 is of course met by the intermediate 17...Bxg2! followed by capturing the knight on a8,
when Black is fine.

177
17...d5!
This precise move hits the b4-pawn.
17...Bxa8 18.Bxa1² gives White a slightly more pleasant position, with Bd4 followed by Nc3 a
natural plan. The problem for Black is that his bishops don’t have much scope now: his d-pawn will
always get in the way of one of them.
 
18.a3
18.Na3 is met by: 18...Bxb4 19.Nxb5 Rxa8 20.Bxa1 Bc6! Activating the bishop! 21.Nd4 Ba4†©
Black’s bishops are excellent. Compared with the previous note, the disappearance of the b-pawns
has clearly improved Black’s prospects.
 
18...Nb3 19.Kc2

178
19...Nc5!
Forcing a change in the pawn structure.
 
20.Nc7 Kxc7 21.bxc5 Bxc5 22.Bxg7 Re8©
The unbalanced pawn structure makes Black’s bishops more powerful. Now Black has ideas like
...a5 and ...b4, and his prospects are good.
 

C) 4.Nc4 Nxe4
 
We have reached the third and final branch of the chapter. The path immediately divides again, into
three main options for White: C1) 5.d4, C2) 5.Nc3 and C3) 5.Qe2!.
 

C1) 5.d4
 
This has been played in quite a few older games but has been completely out of the limelight in
recent years.
 
5...d5 6.Ne3
Black has a number of ways to handle the position, but I prefer:
 
6...Qf6!?

179
 The idea is to lure White’s queen to e2, where it obstructs the bishop on f1, practically forcing it to
develop to g2, which will cost more time.

7.Qe2
We must also consider:
7.Bb5†?!
This move is connected with a piece sacrifice. It is dubious, but Black needs to play some
accurate moves to prove it.
7...c6 8.0-0
8.Qe2?! is met by 8...Be6 9.Nxd5 Bxd5 10.Bd3 Qxd4µ and now 11.0-0? was played in
Fontaine – Milliet, Port Barcares 2005, when 11...f5!N would have left Black simply a piece up.
8...cxb5 9.Nxd5 Qd6 10.Re1 Qxd5 11.Nc3 Qd8

180
12.Qf3!
This is White’s trickiest option.
12.Bg5? Qxg5 13.Nxe4 has not been played and it doesn’t work, but it might looks scary over
the board so it is worth showing the refutation: 13...Qg6 14.Nf6† Kd8 15.Re8† Kc7 16.Nd5†
Kd7 White’s initiative is not enough to make up for the pair of sacrificed pieces. 17.Qe2 Qd6
18.Qg4† f5 19.Qxf5† Kxe8 20.Qxc8† Kf7 21.Qxb7† Now Black can choose between the
flashy 21...Nd7 22.Qxa8 Qxh2†, leading to a queen exchange and a winning endgame, and
21...Qd7 22.Qxa8 Nc6–+ which the engine likes even more.
The alternative is: 12.Rxe4† Be7 13.Bg5 Nc6 14.Bxe7 (14.d5? loses to 14...0-0; 14.Qe2
occurred in Smagin – Makarichev, Moscow 1987, when Black should have played 14...f6!
15.d5 Ne5 16.Rd1 0-0 17.d6 Bxd6 18.Nxb5 Bg4µ, keeping a material advantage) 14...Nxe7
15.Qe2 a6 16.Re1 Be6 17.Rxe6 (17.d5 Nxd5 18.Rd1 0-0 19.Nxd5 Bxd5 20.Re5 Qa5 21.Rexd5
Qxa2µ leaves Black a clear pawn up, as Sakaev points out) 17...fxe6 18.Qxe6 Qd7 19.Nd5
Qxe6 20.Rxe6 0-0-0!µ White’s attack failed in Lipecki – Cilloniz Razzeto, corr. 2015.
12...f5!
12...Be7 is met by 13.Qxe4© when it becomes difficult for Black to complete development.
13.Nxe4 fxe4 14.Rxe4† Be7 15.Bg5 Nc6 16.Rae1

181
16...Rf8!
The rook comes to the defence.
17.Qh5† Rf7 18.Bxe7 Nxe7 19.Rxe7† Qxe7 20.Rxe7† Kxe7³
In this materially imbalanced endgame, Black had more than enough pieces for the queen in
Lipecki – Salvador Marques, corr. 2015.
 
7...Be6
Now Black threatens to take on d4, so White should react accordingly.
 
8.c3 Nd7 9.Nd2
9.Nxd5?! Bxd5 10.f3 does not help White due to 10...0-0-0 11.fxe4 Qh4† when Black regains the
pawn with a significant lead in development.
 
9.g3 is met by 9...c6, supporting the d5-pawn so that the knight can come back to d6. 10.Bg2 Nd6
(10...Qg6 was also not bad in Savchenko – Frolyanov, Ulan Ude 2009) 11.0-0 Be7 Black is doing
fine. For example:

182
12.Nd2 Nf5 13.Nf3 Nxe3 14.Bxe3 h6 15.Rae1 0-0= Evtushenko – Ulasevich, corr. 2013.
 
9...c6
Black plays a generally useful move while waiting for White to make a decision.
 
10.Nxe4
This is the most ambitious and interesting continuation.
 
10.g3 gives Black an easy game after 10...Nxd2 11.Bxd2 Bd6 12.Bg2 0-0 13.0-0 Qg6³ as seen in
Vasquez Nigro – Ramponelli, corr. 2010.
 
10...dxe4 11.g3 Qg6
Moving the queen gives Black the option of supporting the e4-pawn with ...f5; and ...h5-h4 is
another attractive plan.
 
12.Bg2
12.Ng2 Bd6 13.Nh4 Qf6 14.Qxe4 0-0-0© gives Black promising compensation.

183
12...h5!N
12...f5 was played in Yegiazarian – Mamedyarov, Tbilisi 2001, but the text move is more to the
point. Black intends ...h4, gaining space on the kingside.
 
13.f3
13.h4 is met by 13...Bd6³ and Black is doing well.
 
13...exf3 14.Bxf3 0-0-0=
Black has comfortable play, as White’s pieces are still uncoordinated.

C2) 5.Nc3
 
This is the most popular move but it is not dangerous for Black.
 
5...Nxc3 6.bxc3
This is the usual choice.
 
The alternative is not too challenging:
6.dxc3 d5 7.Ne3 c6
With a couple of simple moves, Black gains space in the centre and will look forward to
developing his pieces.
8.Qd4
White prevents the dark-squared bishop from developing for the time being, and also intends to
break with c3-c4 at the right moment.

184
8...Be6!
It is useful to have ...c5 available in case White breaks with c3-c4.
8...Nd7 9.c4! Qf6 10.Qxf6 Nxf6 11.cxd5 cxd5 was also equal in Mamedov – Kuzubov, Lvov
2017, but I would prefer not to adopt an IQP in the endgame without a good reason.

9.Bd2!?N
9.c4?! is convincingly met by 9...c5! 10.Qd3 d4³.
9.f4 f6! 10.c4 was seen in Sulskis – Zulfugarli, Bydgoszcz 1999, when 10...Nd7!N would have
prepared ...Bc5 to good effect; for instance, 11.cxd5 Bc5³ and White’s king will face some
problems.
9.Be2 Nd7 10.c4 Qf6 is a typical way to trade queens, and after 11.Qxf6 Nxf6 12.cxd5 Nxd5=
Black has no problems.
The text move seems like an interesting try but Black is fine after:
9...Qb6! 10.0-0-0 Nd7=
Followed by ...0-0-0.

185
6...g6 7.d4 Bg7
The bishop is well placed on g7 in this structure. Black can consider the ...c5 break, putting
pressure on the d4- and c3-pawns. Also, with the bishop on g7 Black’s king will be safer after
castling.
 
8.Bd3 0-0 9.0-0
Too optimistic is:
9.h4?! c5!
Immediately responding to a flank attack with a counter in the centre!
9...Re8† was slightly less accurate in Inarkiev – Chadaev, Olginka 2011.
10.Bg5

186
 Now in Simacek – Saric, Sibenik 2012, Black should have continued:
10...Qc7!N
It is difficult for White to hold his centre together. Here are some sample variations:
11.Bf4
11.h5 cxd4 12.hxg6 fxg6 13.cxd4 Bxd4µ and Black is simply a pawn up.
11.Ne3 cxd4 12.Nd5 Re8† 13.Kf1 Qc5µ leaves White’s pieces uncoordinated.
11...cxd4 12.0-0 Rd8 13.cxd4 Bxd4³
Black’s extra pawn makes him the slight favourite, although he still needs to complete his
development.

9...Nd7

187
 The knight is heading for f6.
 
10.Bf4
10.f4?! left the e4-square weak after 10...f5! 11.Qf3 Nf6³ in Asplund – Cederlind, Sweden 1979.
 
10.Qf3 Rb8! is a nice move, preparing another fianchetto. After 11.a4 b6 12.Ne3 Nf6 13.Nd5 Nxd5
14.Qxd5 Bb7³ Black’s light-squared bishop was well placed and White remained with a slightly
worse pawn structure in Vavulin – Fridman, Riga 2014.
 
10...Nf6 11.a4 Re8 12.a5 Bd7!
Black is focused on putting his pieces on good squares.

13.Bg5 h6 14.Bh4 g5!

The slight weakening of the kingside is of no consequence here.
 
15.Bg3 Ne4! 16.Bxe4 Rxe4 17.Qd3 Qe8³
With the bishop pair, a better pawn structure and control over the e-file, Black was at least slightly
better in Dubuc – Noble, corr. 2016.
 

C3) 5.Qe2!

188
 This is the critical try: White wants to disturb the coordination of our forces.
 
5...Qe7 6.Ne3
6.d3 Nf6 7.Bg5 Qxe2† (7...Nc6!? is an interesting alternative, not fearing the possibility of
doubled f-pawns) 8.Bxe2 Be7= resulted in a completely harmless endgame in Dreev – Malaniuk,
Odessa 1989.
 
White can also establish his central formation with:
6.d4 Nc6
It makes sense to cover the e5-square before playing ...d5, so that White does not have the
option plonking his knight on e5.
7.c3
7.d5?? Nd4 is obviously not an option for White.
7...d5 8.Ne3 Be6 9.Nd2
In Jansa – Kalod, Brno 2006, Black exchanged on d2. Instead, I think it is best not to react to
the challenge to the knight and instead simply play:

189
9...0-0-0N 10.Nxe4
White should trade off the active knight on e4.
10...dxe4 11.Qb5!?
Threatening d4-d5.
In the event of 11.Qc2 f5 Black intends ...Qf7 followed by advancing on the kingside. White
has nothing better than 12.Bc4 Bxc4 13.Nxc4 Kb8³ when Black has good chances to gain the
initiative on the kingside.
11...Bd7 12.Nd5
12.Qb3 f5ƒ gives Black good chances on the kingside.
12...Qe8 13.Bf4
White’s aggressive play can be conveniently met by:

190
13...Nb4! 14.Qa5 Nxd5 15.Qxd5 Kb8=
With a fine position for Black.
 
6...Be6
6...Nf6 is another good option but it feels more natural to develop a new piece rather than move
one for the second time. Now Black would like to play ...Nc6 next, with ideas of ...Nd4, as well as
the simple ...0-0-0.
 
We will analyse C31) 7.Nc3 and C32) 7.c4!?.
 
7.g3 is met by 7...Nc6! 8.Bg2 d5. Then 9.c3 is necessary to prevent ...Nd4 ideas, and 9...0-0-0 10.d3
Nf6 11.d4 Qd7³ gave Black a useful lead in development in Vrban – Mrkonjic, Mali Losinj 2017.
 

C31) 7.Nc3 Nf6!?

 
I prefer this over exchanging on c3 because once Black gets in ...d5 and ...c6, the knight on c3 will
have poor mobility.
 
8.g3
I also checked 8.d4 d5 9.Qd1 Qd7 10.Bb5 c6 11.Bd3 as played in Iskandarov – Koksal, Mersin
2018, when Black missed a chance to seize the initiative with:

11...c5!N Black should attack the centre immediately before White gets time to arrange his pieces
comfortably. 12.Bb5 Nc6 13.Ne2 a6 14.Bxc6 bxc6³ The bishop pair combined with the central
pressure gives Black the upper hand.
 

191
After the text move, Black should take the opportunity to establish a desirable pawn structure in the
centre.

8...d5 9.Bg2 c6
Black’s pawn chain limits the activity of all three of the minor pieces that White has developed.
 
10.d4 g6 11.0-0 Bg7 12.f4 Na6!
It is important to remember this typical way of developing the b8-knight. Now Black is ready to
shelter his king on the queenside, and White’s early kingside aggression is liable to backfire.
 
13.a3 Nc7 14.Bd2
14.f5 is no big deal, as Black can simply retract the bishop with 14...Bd7³.
 
14...0-0-0
Black has completed development and stands well.
 
15.Rae1
Black continued with 15...Rde8 in Kovalev – A. Smirnov, Batumi (ol) 2018, but it would have
more accurate to play:

192
15...Qd7!N
With the following justification:
 
16.f5?!
White should settle for 16.Qd3, but then 16...Bh3³ gives Black a pleasant edge, as the light squares
in White’s camp will be weak.
 
The text move looks like a logical pawn sac but it only worsens White’s game.
 
16...gxf5 17.Bh3
17.Qf3? Ne4 is winning for Black.
Now Black can make full use of the knight on c7 with:
 
17...f4! 18.Bxe6 Nxe6 19.Nf5 Rhe8µ
Black is significantly better.
 

C32) 7.c4!?

193
7...Nc6!?
This results in a dynamic game where Black also gets his chances to fight for the initiative.
 
8.Nc3
8.Nd5?? Bxd5 9.cxd5 Nd4–+ would be embarrassing for White.
 
Trying to prepare d2-d4 with 8.Nc2? is too slow: 8...d5! 9.cxd5 (9.d4 0-0-0µ leaves White massively
behind in development) 9...Bxd5 10.Nc3 0-0-0!µ White will pay the price for wasting so much time.
 
8...Nf6
8...Nc5? does not work due to the cool move 9.Qd1!, threatening d2-d4. 9...Nd4 (9...Qf6 10.Be2
0-0-0 11.0-0± leaves Black passively placed and a knight is coming to d5) The text move allows
another pawn advance: 10.b4! Nd7 11.Bb2± White’s pieces are excellently placed while Black’s lack
harmony.
 
9.Ncd5
Once again, 9.Nc2?! is inadvisable. 9...Bf5! prevents White’s intended d2-d4 advance, and after
10.Qxe7† Bxe7 11.Ne3 Bg6³ the bishop is well placed on the b1-h7 diagonal while the knight is
ready to jump to d4.
 
9.Qd1?! is not much good either: 9...0-0-0 10.Be2 (10.d4 is met by 10...d5 11.c5 and now the key
move is 11...Ne4!µ, when White’s position comes under serious pressure) 10...Nd4 11.0-0 d5µ
White’s pieces are uncoordinated and his queenside is still undeveloped.
 
9...Nd4!

194
 9...Qd8?! was played in a recent game but 10.b4!N would have been an excellent reply (improving
on 10.b3, as played in Sethuraman – Abasov, Astana 2019). By advancing the pawn two squares
rather than one, White not only prepares to develop his bishop to b2 but also gives himself the option
of kicking the knight with b4-b5. 10...Nd4 (after 10...g6 11.Bb2 Bg7 12.b5± White’s pieces are
optimally placed and he controls more space) 11.Qd3 c5

12.Rb1! (12.Bb2 appears logical but 12...g6! is a good reply, since 13.Bxd4 cxd4 14.Qxd4 Bg7©
leaves White with problems on the long diagonal) 12...g6!? (12...b6 13.Bb2²) 13.bxc5 dxc5 14.Rxb7
Nxd5 15.cxd5 Black cannot regain the d5-pawn (15...Bxd5?? being refuted by 16.Qxd4!+–) so he is
under some pressure.
 
10.Qd3!?
Keeping the queens on leads to a fighting game.
 
The less ambitious alternative is:
10.Nxe7
This should lead to an equal endgame after a series of forced exchanges.
10...Nxe2 11.N7d5 Bxd5 12.cxd5
12.Nxd5?! Nxd5 13.Bxe2 enables Black to seize the initiative with 13...Nb4! 14.Kd1 g6 15.d4
Bg7 16.Be3 d5!³ when White faces some problems connected with the poor position of his
king.
12.Bxe2 is simply met by 12...Be6 13.d4 d5. It is worth adding that in response to 14.c5, Black
should develop his bishop in thematic fashion against the d4-c5 pawn wedge with 14...g6! 15.0-
0 Bg7, with a fine position.

195
12...Nxc1 13.Bb5†
White may as well disturb the black king before recapturing on c1.
13.Rxc1 0-0-0= makes life simpler for Black.
13...Kd8 14.Rxc1 g6 15.0-0 Bh6 16.Rc2 a6 17.Be2 Re8
The rook goes to e7 to defend the c7-pawn.
18.Rfc1 Re7 19.Bf3 a5=
With opposite-coloured bishops on the board and a compromised pawn structure, White can hardly
hope for more than a draw.

10...Bxd5 11.cxd5N
White should continue trying to complicate things and to seize the initiative.

196
 11.Qxd4?! is inconsistent and 11...c5!³ gave Black a good position in Fluvia Poyatos – Mellado
Trivino, Catalonia 2009, with ...Bc6, ...g6 and ...Bg7 the obvious plan.
 
11...Qe5
Threatening ...Nxd5. We have reached a fascinating, irregular and unexplored position, full of
possibilities and challenges for both sides.
White has two interesting continuations: C321) 12.f4!? and C322) 12.g3.
 

C321) 12.f4!? Qxf4

An interesting idea! White sacrifices a pawn to lure the queen away from the e5-square, in order to
develop his bishop to b2.
 
13.b4
13.b3 is also met by 13...c5, supporting the d4-knight. My analysis continues: 14.dxc6 Nxc6
15.Bb2

197
15...d5! A nice move, enabling the bishop to develop freely on c5. 16.g3 Qe4 17.Qxe4† Nxe4
18.Nxd5 (18.Bg2 can be met by 18...0-0-0 19.Nxd5 f5= when Black is doing fine and the d2-pawn
will come under pressure) 18...0-0-0= Material is equal and White has the bishop pair, but his king is
still exposed on e1 and his pieces are uncoordinated.
 
13...c5!
It is logical to protect the knight on d4.
 
13...Qe5!? is also playable but it is somewhat riskier in view of: 14.Bb2 Nf3† 15.gxf3 Qxb2 16.Rb1
Qxa2 17.Rc1© White is two pawns down but he has strong compensation for them, because the black
king may find it difficult find a safe haven.
 
14.dxc6
White has to open the position if he is to find meaningful compensation for the pawn.
 
14...Nxc6

198
15.Bb2!
White values time before material.
 
Saving the pawn with 15.b5? is useless due to 15...Ne5µ and Black seizes the initiative.
 
15.g3!? Qe5 16.Rb1 is playable. Black should take action in centre with 16...d5! 17.Bg2 d4 18.Bb2
Nxb4© when White has some compensation, but proving it will not be easy.
 
15...Be7 16.g3 Qxb4 17.Rb1 Qa4©
The position remains unclear. White surely has compensation for the two-pawn deficit, but Black
also has his chances.
 

C322) 12.g3

199
 White defends the d5-pawn indirectly, by threatening f2-f4. The text move also prepares to develop
the bishop via g2.
 
12...Ne4!
12...Nxd5? is obviously bad due to 13.f4± and Black will have to part with a knight.
 
12...g6!? is a playable alternative, but I prefer the idea of transferring the knight to c5.
 
13.Bg2
Trying to use tactics with 13.Nc4?! does not work so well, due to 13...Qxd5 14.Bg2 and now
Black has the strong 14...Nc5! 15.Qe3† Qe6³ with some advantage.
 
13.f4 leads to a repetition after: 13...Nf3† 14.Ke2 (14.Kd1? Nf2†µ is bad for White) 14...Nd4†=
 
13...Nc5 14.Qc3

200
14...a5!
Stabilizing the c5-knight. Next Black would like to play ...g5, securing the e5-square for the queen
and also intending ...Bg7. Hence White’s next move is forced.
 
15.f4 Qf6 16.Kf2!
This is the critical move, threatening Ng4.
The immediate 16.Ng4 can be met by 16...Qe7† when the knight must go back: 17.Ne3 Qf6=
 
16.b3?! is met by 16...g6 17.Bb2 Bg7³ when Black is the one who is playing for the initiative.

16...h5!

201
 Covering the g4-square.
 
17.b3!
It is important to bring the bishop to b2 in order to put pressure on the d4-knight.
 
17.Re1
This seems like a logical follow-up to White’s previous move, but it is slightly less accurate
than the main line.
17...Be7 18.Nc2
18.b3? does not work here because of 18...0-0 19.Bb2 and now 19...Qg6! is a nice trick, since
20.Qxd4 is now met by 20...Bf6µ.
18.Kf1 is no better than the text move as White will have to exchange the strong knight sooner
or later. For instance, 18...0-0 19.Nc2 Nxc2 20.Qxc2 Rfe8³ and White’s forces are still not
coordinated.
18...Nxc2 19.Qxc2 0-0ƒ
White has to be careful not to become worse, as his king is less safe.
 
17...g5!?
This is the most active choice.
 
17...Be7!? is also playable, as 18.Bb2 Qg6 19.Qxd4 Bf6 20.Qc4 Bxb2 regains the piece.
 
18.Bb2
18.Re1? is met by 18...0-0-0!‚ with a dangerous attack.
 
18...Bg7 19.Nc2! Nxc2 20.Qxf6 Bxf6 21.Bxf6 0-0!
The smoke has more or less cleared, so let us take stock of the position. White has the bishop pair
but Black’s knights are quite annoying for him. There is no time to capture on g5 because the a1-rook
is attacked, and it cannot move to c1 because of the fork on d3. Black wants to play ...Rae8 next,
when White will have to watch out for knights jumps to e4 and d3.

202
22.Raf1
22.Rad1 can simply be met by 22...gxf4 23.gxf4 Rae8= intending ...Ne4† or ...Nd3†.
 
22...Rae8!
This is stronger than 22...gxf4, as it offers White chances to go wrong.
 
23.fxg5!
The tempting 23.Bxg5?? proves to be a mistake after: 23...f6 24.Bh4 (24.Bh6 is met by 24...Rf7
25.f5 Nd3† 26.Kg1 Kh7 27.Bf4 Re2–+ and Black’s pieces are completely dominating) 24...Nd3†
25.Kg1 Nd4µ Just compare the out-of-play bishop on h4 with the mighty knights on d3 and d4!
 
23.Kg1?! is also inaccurate in view of 23...Re2! 24.fxg5 Rxd2 25.Bf3 Ne3³ and Black’s pieces are
too active.
 
23...Nd3† 24.Kg1

203
24...Nce1!„
An excellent move! Black simply wants to eliminate the g2-bishop and invade on the second rank
with ...Re2(†). The position is dynamically balanced, with everything still to play for.
 

Conclusion
 
This chapter has dealt with three important sidelines after 3.Nxe5 d6. Beginning with the Cochrane
Gambit, 4.Nxf7?! Kxf7, White’s piece sac is not fully correct but that does not mean it should be
taken lightly. Both 5.d4 and 5.Nc3!? should be met by 5...c5!, preventing White from establishing his
desired pawn centre, with good chances for Black to capitalize on his material advantage after further
accurate play.
 
4.Nd3 Nxe4 5.Qe2 Qe7 looks slightly odd from White’s perspective, but it can lead to lively play.
Of particular interest are the lines where White continues Nc3-d5 or Nf4-d5, where Black responds
with a mirror attack with ...Nd4.
 
Finally, 4.Nc4 Nxe4 is the most frequently occurring of the three main options in this chapter. Then
5.d4 is sensible but not too threatening, while 5.Nc3 requires you to pay attention to the subtleties of
the queenless middlegame. 5.Qe2! Qe7 6.Ne3 is the other main option, when 6...Be6 is our choice,
with 7.c4!? Nc6!? 8.Nc3 Nf6 9.Ncd5 Nd4! the critical continuation leading to interesting play for
both sides. Most of my analysis in this section is original, so you will have excellent chances to
outprepare an opponent who goes down this path.
 

204
A) 5.Bd3 d5 103
A1) 6.Qe2 103
A2) 6.0-0 Be7 7.Re1 Nd6 8.Nc3 c6 106
A21) 9.Qe2 107
A22) 9.b3!? 109
B) 5.c4 Nc6 111
B1) 6.Be2 111
B2) 6.d4 d5 112
B21) 7.Nc3?! 113
B22) 7.a3 114
B3) 6.Nc3 115
B4) 6.d3 Nf6 7.d4 d5 118
B41) 8.Nc3 119
B42) 8.c5!? 120

205
1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4
This chapter will deal with two interesting sidelines: A) 5.Bd3 and B) 5.c4.
 

A) 5.Bd3
 
The point of this move is to prepare to castle as quickly as possible. There is an obvious issue that the
bishop blocks the d-pawn, but White intends to solve the problem of queenside development with
either c2-c3, Bc2 and d2-d4, or b2-b3 followed by developing the dark-squared bishop to b2 or a3.
However, these ideas take time.
 
5...d5
5...Nf6 is a perfectly playable alternative but I like the text move, unlike Sakaev and Cohen, both
of whom regard it as inferior.
 
White’s two main options are A1) 6.Qe2 and A2) 6.0-0.
 
6.c3?! creates a cheap threat of Bxe4 followed by Qa4† but 6...Nd7! is a convenient reply,
preventing White’s idea while developing a piece. 7.0-0 was seen in Morales – Otalora, Medellin
2010, when 7...Bd6!N would have been the most active choice, followed by ...0-0 with a fine
position. If White tries to win a pawn with 8.Re1 0-0 9.Bxe4?! dxe4 10.Rxe4 then 10...Nc5!µ leaves
him sorely missing the light-squared bishop, and his queenside development is zero.
 
6.Nc3 challenges the knight immediately. Black should respond with 6...Nc5 hitting the bishop and
threatening ...d4. White needs a novelty to avoid becoming worse: 7.Bb5†!N (7.0-0 is simply met by
7...Nxd3 8.Qe2† Be7 9.Qxd3 c6 10.Re1 0-0³ as in Konev – Tydykov, Kemerovo 2007; and 7.Qe2†
Be7 8.0-0 Nxd3 transposes) 7...c6

206
8.d4! Without this intermediate move, White would stand worse. 8...Ne6 9.Bd3 Bd6 10.Ne2 0-0
11.0-0 Nd7 12.c3 Qc7= Both sides are well developed and the position is equal. Black’s further plans
include ...Nd7-f6 and ...Bc8-d7 with a good game.
 

A1) 6.Qe2

White hopes to profit from luring our queen to a suboptimal square.

 
6...Qe7 7.0-0 Nd6

207
 Black retreats the knight before White can develop any unpleasant pressure along the e-file. One
important idea to remember in this variation is that, in the event of a queen exchange, Black has a
strong plan of ...f6 followed by ...Kf7.
 
8.Re1
White can also refuse to exchange queens with:
8.Qd1
Threatening Re1. The response here is easy to remember: just copy White’s move!

8...Qd8!
The queen retreats to safety while making room for ...Be7.
8...Qf6 is slightly inferior due to 9.Nc3! when neither of the following options is great for
Black:
a) 9...Be6 10.Re1 Be7 11.b4! A multi-purpose move, preventing any ...c5 ideas while preparing
to develop the bishop. 11...0-0 12.Bb2 Threatening Nd5. 12...Re8 13.Rb1ƒ White has some
initiative, as Black will need some time to bring his queenside pieces into the game.
b) 9...c6 should be met by 10.b3!, for instance: 10...Be7N (or 10...Bg4, Benjamin – Villanueva,
Internet [rapid] 2017, when 11.Qe1†!N immediately escapes the pin, and after 11...Be7 12.Bb2
we transpose) 11.Bb2 Preventing castling. 11...Bg4 (11...0-0? 12.Nxd5+–; 11...Qh6 12.Re1 Be6
13.Nd4± with unpleasant pressure along the e-file) 12.Qe1! Unpinning while threatening
Nxd5. 12...Qh6 (or 12...Qe6 13.Nd4± and Black will still suffer on the e-file) 13.Nd4± Black’s
king is uncomfortable in the centre.
9.Nc3
9.Re1† Be7 transposes to variation A2 below.
9...c6 10.Re1† Be7
We have transposed to variation A2 below, with two extra moves having been played.
 
208
8.Nc3
In Hovhannisyan – Ter Sahakyan, Yerevan 2019, Black should have immediately exchanged
queens with:
8...Qxe2N 9.Nxe2
9.Bxe2 c6 10.Re1 (10.b3 f6= is similar) 10...Be7 transposes to our main line below.

9...f6!
Again we go for our thematic set-up with the king on f7.
9...Bf5? may seem desirable, but this is not the right time to trade bishops. 10.Bxf5 Nxf5
11.Re1 Be7 12.Nf4 c6 13.g4!± White seizes the initiative as it is not easy for Black to get his
king to safety.
10.Nf4 c6=
Followed by ...Kf7. With the king taken care of, Black will focus on developing his queenside
pieces next.
 
8...Qxe2 9.Bxe2
White may also try:
9.Rxe2† Be7 10.Nc3
10.b3 is met by 10...Be6 intending to castle. If White tries to prevent it with 11.Nc3 then
11...Bf6!= enables ...0-0 next, with a fine position.
10...Be6 11.Nd4
11.b4 covers the c5-square while preparing Bb2 or Ba3, but 11...Bf6! is a good reply, pinning
the knight and preparing to castle (the direct 11...0-0? doesn’t work due to 12.Nxd5±). 12.Rb1
0-0= Having successfully solved the main issue of getting his king to safety, Black was
completely fine in Grischuk – Fedoseev, Riyadh (rapid) 2017.
11...Kd7 12.Nxe6 fxe6

209
 White has the bishop pair but his pieces are not well placed, and he still needs some moves to
develop his queenside.
13.Nb5
13.Nd1 was convincingly met by 13...c5 14.c3 Nc6 15.Bc2 b5!³ in Shirov – P.H. Nielsen,
Drammen 2004. Black has a lot of space on the queenside and his pieces are optimally placed.
13...Nxb5 14.Bxb5† c6 15.Bd3 Bf6=
Black had no problems in Leko – Kramnik, Linares 1999.
 
9...Be7 10.Nc3 c6 11.b3

11...f6!N
It is useful to remember this method of improving the king in these positions.
 
11...0-0?! was played in Godena – Mchedlishvili, Calvia 2007, when 12.Bd3!N would have been
most precise. Following 12...Bd8 13.Ba3 Bc7 14.Re7 Bd7² Black can slowly untangle himself with
careful play, but I would prefer to avoid such a situation.
 
12.Ba3 Kf7 13.Bd3 g6
Black is doing fine, for example:
 
14.Nd4!?
Creating some tactical possibilities such as Rxe7† or Ndb5.
 
14...Rd8!
Avoiding the aforementioned tactics. Now Black can think about ...b6 followed by ...c5, or ...f5
followed by ...Bf6. White’s most logical idea is to play along the e-file with:

210
 
15.Re2 f5 16.Nf3
I also checked 16.Rae1 Bf6 17.Nf3 Ne4 18.Bb2 Nd7= when Black has successfully blocked any
activity along the e-file.
 
16...Bf6 17.Ne5† Kg7 18.f3
Among other ideas, Black could consider expanding in the centre with:

18...d4!? 19.Nd1 b6 20.Nf2 c5=

Having gained space, Black will develop his queenside pieces with a good game.
 

A2) 6.0-0

211
 Keeping the queens on the board also fails to cause Black any serious problems.
 
6...Be7 7.Re1 Nd6 8.Nc3
White has tried two other options:
 
8.c3
This has been tried by a few strong players but is pretty toothless.
8...0-0 9.Bc2 c6 10.d4 Bf5!
Whenever possible in this pawn structure, Black should aim to trade the light-squared bishops.
11.Bf4
11.Bxf5 Nxf5 12.Qb3 Qc7 13.Qc2 g6 14.Nbd2 Re8 15.Nf1 Nd7 16.Ng3 Nxg3 17.hxg3 Nf6
left White without a trace of an advantage in Gunina – Harika, Saint Louis 2019.
11...Bxc2 12.Qxc2 Na6!?
Caruana shows us another fine way of developing the queen’s knight.
13.Nbd2 Nc7 14.Re2
White tries to create some pressure along the e-file but Black can easily nullify it.
14...Re8 15.Rae1 Bf8 16.Ne5

212
16...Qf6!
Black attacks the bishop and gets ready to trade queens.
17.Bg3 Qf5 18.Qxf5 Nxf5=
Black had no problems in So – Caruana, Saint Louis 2018.
 
8.Qe2
The point of this move is to make it harder for Black to castle, but the problem is easily solved.
8...Nc6
Black develops while defending e7.
Cohen covers this line briefly in a note, but only mentions the inferior 8...Bg4?!.
9.Nc3
9.Ne5 was an attempt to disturb the harmony in Black’s camp in Lopez Rodriguez – Adell
Corts, Aragon 1995. A good answer is 9...Nb4!N to eliminate an enemy bishop. For example:
10.Bb5† Nxb5 11.Qxb5† Kf8 The awkward position of the black king is only temporary.
12.Na3 f6 13.Nf3 a6 14.Qe2 Nc6³ Black’s king will go to f7 and he stands slightly better due
to his bishop pair and well-placed pieces.
9...Be6 10.Ne5
Again White tries to disrupt Black’s coordination, but the second player can seize the initiative
with:
 

213
10...Nd4!N
10...Nb4 gave Black a decent game in A. Sokolov – Giertz, Basel 2007, but the text move is
stronger. The powerfully centralized knight will be annoying for White.
11.Qd1
11.Qh5 is met by 11...c5³ when Black has more space in the centre and his pieces are well
placed, while White’s queenside is still far from developed.
11...0-0³
Black’s lead in development and better piece placement give him a slight edge.
 
8...c6
White has two serious options: A21) 9.Qe2 and A22) 9.b3!?.
 

A21) 9.Qe2

214
 White prevents Black from castling for the moment. Cohen gets this far when attempting to show
that 5...d5 is inferior to 5...Nf6, but he only considers 9...Nd7 (technically 11...Nd7 in his book, due
to the transposition noted earlier in the 8.Qd1 line in the notes to variation A1 on page 104), which is
less accurate for Black. Instead I recommend:
 
9...Bg4!
This important move was played by Kramnik more than twenty years ago. It is essential for Black
to start developing his queenside pieces. A deeper point behind the text move is that Black can often
give up the bishop pair without suffering any ill effects.
 
10.b3
10.h3N is obviously met by: 10...Bxf3 11.Qxf3 (weakening the kingside with 11.gxf3? is simply
bad for White: 11...Nd7 12.b4 [or 12.b3 Nc5µ followed by exchanging the d3-bishop, when White
has no compensation for his inferior pawn structure] 12...Kf8!µ White’s weak pawns will soon begin
to tell) 11...0-0= Followed by ...Nd7, with a completely fine position.
 
10...Nd7
Black wants to play ...Nc5 next, exchanging the d3-bishop. The critical try for White is:

215
11.Ba3!?N
Preventing Black’s idea and threatening to take on d6. Sakaev gave this move to support his
opinion that 5...d5 is inferior to 5...Nf6 (like Cohen, he reached the position two moves later due to
the same transposition). I agree that the text move is the most interesting continuation but I do not
concur with Sakaev’s evaluation of the ensuing position.
11.Bb2 proved harmless after 11...Nc5! 12.Na4 Bxf3 13.Qxf3 Nxd3 14.Qxd3 when a draw was
agreed in Morozevich – Kramnik, Wijk aan Zee 2000. The continuation might have been: 14...0-0
15.Nc5 Preventing Black’s intended ...Bf6 because of the fork on d7. 15...b6 16.Qg3 Nf5 17.Qf3
Bxc5 18.Qxf5 Re8= A draw would be the likely outcome.
 
11...Nc8
The knight drops back, protecting the e7-bishop.
 
12.Na4
After 12.Bxe7 Nxe7 Black is ready for ...Nc5. So 13.b4 is logical, but then Black can revert to his
other easy idea:

216
13...Bxf3 14.Qxf3 (14.gxf3?! as usual is bad for White. 14...Nf8! prepares to block the e-file with
...Ne6. After 15.f4 Qd6 16.f5 g6³ the pressure on the e-file is temporary whereas White’s structural
weaknesses are long-lasting.) 14...0-0= Black is completely fine.

12...Bxf3! 13.Qxf3
13.gxf3?! is met by 13...Kf8!³ and White will suffer from his weaknesses on the kingside.
 
13...0-0 14.Bxe7
14.Qh3 g6= is fine for Black.
 
14...Nxe7

217
 The smoke has cleared and Black has no problems.
 
15.Re2 Nc8
Other moves are playable but rerouting the knight to d6 (or possibly b6 in some lines) seems
simplest.
 
15...b5 is unnecessary, as White’s knight is poorly placed and he might consider rerouting it via b2
anyway. 16.Rae1 Ng6 17.Nb2 is given by Sakaev with a “±” assessment, which seems optimistic to
say the least: after 17...Nc5 Black should be fine. The text move seems even easier though.
 
16.Rae1
16.Qg3 Nf6 makes the d6-square available again.
 
16...g6!?=
Blunting the enemy bishop. It is not obvious how White should improve his position. Meanwhile
Black controls all the important squares and will soon achieve full coordination with some
combination of ...Qf6, ...Nd6 and possibly ...b5, followed by trading one or possibly both pairs of
rooks along the e-file.
 

A22) 9.b3!? 0-0 10.Ba3

 
This has only been played once but it demands an accurate response:

10...Re8!N

218
 Protecting the dark-squared bishop and preparing ...Nd7-f8.
 
A vital point is that the immediate 10...Nd7?? loses to 11.Qe2!, since 11...Re8 runs into 12.Bxd6.
 
10...Bg4?! also proved unsatisfactory after 11.h3 Bh5 12.Qe2 Bf6 13.Qe3 Re8 14.Qf4² when Black
was under significant pressure in Grischuk – Caruana, Leuven (rapid) 2018.
 
11.Ne5
11.Nd4 is nothing special after: 11...Bf8 12.Qf3 (12.Rxe8 Qxe8 13.Qf3 is also met by the strong
13...Nf5! when Black has chances to seize the initiative)
 

12...Nf5! A nice trick! It is useful to remember this motif whenever the bishop on a3 and the knight
on d4 are unprotected. 13.Bxf5 (13.Nxf5?! is worse in view of 13...Bxa3 14.Qg3 Bxf5 15.Bxf5 g6³
when Black benefits from superior central control) 13...Bxa3 14.Rxe8† Qxe8 15.Bxc8 Qxc8 In such
positions the bishop tends to be slightly stronger than the knight, so Black is at least equal.
 
With the move in the main line, White is looking for some tactical ideas on the kingside. However,
we can put a stop to his evil intentions with the precise move:

219
11...g6!
The ‘natural’ 11...Bf6? allows devastation on the kingside: 12.Qh5! g6 13.Nxg6! hxg6 14.Bxg6
fxg6 15.Qxg6† Bg7 16.Bxd6+–
11...Nd7? also runs into serious trouble after 12.Bxh7† Kxh7 13.Qh5† Kg8 14.Bxd6±.
 
12.Qf3
12.Qe2?! has no purpose whatsoever, other than setting a single trap. 12...Be6!³ is the correct
reply, followed by ...Nd7 and Black is slightly better. (12...Bf6? 13.Bxd6 Qxd6 14.Nxf7!± was the
potential banana skin that we needed to avoid slipping on.)
 
12...Bf5!
12...Bf6?? still does not work due to 13.Bxd6! Qxd6 14.Ng4! Rxe1† 15.Rxe1 and Black is busted.
 
Instead Black should carry out the typical plan of exchanging the light-squared bishops. White can
try a few tactical shots but none of them offer him any more than equality.
 
13.Bxf5
13.Bxd6 Bxd6 14.Nxf7!? Kxf7 15.Bxf5 leads to a perpetual after: 15...Rxe1† (the immediate
15...gxf5 should lead to a draw in the same way) 16.Rxe1 gxf5 17.Qxf5† Kg7 18.Qg4† Kh8
19.Qd4† Kg8 20.Qg4†=
 
13...Nxf5

220
14.Nxf7!?
This is White’s only serious attempt to cause problems.
 
14...Kxf7 15.Bxe7 Rxe7 16.g4 Rxe1†
16...Nd7 17.gxf5 Rxe1† 18.Rxe1 transposes.
 
17.Rxe1 Nd7
It is high time for Black to develop his queenside.
17...Qg5? looks obvious, but White has 18.Nxd5! cxd5 19.Qxd5†± followed by Qxb7(†) when the
rook will fall.
 
18.gxf5 Qg5†
The queen will recapture on f5 next. White can either settle for an equal endgame or keep the
queens on the board with:

221
19.Qg3!? Qxf5 20.Re3
20.Qc7?! may appear scary but 20...Rf8! sees Black create his own threats on the kingside. For
instance, if 21.Qxb7?? then 21...Kg8!–+ attacks f2 and, with ...Ne5 coming, White cannot save his
position.
 
20...Nf6 21.Rf3
21.Qc7†?! Qd7³ offers Black some winning chances.
 
21...Qe6=
Black has everything under control.

B) 5.c4

222
 This way White controls the d5-square first, making Black think more carefully before supporting
his knight with ...d5, although he often plays it anyway within a few moves.
 
5...Nc6
5...Be7 has been a more frequent choice but I prefer the text move. The point is to avoid
committing the bishop too early. In particular, if White plays d2-d4 and Black plays ...d5, then the
bishop can come directly to b4.
 
We will analyse four main options: B1) 6.Be2, B2) 6.d4, B3) 6.Nc3 and B4) 6.d3.
 

B1) 6.Be2 Be7 7.0-0 0-0 8.d4

 
White usually plays this way.
 
8.d3?! Nf6 9.d4 d5 leaves White with a significantly inferior version of variation B4.
 
8.Nc3 proves to be completely harmless after: 8...Nxc3 9.dxc3 Bf5! This is usually the best square
for the light-square bishop. 10.Be3 Bf6 11.Bd3 Bxd3 12.Qxd3 Ne5= was completely equal and a
draw was agreed in Dubov – Tkachiev, Warsaw 2013.
 
8...Bf6
The dark-squared bishop is well placed on the long diagonal. Next Black would like to play ...Re8
and ...Bf5, with a harmonious set-up.
 

223
9.Be3
White defends the d4-pawn before playing Nbd2.
 
After the committal, 9.d5 Ne7 10.Nd4 Re8 11.Na3?! Nf5!³ Black seized the initiative and went on
to win a nice game in Nisipeanu – Giri, Wijk aan Zee 2010.
 
9...Re8

10.Nbd2
10.Nc3?! needlessly allows the doubling of White’s pawns. 10...Nxc3 11.bxc3 Bf5 12.Re1 h6
13.h3 Na5 The knight on the rim is not so dim when it puts pressure on the c4-pawn. 14.Qa4 b6
15.Bf1 c5!³ Having fixed the weakness on c4, Black was slightly better in K. Rodriguez – Koch,
corr. 2017.
 
White has also temporarily delayed the development of the queen’s knight with 10.Re1 Bf5 11.Qb3
Qd7, at which point he finally decided to move it: 12.Nc3 Nxc3 13.Qxc3 a5= Black was fine in
Vallejo Pons – Mellado Trivino, Salamanca 1998.
 
10...Bf5 11.Rc1
11.d5?! is worse due to 11...Nb4³ when Black’s active minor pieces exert annoying pressure on
White’s queenside.
 
The text move controls the c2-square, making d4-d5 a more viable prospect.

224
11...d5!?N
11...Qd7 gave Black a viable position in Nijboer – Finegold, Wijk aan Zee 1991, but I would prefer
to prevent White’s idea.
 
12.cxd5
12.c5?! is seldom a good idea when Black’s knight is strongly placed on e4. 12...a6! is a useful
move to prevent Bb5, and after 13.a3 g5!ƒ Black has excellent chances on the kingside.
 
12...Qxd5 13.Bc4 Qd7=
Black has a good position playing against the IQP.
 

B2) 6.d4 d5
 
Now we will see the advantage of Black’s previous move: the bishop can go directly to b4.
 
White’s main options are B21) 7.Nc3?! and B22) 7.a3.
7.c5?!
This gives White an inferior version of variation B42, as Black’s knight is evidently more
active on e4 than on f6.
7...Be7 8.Bb5 0-0 9.0-0
9.Bxc6 bxc6 10.Ne5? (White should prefer 10.0-0 when 10...Bg4 transposes to the line below)
is refuted by 10...Ba6!–+ when White’s king will suffer.
9...Bg4 10.Bxc6 bxc6 11.h3 Bh5
The best try for White would have been:

225
12.Nc3!N
12.g4?! weakens the kingside and after 12...Bg6 13.Ne5 Bf6!µ Black went on to win a fine
game in Karjakin – Caruana, Leuven (rapid) 2018.
12...f5
Cementing the knight on e4 while also gaining space on the kingside.
13.Ne2!
White should make the difficult but necessary decision to allow the doubling of his kingside
pawns.
Unpinning the knight with 13.Qd3 is well met by 13...Bf6, covering the e5-square and putting
pressure on the d4-pawn. After something like 14.Bf4 Bxf3 15.gxf3 Ng5µ White has a worse
version of the main line as his knight has a harder time reaching the desired squares f4 and d3.
13...Bxf3 14.gxf3 Ng5ƒ
The computer calls it equal but Black’s position is easier to play due to the kingside structure.
 

B21) 7.Nc3?!

226
7...Bb4
Black already has a lead in development and a lot of activity.
 
8.Qc2
8.Qb3? is much worse due to 8...Qe7 9.Be3 Na5 10.Qc2 (10.Qa4† is met by 10...b5!–+ and White
cannot avoid material loses) 10...dxc4 11.Ne5 Bf5–+ with a winning initiative, Breznik – Vukanovic,
Ljubljana 2000.
 
8.Bd2 was played in Prasad – Ye Rongguang, Kolkata 1991, when 8...Be6!N would have been best,
as Black does not need to hurry to exchange on d2. After 9.cxd5 Bxd5³ Black is ready to exchange
on d2, with an obviously favourable situation playing against the IQP.
 
8...Qe7!
This precise move creates a lot of problems for White.
 
9.Be3
9.Be2? is refuted by 9...Bf5! 10.Qb3 (or 10.cxd5 Nxc3–+) 10...Na5 11.Qa4† b5! and Black wins.
 
9...Bf5!
Black does not waste time and creates a threat on every move!

227
10.Bd3!N
This is the only way to keep the game going for White.
 
10.Qc1? loses to 10...Nxc3! 11.bxc3 Ba3 12.Qd2? (12.Qd1 is the lesser evil but 12...Bb2 enabled
Black to win the exchange for no compensation whatsoever in Longson – Daskalov, England 2018)
and now Black delivers the killing blow:

12...Nb4! 13.Rd1 Nc2† 14.Ke2 dxc4–+ Tolnai – Haba, Kecskemet 1992.

 
10...dxc4 11.Bxc4
Now Black has an enviable choice.

228
 
11...Nxc3
This is the simpler option, playing on Black’s positional advantages.
11...Ng3!? leads to more forcing play: 12.Qa4 Nxh1 13.d5 Bd7 14.dxc6 Bxc6 15.Bb5 Bxb5
16.Qxb5† c6 17.Qc4 0-0 18.0-0-0

18...Nxf2 19.Bxf2 Bxc3³ With a rook and two pawns versus two minor pieces, Black is the slight
favourite.
 
12.Qxf5 g6

13.Qc2

229
 13.Qd3? allows 13...Na4†! 14.Ke2 Nxb2µ when Black is a clear pawn up.
 
13...Nd5† 14.Kf1 0-0-0³
Black’s pieces are well placed while White’s king still needs some care.
 
 

B22) 7.a3
 
Covering the b4-square is a sensible choice, especially having seen the difficulties White experienced
in the variation above.
 
7...Bf5!
It is important to remember this precise prophylactic move. Black defends the knight in advance,
before White can generate pressure against it with Bd3 and Qc2.
 
8.Be3
I checked a few other plausible moves:
 
8.Qb3?!N
This ‘principled’ reaction to Black’s last move allows:
8...Na5! 9.Qa4† c6 10.cxd5?
The lesser evil is 10.c5 Be7³ when, although the knight on a5 may look out of place, Black will
soon break out with ...b5, after which the knight may land on c4.
Black has a powerful intermediate move:

10...b5! 11.Bxb5

230
 After 11.Qd1 Qxd5µ Black’s pieces are active and well placed.
11...cxb5 12.Qxb5† Bd7 13.Qe2 f5µ
White’s three pawns do not fully compensate for the missing piece.
 
8.Nc3N 8...Nxc3 9.bxc3 Be7 10.cxd5 Qxd5 leads to a typical Petroff structure:

Black intends ...0-0 next and is doing fine. If White tries to get too aggressive with 11.c4? then
11...Qe4†!µ is an excellent counter, followed by ...Bf6 when the d4-pawn is an obvious target.
 
Finally, 8.Bd3 is a natural developing move but 8...dxc4 shows the benefit of having the bishop on
f5. 9.Bxc4 Bd6 10.0-0 0-0 11.Nc3 Nxc3 12.bxc3 h6= Black was completely fine in Rasulov –
Godarz, Mashhad 2016.
 
8...Be7

231
9.Bd3
It is important to note that 9.c5?! would be a strategic mistake. For example after, 9...0-0 10.Be2
(10.b4?! Bf6 11.Be2 b6! is a typical break in this structure, and after 12.b5 Na5µ the knight will
come to c4 with strong effect) 10...Bf6 11.0-0 b6!³ Black’s pieces are ideally placed and White’s
queenside will come under pressure.
 
9...dxc4 10.Bxc4 0-0 11.Nc3 Nxc3 12.bxc3 Na5
Black was at least equal in Babic – Repp, corr. 2016.
 

B3) 6.Nc3

232
 This method of challenging the knight is not bad, but Black should have no trouble reaching a safe
position.
 
6...Nxc3 7.dxc3
7.bxc3
This has been tried by some strong players but I consider it slightly inferior, as the bishop
remains blocked on c1 for the time being.
7...Bf5!
Black covers the important b1-h7 diagonal before continuing with the development of his
kingside.
8.d4 Be7 9.Be2
9.Bd3 can simply be met by: 9...Qd7 10.0-0 0-0 11.Ne1 Na5! A typical scheme, putting
pressure on the c4-pawn. 12.Nc2 Bxd3 13.Qxd3 This occurred in Antipov – Huzman,
Jerusalem 2016, and now the simple 13...Rae8!N³ would have enabled Black to combine play
on the e-file with pressure on the c4-pawn.

233
9...0-0 10.Bf4 Bf6 11.0-0 Na5
This is the usual square for the knight in this structure, putting pressure on the weak c4-pawn.
12.Re1 Re8 13.Qa4 b6=
Black was doing fine in Piorun – Fridman, Wroclaw 2009.
 
7...g6
Black can also develop the bishop to e7 but I prefer the fianchetto. There are not so many games
from this position so I have analysed a number of lines myself. I mostly concentrated on set-ups
involving long castling and h4-h5 by White, which seems like the critical way of playing against the
pawn on g6.
 
8.Be3
8.Bg5 Be7 gives White no advantage, for instance: 9.Bh6 Bf8 10.Bxf8 (10.Bg5 Be7 led to an early
draw in some games, including Jones – Mamedyarov, Wijk aan Zee 2018) 10...Kxf8 The king will be
safe on g7. 11.Qd2 Qe7† 12.Be2 Kg7 13.0-0-0 Bf5 14.Nd4 Nxd4 15.Qxd4† Qe5= Black was doing
fine in Khanin – Ovchinnikov, Samara 2017.
 
In the event of 8.Bd3 it is useful to insert 8...Qe7†! before White can castle. 9.Qe2 (9.Be3?! is worse
due to 9...Bh6³; 9.Be2 Bg7 10.Bg5 Qd7 transposes to the 8.Be2 line below) 9...Qxe2† 10.Bxe2 Bg7
Without queens on the board, the doubled pawns on the c-file are more of a liability than an asset for
White. 11.Be3 0-0 12.0-0-0 Re8 13.Rhe1 b6 Now White should take the opportunity to straighten out
his c-pawns with:

234
14.Nd4!N Intending Bf3. (14.Nd2?! occurred in Sjugirov – Rakhmanov, Sochi 2018, when
14...Bd7N followed by doubling rooks along the e-file would have given Black a slight edge)
14...Nxd4 15.cxd4 Bd7= With an equal endgame.
 
A final option is:
8.Be2 Bg7 9.Bg5
Trying to disturb the coordination of Black’s pieces.
9...Qd7
Black should not worry about the light-squared bishop as it can be developed on b7.
9...Bf6?! 10.Bh6! could be problematic.

10.h4!?N

235
 10.0-0 proved harmless after 10...0-0 11.Re1 Re8 12.Nd4 Nxd4 13.cxd4 c5= when Black’s
bishop proved to be well placed on the long diagonal in Lonyuk – Latib, corr. 2013.
In the event of 10.Qd2N Black should fianchetto his other bishop: 10...b6 11.0-0-0 (or 11.Bh6
Bxh6 12.Qxh6 Bb7= followed by ...0-0-0) 11...h6! 12.Be3 (12.Bh4?! leaves the bishop less
than ideally placed, and Black is doing well after 12...Bb7³ and ...0-0) 12...Bb7= Followed by
...0-0-0.
10...h6 11.Be3 b6
With the bishop coming to b7 next, Black is doing fine. A logical continuation would be:
12.h5
12.Nd4 is simply met by 12...Bb7 13.Bg4 f5 14.Bf3 0-0-0= with a comfortable game for Black.
The idea of the text move is to weaken the f5-square before putting the knight on d4.
12...g5 13.Nd4 Nxd4
13...Bb7? is bad due to 14.Bg4± and Black does not have ...f5.
14.Bxd4
14.cxd4 Bb7³ is good for Black, as both his bishops are well placed.
14...0-0 15.Bxg7 Kxg7 16.Qd4† Kg8 17.Bf3!?
17.f4?! Re8!ƒ is dangerous for White.

17...Re8† 18.Kd2 c5!?

Gaining time for ...Bb7.
19.Qd3 Bb7=
The ensuing major-piece endgame is fine for Black.
 
8...Bg7 9.Qd2
A logical alternative is:
9.c5
This leads to simplifications but the following line still contains some interesting points.
236
9...dxc5 10.Bxc5
10.Qxd8† Nxd8 11.Bxc5 Ne6 12.Be3 Bd7= leads to more straightforward equality.
10...Be6 11.Bb5

11...Qd5!
A strong move, forcing the queen exchange on our terms.
12.Qxd5 Bxd5 13.Bxc6†
13.Nd4 is simply met by 13...Bxd4 14.Bxd4 0-0, intending to take on d4 or g2. So White is
forced to go for 15.Bxc6 Bxc6= and a draw was soon agreed in Dijon – Sorcnik, corr. 2017.
13...bxc6!
With the bishop pair in hand, the doubled pawns are no big deal for Black.
13...Bxc6 14.0-0-0 could be slightly annoying.
14.0-0
14.0-0-0 0-0-0 15.Bxa7 Bxa2 is just equal.
14...a5 15.Rfe1†
 

237
15...Kd7!N
15...Kd8 was good enough to hold in Keber – Ciani, corr. 2016, but the text move is more
accurate. The surprising point is that invading on the 7th rank with check actually leads to
problems for White, as shown after:
16.Re7†?! Kc8 17.Bd4 Bxd4 18.Nxd4 Kb7
Black’s bishop is extremely strong.
19.b3 Rae8 20.Rae1 Rxe7 21.Rxe7 Rd8!
It should be a draw, but Black has chances to seize the initiative on the queenside with the help of
his strong bishop.

9...0-0 10.0-0-0!?N

238
 This is the most ambitious move although it comes with a certain amount of risk for White.
 
10.Bd3 can be met by 10...Re8 11.0-0 Ne5!= and Black is doing well.
 
10.Be2 is met by 10...b6 11.0-0 Bb7 with ...Re8 and ...Qd7 coming up. Black was fine in De Carolis
– Palladino, corr. 2011.
 
10...Bf5!
The bishop is well placed here.
 
11.Bd3
11.h4 should be met by 11...h5= as it will be difficult for White to arrange g2-g4 in an effective
way.
If White starts with 11.Bh6 Re8 before playing 12.h4, then 12...Qe7!³ is an excellent reply,
intending ...Qe4.
 
11...Bxd3 12.Qxd3 b6!
The last move prevents any c4-c5 ideas, and Black is at least equal. His plans include ...Qd7,
...Rfe8 and perhaps ...a5. It is hard to imagine that White can create a decent attack on the kingside;
for instance, if 13.h4 then the simple 13...h5³ shuts his play down.
 

B4) 6.d3

Despite being the least popular of White’s four main options, I think it is his best bet.

239
 
6...Nf6 7.d4 d5
We will consider B41) 8.Nc3 followed by the more ambitious B42) 8.c5!?.

B41) 8.Nc3
 
A normal developing move like this allows Black to develop the bishop actively on b4, which
generally gives him a pleasant game.
 
8...Bb4 9.a3
9.Be2 dxc4 10.Bxc4 0-0 is fine for Black, and 11.0-0 Bg4 is covered briefly on page 128 – see
11...Nc6 in the notes to variation B2 in the next chapter.
 
9.Bg5 has been tried in some games but Black can ignore the ‘threat’ with: 9...0-0 10.cxd5

10...Qe8†! Unpinning the f6-knight so that it can take on d5 next. 11.Be2 Nxd5 12.Bd2 The bishop
had nothing left to do on g5. 12...Be6 13.0-0 Rd8= With a firm grip on the d5-square, Black had
nothing to fear in Stoma – Zontakh, Kazimierz Dolny 2001.
 
9.c5?
This attempt to cut the b4-bishop off from the kingside backfires.
9...Ne4! 10.Qb3
10.Bd2 is simply met by 10...0-0 11.Be2 Bg4 (11...b6!?³ is also promising), developing the
bishop and putting pressure on White’s centre. After something like 12.0-0 Bxf3 13.Bxf3 Nxd4
14.Bxe4 dxe4 15.Nxe4 Bxd2 16.Nxd2 the position is close to equal, but 16...Ne6³ favours

240
 Black slightly, as his rooks can come into the game quickly and White’s overextended
queenside may present targets.

10...Bg4!
A typical move, developing another piece and putting pressure on the d4-pawn at the same
time.
11.Bb5
11.Be2!?N is another idea but Black keeps the upper hand after: 11...0-0 (11...Bxf3?! backfires
after 12.gxf3! Nxd4? 13.Qd1!± when all the tactics work in White’s favour) 12.0-0 Now Black
has a nice strategic idea: 12...Bxc3! 13.bxc3 b6!³ After this thematic break against d4-c5 pawn
wedge, White will suffer from structural deficiencies.
11...Bxc3† 12.bxc3
We have been following Santolini – Giuriati, Venice 1996, in which Black made the
understandable but slightly inaccurate exchange on f3. An improvement is:

241
12...Qf6!N
This precise move puts pressure on the knight, thus forcing White’s bishop to retreat.
13.Be2
13.Qxd5? is bad due to 13...Bxf3 14.gxf3 Nxc3 15.Bxc6† bxc6 16.Qe5† Qxe5† 17.dxe5 0-0-
0µ when Black enjoys a much better pawn structure and more active pieces.
13...0-0 14.0-0 b6!³
Black has the upper hand due to his better pawn structure.
 
9...Bxc3† 10.bxc3 0-0
Black’s lead in development and more solid pawn structure are at least as relevant as White’s
bishop pair. A good example continued:
 
11.Bd3 Re8† 12.Be3 dxc4 13.Bxc4 Na5!
Black is fighting for the important c4-square.
 
14.Bd3 Qd5 15.0-0

242
15...Bf5
When your opponent has the bishop pair, it is generally a good idea to trade a pair of bishops.
 
16.Bxf5 Qxf5 17.Ne5³
The players agreed a draw in Siebrecht – Farago, Arco 2010, but Black had every reason to play
on. One promising idea is to play ...Nd7 to trade knights and conquer the c4-outpost; another is to
prepare ...c5.
 

B42) 8.c5!?

243
 This is the critical test. White gains space on the queenside, crucially before ...Bb4 can be played
and without the black knight being actively placed on e4.
 
8...a6!
It is worth investing a tempo to prevent Bb5. If White is allowed to play this move then he will
most probably be able to damage Black’s structure by exchanging on c6. The resulting structure is
not easy to handle for Black, unless you already have some activity on the kingside to compensate
for it.
 
9.Nc3 g6!
It is important to develop the bishop to g7 in order to put pressure on the d4-pawn.
 
After 9...Bg4 10.Be2 Be7 11.h3 Bh5 12.Be3 0-0 13.0-0 Ne4 14.Qb3 Nxc3 15.bxc3 b6 16.Rfe1 Qd7
17.Rad1² Black had failed to equalize in Eljanov – M. Muzychuk, Gibraltar 2019.
 
10.Bd3 Bg7 11.0-0 0-0 12.h3 b6!
With this thematic break, Black eliminates the c5-pawn and with it White’s space advantage on the
queenside, before White is able to consolidate it.
 
13.cxb6
White cannot allow an exchange on c5 as the g7-bishop will become too strong.
 
13...cxb6 14.Bg5
Pinning the knight and hoping to provoke ...h6, which would allow White to gain a tempo with
Qd2 after first retreating the bishop.

244
 
14...Bb7
The bishop may seem passively placed because of the d5-pawn, but from b7 it may help to support
active ideas involving ...Ne4 at some point.
 
15.Qd2 Re8 16.Rfe1 Qd6
The structure is symmetrical and the position is dynamically balanced.
 
17.Rac1
17.Bf4 should be met by 17...Qf8!, and if 18.Bc7 then simply 18...b5= and Black is solid.

17...b5!
Black gains space on the queenside while preventing options such as Na4. Black can consider
following up with ...Na5-c4 at some point, or perhaps ...b4 followed by ...Ne4.
 
18.Bh6 Rac8 19.a3
Having developed all his pieces, Black opted for some exchanges.
 
19...Rxe1† 20.Rxe1 Bxh6 21.Qxh6 Na5
The knight gets ready to jump to c4.
 
22.Ne2
A draw was agreed here in Neto – Finocchiaro, corr. 2017, which is a pity, as the position is full of
life. White wants to transfer his knight to g3 and create threats against the black king, but we have
plenty of resources. My analysis continues:

245
22...Re8!
Interfering with White’s idea of putting the knight on g3.
 
22...Nc4 allows 23.Ng3!, ignoring the b2-pawn and focusing on the kingside. 23...Nxb2 24.Bb1!
The threat of Nf5 is difficult to meet. For instance: 24...Re8 25.Rxe8† Nxe8 26.Ng5 Qf6 (the more
obvious 26...Nf6? loses to 27.Nh5!+–)
 

27.Nh5! Qh8 Now White has the spectacular 28.Bf5!, threatening Ne6! 28...gxf5 29.Qb6 White
wins the piece back with benefits. 29...h6 (29...Bc8? 30.Qc6+–) 30.Nf3 Bc8 31.Qc6 Be6 32.Qxe8†

246
Kh7 33.Qe7± White’s pieces are more active than their counterparts and Black’s king is still under
pressure.
 
23.Rc1 Nc4
The insertion of the respective rook moves makes the knight jump much safer now.
 
24.Rc2
24.Ng3 can be safely met by: 24...Nxb2 25.Nf5 This is the only move that maintains the balance
for White. (25.Bb1? is too slow and 25...Nc4! blocks the c-file, thus permitting the resource of
...Bc8, as seen after 26.Nf5 gxf5 27.Bxf5 Bc8!–+) 25...Qf8 26.Qg5! The game ends in a perpetual
after: 26...Nxd3 27.Nh6† Kg7 28.Nf5† Kg8=

Now Black should fix the queenside pawns with:

 
24...a5 25.Ng3 a4!
Now the b2-pawn remains under pressure and the c4-knight cannot be kicked away by b2-b3.
Black is not worse and the game goes on.
 

Conclusion
 
This chapter has dealt with a couple of interesting sidelines. Against 5.Bd3 I favour the bold 5...d5
(rather than the playable but meek 5...Nf6) when Black needs to avoid a couple of pitfalls, but with
accurate play he obtains a perfectly good game. White faces his own challenges as well: if his early
lead in development does not yield anything tangible, his bishop on d3 and knight on c3 may find
themselves restricted by the b7-c6-d5 pawn chain.
 

247
5.c4 is the other move examined in the chapter, when 5...Nc6 is our active and flexible response.
From there White has tried several options and move orders, but in most cases he will move his pawn
to d4 – either immediately, or after first driving our knight back with d2-d3. In both scenarios, Black
should respond to the arrival of a pawn on d4 by playing ...d6-d5. Generally, if he gets a chance to
follow up with the active ...Bb4(†), he will have an easy game. That is why 6.d3 Nf6 7.d4 d5 8.c5!?
is the most critical variation of the chapter – and note that it is important for White to drive the knight
back to f6 before establishing this pawn structure, as the knight would be a lot more active on e4.
Black responds with 8...a6! to prevent Bb5, followed by a kingside fianchetto and later ...b6 to
challenge White’s pawn wedge, leading to a complex middlegame with mutual chances.

248
A) 7.Nc3 124
B) 7.c4 Bb4† 126
B1) 8.Bd2 126
B2) 8.Nc3 127
C) 7.Bd3 Bd6 128
C1) 8.Qe2† Be6 128
C11) 9.Ng5 129
C12) 9.c4 130
C2) 8.0-0 0-0 131
C21) 9.h3 132
C22) 9.Bg5 Bg4 10.Nbd2 Nbd7 133
C221) 11.c4 134
C222) 11.c3 135

 
1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4 5.d3

249
 In this variation White kicks the knight from e4 first before going for d3-d4. This is a less
ambitious approach than the 5.d4 d5 lines that we will analyse later (see Chapters 8-10), in which
White tries to exploit the placement of the knight on e4 by attacking and undermining it with c2-c4.
 
5...Nf6 6.d4 d5
Sakaev prefers 6...Be7 followed by ...0-0, but I would rather take some space in the centre and
develop the bishop more actively on d6. As you may have noticed, the text move transposes to the
Exchange Variation of the French Defence. This is not considered theoretically critical, but it
certainly has a right to exist.
 
We will analyse three main options: A) 7.Nc3, B) 7.c4 and C) 7.Bd3.
 
7.h3 is hardly necessary at this stage. 7...Bd6 is a good reply, when White has nothing better than
8.Bd3 0-0 9.0-0, transposing to variation C21.
 
7.Be2 is too passive to threaten us. After 7...Bd6 8.0-0 0-0 9.Bg5 Bf5 10.Re1 Nbd7 11.Nbd2 c6
12.c3 Qc7= Black had a pleasant position in Sibriaev – Ivanchuk, Khanty-Mansiysk 2005.
 
7.Bg5 Bd6 8.c4
8.Be2 0-0 9.0-0 transposes to the 7.Be2 line noted above.
8.Bd3 0-0 9.0-0 transposes to the later variation C22.
8...0-0 9.cxd5
9.Bxf6? Qxf6 10.cxd5 proved to be bad for White after the prophylactic 10...Qg6!µ in Mraz –
Fekete, email 2010. Black’s last move threatens ...Re8†, when White won’t be able to play Be2
as the g2-pawn will hang. Thus White’s king will be stuck in the centre and he will have major
problems.

250
9...Qe8†
Conveniently unpinning the knight.
10.Be2 Nxd5 11.Nc3 Be6 12.0-0 Nc6
This is the usual way for the b8-knight to develop in this structure.
13.Ne4 h6 14.Nxd6 cxd6 15.Bd2 Qd7=
Black was completely fine in Cardoso Garcia – Fagerbekk, corr. 2017.
 

A) 7.Nc3

It may seem counterintuitive to put the knight on a square where its movements are limited by the
d5-pawn, but the text is a favourite choice of the engines.
 
7...Bd6
This is usually the best square for the bishop, and the present variation is no exception.
 
8.Bd3
We should also consider:
8.Nb5!?
White aims for an advantageous piece exchange but it costs valuable time, enabling Black to
build up a useful lead in development.
8...0-0 9.Nxd6 Qxd6 10.Be2 Bg4 11.0-0 Nc6!
The knight is better placed here than on d7, as it can be transferred via e7 to f5 or g6.
12.Be3 Rae8
All of Black’s pieces are well placed.

251
13.c3 Bh5
Intending ...Ng4.
14.h3 h6
A useful move, covering the g5-square.
15.Qc2
Now Black went for an ambitious plan.

15...Ne4!?
A solid alternative is 15...Re6N= followed by ...Rfe8, with a fine position.
16.Rae1 f5 17.Bc1 Nd8 18.Ne5 Bxe2 19.Rxe2 c5!„
Black obtained nice play in Nirvana 2.4 – Booot 6.2, engine game 2018.
 
8...0-0 9.0-0 Bg4
One of White’s possible plans is to transfer the c3-knight via e2, so Black prevents it by preparing
to exchange on f3, damaging White’s structure.
 
10.h3
10.Nb5 Nc6 11.Nxd6 Qxd6 gave Black a good game in Walbrodt – Delmar, New York 1893. The
position is similar to the 8.Nb5 line noted above, but here White’s bishop is on d3 rather than e2, so
the pin on the f3-knight may prove annoying for him.
 
10.Bg5 c6 11.h3 Bh5 12.g4 weakens the kingside. A good example continued: 12...Bg6 13.Ne5 Re8
14.Nxg6 hxg6 15.Re1 Rxe1† 16.Qxe1 Nbd7

252
 Black’s position is completely fine. Note that 17.Bxg6!? does not win a pawn due to 17...Qb6!
18.Bf5 Qxd4= as seen in M. Horvath – Latronico, corr. 2011.
 
10...Bh5

11.g4
Breaking the pin is an ambitious choice, although weakening the kingside comes with certain risks.
 
11...Bg6 12.Ne5 c5!
Opening the centre is well timed, due to White’s loose kingside.
 

253
13.Nxg6 hxg6 14.dxc5 Bxc5=
We have been following Wasp 3.2 – DeusX 1.0, engine game 2018. Objectively the position is
equal but, from a human perspective, the pawn on g4 gives Black some cause for optimism.
 

B) 7.c4

This method of fighting for the centre is easily countered by:

 
7...Bb4†
7...Nc6 would transpose to variation B4 of the previous chapter. Although Black’s position there is
perfectly playable, White can try to play somewhat ambitiously with 8.c5!?. The text move simply
gives Black a pleasant game in all variations, so it should be preferred when this move order arises.
 
We will consider B1) 8.Bd2 and B2) 8.Nc3.

B1) 8.Bd2
 
Exchanging a pair of bishops and playing with an isolani is not challenging at all.
 
8...Bxd2† 9.Nbxd2
This is a speciality of GM Miezis.
 
9.Qxd2?! makes life even easier for Black: 9...0-0 10.Be2 dxc4 11.0-0

254
11...Ne4!N The knight it heading for d6, supporting the c4-pawn and making it hard for White to
recover it. (11...Nc6 was also quite pleasant for Black in Bergvoll – Brynell, Copenhagen 2006)
12.Qc2 Nd6 Black is more than okay.
 
9...0-0 10.Be2 Nc6
The knight is usually best developed to c6 in this kind of structure.

11.0-0 Bg4 12.Rc1

12.h3 is simply met by 12...Bh5. A good example continued 13.Rc1 Re8 14.cxd5 Nb4 15.Bb5 Re7
16.Qb3 Nbxd5= and Black was comfortable in Miezis – Ehlvest, Dhaka 2001, with firm control over
d5.

255
 
12...Re8 13.Re1 dxc4 14.h3 Bh5 15.Nxc4 Nd5=
Here too, Black had a comfortable version of the IQP structure in Miezis – Neiksans, Parnu 2016.
 

B2) 8.Nc3 0-0

 
8...Nc6 would transpose to variation B41 of the previous chapter, but it seems natural to castle first
when we have the chance.
 
9.Be2
9.Bd3 dxc4 10.Bxc4 is the same thing.
 
9.a3?!
This move is too slow.
9...Re8† 10.Be3
After 10.Be2 Bxc3† 11.bxc3 dxc4 White will have to work hard to get the c4-pawn back. 12.0-
0 Be6 13.Ne5 was played in Andrijevic – Golub, Bosnjaci 2019, and now 13...Ne4!N 14.Qc2
Nd6³ would have left Black’s extra pawn well protected. Even if White finds a way to capture
it, Black will still end up with a slightly better pawn structure.
10...Bxc3† 11.bxc3

11...Be6!N
This simple developing move should give Black a positional edge.

256
 The tempting 11...Ng4 was played in Borsos – Bareev, Dortmund 1990, but 12.Be2!N
12...Nxe3 13.fxe3 Rxe3 14.0-0© would have offered White reasonable compensation for the
pawn, as Black is behind in development on the queenside.
12.cxd5
12.c5 b6³ gives Black easy play against White’s weakened queenside structure.
12...Bxd5 13.Bd3 c5!
Immediately gaining space on the queenside.
14.0-0 c4³
Followed by ...Ne4, when Black’s active pieces gives him a slight edge.

9...dxc4 10.Bxc4 Bg4 11.0-0

11.Be3 can be met in various ways. 11...Nbd7!? is one option, just to show that sometimes the
knight can go this way rather than via c6. (11...Nc6 is also fine, when 12.0-0 transposes to the note to
Black’s next move in the main line below) 12.0-0 Nb6 13.Bb3 Now in Miezis – Groffen, Utrecht
2004, Black could have gained some space on the queenside with 13...a5!?N, with a fine position.
 
11...Bxf3!?
This leads to a concrete variation, resulting in a more or less forced draw.
 
If Black wants to keep playing in a balanced middlegame, then 11...Nc6 is perfectly good. After the
natural 12.Be3 Black has a few good options, including: 12...Bh5!? (12...Qd6 is a more popular
alternative which has scored excellently) 13.h3 Re8 14.a3 Ba5 (14...Bxc3 is also playable but it is
more ambitious to keep the bishop while keeping the d4-pawn isolated)

257
15.g4 Bg6 16.b4 Bb6 17.b5 Na5 18.Ba2 This position was reached in Miezis – Holm, Sweden 2018,
and now Black could have obtained a lot of activity with 18...c6!Nƒ followed by ...Rc8.
 
12.Qxf3

12...Qxd4! 13.Qxb7
White has to accept the challenge, otherwise he will simply be a pawn down.
 
13...Qxc4 14.Qxa8 Bd6
Black has two threats. The main one is to trap the white queen with ...Nc6 followed by ...Rb8; the
second is ...Bh2†.

258
 
15.Qxa7
White saves the queen but allows the back-up threat.
 
15...Bxh2† 16.Kxh2 Qxf1
Black is doing fine and the game soon ended in a perpetual after:
 
17.f3 Qe1 18.Qxc7 Qh4† 19.Kg1 Qe1†
½–½ Gleizerov – Ulibin, Stockholm 2012.
 

C) 7.Bd3

This is the main move, preparing to castle.

 
7...Bd6
White can try to create some imbalance with C1) 8.Qe2† or simply play C2) 8.0-0.
 

C1) 8.Qe2† Be6

 
8...Qe7 9.Qxe7† Bxe7 10.Bf4 offers White a slight edge.
We will analyse C11) 9.Ng5 and C12) 9.c4.
 

259
9.0-0 0-0 leaves White’s queen misplaced. 10.Ne5 Re8 11.Bg5 c5 12.dxc5 occurred in Hakobyan –
Christiansen, St Petersburg (blitz) 2018, when Black missed a great opportunity:

12...Bg4!N 13.Bxh7† (13.Nxg4 Rxe2 14.Bxe2 Bxc5µ leaves White without enough compensation
for the queen) 13...Kf8!µ The tactics work out in Black’s favour.
 

C11) 9.Ng5

White assures himself of the bishop-pair advantage but Black gets plenty of play in return.
 

260
9...Qe7 10.Nxe6
10.Nd2 has the idea to exchange on e6 and play a quick Nf3 to keep control of the e5-square, but
Black avoids it with: 10...Bd7!N (10...Bg4 was also okay for Black in I. Schneider – Meier, Isle of
Man 2018, but I slightly prefer the text move as the bishop is less exposed on d7) 11.Qxe7† Kxe7
12.0-0 Re8= Intending ...h6 and/or ...Kf8, when Black is doing fine.
 
I also considered:
10.0-0 Nc6!
Developing with tempo.
11.c3
11.Bb5?! is not so good, as Black can simply play: 11...0-0 12.Nxe6 fxe6 13.Bxc6 (13.c3 e5³
leaves White behind in development) 13...bxc6 14.Re1 Rae8³ All Black’s pieces are mobilized
and he is ready to open the position with ...e5 or ...c5, while White still needs to develop his
queenside.
Now in Danin – Martinovic, Austria 2017, Black could have exchanged pieces in style with:

11...Bf5!N
Trading off White’s more active bishop, along with the queens.
12.Qxe7† Nxe7 13.Bxf5 Nxf5 14.Re1† Kf8
The position is essentially equal, but Black has a marginal plus due to his better bishop.
 
10...fxe6 11.Nd2
I checked three other possibilities:
11.c3 0-0 12.0-0 e5 (12...c5!? is possible as well) 13.dxe5 Qxe5 14.Qxe5 Bxe5 15.Nd2 Nbd7
16.Nf3 Bd6= reached a level endgame in Gull 3 – Jonny 8.1, engine game 2018.
 

261
11.Be3 Nc6 12.f4N (12.Nd2 occurred in Suarez Gomez – Peralta, Santiago de Compostela 2018,
when 12...e5!N³ would have been good for Black) 12...0-0 13.c3 White seems to have prevented
both the ...e5 and ...c5 breaks, but Black has the strong 13...Rae8!ƒ intending ...e5, with a good
initiative.
 
Also after 11.0-0 Black should play for the ...e5 break before White is able to prevent it: 11...Nc6!
12.c3 e5! (but not 12...0-0? 13.f4!± and the e6-pawn becomes a weakness) 13.dxe5 In Blomqvist –
Westerberg, Ronneby 2018, Black opted for a queen trade, but a better idea would have been:
13...Nxe5!N 14.Nd2! (14.Bc2? is too slow, and after 14...0-0µ Black’s last piece is ready to join the
action with ...Rae8) 14...Nxd3 15.Qxd3 0-0 16.Nf3 Qf7 With a pleasant position for Black.

11...Nc6 12.Nf3 0-0 13.0-0

13.Bb5?! does not stop Black from carrying out his plan: 13...e5! (13...Ne4!?³ is also good)
14.Bxc6 bxc6 15.dxe5 (15.Nxe5? c5 16.Nf3 [16.c3 Rae8 17.f4 cxd4 18.cxd4 c5µ] 16...Qxe2†
17.Kxe2 Rae8† 18.Kf1 cxd4µ leaves White’s pieces badly coordinated) 15...Rae8 16.0-0 Ng4!³
Black will recapture the e5-pawn and retain the more active position.
 
13...e5 14.dxe5 Nxe5 15.Nxe5 Bxe5
15...Qxe5 16.Qxe5 Bxe5 17.c3 c6 18.Bd2 Ng4 19.g3 Bc7= was also equal in Kovalev – l’Ami,
Wijk aan Zee 2019.
 
16.Bf4 Rae8 17.Bxe5 Qxe5 18.Qxe5 Rxe5 19.Rfe1 Ng4 20.Rxe5 Nxe5
The endgame was equal in Kurmann – Bogner, Lenzerheide 2018.
 

262
 C12) 9.c4

This is the other logical follow-up to White’s previous move. White tries to open the position
before Black is able to complete his development.
 
9...dxc4 10.Bxc4 Qe7 11.0-0
11.Bxe6 led to an equal endgame after 11...Qxe6 12.Qxe6† fxe6 13.Nc3 a6 14.0-0 0-0 15.Re1 Re8
16.Bg5 Nc6= in Kosteniuk – Ju Wenjun, Huaian (blitz) 2017.
 
11...0-0 12.Nc3
12.Bg5 Re8 13.Nc3 transposes.
 
12...Re8 13.Bg5
Preventing exchanges with 13.d5!? is met by 13...Bg4 14.Qxe7 Rxe7 15.Nd4 Nbd7 when the
position is equal. For example, 16.h3 can be met by:

263
16...Be5! 17.Be3 Nb6! 18.Bb3 Bd7 19.a4 a6 20.a5 Nc8= Black is fine in this endgame.
 
Now I think Black’s best option is:

13...Nc6!
A clever move, with two hidden traps!
 
13...h6 is another decent option, which led to an eventual success for Black in Carlsen – Grischuk,
Internet (blitz) 2017.
 
14.Bb5

264
 14.d5? Bxd5! 15.Qxe7 Nxe7³ is the more obvious of the two traps.
The more subtle one is revealed after the natural-looking 14.Rfe1?!, when 14...Nxd4!³ wins a pawn.
 
14.Bxe6 Qxe6 15.Qxe6 fxe6= reaches a typical structure with mutual targets on e6 and d4; Black
was fine in Vaccarella – Sempio, corr. 2017.
 
White intends to exchange on c6 and fracture the queenside structure, but it’s not a big deal for
Black. I recommend:
 
14...h6
It is useful to prod the bishop before taking any further action.
 
15.Bh4N
15.d5 hxg5 16.dxe6 Qxe6 17.Qxe6 Rxe6 18.Bxc6 bxc6 19.Ng5 Ree8 was level in I. Gavrilov –
Simeonov, corr. 2018.
 
Now Black can exchange queens with:
 
15...Bg4 16.Qxe7 Rxe7 17.Bxf6 gxf6 18.Bxc6 bxc6=
Black has two sets of doubled pawns, but as compensation he has the bishop pair and the d4-pawn
is weak; overall, the endgame is balanced.
 

C2) 8.0-0 0-0

265
 White’s two main attempts to get something out of the symmetrical position are C21) 9.h3 and
C22) 9.Bg5.
White has also tried 9.Re1 Bg4 10.Nbd2, but keeping the bishop back on c1 is not such a good
idea. (10.Bg5 Nbd7 followed by ...c6 and ...Qc7 is similar to variation C22 below) 10...Nbd7 11.c3
c6 12.Nf1 Qc7 13.Ng3 Rae8

Black is completely mobilized, and after 14.Bd2 Ne4!³ he seized the initiative in Vaibhav –
Aravindh, New Delhi 2017.
 

C21) 9.h3
 
Having prevented ...Bg4, White plans to put his bishop on g5 next. Black has kept the symmetry with
9...h6 in some games, but a much more interesting continuation is:
 
9...Nc6!?
With the idea of ...Nb4 followed by ...Bf5.
 
10.Bg5
The other natural idea is:
10.c3
Preventing ...Nb4, but the knight has another tempting route available.
10...Ne7
Preparing ...Bf5, which explains White’s next move.
11.Qc2 Ng6 12.Bg5

266
 12.Re1 controls the e-file and prepares Ne5. In Omidi – Pourramezanali, Arad 2016, it would
have been good to play 12...c6N, planning to meet 13.Ne5 with 13...Nh5!, with promising play
on the kingside.
12...h6
It is best to kick the bishop without delay.
13.Be3

13...Nh5!N
13...Nf4 was played in Zhang Lanlin – Padmini, Chengdu 2017, but it’s better to bring the f6-
knight to f4, as the other knight may go to h4 to swap off White’s defending knight on f3.
14.Bxg6 fxg6 15.Qxg6
Black was developing a strong initiative anyway, so White may as well grab a pawn and hope
for the best.
15...Nf4 16.Bxf4 Bxf4³
With ...Bf5 coming, Black has more than enough activity to make up for the sacrificed pawn.

267
10...h6
Once again I suggest forcing the play by prodding the bishop as soon as it arrives on g5, although
10...Nb4 is possible too.
 
11.Bh4 g5!?
This may seem weakening, but it works well for Black.
 
12.Bg3 Bxg3 13.fxg3
The best way to target White’s weakened kingside would be:

13...Nh5!N
268
 13...Ne4 was less accurate in Heberla – Nabaty, Stockholm 2018.
 
14.g4
14.Qe1?! is strongly met by 14...Qd6!, renewing the attack on g3. For instance: 15.Nc3 (or 15.g4
Ng3 16.Rf2 f5³ with a promising initiative on the kingside) 15...Qxg3 (15...Nxg3? would be a good
move if it were not for 16.Nb5! Qf4 17.Rf2±, when Black will soon lose the c7-pawn once the f3-
knight moves and drives the black queen away, leaving White in charge) 16.Nxd5 Qd6³ White
remains under pressure.
 
14...Ng3!
By choosing this square for his knight, Black conveniently gains a tempo before playing ...f5 to
open up the kingside:

15.Rf2 f5!ƒ
White will have to play accurately to avoid becoming worse.
 

C22) 9.Bg5

269
 This is White’s most natural way of continuing his development. The bishop goes to its most active
square and the b1-knight can go to d2 next.
 
9...Bg4
Another possibility is 9...h6 10.Bh4 Nc6!?, intending ...Nb4 followed by ...Bf5, just as in the
previous variation, although the fact that White has not spent a tempo on h2-h3 should slightly
improve his prospects. 11.c3 is the natural choice to prevent ...Nb4, and now 11...g5 12.Bg3 Ne4 led
to a complicated battle in Anand – Caruana, Stavanger 2018.
 
10.Nbd2
White supports the knight on f3 before unpinning it with c2-c3 and Qc2.
 
10...Nbd7
Black continues to mirror his opponent.
 
White may proceed with C221) 11.c4 or the more common C222) 11.c3.
 

C221) 11.c4

270
 I noticed that this move was tried by Giri a few years ago, to create some imbalance from the
symmetrical position.
 
11...dxc4 12.Nxc4 Re8!
Black is not afraid of having his bishop traded for the enemy knight, as after the exchange Black’s
queen will gain access to the b6- and a5-squares.
 
13.h3
White has also tried the immediate exchange:
13.Nxd6 cxd6 14.Qb3!?
White is counting on his bishop pair and is not afraid to weaken his kingside.
14...Bxf3 15.gxf3
I suggest improving Black’s play with:

271
15...Nf8!N
15...h6 was played in Grischuk – Akopian, Enghien-les-Bains 2001, but I prefer to bring the
knight to the ideal e6-square, where it hits not only the bishop but also the d4-pawn. Black’s
main plan is to improve his knights and exploit White’s weakened kingside. For example, let’s
see what happens if White wastes time pawn-grabbing on the queenside:
16.Qxb7?! Ne6 17.Be3 Rb8! 18.Qc6
18.Qxa7?! is worse in view of 18...Nd5µ and the queen comes to h4 with great effect.
18...Nh5ƒ
With the queen coming to h4 and the knights ready to jump all over the weakened squares on the
kingside, Black has more than enough play for the pawn.
 
13...Be6 14.Nxd6 cxd6
White would ideally like to have his knight on c3 instead of f3, in order to cover some ground on
the queenside, but shifting it there would cost three valuable tempos. Giri tries to manoeuvre the
knight to a better square but Radjabov finds counterplay on the queenside.
 
15.Nd2 Qa5 16.Bf4
In the event of 16.Be3N, pre-emptively defending d4, then 16...Nb6 followed by ...Nbd5 is equal.

272
16...Qb4! 17.Bg3 Qxb2 18.Rb1 Qxa2 19.Rxb7 Qd5 20.Rc7 Rec8
It is important to exchange the active white rook.
 
21.Rxc8† Rxc8 22.Qa1

22...Bf5!
Simplifying the position.
 
23.Bxf5 Qxf5 24.Qxa7 Qc2 25.Nf3 Qc6=
The position remained equal and the players agreed a draw a few moves later in Giri – Radjabov,
Shamkir 2018.
273
 C222) 11.c3

This is by far the most popular move in this position.

 
11...c6 12.Qc2 Qc7 13.h3!?
13.Rfe1 Rfe8 14.Bh4 Bh5 15.Bg3 Bg6= maintains the symmetry and it’s no wonder that most
games have ended in draws from here. For instance: 16.Bxd6 Qxd6 17.Bxg6 hxg6 18.g3 Rxe1†
19.Rxe1 Re8= Lu Shanglei – Li Chao, China 2018.
 
13.Rae1 is similar to the above line. 13...Rae8 14.Bh4 Intending to exchange Black’s ‘good’ bishop,
but of course Black copies the same idea! 14...Bh5 15.Bg3 Bxg3 16.hxg3 Bg6= Ramirez – Ali
Marandi, Saint Louis 2018. In positions like this, White’s advantage of being first to move is
virtually worthless.
 
The text move is a more serious attempt by White to unbalance the position. Interestingly, Black has
three bishop moves which all seem playable, so you can choose according to your taste.
 
13...Bh5
13...Bxf3 14.Nxf3 Bf4 gets rid of the bishop pair: 15.Bxf4 (15.Bh4 h6 16.g3 Bd6= [16...g5!? could
also be considered] and the threat of ...g5 means that White will sooner or later have to exchange on
f6, leading to an equal position with opposite-coloured bishops) 15...Qxf4 16.Rae1 Rae8 The position
is essentially equal, but White has a symbolic advantage due to having a knight plus an active bishop
versus two knights.
 

274
13...Be6!? is also possible, for instance: 14.Bxf6!? (14.Rae1 h6 15.Be3 Rae8= leads to symmetrical
equality) 14...Nxf6 15.Ng5 g6 16.Nxe6 fxe6

17.Rae1 Rae8 18.Nf3 c5!„ Black has a slightly inferior structure, but he should not have much
problems in maintaining the balance. (18...e5? doesn’t work due to 19.dxe5 Bxe5 20.Bxg6!+–)

14.Nh4!
This is the key move which makes 13.h3 a credible try for an advantage. White’s ideas include Nf5
and possibly f2-f4 at the right moment, playing against the bishop on h5.
 
14...Bh2†
A typical idea, driving the king into the corner before exchanging the dark-squared bishops.

275
 14...Bg6 15.Nxg6 hxg6 16.Nf3 should be tenable for Black but he still has to work to prove full
equality, due to White’s bishop pair.
 
15.Kh1 Bf4 16.Bxf4 Qxf4

17.g3 Qc7 18.Kg2

White improves his king position and can later think of f2-f4.
 
It would be too hasty to play:
18.f4?!
In Medvedkov – Petrik, Pardubice 1998, Black could have obtained powerful counterplay with:
18...c5!N
Black is ready to gain space on the queenside with ...c4 and/or occupy the e-file with one of his
rooks. My illustrative line continues:
19.g4?!
The immediate attempt to trap Black’s bishop backfires.
In the event of 19.dxc5 Nxc5 White cannot trap the bishop on h5 because his kingside will
become too vulnerable. Meanwhile the black knights are eyeing up the weak e4-square, so I
rate Black’s chances as slightly preferable.
19...Bg6 20.f5
Now Black can take full advantage of White’s exposed kingside with:

276
20...Qg3
Attacking the h4-knight as well as the pawn on h3.
21.Ndf3 c4!
An important move, exploiting the fact that retreating the bishop to e2 would fatally prevent
White’s queen from connecting to the defence of the kingside. Instead, White has to settle for
an inferior endgame with:
22.Qg2 Qxg2† 23.Kxg2 cxd3 24.fxg6 fxg6!³
White’s knights are awkwardly placed while Black is ready to penetrate along the e-file. Black also
has active ideas such as ...Nb6-c4, so White will have to battle for a draw.
 
I also checked: 18.Rae1 Rfe8 19.f4?! (White should prefer 19.Kg2, transposing to our main line
below) The text move was played in Eisen – McCartney, Orlando 2014, when Black missed the most
accurate continuation:
 

277
19...c5!N„ With promising play for Black, just as in the 18.f4?! line noted above.

18...Rfe8
Black takes control of the e-file and gives himself the option of ...Be2.
 
I also considered 18...c5N 19.dxc5 Nxc5 20.Nb3 Nxd3 21.Qxd3 Bg6 22.Nxg6 hxg6 when Black
should be able to hold a draw; nevertheless, with the IQP as a target and full control over the
blockading d4-square, White has a pleasant position with a small but risk-free edge.
 
19.Rae1 Bg6
This seems like a reasonable time to exchange pieces.

278
20.Nxg6 hxg6=

We have been following Glodeanu – Osman, Bucharest 2000. White has a token advantage of a
bishop against a knight but I do not believe this counts for anything significant. Black can break with
...c5 if and when it suits him, or simply trade all the rooks along the e-file.
 

Conclusion
 
This chapter has dealt with the 5.d3 Nf6 6.d4 variation, which we meet with 6...d5, transposing to an
Exchange French. Despite the symmetrical position and somewhat drawish reputation of this
variation, we saw that many of the lines could become rather lively and double-edged. Certainly in
the event of 7.c4 Bb4†, the symmetry is immediately broken and Black has a full share of the
chances, regardless of whether White chooses 8.Bd2 or 8.Nc3.
 
7.Bd3 is the main line, when 7...Bd6 is the natural response. Then 8.Qe2† is an interesting attempt to
exploit White’s advantage of the first move, but after 8...Be6 Black’s position holds up well against
both 9.Ng5 and 9.c4, provided he is well prepared and ready to play dynamically, since the ...fxe6
structure opens the f-file and offers possibilities such as ...e5 to maximize the activity of Black’s
pieces.
 
Finally 8.0-0 0-0 is the main line, when 9.h3 offers Black a convenient way of breaking the
symmetry with 9...Nc6!?, with active prospects for Black. We then turned to 9.Bg5 Bg4 10.Nbd2
Nbd7, when Giri’s 11.c4 is interesting but still equal, and 11.c3 c6 leads to positions where it is hard
for White to turn the advantage of the first move into anything tangible.

279
A) 6.Nc3 140
B) 6.d3 Nf6 142
B1) 7.Nc3 143
B2) 7.Bg5 Qxe2† 8.Bxe2 Be7 144
B21) 9.0-0 145
B22) 9.Nc3 c6 10.0-0-0 Na6!? 146
B221) 11.Rhe1 147
B222) 11.Rde1 Nc7 12.Bd1 Ne6 13.Bd2 Bd7 14.h3 0-0-0 149
B2221) 15.Ng5!? 150
B2222) 15.g4 152

 
1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4 5.Qe2
White forces Black’s queen to e7, which typically leads to an early exchange of queens. This
allows White to develop quickly, castle long and try to seize the initiative by advancing his kingside
pawns and gaining space.
 
5...Qe7

280
 We will analyse A) 6.Nc3 followed by the more popular B) 6.d3.
 
6.d4
This rare move was once tried by Jobava. Black should simply respond with:
6...d5
Reinforcing the knight on e4.
7.Bf4
7.Nbd2 Nc6 sees Black develop actively, exerting pressure in the centre. 8.c3 Bf5 With ...0-0-0
coming next, Black was fine in Voom – Hargreaves, email 2008.
The text move was played in Insalata – Messina, Cortina d’Ampezzo 2002. I think Black
should respond with:

7...Nc6N
Black develops the queenside pieces in order to prepare long castling.
8.Nc3!
An important move, leading to simplifications.
In the event of 8.c3 Bg4 9.Nbd2 0-0-0³ White still needs to find a safe place for his king.
8...Nxc3 9.Qxe7† Bxe7 10.bxc3 Bf5
The endgame is equal after:
11.Bd3 Bxd3 12.cxd3 0-0-0=
 

A) 6.Nc3
 
This is not such a popular way of handling the position, but it was tried by Kramnik in the 2018
Candidates tournament. White opts to exchange knights instead of kicking the black knight away

281
from e4.
 
6...Nxc3 7.dxc3
White is ready to play Bc1-e3, avoiding the queen exchange, followed by 0-0-0. Hence, Black
should exchange the queens without delay.

7...Qxe2† 8.Bxe2 Nc6 9.Be3 Be7 10.0-0-0 0-0

An equally playable alternative is 10...Bg4 followed by ...0-0-0. The engine evaluates it as equal,
but of course there is a lot of play left.
 
11.Rhe1
This was Kramnik’s choice. Two other moves deserve close attention:
 
11.Rde1
Here Black’s play can be improved with:

282
11...Be6!N
In the absence of the rook on d1, the bishop can come to d5 if needed. Black also has ideas like
...a6 and ...b5, or simply bringing a rook to e8.
11...Bf6?! was inaccurate in Kovalev – Saduakassova, Wijk aan Zee 2019. The bishop is not so
well placed on f6 here, as it may become a target for White’s mobile kingside pawns.
12.h4
Gaining space on the kingside is the typical way for White to handle such positions.
12.Nd4 is simply met by 12...Nxd4 13.cxd4 Rfe8= and Black is solid.
12.Ng5 Bd5 13.Bf3 Bxf3 14.Nxf3 Rae8 is also equal.
12...Rae8 13.h5 h6 14.Nd4
14.Rhg1, with the idea to expand on the kingside with g2-g4, can be met by 14...f5!=
preventing it; Black is fine.
14.Nd2 has the same idea, and once again 14...f5!= is fine for Black.

283
14...Bd7 15.Nb5
15.Bf3 Bf6 also leads to a level game.
15...Bd8=
Black has no problems.
 
11.Rhg1
Directly preparing g2-g4 is another logical choice. I believe it’s best to continue developing in
the typical way with:

11...Be6!N
As we saw in the line above, the bishop is well placed on this square.
11...a6 was playable but a bit less purposeful in Artemiev – Khalifman, Moscow 2018.

284
12.g4
White wants to gain further space on the kingside with h2-h4-h5, but Black can fight back with:
12...f5!
This countermeasure on the kingside gains in strength with the rook on f8, which is why I
prefer not to move it to the e-file too soon.
13.g5 f4 14.Bd2
14.Bd4 is met by 14...Rae8, bringing the last piece into action. 15.h4 Bf7= Black is fine.
14...d5!
Preparing to get active with ...Bc5.
15.Nd4 Nxd4 16.cxd4
Black has various ways to continue from here, one of the more interesting options being:
16...f3!? 17.Bd3 Bh3!=
Preventing h2-h4 and fixing the pawn on h2. Black intends to play ...Bd6 next, with complex play
and chances for both sides.

11...Bf6 12.Nd4!?N
12.Nd2 has the idea to jump to e4, but it was easily parried by 12...Re8 13.Bf3 Ne5 14.Bf4 Kf8=
when Black was comfortable in Kramnik – Caruana, Berlin 2018.
 
The text move is a natural improvement, with the idea to play Bf3, putting pressure on the queenside.
Black should respond with:
 
12...Nxd4!
There is another way to go, but it is slightly inferior: 12...Bd7 13.Bf3 Rfe8 14.Nxc6 Bxc6 15.Bxc6
bxc6 16.Kd2² Black should certainly hold this endgame; nevertheless, White has a slight pull
because Black’s pawns on the queenside may become weak.
 
285
13.Bxd4
Now Black can force White to weaken his pawn structure with:
 
13...Bg5† 14.Kb1 c5! 15.Be3 Bxe3 16.fxe3 Rd8 17.Rd2 Kf8 18.Red1 Ke7=
Black has a weak pawn on d6 but White’s structure is also less than ideal, so a draw seems likely.

B) 6.d3

This is the most popular move by far. White kicks the knight away from the centre and prepares to
develop the c1-bishop.
 
6...Nf6
We will consider B1) 7.Nc3 and B2) 7.Bg5.
 
7.Be3 prevents the exchange of queens but gives Black a convenient opportunity to develop his
bishop to the desirable g7-square. 7...g6 8.Nc3 Bg7 9.0-0-0 0-0 Now 10.h3?! is an instructive
inaccuracy: preparing to gain space on the kingside with g2-g4 is a typical plan in such positions, but
it’s too slow and with the bishop on g7 Black doesn’t need to fear White’s usual kingside action.
10...Nc6 11.g4 Now in Vishnu – Gelfand, St Petersburg (blitz) 2018, Black could have seized the
initiative with:

286
11...Be6!N With the idea of ...Nd5, exchanging the c3-knight. For example, 12.Bg2 Nd5 13.Nxd5
Bxd5µ and Black’s bishops exert strong pressure on White’s queenside – especially the g7-bishop,
which has greatly benefited from the knight trade.
 

B1) 7.Nc3
 
This move is playable but it shares a slight drawback with the line in the note above: it allows Black
to fianchetto his bishop to g7, where it will be better placed than on e7.
 
7...Qxe2† 8.Bxe2 g6
Caruana played this way twice in 2018. White has tried various moves but the main ideas for both
sides remain pretty constant.
 
9.Nd4
This was Carlsen’s choice.
 
9.Bg5 prepares to park the king on c1. 9...Bg7 10.0-0-0 0-0 11.Rhe1 h6 12.Bh4 Nc6 13.d4 a6 14.a3
Bd7= Black has developed his pieces comfortably and was doing well in So – Caruana, Berlin 2018.
 
9.h3 was recently played by Topalov. 9...Bg7 10.Nd4 Whenever White places his knight here in this
variation, Black responds in the following way: 10...Bd7 11.Bf3 Nc6 Now after 12.Be3 in Topalov –
Dominguez Perez, Saint Louis 2019, the simplest route to equality would have been:

287
12...Nxd4N 13.Bxd4 Bc6= Black has no problems.
 
9.Nb5 is also nothing special as it can be simply met by 9...Na6. After 10.Be3 Bd7 I analysed:
11.c4!?N (this seems slightly more enterprising than 11.d4 c6 12.Na3 Nc7 13.0-0 Ncd5= when
Black was completely fine in Khukhashvili – Harika, Khanty-Mansiysk 2018)

11...Bg7 12.0-0 (12.Nxa7? is of course bad due to 12...c6µ when the knight is unlikely to see
daylight) 12...0-0 13.d4 Rfe8 14.Rfe1 Ne4= Black is doing fine, and she always has the option of
taking on b5 at a suitable moment.

288
9...a6
Covering the b5-square.
 
10.Bf4
Shirov tried to be aggressive with: 10.h3 Bg7 11.g4 Bd7 As usual in this variation, Black prepares
...Nc6. 12.Bf3 Nc6 13.Nxc6 Bxc6 14.Bxc6† bxc6 15.Bd2

15...h5! Undermining White’s extended kingside. 16.g5 Nd5= Shirov – Kramnik, Cazorla 1998. The
position is objectively equal but I think Black has a slight practical edge in this endgame.
 
10...Bg7 11.h3 Bd7

289
 Once again, we see Black following the prescribed plan.

12.Bf3 Nc6 13.Nxc6 Bxc6 14.Bxc6† bxc6 15.0-0-0 Kd7=

Black had no ill effects from the doubled c-pawns in Carlsen – Caruana, Wijk aan Zee 2018.

B2) 7.Bg5

This is the most precise move order for White, as it more or less forces Black to commit either his
bishop to e7 (after trading queens on e2) or the b8-knight to d7.
 
7...Qxe2†

290
 7...Nbd7 is the other main option but I would prefer not to commit this knight so soon. As we will
see in variation B22 below, the knight has a better route available to participate in the struggle, and it
is worth losing a tempo (by giving White the developing move Be2 for free) to keep the queen’s
knight flexible.
 
8.Bxe2 Be7
White’s opening play has earned him the developing move Bc1-g5 for free, while he keeps the
advantage of the first move in an otherwise symmetrical position. This may sound uninspiring for
Black, but it is not all bad: the bishop is not doing much on g5 and Black may gain time by attacking
it.
 
We will analyse the timid B21) 9.0-0 and the more interesting B22) 9.Nc3.
The passive 9.Nbd2?! can bring no good to White, especially after 9...h6! when White cannot go
back to the desirable d2-square with his bishop. 10.Bh4 g5! (10...Nd5 was playable in Czebe –
Ilincic, Budapest 2018, but Black can play more ambitiously) 11.Bg3 Having driven White’s bishop
to a poor square, Black should now focus on queenside development with:

11...Nc6!N (11...Nh5 was played in Hessel – Pramann, Oberhausen 1998, but Black should not
spend additional time exchanging the bishop which is already badly placed) 12.0-0-0 Be6 Black
develops another piece on its optimal square. 13.h3 0-0-0 14.d4 Rhg8³ All of Black’s pieces are
better placed than their counterparts and he has excellent chances on the kingside, especially playing
against the g3-bishop.
 

B21) 9.0-0

291
 Castling on the kingside cannot be too challenging for Black, as now White cannot hope to create
play on the kingside by freely advancing his pawns.
 
9...h6
Forcing White to declare his intentions with the bishop.
 
10.Bd2
10.Bf4?! proved to be inferior after 10...Nc6 11.Re1 and now Black seized the initiative with:
11...g5! 12.Bc1 Be6 13.c3 0-0-0³ Black was slightly better due to his space on the kingside in
Kristjansson – Istratescu, Rogaska Slatina 2011.
 
10...0-0 11.c4
White gains some space and covers the d5-square before playing Nc3.
11.Re1 is simply met by 11...Re8 12.Bf1 Nc6 13.g3 Bd7= when Black had completed his
development and was completely fine in Burg – Fridman, Belgium 2018.
 
11...d5
Breaking in the centre without delay.
 
12.Nc3 Rd8 13.Rfe1
Now Black developed his knight to a6 in Tabatabaei – Jobava, Bandar e Anzali 2017, but I prefer
the following straightforward approach:

292
13...c6N
Bolstering the d5-pawn.
 
14.Bf1 Bf8 15.h3 Na6=
Black has a safe position with some chances to seize the initiative, due to White’s slightly
weakened queenside.
 

B22) 9.Nc3

293
 White develops his fourth and final minor piece, and is now ready to castle on either side.
Meanwhile, Black is solid and should focus on arranging his pieces optimally.
 
9...c6
Covering the d5- and b5-squares, and also freeing the c7-square for the knight.
 
10.0-0-0
Castling long is of course the most challenging way to play.
 
Sometimes White starts with 10.h3, but after 10...Na6 11.0-0-0 the play quickly transposes to normal
paths.
 
10.Ne4 leads to simplifications after: 10...Nxe4 11.dxe4 Bxg5 12.Nxg5 h6 13.Nf3 Ke7 14.0-0-0
Rd8= Black is completely fine and the players agreed a draw just one move later in Svidler –
Grischuk, Moscow 2009.
 
10.0-0 was played in some old games but is less challenging, as White deprives himself of the option
of hurling his kingside pawns up the board. Black can continue with his typical manoeuvre: 10...Na6
11.Rfe1 Nc7 12.Bf1 (12.Nd4 can be met by 12...Ne6 13.Be3 g6, covering the f5-square, and after
14.Bf3 Bd7 15.Nxe6 Bxe6= Black had no problems in Sadvakasov – Timman, Hoogeveen 1999)
12...Ne6 13.Be3 0-0 14.d4 This position was reached in Spassky – Karpov, Turin 1982. White
intends to play d4-d5 next, so a logical reply is:

14...Nc7N Controlling the d5-square and freeing the way for the bishop to come to f5. 15.Bd3 Re8
16.h3 Bd7= Black is doing fine.

294
10...Na6!?
This is a typical way of developing the knight in many variations of the Petroff. The knight heads
for the e6-square via c7.
We will analyse two main options: B221) 11.Rhe1 and B222) 11.Rde1.
 
He can also start with:
11.h3
White just makes a useful move before committing one rook or the other.
11...Nc7
Black of course continues with his plan. Whichever rook White places on the e-file, Black will
react exactly as he does in the respective main lines below. Sometimes White ignores the e-file
entirely, but this is also no problem for us.
12.g4N
12.Rde1 Ne6 13.Bd2 Bd7 14.Bd1 transposes to variation B222.
12.Rhe1 is even less threatening than in variation B221 below, as White is too slow to create
meaningful pressure along the e-file and h2-h3 is not particularly useful. After 12...d5 13.Bf1
Ne6 14.Be3 0-0= Black had an easy game in Anisimov – Sadhwani, Internet 2019.
12...Ne6 13.Be3 0-0
Black is doing fine, for example:

295
14.Nh4 g6
Covering the f5-square.
15.f4
White is ready to seize the initiative on the kingside with f4-f5, so Black should play
15...Ne8! 16.Nf3 f5!³
White’s overextended pawns on the kingside will become a target.
 

B221) 11.Rhe1

296
 This move is usually associated with the idea of Ne4, provoking an exchange on that square and
then playing against the d6-pawn.
 
11...d5!
I recommend preventing White’s idea before resuming our plan of bringing the knight to e6.
 
11...Nc7 is the main move, but it allows White to execute his idea with 12.Ne4! Nxe4 13.dxe4
Bxg5† 14.Nxg5 Ke7 15.f4 Ne6 16.Nxe6 Bxe6 17.f5² when White was able to put some pressure on
the d6-pawn in M. Novikov – Qashashvili, Rilsk 2017. A single weakness should be manageable and
Black should hold the endgame without major problems – but still, I prefer to prevent White’s plan
altogether.
 
12.Bf1
Centralizing the knight with 12.Ne5 can be simply met by 12...Nc7 13.Bh4 Ne6 14.d4 0-0= when
Black is fine and the knight on c3 is misplaced.
 
After 12.d4 Nc7 it is logical for White to try:

13.Bd3 (13.Bd2 was hardly scary in Khan – Guendogdu, corr. 2012) 13...Ne6 14.Bh4 0-0 Black was
fine in Grandins – Jegorovas, corr. 2005. Blac intends ...Re8 and ...Bd7, with a fine position.
 
12.Nd4 is also nothing special. Black can simply go for: 12...0-0 13.Bf3 Bd8 (13...Re8?! 14.Bxf6
gxf6 looks slightly uncomfortable for Black) 14.Nce2 White’s last move took away a retreat square
of the f3-bishop. In Ninov – Popchev, Bankia 2011, Black could have exploited this with:

297
14...h6!N Exploiting the fact that the bishop cannot go to f4 or h4 due to ...g5-g4, trapping the other
bishop. 15.Bd2 And now the strong follow-up 15...Ng4 16.Bxg4 Bxg4 leaves Black with the bishop
pair, and with it the more comfortable side of the near-equal position.
 
The text move stops Black castling, but he can simply continue his plan with:

12...Nc7!
On this occasion, Black should not fear the doubling of his pawns.
 
13.Bxf6 gxf6

298
 Black’s structure has been weakened but he has the two bishops as compensation, and White’s
pieces are far from ideally placed.
 
14.Ne2 Bg4 15.Ned4 Ne6 16.Rd2?!
We have been following Kontic – Petijevic, Bari 2017, in which Black inexplicably played
16...Bxf3?. A much saner continuation would be:

16...Kd7N
The king is ideally placed here. A logical continuation might be:
 
17.h3 Bh5 18.g4 Bg6 19.Rde2
White logically increases the pressure along the e-file, but Black has a good way to counter it:
 
19...Bb4! 20.c3 Nxd4 21.Nxd4 Bd6³
Only Black has any real winning chances. His bishops are strong and he can open things up for his
rooks with ...h5.
 

B222) 11.Rde1

299
 It is better to bring this rook to the open e-file rather than the h1-rook, because the latter will be
useful when White starts pushing his kingside pawns.
 
11...Nc7
11...h6 has been played in some games, but it is better to gain time against the g5-bishop by
bringing the knight to e6.
 
12.Bd1
This is the most logical sequel to White’s previous move. White tucks his bishop out of the way
and hopes that his rook will cause some annoyance along the e-file, before advancing his kingside
pawns, after which the other rook may come into its own.
 
12.Ne4 has been played but, unlike the equivalent position from variation B221, Black can simply
exchange on e4 and g5 without his d6-pawn coming under fire, so he is easily equal.
 
An older try for White is:
12.Nd4
White gets ready to exchange the knight on e6, while also unblocking his f-pawn.
12...Ne6 13.Nxe6 Bxe6 14.f4
White starts expanding on the kingside.

300
14...d5!
Black can afford to take a tempo to improve his position in the centre.
15.Bf3
This stops Black from castling but, without queens on the board, this is of no great concern.
The overaggressive 15.g4?! is met by: 15...h6 16.Bh4 g5! An important move, halting White’s
progress on the kingside. 17.fxg5 Nd7 18.Bg3 hxg5³ Followed by ...Nf8-g6, with excellent
chances for Black.
15.Bxf6N 15...Bxf6 16.Bg4 is not a problem in view of 16...0-0 17.Bxe6 fxe6 18.Rxe6 Bxc3
19.bxc3 Rxf4= with a completely equal rook endgame.
15...Kd8 16.Ne2 Re8=
In Timman – Karpov, Tilburg 1980, Black’s pieces were well coordinated and he was doing fine.
 
12...Ne6
The knight is well placed on this central post, where it controls some important squares and limits
the scope of the rook on e1.
 
13.Bd2
White has retracted both his bishops, and is now ready to advance on the kingside, while Black still
needs two tempos to take his king to safety.
 
13...Bd7
Most of the action is going to take place on the kingside, so it is better to castle long.
 
14.h3 0-0-0
Now there are two critical ways for White to continue: B2221) 15.Ng5!? and B2222) 15.g4.

301
 

B2221) 15.Ng5!?

White takes the opportunity to attack the f7-pawn, winning a tempo to advance his f-pawn, while
also highlighting the fact that the e7-bishop is undefended.
 
15...Rhf8 16.f4
White threatens f4-f5, so it is essential to defend the e7-bishop.
 
16...Rde8 17.Nf3!?
This is the most ambitious choice: White avoids exchanging pieces and is now ready for f4-f5
followed by g2-g4.
 
I also considered:
17.Rhf1N 17...d5
With the idea of ...Bd6, improving the placement of the bishop.
18.Nxe6 Bxe6 19.g4
19.Ne2 Bd7 20.f5 Bd6 21.g4 transposes.

302
19...Bd6 20.f5 Bd7 21.Ne2
White improves his knight.
The overaggressive 21.g5? is too risky, as White’s advanced pawns will turn out to be a liability
after: 21...Rxe1 22.Bxe1 Ng8 23.h4 f6! The key move, leading to big problems for White.
24.g6 Nh6 25.gxh7 Rh8µ Black will regain the h7-pawn, leaving White’s remaining pawns on
the kingside very weak.
21...Re7=
Intending ...Rfe8. Even though White has managed to gain space on the kingside, Black has good
play along the e-file. White has no advantage and he should be mindful of the fact that his advanced
kingside pawns have the potential to become weak.
 
17...Nc7
Black stops f4-f5 for the moment and plans ...Nfd5 followed by ...f5.
 
18.Ne2
White can also start with a pawn advance:
18.g4 Nfd5
Freeing the way for the f7-pawn.
19.Ne2
I also checked 19.Nxd5N 19...Nxd5 20.c4 Nc7 21.f5, when it is advisable to break with
21...g6!= before White can build upon his extra kingside space, and the position remains
balanced.
The text move was played in V. Onischuk – Semcesen, Lvov 2009. White continues his strategy
of not exchanging pieces, but Black would have been fine after:

303
19...f5!N
An important move, preventing White from gaining additional space with f4-f5.
20.Ng3
20.g5 is met by the simple 20...Bd8= with the knight coming to e3, when Black is doing fine.
20...fxg4 21.hxg4 Bxg4 22.Rxh7 Nxf4 23.Rxg7 Bxf3=
All the pawns on the kingside have disappeared, and the position liquidates into equality.
 
18...Bd8 19.g4
If White is given time to put his knight on g3, he will be perfectly coordinated and it will be hard to
stop his play on the kingside. Thus, the following novelty is quite a significant improvement:

304
19...h5!N
19...c5?! 20.Ng3² gave White a pleasant advantage in Bruzon Batista – Dragun, Burlingame 2019.
 
20.g5 Nfd5
By forcing the g4-g5 advance, Black has rendered his opponent’s pawn structure more rigid.
Another benefit is that the d7-bishop enjoys more scope.
 
21.Ng3 Rxe1 22.Nxe1 h4!
Fixing the h3-pawn as a target for Black’s light-squared bishop.
 
23.Ne4 Ne8 24.Bg4
24.Nf3 is met by 24...Rh8= followed by ...Ne7-f5 when Black has nothing to fear.

24...Ne7
24...f5!? is another interesting possibility.
 
The text move takes control over the important f5-square. Play may continue:
 
25.Nf3 Rh8 26.Be1 Nd5=
With a fine position for Black.
 

B2222) 15.g4

305
15...Rde8
It is useful to defend the e7-bishop while also fighting for the e-file.
 
16.Ne2
I also considered:
16.g5
Black retreated the knight to g8 in Gabrielian – Lintchevski, Kazan 2018. A simple
improvement is:
16...Nh5!N
It is important to keep the knight active, and Black can follow up with ...f6. Perhaps in the
above game Black was worried about placing the knight in the firing line of the bishop on d1,
but White has no way to exploit this and the knight may jump to the f4-square.
17.Ne2
17.Nh4 Nef4 18.Bg4 g6 is good for Black, who can follow up with ...f5.
17.h4 g6 18.Ne4 Rhf8= and Black is ready to break with the f-pawn.
17...f6 18.h4 Bd8=
Black’s pieces are well placed and White cannot advance further on the kingside.
 
16...g6
This is a useful preparatory move before ...h5.
 
16...Bd8N is also perfectly playable.
 
17.Rhg1 Bd8
Black is arranging his pieces ideally.

306
 
18.c4 h5!
A typical thrust on the kingside, challenging White’s advanced pawns.
 
19.g5 Nh7 20.h4
We have been following Demchenko – Visakh, Konya 2018. This would have been the perfect time
to carry out the thematic pawn break:

20...f6!N
Giving life to Black’s minor pieces, while opening the f-file for the rooks.
 
21.d4
 
The main justification of Black’s previous move is that the obvious 21.gxf6?! Nxf6 22.Rxg6
enables Black’s pieces to springs to life:

307
22...Ng4 23.Rf1 Nc5! 24.Rxd6 Bf5ƒ Black has a strong initiative for the two sacrificed pawns, while
White’s pieces are passive and uncoordinated.
 
21...fxg5 22.hxg5

22...Rhf8 23.Nc3 Nf4=

Black is doing well: the g5-pawn is under pressure and the protected passer on h5 is an asset which
is like to become stronger as more pieces are exchanged.
 

Conclusion
 

308
The 5.Qe2 Qe7 variation almost always implies an early queen exchange. We started with the
slightly unusual 6.Nc3 line, which should not be underestimated, although Black should be fine with
accurate play.
 
The main line continues 6.d3 Nf6 7.Bg5 Qxe2† 8.Bxe2 Be7, when we looked at a number of
options, but White’s only truly challenging plan is to castle on the queenside and prepare to advance
his kingside pawns. Thus, 9.Nc3 c6 10.0-0-0 Na6!? 11.Rde1 Nc7 is our main line, when Black
intends to place his knight on e6 and castle on the queenside. The ensuing queenless middlegame
holds chances for both sides. From Black’s perspective, the most important thing is to counter
White’s f4-g4 pawn advance before his kingside initiative gathers too much steam. This can often be
achieved with a timely ...h5, when g4-g5 makes some squares available for Black’s pieces, as well as
giving him the option of ...f6 as a subsequent pawn lever to open lines for counterplay.

309
A) 9.cxd5 157
B) 9.Re1 Bf5 159
B1) 10.Qc2 Na6 11.a3 Bg6 12.c5 Bb8 13.Nc3 f5 14.Qb3 Bh5! 160
B11) 15.Bxe4 161
B12) 15.Nxe4 162
B2) 10.Qb3 Qd7 11.Nc3 Nxc3! 12.Bxf5 Qxf5 164
B21) 13.bxc3 165
B22) 13.Qxb7 166

 
1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4 5.d4

310
 This move is the clear main line in terms of the overall number of games played with it.
 
5...d5
Sometimes Black prefers to play with a ‘small centre’ but I prefer the traditional approach of
gaining space and supporting the knight on e4.
 
6.Bd3
This is clearly the most active and logical way of developing the bishop in preparation for castling
followed by c2-c4, undermining the knight on e4. Other moves exist of course, but none of them
require any special preparation. The entire point of 5.d4, leaving the knight on e4 – as opposed to
5.d3 Nf6 6.d4, as covered in Chapter 6 – is to treat Black’s ‘active’ knight as a target to be
undermined, and putting the bishop on d3 is the only effective way of pursuing that goal.
 
I will mention briefly that 6.c4?! Nc6 transposes to variation B2 of Chapter 5 on page 112, but
6...Bb4†! is more accurate, when Black is already fighting for the advantage.
 
6.Nbd2 can be met by 6...Bf5 when White hardly has anything better than 7.Bd3 Bd6, transposing to
7.Nbd2 Bf5 in the notes below.
6...Bd6
6...Be7, 6...Nc6 and even 6...Bf5!? are all serious alternatives, but the text move is the one I have
chosen to recommend.
 
7.0-0
7.c4 c6 is bound to transpose to one of the lines below, as both sides will surely castle in the near
future.

311
 7.Nbd2 Bf5 8.0-0 0-0 transposes to the note on 8.Nbd2 below.
 
7...0-0 8.c4
This is the most popular and principled move. I checked three sidelines:
 
8.Nbd2 should be met by 8...Bf5, developing while supporting the knight on e4. 9.Re1 Re8 10.Nf1
Bg6 The bishop retreats to a safer square. 11.Ne3 c6

12.c4 White tries to shake the stability of the central knight but 12...Nf6 is a convenient reply, and
after 13.Bxg6 hxg6 14.Qb3 Qb6= Black had no problems in Anand – Caruana, Saint Louis 2018.
 
8.Re1 Bf5
This is our typical reaction whenever the e4-knight is attacked.
9.Nc3!?
9.Nbd2 transposes to the 8.Nbd2 line above.
9.c4 c6 transposes to variation B.
9...Nxc3 10.bxc3 Bxd3
10...Qd7 and 10...Bg6 seem playable as well.
11.Qxd3
11.cxd3 Re8=
11...Nd7 12.Rb1

312
 In Carlsen – Shirov, Moscow 2006, the simplest continuation would have been:
12...Re8!N=
White cannot really justify taking on b7 as his rook will be cut off after ...Nb6, so Black has no
problems.
 
8.Nc3
This move has also been tried by some strong players but it is not at all challenging for Black.
8...Nxc3 9.bxc3 Bg4
This pin can be annoying for White.
10.h3 Bh5 11.Rb1 b6 12.Re1
12.c4 is met by 12...c6 13.cxd5 cxd5 14.c4 dxc4 15.Bxc4 Nc6= followed by ...Rc8 and Black
was doing well in Obregon – Proleiko, Internet 2018.

313
12...c6
Black intends to complete development with ...Nd7 followed by ...Qc7.
13.c4
White has to go for this break sooner or later. If he delays it, Black may consider ...b5 at some
point, eliminating the possibility of c3-c4 once and for all.
13...dxc4 14.Bxc4 Nd7 15.Bd3 Re8=
Black is fine and the players soon agreed a draw after:
16.Bg5 Qxg5 17.Nxg5 Bxd1 18.Rbxd1 Nf6 19.Ne4
½–½ Fedoseev – So, Tbilisi 2017.

8...c6

314
 White has tried several moves from here; the remainder of the current chapter will focus on A)
9.cxd5 and B) 9.Re1.
 
The two most popular and challenging options are 9.Qc2 and 9.Nc3, which will be covered in
Chapters 9 and 10 respectively.
 
I also considered:
9.Qb3 dxc4
This seems like the simplest reply.
10.Bxc4 Nd7
Intending ...Nb6 or ...Nf6.
11.Re1 Ndf6
11...Nef6!?N is also possible, and if 12.Ng5 Nd5 13.Bxd5 cxd5 14.Qxd5 Nf6 15.Qb3 h6
16.Nf3 Be6 17.Qd3 Qd7© Black has good compensation for the pawn.
12.Nbd2
12.Nc3 can be met by 12...Nxc3 13.bxc3 b5! 14.Bd3 Be6 15.Qc2 (or 15.Rxe6!? fxe6 16.Qxe6†
Kh8© when White has enough compensation to maintain the balance, but nothing more)
15...h6= and Black is fine.
12...Nxd2 13.Bxd2 Qb6!
Breaking the battery on the a2-g8 diagonal.
14.Qd3
Now an important improvement is:

14...Bg4!N
Completing development as quickly as possible.
14...Qxb2? allows White to gain a strong initiative, and he went on to win a spectacular
attacking game in Paravyan – Golubov, St Petersburg 2018.

315
15.Ne5 Rad8!=
Black is doing fine in this IQP position.
 

A) 9.cxd5 cxd5
 
I find it hard to believe that releasing the tension so soon can offer White real chances for an
advantage.
 
10.Nc3
10.Qb3 Nc6 11.Nc3 gives Black a choice between 11...Re8, transposing to the 11.Qb3 line in the
notes below, and 11...Nxc3, which is also perfectly playable.
10...Re8!?
10...Nxc3 has been more popular but I prefer to support the knight rather than exchange it, for the
moment at least.
 
11.Qc2
11.Nxd5?! Bxh2† 12.Kxh2 Qxd5 13.Re1 Nc6 gave Black the easier game due to his safer king
and pressure against the d4-pawn in Van Tricht – Hofer, corr. 2014.
 
11.Re1 invites simplifications with 11...Nxc3 12.bxc3 Rxe1† 13.Qxe1 Nc6= when Black was fine in
Lagno – Koneru, Doha 2016.
 
11.Qb3 Nc6! is an important detail, when White should try:

316
12.Re1N (12.Nxd5?! is bad due to 12...Bg4!³ and Black seizes the initiative, Sandaogo –
Oppermann, corr. 2005) The text move puts pressure on the e4-knight and forces Black to take a
decision, but the cool 12...Be6! makes it impossible for White to take on e4, and after 13.Qxb7 Nb4
the game is likely to end in a repetition after: 14.Bxe4 Rb8 15.Qxa7 Ra8=
 
11...Bf5
Black defends his knight in the usual way and intends ...Nc6 next.
 
12.Qb3
12.Re1?! is met by 12...Nc6! threatening ...Nb4. 13.a3 Rc8

14.Be3N (14.Qb1?? ran into the crushing 14...Nxd4!–+ in Kunschek – R. Fritz, Vienna 2012)
14...Bg6³ All the black pieces are well placed, especially the knight on e4.
 
The text move hits both d5 and b7, but Black can calmly ignore both attacks with:

317
12...Nc6! 13.Qxd5?!
This seems like the most principled follow-up to White’s previous move, but the whole concept is
flawed.
 
13.Qxb7?! would also be a mistake due to 13...Nb4!, threatening not just the d3-bishop, but also
...Re7 trapping the queen. 14.Bxe4 dxe4 15.Ne5 Bxe5 16.dxe5 Nd3³ Black’s pieces are too active.
13.Be3 is the best that White can do, although 13...Nd2! 14.Nxd2 Bxd3 still reached a position
where Black had the slightly more comfortable side of equality due to his bishop pair in Dunn –
Shchepetnev, email 2002.
 
13...Bg6
There are too many threats for White to handle comfortably.
 
14.Bxe4
14.Qb3? does not work due to 14...Nxd4! 15.Nxd4 Nc5µ.
 
14...Bxe4 15.Qh5
Now we have a chance to improve Black’s play with:

318
15...g6!N
Black took on f3 immediately in Rosch – Portune, email 1997, but it is worth making a slight
improvement to the kingside structure first.
 
16.Qh3 Bxf3 17.Qxf3 Nxd4ƒ
Black has a nice initiative as White still needs to develop his queenside.
 

B) 9.Re1

319
 This move is consistent with White’s aim of targeting the e4-knight. Over the next few moves, he
will continue to put pressure on this piece, and possibly the black queenside as well.
 
9...Bf5
White’s two main options are B1) 10.Qc2 and B2) 10.Qb3.
 
10.Nc3 is a harmless move which leads to nothing but piece exchanges: 10...Nxc3 11.bxc3 Bxd3
12.Qxd3 dxc4 13.Qxc4 Nd7

Black’s position is fine, and after the further 14.Rb1 Qc7 15.Qb3 b5 16.c4 bxc4 17.Qxc4 c5
18.dxc5 a draw was agreed in Anand – Caruana, Saint Louis 2016.
 
10.cxd5 cxd5 11.Nc3
White has tried this in a few top-level games. Black is of course happy to trade some pieces
with:
11...Nxc3 12.bxc3 Bxd3 13.Qxd3
It is important to note that White only has one truly threatening idea, which is to play Ng5,
intending to meet ...g6 with Qh3, forcing a further weakening of Black’s kingside with ...h5.
The simplest way to prevent this is:
13...Re8!
Challenging White along the e-file.
14.Bg5
14.Rxe8† Qxe8 15.Rb1 is simply met by 15...b6 16.Qb5 Qe6. It is worth adding that 17.Bf4!?
is a creative attempt to get the rook to the e-file with gain of tempo, but 17...Bf8!= keeps things
solid, with a fine position for Black.
14...Rxe1† 15.Rxe1 Qd7 16.Ne5

320
 White has to create some imbalance before Black can catch up in development and exploit the
weaknesses on White’s queenside.
16...Bxe5 17.dxe5 h6

18.e6!?
A clever tactic, but it does not give White any advantage.
18...fxe6 19.Bxh6 Nc6=
The players soon agreed a draw in Blanco Gramajo – Straka, corr. 2014.
 

B1) 10.Qc2 Na6

321
 We have seen the knight being developed to a6 to good effect in a number of the earlier chapters,
and here it has the additional benefit of winning a tempo thanks to the threat of ...Nb4.
 
11.a3 Bg6
The bishop moves to a safe square, ensuring that Nc3 will not be a problem.
 
12.c5
12.Nc3?! simply offers to exchange two minor pieces in a way that cannot possibly benefit White:
12...Nxc3 13.Bxg6 hxg6³ followed by ...dxc4 and White was positionally worse in Galliamova –
Kosteniuk, Kazan 2004.
 
12...Bb8
Black should choose this square for the bishop in order to keep c7 free for the knight.
 
13.Nc3
The other way to challenge the e4-knight is:
13.Nbd2
I find this less logical than the main line, as the bishop will be blocked on c1. Black of course
responds with:
13...f5
Cementing the knight on e4.
14.Qb3 Rf7
Black has other possible ways to continue but I would prefer this solid move.

15.Bxa6

322
 15.Qa4 was played in a blitz game between two strong engines. White forces the a6-knight to
go back to c7 which will allow Ne5. However, after 15...Nc7 16.Ne5 Rf8 17.Nxg6 hxg6 Black
has good attacking prospects on the kingside, as shown after 18.Be2 Ne6 19.Nxe4 fxe4‚ in
Rybka 4.1 – Booot 6.2, engine game 2017.
15...bxa6
Black’s damaged queenside structure is of little concern, as he has two bishops and nice play on
the kingside.
16.Qa4 Qc8 17.b4 f4„
Black’s play on the kingside more than compensates for the weakened queenside pawn structure,
Williamson – Straka, corr. 2017.

13...f5
Black provides full support to his centralized knight. The critical reply is:
 
14.Qb3
14.b4? was a mistake in Holemar – Bartos, Brno 2007, when 14...Nc7!N intending ...Ne6 and
...Bh5 would have given Black great play.
 
14.Ne2 Bh5„ gives Black nice prospects on the kingside, especially after 15.Bf4? Bxf3 16.gxf3
Nxf2!µ as seen in Aguero Jimenez – Ippolito, Barcelona 2018.
 
14...Bh5!
Black sacrifices a pawn and goes for a direct assault on the kingside.
 
White has to accept the challenge by capturing e4 twice, which he can do by means of B11) 15.Bxe4
or B12) 15.Nxe4.

323
 
15.Qxb7?? just loses to 15...Bxf3, and if 16.gxf3 Qh4–+ Black will crush on the kingside.
 

B11) 15.Bxe4 fxe4 16.Nxe4

Black should weaken the white king’s defences with:

 
16...Bxf3 17.gxf3 Qd7
The queen is well placed here, supporting the b7-pawn while being ready to shift to the kingside.
 
18.Ng3
White settles his knight on an excellent defensive square. White is a pawn up but the doubled
pawns on the f-file do not add much value to his position.
 
18.Nd6? would be a misguided attempt to provoke the exchange of Black’s strong bishop. 18...b6! is
a strong reply, and after 19.Be3 Rxf3µ Black wins back the pawn while White’s kingside is too weak.
 
18.Ng5? with the idea to place the knight actively on e6 is also bad. Play continues 18...h6 19.Ne6
Rf6 20.Bd2 and now a strong move is:

324
20...Bc7!µ Black is ready to bring the a8-rook into action, and the b7-pawn is untouchable due to
...Bxh2†.
 
Now Black has two ways to continue, my favourite being:

18...Bf4!
Black offers the exchange of bishops, but the main purpose of the text move is to bring the a8-rook
into action.
 
18...Nc7!? is an interesting dynamic possibility, whereby Black sacrifices another pawn in order to
bring the knight to e6. 19.Qxb7 White should accept the challenge, otherwise the knight comes into

325
action free of charge. 19...a5! Black wants to play ...Ra7, driving the enemy queen away in order to
facilitate ...Ne6. 20.Bg5 Ra7 21.Qb3 a4 Fixing White’s queenside pawns. (The immediate 21...Ne6
is not fully convincing after 22.Qe3.) 22.Qe3 White stops ...Ne6 for now. 22...Rb7©

Even though Black is two pawns down for the moment, he has strong compensation. The b2-pawn
is weak, Black has pressure down the b-file and ...Rb3 is on the cards as well, hitting the queen and
the f3-pawn. My illustrative line continues 23.Re2 h6 24.Be7 Rb3 25.Qd2 Rfxf3 26.Bd6 Ne6
27.Rae1 Rf6© and Black still has full compensation.
 
19.Re2
19.Be3?! is inferior as it allows Black to bring his knight into play: 19...Nc7! 20.Bxf4 Rxf4
21.Qxb7 Rff8 Threatening to trap the queen with ...Rfb8. 22.Qb3 Ne6³ The knight is ready to wreak
havoc on White’s kingside.
 
19...g5!?©
In Kratz – Rennert, email 2006, Black was ready to bring the a8-rook into play, either by putting it
on e8 or by doubling along the f-file.
 

B12) 15.Nxe4 fxe4

326
16.Bxe4
A riskier alternative is:
16.Rxe4?!
Black should do well against this move, as long as he takes on f3 with the correct piece:
16...Rxf3!
16...Bxf3? would be a serious mistake, as after 17.gxf3 Rxf3 18.Rg4! White’s rook is ideally
placed for both attack and defence. White threatens Bxh7†, so 18...Qf6 is logical, but after
19.Be3± White’s pieces are coordinating well while most of Black’s army remains stuck on the
queenside.
17.gxf3 Bxf3
You can see how important it was to take on f3 with the correct piece: the bishop on f3 is far
more effective than a rook on the same square, as it threatens the weak light squares around
White’s king as well as preventing the rook from moving to g4. The next few moves are more
or less forced for both sides.

327
18.Re1 Bxh2† 19.Kf1 Bg4!
The direct 19...Qh4?? is refuted by 20.Bxh7†.
Black now threatens ...Qh4, forcing White to play:
20.Be2 Bd7
Of course Black avoids exchanging his powerful light-squared bishop.
21.Kg2! Qh4 22.Bg5 Qxg5† 23.Kxh2 Rf8‚
Despite White’s accurate defence, Black retained excellent attacking chances in Rimkus –
Sneddon, corr. 2018.
 
16...Kh8 17.Bg5!
A vital intermediate move, stopping the black queen from landing on h4 with devastating effect.
 
17.Bc2? comes unstuck after 17...Bxf3 18.gxf3 Qh4 19.h3 and now the brilliant:

328
19...Bg3!! Black finds a way to bring his queenside pieces into action with tempo. 20.Qe3 Nc7
21.Kg2 Bf4 22.Qe7 Rf6! Of course Black doesn’t exchange queens. 23.Bf5 h5!–+ Guarding the g4-
square in preparation for ...Qg5†. White was unable to defend his kingside in Schmandt – Gokerman,
corr. 2012.

17...Qd7
The queen is ready to go to h3.
 
18.Bc2 Bxf3 19.gxf3 Qh3
Fortunately for White, he can save himself with:
 

329
20.f4!
Forcing the exchange of queens.

20...Qxb3
The players agreed a draw here in Postl – Ramik, corr. 2011. Play would have continued:
 
21.Bxb3 Bxf4=
The endgame is balanced.
 

B2) 10.Qb3

330
 Hitting the b7-pawn is the most critical move.
 
10...Qd7!
This may seem like a tactical oversight, but everything will work out fine for Black.
 
11.Nc3
I also checked a couple of less critical alternatives:
 
11.Be3 is a simple developing move which was once tried by Leko. Anand replied with the logical
11...dxc4 12.Bxc4 Qc7, vacating the d7-square for the knight. After the further 13.Nc3 Nxc3
14.bxc3 Nd7= Black completed his development and was doing fine in Leko – Anand, Leon 2001.
 
11.c5
This move releases the central tension and enables Black to focus on the kingside.
11...Bc7
Of course the bishop remains on its best diagonal.
12.Nbd2 Be6!
Black uses simple tactics to prepare ...f5, cementing the knight on e4.
13.Nf1
13.Qc2 f5 14.b4 allows 14...a5!, creating discomfort for White on the queenside.
After the text move, White has a positional threat of Ng5, effectively forcing Black to trade off
his active knight. I recommend a novelty to prevent that idea:

331
13...h6!N
Now Black is ready to play ...f5 whenever it becomes necessary.
13...f5?! 14.Ng5! was good for White in Belov – Kosteniuk, Kazan 2005.
13...Na6?! was played in Navara – Jakovljevic, Budva 2009, when 14.Ng5!N would have been
strong.
14.Ng3 Na6!
Black is ready to play ...Rae8 followed by ...f5, with good attacking chances on the kingside.
 
11...Nxc3! 12.Bxf5 Qxf5
White has two options: B21) 13.bxc3 is playable but harmless, while B22) 13.Qxb7 is the more the
challenging move.
 

B21) 13.bxc3

332
13...b6
White has tried various moves here but Black holds comfortably against all of them.
 
14.Bg5
White intends Be7, exchanging bishops and infiltrating on the 7th rank.
 
14.h3 dxc4 15.Qxc4 b5 16.Qb3 a5 (16...Nd7 also looks fine) 17.Ba3 a4 18.Qb2 Bxa3 19.Qxa3 Nd7
20.Ne5 Rac8 21.Nxd7 Qxd7 22.Re7 Qd5= and the players soon agreed a draw in Brkic – Papp,
Austria 2018.
 
Adams tried to manoeuvre his knight with 14.Nh4 Qh5 15.g3, intending Ng2-e3 with pressure on
d5. However, after the simple 15...Nd7 16.cxd5 cxd5 17.Ng2 Rfe8= Black was perfectly okay in
Adams – Papp, Gibraltar 2019.
 
I also considered:
14.cxd5 cxd5 15.Qb5
White tries to impede the development of the b8-knight.
Trading bishops with 15.Ba3 is harmless: 15...Bxa3 16.Qxa3 Nc6 17.Qd6 Rac8 18.Ne5 White
wants to win the d5-pawn, but he will lose c3 in return. 18...Nxe5 19.Rxe5 Qc2= With an equal
endgame, Liu Qingnan – Xu Xiangyu, China 2018.
15...Qd7!
Offering a queen trade while preparing to develop the queenside.

333
16.Rb1
16.a4 Rd8 17.g3 Bc7 18.Bg5 f6 19.Bd2 Nc6 was equal in Adams – Morozevich, Wijk aan Zee
2005.
16...Rc8 17.g3 Bf8 18.Bd2 Nc6
Black has successfully developed his queenside and his position is quite comfortable.
19.Ne5 Nxe5 20.Rxe5 Rd8 21.Qxd7 Rxd7 22.Rb5 Rad8=
The endgame was level in Vachier-Lagrave – Caruana, Stavanger 2017.

14...dxc4 15.Qxc4 b5!

It is useful to cover the light squares on the queenside, while also kicking the white queen away
from the active c4-square.

334
 
16.Qb3 Nd7
Black should not worry about the imminent infiltration on the 7th rank.
 
17.Be7 Bxe7 18.Rxe7 Rae8 19.Rae1
White’s activity can be neutralized with:

19...Nf6! 20.h3
White should take a moment to solve his back-rank issues.
 
20...Rxe7 21.Rxe7 Nd5!
Black doesn’t care for the a7-pawn, but instead focuses on creating kingside counterplay.
 
22.Rxa7 Nf4
Black’s initiative enables him to regain the pawn with ease, and the game was soon drawn after:
 
23.Qd1 Qg6 24.g3 Nxh3† 25.Kf1 Qe4 26.Qe2 Qd5

335
27.Qe3 h6 28.Re7 Ng5 29.Nxg5 hxg5 30.Re8 Qxa2
½–½ Naiditsch – Wang Hao, China 2019.
 

B22) 13.Qxb7

This is the more critical test of Black’s play, leading to an unbalanced pawn structure.
 
13...Qd7!
Black forces a queen trade.

336
 
14.Qxd7
Nobody has tried 14.Qxa8? yet, with good reason, as it results in the loss of the queen after:
14...Na4! Threatening ...Nb6. 15.c5 (15.cxd5 cxd5–+ followed by ...Nb6 wins the queen for Black.)
15...Na6 16.Qxf8† Bxf8µ Black is clearly ahead in material.
 
14...Nxd7 15.c5!
This is White’s best attempt to cause problems.
 
15.bxc3 dxc4 16.a4 was seen in Volokitin – Papp, Austria 2018, when 16...a5N= would have
prevented White from gaining further space on the queenside, with a comfortable position for Black.

15...Bxh2†!
Black should avoid 15...Bxc5?! 16.dxc5 Ne4 17.Nd4 Ndxc5 18.Nxc6² when the isolated d-pawn
and White’s superior minor piece will give Black some problems in this endgame.
 
16.Nxh2 Ne4
White has a queenside pawn majority and a bishop against a knight. However, it will be difficult
for him to create a passer after Black plays ...a6, and the bishop’s mobility is limited by White’s own
pawns on d4 and c5. Black will look to deploy his knights to optimal squares and the chances are
approximately equal.
 
17.Bf4
Covering the important b8-square.
 

337
Another game continued 17.Nf1 Rfe8 18.f3 Nef6 19.Rxe8† Rxe8 20.Bd2 Nf8 21.Ne3 Ne6 and
Black got a fine position in Grischuk – Safarli, Astana 2019.
 
17...Rfc8!?
Freeing the f8-square for the knight to travel to e6.
 
17...Rfe8 has also been played, with the same idea.

18.f3 Nef6 19.Re3

19.Re7N can be met by 19...Re8 20.Rae1 Nf8= followed by ...Ne6, when the active rook will be
exchanged and the position is equal.
 
19...Nf8 20.Rd1 Ne6 21.Be5 Nd7=
Having arranged his knights properly, Black had no problems in Anand – Fedoseev, Wijk aan Zee
2019.

Conclusion
 
This chapter has introduced the important topic of 5.d4 d5 6.Bd3, when 6...Bd6 7.0-0 0-0 8.c4 c6
reaches the main tabiya for this and the next two chapters. We saw that relieving the tension with
9.cxd5 cxd5 leads to nothing special for White, but 9.Re1 Bf5 leads to more complex play. The first
of our two main lines continued 10.Qc2 Na6 11.a3 Bg6 12.c5 Bb8 13.Nc3 f5 14.Qb3, when
14...Bh5! offers a promising pawn sacrifice. We analysed 15.Bxe4 and 15.Nxe4 and found that Black
gains at least equal chances in all variations; naturally it will be useful to memorize some of the
forcing variations that can occur here.
 

338
We then moved on to 10.Qb3 Qd7 11.Nc3 Nxc3! 12.Bxf5 Qxf5, when 13.bxc3 leads to a fairly quiet
positional game. The more ambitious 13.Qxb7 leads to an unbalanced pawn structure, with White
hoping to profit from having a bishop against a knight plus a queenside pawn majority, but Black
should not be worse at all, provided he arranges his pieces correctly.

339
A) 10.Bxe4 170
B) 10.a3 171
B1) 10...Bg4 171
B2) 10...f5!? 174
B21) 11.c5?! 174
B22) 11.b4 175
B23) 11.Nc3 Nc7 177
B231) 12.c5 178
B232) 12.b4 Bd7! 181
B2321) 13.Bb2 182
B2322) 13.c5 183

 
1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4 5.d4 d5 6.Bd3 Bd6 7.0-0 0-0 8.c4 c6 9.Qc2
White puts pressure on the e4-knight, hoping to provoke ...f5 or send the knight back from e4.
 
9...Na6!

340
 Black ignores the threat and develops quickly, while offering a pawn sac.
White can either accept the offer with A) 10.Bxe4 or defend against Black’s threat with B) 10.a3.
 

A) 10.Bxe4 dxe4 11.Qxe4

 
This gives Black good compensation after:
 
11...Re8 12.Qd3
Exchanging queens with 12.Qh4?! allows Black to seize the initiative: 12...Qxh4 13.Nxh4 Nb4
14.Na3 Re4 15.Nf3 Bg4³ Black threatens ...Bxf3 followed by ...Rh4, and he will inevitably regain
the pawn with a better position, Novotny – Bartos, Chrudim 2004.
 
12...Nb4
Black uses the exposed position of the white queen to put his pieces on active squares.
 
13.Qb3 Bf5 14.Na3
White has also tried:
14.Bg5 Qd7 15.Nbd2?!
15.Na3 is preferable, when 15...b6 transposes to the main line below.
15.Nc3? is a mistake in view of 15...Bc2 16.Qa3 Bd3µ when Black regains the pawn while
maintaining all his positional trumps, Albesa – A. Anand, corr. 2006.
The text move is inaccurate, as White develops his knight but fails to control certain key
squares. He does threaten c4-c5, but Black can himself play:
15...c5! 16.d5
Now in Ryska – Dearnley, corr. 1994, Black could have obtained a superb position with:

341
16...a5!N
Intending to gain more space on the queenside with ...a4.
17.a4
17.a3 a4 18.Qc3 Nd3µ leads to an obvious disparity in activity between the respective pieces.
17...h6 18.Be3 Ra6!µ
The rook lift reveals another point behind Black’s 16th move. The rook can go to b6, or to the
kingside after the bishop drops back from d6. Black is in full control and White’s extra pawn is not
felt at all.

14...b6N

342
 This seems like the simplest way of preventing White’s threat of c4-c5, although 14...a5!? also led
to interesting play in Svoboda – Vlasak, Sec u Chrudimi 2005.
 
15.Bg5 Qd7©
Black has full compensation for the pawn. His pieces are well placed and his further plans include
gaining space on the queenside with ...a5-a4. It’s also important to note that the knight on a3 will find
it hard to find an active role in the game.
 

B) 10.a3

In this position I took the slightly unusual step of covering two options for Black: B1) 10...Bg4 and
B2) 10...f5!?. Both are decent and may appeal to different types of players, although I must say the
latter would be my personal preference.
 

B1) 10...Bg4 11.Ne5

 
11.Bxe4?! is not good because of 11...dxe4 12.Qxe4 Bxf3 13.Qxf3 (13.gxf3? f5!µ and White will
face a lot of problems on the kingside) 13...Qh4 14.g3 Qxd4³ when Black has slightly the more
comfortable game.
 
11...Bf5
This is the safest and best choice.
 
The critical alternative is:

343
11...Bxe5
This was my initial preference but a more thorough analysis convinced me that White ends up
on top. The main line goes:
12.dxe5 Nac5 13.f3!
13.cxd5 Qxd5! (this is more straightforward than 13...cxd5 as played in Martinez Duany –
Enchev, Rochefort 2018) 14.Bxe4 Nxe4 15.Nc3 (trying to win the material with 15.f3?
backfires after 15...Nc5! 16.fxg4? [16.Rd1 Qb3µ] 16...Nb3–+) 15...Nxc3 16.Qxc3 Rfd8= and
Black clearly has nothing to worry about.
The text move leads to a series of forced moves.
13...Nxd3 14.Qxd3 Nc5 15.Qd4
It is no good for White to play 15.Qe3? d4µ or 15.Qc2 Bh5µ intending ...d4, when he is far
behind in development.
15...Nb3 16.Qxg4 Nxa1 17.Bh6 g6

18.Nc3!
It is not so good for White to regain the exchange immediately: 18.Bxf8?! Qxf8 19.cxd5
Qc5†!N Black must activate the queen before White locks it in with d5-d6. (19...Nc2? allows
20.d6! Ne3 21.Qf4 Nxf1 22.Kxf1± when White’s central pawns are too strong, Ma Qun – Xu
Xiangyu, China 2017) 20.Kh1 Nc2 21.Qd7 Ne3 22.Rg1 cxd5 23.e6 fxe6 24.Qxe6† Kf8
25.Qf6† Kg8= With perpetual check.
18...Qb6†
18...Nb3 19.Bxf8 Qxf8 20.cxd5 cxd5 21.Kh1!² left Black with a weak d-pawn and holes on
the kingside in Baklan – T. Petrosian, Internet 2004.
19.Kh1!
19.Rf2 was Khalifman’s recommendation in Opening for White According to Anand, but it later
became clear that White does not have to defend b2.
19...Qxb2 20.Qf4! f5!

344
 20...f6? was once played at the elite level, when both players missed White’s winning
continuation: 21.e6! (21.Na4? was played in Jakovenko – Kramnik, Dortmund 2009) 21...Qxc3
22.Qd6!+– With unstoppable threats.

21.Na4 Qe2 22.Rxa1 Rfe8 23.cxd5 Qxe5 24.Qxe5 Rxe5 25.dxc6²

Black has managed to hold this endgame in correspondence games (see for instance Manso Gil –
Brodda, email 2016) but I believe his defensive task would be unpleasant in a practical game.
 
12.b4!
Black was threatening ...Bxe5 followed by ...Nac5, so White covers the c5-square.
 
It is important to note that hitting the central knight does not work here:
12.f3? Bxe5 13.dxe5 Qb6† 14.c5
14.Kh1?! Nf2†!µ followed by ...Bxd3 clearly favours Black.
The text is trickier but it still turns out well for Black after:

345
14...Nexc5! 15.Bxf5 Nb3† 16.Kh1 g6!
An important intermediate move.
17.Ra2 Nd4 18.Be3 Qb5! 19.Bxd4 Qxf1† 20.Bg1 gxf5 21.Qxf5
Black has a draw in his pocket but he can also play more ambitiously with:

21...f6!N
21...Rae8= forced White to take the perpetual check in Baramidze – Senff, Germany 2008.
22.exf6 Rf7 23.g4
Intending g4-g5.
23...Re8³
The position is irregular and double-edged, but objectively Black stands better and has good
winning chances with accurate play.
346
 
12...Nc7!
Black prepares to bring his knight to a more active post on e6.
 
12...Qh4 is the only move mentioned by Khalifman, whose analysis was at the forefront at the time,
but obviously a lot has changed since 2003.
 
12...c5!? was another move that caught my attention, but here too White keeps some pressure, after
the following computerish variation: 13.cxd5 cxd4 14.Bxe4 Rc8 15.Nc6! Bxe4 16.Qxe4 bxc6
17.dxc6 Re8 18.Qf3 Qb6 19.Nd2 Rxc6 20.Nb3! Rc3 21.Qd1 d3 22.Bd2 Rc2 This occurred in
Dominguez Perez – Liang, Saint Louis 2019, and now 23.Qf3!N² would have left the d3-pawn
looking like more of a burden than an asset for Black.
 
13.f3 Bg6!
This is a vital tactical resource in the present variation.

14.Be3!?N
This seems to be the critical try for an advantage. Now White really threatens to take on e4.
 
The point of Black’s previous move is that 14.fxe4? doesn’t work due to 14...Bxe5! 15.dxe5 dxe4–+
followed by ...Qd4†.
 
14.c5 forces Black to exchange on e5, but after 14...Bxe5 15.dxe5 Ng5 16.Bb2 Black can create
some play on the queenside with:

347
16...a5!N (16...d4? was not so good in Anand – Caruana, Wijk aan Zee 2018) 17.f4 Bxd3 18.Qxd3
Ne4 19.Nd2 axb4 20.axb4 Rxa1 21.Bxa1 f5!= Black has a solid position with good prospects for his
knights.
 
14...Ng5
14...Nf6?! is inferior due to 15.Nxg6 hxg6 16.c5 Be7 17.Nc3² when Black is passive and White
has an easy plan of expanding on the queenside.
 
15.Nxg6 fxg6!
A surprising decision but an important detail, giving Black the opportunity to exchange the dark-
squared bishops with ...Bf4.
 
16.Nc3
16.Qd2 attacks the knight on g5 and momentarily prevents ...Bf4, but 16...Qf6! is a good answer.
Play continues 17.Nc3 Bf4 18.cxd5 Bxe3† 19.Qxe3 cxd5 (19...Nxd5? gives away the e4-square,
allowing 20.Bc4! Rad8? 21.Ne4+– when Black pays a heavy price) 20.Bc2 Followed by Bb3. White
maintains slight pressure although Black should be able to hold.
 
16...Bf4 17.Bxf4
17.Bf2 is met by the strong sacrifice 17...Nxf3†! 18.gxf3 Qg5† 19.Kh1 (19.Bg3 leads to a draw
after 19...Bxg3 20.hxg3 Qxg3† 21.Qg2 Qf4 22.Qf2 Qg5†= with a repetition) and now a follow-up
sacrifice: 19...Bxh2! The position is complicated but Black maintains the balance in all variations.
 
17...Rxf4 18.cxd5 Nxd5 19.Ne2 Rf8 20.Bc4 Kh8

348
 If the g6-pawn was back on f7, Black would have no worries whatsoever. As things stand though,
White’s knight is just a couple of moves away from reaching an unshakable post on e4, which gives
him slightly better chances.
 

B2) 10...f5!?

This move has taken over as the main line at the top level. Black strengthens the knight on e4 and
obtains some attacking chances on the kingside. The big question is whether or not White will be
able to exploit the weakened e5-square.
We will analyse three main options: the inaccurate B21) 11.c5?!, followed by the more serious B22)
11.b4 and B23) 11.Nc3.
Jumping straight away to the e5-square with 11.Ne5N is not dangerous for Black, who plays
11...Bxe5 12.dxe5 Nac5 13.cxd5 Nxd3 14.Qxd3 Nc5= followed by ...Qxd5 with easy equality.
 

B21) 11.c5?!
 
Expanding on the queenside immediately is inferior, as it allows Black to keep his bishop pointing
towards h2.
 
11...Bb8
The bishop is powerful on the b8-h2 diagonal.
 
12.Bxa6
349
 White should at least gain something in return by ruining Black’s pawn structure.
 
12.Nc3?! is simply inferior in view of: 12...Nc7 13.Bf4 (13.Ne2 Ne6µ gives Black a much-
improved version of variation B231 below, as the dark-squared bishop is far better on b8 than on e7,
Morozevich – Piket, Wijk aan Zee 2002) Now a significant improvement is:

13...g5!N (gaining a useful tempo over 13...Ne6, as played in Ni Hua – Vakhidov, Bled [ol] 2002)
14.Be5 Ne6 15.Bxb8 Rxb8 16.Ne2 Qf6µ Followed by ...g4 with a strong attack.
 
12...bxa6 13.Nc3
13.Ne5?! Bxe5 14.dxe5 was played in Baklan – Fressinet, Kallithea 2002, when 14...f4!N 15.f3
Ng5 16.Nc3 a5!³ would have been excellent for Black, whose bishop will be superbly placed on a6.
 
13...f4 14.Re1
14.Nxe4 Bf5 15.Nfd2 f3!‚ opens things up for the dark-squared bishop, giving Black a powerful
attack.
 
14...Bf5
Unfortunately, a rather premature draw agreement took place here in Ponomariov – Motylev,
Moscow 2002. Black’s chances are higher, as his weak pawn structure on the queenside is
outweighed by his great attacking chances on the other wing. A sample continuation could be:

350
15.Qa4 Qf6 16.Qxa6
White wins a pawn but Black will continue to focus on bringing pieces to the kingside.
 
16...Bc7
Intending to bring the last piece into play with ...Rae8. Instead of just waiting, White should try to
create some complications with:
 
17.Qb7 Rac8 18.Nxd5!? cxd5 19.Qxd5† Rf7µ
White has three pawns for the piece but Black still has the initiative on the kingside, which he
plans to increase by means of ...g5-g4. A final point worth mentioning is that White is unable to win
material, as shown after:

351
20.Rxe4? Rd8 21.Qb7 Bb6!–+
And Black wins.
 

B22) 11.b4 Nc7

12.Bb2?!
This move appears consistent and has been played several times, but the problem is that White
invites the black knight to head for the f4-square.
12.Nc3 is preferable, transposing to variation B232 below.

352
 
12...Ne6 13.Re1
Even worse is:
13.Ne5?
White tries to block the d6-bishop but allows the black queen to join the attack.
13...Qg5!
The queen is annoying for White to deal with.
14.f4?!
We have been following Curi – Roselli Mailhe, Montevideo 2006. White’s last move was an
understandable but wrong decision, which could have been punished by:

14...Nxf4!N
Presumably both players missed some details in the tactical line that follows.
15.Bc1 Bxe5 16.dxe5 Nh3† 17.Kh1 Nef2†! 18.Rxf2 Nxf2† 19.Kg1 Nh3† 20.Kh1
Now Black does not have to take a draw but can instead play:

353
20...Qh4! 21.gxh3 Qe1† 22.Kg2 f4! 23.Nd2
This is White’s only defence against ...f3†, but it is not really a defence at all.
23...f3† 24.Nxf3 Bxh3† 25.Kxh3 Rxf3† 26.Kg2 Raf8–+
Black crashes through.
 
The text move prepares to meet ...Nf4 with Bf1. Black is already better and has several possible
ways to continue, but I believe the strongest is:

13...N6g5!N
13...Qf6 was also fine in Charbonneau – Bluvshtein, Montreal 2005, but I prefer the idea of
challenging the f3-knight, which is holding White’s kingside together quite well.

354
 
14.Ne5
14.Nbd2 with the idea to maintain the knight on f3 can be met by 14...Nxd2 15.Nxd2 Qf6³ when
Black controls more space on the kingside and his pieces are well placed.
 
14...Rf6!
The rook lift shows another advantage of 10...f5. A logical continuation could be:
 
15.cxd5 cxd5 16.Nc3 Be6 17.f4
Now it’s important for Black to play:

17...Bxe5! 18.fxe5
18.dxe5? is punished severely by 18...Qb6† 19.Kf1 Nf3!! 20.gxf3 Rg6–+ when Black has a
crushing attack.
 
18...Rg6µ
Even after the departure of the dark-squared bishop from the board, Black still has excellent
attacking chances on the kingside.
 

B23) 11.Nc3

355
 This developing move is the clear main line.
 
11...Nc7
Black brings his knight into the game while defending the d5-pawn.
We will analyse, as the two main options, B231) 12.c5 and B232) 12.b4.
 
I also checked:
12.Be3
Now Black can improve on Idani – Jobava, Bandar e Anzali 2017, with the aggressive
continuation:
12...f4!?N 13.Bc1
The piece sacrifice 13.Nxe4?! is not good enough: 13...dxe4 14.Bxe4 fxe3 15.Bxh7† Kh8
16.fxe3 Bg4! It is important to eliminate the f3-knight before it can cause trouble. White does
not have enough compensation for the piece; for example, 17.c5 Bxf3! 18.cxd6 Qxd6 19.Rxf3
Rxf3 20.gxf3 g6! 21.Qxg6 Qxg6† 22.Bxg6 Rg8³ and Black has good winning chances in the
endgame.
13...Ng5!
Sacrificing a pawn in order to exchange the key defender on the kingside.

356
14.Nxg5 Qxg5 15.Bxh7† Kh8 16.Qg6!
It is important for White to exchange queens before Black creates serious threats against the
king.
16...Qxg6 17.Bxg6 dxc4
The pawn structure is imbalanced but Black is doing fine. For example:
18.Ne4 Be7 19.Re1 Be6=
Followed by ...Rad8.
 
The other possibility is:
12.Ne2 dxc4
12...Ne6 was the choice of Morozevich and some other strong players, but I prefer to leave
White with an IQP.
13.Bxc4† Be6

357
14.Re1
Bringing the rook to the e-file, so that when White plays Nc3 it will create strong pressure on
the e4-knight.
Another idea is 14.Bxe6† Nxe6 15.Nc3, challenging the e4-knight, as played in Laine –
Johansen, email 2006. Black’s best reply would have been 15...Bc7!N, ignoring the threat to the
knight and preparing ...Qd6, exerting pressure on the b8-h2 diagonal and making room for the
a8-rook to come to d8 or e8. After 16.Nxe4 fxe4 17.Qxe4 Qd6© Black has good compensation
for the pawn: he threatens ...Rxf3 and is ready to bring the a8-rook into action next.
14...Bxc4 15.Qxc4† Kh8 16.Nc3
Black should be fine thanks to the following accurate move:
16...Qe8!
Despite the self-pinning aspect of this move, it’s worth it to support the knight on e4 while
preparing to transfer the queen to the kingside.
17.Nxe4
17.Qd3 is met by 17...Qh5 18.Nxe4 fxe4 19.Rxe4 and now a typical sacrifice: 19...Rxf3!
20.gxf3 Qxh2† 21.Kf1 Nd5© Black has excellent compensation for the exchange. The rook is
about to join the party with ...Rf8, the f2- and f3-pawns are vulnerable and it is difficult to find
a safe shelter for the white king.
17...fxe4 18.Ne5 Bxe5 19.dxe5 Qxe5 20.Rxe4 Qd5
An equal endgame has arisen and the players quickly agreed a draw after:
21.Qxd5 Nxd5 22.Bd2
½–½ Pedersen – Koesebay, corr. 2004.
 

B231) 12.c5

358
 This is the most popular continuation. White gains space on the queenside and forces the bishop to
vacate the b8-h2 diagonal. On the other hand, White releases the pressure in the centre, and Black’s
bishop will find another good post on f6.
 
12...Be7 13.b4
13.Bf4?! Ne6 14.Be5 Bf6 15.Ne2 transposes to the 13.Ne2 Bf6 14.Bf4?! line just below.
 
White can try to establish his knight on e5 with:
13.Ne2 Bf6 14.Ne5
14.b4 Ne6 transposes to the main line below.
14.Bf4?! was well met by 14...Ne6 15.Be5 g5!³ in Zeberski – Wlodarczyk, Opole 2006. Black
intends ...g4 with strong pressure on the kingside.
14...Ne6

359
15.Be3!?N
This interesting novelty leads to an unclear position.
15.f4 Bxe5! seems counterintuitive, but it has a hidden positional idea! 16.fxe5 Now in Rossetti
– Maccarini, Bratto 2008, Black failed to follow up in the best way: 16...b6!N Challenging the
c5-pawn, more or less forcing White to respond with 17.b4, and now 17...bxc5 18.bxc5 Qa5!³
is the point of the whole operation starting at move 15. Black prepares to exchange the light-
squared bishops with ...Ba6, which will give him at least a slight edge.
15...f4!
This is definitely the critical move.
16.f3!
16.Nxf4? does not work due to 16...Bxe5 17.Nxe6 Bxh2† 18.Kxh2 Qh4† 19.Kg1 Bxe6µ and
White will have problems on the kingside.
16...fxe3 17.fxe4 Nxd4! 18.Nxd4 Bxe5 19.Rxf8† Qxf8 20.Nf3÷
The position is messy, with chances for both sides.
 
13...Ne6 14.Ne2 Bf6 15.Bb2
White’s idea is to play Ne5 followed by f2-f4.
15.Be3N is simply met by 15...g5!‚ intending ...f4 and ...g4, with good attacking chances on the
kingside.
 
15...Qc7!
Black prevents the enemy knight from going to e5.
 
16.Ne1
16.Ne5? doesn’t work due to 16...Nxd4 17.Bxd4 Bxe5µ and Black is simply a pawn up.

360
 
16.b5?! is too optimistic: 16...Bd7 17.Rab1?! (17.a4 defends the pawn but 17...Rae8µ gives Black
excellent chances on the kingside) Now in Macieja – Pashikian, Stepanakert 2004, Black should have
simply taken the free pawn with:

17...cxb5!N 18.Qb3 Qc6 19.Bc1 a6µ White is not going to get his pawn back.
 
16.Rad1 g6!
Black intends ...Qg7, putting pressure on the d4-pawn, followed by developing with ...Bd7.
17.Qb1 Qg7 18.Ba1 Bd7 19.Kh1
19.Ne5? loses a pawn to 19...Nxd4!, so White makes a prophylactic move, ensuring that Black
will not be able to take on e2 with check. Black’s best response is:
 

361
19...Rae8!N
Enabling the bishop to drop back to c8, after which Black can play freely on the kingside with
all his pieces in the action.
19...Be8 was played in Tseshkovsky – Motylev, Ekaterinburg 2002, when 20.Ne5!N would
have been best, as 20...Nxd4? 21.Nxd4 Bxe5 22.Bxe4 fxe4 23.Ne6± turns out badly for Black.
20.Ne5 Bc8 21.f3
21.f4 is met by 21...g5!‚ with good attacking prospects on the kingside.
Now Black has an interesting idea:

21...Nxd4!? 22.Bxd4 Bxe5 23.Bxe5 Qxe5 24.fxe4 fxe4 25.Bc2 Bg4©

Black has fine compensation for the sacrificed piece, as the pawns on e4 and d5 form a strong
barrier against the c2-bishop.
362
 
Returning to our main line, after dropping the knight back to e1 White intends f2-f3, when moving
the knight would hang the f5-pawn. Hence Black should simply play:

16...g6
16...Bg5?! was not so good in Wan Yunguo – Liu Qingnan, China 2017.
 
The text move bolsters the f5-pawn, enabling Black to follow up with ...Bd7 and ...Rae8, bringing all
the pieces into the action zone. It’s important to note that White’s natural and generally desirable
move leads him into trouble:
 
17.f3?! Qg7!
Forcing White to make a difficult decision.
 
18.Rd1!
Sacrificing a pawn seems like White’s best chance to hold.
 
18.fxe4 maintains the material balance but White ends up being positionally worse: 18...fxe4 19.Qc3
(19.Bxe4? is easily refuted by 19...Nxd4–+ and Black wins material) 19...exd3 20.Nxd3 Ng5!³ The
knight looks forward to landing on the e4-outpost. Black also has the bishop pair, so White has a
difficult time ahead.
 
18...Nxd4 19.Bxd4 Bxd4† 20.Nxd4 Qxd4† 21.Kh1 Ng5 22.f4 Ne6 23.g3
White has stabilized his position and intends Nf3-e5.
 
23...Qg7³

363
 White’s knight will soon make its way to e5 and it will not be easy for Black to realize his extra
pawn. Nevertheless, he has realistic winning chances and has certainly won the opening battle.
 

B232) 12.b4

Keeping the tension for another move forces Black to come up with an accurate response.
 
12...Bd7!
A nice little move! Now Black can think of playing ...Ne6 next. If White meets that move by
taking twice on d5, then Black can bring his rook to c8 with tempo and perhaps play ...Bc6, when
both of his bishops will point menacingly towards White’s kingside.
 
Play against an IQP after 12...Nxc3 13.Qxc3 dxc4 would be positionally desirable but White has a
strong intermediate move in 14.Bg5! (14.Bxc4† Be6= would be comfortable for Black) which
disturbs the harmony in Black’s camp. Best play continues: 14...Nd5 15.Bxc4 Qc7 16.Bd2
Retracting the bishop before it gets into trouble on the kingside. 16...Be6 17.Qd3 h6 18.Rfe1² The
weakness of the e5-square gives White the upper hand.
 
We have a final split, with B2321) 13.Bb2 and B2322) 13.c5 the main options.
 
13.Ne5? runs into 13...Ne6! and it will be difficult to protect the d4-pawn.
 
13.Ne2 lifts the pressure from the d5-square, so 13...dxc4 14.Bxc4† Be6= gives Black a fine
position.
 

364
13.Be3?! gives Black a favourable version of the aforementioned pawn sacrifice: 13...Ne6! 14.cxd5
cxd5 15.Qb3 (15.Nxd5 Bc6 16.Bc4 Bxd5 17.Bxd5 Bxh2†³ regains the material with a better
position for Black) 15...Nxc3 16.Qxc3 f4 17.Bd2 Ng5ƒ With a good initiative for Black.
 
13.Rb1 Ne6 14.cxd5 cxd5 15.Nxd5
Now in Tsydypov – Frolyanov, Internet 2019, Black’s best idea would have been a prophylactic
move:

15...Kh8!N
Removing the king from the exposed diagonal before choosing an active plan.
16.Qb3
The queen moves away from the vulnerable post on c2.
The greedy 16.Bxe4? fxe4 17.Qxe4 is met by the typical sacrifice 17...Rxf3! 18.Qxf3 Bc6–+
and White is about to lose material.
16.b5 covers the c6-square but the bishop has another route to activate itself: 16...Be8!
Intending ...Bh5. A sample variation goes 17.Bxe4 Rc8 18.Nc3 fxe4 19.Qxe4 Rxc3 20.Qxe6
and now the thematic 20...Rcxf3! 21.gxf3 Bg6© gives Black a strong initiative for the material,
while White is left with a ruined structure on the kingside.
16...Bc6 17.Ne3 Nf4
White is under strong pressure from Black’s active pieces, and has to come up with:

365
18.d5!
Returning the extra pawn to release the tension.
18...Nxd5 19.Nxd5 Bxd5 20.Bc4=
With an equal position.
 

B2321) 13.Bb2

This was Khalifman’s recommendation, based on a high-level game between Leko and Morozevich
in 2002. The bishop may not look so effective facing its own d4-pawn, but it allows White to support
an invasion on the e5-square.
366
 
13...Ng5
Black moves the knight away from the e4-square to prepare ...dxc4, in order to play against the
IQP.
 
An interesting alternative is to sacrifice a pawn with 13...Ne6!? 14.cxd5 cxd5 15.Nxd5 and now
15...N4g5! exchanges the important defender of White’s kingside. 16.Nxg5 Qxg5© Black has nice
compensation due to his activity on the kingside.
 
14.Ne5
White has to place the knight actively if he is to hope for an advantage.
 
14...dxc4
Khalifman only mentioned 14...Bxe5 which was played in Leko – Morozevich, Dortmund 2002,
but later games revealed that the text move is preferable.
 
15.Bxc4† Be6
As is well known, exchanging minor pieces tends to benefit the side playing against the IQP.
 
16.Na4
White tries to put pressure on the queenside by activating his knight. Black responded by
exchanging on c4 in Rehtman – Mayer, corr. 2010, but I think it would have been simpler to play:
 

16...Bxe5!?N

367
 Changing the pawn structure in order to play against White’s dark-squared bishop, bearing in mind
that the ideal e6-square is available as a blockading point for the knight.
 
17.dxe5 Bxc4 18.Qxc4† Qd5 19.Qxd5†
19.Qc2 is conveniently met by 19...Qf7= followed by ...Nd5, with the other knight coming to e6.
 
19...Nxd5 20.Nc5 b6
White’s only chance to cause problems is:

21.Nb7!?
Intending to place the knight on the active d6-square. However, a good reply is:
 
21...a5!=
Forcing White to open the b-file, which will be useful for the black rooks to create counterplay.
The ensuing endgame should be fine for Black.
 

B2322) 13.c5

368
 White gains space on the queenside and also eliminates the possibility of an IQP position after
...dxc4.
 
13...Be7 14.Ne5 Ne6
Improving the knight while attacking the d4-pawn.
 
15.Ne2
White intends to develop the dark-squared bishop to b2, strengthening the e5-knight.
 
15.Be3N should be met by 15...Bf6 16.f4 a5!, taking action on the queenside before White is able to
build up pressure there. Play may continue:
 

369
17.Ne2 (17.b5?! can be met by 17...Bxe5 18.dxe5 Nxc3 19.Qxc3 cxb5³ when, following the
exchange of the light-squared bishops on b5, Black’s knight on e6 will be stronger than White’s
remaining bishop on e3) 17...Be8= The position is about equal. Both sides have their knights firmly
established on the central squares. Black’s other knight is well placed on e6, especially in case Black
decides to take on e5, when the ensuing passed e-pawn will be well blockaded. White has more space
on the queenside, but it is not much to worry about as Black always has the option of blockading the
queenside with ...b5.

15...Bf6 16.Bb2 Qc7

Threatening ...Nxd4.
 

370
17.f4
White strengthens the e5-knight but also leaves its counterpart firmly planted on e4.
 
White can try to win a pawn with 17.Nxd7!? Qxd7 18.f3 but Black has a strong reply:

18...Bg5! Threatening ...Be3†. 19.fxe4 (19.Bc1 is met by 19...g6!, supporting the f5-pawn and
preparing to evacuate the knight from e4. 20.fxe4 fxe4 21.Rxf8† Rxf8 22.Bxe4 dxe4 23.Qxe4

White has won a pawn, but after 23...Qf7„ Black has nice counterplay on the kingside.) 19...fxe4
20.Bxe4 dxe4 21.Qxe4 Nc7„
 

371
 White is a pawn up but Black has a nice outpost on d5 and counterplay coming up on the f- and e-
files.
 
17...Be8 18.Ng3
Now in Ter Sahakyan – Frolyanov, Internet 2019, the simplest option would have been:

18...g6N
Securely supporting the f5-pawn, with a comfortable position for Black.
 

Conclusion
 

372
This chapter has dealt with the important option of 9.Qc2, putting pressure on the e4-knight. 9...Na6!
is our reply, when taking on e4 wins a pawn for White but offers Black plenty of compensation, so
10.a3 is deservedly the main line. I decided it was worth analysing 10...Bg4, as historically it’s the
most popular move and it is certainly playable; nevertheless, White does keep a small advantage, so I
would not recommend this, although you may wish to keep it as a back-up or a surprise option.
 
10...f5!? has taken over as the main line for Black, and my analysis shows that it is in good
theoretical shape. That is especially true after 11.c5?!, when White releases the tension prematurely
and allows Black to organize his pieces in the best possible way, with the bishop pointing menacingly
towards h2. 11.b4 is not so bad, although after 11...Nc7 White had better transpose to the main line
with 12.Nc3, as the alternative 12.Bb2?! invites 12...Ne6 when the knight eyes up the juicy f4-
square.
 
The main line continues 11.Nc3 Nc7 12.b4, when 12...Bd7! is most accurate. Then 13.Bb2 gives
Black a pleasant choice between the solid 13...Ng5 and the interesting pawn sacrifice 13...Ne6!?, so
13.c5 seems best, when 13...Be7 14.Ne5 Ne6 leads to a complex positional struggle. Each side
boasts a strong knight on the ‘king five’ square, and while White has gained some queenside space
with c4-c5, the same advance has left his d4-pawn slightly weak, giving Black tactical resources
involving ...Nxd4 in some lines. Overall the positions seem dynamically equal and I have no
hesitation in recommending this variation for Black.

373
A) 12.Ne5 187
B) 12.Re1 189
C) 12.Ng5 h6 190
C1) 13.Nxf7N 190
C2) 13.Qf3! 192
D) 12.Bg5 Qa5! 194
D1) 13.Re1 195
D2) 13.d5!?N 196
D3) 13.Nh4 198

 
1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4 5.d4 d5 6.Bd3 Bd6 7.0-0 0-0 8.c4 c6 9.Nc3
This move has been White’s most frequent choice and it remains a regular guest at top
tournaments.
 
9...Nxc3 10.bxc3 dxc4 11.Bxc4 Bf5

374
 The bishop is well placed here, covering the b1-square and also defending against any future
attacks along the b1-h7 diagonal.
 
I analysed 11...Bg4 as well, but found 12.h3 Bh5 13.g4 Bg6 14.Ne5 to be uncomfortable for Black.
 
White has four main candidates: A) 12.Ne5, B) 12.Re1, C) 12.Ng5 and D) 12.Bg5.
A relatively harmless alternative is:
12.Bd3 Bxd3 13.Qxd3 Nd7

14.g3
White improves the position of his king before taking further action.

375
 In another game White tried 14.c4 Re8 15.Bb2, when Black found a nice manoeuvre: 15...Be7!
16.Rfe1 Bf6 From this square, the bishop effectively neutralizes all White’s hopes of activating
his bishop with d4-d5. 17.Rxe8† Qxe8= Black was fine in Brkic – Banusz, Austria 2018.
14.Rb1 can be simply met by 14...b5, when White should play 15.c4 before Black conquers the
light squares on the queenside with ...Nb6. 15...bxc4 16.Qxc4 Qc7 17.Bd2 White can try to put
pressure on the c6-pawn but Black will simply exchange it with ...c5. 17...Rab8 18.Rbc1 Rfc8
19.Rc2 c5! 20.Rfc1 Nb6 21.Qf1 Qe7 22.dxc5 Rxc5= The endgame was completely equal in
Paichadze – Ter Sahakyan, Internet 2019.
14...Re8 15.Kg2 Nf6 16.c4 Qd7 17.Bb2 Rad8=
Having fully mobilized his forces, Black had no problems whatsoever in Vachier-Lagrave –
Caruana, Paris 2018.
 

A) 12.Ne5 Nd7
12...Bxe5 is another way to play, but personally I prefer not to exchange the dark-squared bishop.
 
13.f4
White firmly supports the e5-knight and aims to seize the initiative on the kingside with moves like
Qh5 or g2-g4.
 
13.Qf3 is met by: 13...Qf6! (13...Be6 is slightly less accurate in view of 14.Bxe6 fxe6 15.Qg4 Rf5
16.Nc4 Qc7 17.g3 Nb6 18.Ne3² when White may have some pressure) 14.Nxd7 Bxd7 15.Qxf6
gxf6= The doubled pawns are of little consequence and the position is balanced.
 
Now in Esenov – Gunbayar, Beijing (rapid) 2008, Black should have improved the knight with:

376
13...Nf6!N
Covering a lot of squares on the kingside. Next Black can prepare a timely ...c5 to obtain
counterplay in the centre.
 
14.g4
This aggressive move is the only critical try, but Black has a good answer.
 
14...Be6! 15.Bd3
In the event of 15.Bxe6 fxe6 16.Qe2 Qe7³ Black is doing fine. His ideas include breaking with
...c5, and the overexposed white pawns are liable to become weak.
Now Black has another strong move:
 
15...c5!
It is inferior to grab a pawn with 15...Bxe5?! 16.fxe5 Nxg4 17.Qe1 when the position is messy.

16.g5
After 16.f5 Bd5µ it is clear that White’s aggressive kingside strategy has backfired. For example:
17.g5 cxd4 18.cxd4 Qb6!–+
 
16...Nd5 17.f5
There follows a forced sequence of moves:
 
17...Nxc3 18.Qe1 Bxe5 19.Qxe5 Bd5 20.dxc5
The position remains sharp, but Black will emerge with the better chances after a precise sequence:

377
20...Bc6! 21.Bc4!
21.Qxc3?? Qd5 is terminal, so White covers the d5-square.
 
21...b5!
Another important resource.
 
22.cxb6 Qxb6† 23.Qe3
23.Be3? Qb2!µ gives Black a larger advantage.
 
23...Rfd8³
Black has slightly better chances in the endgame, as White will continue to suffer from his exposed
kingside.
 

B) 12.Re1

378
 This is the most popular move in this position, although it usually transposes to one of the main
lines later in the chapter. In this section we will consider a few independent possibilities which are
generally not so challenging.
 
12...Nd7 13.h3
Such a slow move cannot be a serious concern for Black.
 
13.Bg5 is the main line, when 13...Qa5 transposes to variation D1 on page 195.
I also checked an energetic but dubious novelty:
13.Ng5?!
Black can simply respond with:
13...Nf6!
13...h6 14.Qf3!÷ is more complicated.
14.Nxf7
We will see a similar sacrifice to this in the later variation C1. Neither version works out well
for White!
14.Qf3 is met by 14...Bg6³ when Black’s kingside is well fortified by his minor pieces, so
White has virtually no chance of getting a dangerous attack.
14...Rxf7 15.Bxf7† Kxf7 16.Qb3† Kg6 17.Qxb7

379
17...h5!
Making a useful move on the kingside and waiting for White to take on c6.
18.Qxc6
18.h3 is met by 18...Rc8 intending 19.Qxa7?! Ne4µ with a strong attack.
18...Rc8 19.Qf3 Qc7 20.g3 Qxc3³
Black has better prospects in the endgame.
 
Now in Zawadzka – Kubicka, Warsaw 2018, Black’s best continuation would have been:
 

13...Re8!N
Challenging for the open file.
380
 
14.Rxe8†
Trying to win a pawn with 14.Qb3?! doesn’t work after 14...Rxe1† 15.Nxe1 Qe7³ followed by
...b5.
 
14...Qxe8 15.a4
It seems logical to prevent ...b5.
 
15...h6 16.Qb3 Rb8
Black has a good position, for instance:

17.Ba3 Bxa3 18.Qxa3 Qf8! 19.Qa2 Be4=

Black’s pieces are well placed and the pressure on the f7-point is not an issue for him.

C) 12.Ng5

381
 MVL employed this move twice in 2018.
 
12...h6
The knight should be challenged. We will analyse C1) 13.Nxf7N and the more accurate C2)
13.Qf3!.
 

C1) 13.Nxf7N
 
Creating complications in this way does not work out well for White, which is no doubt why MVL
avoided it.
 
13...Rxf7 14.Bxf7†
No better is:
14.Qb3 Bg6!
14...Qd7 also turns out well after 15.Re1 Na6 16.Ba3 Bxa3 17.Qxa3 Re8³ when the two minor
pieces are more valuable than the rook.
The text move is even stronger; Black is inviting White to capture his queenside pawns.
15.Qxb7 Nd7 16.Qxc6 Nb6 17.Bxf7† Bxf7
White has a rook and three pawns for two pieces, but Black’s pieces are extremely well placed
whereas White’s are passive and lack coordination. Black can create a lot of pressure on the
kingside.

382
18.Qf3 Qc7 19.Qh3 Rf8
It may not be apparent at first sight, but Black stands clearly better here. For example:
20.Bd2 Nc4 21.Rad1 Bd5µ
White’s pawns are completely blocked, while the knight on c4 and bishops on d6 and d5 look
menacing to say the least!
 
14...Kxf7 15.Qb3†

15...Kg6!
The king is best placed on g6 for now, avoiding any further checks by the white pieces.
 

383
16.Qxb7
With a series of forcing moves White has managed to win material, but he has lost a lot time in the
process. Black’s bishops are well placed and in the absence of any defenders on the kingside, Black
can create a lot of play there.
 
16...Nd7 17.Qxc6
17.Re1 sees White wait for Black to force him to take on c6, but 17...Qc8! 18.Qxc8 Rxc8³ gives
Black the better endgame, as he has excellent squares such as d5 and c4 for his pieces.
 
17...Nf6
Black wants to play ...Rc8 followed by ...Qc7, hitting both the h2- and c3-pawns.
 
18.Bd2
18.Re1 gives up a pawn without a fight after 18...Rc8 19.Qf3 Qc7 20.h3 Qxc3 21.Qxc3 Rxc3³,
reaching an endgame where only Black can play for a win.
 
The prophylactic retreat 18.Qf3 is met by: 18...Be4 19.Qh3 Qc7 Black continues to put his pieces on
good squares, creating maximum pressure. 20.Re1 Threatening to take on e4, but Black has a
convenient reply.

20...Rd8! This defends against the threat of Rxe4, as now the rook will not be hanging on a8. 21.Bd2
(21.Rxe4 Nxe4 22.Qg4† Ng5 does not work for White, as he cannot win material with Bxg5
followed by Qe4†, and if 23.h4? then 23...Qxc3–+ wins for Black) 21...Bf5 22.Qh4 Kf7µ Black has
full control over the d5- and c4-squares, giving him clearly better chances.
 
18...Rc8 19.Qf3 Qc7 20.h4 Be4

384
 Pushing the white queen to an inferior position.
 
21.Qh3 Rf8 22.Rfe1 Qc6³
Even though it is not clear how Black will break through, all of his pieces are well placed and
White is under pressure on the kingside. Black can even think of walking his king back to g8 or f8
before taking action.
 

C2) 13.Qf3!

This is the main point behind the knight’s jump to g5.

 
13...Bg6 14.Ne4
Wang Hao tried to manoeuvre the knight along a different path with 14.Nh3, and after 14...Nd7
15.Nf4 Bh7 16.Nh5 White threatens Qg4. But of course Black does not allow it and defends actively
with:

385
16...Qh4! After the further 17.Bf4 Bxf4 18.Nxf4 Rae8 19.Rae1 Nf6= Black had fully mobilized his
forces and was completely fine in Wang Hao – Nisipeanu, Bucharest 2013.
 
The text move offers to sacrifice a pawn for the initiative. Black should accept the challenge with:
 
14...Bxh2†!
14...Bc7 seems inferior in view of 15.Nc5 b6 16.Nd3² with good play for White.
 
15.Kxh2 Qh4† 16.Qh3
16.Kg1?! Bxe4 17.Qf4 Qxf4 18.Bxf4 Nd7³ gives White an inferior version of the main line, as the
g-file is closed.

386
16...Qxh3†
Maintaining the queens with 16...Qxe4? allows White to seize the initiative with 17.Ba3 Rd8
18.f4± when the two white bishops are annoying, while Black’s queenside is yet to be developed.
 
17.gxh3!
White intends to attack down the g-file.
 
17...Bxe4 18.Rg1
Now it’s important not to cling to the extra pawn, but to give it back with:
 
18...Re8!
Protecting the h6-pawn with 18...Kh7? runs into 19.Ba3! Rd8 20.Bxf7± when White’s powerful
bishop pair combined with the pressure along the g-file is too much for Black to handle.
18...Kh8?! 19.Ba3² is not quite as bad as the line above, but Black should avoid it all the same.
 
19.Bxh6 g6
Black has neutralized White’s threats on the kingside and can now focus on completing his
development.

387
20.h4!?N
This seems like an interesting try, intending to trade off the weak h-pawn. Two other moves have
been tested in practice:
 
20.a4 gains space on the queenside and prevents ...b5, but it allows Black to develop comfortably
with: 20...Nd7 21.Bf1 Nf6 22.Be3 Rad8 Black has easily completed his development while White
did not manage to pose any problems at all. After 23.c4 Black executed a nice knight manoeuvre
with 23...Nh5! 24.Ra3 Ng7= intending ...Nf5 with a fine position, and the players agreed a draw
after a few more moves in Bernadskiy – Ter Sahakyan, Villorba 2019.
20.Rge1
This was MVL’s choice but he found no advantage with it.
20...Nd7 21.Bf1
White’s idea is to play a long game and create chances with his two bishops, but his chances for
an advantage are limited by his scattered pawn structure on both flanks.
21...c5
Black breaks in the centre immediately before White gets a chance to play c3-c4. (If he was
allowed to play that move, then ...c5 could be met by d4-d5.) Now White went for
simplifications with:
22.Bb5 Bc6 23.Rxe8† Rxe8 24.Bxc6 bxc6

388
25.Be3
White can also try 25.dxc5N but Black is fine after: 25...Nxc5 26.Be3 Ne4 27.c4 Rb8!
Intending to activate the rook via b2. 28.Bxa7 (28.Rd1 a5= is also fine for Black) 28...Ra8
29.Be3 Ra4= Black regains the pawn with a good position.
25...cxd4 26.Bxd4 a6=
The position remained equal in Vachier-Lagrave – Wang Hao, Isle of Man 2018.
 
20...Nd7
Black has other ways to play, but I feel that the text move would be the most practical choice.
 
A playable alternative is:
20...b5 21.Bb3 a5
Intending to drive White’s bishop away from the a2-g8 diagonal.
22.a4
I also checked 22.a3 a4 23.Ba2 Nd7 24.h5 Kh7= when the position is balanced.
The text move enables Black to create a passed pawn on the a-file.
22...Nd7
Intending ...Nb6.

389
23.axb5
23.h5 is met by 23...bxa4 24.Rxa4 Kh7 25.Bg5 f6 26.Be3 Nb6 27.hxg6† Bxg6= followed by
...a4.
23...cxb5 24.h5 a4 25.Ba2 Kh7 26.Bf4 Kg7=
Even though White has the bishop pair, Black is doing fine due to his play against the c3-pawn and
control of key squares such as d5 and c4.
 
21.h5 Kh7
Supporting the g6-pawn while also pushing the enemy bishop away.
 
21...Nb6 is also possible, but after 22.Bb3 Nd5 23.Rg3 Bf5 24.hxg6 fxg6 White has slightly the
more pleasant side of a probable draw.
 
22.Be3 Bd5!
The point behind Black’s previous move.
 
23.hxg6†
23.Bd3 is of course met by 23...Be4 with equality.
 
23...fxg6 24.Bxd5 cxd5
White’s only real attempt to cause problems is:

390
25.Rab1 b6 26.Rg5
Trying to exploit Black’s loose pawns, but Black can keep everything under control.
 
26...Nf6 27.Rbg1 Rg8 28.Re5 Rae8 29.a4 Kg7 30.f3 Kf7=
Black has no problems in this endgame.
 

D) 12.Bg5

391
 White develops the bishop with gain of tempo, and may think about harassing the f5-bishop with
Nh4 next.
 
12...Qa5!
I prefer this active, counterattacking move over the more timid ...Qc7. One of the themes from this
point is that if and when White plays Nh4, Black will look to counter with ...Be6, and if Bxe6 then
...Qxg5 shows the power of Black’s queen.
 
White has three main options: D1) 13.Re1, D2) 13.d5!?N and D3) 13.Nh4.
 
13.Qd2 is simply met by 13...Nd7, intending ...Rae8 and ...Nb6-d5. After the further 14.Bf4 Qc7
15.Bxd6 Qxd6= Black had no problems in Kacharava – Zarubitski, Kirishi 2018.
 

D1) 13.Re1
 
With this tricky move, White waits for Black to develop his knight to d7, so that after Nh4 Black will
not have the typical ...Be6 available for tactical reasons, as explained below. Fortunately Black has
other resources he can rely upon.
 
13...Nd7 14.Nh4
14.Qd2 can be met by the usual 14...Nb6 15.Bb3 Nd5 with a good game for Black.

White decided to occupy the 7th rank with 16.Bxd5 cxd5 17.Be7 Bxe7 18.Rxe7 in Bacrot –
Bogner, London 2016, at which point the simple 18...Qb5N, with the idea of ...Rae8 to exchange

392
rooks, would have been fine for Black.
 
White has also tried:
14.Qb3 b5 15.Bf1 Rae8
Black’s moves are simple and strong.
16.c4
16.Bd2 can be met by 16...Be6 17.c4 bxc4 18.Bxc4 Bxc4 19.Qxc4 Qd5= and Black is
comfortable.
In Kantans – Khalifman, Liepaja 2018, Black’s most accurate option would have been:

16...bxc4N 17.Bxc4
17.Qxc4 Qd5 is fine for Black.
It seems that White has annoying pressure along the a2-g8 diagonal, but Black can neutralize it
with:
17...Be4!
Threatening to win a piece with ...Bxf3, while also preparing to consolidate with ...Bd5.
18.Nd2
Black also has no problems after 18.Bd2 Qc7= followed by ...Nb6.
18...Bd5=
White has no trace of an opening advantage.
 

393
14...Nb6!
This important resource enables Black to save the bishop pair.
14...Be6? is no good here due to 15.Rxe6! fxe6 (15...Qxg5? 16.Nf3+– is even easier for White)
16.Bxe6† Kh8 17.Qh5! and Black is in deep trouble.
 
15.Bb3 Bd7 16.Bd2 Rae8 17.Rxe8
17.Qf3 was played in Clair – Armand, corr. 2002, when 17...Be6!N would have neutralized the b3-
bishop while preparing ...Nc4; Black is doing fine.
 
17...Rxe8 18.Rb1 Qa6!
Intending to land the queen on d3 or e2.
 
19.Qh5

394
 The time has come to improve upon Black’s play.
 
19...Nd5!N
More precise than 19...Nc4 as played in Liang – Lenderman, Saint Louis 2018.
 
20.g3
White cannot grab a pawn, as 20.Bxd5? cxd5 21.Qxd5? Qd3 wins for Black.
 
20...Be6³
Black has done more than equalize, as White’s knight and queen are useless on the kingside.
 

D2) 13.d5!?N

395
 This untested move is definitely worth analysing, although Black has more than one satisfactory
reply.
 
13...Be4
Forcing White to take on c6 seems easiest.
 
The alternative is:
13...Qc5 14.Qb3!?
14.dxc6 Nxc6= is easily equal.
After 14.Bb3 cxd5 15.Be3 Qc7 16.Qxd5 White’s pieces look active while Black still needs to
complete his queenside development, but 16...Be6 conveniently enables him to exchange the
strong bishop on b3: 17.Qe4 Bxb3 18.axb3 Nd7! It’s better to place the knight on f6 than on c6.
19.Rfd1 Nf6 20.Qd3 Rfd8= Black is fine.
The text move is an interesting attempt to seize the initiative, but Black can equalize as follows:
14...b5 15.Be2 Be4!
The most precise, not allowing Black’s pawns to become disconnected.

396
16.c4!
This is the best attempt to create problems for Black.
16.Be3 gets nowhere after 16...Qxd5 17.Rfd1 Qxb3 18.axb3 Be7 19.Rxa7= with an equal
endgame.
16...bxc4 17.Qb7 cxd5 18.Qxa8 Qc7
The queen is in danger on a8, so White has to play:
19.Rfb1!
But not 19.Rab1? due to: 19...Bxb1 20.Qxd5 Bxa2µ
19...Bxb1 20.Qxd5 Bd3=
The position becomes drawish.
 
14.dxc6 Bxf3 15.Qxf3 Nxc6
White has two good bishops but, as usual, the weak pawn structure limits his chances.
 

397
16.Bd5 Rae8
Intending ...Re5.
 
17.Rfd1
17.Bh4 Ne5³ with the idea of ...Ng6.
 
17...Ne7!
Forcing the exchange of the knight for one of the enemy bishops.
 
18.Be3
18.Bxe7 Rxe7 19.Bxb7 doesn’t win a pawn, as Black has 19...Bxh2†! 20.Kxh2 Qc7†= leading to a
drawish endgame.

398
18...Nxd5 19.Qxd5 Qxd5 20.Rxd5 Bb8=
 
White has a slight edge in activity but his split queenside pawns and the reduced amount of
material make a draw a near certainty.
 

D3) 13.Nh4

This has been White’s most popular choice.

 

399
13...Be6!
As mentioned earlier, this resource is one of the major advantages of having the queen on a5.
 
14.Bxe6 Qxg5 15.Nf3 Qa5
15...Qd8 was the choice of an elite player and Petroff expert in Giri – Caruana, Baden-Baden 2018,
but it is too passive for my taste; I prefer to put the queen back on the more prosperous a5-square.
 
16.Bb3
16.Qb3 was played in Conde Poderoso – Hofer, corr. 2018, when the simplest reply would have
been 16...Qb6N, exchanging queens and leading to an equal endgame with opposite-coloured
bishops.
 
16.Bh3!?N prevents ...Nd7 but does not exert any pressure on Black’s kingside, so Black can safely
grab a pawn with 16...Qxc3. White has some compensation after 17.Rb1 b6 18.Re1, but Black can
return his extra material and obtain a comfortable position after:

18...Na6 19.Qa4 Qa5 20.Qxc6 Rad8= Intending either ...Nb4 or ...Nc7, followed by ...Qd5, with a
fine position.

400
16...Nd7
Please be aware that 16...Qxc3?? is too greedy here, as 17.Ng5!+– gives White a lethal attack.
 
Instead Black should simply focus on completing his development with ...Rae8 and ...Nf6. The
presence of opposite-coloured bishops favours White slightly, but Black is solid enough on the
kingside and full equality should not be too far away.
 
17.Qd3 Rae8
17...Nf6 is also perfectly playable, and after 18.Rfe1 Rae8 19.g3 g6 20.Kg2 Kg7= White had no
advantage in Zawadzka – Koneru, Khanty-Mansiysk 2018.
 
18.g3 g6
Both sides prepare to improve their king positions.
 
19.Kg2
19.Nd2 does not change much. 19...Qd8 The queen moves away from her slightly exposed
position. 20.Rfe1 Kg7 21.Kg2 Rxe1 22.Rxe1 Re8 23.Rxe8 The players agreed a draw in Bellahcene
– Tregubov, Brest 2018.

401
19...Nf6 20.Rae1 Kg7 21.Rxe8 Rxe8 22.Rb1 c5!
Black is fine after this typical central break, and White found nothing better than forcing a draw
with:
 
23.Bxf7 Kxf7 24.Rxb7† Re7 25.Ng5† Kg7 26.Rxe7† Bxe7 27.Qc4 Kh6 28.Nf7† Kg7 29.Ng5=
The players repeated a few more times before calling it a day in Massy – Deskin, corr. 2018.
 

Conclusion
 
In this, the final chapter on 5.d4, we analysed the important option of 9.Nc3, when 9...Nxc3 10.bxc3
dxc4 11.Bxc4 Bf5 is our choice. Then 12.Ne5 Nd7 13.f4 is an aggressive choice, but 13...Nf6!N
leads to a complex position where my analysis shows that Black more than enough resources to deal
with White’s pawn advances on the kingside. We then turned to 12.Re1, which is a sensible move but
after 12...Nd7 White should really transpose to a later variation with 13.Bg5, as independent
alternatives give him nothing.
 
12.Ng5 is the first of two theoretically critical options. 12...h6 is our reply, when the 13.Nxf7N
sacrifice does not work as White would like it to, but 13.Qf3! is more challenging. The main line
sees Black grab an extra pawn before returning it, resulting in an endgame where any advantage of
White’s bishop pair is cancelled out by his inferior pawn structure.
 
Finally 12.Bg5 is an active move, when 12...Qa5! is our energetic reply, with veiled threats against
the bishop which White just moved. Then 13.Re1, 13.d5!?N and 13.Nh4 all demand attention but my
analysis shows that the Petroff remains bulletproof in all variations.

402
A) 7.Bf4 0-0 8.Qd2 Nd7 9.0-0-0 Nc5! 201
A1) 10.Nd4 202
A2) 10.h4 204
A3) 10.Be3 Ne4! 205
A31) 11.Qe1 206
A32) 11.Qd4 207
B) 7.Be3 0-0 209
B1) 8.Bd3 Nd7 210
B11) 9.Qd2 210
B12) 9.Qe2 211
B2) 8.Qd2 Nd7 9.0-0-0 Nf6 10.Bd3 c5 212
B21) 11.c4?! 216
B22) 11.Bg5 218

 
1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4 5.Nc3

403
 This move has been known for well over a century but it has been regarded as a sideline for most
of that period. However, more recently it became clear that White’s plan involving dxc3 followed by
quick development and long castling has real bite. At the time of writing it is White’s most
fashionable try for an opening advantage, so we may think of it as the ‘modern main line.’
 
5...Nxc3 6.dxc3 Be7
White’s only ambitious idea is to arrange long castling, so his choice boils down to A) 7.Bf4 and
B) 7.Be3.
 
Castling on the kingside is less challenging, but still playable of course, so we should take a quick
look at it:
7.Bd3 0-0 8.0-0
The simplest way to deal with this set-up is to bring our knight to c5, where it is not only well
placed but also proves to be a source of trouble for the d3-bishop.

8...Nd7
It may look tempting to pin the enemy knight with 8...Bg4?! but White has a strong response in
9.Qe2!, when the threat of Qe4 provokes weaknesses in Black’s camp. 9...d5 (I also checked
9...Nc6N 10.Qe4 f5 11.Qd5† Kh8 12.Nd4² when the weakness of the e6-square causes Black
some problems) Now in Guzman Freire – J. Sanchez, Santo Domingo 2010, the most energetic
continuation would have been 10.h3N 10...Bh5 11.g4! Bg6 12.Bxg6 hxg6 13.Re1² when White
may claim a slight pull due to his better development.
9.Be3
White plans to exchange on c5 rather than moving his bishop away from d3.
9...Nc5 10.Bxc5 dxc5 11.Re1
White enjoys a slight lead in development but, once Black is able to coordinate his pieces, his
two bishops may make their presence felt.

404
11...g6
Playing against the enemy bishop.
12.Qe2 Bd6 13.Nd2 Kg7 14.g3 Bd7
The position was equal in Mirkowski – Sorcnik, corr. 2016, as White can trade his knight for one
of the bishops.
 

A) 7.Bf4 0-0
 
Black has mostly reacted to White’s last move by arranging long castling. However, for the sake of
clarity and comfort, I recommend a set-up with ...0-0, similar to how we will play against the more
popular 7.Be3. The main difference is that without a bishop on e3, Black will be able to conveniently
deploy his knight on c5.
8.Qd2 Nd7 9.0-0-0

405
9...Nc5!
With the bishop on f4, it’s important that the knight goes to c5 rather than f6.
 
If we continue in the same way as in variation B2 with 9...Nf6, it will be harder to create counterplay
on the queenside. The reason is that the ...c5 advance is an integral part of Black’s plans, but it will
be difficult to execute because the d6-pawn is already under pressure from the bishop on f4.
After the text move White has three main tries: A1) 10.Nd4, A2) 10.h4 and A3) 10.Be3.
 
I also considered:
10.Kb1 Bf6
10...Re8 is also playable and is quite likely to transpose to the line below after a few more
moves.

406
11.Be3
11.Nd4 prepares f2-f3, so Black should play: 11...Ne4! 12.Qe3 Re8 13.Qf3 Bd7 14.Bd3 d5
Cementing the outpost on e4. 15.Bxe4 Rxe4 16.Rhe1 c5 17.Nb3 Qe8³ I. Agrest – Duda,
Internet 2019.
11...Ne4
This is the typical reaction whenever White drops the bishop back to e3.
12.Qd5 Re8 13.Bc4
After 13.Bd3 c6 14.Qh5 d5= Black has arranged his pieces well and is completely fine, Salcedo
Mederos – Gonzalez, corr. 2011.
13...Re7 14.Bd3 Qe8!
I consider this move the most precise, although 14...c6 was also playable in R. Smith –
Verducci Silvelo, corr. 2006.

407
15.Qh5 Bd7N
15...b6 was seen in Churkina – Abdulmanov, corr. 2014, but I slightly prefer the text move.
16.h4 g6 17.Qh6 Bg7 18.Qf4 h5=
Black’s knight is powerful and the rest of his position is firmly under control.
 

A1) 10.Nd4
 
This has been tested by some strong players, but no one at the elite level has played it in almost a
decade. The correct antidote is:
 
10...Re8 11.f3 Ne6 12.Be3

408
12...Bg5!
Provoking a weakening of the e4-square.
 
13.f4
Other moves fail to challenge Black. For instance, 13.h4 Bxe3 14.Qxe3 Nxd4 15.Qxd4 Qe7
16.Bc4 Qe3† 17.Qxe3 Rxe3 reached an equal endgame and the players agreed a draw in Polgar –
Ivanchuk, Hoogeveen 2009.
 
13...Bf6
After 13...Nxd4 14.fxg5 Nc6 15.h4 Black was able to hold a draw in Adams – Kramnik, Wijk aan
Zee 2008, but I would prefer not to give White the bishop pair and chances to cause problems on the
dark squares.
 
The text move avoids any such problems, and Black intends to exploit the newly created weakness
with ...Nc5-e4.
 
14.Nf3!?
White resorts to prophylaxis to prevent Black’s idea. White also intends to gain space on the
kingside with f4-f5 or g2-g4 next.
 
14.h3?! completely ignores Black’s idea. After 14...Nc5 White resorted to 15.Bd3 to cover the e4-
square in Grabics – Srebrnic, Rijeka 2005, and now Black could simply have played 15...Nxd3†N
16.Qxd3 b6!³ intending ...Bb7, when the bishop will exert a powerful influence on the long diagonal.

409
14...c6!?N
This is my improvement on 14...Nc5?! 15.Bxc5 dxc5 16.Qf2! when White wins a pawn. True,
after 16...Qe7 17.Re1 Qf8 18.Rxe8 Qxe8 19.Qxc5 Black had some compensation in Neelotpal – S.
Singh, Calicut 2003, but I’m not convinced that it’s enough.
The text move has a few purposes: Black frees the path of the queen to go to the queenside, as well
as the c7-square for the knight to go to d5. There is also the matter of leaving the d6-pawn en prise –
but as we will see, capturing it does not give White as easy game.
 
15.f5
This seems logical, as White gains space on the kingside and drives the knight back to c7, where it
blocks the intended path of Black’s queen.
 
15.Qxd6 is met by 15...Qxd6 16.Rxd6 Nc7 17.Bd2 Nd5 when the position is dynamically balanced:
White has an extra pawn but his rook is in danger of being trapped.
 
15.Bd3 is met by 15...Qa5 16.a3 Nc5= with good prospects on the queenside.
 
15...Nc7 16.g4

410
16...Nd5 17.Bf2 Qa5„
Black has good counterplay on the queenside.
 

A2) 10.h4

Black should respond to this by continuing with his normal development.

 
10...Re8 11.Bc4

411
 11.Be3 can be met by 11...Ne4 12.Qd3N (12.Qe1?! is too passive, and after 12...Bg4 13.Bd3 Bf6³
Black already had the easier game due to having such active pieces in C. van Oosterom – Skatchkov,
Groningen 2005) 12...Bf6 13.Ng5 Bf5= and Black is fine.
 
11...Be6
Simply exchanging White’s active bishop.
 
12.Bxe6 Nxe6

13.Be3 Qd7
With the idea to go to c6, activating the queen. Once that has been done, Black can start pushing
his pawns with ...b5 and ...a5, with good prospects on the queenside.
 
14.Qd3
I also checked:
14.Kb1 Qc6 15.Rh3
15.Ng5?! Bxg5 16.hxg5 was seen in Elistratov – Khajrutdinov, Voronezh 2016, and now Black
could have safely grabbed a pawn: 16...Qxg2!N 17.Rh4 Nf8³ With the knight guarding the h7-
pawn, Black’s king is safe.
The main idea of the text move is to prepare g2-g4. My suggestion for Black is:

412
15...b5!?N
15...Nc5 and 15...Bf6 have been tested at GM level but I see no reason to delay our attack.
16.g4 a5 17.Nd4 Qe4 18.Nxe6 Qxe6 19.Rg3 b4
Black has fine play on the queenside.
 
Cohen evaluates the present position as slightly better for White (the Israeli FM recommends 10...c6
for Black instead of 10...Re8) but Black is doing fine after:
 
14...Qc6 15.Qf5
Black has no problems after 15.Rde1 Nc5 16.Bxc5 Qxc5 17.Re3 Bf6 18.Rhe1 Rxe3 19.Rxe3 Qc6=
intending ...Re8, and after 20.g3 the players agreed a draw in Almasi – Fridman, Germany 2005.
 
15.Ng5 is simply met by 15...Nf8 16.f3 Rad8 intending ...d6-d5. A good example continued: 17.c4
a6 Now Black is ready to break with ...b5. 18.Ne4 White wants to bring the knight back to c3.
18...Nd7 19.Qd5 Now Black can improve on Rehorek – Vojta, Czech Republic 2006, with:

413
19...Qxd5N 20.cxd5 (or 20.Rxd5 Nf6³) 20...f5 21.Ng3 g6³
 
15...a5
A typical idea, gaining space on the queenside.
 
16.h5
16.Ng5 can be met by 16...Bxg5 17.hxg5 Nf8 and with the h7-pawn safely guarded, Black is ready
for active play on the queenside, as shown after: 18.Rh2 Re5 19.Qd3 b5„

16...a4 17.a3 b6!

Preparing to bring the rook into action via the a5-square.

414
 
18.Nd4 Nxd4 19.cxd4 d5=
Black was completely fine in Krebs – Hauenstein, corr. 2010.
 

A3) 10.Be3

This is the most frequently played move in this position, but I am not impressed by it and I believe
Black has a simple antidote.
 
10...Ne4!
This move is rarely played but highly effective! Black attacks the queen and forces it to move to a
more exposed square. In some cases our knight will drop back to f6, leading to something similar to
the main line (see variation B2 on page 212) but with White’s queen on a slightly worse square.
 
We will analyse A31) 11.Qe1 and A32) 11.Qd4.
 
11.Qd5? simply loses a tempo, as Black is happy to play 11...Nf6 anyway. 12.Qd4 (12.Qd2 reaches
the exact position of variation B2 but with Black to move instead of White! Instead of repeating with
...Ne4, Black can and should aim for more with 12...c5!³ as in Moser – Babic, corr. 2013, when the
extra tempo gives Black good prospects to seize the initiative on the queenside.) 12...c5 13.Qf4 In T.
Petrosian – Rabiega, Internet 2004, the players reached the same position as in variation A32 below,
but with White having wasted a tempo on Qd5-d4-f4. Black should have continued:

415
13...d5!N This is the thematic follow-up to ...c5, sidestepping the threat of Bxc5 and gaining space in
the centre. White has wasted too much time with his queen and Black is clearly better.
 

A31) 11.Qe1

This rather invites Black to target the queen with:

 
11...Re8 12.Bd3 Bf6
Black has a pleasant game with better-placed pieces.

416
 
13.Qe2
Another game continued 13.Kb1 Bd7 14.Bxe4 Rxe4 when White eliminated the annoying knight,
but after trading off his light-squared bishop he can hardly hope for a successful attack on the
kingside. 15.Qd2 Now in Koch – Koronowski, corr. 2007, Black’s best would have been:

15...Qe7!N 16.Rhe1 Re8³ With the bishop pair and triple power on the e-file, Black is obviously
doing well.
13.Qg1?! is a creative way to sidestep the black rook and support the advance of the g-pawn, but
these factors do not justify placing the queen on such a square. 13...Qd7! is a strong reply, preventing
any g2-g4 ideas and also planning to go to a4. 14.Nd4 Now in Evchin – Karnaukh, Kiev 2004, the
best continuation would have been:

417
14...Qa4!N With a strong attack. For instance, 15.Kb1 Be6 16.Nxe6 Nxc3†! and White is crushed.
 
13...Nc5!
Preparing to eliminate the important bishop on d3.
 
14.Rhe1
Now in Krueger – Egoshin, corr. 2015, Black should have simply played:

14...Nxd3†N 15.Qxd3 a6³

With slightly better chances due to the bishop pair.
 

418
 A32) 11.Qd4

11...Nf6
If you compare this position to variation B2 on page 212, you will see that both sides have ‘wasted’
time: Black’s knight has spent four tempos rather than two to get from b8 to f6, while White has
spent one extra tempo with each of his dark-squared bishop and his queen. The bishop is on e3 in
both lines, but in the present variant the white queen is on d4 instead of d2. Black’s plans generally
involve ...c5 anyway, so it is obvious that White’s queen is exposed on d4 and will have to waste
additional time. The only exception to this rule would be if White could profit from moving the
queen to f4 or h4, but that does not seem to be the case.
 
12.h3
Preparing g2-g4.
 
12.h4 c5 13.Qf4 was seen in Osmak – Fridman, Skopje 2019, and now Black can improve with:
 

419
13...d5!N This is often a desirable follow-up to ...c5 in these positions, as Black continues to gain
space in the centre. 14.h5 Bd6 15.Qh4 Re8µ Black has achieved his desired c5-d5 pawn duo, while
White has difficulties developing his kingside attack.
 
12.Qf4 was seen in Dauga – Roos, corr. 2011. White plays prophylactically to prevent ...c5, but
Black can continue attacking the queen with: 12...Nh5!N 13.Qe4 c6 With ...c5 still not possible,
Black tries to get in ...d5.

14.c4 (also after 14.Bd3 g6 15.Bh6 Re8³ the queen on e4 remains a headache for White, being in the
firing line of Black’s pieces) 14...Re8³ Black develops swiftly, while White still has problems
starting action on the kingside because of the exposed position of his queen.

420
12...c5!
This thematic move becomes even stronger when the queen is on d4.
 
13.Qh4
In the event of 13.Qf4 d5 14.Bd3 Re8³ White’s queen is still exposed while Black’s pieces are
optimally placed, and Black can slowly build up pressure against the white king.
 
13...d5
This is the simplest move, although it does allow White to more or less force a draw if he wants it.
 
13...Re8!? occurred in Krueger – Schmitz, corr. 2014, and is a completely playable alternative if
Black wants more than a draw.
 
14.g4!?N
The drawing line is 14.Bd3 Ne4 15.Qh5 Nf6 16.Qh4 Ne4 17.Qh5 Nf6 18.Qh4 Ne4 ½–½ as seen
in Krüger – Sugonyak, corr. 2018.
 
The text move is a natural attempt to grab the initiative on the kingside. Black should respond
equally aggressively with:
 

421
14...Ne4 15.Qh5 Qa5! 16.a3
White cannot play 16.Qxd5?? due to 16...Nxc3! 17.bxc3 Qa3† 18.Kd2 (or 18.Kb1 Be6–+)
18...Rd8 and Black wins.
 
16...Be6 17.Bd3
White’s idea is to take on e4 and then play Nd4, exploiting the pin on the c5-pawn as the queen on
a5 is unprotected. Therefore we should move the queen to a safe square.
 
17...Qa4
White is under pressure on the queenside so he should try to liquidate with:

422
18.Ng5 Bxg5 19.Bxg5 f5!
Covering the b1-h7 diagonal and threatening ...Nxf2.
 
20.Be3 b6 21.gxf5 Bxf5=
The computer calls it equal but, thanks to the strong knight on e4 and better pawn structure on the
kingside, I would slightly prefer Black’s chances in a practical game.
 

B) 7.Be3

This move is deservedly the main line.

 
7...0-0
White’s most important options are B1) 8.Bd3 and B2) 8.Qd2. In both cases he generally castles on
the queenside, but the former move gives him the option of placing his queen on e2 instead of d2.
 
Black has an easier time after:
8.Bc4
This invites the simple plan of ...c6 followed by ...d5, when White will have to waste time
moving the bishop again.
8...c6 9.Qd2
9.0-0 cannot threaten us of course, and after 9...Nd7 10.Re1 d5 11.Bf1 Nf6 12.h3 Re8= Black
was fine in Aronian – Giri, Internet 2018.
9...d5 10.Bd3
Also after 10.Be2 Nd7 11.0-0-0 Nf6 Black has no problems.

423
10...Nd7 11.0-0-0
You can find many games in the database that reached this position, but in all of them White
was to move! Black can use his extra tempo to play:
11...Nc5=
Exchanging the d3-bishop and with a swift development of his pieces to follow, Black has good
prospects.
 

B1) 8.Bd3
 
The main significance of this move is that White can follow it by developing his queen to e2 rather
than d2 as in the main line. However, the drawback is that Black will be able to eliminate one of the
enemy bishops by bringing his knight to c5, which will substantially reduce White’s attacking
potential.
 
8...Nd7
Despite the previous comment, White often plays B11) 9.Qd2, even though B12) 9.Qe2 is clearly
the consistent follow-up to his last move.
 

B11) 9.Qd2 d5!

 
Black wants to play ...Nc5 next, without getting his structure damaged.
 
9...Nc5?! works well with the queen on e2, but here it is not so effective due to 10.Bxc5 dxc5 11.0-0-
0 when White’s queen is better placed on d2 than on e2.
 
There is nothing wrong with 9...Nf6, when the natural 10.0-0-0 transposes to the main line, as
covered in variation B2 below. However, since White has developed his bishop to d3 prematurely, I
think we should take the opportunity to target the bishop.
 

424
10.0-0-0 Nc5 11.Kb1
11.Bxc5 Bxc5 12.c4 c6 13.cxd5 Qxd5³ favoured Black due to the bishop pair in Weishaeutel –
Horn, Kiel 2014.
 
11...c6
Black does not need to hurry to exchange on d3, as moving the bishop will waste time for White.
 
12.h4 Nxd3 13.Qxd3 Qd6
Black plans to exchange queens with ...Qg6.

14.Rde1!?
425
 This prevents Black’s idea, for now.
 
14.c4 was played in a high-level game, when 14...Rd8 15.cxd5 Qxd5 16.Qxd5 cxd5 17.Nd4 Bd7
reached an equal position, since Black’s two bishops fully compensate for the isolated pawn. The
game soon ended in a draw by repetition after:

18.Nb3 Bf5 19.Nd4 Bd7 20.Nb3 Bf5 21.Nd4 ½–½ Aronian – Caruana, London 2018.
 
14...Re8
14...Qg6? would be a mistake due to 15.Qxg6 hxg6 16.Bxa7!.
 
15.Ng5 Qg6 16.Qxg6 hxg6 17.Bc5 Be6 18.Nxe6 fxe6
White has no advantage in this endgame, as the following example demonstrates:

426
 
19.Bd4 Kf7 20.Rh3 c5 21.Be5 Rh8 22.Rf3† Kg8 23.g3 g5 24.hxg5 Bxg5=
Black went on to draw quite comfortably in Krivic – Moll, corr. 2013
 

B12) 9.Qe2

This is the more logical follow-up to 8.Bd3.

 
9...Nc5!?

427
 9...d5 10.0-0-0 Nc5 is playable here too. However, the text move seems simpler as Black has
nothing to fear from the doubling of his pawns.
 
10.Bxc5
It’s better for White to exchange his dark-squared bishop, as its colleague on d3 is needed to create
some pressure on the kingside.
 
Ponomariov tried to save the bishop pair with 10.Bc4?! but it proved futile after: 10...d5 11.Bd3 (or
11.0-0-0 c6³ and Black will kill the bishop sooner or later) 11...Nxd3† 12.Qxd3 c5³ White’s eventual
victory was certainly not due to the quality of his opening play in Ponomariov – Ni Hua, Riyadh
(rapid) 2017.
 
10...dxc5 11.0-0-0 Bd6

12.Rhe1
White centralizes the rook and prevents ...Re8, making Black work to develop his pieces.
 
12.Kb1?! was played in Mamedov – Gelfand, Moscow 2010, when the simple 12...Re8N 13.Qd2
Bg4³ would have given Black a pleasant position with the bishop pair.
 
12.h4?! was played in Al Qudaimi – Eid, Beirut 2007, when the same recipe would have worked
well: 12...Re8N 13.Qf1 Bg4³ and Black is doing well.
 
12...Qf6 13.Kb1 Bd7
Now Black is ready for ...Rfe8.

428
14.Ne5 Be6 15.Qe4 g6 16.f4 Rad8 17.g3 c6=
Black had no problems in Nestorovic – Lenderman, Washington DC 2018.
 

B2) 8.Qd2

This is the main line, postponing the development of the light-squared bishop and prioritizing long
castling.
 
8...Nd7 9.0-0-0

429
 9.Bd3 leads back to variation B11.
 
The text move brings us to the main tabiya of the Modern Main Line. Black has tried numerous
moves here; for instance, one popular option is 9...c6 intending ...d5. By playing that way, Black
takes control over some central territory and aims to neutralize White’s initiative on the kingside by
exchanging some pieces, especially with ...Bc5 or ...Nc5. This way of handling the position has a
solid reputation, but I eventually decided to recommend an equally sound but more active and
dynamic set-up.
 
9...Nf6
Until just a few years ago, this move had mostly flown under the radar. For instance, John Shaw
did not mention it at all in his 2016 book Playing 1.e4 – Caro-Kann, 1...e5 & Minor Lines. However,
a few strong GMs started playing it in 2017, and when Caruana used it to equalize effortlessly against
Carlsen in their 2018 World Championship match, a lot of players took notice. Since then, it has
continued to prove its worth at the highest level. The knight is useful on f6 for defensive purposes on
the kingside, but it also supports the ...d5 advance. Moreover, by controlling the d5-square Black
facilitates the ...c5 advance, which is the key to his counterplay in many lines, as we will see.
 
10.Bd3
White usually develops the bishop here before deciding how to move forward on the kingside. I
also considered a couple of ways in which White may start attacking while leaving the bishop on f1:
 
10.h3 c5
Black plays just as in the main line. Now White can start gaining space on the kingside
immediately with:
11.g4
11.Bd3 transposes to the 11.h3 line in the notes to the main line below.
11.Bf4?! was once played by Caruana, but it seems like a waste of time because the threat of
Bxd6 can be parried easily by: 11...Be6! 12.a3 (12.Bxd6? would be a mistake due to 12...Bxd6
13.Qxd6 Qa5µ when the threats of ...Qxa2 and ...Ne4 are too strong) 12...d5³ Black has
developed normally while the Bf4 move seems useless, Caruana – Hou Yifan,
Karlsruhe/Baden-Baden 2018.
Now in Atlas – Kreisl, Austria 2019, Black should have continued developing in the normal
way with:

430
11...Be6N 12.a3 d5
Black has achieved his optimal pawn structure in the centre, which we will witness in the main
line as well.
13.Bg2 Ne4 14.Qe1 Qc7=
Black has a good game and is ready to develop his attack on the queenside.
 
10.h4!? c5
Black continues in the usual way, freeing the queen’s path to a5.

11.Kb1
This prophylactic move seems sensible.

431
 11.h5N can be met by 11...h6 when White has no obvious way through on the kingside. Black’s
last move not only prevents h5-h6 but also prepares ...Be6 without allowing Ng5. With ...Qa5
coming as well, Black has at least equal chances.
11...Qa5 12.a3
12.c4 is a typical reaction when the queen arrives on a5: White plays it safe and Black generally
gets a comfortable endgame. 12...Qxd2 13.Rxd2 (13.Nxd2 Ng4 would be inconvenient for
White) Now 13...Bd7 seems best, intending to place the bishop on the long diagonal. 14.Bd3
Bc6= Black will put his rooks on e8 and d8 next, with a pleasant position.
12...d5
Gaining space in the centre.
13.Ne5 Be6
Defending the d5-pawn and intending ...Ne4.
14.f3 Rfd8 15.g4

15...d4!
An essential strike in the centre before White’s flank attack can gather momentum.
16.cxd4 Qa4!
Of course Black avoids exchanging queens in order to maintain the pressure.
17.g5
Now Black’s play can be improved with:
17...cxd4!N
After 17...Nd5?! things started to go downhill for Black in Fedorchuk – Di Nicolantonio, Cap
d’Agde 2018.
18.Bxd4 Bc5!
Keeping up the momentum.
19.c3 Bxd4 20.cxd4

432
20...Nd7! 21.Nxd7 Qxd7ƒ
Black has a strong initiative, with ideas like ...Qd5 in the air.

10...c5
This move is an integral part of Black’s scheme. It covers the d4-square and gives the queen a path
to a5, from where she can create threats against the white king. Another important idea for Black is
to play ...Be6 and then gain space in the centre and on the queenside with ...d5 and ...c4.
 
In the remainder of this chapter, we will focus on two quite important sidelines: B21) 11.c4?! and
B22) 11.Bg5. As you can tell from the punctuation, the former is objectively weaker, but I analysed it

433
as a main line because the way Black meets it is highly instructive, and important to one’s overall
understanding of our chosen set-up.
 
The most important options are 11.Kb1 and 11.Rhe1, and we will analyse them in the next chapter.
White can also move his h-pawn either one or two squares:
 
11.h3
White covers the g4-square and aims to launch his g-pawn at some point.
11...Be6
11...Qa5 is possible as well, but in general I prefer to place the bishop on e6 first and only then
think about the queen.
12.Kb1
12.Ng5? was a mistake in Kopasz – Papp, Hungary 2019, due to the blunt 12...Bxa2!N and if
13.b3 a5 14.Kb2 a4µ White may regret trying to trap the bishop.
12.a3?! is also bad as it creates a ‘hook’ which Black can use to open the queenside, either with
...b5-b4 or a ...Bxa3 sacrifice. 12...d5 13.Ng5 (13.Bg5 b5!‚) 13...b5!µ
12...d5 13.Ng5 c4 14.Be2

14...Rc8!
It is well worth remembering this motif! The rook goes to c6 and the queen will go to a5,
putting pressure on White’s queenside. Play might continue:
15.Bd4 Rc6 16.Rhe1 Qa5
Threatening ...Ra6.
17.Qc1 Bc5!³
Exchanging the strong bishop on d4, with better chances for Black.
 
11.h4

434
 This move can be met in much the same way as the line above.
11...Be6 12.Kb1 Qa5 13.a3
Perhaps White should aim to reduce risk with 13.c4, which leads to an equal endgame after:
13...Qxd2 14.Bxd2 h6 Preventing any Ng5 ideas. 15.Rhe1 Rfe8 16.b3 d5= Mishra –
Balakrishnan, Las Vegas 2018 (Black can also consider 16...Rad8N= followed by ...d5 next).
13...d5!
13...c4 has also been played but it is usually preferable to play ...d5 before ...c4.
14.Ng5
14.h5? is not possible due to 14...c4 15.Be2 Ne4 16.Qe1 Bxa3! with a crushing attack.

14...c4 15.Be2
Now in Fong – Nazario, Pune 2014, Black missed the strongest continuation:
15...Rac8!N
Black enjoys some initiative. The idea of the last move is exactly the same as in the previous note:
the rook is heading for c6, where it will be optimally placed for attacking on the queenside and
defending the e6-point.
 

B21) 11.c4?!

435
 This move prevents ...Qa5 and also turns the d6-pawn into a long-term target. However, it has a
major drawback which Black should exploit immediately with:
 
11...b5!
Black strikes immediately to open lines against the enemy king.
 
12.cxb5
12.Rhe1 was played in Mascarenhas – Chaves, Rio de Janeiro 2013, when 12...Be6N would have
forced White to take on b5: 13.cxb5 a6!

436
 Opening lines on the queenside favours Black, so the best White can do is to block things up as
much as possible with 14.b6 Qxb6 15.c4. Still, after 15...Rfe8³ Black is slightly better as he has
opened a file on the queenside, whereas White has yet to achieve anything on the kingside.
 
12.Bf4?
With this move, White naively hopes for ...bxc4, after which Bxc4 will give White ready-made
pressure on the d6-pawn.
12...Be6!
Black insists on sacrificing the pawn.
13.Rhe1?
13.cxb5N can be met by 13...c4 14.Be2 a6! 15.b6 (15.Bxd6? axb5 16.Bxe7 Qxe7 is simply
winning for Black) 15...Qxb6µ and Black’s attack is too strong. The text move is even worse
though.
13...bxc4 14.Bf1
Black has a choice of winning lines:

14...Rb8!N
This is the fanciest and most efficient solution, although there was nothing at all wrong with
14...d5–+ in Bystrov – Slepov, Litomysl 2019.
15.Bxd6 c3! 16.bxc3
16.Qxc3 Bxd6 is an extra piece.
16...Qa5
With a winning attack.
 
12...a6!
In the spirit of the Benko Gambit! Obviously White has already castled queenside here, making
Black’s initiative much more dangerous.

437
 
13.b6
13.bxa6
In Gueci – Estremera Panos, Stockholm 2019, Black recaptured on a6 with his bishop, but it
would have been stronger to play:
13...Be6!N
Threatening ...c4 followed by ...Rxa6.
14.Bb5
14.a3? is met by 14...c4 15.Be2 Rxa6–+ followed by ...d5 when the e7-bishop will also join the
attack.
14.c4? is also no good in view of 14...Rxa6 15.a3 Qa8 16.Ng5 Rb8–+ with the devastating
threat of ...Rxa3.
The text move prevents the ...c4 idea, but Black can simply play:

14...Bxa2! 15.c4 Bb3 16.Rde1 d5!ƒ

With excellent play on the queenside.
 
13...Qxb6 14.c4
Now in Lagarde – Michalik, Germany 2018, Black broke in the centre with 14...d5?! which seems
a bit premature. It is better to prepare the central break with:
 

438
14...Be6!N
Intending to play ...d5 with added force.
 
15.Ng5
15.Rhe1!? disrupts Black’s plan in a subtle way but 15...Rfe8³ is a good reply, when it is clear that
Black’s chances on the queenside are better than White’s on the kingside. (Instead 15...d5?! would be
met by 16.Ng5! when 16...dxc4? is not working due to 17.Nxe6 Qxe6 18.Bxc5± when the value of
the rook on e1 is clear.)
 
15...Bd7
Black has better chances, as shown by the illustrative line:
 
16.Qc2 h6 17.h4!
White’s best chance is to maintain the knight on g5 with this thematic move.
 
17...Rab8 18.f3 d5!?
Black sacrifices a pawn to activate his dark-squared bishop.
 
19.cxd5 Bd6
Intending ...Be5.
 
20.Ne4 Nxe4 21.fxe4 Rfc8©
Black has superb compensation for the pawn, with ideas such as ...c4 and ...Be5 in the air.

439
 B22) 11.Bg5

The main idea behind this move is to transfer the queen to the kingside with Qf4-h4. In reply to
...h6 White can always think of sacrificing his bishop, so it’s better to keep the h-pawn at home for
the moment.
 
11...Be6 12.Qf4
White goes for a direct attempt to transfer the queen to the kingside, without spending a tempo
developing the h1-rook.
 
12.Rhe1 transposes to one of the main lines – see variation B3 of the next chapter on page 235.
 
12...h6!N
This is not the only decent move but it seems like the cleanest equalizer, as it leads to a forced
variation which ultimately concludes in perpetual check.
 
12...d5 was a playable alternative seen in Svidler – Duda, Riga 2019, but White has a few more
options at his disposal against this move.
 

440
13.Bxh6!?
This sacrifice is very much in the spirit of the 11.Bg5 variation.
 
The alternative is: 13.Bxf6 Bxf6 14.Qe4 g6 15.Bc4! White must maintain control over the d5-square.
(15.Qxb7? is too greedy and 15...Rb8 soon leads to a crushing attack) 15...Re8 16.Bxe6 Rxe6
17.Qxb7 Rb8 Black has full compensation for the pawn due to his initiative on the queenside.
 
13...gxh6 14.Qxh6 Re8 15.Qg5† Kf8 16.Qh6† Kg8
White can either take a draw immediately or play for more with:
 
17.Ng5!?

441
 This looks tempting but Black has adequate defensive resources, and White also has to play
accurately to maintain his initiative. The following line may seem rather long, but most of the moves
are forced on both sides so it should not be too difficult to learn.
 
17...Bf8 18.Bh7† Kh8 19.Qh4 Ng4!
An important move, allowing Black to insert the bishop on h6.
 
20.Bf5†
It’s important to place the bishop on f5, as this creates an additional threat of Bxe6 followed by
Nf7†.
 
20...Bh6 21.f4 Qf6!
Another important move, exploiting the unsupported f4-pawn.

442
22.Bxe6 Qxf4† 23.Kb1 fxe6 24.Nh3 Qf5
24...Qe4? is inferior due to 25.Nf2 Nxf2 26.Qxh6† Kg8 27.Rhf1!±.
 
25.Rdf1 Qg6 26.Rf4!
A strong resource.
 
26...Ne3 27.Rf6 Qxc2† 28.Ka1 Nf5 29.Qh5
 

29...Re7!
The rook comes to the rescue.
443
 
30.Ng5!
30.g4? loses to 30...Rh7! with the terrible threat of ...Bc1.
 
30...Qxg2
Many pieces have been eliminated and the game is approaching its conclusion.
 
31.Rxh6† Nxh6 32.Qxh6† Kg8 33.Qg6†

33...Rg7!
Other moves would lose the queen to a discovered attack.
 
34.Qxe6† Kh8 35.Qh6† Kg8 36.Qe6†
The game ends in a perpetual.
 

Conclusion
 
5.Nc3 Nxc3 6.dxc3 has taken over as the Main Line of the Petroff at all levels. I recommend 6...Be7
followed by short castling no matter how White plays. He almost always castles on the queenside,
and the only question is how he will prepare it. First we analysed 7.Bf4 0-0 8.Qd2 Nd7 9.0-0-0,
when 9...Nc5 is the correct choice, taking into account the placement of the bishop on f4. There are
some lines worth learning, but Black essentially has no problems.
 
7.Be3 is the more popular and challenging move. Again 7...0-0 is our choice, but this time after
8.Qd2 Nd7 9.0-0-0 we play 9...Nf6, taking the knight on a different route. The general idea is to

444
follow up with ...c5, taking control over the d4-square and gaining space on the queenside while
preparing ...Qa5. We saw a number of lines where Black obtains nice counterplay, and we will
continue our investigation in the next chapter where we analyse White’s most popular attempts to
find an advantage.

445
A) 11.Kb1 b5! 12.Rhe1 Rb8 13.Bg5!? Be6 14.Qf4 Qa5 222
A1) 15.b3 224
A2) 15.Rxe6!N 226
B) 11.Rhe1 Be6 227
B1) 12.a3 228
B2) 12.Kb1 Qa5 229
B21) 13.c4 229
B22) 13.a3 231
B3) 12.Bg5 235

 
1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4 5.Nc3 Nxc3 6.dxc3 Be7 7.Be3 0-0 8.Qd2 Nd7 9.0-0-0
Nf6 10.Bd3 c5
This chapter will deal with two important options: A) 11.Kb1 and B) 11.Rhe1.
 

A) 11.Kb1

446
 This is generally a useful move, but in this instance it allows Black to go on the offensive
immediately with:
 
11...b5! 12.Rhe1
12.Ng5? seems pointless. Black can continue logically with: 12...Rb8 13.f3?! (13.Ne4N is the
lesser evil although the simple 13...Be6³ leaves Black with a useful initiative on the queenside)
13...d5! 14.h4 c4 15.Be2 b4 Black already had a winning attack in Zatonskih – Kosteniuk, Internet
2019.
 
12.Bxb5?!
Taking the pawn is extremely risky for White. In Sanal – Antic, Belgrade 2018, the best way to
build Black’s attack would have been:
12...Be6!N
Threatening ...Bxa2† followed by ...Qa5†.
13.Rhe1
13.Qe2 can be met by 13...Rb8 14.c4 Ne4‚ when Black threatens ...a6, and also plans to bring
the dark-squared bishop into the attack with ...Bf6, hitting the b2-pawn.

447
13...Rb8!
13...Bxa2†?! should be avoided due to 14.Kxa2 Qa5† 15.Kb1 Qxb5 16.Bxc5!² when Black
will be a pawn down for insufficient compensation.
14.c4
14.Bf1?? Bxa2†! is crushing.
14...a6 15.Ba4 Bxc4 16.Bb3 Bb5µ
Black has a strong initiative on the queenside, the immediate threat being ...c4.
 
12.Bg5
The idea of this move is to exchange on f6 and play Be4, attacking both the a8-rook and the d6-
pawn.
12...Be6 13.Bxf6?!
It turns out that taking immediately is bad for White.
13.Rhe1N is preferable, when 13...Rb8 transposes to the main line below.
13...Bxf6 14.Be4

448
14...b4!
Black ignores the attack on his rook and immediately goes for the kill!
15.c4
15.Bxa8? loses to 15...bxc3 16.Qe3 (16.bxc3? Qa5 is already resignable for White) 16...Bxa2†!
17.Kxa2 Qa5† 18.Kb1 and now the cool 18...Rxa8 leaves Black with a winning attack.
15...Rb8 16.Qd3 h6 17.Rhe1 Rb6³
Black had good attacking chances on the queenside in Tukhaev – Sanal, St Petersburg 2018.

12...Rb8
12...Be6?! is strongly met by 13.Ng5 Bd7 14.Ne4! and now after 14...Bc6 15.Ng3!ƒ intending
Nf5 White will gain a strong initiative on the kingside.

449
 
13.Bg5!?
This nice idea by Vidit worked wonders, as Gelfand fell into the trap straight away. White plans to
transfer his queen to h4, creating strong attacking chances. Vidit’s success could inspire other players
to try the same idea for White, so I advise the reader to pay special attention to this line.
 
A previous game continued: 13.Bf4 Threatening Rxe7 followed by Bxd6. 13...Re8! 14.Ng5 (the
point of Black’s last move is that 14.Rxe7? doesn’t work due to 14...Rxe7 15.Bxd6 Qxd6 16.Bxh7†
Kxh7 17.Qxd6 and now 17...Rd7 wins for Black) Now in Rodshtein – Hoolt, Stary Smokovec 2018,
Black would have done well to utilize the thematic rook lift:

14...Rb6!N The rook is well placed on b6, defending the d6-pawn and also unpinning it so that it can
advance to d5 at some point. A possible continuation is 15.f3 h6 16.Nh3 c4 17.Bf1 d5³ when Black
has plenty of space and attacking chances on the queenside.

450
13...Be6
13...c4? was Gelfand’s choice, but he had missed 14.Nd4! Bd7 15.Nf5 Bxf5 16.Bxf5± when White
was clearly better in Vidit – Gelfand, Prague 2019.
The text move is an improvement which was played by Fridman a few months after the above game.
 
14.Qf4 Qa5
The position is becoming rather sharp. White can meet the threat to the a2-pawn either by
defending with A1) 15.b3 or by going on the offensive with A2) 15.Rxe6!N.
 

A1) 15.b3

451
 Black has more than one playable continuation but my preference is:
 
15...c4!N
Breaking up White’s queenside.
 
The game continued:
15...h6
This should lead to a forced draw.
16.Bxh6 Nd5 17.Bh7† Kxh7 18.Rxd5 Bxd5 19.Bxg7 Kxg7 20.Qg4†
My guess is that Fridman forgot his home analysis when he played:
 

452
20...Kh7?
A losing move!
20...Kh8! draws the game. The difference is that after 21.Rxe7 Bxf3 22.Qxf3 Black has the
resource 22...Rbe8! (which is not possible with the king on h7, as shown below) and the game
ends in a perpetual after: 23.Qh5† Kg7 24.Qg5† Kh8=
21.Rxe7 Bxf3 22.Qxf3 Qd8
22...Rbe8 loses to 23.Rxe8 Rxe8 24.Qxf7† with the king on h7.
23.Qh5† Kg7 24.Qg5† Kh7 25.f4 d5 26.f5 Rb6 27.f6 Rxf6 28.Qxf6 Kg8 29.Qg5† Kh8 30.Qh6†
Kg8 31.Re3
1–0 Nisipeanu – Fridman, Magdeburg 2019.
 
16.Bf5 cxb3 17.axb3

17...b4!
Marching forward – there’s no time to relax!
 
18.c4
18.Bxe6 fxe6 19.c4 gives Black the additional option of 19...Nd5!? (19...Rb6 simply transposes to
the main line below) 20.Rxd5 exd5 21.Qg4÷ when the position is messy.
 
18...Rb6
Black continues to create threats against the white king.
 
19.Bxe6 fxe6
19...Ra6? is a mistake, as after 20.Qd4 fxe6 21.Rxe6± White is on top. It’s essential to keep the
white queen on f4 in order to maintain tactical opportunities involving the knight jumping from f6.
453
 
20.Rxe6

20...Nh5!
Now White has two acceptable continuations, both of which lead to equality with accurate play.
 
21.Qc1!?
Despite being a bit passive, this works out okay for White.
21.Qd4!? sees White keeping his queen centralized while counting on 21...Bxg5 22.c5! as the tactical
justification to recover the piece. Still, Black has no problems after:

454
22...Qxc5 (22...dxc5 is also fine, and 23.Qc4 Qb5 24.Qxb5 Rxb5 25.Nxg5 c4= reaches an equal
ending) 23.Qxc5 dxc5 24.Nxg5 h6 25.Rxb6 axb6 26.Ne4 Nf4= The endgame is balanced.
 
21...Ra6 22.Qb2 Bxg5
Here too, Black temporarily wins a piece but White immediately recovers it.

23.Rd5 Qb6 24.Rxg5

24.Nxg5 is playable, though after 24...Qxf2 White needs to play a series of ‘only moves’ to hold
the balance: 25.Qd4! Qxg2 26.Kb2! Nf6 27.Qd2! Qxd2 28.Rxd2 h6

29.Nf3 Ne4! 30.Rxe4 Rxf3= Black is a pawn up but White’s rooks are active and the d6-pawn is
weak, so White should not have too much trouble holding the draw.

455
 
24...Qxf2 25.Qc1
Defending against checks on the first rank. Now Black has an interesting resource:
 
25...Rxf3!? 26.gxf3 Qd4 27.c3 Qd3† 28.Kb2 bxc3† 29.Qxc3 Ra2† 30.Kxa2 Qxc3 31.Rxh5=
Black’s queen is a match for White’s rooks, and the endgame should be drawn.
 

A2) 15.Rxe6!N

White eliminates the dangerous bishop before it can create more threats against the white king.
This exchange sacrifice also weakens Black’s kingside.
 
15...fxe6 16.Qh4 g6!
16...h6 is also playable but I believe that the text move is better.
 
17.Qh3!
The benefits of sacrificing on e6 are clear: Black’s forces are not well coordinated and the squares
around his king are weak, so precise play is needed.
 
17...Rf7
Black cannot afford to waste time defending the pawn, so instead he should focus on arranging his
defensive set-up.
 
18.Qxe6
456
18...Qd8!
The queen falls back in defence of the king. Black intends ...Qd7 to remove the active queen from
e6, and now it is White who needs to find a strong yet subtle move to avoid becoming worse.
 
19.c4!
The idea is to take away the option of ...c4 from Black. This may not seem relevant now, but it’s
actually rather important!
The immediate 19.Re1?! prevents ...Qd7 but after 19...Rb6! White’s resources are not good enough:

20.Qh3 (20.Qe2? runs into 20...c4µ which explains why 19.c4! is more precise) 20...Bf8 21.Re6 d5µ
The rook defends successfully along the 6th rank.

457
 
19...b4!
19...Qd7 is met by 20.Qxd7 Nxd7 21.Bxe7 Rxe7 22.cxb5² with slightly better play for White.
 
20.Re1
Preventing ...Qd7.

20...Rb7!
Unlike the note to White’s 19th move above, 20...Rb6? doesn’t work here due to 21.h4 d5 22.Qe2
Ne4 23.Bc1!± when Black is unable to maintain his strong knight on e4 any longer.
 
The text move makes ...Qd7 possible again. A possible continuation is:
 
21.Qh3 Qd7 22.Qh4 Rb8 23.Bc1
The threat of Ng5 seems awkward, but Black holds the balance after:

458
23...Nh5! 24.Ng5 Bxg5 25.Qxg5 Re8 26.Rxe8† Qxe8 27.g4 Nf6 28.f4 Qe7=
White has sufficient compensation for the exchange but Black has no problems holding the
balance.
 

B) 11.Rhe1

11...Be6
Black has placed his pieces optimally and can now think about expanding with ...d6-d5.

459
 We will analyse three main options:B1) 12.a3 is slow but playable; B2) 12.Kb1 has been the most
popular choice; andB3) 12.Bg5 is the most energetic and challenging option.
 
12.c4?! should be met by the typical 12...b5! 13.cxb5 as in Pobor – Krstic, Niksic 2018, when
13...a6!N would have given Black excellent chances on the queenside.
 

B1) 12.a3

This slow move gives Black a few good options. My favourite is:
 
12...c4!N
Immediately seizing the initiative. White is still not worse, but he will have to be careful.
 
12...d5 is a sensible alternative which was played in Paravyan – Gelfand, Moscow 2018.
 
13.Be2
13.Bf1 is also met by 13...Ne4 and if 14.Qe2?! (White should settle for 14.Qd4 d5 15.Bxc4,
transposing to the main line below) 14...d5³ White’s pieces are strangled while Black enjoys a space
advantage.
 
13...Ne4 14.Qd4 d5
Black has several nasty ideas such as ...b5, ...Qa5 and ...Bf6 on the cards, so White should head for
complications with:
 

460
15.Bxc4! Bf6!
White will have to play accurately to avoid being worse.

16.Ne5!
16.Qd3?! Nxc3 17.bxc3 dxc4³ is good for Black.
 
16...Qc7 17.Bxd5
17.f4?? is not possible due to 17...Bxe5 18.fxe5 Qxc4 when Black keeps an extra piece.
17.Bf4!? is strongly met by: 17...Nxc3! 18.bxc3 (18.Qxc3 dxc4 19.Qg3 Qc5³ leaves Black with
two well-placed bishops and a space advantage on the queenside) 18...dxc4

461
19.Nd7! A strong resource, liquidating the position. 19...Bxd4 20.Bxc7 Bxf2 21.Nxf8 Now it is
Black who needs an accurate move: 21...Bg4!= The position remains balanced.
 
17...Bxd5 18.Qxd5 Qxe5

19.f3!
Surprisingly, the knight has nowhere to go and must sacrifice itself!
 
19...Nxc3 20.Qxe5 Bxe5 21.bxc3 Bxc3 22.Re2 Rfe8=
The arising endgame is equal.
 

B2) 12.Kb1

462
 This natural move was Carlsen’s choice against Caruana.
 
12...Qa5
White may respond to the threat with B21) 13.c4 or B22) 13.a3.
 

B21) 13.c4 Qxd2

 
After the exchange of queens Black is usually fine, and he will prepare for an eventual break in the
centre with ...d5.
 
14.Bxd2
White counts on his well-placed pieces and the potential target on d6. However, Black can solve
this problem by centralizing his rooks and executing the ...d5 break.
 
14.Nxd2 enabled Black to carry out the same general plan after: 14...Rfe8 15.f3 Rad8 16.Bf2 b6
Black takes his time before carrying out the pawn break.

463
17.Ne4 Kf8 18.a4 h6 19.b3 d5= Black had no problems in Rudenko – Onoprichuk, corr. 2017.
 
14...h6
14...Rfe8 is also possible and has been played in a few games.
 
15.h3!?
This is White’s latest attempt to squeeze something out of the position.
 
The high-profile game between Carlsen and Caruana continued with: 15.Nh4 Rfe8 16.Ng6 Ng4
17.Nxe7† Rxe7 18.Re2 Ne5

464
19.Bf4 Nxd3 20.Rxd3 Rd7 21.Rxd6 Rxd6 22.Bxd6 Rd8 23.Rd2 Bxc4 24.Kc1 b6 (24...Rc8!?N=
avoids losing a pawn but Caruana has correctly assessed the endgame as drawing) 25.Bf4 Rxd2
26.Kxd2 Despite White’s extra pawn, the opposite-coloured bishops enabled Black to hold the draw
with no real difficulty in Carlsen – Caruana, London (11) 2018.
 
15.b3 brought White success in one game but Black’s play can easily be improved, and he has a
choice of decent continuations: 15...Rad8!? (15...Rfe8N= is a simple alternative to prevent White’s
idea in the game) 16.Bf5 Bxf5 17.Rxe7 This occurred in Nepomniachtchi – Duda, Paris (rapid) 2019,
and now 17...Rfe8!?N 18.Rxb7 Re2 would have given Black plenty of counterplay.

15...Rfe8!
It is best to centralize the rooks before breaking in the centre.
 
15...d5 is playable but White may develop some pressure after: 16.cxd5! (16.Ne5 Rfd8 17.cxd5 Bxd5
18.g4 Kf8 19.g5 hxg5 20.Bxg5 Rd6 was equal in Praggnanandhaa – Vocaturo, Sitges 2018)
16...Nxd5 17.c4 Nf6 (I also checked 17...Nb4N 18.Bxb4 cxb4 19.Nd4 Bc5 20.Nxe6 fxe6 21.f3 Rf6
22.Be4 Rf7² when Black should be able to hold, but he is not yet fully equal) Now in Mosadeghpour
– Lorparizangeneh, Ahvaz 2019, White could have kept a slight edge with 18.Bc3N 18...Rfe8 19.b3
Rad8 20.Ne5, when Black still has to do some work to neutralize White’s active pieces.
 
16.g4!?
16.Nh4?! was strongly met by 16...Ng4! 17.hxg4 Bxh4 18.g3 Bf6 19.Be2 Bd4³ and it’s Black who
is slightly better here due to the better pawn structure, as in Sanal – Bindrich, St Petersburg 2018.
 
16.Bc3 was played in Nihal – Sadhwani, Internet (rapid) 2019. Although the immediate 16...d5 was
quite playable in the game, I would suggest the preparatory 16...Rad8!?N if only for the sake of

465
consistency: Black brings the last piece into play before breaking in the centre, with no problems
whatsoever.
The text move was played in Arakhamia-Grant – Ushenina, Antalya 2019. White is clearly aiming
for a g4-g5 break, which prompted Black to retreat her knight to d7 in the game. Instead I
recommend:

16...Rad8!N
Again this preparatory move is our choice. Now Black is poised for ...d5, so White should carry
out her own pawn break without delay.
 
17.g5 hxg5 18.Nxg5 Bd7 19.Rg1 Bc6 20.Bc3
20.h4 is simply met by 20...d5 21.cxd5 Bxd5 22.Bc3 Bc6= with the idea of ...Nd5 and Black is
fine.

466
20...d5
Finally!
 
21.cxd5 Nxd5=
Black has a good game.
 

B22) 13.a3

467
 My original analysis contained a brief note on this move – but after Caruana played it in 2019, it
became clear that more serious work was required.
 
13...c4 14.Be2 Ne4
My initial impression of this position was that Black was developing plenty of activity and had
good attacking chances. This is essentially true, but White has some interesting ways of fighting
back, as we will see.
 
15.Qc1
15.Qd4? is not the way to go, as after 15...d5µ White’s queen is too exposed.
 
15...d5
Black has a nice structure with plenty of space on the queenside, and his knight is strong. White
has an excellent outpost on d4 but he needs to find a way to unravel his pieces.
 
16.Bd4
Threatening Bxc4. The time has come to improve Black’s play with:

16...b5!N
A logical move, not only preventing White’s threat but also preparing a future break with ...b4.
 
16...Rac8 was played in Caruana – Duda, Paris (rapid) 2019. This move also seems playable, Black’s
eventual defeat being caused by subsequent errors rather than by any fault in his play up to now; but
still, I prefer the text move.
 
17.Nd2!

468
 I believe this to be the critical continuation, and it’s possible that Caruana prepared it. The knight
retreat looks modest and it seems to walk into a nasty pin, but there is more to it than meets the eye!
 
17.Ne5?! was Caruana’s choice in the above game but it does not work as well here due to 17...Qc7!
when White has some problems. Firstly, Black’s last move prepares the blunt threat of ...a5 followed
by ...b4 with a huge attack. Secondly, White’s active possibilities can be thwarted by ...Bg5. A
sample variation showing Black’s idea could be:
 

18.Bg4? (18.Bf3?? is even worse, as 18...Bg5 wins on the spot) 18...Bg5! 19.f4 Bxg4 20.Nxg4
Qxf4µ Black wins a pawn.
 
A more serious alternative is:
17.h4
After seeing the note above, the value of taking the g5-square under control is obvious.
However, Black now gets to show another nice feature of his 16th move.
17...Rab8!
It transpires that Black does not need the support of his a-pawn, and immediately threatens the
brutal sequence of 18...b4!! 19.axb4 Bxb4 20.cxb4 Rxb4–+ when White has no good defence
against ...Ra4.
18.Ne5
With the g5-square under control, this is White’s best way to distract Black from his attacking
plan on the queenside.
18.Ng5? may seem like a logical follow-up to White’s previous move, but Black can break
through with a splendid sacrifice: 18...Nxc3†!! (the aforementioned sequence beginning with
18...b4?? does not work here because White will take twice on b4 and then trade off the
attacking knight on e4, so Black does better to sacrifice it!) 19.Bxc3 (19.bxc3? is out of the

469
 question as after 19...Bxa3 20.Qe3 b4–+ Black easily crashes through on the queenside) 19...b4
20.axb4 (20.Bd4 bxa3 followed by ...axb2 is winning for Black) 20...Bxb4 21.Bxb4 Rxb4–+
The important point is that 22.Nxe6 is met by the killer 22...Ra4! and Black gets a winning
material advantage.

18...Qc7 19.Bf3
19.Bg4?! f5!³ just helps Black to stabilize his central position.
19...a5
19...f5 allows 20.g4„ with counter-chances on the kingside.
Instead Black remains focused on the enemy king. White would be in trouble, if it were not for
the following sacrifice:
20.Rxe4!
20.Bxe4? would be a mistake due to 20...dxe4 21.Rxe4 b4 22.cxb4 axb4 23.axb4 Rxb4µ when
Black’s attack is too strong.
20...dxe4 21.Bxe4

470
 White’s bishop on e4 is more valuable than a rook on the same square, as the two bishops on e4
and d4 work in tandem to create threats against the black king. White has compensation for the
exchange and, with precise play, he can neutralize Black’s attack on the queenside. A sample
variation is:
21...b4 22.axb4 axb4 23.Nc6 bxc3 24.Bxc3 Rb5
All logical moves up till now.
25.Qe3!
White’s queen gets ready to work with the bishop pair.
25...Bc5 26.Qf3 Re8

Now White can and should force a draw with:

27.Bxh7†! Kxh7 28.Qh5† Kg8 29.Bxg7 Kxg7 30.Qg5† Kf8 31.Qh6† Kg8 32.Qg5†=

471
17...Bg5
This is the logical response, pinning the knight.
 
18.Bf3!
This exchange sacrifice is the big idea behind White’s previous move. Now Black is at a crossroads
and must decide whether or not to grab the material. Both options lead to some sort of dynamic
equilibrium and the choice comes down to personal taste.
 
18...f5!?
This way Black keeps the material equal and avoids giving White a kingside initiative.
 
Black can also accept the sacrifice with:
18...Bxd2 19.Rxd2 Nxd2† 20.Qxd2

472
 The ideally-placed bishop on d4 is no less valuable than either black rook. Even though the
engines tend to prefer Black, I believe that White has fair compensation, especially in a
practical game. A sample variation could be:
20...Qd8
It is time to bring the queen back to the defence.
21.Qf4 h6
Intending ...Qg5.
22.h4 Re8 23.Bh5!
Probing at the kingside and threatening to take on e6.
23...Kh7!
Black keeps things solid.
 

473
24.Rxe6!?
A double exchange sacrifice!
24.Re3!? is an interesting option which maintains the tension, although Black is not worse by
any means.
24...Rxe6
24...fxe6? loses on the spot to: 25.Bg6†!! Kxg6 (25...Kh8 26.Qxh6†+–) 26.Qg4†+–
25.Qf5†
25.Qxf7 Qg8 26.Qb7 Rf8 27.Qxa7µ reaches a position where the engine heavily prefers Black,
although converting his advantage would not be easy in a practical game.
25...Rg6 26.Bxg6† fxg6 27.Qf7 Qg8 28.Qd7
White’s bishop is in full flare, tying down the black queen.

474
28...h5!
Preventing any kingside advances from White.
29.Qxb5 Qf7=
White has won back a pawn but allows Black to improve his queen. Material is limited and Black’s
weaknesses are well defended; so although White has fair compensation for the exchange, Black is
not worse at all.
 
Returning to the main line, Black reinforces his strong knight and he would stand clearly better, if it
were not for the following devious trick:

19.Bxe4 fxe4 20.Nxc4!

White wins a pawn, but only temporarily.
 
20...Qd8! 21.Be3!
White should take the opportunity to trade off one half of Black’s bishop pair.
 
21...Bxe3 22.Nxe3 Rxf2

475
 The tactical sequence has ended and it is time to take stock of the situation. Black has a weakness
on d5 which will need protection, and White has the theoretically superior minor piece. On the other
hand, Black has the better pawn structure, including a protected passer on e4 which will always give
him chances for counterplay, even through it is currently blocked by the strong knight on e3. The
rook on f2 is also extremely active. A sample variation could be:
 
23.Rd2 Qf8 24.Qd1 Rd8=
With balanced play.
 

B3) 12.Bg5

476
 White goes for the typical idea of Qf4-h4, transferring the queen to the kingside.
 
12...d5
Black continues with his normal plan.
 
12...Bxa2?? just loses to 13.b3 when Black does not have the resources needed to save his trapped
bishop or obtain compensation for it.
 
12...h6? was an instructive mistake made by Harikrishna. One should always be careful about
playing this move when the rook is still on f8. 13.Bxh6! c4 (13...gxh6? 14.Qxh6+– gives White a
crushing attack) 14.Bxg7! White won a nice attacking game in Karjakin – Harikrishna, Shamkir
2016.
 
12...Qa5?! is inaccurate due to 13.a3 Rfe8 14.Qf4! (14.Bxf6 was less convincing in Giri – Motylev,
Tbilisi 2017) when White is better. 14...h6? only makes things worse after 15.Rxe6! fxe6 and now in
Kantor – Bernabeu Lopez, Roquetas de Mar 2019, 16.Bxh6!N would have been crushing for White.
 
13.Qf4
All of White’s pieces are in play and he intends Qh4 with mating threats.

477
13...Re8!
An important defensive move, freeing the f8-square for the bishop.
 
13...Qa5? is refuted by the simple 14.Qh4 h6 15.Bxh6+– when Black cannot defend his kingside.
 
13...Qb6?! is met by the strong 14.Re5!± with threats including Bxf6 followed by Rh5.
 
14.Re5!?
This move sharpens the play: White hopes to swing the rook into the attack along the 5th rank, or
possibly to double along the e-file, but at the same time the rook is rather exposed.
 
14.Bb5 should be met by 14...Rf8 when it may seem as if White has gained time but his bishop is
misplaced and exposed to ideas involving ...Qa5 or ...Qb6, and he found nothing better than 15.Bd3
leading to a short draw in Karjakin – So, Riga 2019.
 
14.Kb1 is not so effective now. My suggestion is: 14...Qb6!?N (14...h6 was also fine in Harika –
Kosteniuk, Khanty-Mansiysk 2017)

478
15.Re5 (15.Qh4? does not work here due to 15...h6 16.Bxh6 gxh6 17.Qxh6 and now 17...Ne4!
18.Bxe4 dxe4 19.Rxe4 Bxa2†!–+ shows one of the hidden points behind putting the queen on b6)
15...c4 16.Bf5 Bxf5 17.Qxf5 g6 18.Qh3 Ba3 19.Bc1 h5³ Black has slightly better chances due to his
pressure on the queenside and control of important squares such as g4 and e4.
 
14.Qh4 is safely met by 14...h6 15.Bxh6 when I suggest: 15...gxh6!?N (15...Ne4 16.Qh5 Nf6= led to
a draw in Sneddon – Bysina, corr. 2018, but the text move is simpler and reduces White’s options)

16.Qg3†16.Qxh6? doesn’t work due to 16...Bf8 17.Qg5† Bg7–+ which shows the value of the rook
on e8!) 16...Kf8 The game ends in a repetition after: 17.Qf4 Kg7 18.Rxe6! (17.Qf4 gives Black the
extra option of avoiding the immediate draw with 17...Bg4!? if he feels like it) 18...fxe6 19.Qg3†
Kf8 20.Qf4=
479
 
14.c4!?
This is an interesting try, undermining the e4-square.
14...Qa5!
14...d4?! is an instructive mistake: Black no longer has the e4-square available for the knight,
so 15.Re5! gives White an improved version of the main line, due to the following point:
15...Qc7? (the lesser evil is 15...Bd6 16.Bxf6 Qxf6 17.Qxf6 gxf6 18.Rh5 h6 19.h3² with a
slightly better endgame for White) 16.Qh4 h6 17.Bxh6!+– Unlike the main line below, Black
does not have the ...Ne4 move available, so he is simply busted.

15.cxd5
White can win a pawn with 15.Bxf6N 15...Bxf6 16.cxd5 Bxd5 17.Rxe8† Rxe8 18.Bxh7† Kxh7
19.Qf5† Kg8 20.Qxd5 but, after 20...Qa4 intending ...Rd8, Black’s active pieces fully
compensate for the small material deficit.
15...Nxd5 16.Qh4 h6 17.Bxe7 Rxe7 18.Bc4 Rd7 19.Ne5 Rd6 20.f4 Rad8=
Black was doing fine in Yloenen – Gerzina, corr. 2014.

480
14...Qc7!
An important move to remember! Now Black threatens ...Ng4.
 
14...Bd6?! doesn’t work due to 15.Bxf6 Qxf6 16.Bxh7†! Kxh7 17.Rh5† Kg8 18.Qxd6² when White
picks up a pawn for questionable compensation.
 
15.Qh4!?
This leads to messy positions where both sides will have to play accurately to avoid becoming
worse.
 
15.Bf5 is comfortably met by 15...Nh5! (but not 15...Bd6? 16.Bxf6 gxf6 17.Qg4† Kf8 18.Ree1+–)
16.Qh4 Bxg5† 17.Qxg5 Nf6 18.Bxe6 Rxe6= and Black is doing fine.
 
15...h6 16.Bxh6 Ne4!
An important resource!

481
17.Qf4!
17.Bg5? allows a nice tactic: 17...Qxe5! 18.Bxe7 Qc7µ
 
17...Nxf2 18.Qg3
18.Rde1?! is met by 18...Nxd3† 19.cxd3 Bf8 20.Bg5 Bd7³ and Black is slightly better due to the
bishop pair.
 
18.Bxg7!? Kxg7 is playable for White as long as he follows up with:

19.Kb1!! A very cool move, threatening Rg5†. 19...Nxd1 20.Rg5† Bxg5 21.Qxc7 Bh6÷ Materially
Black has more than enough for the queen, but the position remains dynamically balanced.

482
 
18...Ng4 19.Bf4 Bd6 20.Bh7†!
20.Rde1? is inadequate due to 20...Nxe5 21.Bxe5 Bxe5 22.Rxe5 Qe7µ when Black has parried the
attack successfully.
 
“This and the next move are very difficult for White to find over the board!” were my words when I
first analysed this variation while writing this book. But just a few days later, it was played by
Navara against Gelfand!

20...Kh8!N
20...Kf8? was the wrong choice in Navara – Gelfand, Prague 2019.
 
21.Re4!
This is White’s only decent move. The crazy but more or less forced sequence continues with:
 
21...dxe4 22.Bxd6 Qc8 23.Qh4
23.Bxe4?! is strongly met by 23...Bf5! when the exchange of light-squared bishops kills White’s
chances for a successful attack: 24.Bxf5 Qxf5 25.Bxc5 Qxc5 26.Qxg4 Kg8³ White has two pawns
for the exchange but the doubled c-pawns are not the best, so White is on the defensive.
 

483
23...exf3
This is the safest continuation, leading to a forced draw.
 
23...Nh6!? is another possibility, and after 24.Bxe4 Bf5 25.Be5 Kg8 26.Bd5 Bg6 27.Rd2÷ the
position remains complex.
 
24.Be4† Nh6 25.Be5 Kg8 26.Bxg7 Kxg7 27.Qg5† Kf8 28.Qxh6† Ke7 29.Qg5† Kf8=
The sharp line ends in a draw.
 

Conclusion
 
This chapter has concluded our coverage of the highly topical 5.Nc3 system. We focused on two
important options for White on move 11, the first being 11.Kb1, when 11...b5! is a good reply.
Taking the pawn is too risky for White, so 12.Rhe1 Rb8 13.Bg5!? Be6 14.Qf4 Qa5 is the main line.
Then 15.b3 is a reasonable move but 15...c4!N gives Black plenty of counterplay. The exchange
sacrifice 15.Rxe6! fxe6 16.Qh4 is trickier, although 16...g6! gives Black fair chances.
 
We then moved on to the natural developing move 11.Rhe1, which can be considered the ultimate
main line of the 5.Nc3 system. 11...Be6 is our reply, leading to a further split. 12.a3 is the least
dangerous move, when 12...c4!N gives Black a lot of activity. We then moved on to 12.Kb1 Qa5,
when 13.c4 leads to a queen trade followed by relatively few problems for Black, while 13.a3 is a
recent innovation by Caruana which might attract more followers, but you will have little to fear
when armed by my analysis. Finally, 12.Bg5 is a critical option, aiming for a kingside attack based
on piece activity. There are many tactical and sacrificial possibilities, and the cost of a mistake is

484
likely to be high on either side. However, my analysis once again shows that Black’s position is
completely sound, and with accurate play we should have nothing to fear.

485
;
A) 3.c3 dxc3 241
A1) 4.Nxc3 241
A2) 4.Bc4 243
B) 3.Nf3 Bb4† 246
B1) 4.Bd2 Bxd2† 246
B11) 5.Qxd2 246
B12) 5.Nbxd2 248
B2) 4.c3 dxc3 249
B21) 5.bxc3?! 249
B22) 5.Nxc3 250
C) 3.Qxd4 Nc6 252
C1) 4.Qc4 252
C2) 4.Qe3 Nf6 254
C21) 5.e5?! 255
C22) 5.Bd2 255

486
 C23) 5.Nc3 Bb4 6.Bd2 0-0 7.0-0-0 Re8 257
C231) 8.Qf4!? 257
C232) 8.Bc4 258
C233) 8.Qg3 Nxe4 9.Nxe4 Rxe4 260
C2331) 10.c3 260
C2332) 10.Bf4 262

 
1.e4 e5 2.d4 exd4
Following his early central break, White has three main options: A) 3.c3, B) 3.Nf3 and C) 3.Qxd4.
 

A) 3.c3
 
White offers a gambit to obtain a lead in development.
 
3...dxc3
We will analyse A1) 4.Nxc3 and A2) 4.Bc4.
 

A1) 4.Nxc3 Nc6

Black develops solidly and takes the e5-square under control, giving himself the option of ...Nf6
without having to worry about e4-e5 in response.

487
 
5.Nf3
White makes the obvious developing move, which also happens to regain control over e5, thus
making it harder for Black to arrange ...Nf6.
 
5.Bc4 allows 5...Nf6, and after 6.Nf3 Bd6!? Black prevents e4-e5 and prepares to castle before
rerouting his bishop to a better square. For example: 7.0-0 0-0 8.Bg5 h6 9.Bh4 Be7 Breaking the pin
and preparing to develop the queenside. 10.Re1 d6³ White did not have enough compensation in
Nogler – Richter, email 2013.
 
5...Bc5
Before playing ...d6, I prefer to develop the bishop actively.
 
6.Bc4 d6 7.Bg5
7.Ng5?! was safely met by 7...Ne5 8.Bb3 h6µ in Fusco – Monzon, Villa Martelli 2000.

7...Nge7
This time we should develop the knight to e7 rather than f6, to avoid stepping into a pin.
 
8.0-0
8.Nd5 aims to put direct pressure on the knight. After 8...h6 9.Bh4 Black should break the pin
immediately with:

488
9...g5! 10.Bg3 (10.Nxg5? occurred in Turati – Macchiagodena, Italy 1992, when 10...Ng6!N would
have led to decisive material gains, for instance: 11.Ne6 Qxh4 12.Nxc5 dxc5 13.Nxc7† Kf8
14.Qd6† Kg7 15.Nxa8 White is briefly ahead in material, but after 15...Qxe4† Black is winning.)
10...Nxd5 11.exd5

11...Qe7† This active move is obviously stronger than retreating the knight. 12.Kf1 Ne5µ White was
far away from achieving any compensation for the pawn in Fabig – Herfurth, corr. 2017.

489
8...0-0
Both sides have completed development. The only way that White can hope for compensation is to
put pressure on the e7-knight with:
 
9.Nd5 Be6
Weakening the kingside with ...h6 and ...g5 would be riskier when Black has already castled, so I
prefer the text move.
 
10.b4!?
10.Nf6†? proved to be a useless waste of time after 10...Kh8µ in Puechert – Kahl, email 2001; I
doubt that White has anything better than immediately returning the knight to d5.
 
The simple 10.Rc1 may be the best that White can do, hoping for some pressure along the c-file
rather than a direct attack. Play may continue: 10...Qd7 11.Bb5 Pinning the other knight this time.

490
11...f6 12.Bf4 Bb6 Black kept a solid position in Versavel – Vertongen, Belgium 2001, although
White had a certain amount of compensation due to his active pieces.
 
The text move is an energetic try: White sacrifices a second pawn in the hope of increasing his
initiative. Fortunately we can follow a correspondence game where Black gained the upper hand.
 
10...Bxd5 11.exd5 Nxb4
 

12.Re1 f6 13.Be3 Bb6

491
 Given enough time, Black will consolidate, so White should force a position with two bishops
against two knights with:
 
14.Nd4 Bxd4 15.Bxd4 Kh8³
We have been following Fabig – Kuhl, email 2014. White’s bishops are strong and there are some
weak squares in Black’s camp. Nevertheless, I do not rate his compensation as sufficient for the two-
pawn deficit, so I prefer Black’s chances.
 

A2) 4.Bc4

This is known as the Danish Gambit.

 
4...cxb2 5.Bxb2
White has sacrificed two pawns for a significant lead in development and attacking chances, but
Black has a well-known route to safety.
 
5...d5!
This is the most popular move, and not without reason. Black gives back the material to reach an
endgame with at least equal chances.
 
6.Bxd5
White can avoid the endgame with:
6.exd5?!

492
 The problem with this option is that White closes the diagonal of his light-squared bishop – one
of his main trumps in the Danish Gambit.
6...Nf6 7.Nf3
7.Ne2 doesn’t seem challenging at all, for instance: 7...Bd6 8.h3 Preventing any possibility of a
sacrifice on h2. 8...0-0 9.0-0 Nbd7µ Followed by ...Nb6 when White had no real compensation
for the pawn in Sanchez Fleire – Abatantuono, Caracas 2015.
7...Bb4† 8.Nbd2
8.Nc3 is met in a similar fashion with 8...Qe7†, forcing White’s bishop into a passive retreat.
9.Be2 0-0 10.0-0 Now in I. Horvath – Novaky, Hungary 2012, the strongest would have been
10...Nbd7Nµ followed by ...Nb6 when White has little to show for the missing pawn.
8...Qe7†!
It is worth including a disruptive check before castling.
9.Ne5!?
White blocks the check in an unconventional way. Black’s simplest reply is:
 

9...Nbd7N
9...Ne4?! looks tempting but after 10.0-0 Nxd2 11.d6!÷ the position became messy in Kobus –
Lukas, Erle 2000.
10.0-0 Nxe5 11.Bxe5 0-0 12.Re1 Bf5µ
Black keeps everything under control while remaining a clear pawn up.

493
6...Nf6
Black ‘blunders’ his queen but regains it immediately.
 
7.Bxf7†
7.Nc3?! Nxd5 8.Nxd5 Nd7! followed by ...c6 leaves Black a clear pawn up, with the bishop pair
as a bonus (8...c6? 9.Nf6†! is the one big trick that Black must avoid).
 
7...Kxf7 8.Qxd8 Bb4† 9.Qd2 Bxd2† 10.Nxd2
After a few forced moves, we reach an endgame with equal material and opposite-coloured
bishops. However, with many other pieces on the board and each side having a ‘plus two’ pawn
majority on one flank, we are a long way from a draw. Objectively the position is equal, but Black
has scored well in practice and you can see from the following analysis that his play is quite easy to
understand.

494
10...Re8
Black should obviously bring his rook to the e-file, not only attacking the e4-pawn but also
enabling the king to retreat to g8 without any loss of harmony, if the need arises.
 
11.Ngf3
This tactical defence of the e-pawn seems best.
 
Black has excellent chances after:
11.f3 Nc6
The knight can head for d3, via b4 or e5.
12.Ne2
12.a3 prevents ...Nb4, so Black goes the other way. 12...Ne5 13.Bxe5 Rxe5 14.Ne2 Now in
Cetynkaya – Dinc, Manavgat 2016, the simple 14...b6!N would have been strongest, intending
to develop the bishop to a6, with a promising endgame for Black.
12.Rc1 pins the c6-knight to the c7-pawn, but 12...Nd5! activates the other knight. A good
example continued: 13.a3?! (White should settle for 13.Ne2N 13...Ndb4³ when Black wins a
pawn, although White will be able to claim a bit of compensation) 13...Nf4 White is yet to
develop his kingside pieces and the knight is already a source of troubles for him. 14.Rc3 Rd8!
Increasing the pressure. 15.Ba1 Nxg2†–+ Black soon won in N. Ivanov – Johnson, USA 2000.

495
12...Nb4 13.0-0 Nd3
The knight is beautifully placed on d3, covering a lot of important squares.
14.Ba3 b6³
With a strong knight on d3 and ...Ba6 coming next, Black had the better game in Bykov – Erofeev,
St Petersburg 2010.

11...Kg8
The king steps away from any checks.
11...Nxe4? loses material to 12.Ne5†± as in Lasley – Leber, corr. 2012.
 
12.Ne5

496
 12.0-0N can be met by 12...Na6! with dual ideas of ...Nb4 and ...Nc5 (but not 12...Nxe4? 13.Nxe4
Rxe4 14.Rfe1ƒ). Play may continue: 13.Ne5 Nd7 14.Nxd7 Bxd7 15.f3 Be6=
 
12...Nbd7 13.f4?
We are following a game in which White stabilized the knight on e5 in this ambitious way, but ran
into problems because the e4-pawn can no longer be supported by f2-f3.
 
White should settle for 13.Nxd7N 13...Bxd7 14.f3 with equal chances.

13...Nc5!
Immediately targeting the weak central pawn.
 
14.Rc1 b6 15.Ba3 Nfxe4 16.Nxe4 Nxe4 17.Rxc7
White has managed to maintain material equality, but that is not the end of the story.
 
17...Ba6!ƒ
White’s king in the centre became a burden in Martinez Romero – Yuan, Kemer 2009.
 

B) 3.Nf3

497
 Sometimes White uses this move order in an attempt to lure Petroff players out of their preferred
opening. However, Black can avoid an unfavourable transposition with:
 
3...Bb4†
Black is ready to accept the pawn sacrifice.
 
3...Nc6 transposes to 2...Nc6 3.d4 exd4, when 4.Nxd4 is the Scotch Game and 4.Bc4 the Scotch
Gambit – perfectly playable for Black, but hardly a logical choice for a Petroff player.
 
B1) 4.Bd2 is worth considering, but B2) 4.c3 is the most critical choice.
 

B1) 4.Bd2 Bxd2†

 
I think B11) 5.Qxd2 is a slightly worse way of recapturing than B12) 5.Nbxd2.
 

B11) 5.Qxd2 c5!

 
Black makes it difficult for White to regain the missing pawn.
 
6.c3
White can also try:
6.Bd3!?N

498
 First protecting the e4-pawn in order to take the impact out of ...Nf6, before breaking with c2-
c3.
6...Nc6 7.c3 d5!
This typical freeing central break gives Black an easy game, and White still has to work to
regain his pawn.
7...dxc3 8.Nxc3© gives White decent compensation.
8.exd5
8.cxd4 is met by 8...dxe4 9.Bxe4 Nxd4 10.Nxd4 cxd4 when White still has to work to regain
the d4-pawn. My line continues: 11.0-0 Nf6 12.Bf3 0-0 13.Rd1 Re8! It is not so easy for White
to win his pawn back due to the back-rank issues. Instead he should play 14.Nc3!, exploiting
the pin on the d4-pawn in order to develop the knight to its optimal square. 14...Bf5 15.Qf4
Qb6! 16.Qxf5 dxc3 17.bxc3 Qc7 The position is essentially equal although Black may have a
slight practical edge due to his better pawn structure on the queenside.
8...Qxd5 9.cxd4 cxd4
9...Nxd4!? is ambitious, although 10.Nxd4 Qxd4 11.0-0 Nf6 12.Bb5† Ke7 13.Qg5© leaves
White with obvious compensation due to Black’s weak king position.

10.0-0 Nge7 11.Na3 0-0 12.Nb5 Qd8

Finally White regains the d4-pawn with:
13.Nbxd4 Nxd4 14.Nxd4 Ng6=
Black is doing fine.

499
6...Nf6!
An important move to remember! While White is busy regaining the d4-pawn, Black develops his
kingside and attacks e4.
 
6...dxc3 7.Nxc3© gave White good compen-sation in Neuberger – Abram, Vienna 2003.
 
7.e5
I also considered:
7.Bd3 d5!
Black is consistent in getting his knight on e4.
8.exd5
8.e5?! Ne4 9.Qf4 dxc3 10.Nxc3 Nxc3 11.bxc3 Qe7µ is clearly better for Black.

500
8...Qxd5!N
A natural improvement over 8...0-0, played in D. Fischer – Langheld, email 2010.
9.Bb5†
9.cxd4 cxd4 10.0-0 0-0 11.Nxd4 is met by 11...Rd8 12.Rd1 Qxd4 13.Bxh7† Kxh7 14.Qxd4
Rxd4 15.Rxd4 Nc6³ when Black has two pieces for the rook, and better chances in the
endgame.
9...Nc6 10.cxd4 cxd4 11.Bxc6† bxc6 12.Qxd4 Ba6!³
The white king will face some problems.
 
7...Ne4 8.Qf4 d5 9.exd6
Now it is important not to retreat the knight immediately from e4, but maintain it on its active post
with:

501
9...0-0! 10.cxd4 cxd4 11.Be2 Re8 12.Nxd4?
We are following a game in which the text move was accurately punished by Black.
 
12.0-0N is better, but the simple 12...Qxd6 13.Qxd6 Nxd6 14.Nxd4 Nc6! 15.Nxc6 Rxe2³ reaches an
endgame where Black is pressing.
 
12...Qb6 13.Nb3
And now comes the strong:
 
13...Bg4! 14.Qxg4 Qxf2† 15.Kd1 Qf6µ
Black has a strong initiative due the poor position of the white king. In the game White simply
blundered with:
 
16.Rf1?? Qxf1† 17.Bxf1 Nf2† 18.Kd2 Nxg4–+
Tokhirjanova – Ni Shiqun, Sharjah 2014.
 

B12) 5.Nbxd2

502
 As mentioned earlier, I think this is a better option for White than the previous variation, but Black
still has a choice of decent continuations.
 
5...Nc6
This move offers a simple path to a comfortable position.
 
5...c5!?
This is the more complex and ambitious way to continue. White has to play accurately to
maintain the balance.
6.c3!
White should break immediately before Black consolidates.
6.Bc4 has been played a few times but I don’t think it is critical at all, as Black simply gets
good play with an extra pawn. 6...d6 7.0-0 Nc6 8.Qe2 Nf6µ followed by ...0-0 and Black is
clearly better, as in Daniel – Dettmann, email 2001.
6...dxc3 7.bxc3 Ne7 8.Nc4!
8.e5?! is well met by 8...d5! before White can fully clamp down on the d6-square. 9.exd6 Qxd6
10.Ne4 Qxd1† 11.Rxd1 0-0³ Even if White wins the c5-pawn, his pawn structure on the
queenside will remain weak, giving Black an edge.
8...d5 9.Ne3 Be6
9...dxe4 leads to an unclear ending after 10.Qxd8† Kxd8 11.Ng5 Rf8 12.Nxh7 Re8 13.Ng5 f5
14.0-0-0† Kc7÷ when Black has good control in the centre, but is behind in queenside
development.
10.exd5 Nxd5 11.Qa4†!
An important move, preparing long castling in order to exert pressure along the d-file.
11...Nc6 12.0-0-0 0-0÷

503
 No matter how scary the pin on the d-file looks, Black is doing fine due to tactical resources such
as ...b5.
 
6.Nb3 Nf6 7.Bd3 0-0 8.0-0
Both sides have castled comfortably and White is ready to pick up the d4-pawn, restoring material
equality.
 
8...Re8 9.Re1 d6 10.Nfxd4
Now in Savchenko – Chadaev, Olginka 2011, the simplest option would have been:

10...Nxd4N 11.Nxd4 Bd7=

Black has a comfortable game.
 

B2) 4.c3

504
 White sacrifices a pawn with the aim to get compensation by developing quickly.
 
4...dxc3
We will consider both of the recaptures: B21) 5.bxc3?! and B22) 5.Nxc3.
 

B21) 5.bxc3?!
 
This move is clearly inferior, as the knight on b1 remains undeveloped.
 
5...Bc5
A logical retreat: the bishop is well placed and active on c5.
 
6.Bc4 d6
Black delays the development of the g8-knight so that Ng5 can be met by ...Nh6.
 
Preparing to castle with 6...Ne7?! is worse in view of 7.Ng5 d5 (7...0-0?? 8.Qh5+–) 8.exd5 0-0 9.0-
0= as seen in Krakops – Rytshagov, Mezezers 2000. White is no longer a pawn down and he can
hope to create some play on the kingside.
 
7.0-0
7.Ng5?! is no good here due to 7...Nh6 8.0-0 0-0µ when Black is a stable pawn up and can play
...Ng4 followed by kicking the knight from g5 with ...h6.
 

505
7.Qb3?! has been played in a few games but 7...Qe7 8.0-0 Nf6µ left Black completely in control and
a clean pawn up in David Bordier – Walther, corr. 2013.
 
7...Nc6 8.Nbd2
8.Qb3 is, just like on the previous move, met by 8...Qe7 9.Bg5 Nf6 10.Nbd2 0-0µ when White has
no compensation for the pawn.
 
8.Bg5 Nf6 9.e5 was tried in Jeknic – Babic, Cetinje 2010, when the simplest would have been
9...dxe5N 10.Qxd8† Nxd8 11.Nxe5 Be6µ and Black is simply a pawn up in the endgame.
 
8...Nf6
This is the right time to develop the knight to f6, as Black is ready to castle and has avoided any
e4-e5 tricks.
 
9.Nb3 Bb6 10.a4 a6 11.Bg5 h6 12.Bh4
Now in Sammour Hasbun – Ippolito, Internet 2007, Black could have simply played:

12...0-0Nµ
Followed by removing the pin on the f6-knight, either by ...g5 or ...Ne7-g6. Black has a safe extra
pawn and is doing great.

B22) 5.Nxc3

506
5...Bxc3†!
Black should take the opportunity to damage White’s pawn structure before White gets a chance to
support his knight.
5...Ne7 is well met by 6.Qd4! Bxc3† 7.Qxc3 when, with White’s pawn structure on the queenside
intact, he could hope for some compensation due to his bishop pair in Zvjaginsev – Beliavsky,
Portoroz 1999.
 
6.bxc3 Ne7
This is the safest square for the knight to develop, as on f6 it could be hit by e4-e5.
 
7.Bc4
This is White’s usual choice, developing the bishop to its most active square and preparing to
castle.
 
7.Ba3 d6 8.e5!? can be met by 8...0-0, when 9.exd6 cxd6 10.Qxd6 Qxd6 11.Bxd6 transposes to the
7.e5 line below.
 
7.e5 d5
7...0-0 8.Ba3 d5 leads to the same thing.
8.exd6 Qxd6 9.Qxd6 cxd6 10.Ba3
Now it’s important not to waste time saving the d6-pawn but instead focus on quick
development with:

507
10...0-0!
10...d5?! was riskier in Zezulkin – Karpov, Bastia 2002.
11.Bxd6 Re8 12.0-0-0 Nbc6
Black is completely fine in this endgame, as White’s weak pawn structure on the queenside
limits the potential of his bishop pair. The game continued:
13.Bc4
The logical square for the bishop, so that it can put pressure on f7.
13...Nf5 14.Ng5 Nxd6 15.Rxd6
Now in Zezulkin – Przewoznik, Ustron 2008, it would have been simple and good to play:

15...Re7!N 16.Rhd1 Bf5 17.h3 Rae8=

Black will soon drive the knight back from g5, with a good position.

508
 
7...0-0 8.0-0
A reasonable alternative is:
8.e5!? d5 9.exd6 Qxd6
Black might also consider 9...cxd6!?÷ when he has an extra pawn, but White’s bishop pair and
lead in development compensate for it.
10.Qxd6 cxd6
Black cannot possibly be worse, and the onus is on White to regain his pawn.
11.Ba3 Re8
11...Nbc6 12.Bxd6 Re8 13.0-0-0 transposes to the previous note.
12.0-0-0 d5!?
12...Nbc6 13.Bxd6 would transpose again.

13.Rhe1 Be6 14.Bb3 Nbc6 15.Ng5 Bf5

Now White has to trade off his dark-squared bishop in order to regain the pawn:
16.Bxe7 Nxe7 17.Ba4 Nc6 18.Rxe8† Rxe8 19.Rxd5 g6=
Black had no problems in Hector – Ax. Smith, Ronneby 2018.

509
8...Nbc6
8...d6 9.e5 Nbc6 would lead to the same thing.
 
9.Bg5!?
White struggles to prove his compensation after other moves. For example:
 
9.e5 d5! is the typical central break which frees Black’s position. 10.exd6 Qxd6 11.Qxd6 cxd6
12.Ba3 d5 13.Bb5 Rd8³ White’s bishop pair was not enough to compensate for the deficit of a pawn
in Velimirovic – Motylev, Herceg Novi 2000.
 
Trying to go on the offensive immediately with 9.Ng5?! is not a good idea. 9...h6 10.Nh3 occurred in
Deak – Jamrich, Hungary 1995, when the thematic 10...d5!N would have been even stronger than
normal, as Black also threatens to exchange on h3, ruining White’s pawn structure. For instance:

510
11.Bxd5 (11.exd5? is even worse in view of 11...Na5 12.Bb3 Bxh3 13.gxh3 Nxb3 14.axb3 Nxd5µ
when White is simply a pawn down with a weak kingside to boot) 11...Bxh3 12.gxh3 Nxd5 13.exd5
Ne5³ Material is level but White’s pawn structure on both the queenside and kingside is ruined.
 
The text move seems like White’s best attempt to obtain active compensation for the pawn. He
refuses to play e4-e5 which, as we have seen, enables Black to advantageously simplify the position
with the ...d5 break.
 
9...d6 10.Nd4

10...Kh8

511
 A simple and strong move: Black gets ready to play ...f6, unpinning the e7-knight.
 
11.Rb1
We are following a game in which Black went for simplifications with:
 
11...Nxd4
11...f6!?N is a more ambitious alternative, when White still has to prove his compensation.
 
12.cxd4
12.Qxd4?! f6 13.Bh4 Nc6³ leaves White with insufficient compensation for the pawn.
 
12...f6 13.Bf4 d5!
The typical central break gave Black full equality in Vitoux – Kunin, Barcelona 2009.
 

C) 3.Qxd4 Nc6

This is the Centre Game, where White develops his queen early and loses some time, but hopes to
make up for it by castling on the queenside and launching some kind of attack.
 
We will consider C1) 4.Qc4 followed by the main line of C2) 4.Qe3.
4.Qa4?! misplaces the queen. A good example from correspondence praxis continued: 4...Bc5
(4...Nf6 is fine too) 5.Nf3 Nge7 6.Be2 d6 7.Nc3 0-0 8.0-0 As you can see, the queen on a4 is doing
absolutely nothing. 8...a6 9.Qc4 Bg4³ Black’s pieces are harmoniously placed, while White will have
to waste additional time finding an optimal square for his queen, Mulde – Landes, corr. 2017.

512
 

C1) 4.Qc4
 
This was recently played by Wei Yi against Karjakin. Black can continue in the same way as in
variation C23.
 
4...Nf6 5.Nc3
5.Bg5?! is not so good, as after 5...h6 White should retreat his bishop along the c1-h6 diagonal,
which will amount to a waste of time. If White decides to maintain the pin with 6.Bh4? then he is
brutally punished by:

6...g5! 7.Bg3 d5! 8.exd5 Nxd5–+ Black is ahead in development and White’s queen is misplaced, so
Black will build a decisive initiative.
 
5...Bb4 6.Bd2 0-0 7.0-0-0 d6 8.f3 Re8
The rook is well placed on e8. The white queen is misplaced on c4 and has little option but to
retreat to e2, after which Black can always think of breaking with ...d5 at an appropriate moment.
 
9.Qe2
9.a3 should be met by 9...a5! when White has nothing better than transposing to our main line with
10.Qe2.

513
9...a5!
9...Be6 was played in the high-level game Wei Yi – Karjakin, Jerusalem 2019, but the text move is
more accurate.
9...d5 is playable but the timing of this pawn break is not ideal: Black also has his queen sitting
opposite the enemy rook, and 10.a3 a5 11.Nh3 leads to complicated play.
 
The text move is always useful in these positions, as after White plays a2-a3 Black will not need to
move the bishop from b4. Black’s subsequent plans may involve putting his bishop on e6 and knight
on e5, followed by advancing on the queenside with ...c6 and ...b5.
 
10.a3
If White decides not to move on the queenside, play may continue 10.g4 Be6 11.Kb1 Ne5 12.g5
Nfd7 13.Qf2 Nc5ƒ followed by ...c6 and ...Qb6 with a nice initiative on the queenside.
 
10...Ne5
Black does not fear losing his bishop.
 
11.g4
White should advance on the kingside sooner or later.
 
Capturing the bishop with 11.axb4? would be a mistake due to 11...axb4 12.Nb1 c5!µ followed by
...Ra1, with a nasty attack.

514
11...Be6 12.Qg2 Nc4!
Forcing a favourable trade.
 
13.Bxc4 Bxc4³
In Ponkratov – Matlakov, Moscow 2019, Black enjoyed the bishop pair and good attacking chances
on the queenside, while White’s kingside attack has not really started yet.
 

C2) 4.Qe3

515
 This is by far the most popular retreat of the queen.
 
4...Nf6
This natural developing move leads to a further split between C21) 5.e5?!, C22) 5.Bd2 and C23)
5.Nc3.
 

C21) 5.e5?!
 
This overoptimistic move will backfire.
 
5...Ng4 6.Qe4
The alternative is:
6.Qe2 d6!
A strong sacrifice to open up the position and exploit White’s lag in development.
7.exd6†
7.h3 doesn’t work, as 7...Ngxe5 sees Black safely grab a pawn without losing his knight. 8.f4
was seen in Khandelwal – Addison, Swansea 2006, and now 8...Nd4!N would have been best,
and after 9.Qf2 Nec6µ Black is simply a pawn up.
7...Be6 8.h3 Nge5 9.f4 Nc4

10.dxc7?
10.Nf3N is the lesser evil although White will still suffer after: 10...Bxd6 11.Ng5 Be7!µ A
strong move, preventing White from taking on e6 due to the deadly check on h4. Black also
threatens to exchange on g5 followed by ...0-0, when White’s lack of development will tell.

516
10...Qxc7 11.Nf3 0-0-0–+
Black had a crushing initiative in Gagunashvili – Sanikidze, Tbilisi 2016.
 
The text move is met by a strong sacrifice:

6...Ngxe5!
Black gives up a piece to obtain a fierce initiative.
 
7.f4 d5 8.Qe2
8.Qe3 d4 9.Qe2 Bd6 10.fxe5 Nxe5µ followed by ...0-0 leaves Black with two pawns for the piece,
plus a potent attack against the enemy king.
 
8...Bg4 9.Nf3 Bc5 10.Kd1
White removes his king from the e-file, but it doesn’t help much.
 
10.fxe5? Nd4–+ is crushing.
 
10.h3 is met by 10...Bxf3 11.gxf3 0-0 12.fxe5 Nxe5© with amazing compensation for the piece.
 
10...0-0 11.fxe5 Re8 12.Qb5
12.Bf4 is met by 12...Bd4! 13.Nbd2 Nxe5 14.Bxe5 Rxe5µ with big problems for White.

517
12...Nxe5 13.Be2
Black has no immediate win but he can play it cool with:
 
13...Bxf3 14.Bxf3 Bb6 15.Qf1 c6µ
Black has great compensation, with two pawns for the piece and a lasting initiative. In Soszynski –
Nichols, email 2010, White never managed to solve the problem of his unsafe king.
 

C22) 5.Bd2

518
 White can opt for a slight change in the move order with this move, which is designed to prevent
the pin with ...Bb4 before playing Nc3. However, it gives Black another tempting option:
 
5...Be7!?
Preparing for a quick break in the centre with ...d5.
 
5...Bb4 is a decent move which is likely to transpose to variation C23 after 6.Nc3. However, White
also has alternatives such as 6.a3 Bxd2† 7.Nxd2 followed by 0-0-0, with a playable position for both
sides.
 
6.Nc3 d5!
It is important to carry out this move immediately, before White prevents it with Bc4.
 
7.exd5
7.Nxd5 Nxd5 8.exd5 Qxd5 transposes.
 
7...Nxd5 8.Nxd5
A popular but misguided alternative is:
8.Qg3?
This is practically refuted by:
8...Ncb4!
Hitting c2 while preventing 0-0-0.
9.Bd3
9.0-0-0?? loses outright to 9...Nxc3 10.bxc3 Nxa2† 11.Kb2 and now 11...Nxc3!–+ is the key
detail.
9.Bb5† c6 10.Ba4 is met by 10...Bf5µ when all Black’s pieces are aggressively placed, and he
was clearly better in Rodriguez Garcia – Ibarra Jerez, Las Palmas 2017.
9...0-0 10.Nge2 Nxd3† 11.Qxd3
Now in Bergsson – Cohen, Reykjavik 2009, the strongest continuation would have been:

519
11...Nb4!N 12.Qxd8 Rxd8 13.0-0-0 Bf5µ
White will lose the c2-pawn.
 
8...Qxd5

9.Ne2 0-0 10.Nf4

Driving the queen away from d5 so that White can castle.
 
10.Nc3?!
This method of attacking the queen is worse due to:
10...Qh5!

520
 White still cannot castle.
11.Be2 Qg6 12.Bd3
White sacrifices the g2-pawn, hoping that the g-file can be used for attacking purposes.
However, Black can safely grab it as he can defend his kingside relatively comfortably.
12.0-0 Bf5³ puts the c2-pawn under pressure and Black is clearly on top, Sanchez Castillo –
Houska, Turin (ol) 2006.
12...Qxg2 13.0-0-0 Qh3 14.Qe4
14.f3?! was met strongly by 14...Bf5 15.Ne4 Bg6µ and with the bishop solidly blocking the g-
file, Black was clearly better in Atabayev – Kayumov, Dubai 2011.
We have been following Trifunovic – Aleksandrov, Bad Woerishofen 2001. I suggest:

14...Be6N³
White has some chances on the kingside, but not enough to provide full compensation for the
missing pawn.
 
10...Qc5 11.0-0-0 Bf5

521
12.Qxc5 Bxc5 13.Bd3 Bxd3 14.Nxd3 Bb6=
The endgame was completely equal in Artemiev – Fedoseev, Poikovsky 2018.
 

C23) 5.Nc3

5...Bb4
Black develops his bishop actively and pins the knight. His further plans include ...0-0 followed by
putting pressure on the e4-pawn.
 

522
6.Bd2 0-0 7.0-0-0 Re8
We will analyse C231) 8.Qf4!?, C232) 8.Bc4 and C233) 8.Qg3.
 

C231) 8.Qf4!?
 
Nepomniachtchi has played this move a few times in recent years, though always at fast time limits.
White moves his queen away from the e-file and keeps the e4-pawn supported. Compared to the
main line, Black can still win the central pawn but here he has to part with the dark-squared bishop,
which White hopes to exploit in some way.
 
8...Bxc3 9.Bxc3 Rxe4
9...Nxe4 is also playable, when 10.Bc4 is messy. Black should be doing well with accurate play,
but the text move seems simpler.
 
10.Qg3

This is the position White was aiming for. The combination of the queen on g3 and bishop on c3
gives White a certain amount of pressure against the kingside and on the dark squares, which he
would like to increase. I really like how Anand neutralized White’s play in the game which we will
follow.
By the way, an almost identical situation can arise after the more common 8.Qg3 Rxe4 (rather than
8...Nxe4, which I recommend in variation C233 below) 9.a3 Bxc3 (Black tends to do better when he
preserves this bishop in some way) 10.Bxc3, when the position is the same except that White’s pawn
is on a3 instead of a2. The present position with the pawn on a2 is a slightly worse version for White,
as we will see below.
523
 
10...d5
Black prepares to develop his bishop, while preparing an important tactical motif.
 
11.f3 Re6!
Black was successful with 11...Re8 in Nepomniachtchi – Carlsen, Leuven (blitz) 2017, but the text
move is more accurate.
 
12.Ne2 Rd6
Black already threatens ...d4, driving the bishop back and killing White’s pressure on the dark
squares.
 
13.Qf4
White tries to prevent Black’s idea but he can play it anyway!
 
13...d4! 14.Nxd4 Nd5 15.Nxc6 bxc6
White cannot avoid material losses; Nepo decided to sacrifice the exchange but to no avail.

16.Rxd5 cxd5 17.Be5 Re6µ

Nepomniachtchi – Anand, Saint Louis (blitz) 2017.

C232) 8.Bc4

524
 White develops another piece and prevents ...d5 for the moment, but the drawback is that the
bishop may become a target.
 
8...d6 9.f3
Reinforcing the e4-pawn while preparing g2-g4.
 
The alternative is:
9.Nf3
White develops his knight but does nothing to create play on the kingside. Black can simply go
for a favourable exchange of bishops with:
9...Be6 10.Bxe6 Rxe6
The e4-pawn is under pressure, so the logical continuation seems to be:
11.Ng5 Re8

525
12.f3
The aggressive 12.f4?! turned out badly after 12...h6 13.h4 Qc8!µ in Tartakower – Reshevsky,
Stockholm 1937. Black covers the h3-square and thus threatens to take on g5. White’s play on
the kingside has come to a halt, and Black’s pressure on the e4-pawn will begin to tell.
12...Ne5
Black moves his knight to a strong central outpost, eyeing the c4-square and preparing a
general queenside advance with ...c6, ...b5 and so on.
13.Qe2
13.Nh3 enabled Black to gain a strong initiative with 13...Nc4 14.Qd4 b5!, and after the further
mistake 15.Bg5? c5–+ Black already had a crushing position in A.P. Smith – Turner, London
2018.
13...Qe7 14.Nh3 c6 15.Nf4 a5 16.Nd3 b5µ
Black clearly had the better attacking prospects in Munoz Pantoja – Vidal Zamora, Barcelona 2017.
 
9...Ne5
9...Na5 is also possible, although after 10.Be2 Black should be sure to play 10...Be6 with a fine
position, rather than 10...d5?! 11.Nxd5! Nxd5 12.Qd3² when White was poised to regain the piece
with a slightly better position in A.P. Smith – Flear, Torquay 2013.

526
10.Bb3
10.Be2 can be met by 10...Bc5 11.Qg5 c6 when Black is ready to advance on the queenside with
...b5. The game continued 12.h4 h6 13.Qg3 b5!µ when Black had the faster attack in Groot –
Hlavacek, corr. 2017.
 
10...c6
Here too, Black has the simple plan of advancing his queenside pawns.
 
11.g4 a5 12.g5 Nfd7 13.f4
White wastes no time and counterattacks on the kingside.
 
13...Bxc3!
Eliminating the knight in preparation for the following pawn move.
 
14.Bxc3 a4

527
15.Bxf7† Nxf7 16.Qd4
This may look scary but Black is doing well after:
 
16...Nde5! 17.g6 hxg6 18.fxe5 c5 19.Qd3 b5!µ
Materially White is fine, but Black’s queenside attack was too fast and too strong in Cardelli –
Koskela, corr. 2011.
 

C233) 8.Qg3

528
8...Nxe4
8...Rxe4!? is the most popular move and is also good for Black, but I believe the text is stronger.
 
9.Nxe4 Rxe4
We have reached a final split. C2331) 10.c3 has been the most popular choice for White, but
C2332) 10.Bf4 also requires attention.
 

C2331) 10.c3 Be7!

 
This is the most accurate retreat, intending to annoy the queen with ...Bh4.
 
11.Nf3!?
This rare but logical developing move controls the h4-square.
 
11.Bd3 Bh4! creates space for the rook to come back. After 12.Qf3 Re8 White’s queen is on a poor
square and Black threatens ...Ne5. A good example continued:
 

13.Qf4 Ne5 14.Bc2 d5 15.Nf3 Nxf3 16.Qxf3 Be6µ Black was fully developed and White did not
have the slightest compensation for the pawn in Belka – Koch, corr. 2011.
 
11.f4
This has been the most common continuation: White marches his f-pawn forward before
developing his kingside pieces, hoping that the f-pawn will help in creating some play.
11...d5 12.Bd3

529
 Unfortunately for White, his play is more or less refuted by a powerful exchange sacrifice:

12...Bf5! 13.Nf3
13.Bxe4 Bxe4 14.Nf3 Qd7 transposes.
13...Qd7 14.Bxe4 Bxe4 15.Be3 Na5
Black has a nasty attack on the light squares, with ideas including ...Qa4 and ...Nc4.
 

16.Bd4 f6
This not only prevents the mate but also covers the e5-square.
16...Bf8 is a good alternative which led to another convincing win for Black in Rudd – Avrukh,
London 2010.
17.Rhe1 Qa4–+
530
 Black had a winning advantage in Kriebel – Michalik, Slovakia 2017.

11...d6
11...d5?! is inaccurate due to 12.Bf4, with pressure on the c7-pawn.
 
The text move avoids that problem and threatens ...Rg4.
 
12.h3 Ne5!
Black not only offers the exchange of knights but also prevents Bd3 for a while.
 
13.Nxe5
Piece exchanges generally favour Black, as they take him closer to a winning endgame while also
reducing White’s attacking potential, if there is any! Despite all that, the text move seems to be
White’s best try.
 
His position comes unglued more quickly after:
13.Ng5?!
A rather desperate attempt to create problems on the kingside.
13...Ra4 14.a3
14.f4 Bxg5 15.fxe5 Bxd2† 16.Rxd2 Rxa2 was simply winning for Black in Rudd – Haslinger,
Liverpool 2008.
14...Bf5 15.f4
Here Black played 15...Nc6 and went on to win in Langschmidt – Hahner, Remote email 2013,
but he could have won more quickly and stylishly with:

531
15...Qd7!!N 16.fxe5 Rxa3!! 17.bxa3 dxe5
Despite being a rook and a piece up, White cannot cope with the fire-breathing bishops which
point towards his king.
18.Kb2 Qa4–+
White can resign.
 
13...Rxe5
Black remains a pawn up and is ready to develop his light-squared bishop next.
 

14.f4

532
 Trying to make the rook’s life uncomfortable.
 
14.Bc4 is well met by: 14...Bf5 15.Rde1 Rxe1† 16.Rxe1 d5! Blocking the bishop’s diagonal. 17.Bb3
a5µ White had no compensation whatsoever in Relyea – Landes, corr. 2014.
 
14...Ra5 15.a3 Bd7
Black develops the bishop to a safe square where it will not be harassed by the enemy pawns.
Black has no weaknesses and several pieces have been exchanged, so White’s compensation is in
great doubt. A good example continued:
 
16.Bd3 Bh4!?
Forcing the queen to move to an inferior square.
 
17.Qe3 Qe8!
Black would obviously be happy to simplify towards an endgame.
 
18.Qf3 Bc6 19.Qg4 Qe7µ
The bishop on h4 meant that White was unable to take control over the open e-file, so White had
no real compensation for the pawn in Relyea – Renard, corr. 2018.
 

C2332) 10.Bf4

10...Qf6! 11.Nh3
The c7-pawn is poisoned due to 11.Bxc7? d6–+ and the bishop will be trapped.

533
 
11...d6 12.Bd3
White seems to be gaining some activity but Black has a strong retort:
 
12...Nd4!
Bringing the knight into action. This precise continuation was shown by the Cuban genius
Capablanca back in 1913!
 
13.Kb1
13.Bg5 Qe6 14.Kb1 Rg4 15.Qe3 was seen in Palomino – Pachon, Bogota 2011, when 15...h6!N
16.Bf4 Bc5µ would have left Black a clear pawn up.
 
13.Be3 should be met by 13...Nf5!N (rather than 13...Bg4?? 14.Ng5! when Black was in trouble in
Mieses – Capablanca, Berlin 1913, although he later outplayed his opponent from a lost position and
won) 14.Qf3 Re8µ when Black keeps a safe extra pawn.
 
13...Ne2 14.Qf3
14.Bg5N 14...Nxg3 15.Bxf6 Nxh1 16.Bxe4 Bxh3 17.Rxh1 Be6µ leaves Black with a clear extra
pawn in the endgame.
Now Black can secure a big advantage with the help of an amazing tactical resource:

14...Ba3! 15.bxa3
15.Bc1 Qxf3 16.gxf3 Re8µ left Black a clear pawn up in Brankov – Cemhan, Belgrade 2013, and
16...Re7!N–+ would have been even more precise.
 
15...Nc3† 16.Kc1 Nxa2† 17.Kb1 Nc3† 18.Kc1 Bxh3 19.Bxe4
534
 After a mostly forced sequence, the final tactical detail is:

19...Bxg2! 20.Qxg2 Qxf4†–+

Black had a decisive material advantage in Nepomniachtchi – Vallejo Pons, Moscow 2007.

Conclusion
 
This chapter has dealt with a variety of possibilities after 1.e4 e5 2.d4 exd4. Firstly 3.c3 dxc3 leads to
gambit play: 4.Nxc3 gives White a lead in development but Black remains solid, with decent chances
to consolidate his extra pawn; and 4.Bc4 sacrifices a second pawn, but Black quickly returns the
material to reach a comfortable endgame.
 
3.Nf3 is an attempt to reach some sort of Scotch-type opening, but 3...Bb4† is a convenient reply.
Then 4.Bd2 Bxd2† gives Black a pleasant game after either recapture, while 4.c3 is a somewhat
speculative pawn sacrifice, for which White will do well to prove his compensation.
 
Finally, 3.Qxd4 Nc6 sees White lose time with his queen in order to facilitate long castling, which he
hopes will unsettle Black in some way. The main line continues 4.Qe3 Nf6 5.Nc3 Bb4 6.Bd2 0-0
7.0-0-0 Re8 when Black has pressure against the e4-pawn and White usually sacrifices it in some
way. Although his scheme is objectively rather dubious, a number of the ensuing variations are
difficult to find over the board, so it is worth making the effort to memorize some of the main details.
If you are well prepared with the recommendations I have provided, you will have excellent chances
to score a full point.

535
A) 4.d4? 265
B) 4.Bc4?! 266
C) 4.Qf3 267
D) 4.Bb5† 269
E) 4.Nf3 Nf6 271
E1) 5.Be2 271
E2) 5.c4 273
E3) 5.Bc4 Nxd5 275
E31) 6.Bxd5 275
E32) 6.0-0 277
E4) 5.Bb5† c6! 6.dxc6 Nxc6 7.d4 Bd6 279
E41) 8.Qe2†?! 280
E42) 8.0-0 281

 
1.e4 e5 2.f4
Welcome to the ever-romantic King’s Gambit.

536
 
2...d5
2...exf4 is by far the main reply, but it has already been well analysed in many books, one example
being Ntirlis’s book, also published by Quality Chess. I was keen to analyse something a little
different, and I found the text move (with ...exf4 to follow) a good choice. Another nice feature of
my recommended line is that Black gets to dictate the play, as opposed to 2...exf4 when it is White
who gets to choose between 3.Nf3 and 3.Bc4.
 
3.exd5 exf4

This is the main tabiya for the chapter. Black threatens ...Qh4†, creating long-term problems for
the enemy king. White can ignore the danger with A) 4.d4? or B) 4.Bc4?!, or play more prudently
with C) 4.Qf3, D) 4.Bb5† or E) 4.Nf3.

A) 4.d4?
 
Allowing ...Qh4† is not a good idea, especially when the king has to step forwards rather than to f1.
 
4...Qh4† 5.Kd2
5.Ke2 Bg4† 6.Nf3 occurred in Bove – Ragger, Budva 2003, when 6...Bxf3†!N would have been
the right choice to force White to compromise his pawn structure. After the further 7.gxf3 Nd7 8.c4
0-0-0 9.Kd2 Ne7µ White’s king must feel miserable while his central pawns will soon be put under
pressure by Black’s army.
 
After the text move Black has several ways to maintain a favourable position, but the best is:

537
5...Qd8!N
Having disturbed the white king, Black returns the queen to base and will now focus on developing
his pieces.
 
6.Nf3
6.Nc3 Nf6 7.Bc4 c6µ wins a pawn.
Also after 6.c4 c6 7.Nc3 Nf6 8.Nf3 cxd5µ Black is clearly better.
 
6...Nf6!
In the event of 6...Qxd5?! 7.Nc3 Bb4 8.a3 Bxc3† 9.Kxc3³ White is not doing all that badly,
despite the unusual king position.
 
7.Bb5† c6 8.Re1† Be7 9.dxc6 Nxc6µ
White has no queenside development and it will take him some time to bring his king to safety. A
sample continuation could be:

538
10.Ke2 0-0 11.Bxf4
11.Kf1 Bg4 12.Bxc6 bxc6 13.Bxf4 transposes.
 
11...Bg4 12.Bxc6 bxc6 13.Kf1
White has won a pawn and brought his king to safety, and he just requires a few spare tempos to
develop his queenside. However, Black can capitalize on his dynamic advantages as follows:
 
13...Nh5! 14.Be3 Bh4!
The key move, disturbing the coordination of White’s pieces. The tactical justification is:
 
15.g3

539
15...Bxg3!! 16.hxg3 Qd6! 17.Bf2 f5µ
Black is ready to break open the king’s fortress with ...f4, with a formidable attack.
 

B) 4.Bc4?!

White develops his bishop and defends the d5-pawn, but the drawback is that it allows Black the
annoying check on h4. Even with the f1-square available for the king, this is not a good idea for him.
 
4...Qh4†

540
 As usual, whenever Black gets a chance to prevent White from castling, he should go for it.
 
5.Kf1 Bd6
Black has a simple plan of ...Ne7 and ...0-0, while it is not so clear how White should coordinate
his forces.
 
6.Nc3
6.Nf3 Qh6 7.Nc3 Ne7 8.Ne4 0-0 leaves White with nothing better than 9.d4, transposing to the
note on 7.d4 below.
 
6.d3 Ne7 7.Nc3 0-0 8.Ne4 Nd7 9.Qe1 occurred in Bonafede – Cooklev, Albena 2014, and now
9...Qh6!N would have kept the queens on the board, with clearly better play for Black, who can
consider playing ...b5 next.

6...Ne7 7.Ne4
7.Nf3 Qh6 transposes to the above note on 6.Nf3.
7.d4 is met by 7...0-0 8.Nf3 Qh6 9.Ne4 and now 9...Nd7! is best, threatening ...Nb6. 10.Bb3 Nf6
11.Nxf6† Qxf6–+ Kavcic – Pavasovic, Bled 1998. All Black’s pieces will soon gang up against the
white king.
 
7...0-0 8.Qe1
Westerinen – Agdestein, Espoo 1989. Just as in the lines noted above, Black should not exchange
queens as his king is much safer than its counterpart. The best choice is:

541
8...Qh6N
White will have a lot of troubles due to his unsafe king.

C) 4.Qf3

White defends d5 and guards against the check, but loses time moving his queen to a rather
exposed square.
 
4...Nf6
There is little to be gained from 4...Qh4†?! now, due to the simple 5.g3.

542
 
5.Nc3
I also considered:
5.Bb5†
This is well met by:
5...c6! 6.dxc6 Nxc6 7.Ne2
Grabbing the pawn is of course bad for White after 7.Bxc6†? bxc6 8.Qxc6† Bd7–+ when
Black’s pieces will develop quickly, creating deadly problems for the white king, Daikeler –
Augustin, corr. 1987.

7...Be7 8.Nbc3
8.0-0?? Qb6†–+ wins a piece.
8.Qxf4 0-0 9.Bxc6 bxc6 10.0-0 Bd6 11.Qh4 Rb8!ƒ followed by ...Rb4 gives Black a powerful
initiative.
8...0-0 9.Bxc6 bxc6 10.Qxf4N
10.0-0 occurred in Khan – Bascetta, email 2009, when 10...g5!Nµ would have supported the
valuable f4-pawn.
The text move is worth checking: White captures the pawn but Black can use the fact that
White has not yet castled to gain something in return.

543
10...Bd6 11.Qh4 Qb6 12.d4
I also checked 12.Qf2 Re8 13.Qxb6 (13.0-0? Bxh2†–+) 13...axb6 14.0-0 b5! when Black is
clearly better. For example, 15.Ng3 b4 16.Nce2 and now 16...b3! is a nice touch, the point of
which is revealed after 17.cxb3 Ba6–+ when White cannot defend with d2-d3.
12...Bf5
12...c5!? is also good.
13.0-0 Bxc2µ
Black has the bishop pair and the better pawn structure.
 
5...Bd6 6.Bb5†
White could also try:
6.d4
Aiming to capture the f4-pawn immediately, but Black can play a strong gambit.
6...0-0 7.Bxf4 Bg4 8.Qf2 Re8† 9.Nge2
9.Be2N is strongly met by 9...Bb4! when the d5-pawn comes under fire. A sample continuation
could be: 10.0-0-0 Bxc3 11.Bxg4 (11.bxc3? Bxe2 12.Nxe2 Qxd5–+ is crushing) 11...Nxg4
12.Qg3 Bxb2† 13.Kxb2 Nf6 14.Be5 Nbd7 15.Nf3 (winning the c7-pawn is not good either:
15.Bxc7 Qc8 16.d6 Nb6!–+ and with ...Nbd5 coming next, Black will soon recover the pawn
with a strong attack to boot) 15...Qe7 16.Ka1 Nxe5! 17.dxe5 Ne4 18.Qe1 (18.Qf4 Nf2–+)
18...Ng5µ Black wins the e5-pawn.
Now in Golubov – Bindrich, Batumi 2018, Black exchanged on f4, which was not a bad move
in itself. However, a more accurate continuation is:

544
9...Bb4!N
We saw the same idea in the 9.Be2 line just above.
10.Bg5
10.0-0-0? Bxc3 11.bxc3 Qxd5–+ is crushing.
10...Bxc3† 11.bxc3 Nbd7 12.h3 Bh5 13.g4 Bg6 14.Bg2 Qe7µ
White still has to solve some problems with his king.
 
6...c6!
A typical pawn sacrifice.
 
7.dxc6 0-0! 8.cxb7 Qe7†
A simple but useful tactical trick to have in your arsenal for such positions.
 
9.Qe2 Qxb7 10.Nf3 Nc6N
Black opted to chase the bishop with 10...a6 in B. Savchenko – Torosyan, Batumi 2018, but I
would prefer to develop the knight first.

545
11.0-0 Re8 12.Qc4
If 12.Qd1 Bg4 13.d4 Rad8ƒ Black’s forces are fully mobilized and ready for action.
 
12...Bd7
Black has a fine position and White will have to play precisely to avoid landing in a significantly
worse position. A sample line could be:
 
13.d4 a6 14.Ba4 Rac8

15.Ng5!
The most accurate, but it does not fully solve White’s problems.
546
 
15...Nd8 16.Qb3 Qa7!? 17.Bxd7 Qxd4† 18.Kh1 Nxd7³
White still has to take care of his queenside.
 

D) 4.Bb5†
 
White develops his bishop with tempo and also eliminates his weak pawn on d5 after Black’s reply.

4...c6!
This is our typical reaction whenever the bishop check occurs: White gets to exchange off his weak
pawn, but it is worth it for Black to speed up his development and open lines for his pieces.
 
5.dxc6 Nxc6 6.d4
6.Qf3 Nf6 transposes to the note to White’s 5th move in variation C above.
 
6.Nf3 Bd6 7.d4 (7.0-0?? Qb6†–+) transposes to variation E4 after 7...Nf6.
 
6...Bd6 7.Nc3
7.Nf3 Nf6 is another transposition to variation E4.
7.d5?! doesn’t win a piece due to 7...a6 8.Ba4 b5 9.dxc6 bxa4µ as seen in Mary – Vasseur, France
1992; White will have a hard time coping with Black’s bishop pair.
 
7...Nf6
7...a6?! was played in B. Savchenko – Shcherbakov, Sochi 2018, but it wastes valuable time.

547
 
8.d5!
White has to play a series of precise moves to avoid ending up in a clearly worse position.
 
8.Nge2?!N with the intention to grab the f4-pawn cannot be recommended in view of 8...0-0 9.Bxc6
(9.0-0 g5µ) 9...bxc6 10.Bxf4 Bxf4 11.Nxf4 Re8† 12.Nfe2 Ng4µ when the dark squares in White’s
camp are too weak.
 
Simple development with 8.Nf3?! is also not enough: 8...0-0 9.0-0 h6µ Followed by ...g5 and White
will be under pressure on the kingside, Campbell – Romilly, Dundee 2011.

8...a6 9.dxc6
I also considered:
9.Ba4N 9...b5 10.dxc6 bxa4
Black is doing well, for example:
11.Nge2 Qc7 12.Qd4!
Avoiding 12.0-0? 0-0 13.Bxf4 Bc5† 14.Kh1 Qxc6µ when it will be difficult for White to
handle the pressure on both the a8-h1 and a7-g1 diagonals.

548
12...a3!
Ruining White’s pawn structure.
12...Qxc6 leads to approximate equality after: 13.Bxf4 Bxf4 14.Qxf4 0-0 15.Qxa4=
13.Bxf4
13.bxa3? is bad due to 13...Qxc6 14.Bxf4 Bc5µ.
13...axb2 14.Rb1 Bxf4 15.Nxf4 Qxc6³
Black’s better pawn structure gives him the upper hand.
 
9...axb5 10.Nxb5
Now we can improve with:

549
10...Qe7†N
10...Qa5† was less precise in Koestner – Teske, email 2012.
11.Qe2
In the event of 11.Ne2? Be5µ Black’s two bishops are even more dangerous with queens on the
board.
 
11...0-0 12.Qxe7 Bxe7 13.Bxf4 bxc6
Black is a pawn down for now, but the bishop pair offers plentiful compensation. A sample line
could be:
 
14.Nc3 Re8 15.0-0-0
15.Nge2 Bb4 16.0-0-0 transposes.
 
15...Bb4 16.Nge2 Bg4
The white knights are feeling the heat.
 
17.Rde1

17...Ne4!
Stepping up the gas on every move!
 
18.a3 Ba5 19.b4 Bb6 20.Kb2 g5©
Depending on where White moves his bishop, Black will follow up with either ...Nd6 or ...Nd2,
with strong compensation.
 

550
 E) 4.Nf3
 
This is the most natural and popular move. White develops the knight to its best square and prevents
the check on h4. This position can also be reached via 2...exf4 3.Nf3 d5 4.exd5, but our chosen move
order has avoided the option of 3.Bc4.
 
4...Nf6
4...Qxd5?! grabs a pawn but wastes precious time, and 5.d4 Bd6 6.c4 Qe6† 7.Be2 c6 8.0-0© gave
White strong compensation due to his lead in development in Ribeiro – Teshima, Baena Neves 2015.
 
We have reached another important branching point, where White can choose between E1) 5.Be2,
E2) 5.c4, E3) 5.Bc4 and E4) 5.Bb5†.
Let’s have a brief look at a very harmless option:
5.Nc3 Nxd5 6.Nxd5
6.Bc4 Nxc3! 7.dxc3 (7.bxc3 Bd6 and White lacks any reason for his position) 7...Qxd1†
8.Kxd1 Bd6 9.Re1† Kf8 has been played a number of times. Black is much better.
6...Qxd5 7.d4 Be7 8.c4 Qd6!
This is the only key move to know.
9.Kd2
So far Hector – A. Smith, Stockholm 2017. Here there is an obvious improvement:
9...g5!N
We are in uncharted territory, but it is safe to say that no one would prepare to play this with White.
Black is much better.
 

E1) 5.Be2
 
This modest developing move has been played twice by Rapport. White waits for Black to capture on
d5 with the knight before pushing his pawns to c4 and d4.
 
5...Nxd5 6.c4
6.0-0?! is inferior due to 6...h6N when Black will comfortably defend his extra pawn with ...g5,
and it will be tough for White to get compensation.

551
6...Ne7
This is a relatively simple and practical approach.
 
6...Nf6?! makes it too easy for White to regain the f4-pawn. 7.d4 Bd6 8.c5 Be7 9.Nc3 0-0 10.Bxf4
Be6 11.0-0 was close to equal in Rapport – Gelfand, Hengshui 2019, but I slightly prefer White’s
chances.
 
6...Nb4!?
None of the top players have opted for this move, but the engines rate it most highly. If you
wish to punish White’s opening choice and are not afraid of tactics, then you may wish to
consider it. Here is my analysis:
7.d4 Bf5
Forcing White’s knight to go to a3, after which the knight cannot be kicked from b4.
8.Na3 g5!
Black should not give away the f4-pawn too easily – not only for materialistic reasons, but also
because it severely limits the c1-bishop and blocks any play along the f-file.

552
9.0-0
9.h4N may seem frightening but 9...N8c6! is an excellent reply, preparing to castle as fast as
possible. 10.d5?! (10.hxg5 is met by 10...Qe7µ intending ...0-0-0, when White is far behind in
development and has a lot of weaknesses) The text move appears critical but it is refuted by a
cool move: 10...Bc5!! Locking down the king on e1. 11.dxc6 Qxd1† 12.Kxd1 0-0-0† Black has
a winning initiative.
The text move was played in Triger – Tardif, Torcy 1991. A simple improvement is:
9...h6N
Solidifying the kingside pawn chain and preventing any Nxg5 tricks. My illustrative line
continues:
10.g3 N8c6 11.d5
11.gxf4 g4–+ is horrible for White.
11...Bc5† 12.Kh1
Once again Black should have no qualms about sacrificing the knight:
12...Qe7! 13.dxc6 bxc6–+
All Black’s pieces are active and White’s position is riddled with weaknesses.
 
7.Nc3
7.d4 Ng6 8.Nc3 Be7 threatens a check on h4. 9.Nd5 (9.0-0 is met by 9...0-0 and after: 10.Nd5?
[White should prefer 10.Kh1N although 10...c5!³ still favours Black] 10...Bd6 11.c5 Now 11...c6!µ
is an important detail which led to a big advantage for Black in Otten – M. Bensdorp, Hilversum
2007)

553
9...Bh4† 10.Nxh4 Qxh4† 11.Kf1 Na6 12.b4 0-0 13.b5 c6!–+ Black already had a winning advantage
in Grigoryan – Pashikian, Yerevan 2018.
 
7...Ng6 8.h4!?
White does not want Black to develop smoothly.
 
8.d4 transposes to the previous note.
 
8...Be7 9.h5 Nh4 10.Nd5 Nc6
10...Bg5!?N³ is another good option when White will have to play accurately to prove any
compensation.

554
11.d4 Nxg2† 12.Kf1 Ne3† 13.Nxe3 fxe3 14.d5 Nb4 15.a3 Na6 16.Bxe3 0-0³
White’s compensation was still in question in Rapport – Adams, Biel 2015.

E2) 5.c4

This move is met convincingly by:

 
5...c6!
Immediately breaking up White’s centre.
 

555
6.d4
6.dxc6?! hands Black a lead in development: 6...Nxc6 7.d4 Bb4† 8.Nc3 0-0 (I find this more
accurate than 8...Ne4 as played in Grafl – Berzinsh, Germany 2003) 9.Be2 Ne4 10.Bd2 This position
was reached in Dedic – Halilovic, Sarajevo 2018, when Black opted to exchange on d2. Instead I
recommend supporting the powerful knight with:

10...f5!Nµ Apart from his uncastled king, White has to take care of his d4-pawn as well.

6...Bb4†
6...cxd5 is also possible, but after 7.c5 the bishop gets locked in on f8. Still, after 7...b6 the position
remains unclear, with chances for both sides.
 

556
The text move seems preferable to me: the idea is to develop as fast as possible while activating the
bishop before White plays c4-c5.
 
7.Nc3
White can also try: 7.Bd2 Bxd2† 8.Qxd2 cxd5 9.Nc3 (9.c5?N makes no sense as the bishop has
already escaped from f8: 9...0-0 10.Be2 b6! 11.0-0 [11.cxb6 Qxb6 12.0-0 Ne4 13.Qxf4 Nc6!µ leaves
White with weak pawns] 11...bxc5 12.dxc5 Qe7 13.Rc1 Nc6µ White’s pieces are uncoordinated and
Black is in complete control)

9...0-0 10.Qxf4 Nc6 11.0-0-0 Be6 Black had the better chances in Martinez Martin – Meszaros,
Peniscola 2002, due to his ideally-placed pieces and attacking prospects against the white king.
 
7...0-0 8.Be2
White prepares to castle before taking on f4.
 
8.dxc6?! Nxc6 transposes to the note to White’s 6th move above.
 
White can also take the pawn back immediately:
8.Bxf4 cxd5 9.Be2
9.Bd3 dxc4 10.Bxc4 transposes.
9...dxc4 10.Bxc4
10.0-0?! is well met by 10...Bxc3! (the immediate 10...Be6 is countered by 11.Ng5) 11.bxc3
Be6³ with good control over the light squares in the centre, so White will find it hard to prove
his compensation.

557
10...Qe7†!
10...Re8†?! is inaccurate due to 11.Ne5= when the attack on f7 gave White time to castle next
in Stojcevski – Dorfanis, Skopje 2012.
11.Be5
11.Qe2 Qxe2† 12.Kxe2 Nc6³ left Black with the safe king and preferable pawn structure, and
White eventually went down in Gaviota 1.01 – Gull 3, Engine Game 2017.
11...Ng4 12.0-0
12.Qd3N covers the e3-square but 12...Nc6 13.0-0 Ncxe5 14.Nxe5 Nxe5 15.dxe5 Bxc3
16.Qxc3 Be6³ leaves Black with the superior pawn structure.

12...Bxc3!

558
 12...Ne3?! is dubious due to 13.Qd3, and if 13...Nxf1? 14.Nd5 White wins.
13.bxc3 Ne3 14.Qd3 Nxf1³
White’s piece activity was not enough to claim full compensation for the exchange in Krol –
Morcin, email 2014.
 
8...cxd5 9.0-0 Nc6 10.Bxf4
10.c5?! is strongly met by 10...Ne4³.
 
10...dxc4 11.Bxc4 Bg4
Creating pressure against the d4-pawn.
 
12.a3 Bd6
Black has a pleasant position, for example:

13.Be3 Rc8 14.Ba2 Rc7 15.Kh1 Rd7=

Black was completely fine in Salvador Marques – Slawinski, corr. 2017.
 

E3) 5.Bc4

559
5...Nxd5
After this natural move White has a choice between two main options: E31) 6.Bxd5 and E32) 6.0-
0.
 

E31) 6.Bxd5
 
White wants to develop his knight with tempo, but trading off the bishop is a concession.
 
6...Qxd5 7.Nc3
7.0-0?! is inferior due to the simple 7...Bd6 8.Nc3 Qf5 9.d4 0-0µ when White had no easy way to
regain the f4-pawn in Schieder – Menezes, Graz 2017.
 
7...Qf5
Black anticipates an attack on the f4-pawn and protects it in advance, making it difficult for White
to capture it.
 
7...Qh5 8.d4 Bb4 9.0-0 0-0 10.Bxf4 was approximately equal in Cheparinov – Guliev, Baku 2011.
 
8.d4
8.0-0?! is worse in view of 8...Nc6! 9.d4 Be6 when Black is all set for ...0-0-0. Play may continue:

560
10.Ne5 (after 10.Nh4? Qa5–+ Black threatens the simple ...Nxd4 winning a pawn, or simply ...0-0-0
with an overwhelming advantage, Feicht – Baldauf, Zirndorf 2017) 10...Nxe5 11.Rxf4 Qg6 12.dxe5
Bc5† 13.Kh1 0-0µ Black benefited from the bishop pair and the better pawn structure in Barle –
Delchev, Ljubljana 2003.
 
8...Bd6
The bishop should usually go to d6 in this variation, overprotecting the f4-pawn.

9.Qe2†!
White should try to create some problems in Black’s coordination.
9.0-0?! transposes to the note to White’s 7th move.

561
 
9...Qe6
This is a simple route to an easily equal position.
 
The maximalist try is:
9...Kd8!?
If you want to press for an advantage, this is the way to do it. However, you should know that
the ensuing positions are complicated and a good deal of precision is necessary.
10.0-0 Re8 11.Ne5
White has to go for this, as moving the queen from e2 will waste more time.
11...g5! 12.Qh5N
12.g3? Nc6–+ was simply winning for Black in Schuerman – Beukema, Brasschaat 2014.
The text move is the best try for White, moving the queen away from the e-file so that the
knight is no longer pinned, while also attacking f7.

12...Bxe5 13.dxe5 Rxe5 14.g3

Now the variations become a bit computerish!
14...Nc6 15.gxf4 gxf4 16.Qd1†!?
16.Qxf5 Rxf5 17.Bxf4 Be6³ gives Black chances to convert his extra pawn.
The text move is trickier: White prefers to keep the queens on the board as Black’s king is still
in the centre.
16...Ke8 17.Bxf4

562
17...Qg6†! 18.Kh1 Bg4!
Bringing the pieces into the game with tempo.
19.Qd2 Rd8 20.Qg2 Re6 21.Bxc7
The direct 21.Nd5? is bad due to 21...Rxd5! 22.Qxd5 Bf5.
21...Rd7µ
Black emerges with a clear advantage, as his king is safe enough on e8, surrounded by his own
pieces.
 
10.d5 Qxe2† 11.Nxe2
Now in Grischuk – He. Gretarsson, Reykjavik 2000, Black’s simplest option would have been:

563
11...Bg4N 12.Bxf4 Nd7=
Black has no problems in the endgame.

E32) 6.0-0 Be6

 
Some other moves have been played but I believe this to be the most accurate.
 
7.Qe2
I considered two other options:
 
7.d4 Ne3 8.Bb5† (8.Bxe3? is of course bad because Black grabs not the e3-bishop, but the other one
with 8...Bxc4µ followed by picking up the exchange, as in Czestochowski – Kocher, corr. 2016)
8...c6 9.Bxe3 fxe3³ Schwarhofer – Alvir, Austria 2011. Even if White manages to win the e3-pawn,
Black will have the bishop pair which will guarantee him at least a slight edge.
 
7.Bb3 c5!
Black controls the d4-square before playing ...Nc6.
8.d4 cxd4 9.Nxd4 Bc5 10.Kh1 Bxd4 11.Qxd4 0-0

12.Qf2!
White’s other options are slightly less precise:
a) 12.Bxd5 Nc6 13.Qxf4 (13.Qf2 Qxd5 14.Nc3 Qc4 15.Bxf4 Rfe8 16.Qg3 Ne7 was similarly
pleasant for Black in Firsching – Wegelin, email 2012) 13...Qxd5 14.Nc3 Qc4 15.Be3 Qxf4
16.Rxf4 Rfd8 Black had slightly the more pleasant side of equality in Plenca – Kuljasevic, Mali
Losinj 2018.

564
 b) 12.Bxf4 Nc6 13.Qf2 Nxf4 14.Qxf4 occurred in Gabrielian – Schroeder, Moscow 2015. This
would have been a good time for Black to simplify with 14...Bxb3 15.axb3 Qd4! 16.Nc3 Qxf4
17.Rxf4 Rfd8 when Black once again enjoys a miniscule edge in the endgame.
12...Nc6 13.Rd1 Qg5 14.Bxd5 Bxd5 15.Nc3 Be6 16.Bxf4 Qf5 17.Rd2 Rfd8 18.Rad1 Rxd2 19.Rxd2
Rd8 20.Rxd8† Nxd8=
The position was completely equal in Marchisotti – Kolpak, corr. 2016.

7...c6!
7...Be7 has been the most popular choice, when 8.d4 leads to an unbalanced game with chances for
both sides.
 
The text move is more precise: Black reinforces the knight and thus prepares to develop his bishop
on the more desirable d6-square, where it will defend the f4-pawn.
 
8.Nc3
I also checked:
8.d4 Bd6 9.Bb3 0-0 10.c4N
This is better than 10.c3? as played in Toelg – Mierzwa, Stuttgart 2015, when 10...Re8N would
have left Black clearly on top.
10...Nf6
 

565
11.c5!?
White should try to create some play immediately, otherwise Black will consolidate easily.
11...Bxb3 12.cxd6 Re8 13.Qd3 Bd5
The bishop is excellently placed here.
14.Bxf4 Nh5
Black is poised to win the d6-pawn. White should try to create some complications with:
15.Be5! f6 16.Nc3!
Black needs a precise reply:

16...Nd7!
16...fxe5? would be a mistake, as 17.Nxd5 cxd5 18.Qb5!© brings White powerful
compensation for the piece.
566
17.Nxd5 cxd5 18.Qb3 Nb6 19.g4 fxe5 20.gxh5 e4 21.Ne5 Qxd6³
Black is a pawn up and in control on the kingside, though the strong knight on e5 offers White a
little bit of play.
 
8...Bd6
This is consistent with Black’s previous move, although it allows White to take twice on d5,
leading to simplifications and eventual equality.
 
8...Be7!? is an interesting way to keep more complexity in the position. A correspondence game
continued: 9.d4 g5! Black should try to defend the f4-pawn if he is to hope for an advantage. (9...0-0
10.Nxd5 cxd5 11.Bd3 is just equal, as the f4-pawn will soon fall) 10.Bd2 Nd7 11.Rae1 h6 12.Nxd5
cxd5 13.Bd3

White’s pieces are all taking part in the game while Black still needs to find shelter for his king.
After 13...Qb6 14.Bf5! Rc8 15.g3 fxg3 16.Ba5!÷ the game exploded into complications in
Nightingale – Van Unen, corr. 2013.
 
9.d4?!
The untested 9.Nxd5N is White’s best, and after 9...cxd5 10.Bxd5 Bc5† 11.d4 Qxd5 12.Qb5† Nd7
13.dxc5 a6 14.Qb4 Qxc5† 15.Qxc5 Nxc5 16.Bxf4= we reach an equal endgame.
 
9...0-0 10.Ne4
10.Nxd5 cxd5 11.Bb3 Nc6 left White a pawn down for not much in Steinbruegge – Walter,
Willingen 2004.
 

567
10...Bc7N
Black retreated the bishop to e7 in Tsydypov – Badmatsyrenov, Ulan Ude 2014, but it is better to
keep the f4-pawn defended. Given the chance, Black will follow up with ...Nd7 with an ideal
position, so White’s next move is clearly the best try.
 
11.Nc5 Bg4! 12.h3
12.Nxb7?! Qf6 13.c3 Nd7ƒ gives Black a strong initiative, with ...Rae8 coming next.
 
12...Re8 13.Qf2 Bc8 14.Ne5 g5!
Black is ready to kick the knight away from the centre with ...f6, and he stands at least slightly
better.
 

E4) 5.Bb5†

568
5...c6!
Once again, Black should have no hesitation in trading off White’s pawn weakness in order to
speed up development.
 
6.dxc6 Nxc6 7.d4
7.0-0?? lost a piece to 7...Qb6†–+ in Weber – Bergmann, Frankfurt 2018.
 
7.Ne5? is also bad, as it wastes time and helps Black to complete his development: 7...Qb6 8.Qe2
Be6! Preparing long castling. 9.c3 0-0-0 10.Bxc6 (10.d4? Nxe5–+) 10...bxc6 11.d4 g5!µ Black was
clearly on top in Richter – Denaro, corr. 1990, his next objective being to break up the centre with
...c5.
 
7...Bd6
We have reached a final split. Sometimes White gets tempted to force the play with E41) 8.Qe2†?!,
but he would be better off settling for simple development with E42) 8.0-0.
 

E41) 8.Qe2†?!

569
 This is an attempt to create disharmony in Black’s camp, but it backfires after precise play.
 
8...Be6 9.Ng5
This is the logical sequel to White’s previous move.
9.c4? is a waste of time. Black simply plays 9...0-0, and if 10.d5 Re8! White is already lost.
 
9...0-0 10.Nxe6
Instead of the simple recapture, Black has the strong resource:
 
10...Qb6!
Black refuses to take the knight with the pawn, but rather prepares to bring a rook to e8, exploiting
the alignment of White’s queen and king on the open e-file.

570
11.Bxc6
11.Nxf8? is met by 11...Nxd4–+ followed by capturing the b5-bishop, as in Nocci – Meifert, corr.
2008. The poor knight on f8 is trapped and will be captured next.
 
11...bxc6 12.Nxg7
White has tried various alternatives, but he is in trouble in all lines.
 
12.Nxf8 is met by 12...Re8, going after the queen: 13.Qxe8 Nxe8 14.c3 Kxf8–+ With ...c5 coming
next, Black was winning in Groszpeter – Hazai, Budapest 1980.
 
12.0-0 Rae8 13.Qd3N (13.Nc3 Rxe6 14.Qc4 occurred in Sandomirsky – Sfyrney, corr. 1982, when
defending the f4-pawn with the simple 14...g5N would have given Black an almost decisive
advantage) 13...Rxe6µ Black’s pieces are much more active than their counterparts, many of which
are still on their initial squares.
 
12.Nd2 Rae8 13.Nc4 Qb8 14.0-0 runs into a nice trick:

571
14...f3! Opening the b8-h2 diagonal. 15.Qxf3 Bxh2† 16.Kh1 Rxe6 17.Bf4 Bxf4 18.Qxf4 Now in
Mimbela Lopez – Guzman Moneo, Aragon 2014, the most precise would have been 18...Qb5!N
19.b3 Nh5–+ followed by ...Rg6 or ...Qg5.

12...Rae8! 13.Nxe8 Rxe8 14.Qxe8† Nxe8 15.c3

Even worse is 15.0-0?N 15...Qxd4† 16.Kh1 f3!–+ when White’s king find himself alone on the
kingside.
 
15...c5–+
White’s problems were insurmountable in Mari Ooriad – Lorparizangeneh, Sari 2017. Materially
he is okay, but his exposed king and undeveloped queenside pieces make all the difference.

572
 

E42) 8.0-0 0-0

9.c3
White solidly supports the d4-pawn and will follow up by developing his pieces. Three alternatives
are worth considering:
 
9.c4?! is too ambitious. White should generally avoid placing his pawns on d4 and c4 in these
positions, as the d4-pawn will be inadequately defended. 9...Bg4 10.Nc3 Bc7!? 11.Bxc6?! (White
should try 11.Kh1N 11...a6 12.Bxc6 bxc6 although Black is certainly better here too, due to the
simple plan of putting pressure on the d4-pawn with ...Qd7 and ...Rad8) 11...bxc6 12.Ne2 Nh5µ
Black’s bishop pair and space on the kingside gave him great play in Rrhioua – Upton, Gibraltar
2004.
 
9.Nc3?!
The main idea of this move is to transfer the knight to e2 at some point, attacking the f4-pawn.
However, it just doesn’t work out the way White would like.
9...Bg4 10.Bxc6
White has to exchange on c6 sooner or later.
10.Ne2 can be simply met by: 10...Qc7 11.c4 Rad8! Whenever White places his pawns on d4
and c4, this rook usually belongs on d8 where it puts pressure on the d-pawn. 12.Bxc6 White
rules out any ...Nxd4 tricks. 12...bxc6 13.Qc2 c5!? 14.Kh1 cxd4µ The opening of the position
favoured Black’s bishop pair in Drlje – Trbojevic, Pula 1999.
10...bxc6 11.Qd3 Re8 12.Bd2 Qb6

573
13.Na4
Other moves are no better, for instance:
a) Going after the f4-pawn immediately with 13.Ng5? is a mistake. In Zlatic – Lazic, Zlatibor
2006, the simple 13...Rad8!N would have refuted White’s play. For instance, 14.Bxf4 (or
14.Na4 Qb5 15.Qxb5 cxb5 16.Nc3 b4–+) 14...Bc5! 15.Kh1 Bxd4–+ The two bishops are too
strong in such an open position.
b) 13.b3N 13...c5 gives Black a nice initiative, for example: 14.Na4 Qc6 15.dxc5 Bxc5†
16.Nxc5 Qxc5† 17.Qd4! (17.Kh1? Ne4µ) 17...Qxc2³ Black’s pieces are actively placed and
much better coordinated.
13...Qc7 14.Rae1 Rad8³
White was under some pressure in Van Hul – Wantiez, Belgium 1997.
 
9.Nbd2
This is similar to the main line.
9...Bg4 10.Nc4
10.c3 transposes to the main line.
10...Bc7
Releasing the tension with 10...Bxf3? proved to be premature after 11.Rxf3 Bc5 12.Bxc6
Qxd4† 13.Qxd4 Bxd4† 14.Kf1 bxc6 15.Bxf4² in Renet – Van der Sterren, Budel 1987.
11.Bxc6
11.c3 Re8 transposes to the position after 11.Nc4 Bc7 in the note to White’s 11th move in the
main line below.
11...bxc6 12.Qd3
12.Nce5 occurred in Formanek – Mulyar, Connecticut 2007, when Black should reply:
12...Bxf3!N Black gives up one of his strong bishops in order to secure a strong outpost on e4

574
 for his knight. 13.Nxf3 Ne4 14.Qd3 Re8 15.b3 g5 16.Bb2 f5µ Black has a super-strong knight
on e4 and attacking chances on the kingside.
12...Re8 13.Nfe5!
13.b3?! is well met by 13...Bxf3 14.Rxf3 as in E. Pedersen – Pordzik, corr. 1995, and now
14...Re4! 15.Bb2 Qe7µ would have seen Black make good use of the same e4-outpost, but this
time with his rook!
The text move is a better try, but Black can counter it with a nice tactic.

13...Bf5! 14.Qxf5N
14.Qd2? is bad due to 14...c5!, opening up the position so that the bishops will become more
effective. 15.Rxf4 Be6 16.Ne3 cxd4 17.Rxd4 Qb8 The powerful bishops and White’s lack of
development on the queenside amounted to an almost decisive advantage for Black in Perez –
Foy, email 2002.
14...Qxd4† 15.Kh1 Bxe5 16.Nxe5 Rxe5 17.Qxf4 Re4 18.Qf3 Rae8³
Black’s active pieces guarantee him at least a slight edge.

575
9...Bg4
This is the normal square for the bishop, where it pins the knight and stands actively on the
kingside.
 
10.Nbd2 Re8
The rook will be useful on the open file.
 
11.Bd3
The bishop was not doing anything on b5, so it makes sense for White to bring it to a central square
where it points towards the kingside. A possible plan for White is Re1 followed by Ne4, exchanging
a few pieces.
 
11.Re1 Qd7 12.Bd3 was played in Schulz – Miller, Bad Kissingen 2018, when 12...Rxe1†N 13.Qxe1
Re8 14.Qf1 h6³ followed by ...g5 would have left Black with a useful space advantage on the
kingside.
11.Nc4 Bc7 12.Bd2
Trying to escape the pin with 12.Qd3?! is well met by the active 12...Qd5!µ.
12.Nce5N may seem logical but the forcing sequence of 12...Bxe5 13.Bxc6 bxc6 14.dxe5
Qxd1 15.Rxd1 Bxf3 16.gxf3 Nd5! soon enables Black to win a pawn: 17.c4 Ne3 18.Bxe3 fxe3
19.f4 g5!µ White is about to lose one of his pawns, and Black will have good chances to
convert his advantage in the double-rook endgame.

576
12...Ne4 13.Be1 Qd5 14.Na3
We have been following Ginzburg – Zarnicki, Villa Martelli 2002. All the black pieces are
actively placed, and now he should have transferred his queen to the kingside with:
14...Qh5!N
With unpleasant threats such as ...Ng5 and ...Re6. White’s best defence is:

15.Bc4!
Stopping ...Re6 and also creating tactical defensive ideas based on the f7-point.
15.Be2? Ng3!–+ is crushing.
15.h3? Bxh3 16.gxh3 Qxh3–+ leaves White without a good answer to ...Re6.
15...Ng5 16.Nxg5 Bxd1 17.Bxf7† Qxf7 18.Nxf7 Be2µ
Black wins an exchange, with great chances to win the ensuing endgame.
577
11...Qd7!
Black connects his rooks and keeps various options open, including doubling on the e-file and
gaining space on the kingside with ...h6 and ...g5.
 
11...h6 enabled White to carry out his plan with 12.Re1 g5 13.h3 Bh5 14.Ne4! when the position was
about equal in C. Sadler – Brooks, corr. 2014.
 
12.Qc2!
Persisting with the idea of Ne4.
 
12.Re1 transposes to 11.Re1 Qd7 12.Bd3 in the note to White’s 11th move above.
 
12.Nc4 Bc7 13.a4 Rad8 14.Bd2 is another logical continuation, when I like the typical plan of
gaining space on the kingside with:

578
14...h6!N (14...g6 has been played in a few correspondence games but I prefer the text) 15.Qc2 g5
16.Rae1 Kg7³ Black keeps a slight edge due to his space on the kingside.

12...Rac8 13.Qb1
Another possibility is 13.Nc4 Bb8 14.Nfe5 in order to simplify. However, after 14...Nxe5 15.Nxe5
Bxe5 16.dxe5 Rxe5 17.Bxf4 Rd5 18.Be2 Bxe2 19.Qxe2 Re8³ Black kept a slight edge due to his
control over the open files in Volotovsky – Hoeglauer, Pardubice 2017.
 
13...h6!N
13...g6 was played in Ketelhöhn – Harabor, corr. 2018, but once again I prefer a more ambitious
space-gaining operation.

579
 
14.Ne4 Nxe4 15.Bxe4 g5
I believe Black keeps a slight edge in all variations. For instance:

16.Bd2 Ne7
The knight was not doing much on c6, so it drops back and prepares ...Bf5, exchanging the more
active of White’s bishops.
 
17.Qc2 Bf5 18.Rae1 Kg7 19.Qd3
Preparing c3-c4.
White cannot play 19.c4? immediately due to 19...Bxe4 20.Rxe4 (20.Qxe4 Rxc4–+) 20...Qf5!,
pinning the rook and threatening ...Rxc4. My analysis continues:

580
21.Rfe1 Nd5! Exploiting another pin, this time on the c4-pawn. Now Black threatens ...Nf6. 22.Qd3
Rxe4 23.Rxe4 Rxc4–+ Black is a healthy pawn up, with a clear positional advantage as well.

19...b5 20.b3
White is all set to advance his c-pawn, but Black maintains his slight initiative with:
 
20...Bxe4 21.Rxe4 f5! 22.Re2 Nd5
With the idea of ...Nf6-e4, establishing the knight on its perfect outpost.
 
23.Rxe8

581
 23.Ne5? is bad due to 23...Bxe5 24.dxe5 Qe6 when Black controls the light squares and the bishop
on d2 is dominated.
 
23...Rxe8

24.Re1 Rxe1† 25.Bxe1 Nf6³

Despite the simplifications, Black keeps a slight edge due to the excellent outpost on e4.

Conclusion
 
The King’s Gambit is not so common these days, but you can still expect to encounter it from time to
time. After 1.e4 e5 2.f4, I recommend 2...d5 3.exd5 exf4 as a good way of obtaining a thematic pawn
structure which offers good play for Black. White usually responds with 4.Nf3 to prevent the check
on h4, when 4...Nf6 is our reply. Given the chance, Black will follow up with ...Nxd5, obtaining a
good game. If ever White defends the d5-pawn with c2-c4, we should have no hesitation in playing
...c6 to open lines for our pieces. If White opts for 5.Bb5†, we once again play 5...c6! 6.dxc6 Nxc6
with quick development and active piece play. The King’s Gambit is an aggressive opening favoured
by attacking players, so it’s an especially nice feature of our chosen set-up that Black often ends up
with the safer king.

582
A) 3.g3 288
B) 3.f4 d5! 4.fxe5 Nxe4 290
B1) 5.Qf3 290
B2) 5.d3 292
B3) 5.Nf3 294
C) 3.Bc4 Nxe4 4.Qh5 Nd6 5.Bb3 Be7 295
C1) 6.Nf3 297
C2) 6.Qxe5 298

 
1.e4 e5 2.Nc3
The Vienna Game is characterized by the development of this knight followed by something other
than a quick Nf3.
 
2...Nf6
This is generally the preferred move, and it also feels like the natural choice for a Petroff player.
White has three main ways to interpret the Vienna, all leading to completely different types of
positions: A) 3.g3, B) 3.f4 and C) 3.Bc4.
3.Nf3 leads back to variation B of Chapter 1.

583
 

A) 3.g3
 
I recommend meeting this with an active response in the centre:
 
3...d5 4.exd5 Nxd5 5.Bg2 Nxc3 6.bxc3 Bc5
The bishop is well placed on the a7-g1 diagonal, keeping an eye on key squares like d4 and f2. The
bishop can also be useful in blocking White’s play along the b-file after ...Bb6.
 
7.Nf3
This is the most active way of developing the knight, hitting the e5-pawn and possibly going via
g5-e4 later.
 
The alternative is:
7.Ne2 Nc6 8.0-0 0-0 9.d3
And now an important detail is:

9...Bg4!
The idea behind this is to provoke h2-h3, when the bishop will drop back to e6 and the useful
move ...Qd7 will come with gain of tempo.
10.h3 Be6 11.Be3
Another nice example continued 11.c4 Qd7 12.Kh2 Rad8 13.Be3 Bxe3 14.fxe3 Ne7 15.Nc3
f5³ when Black enjoyed the better pawn structure and controlled plenty of space on the
kingside in Liu Qingnan – Harikrishna, Beijing 2012.
11...Bxe3
584
 This is the typical response when the bishop arrives on e3, just as Harikrishna showed in the
example above.
12.fxe3 Qd7 13.Kh2 b6³
Black had the better game in Milde – Ellis, corr. 2019, with ...f5 coming sooner or later.

7...Nc6 8.0-0 0-0 9.Re1

White usually attacks the e5-pawn immediately.
 
9.d3 can be met by 9...h6, when 10.Re1 Qf6 transposes to the main line, and 10.Rb1 Bb6 does not
change much.
 
9...Qf6!
This is the best way to defend the pawn.
 
9...Re8 10.d4! exd4 11.Rxe8† Qxe8 12.cxd4² gave White good chances in Adhiban – Seel, Berlin
(rapid) 2015.
 
9...Bg4 was initially going to be my recommendation but, after a deeper examination, I was not quite
satisfied with Black’s position if White plays precisely: 10.h3 Bh5 11.g4 Bg6 12.Nxe5! Nxe5 13.d4!
(but not 13.Rxe5?! Bd6ƒ followed by ...f5 with a promising initiative on the kingside) 13...f5! Black
should waste no time in creating counterplay on the kingside. 14.g5 f4 15.dxc5 Although Black’s
knight is actively placed, the bishop is the more valuable piece which should be eliminated.
15...Qxg5 Now White should force the exchange of queens with: 16.Qd5† Nf7 17.Qxg5 Nxg5
18.Rb1 Rae8 19.Rxe8 Rxe8 20.Bxf4 Ne6 Even though the endgame should ultimately be a draw,
White’s bishop pair and (admittedly doubled) extra pawn gave him some chances to press in Howell
– Matlakov, Gibraltar 2017.
 

585
10.d3 h6
Covering the g5-square.
 
11.Be3
11.Rb1 is met by 11...Bb6, blocking the open file. A good example continued: 12.Be3 Bg4!? Just
as in the note to move 7 above, Black provokes h2-h3. Here the queen cannot go to d7 to gain a
tempo, but it is still worth provoking a slight loosening of White’s kingside. 13.h3 Be6 14.c4 White
would like to manoeuvre his knight to e4 via d2, and in general his plans include putting pressure on
the b-file. Therefore Black opens up the centre with:

14...e4! 15.Nd2 exd3 16.c5 (16.cxd3?! Rad8³ gives Black easy play against the weak d-pawn)
16...Ba5 17.Rxb7 Nb4= Black had a fine position in Nakamura – Xiong, Saint Louis 2018.

586
11...Bb6 12.Nd2
12.Rb1 transposes to the previous note.
 
12...Re8
The rook is useful here, overprotecting the e5-pawn and being ready for a possible opening of the
e-file.
 
13.Rb1
13.Nb3 Bf5 14.c4 threatens c4-c5, so Black should trade bishops. 14...Bxe3 15.Rxe3 Now in Glek
– Tkachiev, Bastia 2002, a simple route to a good game would have been:

587
15...Rad8N Developing the last piece. Play may continue: 16.Qe2 Intending not just Re1 but also f2-
f4. 16...Bd7 Supporting the e8-rook. 17.Re1 (if 17.c3 then 17...b6= covers the c5-square, maintaining
the balance) 17...Nd4= Black has no problems.

13...Bf5 14.c4 Rad8 15.Ne4

15.c5N 15...Ba5 16.Rxb7 is strongly met by 16...e4! 17.dxe4 Be6© when Black has amazing
compensation for the sacrificed pawns.
 
15...Qg6
Black had a fine position in Lu Shanglei – Harikrishna, Beijing 2012.
 

B) 3.f4

588
 White plays in the style of the King’s Gambit.
 
3...d5!
Black should strike in the centre too!
 
3...exf4? is bad due to the obvious 4.e5.
 
4.fxe5
4.exd5 gives Black a pleasant choice: 4...exf4 is simplest, reaching something similar to our King’s
Gambit coverage in the previous chapter. Alternatively, 4...Nxd5 is a good move, and Black could
even consider 4...e4!?, reaching a version of the Falkbeer Countergambit (1.e4 e5 2.f4 d5 3.exd5 e4)
where White has opted for 4.Nc3 instead of the more critical 4.d3.
 
4...Nxe4
I have analysed White’s three main options: B1) 5.Qf3, B2) 5.d3 and B3) 5.Nf3.
 

B1) 5.Qf3
 
I believe Black’s most accurate option to be:
 
5...Nc6!?
By playing this move before exchanging on c3, we forces the white bishop to go to b5.

589
6.Bb5
6.Nxe4?! is bad due to 6...Nd4! 7.Qc3 dxe4 8.Ne2 Nc6³ as seen in Tzermiadianos – A.
Mastrovasilis, Athens 2002. White’s pieces are misplaced and the e5-pawn can come under pressure.
 
6...Nxc3 7.dxc3 Qh4†
Provoking a slight weakening of White’s kingside.
 
8.g3 Qe7!?
Black was fine in the endgame after 8...Qe4† 9.Qxe4 dxe4= in Nakamura – Yermolinsky,
Stillwater 2007, but I find it more interesting to keep the queens on the board.

590
9.Bf4!?N
Preparing long castling seems like the most natural plan to me.
 
I only found one game from the above position, which continued: 9.Ne2 Bd7 (9...a6!?N 10.Bxc6†
bxc6 leads to double-edged play) 10.Bxc6 Bxc6 11.Nd4

11...Qxe5†N (11...Bd7 12.0-0 0-0-0 13.Qxf7 Qxe5 14.Bf4 Qe7 was playable for Black in
Heinemann – Krueger, Germany 2019, but the text move seems simpler) 12.Kd1 0-0-0 13.Re1 Qd6
14.Qxf7 Qd7 15.Qxd7† Bxd7 The position is essentially equal but Black can try to make something
of his bishop pair.
 
9...Bd7 10.0-0-0 0-0-0
Black is doing fine, although it is worth mentioning an interesting tactical line:
 
11.Ne2 Nxe5 12.Qxd5!? c6 13.Qd4 cxb5 14.Qxa7
The sharp line just leads to an early perpetual after:
 
14...f6 15.Qa8† Kc7 16.Qa5† Kc8=
 

B2) 5.d3

591
 This best formula against this move has been known for a long time.
 
5...Nxc3
The tempting 5...Qh4†?! is not so good due to 6.g3 Nxg3 7.Nf3 Qh5 8.Nxd5² with messy
complications which favour White.
 
6.bxc3 d4!
Black takes the opportunity to attack his opponent’s weaknesses while preventing White from
setting up his ideal pawn structure with d3-d4.
 
7.Nf3 Nc6
7...dxc3 is also pretty good although 8.d4 gives White a certain amount of activity for the pawn, so
I prefer to maintain the tension for the moment.
 
8.Be2
The alternative is:
8.cxd4 Bb4†!
Black exchanges the dark-squared bishops, which weakens the dark squares in White’s camp,
and also helps Black to castle more quickly.
Black should not recapture the pawn immediately with 8...Nxd4?! due to 9.c3 Nxf3† 10.Qxf3²
as seen in Nava – Demchenko, Ho Chi Minh City 2019. White will follow up with d3-d4
followed by Bd3 (or Bc4) and 0-0, with a strong centre and some pressure on the kingside.

592
9.Bd2 Bxd2† 10.Qxd2 Nxd4 11.c3 Nxf3† 12.gxf3 0-0³
White’s king was exposed and his pawn structure overextended in Hering – Yloenen, corr. 2013.
 
The text move has been White’s most popular choice, avoiding any reaction in the centre in favour of
developing. I believe this is the right time for Black to take a greedy approach with:

8...dxc3!
Surprisingly, this natural capture has not been all that popular. In most games Black has been
focused on preparing ...0-0, but after analysing this variation I have concluded that Black should
instead aim for long castling and seizing the initiative for himself!
 

593
9.0-0 Qd5!
Black centralizes the queen, putting pressure on the e5-pawn and already hinting at ...0-0-0.
 
10.d4
This seems the most logical. White guards the e5-pawn and hopes to win the c3-pawn somehow.
The drawback is that the d4-pawn becomes a target.
 
10.Bf4 is another way to protect the e5-pawn, which was seen in Vedral – Masek, Czech Republic
2018. I recommend 10...h6!N, covering the g5-square in order to prepare ...Be6 without allowing any
Ng5 ideas. Play may continue: 11.Qe1 Qa5

Black stands clearly better: he keeps an extra pawn and is ready for ...Be6 followed by ...0-0-0. In
the event of 12.Rb1 Black can play 12...Be6 anyway, as 13.Rxb7? is met by 13...Bc5† 14.Kh1 Bb6–+
and White’s rook is trapped.

594
10...Be6
Continuing with the plan.
 
11.Be3 0-0-0
Now Black has a positional threat of ...Ba3-b2, when the bishop on b2 will support the c3-pawn,
block the b-file and win the a2-pawn, as the rook will have to move away.
 
12.Qd3
12.a3 stops Black’s plan but now the a1-rook will have to keep watch over the weak pawn. 12...h6!
is a good reply, not only covering the g5-square but also intending ...g5-g4 to undermine the d4-
pawn. A good example continued:

595
13.Qe1 g5 14.h3 Qa5 15.Qf2 Ne7!µ With an extra pawn and the knight heading for the perfect g5-
square, Black was well and truly in control in Tedesco – Zielasko, corr. 2014.
 
12...Qa5 13.Rfb1

13...a6!
Covering the b5-square.
 
14.a4
The time has come to break down White’s central structure with:
 

596
14...f6!µ
White’s compensation for the pawn was highly doubtful in Sisak – Abolins-Abols, corr. 2017.
 

B3) 5.Nf3

5...Be7
Initially I wanted to recommend 5...Bc5. However, after analysing it deeply, I realized the engine
overestimates Black’s chances in the resulting position. The main line of my analysis goes: 6.d4 Bb4
7.Bd2 c5 8.Bd3 Nxd2 9.Qxd2 Nc6 Even though Black has the bishop pair and seems to be putting a
lot of pressure on the enemy centre, things are not so clear after 10.a3 Ba5 and now the pawn
sacrifice 11.b4! cxb4 12.Nb5© gave White strong compensation in Hoang – Cheparinov, Ho Chi
Minh City 2019.
 
6.Qe2
This is the usual move, aiming for the dxc3 pawn structure.
 
6.d4 c5 gives Black easy play, as does 6.d3 Nxc3 7.bxc3 c5.
 
6...f5!?
6...Nxc3 is the main line and is covered very well from Black’s perspective in Playing 1.e4 e5 by
Ntirlis.
 
The text move is an interesting alternative: Black waits for White to force the knight’s exchange with
d2-d3, leading to a different pawn structure.

597
 
7.d3
7.exf6?! Nxf6ƒ leaves White behind in development.
 
7...Nxc3 8.bxc3 0-0
Black has a simple plan of ...c5, ...Nc6 and ...Be6. White’s queenside pawn structure contains some
weaknesses, which Black hopes to exploit later.
 
9.g3
White prepares to develop his light-squared bishop while preventing ...f4.
9.h4?! f4³ leaves the bishop locked on c1.
 
9...c5 10.Bg2 Nc6 11.h4
I also considered 11.0-0 Be6 when Black has achieved the optimal placement of his pieces and is
ready to start creating play on the queenside. 12.d4 Qa5 13.Bd2 Rac8³ Black’s queenside pressure
gave him the better chances in Rydholm – Duliba, corr. 2012. It is important not to close the
queenside with ...c4, but rather just maintain the tension.

11...h6
Covering the g5-square. Another correspondence game continued:
 
12.0-0 Be6
Again, Black completed development in the ideal way and is ready to focus on the queenside.
 
13.Rb1 Rb8 14.Nd2 b5 15.Nb3 c4 16.Nd4 Nxd4 17.cxd4 c3!

598
 Creating a pawn majority on the queenside.
 
18.Qh5 Qe8 19.Qxe8 Rfxe8³
Black’s superior queenside structure gave him the better endgame in Pichushkin – Abramov, corr.
2013.
 

C) 3.Bc4

Finally we come to this simple developing move, which Black can meet with a standard ‘fork
trick’.
 
3...Nxe4 4.Qh5
This is by far the main move. Alternatives are weaker:
 
4.Nf3?! transposes to variation A of Chapter 1.
 
4.Bxf7†?! Kxf7 5.Nxe4 d5 6.Ng3
6.Qf3† Kg8 7.Ng5 is a flashy show of aggression, which can easily be rebuffed by 7...Qd7!µ
followed by ...h6, as in Gal – Eismont, Gyula 1993.
6.Qh5† only helps the black king to go to a safe square. 6...Kg8 7.Ng3 Nc6 8.d3 g6 9.Qd1
Now in Lin Chen – Shen Yang, Hefei 2010, the best move was 9...h6!N, not only covering the
g5-square but also planning to activate the rook: 10.N1e2 Rh7 11.0-0 Rf7³ Black is clearly on
top.
6...Nc6

599
 Black has a strong centre and the bishop pair, and it will just take a few moves to regroup and
get his king to safety.
7.d3 Be7 8.Nf3 Rf8 9.0-0 Kg8³
Having castled artificially, Black was ready to enjoy his positional advantages in Ni Hua –
Vladimirov, Tripoli 2004.
 
4.Nxe4?! d5 5.Bd3 dxe4 6.Bxe4
Black already has a slight edge in central control, and now he increases it with:
6...f5!
Compared to the more popular ‘fork trick’ variation where the moves Nf3 and ...Nc6 have
already been played, it is obvious that Black benefits from being able to chase the bishop
without having to worry about Bxc6†.
7.Bf3
7.Bd3 Be6 8.Qe2 should be met by 8...Qf6! preparing long castling. 9.Bc4 Nc6 10.Bxe6 Qxe6
Now in Moreno Garrido – Navarro Perez, Barcelona 2001, White had to play 11.Nf3N to
prevent ...Nd4, nevertheless, after 11...0-0-0µ Black is clearly ahead in development and piece
activity, as well as enjoying a space advantage.
7...e4 8.Be2 Nc6
Having gained space in the centre, Black now focuses on completing development.
9.d3 Bd6 10.Nh3 0-0

11.dxe4 fxe4 12.Be3 Bxh3 13.gxh3 Kh8µ

Black was clearly better due to his extra space in the centre and White’s weakened kingside in
Safarli – Mamedyarov, Khanty-Mansiysk 2013.

600
4...Nd6 5.Bb3
5.Qxe5† Qe7 leads to a queenless middlegame which is completely harmless for Black. 6.Qxe7†
Bxe7 7.Bb3 Nf5! An important move. The knight was badly placed on d6 where it blocked the d-
pawn. Hence Black moves it in preparation for ...c6 followed by ...d5 – not only opening a path for
the c8-bishop, but also restricting White’s minor pieces on the queenside.

8.Nf3 (transferring the knight to e3 with 8.Nd5 is nothing special: Black simply sticks to her plan
and gets a comfortable position after 8...Bd8 9.Nf3 c6 10.Ne3 d5=) 8...c6 9.0-0 d5 10.Re1 0-0=
Black was doing fine in Harika – A. Muzychuk, Baku (ol) 2016.
 
5...Be7

601
 5...Nc6!? leads to a messy game after 6.Nb5! g6 7.Qf3 f5 8.Qd5 Qf6 9.Nxc7† Kd8 10.Nxa8 b6©
when Black has compensation for the exchange. This position has been debated for many decades
and it makes an interesting topic of analysis with modern engines, but such a sharp and theoretically
demanding variation is hardly the most practical choice when facing a relatively rare sideline.
 
The text move is the simple and solid choice, which White can meet with either C1) 6.Nf3 or C2)
6.Qxe5.
 

C1) 6.Nf3

Developing a piece and preparing to castle as well as intending to take on e5 with the knight.
 
6...Nc6
This is more precise than 6...0-0.
 
7.Nxe5
7.0-0? is not possible due to 7...g6µ.
 
7.Nb5? is refuted by 7...g6!, driving the queen away without making any positional concessions.
(7...0-0? enabled White to get compensation by weakening Black’s pawn structure with 8.Nxd6
cxd6© in Mamedyarov – Giri, Internet [bullet] 2018; note that 8...Bxd6? would be even worse due to
9.Ng5) 8.Nxd6† Bxd6 9.Qh6

602
9...Bf8 10.Qe3 Bg7 White is a pawn down for no compensation. Note that the attempt to regain the
pawn with 11.d4 doesn’t work due to 11...0-0 12.dxe5 Nxe5 13.Nxe5 Bxe5µ.
 
7...g6!
7...Nxe5?! is inferior, as after 8.Qxe5 0-0 9.0-0² Black will have to waste time relocating the d6-
knight.
 
The text move is an important detail which ensures Black of at least equal chances. White must either
exchange on c6 or return with his queen to e2, and both options give Black something to cheer for.

8.Nxc6

603
 This allows Black to open a path for the c8-bishop.
 
8.Qe2 is well met by 8...Nd4! – Black should be more interested in exchanging the strong b3-bishop
than the knight on e5. 9.Qd3 Nxb3 10.axb3 0-0 11.0-0 Bf6 Kicking the knight from the centre.
12.Nf3 This position was reached in Shanava – Mchedlishvili, Tbilisi 2013, and here I suggest:

12...b6N Preparing to complete development, rather than putting too much emphasis on the bishop
pair. 13.Nd5 Bb7 14.Nxf6† Qxf6 15.Qc3 Kg7 The position is close to equal but slightly more
pleasant for Black.
 
8...dxc6 9.Qf3
9.Qd1 should be met by the typical idea of 9...Nf5, intending to centralize the knight on d4 without
giving White a chance to play d2-d4 himself. 10.a3 Nd4 11.Ba2 occurred in Raetsky – Ubilava,
Llucmajor 2017, and now it would have been best to play:

604
11...h5!N 12.h3 (12.0-0 is met by 12...Bg4 when White must weaken his kingside a bit: 13.f3 Be6
14.Bxe6 Nxe6³ Black intends ...Qd7 followed by long castling) 12...Be6 13.Bxe6 Nxe6³ Followed
by ...Qd7 and ...0-0-0 with the makings of a good initiative on the kingside.

9...0-0 10.d3
Now we should improve the knight in thematic style with:
 
10...Nf5! 11.0-0 Nd4 12.Qg3 a5ƒ
Black’s knight is extremely strong and he has a nice initiative on the queenside, Mista – Hera,
Ustron 2009.
 

605
 C2) 6.Qxe5

White regains the pawn but will have to lose additional time with his queen.
 
6...0-0 7.Nf3
7.Nge2 Nc6 8.Qf4 has been played in a bunch of games. White’s idea is to keep the f3-square free
for the light-squared bishop, in case Black chases it with his knight. However, I think Black should
do it anyway!

8...Na5!?N The point is revealed after 9.Bd5 c6 10.Bf3 when 10...g5! is a strong move. The white
queen will be kicked around the board while Black uses the tempos gained to establish harmony

606
between his pieces. 11.Qd4 Nf5 12.Qd3 d5 13.Bg4 (13.0-0? is bad due to 13...Ng7!–+ threatening
both ...g4 and ...Bf5) 13...Nh4 14.Bxc8 Rxc8ƒ It is clear that Black has a wonderful initiative;
White’s queen is terribly misplaced and his queenside is still underdeveloped.
 
7.d4
This move should be met in a similar fashion:
7...Nc6 8.Qf4 Na5 9.Bd5
Safeguarding the bishop seems logical and principled, but it brings certain risks.
9.Nf3 is safer although 9...Nxb3 10.axb3 b6 gave Black pleasant equality in Ding Yixin – Tan
Zhongyi, Xinghua 2015.
9...c6 10.Bf3
Whenever White’s bishop retreats to f3 with the queen standing on f4, Black can always
consider pushing his pawns on the kingside in order to gain space and unsettle White’s pieces.

10...f5!N
10...Ne8 was less to-the-point in Skliarov – Sivuk, Lvov 2016.
11.Nge2 Nf7!³
Black is ready to play ...d5, with a preferable position.

607
7...Nc6
Black develops the knight with tempo, while also preparing the familiar plan of hunting down the
b3-bishop.
 
8.Qf4 Na5 9.0-0
9.d4 transposes to the position after 9.Nf3 in the note on 7.d4 above.
 
9...Nxb3 10.axb3
Black’s next task is to develop his light-squared bishop, so his next move makes perfect sense.
 
10...b6 11.Nd5

608
11...Re8!
This is more accurate than 11...Bb7, as now Black will be able to recapture on e7 with the rook.
 
12.Nxe7†
I also considered 12.d3 Bf8 13.Qg4, when the calm 13...f6! prevents White’s intended Bg5. After
the further 14.Bd2 Bb7= Black’s pieces were well placed and he had no problems in Forcen Esteban
– Rakhmanov, Linares 2018.
 
12...Rxe7 13.d3
We are following a correspondence game in which Black found an interesting manoeuvre.
 

609
13...Re6! 14.Nd4 Rf6=
Black is ready to push the knight back with ...c5 and also develop his bishop to b7. White had no
trace of an advantage in Gibas – Young, corr. 2018.
 

Conclusion
 
The Vienna Game is rarely seen at the top level, although several strong players use it occasionally as
a surprise weapon. After 1.e4 e5 2.Nc3 Nf6 we examined three main options. Starting with 3.g3 we
should take the opportunity to play 3...d5, leading to an active game with free piece play.
 
Next we looked at 3.f4, when 3...d5! 4.fxe5 Nxe4 is a reliable antidote. White has a few options, but
Black has a safe position and in most variations his superior pawn structure gives him excellent long-
term prospects.
 
Finally we considered 3.Bc4, when 3...Nxe4 4.Qh5 Nd6 5.Bb3 Be7 is a safe and reliable choice.
White regains his pawn easily enough, but the resulting positions are pleasant enough for Black.

610
A) 4.Qe2 302
A1) 4...Bc5 303
A2) 4...d5!? 303
B) 4.Nf3 d5 5.Bb3 Bb4† 305
B1) 6.c3 305
B2) 6.Bd2 Bxd2† 308
B21) 7.Nbxd2 308
B22) 7.Qxd2 0-0 309
B221) 8.exd5 310
B222) 8.Nxe5 313

 
1.e4 e5 2.Bc4
The Bishop’s Opening is a respectable opening which is sometimes used by fans of the Italian
Game (1.e4 e5 2.Nf3 Nc6 3.Bc4) as a transpositional device against Petroff players.
 
2...Nf6 3.d3
3.Nc3 would of course lead back to variation C of the previous Chapter 1.

611
 
3...c6
Black prepares to build a strong centre while gaining a tempo against the bishop.
 
We will analyse A) 4.Qe2 followed by the main line of B) 4.Nf3.
 
4.Bb3 d5 gives White nothing better than 5.Nf3, leading straight to variation B.
 
The only other option worth mentioning is:
4.Ne2
White intends to castle and play a quick f2-f4 at an appropriate moment.
4...d5 5.exd5 cxd5 6.Bb3 Be7 7.0-0 0-0 8.Bg5

8...Nc6
White’s idea is well illustrated after 8...h6?! 9.Bxf6 Bxf6 10.Nbc3 Be6 11.f4! exf4 12.Nxf4
when White gets strong pressure against the d5-pawn. After the further 12...Bxc3?! 13.bxc3
Nc6 14.Qf3 Ne7 15.Rae1 Qb6† 16.Qf2 Qd6 17.Qe2± White was much better in Khachiyan –
Ni Hua, Internet 2018.
9.Nbc3 Be6 10.d4
Now White doesn’t have 10.f4?N due to 10...Qb6† 11.Kh1 Ng4µ; and 10.Kh1? is answered by
10...d4!µ.
10...e4 11.Nf4

612
11...Qd6!N
11...Ne8 was less good in Sarmadi – Florentiades, Pervolia 2010.
11...Bf5!?N= is possible though.
12.Nxe6 fxe6=
Black has a good game, and White cannot break with 13.f3? immediately due to 13...Ng4! when
Black gains a decisive advantage.
 

A) 4.Qe2

613
 White ‘prevents’ ...d5 – but in fact Black can play it anyway!
We will look at two options: A1) 4...Bc5 is a simple and sound approach, whereas A2) 4...d5!? is
an interesting alternative if you want to play more dynamically.
 

A1) 4...Bc5
 
Black develops and prepares to castle.
 
5.Nf3 0-0
Now Black is ready for ...d5.
 
6.Bg5
6.Nxe5? is bad due to 6...d5 7.Bb3 Re8 when White has three pieces on the e-file, which is hardly
a good sign. 8.f4 Nbd7µ Kulczycki – Pereira, email 2000.

6...d6 7.Bb3 a5!

A typical move in this structure, gaining space on the queenside.
 
8.c3
Now in Sek – Yeletsky, Sochi 2017, it would have been logical for Black to continue his queenside
expansion before resuming other plans. Therefore I suggest:
 
8...a4N 9.Bc2 h6 10.Bh4
And now the most precise continuation is:

614
10...Qb6! 11.b4
11.Bxf6? loses material to 11...Qxb2 of course.
 
11...axb3 12.Bxb3 Nbd7³
Black’s better pawn structure on the queenside gives him a slight edge.
 

A2) 4...d5!?

615
 Although the previous line is perfectly decent, it is also quite tempting to sacrifice a pawn for the
sake of the initiative.
 
5.exd5
5.Bb3 is legal but, since the whole point of White’s 4th move was to discourage ...d5, the text
move is the only option we should concern ourselves with.
 
5...cxd5 6.Qxe5†
6.Bb5† Nc6 7.Qxe5† Be7 transposes to the main line below.
 
6...Be7 7.Bb5†
Inferior is:
7.Bb3?! 0-0 8.Ne2
8.Nf3 Bb4† 9.Kf1 Nc6 10.Qg5 h6 11.Qh4 occurred in Suchorukow – Bezgodov, Orel 1992,
when 11...a5!N would have given Black an excellent initiative for the pawn, as White’s king is
misplaced and he is behind in development.

8...Bb4†!N
8...Re8 led to a quick victory after 9.0-0?? (9.Qf4N was necessary, when 9...Nc6ƒ gives Black
a nice initiative for the pawn) 9...Bd6–+ and White lost the e2-knight in Ferrandi – Karpatchev,
Calvi 2006. The text move is even better though.
9.Kf1
9.c3 is no better in view of 9...Re8 10.Qf4 Bd6µ when, even if White somehow manages to
castle, he will be far behind in development.
9...Re8ƒ
White has lost the right to castle and is behind in development, giving Black a strong initiative.

616
7...Nc6 8.Nf3 0-0 9.Bxc6 bxc6 10.0-0 Bd6 11.Qe2
11.Qd4 was played in Ratsch – Westin, corr. 1984, and here I suggest 11...Bg4!N 12.Nbd2 c5
13.Qa4 Rb8© when Black’s pieces are active while White struggles with development.
 
11...Re8 12.Qd1
We have been following Hou Yifan – Salem, Moscow 2010. Black’s best would have been:

12...Bg4N
Developing actively while setting up an awkward pin. Play may continue:
 

617
13.h3 Bh5 14.Nbd2
And now a nice idea is:
 
14...Nd7!³
The knight aims to go to e5, creating more pressure on the f3-knight. White’s pieces are rather
passive and she has an underdeveloped queenside. Black has more than enough compensation for the
pawn, and can look to create play on the kingside.
 

B) 4.Nf3 d5 5.Bb3

5...Bb4†
The idea of this move is to provoke c2-c3, so that after ...Bd6 White will no longer be able to play
Nc3, putting pressure on d5.
 
5...Bd6 6.Nc3 is certainly playable for Black but personally I prefer to avoid this.
 
5...a5!? is another interesting option, provoking a2-a4 before going for the same idea as in our main
line. However, it has a small drawback: 6.a4 Bb4† 7.c3 Bd6 and now White can play 8.exd5 cxd5
when the b1-knight can later go via a3 to b5, exploiting the hole created by Black’s 5th move. It’s not
too drastic and Black’s position remains quite playable, but I would prefer not to give White that
opportunity.
 
White has two main options: B1) 6.c3 and B2) 6.Bd2.
 

618
 B1) 6.c3 Bd6 7.0-0
 
I believe this is the best move order for White, as it keeps his options open regarding how to defend
the e4-pawn next.
 
7.Qe2
This can hardly challenge Black.
7...0-0 8.Nbd2 Nbd7 9.Nf1
9.0-0 occurred in Kovalenko – Petrosyan, Internet 2018, when 9...b6!N intending ...Ba6 would
have been annoying for White.
The idea of the text move is to transfer the knight to g3 before castling, thus saving a move.
9...Nc5 10.Bc2 b6!

Thematically developing the bishop to a6, where it will be most annoying for White.
11.b3 dxe4 12.dxe4 Ba6 13.c4
This is practically forced, but now the d4-square is too weak.
13...Ne6 14.Ng3 Bb4† 15.Kf1 b5!µ
Black had clearly won the opening battle in Tiviakov – Lagno, Benidorm 2008.
 
7.Nbd2 0-0 8.0-0
8.Qe2 transposes to the line above.
8...Nbd7 9.Re1 Re8
Black continues developing his pieces to their optimal squares.
10.Nf1

619
10...h6
White may intend to play Bg5 at some point, so it makes sense to cover the g5-square before
doing anything else.
11.Ng3 Bf8 12.d4
This seems like White’s best attempt to make something happen, but Black is well equipped to
deal with it.
12...dxe4 13.Nxe5 Nxe5 14.dxe5 Qxd1 15.Rxd1
Now in Alekseev – Landa, Taganrog 2011, the most accurate would have been:

15...Rxe5N 16.Rd8
Presumably Landa did not like the look of this, but there is a simple solution:
16...Re8 17.Rxe8 Nxe8 18.Nxe4 Bf5=
620
 Black has no problems.
 
7.Bg5
This has been a popular choice: White pins the knight and puts pressure on the d5-pawn, thus
forcing Black to make a decision in the centre.
7...dxe4 8.dxe4 h6 9.Bh4
Instead of castling immediately, Black should focus on development with:

9...Qe7!
Supporting the d6-bishop and thus preparing ...Nbd7.
10.Nbd2 Nbd7 11.0-0
11.Bg3?! was a strange choice from the World Champion, and after 11...Bc7 12.0-0 Nh5³
Black was able to trade off the g3-bishop without weakening anything on the kingside in
Carlsen – Caruana, Saint Louis 2014.
11...Bc7
A useful prophylactic move, ensuring that Nc4 will not come with tempo – but White plays it
next anyway.
12.Nc4 0-0 13.Ne3 Nc5 14.Bc2

621
14...Rd8 15.Qe2 g6!=
This final accurate move prepares ...Kg7 to supporting the f6-knight, thus enabling the queen to
move away from e7. Black was doing fine in Hall – Semrl, corr. 2017.

7...0-0 8.Re1
8.Bg5 is met by 8...dxe4 9.dxe4 Na6, when the knight heads for its usual home on c5. 10.Nbd2
Bc7 Preventing Nc4. 11.Bh4 Nc5 12.Nc4 Nxb3 13.axb3 Qxd1 14.Raxd1 Be6= Black was fine in
Ponkratov – Khairullin, Kaliningrad 2015.
 
8...Nbd7 9.Bg5
White may also leave the bishop at home for the time being with:

622
9.Nbd2
Intending to open the centre with d3-d4. The most accurate reply is:
9...Qc7!
Preventing White’s plan.
9...Re8?! 10.d4! looks annoying for Black. For example, 10...exd4 11.cxd4 dxe4 12.Ng5² and
White obtained some pressure in Gattenloehner – Goreskul, Winterthur 2003.
10.Nf1?!
10.d4 doesn’t work due to 10...exd4 11.cxd4 dxe4³.
The text move appears natural but Black shows a strong reply against it in the game we are
following.
10...dxe4!
This exchange is well timed, now that the knight has moved away from the c4-square.
11.dxe4 Nc5 12.Bc2 Rd8 13.Ng3
Now in Duncker – Feicht, Giessen 1996, the strongest continuation was:

13...Bf8!N
Forcing the queen to e2, when it becomes a target for the bishop arriving on a6.
14.Qe2 h6
Covering the g5-square.
15.h3 a5!³
Another typical move, gaining space on the queenside. Black has an excellent position and a
natural plan of ...b6 followed by ...Ba6.

623
9...dxe4 10.dxe4 Bc7 11.Nbd2 Nc5 12.Bc2 a5
Black was comfortable in Hall – Kribben, corr. 2017.

B2) 6.Bd2

6...Bxd2†
After this natural move, both B21) 7.Nbxd2 and B22) 7.Qxd2 deserve attention.
 

624
 B21) 7.Nbxd2 a5!
 
This move is well timed, with the knight far away from the b5-square.
 
8.c3
8.a3 allows Black to gain more space on the queenside, which should be of some benefit. 8...Nbd7
9.0-0 a4 10.Ba2 0-0 11.Re1 Re8 Now Aronian went for:

12.d4!? Black should always be aware of this method of opening up the centre. Caruana shows a
good reply though: 12...exd4 13.exd5 Rxe1† 14.Qxe1 cxd5 15.Rb1 White must support the b2-pawn
before taking on d4, otherwise ...Qb6 will hit both the d4-knight and the b2-pawn. 15...Qb6 16.Nf1
Nf8 17.Qd1 Ne6 18.Ng3 Bd7 19.Ne2= White was finally ready to recover the pawn, but Black
remained solid enough in Aronian – Caruana, Paris 2018.
 
8.a4 is the other obvious candidate. 8...Nbd7 (8...Bg4 was also perfectly playable in Marbourg –
Caron, corr. 2018) 9.0-0 0-0 10.Re1 Qc7 11.d4!? Again we see White’s typical central break in this
structure. If he is going to try this move, he needs to do it before Black is able to develop fully and
consolidate.

625
11...exd4 12.exd5 cxd5 13.Nxd4 Nc5 14.Nf1 Now in Moll – Vesely, corr. 2016, the simplest option
would have been 14...Bd7N 15.Ne3 Nxb3 16.Nxb3 Rfe8= when Black is solid.

8...Nbd7
8...0-0!?N also deserves consideration, as 9.Nxe5 is not dangerous: 9...Re8 10.d4 (10.f4 Qb6³) and
now a nice tactical resource is:

626
10...Nxe4! 11.Nxe4 f6 12.0-0 fxe5 Black is doing fine. For example, 13.Ng3 Nd7 14.Re1 exd4
15.Qxd4 Nf6= and the position remains balanced.
 
9.exd5!?
9.0-0 is the most popular move but Black is comfortable after: 9...0-0 10.Re1 Re8 11.Nf1 dxe4
Black wants to play ...Nc5 next without allowing exd5. 12.Ng5 Re7 13.Nxe4 Nxe4 14.dxe4 Qc7=
Black was doing well in Drabke – Mons, Austria 2017.
 
9...cxd5 10.0-0 0-0 11.Re1 Qc7!
11...Re8 was played in Carlsen – Caruana, Stavanger 2018, but the text move is more precise.

627
12.c4!?
12.Nf1N can be met by 12...b5! 13.a4 bxa4 14.Bxa4 e4 with a fine position for Black.
 
12...d4
The structure is similar to a Benoni with colours reversed. However, an important difference is that
White does not have the strong bishop on g2.
 
13.Ba4
13.c5!? should be met by the simple 13...Nxc5 (13...Qxc5?! 14.Rc1 gives White a useful initiative)
14.Nxe5 and now Black should eliminate the strong bishop: 14...Nxb3 15.Nxb3 Qb6 16.Nc4 Qd8=
Black had no problems in Röckendorf – Rio, corr. 2018.
 
13...b6 14.Qe2 Re8 15.Ne4 Nxe4 16.Qxe4 Bb7
Black is doing fine and the game soon ended in a repetition after:
 
17.Qf5 Bc8 18.Qe4 Bb7
½–½ Kribben – Pierzak, corr. 2016.
 

B22) 7.Qxd2

This is less popular than the previous recapture but it is more challenging.
 
7...0-0!
I believe Black must sacrifice a pawn in order to obtain fully equal chances.

628
 
7...Qc7 8.Qg5 0-0 9.Qxe5 Qxe5 transposes to 8...0-0 9.Qxe5 Qxe5 in the note just below.
 
The main alternative is:
7...Qd6
This has scored pretty well for Black but I like White’s chances against it.
8.Qg5 Nbd7
8...0-0 is met by 9.Qxe5 Qxe5 10.Nxe5 dxe4 and now 11.d4! is an important move, after which
I believe White has slightly better chances. For instance, 11...Nbd7 12.Nc3 Nb6 13.0-0-0 Nbd5
14.Rde1² and Black needed to take care of the e4-pawn in Godena – Zubov, Tel Aviv 2013.
9.exd5
9.Qxg7?! Rg8 10.Qh6 occurred in Pikula – Leventic, Bol 2014. Instead of taking on g2, Black
should have continued 10...a5!N 11.a4 Nxe4! 12.Qxh7 Ndf6 13.Qh4 Nc5ƒ when the bishop on
b3 is badly placed and it is obvious that Black’s initiative is worth at least a pawn.
9...cxd5
Now White has an annoying move:

10.d4! e4
10...exd4 is safer although 11.0-0 0-0 12.Nxd4 gave White a pleasant edge playing against the
IQP in Figlio – Hill, corr. 2018.
11.Ne5 0-0 12.Nc3 Nb6 13.f3!
Hitting Black’s pawn chain.
13...Be6
This position was reached in Vitiugov – Zubov, Gjakova 2016. My suggestion for White would
be:
14.0-0-0!?N 14...a5 15.a4 Rac8 16.Qe3 Qb4 17.Rhe1²
White is better due to the pressure on Black’s centre, although the position is still complex.

629
 
After the text move White has a choice between B221) 8.exd5 and B222) 8.Nxe5.
 
8.0-0 and 8.Nc3 are playable but in both cases 8...Re8= gives Black a comfortable game, so we will
focus on what happens when White plays ambitiously and captures the pawn.
 

B221) 8.exd5 cxd5 9.Nxe5

9...a5!
An important move, provoking either a2-a4 or c2-c3, both of which have their own drawbacks.
 
The more obvious 9...Re8 10.d4 Nc6 gives White a pleasant position after: 11.0-0 Nxe5 12.dxe5
Rxe5 13.Nc3²
 
10.a4
This move weakens the support of the b3-bishop, which Black will soon exploit.
 
10.c3N deserves attention. Now White’s bishop is happier but his knight does not have the c3-square
available, so Black can play: 10...Re8 11.d4 Nc6

630
12.0-0 (12.f4? is bad due to 12...Ne4 13.Qe3 a4 14.Bc2 f6 15.Nxc6 bxc6 16.0-0 Qb6µ when Black
is clearly on top due to the activity of his pieces) 12...Nxe5 13.dxe5 Rxe5= Black has a reasonable
IQP position.

10...Nc6!
Instead of trying to recover the pawn, Black should strive for positional compensation with a
lasting initiative.
 
10...Re8?!N is not so good due to 11.d4 Nc6 12.0-0 Nxe5 13.dxe5 Rxe5 14.Nc3² when the knight
comes to its best square and White has meaningful pressure against the d5-pawn.
 

631
11.f4
11.Nxc6N is strongly met by an important intermediate move: 11...Re8†! 12.Ne5 (12.Kf1?! bxc6³)
12...Rxe5† 13.Kf1 Black has good compensation for the pawn and he has a few different ways of
handling the position. My illustrative line continues:

13...Qb6 14.Nc3 (14.d4? is bad due to 14...Ne4 15.Qe3 Rf5 16.f3 Bd7µ followed by ...Re8) 14...d4
15.Ne2 (15.Nb5 Bd7³) 15...Bg4ƒ Followed by ...Rae8. Black enjoys nice compensation for the
pawn, and this would be especially true in a practical game without an engine showing the best
defensive moves for White.
 
I also considered:
11.d4 Qb6!
The drawback of a2-a4 becomes clear, as White cannot defend his central pawn with c2-c3.
12.Nxc6
Another possibility is 12.Nc3N which leads to exchanges: 12...Nxd4 13.Nxd5 Nxd5 14.Bxd5
Qc5 15.c3 Qxd5 16.Qxd4 Qxd4! 17.cxd4 Rd8 18.0-0 Be6! This is more precise than taking on
d4 immediately. 19.Rfd1 f6 20.Nf3 Ra6³ Black will regain the pawn sooner or later, when the
resulting endgame with bishop versus knight will obviously be in his favour.
12...bxc6

632
13.0-0
White has managed to bring his king to safety while remaining a pawn up. At the same time,
his queenside pieces are yet to be developed while the bishop on b3 merely stares at the d5-
pawn. Also note that White has no pieces near the kingside which may give Black chances to
start an attack there.
13...Ne4 14.Qd1 Rb8
14...Nd6 was also good for Black in Waters – Miettinen, corr. 2009.
15.Nc3 Nd6 16.Qd2 Nf5 17.Rad1 Ba6 18.Rfe1 Rfe8
Black has excellent compensation for the pawn, due to his pressure on the d4-pawn and the
useless white bishop on b3. For example:

19.Rxe8† Rxe8 20.h3

633
 20.Qf4? Nxd4–+
20...Re6 21.Qf4 g6³
Black’s positional advantages were worth more than a pawn in Kolanowski – Stojanovic, corr.
2018.

11...Re8 12.0-0 Nxe5 13.fxe5 Rxe5 14.Nc3

White seems to be getting nicely coordinated, with the knight eyeing the b5-outpost and Rae1
coming next. However, Black has a fine idea.
 
14...h6!
Preparing to place the rook on a rather unusual but effective outpost.

634
15.Rae1 Rg5!
The rook is ideally placed here, where it guards the d5-pawn as well as putting pressure on White’s
kingside. Black was at least equal in Lafarga Santorroman – Hall, corr. 2016.
 

B222) 8.Nxe5

I consider this the more challenging way of taking the central pawn. White will not be a pawn up
for long, but he hopes to obtain positional pressure with the help of the strong knight.
 
8...Qe7
8...dxe4 is well met by 9.Nxf7! Rxf7 10.dxe4 Qe7 11.Bxf7† Kxf7 12.f3² when White’s rook and
two pawns are worth slightly more than the two minor pieces, Chen Qi – Zhao Yuanhe, Jiangmen
2018.
 
9.f4 dxe4 10.d4!
This strong move was first played by Karjakin against Fridman in 2012.
 
10.0-0 is nothing special, and after 10...exd3 11.Nxd3 Bf5 Black was doing fine in Enescu –
Vinchev, corr. 2008.
 
10.Nc3!?
This is an interesting sacrificial attempt to seize the initiative on the kingside.
10...exd3 11.0-0-0 dxc2

635
12.Rde1
12.Bxc2N is met by 12...Nbd7 13.Nc4 (13.Rde1?! Nxe5 14.Rxe5 Qd8µ simply leaves Black a
clear pawn up) 13...Qc5 14.Nd6 Nb6³ and White’s compensation for the pawn is not sufficient.
12...Qc7 13.g4
Now it’s important to challenge White’s strong bishop with:
13...Be6! 14.g5 Nfd7

15.Bxc2
15.g6? doesn’t work due to 15...Nxe5 16.gxh7† Kh8 17.fxe5 Bxb3 18.Rhg1 and now in Bartel
– Rausis, Czech Republic 2017, the most accurate was 18...Qe7!N, avoiding any weakening of

636
 the kingside with ...g6. After 19.axb3 (or 19.Qg2 f6µ) 19...Na6µ the black king is safe behind
the enemy pawn on h7, and Black has good chances to convert his extra material.
15...Nxe5 16.Rxe5
16.fxe5 Nd7 17.Qf4 Rfe8! Freeing the f8-square, so that after 18.Qh4 Nf8µ Black’s kingside is
secure.
16...Nd7 17.Re3 Nb6 18.Qf2 Nc4 19.Re2 Qb6³
White does not have sufficient compensation for the pawn.

10...Be6!
It’s important to neutralize White’s powerful bishop.
 
11.0-0
11.c4 is met by 11...Nh5! when ...f6 becomes a serious threat, as White’s knight has nowhere to
retreat. (11...b5!?N is an interesting option as well, for example: 12.cxb5 Bxb3 13.axb3 cxb5 14.Nc3
b4 15.Nd1 The knight goes to the strong blockading square on e3. 15...Nd5 16.Ne3 Rd8 17.Nxd5
Rxd5= Black is doing fine.) 12.f5 Bxf5

637
13.0-0 (13.g4? is refuted by 13...f6 14.c5† Be6–+) 13...g6 14.g4 f6! The tactics were not
unfavourable to Black in Cruzado Duenas – Hatzl, corr. 2016.
 
11...Bxb3 12.axb3

12...c5!
12...Rd8 is also acceptable, when 13.c4 c5! 14.d5 transposes to the main line below.
 
13.d5
13.dxc5?! Qxc5† only helps Black: 14.Qf2 (14.Kh1N can be met by: 14...Nc6 15.Qc3 White
needs to exchange the active queen on c5. 15...Qxc3 16.Nxc3 Nd4! The knight is excellently placed

638
here. 17.Rac1 Rad8 18.Rfe1 Rfe8 19.g4 h6³ Followed by ...b5, when Black’s active pieces offer him
at least slightly better chances.)

14...e3 15.Qe2 Nc6 16.Nxc6 Qxc6 17.Nc3 Rfe8³ Black was pushing for an advantage thanks to the
strong passed e3-pawn in Ragger – Nisipeanu, Graz 2018.
 
13...Rd8 14.c4
14.Nc3 is met by 14...Nc6!, challenging the e5-knight and also intending to go to d4. 15.Rad1
Nxe5 16.d6 Qe8 17.fxe5 Qxe5 Now White has to find some accurate moves if he is to maintain the
balance:

639
18.Qf4! Qxf4 19.Rxf4 Re8 20.d7 Red8 21.Nxe4 Nxd7 White is still a pawn down, but after 22.Rf5!
Nf8 23.Rxd8 Rxd8= he wins it back and the endgame is equal.
 

14...Nc6!
Just as we saw in the note above, Black challenges the e5-knight and also intends to go to d4.
 
15.Qc3N
I believe it is better for White to allow Black’s knight to go to d4 than to let his own knight
disappear from e5.
 
15.Nc3?! has been tested in practice but after 15...Nxe5 16.fxe5 Qxe5 17.Qf4 Qd4† 18.Kh1 e3³
Black was slightly better in Stukopin – Bai Jinshi, Burlingame 2019, as it was not so easy to regain
the e3-pawn.
 
15...Nd4
The knight is extremely strong here.

640
16.Re1 b5! 17.Nd2 b4
White’s best approach is to sacrifice the exchange in some way.
 
18.Qg3 e3! 19.Rxe3
I also checked 19.Qxe3 Nc2 20.Qf2 Nxe1 21.Rxe1 Re8= when White has enough compensation
for equality, but nothing more.
 
19...Nf5 20.Qh3 Nxe3 21.Qxe3 a5©

Next Black will play ...Qd6 and ...Re8. White surely has compensation for the exchange, but not
enough to claim any advantage.

641
 

Conclusion
 
1.e4 e5 2.Bc4 is a pretty respectable way to start the game. We respond with 2...Nf6 3.d3 c6, and in
the event of 4.Qe2 we have a pleasant choice between 4...Bc5, leading to a solid position with no
problems, and 4...d5!?, sacrificing a pawn for a good initiative.
 
4.Nf3 d5 5.Bb3 is the main line, when 5...Bb4† is a useful finesse. If White hits the bishop with 6.c3
then we simply retreat it to d6, when White is unable to bring his knight to c3 and Black generally
has a comfortable game. 6.Bd2 Bxd2† is the main line. Then 7.Nbxd2 a5! is pleasant for Black, but
7.Qxd2 is more challenging, when Black’s best approach is to offer a pawn sacrifice with 7...0-0.
Then 8.exd5 is the materialistic option, when Black’s initiative seems like more than enough to offset
the small material deficit. 8.Nxe5 seems like a better bet, with White aiming for a positional edge,
but we saw that Black has plenty of resources and does not stand worse at all.

642
 Appendix

 
 
As mentioned in the Preface, after analysing the Petroff for this book I was inspired to test my
analysis over the board. The following game took place in September 2019.
 

Konguvel Ponnuswamy – Swapnil Dhopade

Udaipur 2019
 
1.e4 e5 2.Nf3 Nf6 3.d4 Nxe4 4.Bd3 d5 5.Nxe5 Nd7 6.Nc3 Nxc3 7.bxc3 Nxe5 8.dxe5 Bc5 9.0-0 0-
0

10.Kh1
Black’s superior pawn structure gives him good long-term prospects, so White should play
energetically if he wishes to cause problems. That’s why 10.Qh5 is a more important option,
although Black is doing fine – turn back to page 42 for details.
 
10...c6
This is always a useful move, although the machine’s suggestion of 10...Qh4!?N also looks
promising.
 

643
11.Qh5 g6 12.Qh6 Be7
Covering the dark squares.
 
13.f4 f5!
This is the typical way to block White’s actions on the kingside. On page 42 we reached the same
position but with White having played Be3 instead of Kh1, so I was essentially still following my
analysis.

14.c4
Exchanging one of the doubled pawns is risky: theoretically White improves his structure, but in
doing so he enables Black to get closer to the a- and c-pawns, which will remain weak for a long
time.
 
14...Be6 15.cxd5 Qxd5 16.Be3 c5 17.Bf2
White would like to exchange the bishops with Bh4, but I was able to avoid this quite conveniently.
 
17...Rf7 18.Bh4 Bf8 19.Qg5

644
19...c4
Gaining space on the queenside and driving White’s bishop to a passive position.
 
20.Be2 Qe4 21.Bd1 Rd7
Black’s pieces are more active than their counterparts and White faces a difficult defensive task.
 
22.Re1 Qd4 23.Bf3 Qc3

24.Rad1
Rather than passively defending c2, White tries to obtain some counterplay, but is unable to create
any serious threats.
645
 
24...Rxd1 25.Rxd1 Qxc2 26.h3
26.Qf6 gets nowhere after 26...Re8.
 
26...c3

27.Kh2?
27.Qf6 Re8 28.Rd8 was the last chance. 28...Qa4! is the key resource, when Black should win with
careful play, but White can keep playing for a while. After the text move, he has nothing and the
game is soon over.
 
27...Qxa2 28.Bxb7 c2 29.Re1 Rb8 30.Bf3 Rb1 31.Qd8 Qc4 32.Be7 Qc8 33.Qxc8 Bxc8 34.Bd5†
Kh8 35.Bf6† Bg7
0–1

646
 Variation Index

Chapter 1
1.e4 e5 2.Nf3 Nf6
A) 3.Bc4?! 8
B) 3.Nc3 Bb4!? 4.Nxe5 0-0 5.Be2 Re8 6.Nd3 Bxc3 7.dxc3 Nxe4 8.0-0 d6 10
B1) 9.f3 11
B2) 9.Re1 Nd7 13
B21) 10.Bf1 13
B22) 10.Bf3 14
B3) 9.Nf4 Nd7 16
B31) 10.f3 17
B32) 10.c4 18
 
Chapter 2
1.e4 e5 2.Nf3 Nf6 3.d4 Nxe4
A) 4.dxe5 d5 5.Nbd2 Nxd2 22
A1) 6.Bxd2 Be7 23
A11) 7.c4 24
A12) 7.Bd3 26
A2) 6.Qxd2 27
B) 4.Bd3 d5 31
B1) 5.dxe5 Nc5 31
B11) 6.0-0 Be7 7.Nc3 c6 31
B111) 8.Nd4 32
B112) 8.Re1 33
B12) 6.Be2 Be7 7.0-0 0-0 35

647
 B121) 8.Nc3 35
B122) 8.Be3 37
B2) 5.Nxe5 Nd7 40
B21) 6.Nc3 41
B22) 6.Qe2 43
B23) 6.0-0 Nxe5 7.dxe5 Nc5 46
B231) 8.Nc3 47
B232) 8.Be2 50
B233) 8.Be3 51
 
Chapter 3
1.e4 e5 2.Nf3 Nf6 3.d4 Nxe4 4.Bd3 d5 5.Nxe5 Nd7 6.Nxd7 Bxd7 7.0-0 Bd6
A) 8.Qh5 Qf6! 57
A1) 9.Qxd5 57
A2) 9.Nc3 58
B) 8.Nc3!? Nxc3 9.bxc3 0-0 10.Qh5 f5 61
B1) 11.Bg5 62
B2) 11.Rb1 63
B3) 11.Re1 c6 12.Bg5 Qc7 64
B31) 13.c4 65
B32) 13.Re3!? 66
C) 8.c4 c6 9.cxd5 cxd5 67
C1) 10.Qh5 0-0 11.Qxd5 Bc6 12.Qh5 g6 68
C11) 13.Qh6 68
C12) 13.Qh3 68
C2) 10.Nc3 Nxc3 11.bxc3 0-0 12.Qh5 g6! 13.Qxd5 Qc7 14.Bh6!? Rfd8 71
C21) 15.Qg5 72
C22) 15.Qf3 73
  

648
Chapter 4
1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6
A) 4.Nxf7?! Kxf7 76
A1) 5.d4 c5! 76
A11) 6.dxc5 77
A12) 6.Nc3 78
A2) 5.Nc3!? 79
B) 4.Nd3 Nxe4 5.Qe2 Qe7 82
B1) 6.b3 83
B2) 6.Nc3 84
B3) 6.Nf4 Nc6 86
B31) 7.c3 87
B32) 7.Nd5 88
C) 4.Nc4 Nxe4 90
C1) 5.d4 90
C2) 5.Nc3 92
C3) 5.Qe2! Qe7 6.Ne3 Be6 94
C31) 7.Nc3 95
C32) 7.c4!? Nc6!? 8.Nc3 Nf6 9.Ncd5 Nd4! 10.Qd3!? Bxd5 11.cxd5N Qe5 96
C321) 12.f4!? 98
C322) 12.g3 99
 
Chapter 5
1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4
A) 5.Bd3 d5 103
A1) 6.Qe2 103
A2) 6.0-0 Be7 7.Re1 Nd6 8.Nc3 c6 106
A21) 9.Qe2 107
A22) 9.b3!? 109

649
B) 5.c4 Nc6 111
B1) 6.Be2 111
B2) 6.d4 d5 112
B21) 7.Nc3?! 113
B22) 7.a3 114
B3) 6.Nc3 115
B4) 6.d3 Nf6 7.d4 d5 118
B41) 8.Nc3 119
B42) 8.c5!? 120
 
Chapter 6
1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4 5.d3 Nf6 6.d4 d5
A) 7.Nc3 124
B) 7.c4 Bb4† 126
B1) 8.Bd2 126
B2) 8.Nc3 127
C) 7.Bd3 Bd6 128
C1) 8.Qe2† Be6 128
C11) 9.Ng5 129
C12) 9.c4 130
C2) 8.0-0 0-0 131
C21) 9.h3 132
C22) 9.Bg5 Bg4 10.Nbd2 Nbd7 133
C221) 11.c4 134
C222) 11.c3 135
  
Chapter 7
1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4 5.Qe2 Qe7
A) 6.Nc3 140

650
B) 6.d3 Nf6 142
B1) 7.Nc3 143
B2) 7.Bg5 Qxe2† 8.Bxe2 Be7 144
B21) 9.0-0 145
B22) 9.Nc3 c6 10.0-0-0 Na6!? 146
B221) 11.Rhe1 147
B222) 11.Rde1 Nc7 12.Bd1 Ne6 13.Bd2 Bd7 14.h3 0-0-0 149
B2221) 15.Ng5!? 150
B2222) 15.g4 152
  
Chapter 8
1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4 5.d4 d5 6.Bd3 Bd6 7.0-0 0-0 8.c4 c6
A) 9.cxd5 157
B) 9.Re1 Bf5 159
B1) 10.Qc2 Na6 11.a3 Bg6 12.c5 Bb8 13.Nc3 f5 14.Qb3 Bh5! 160
B11) 15.Bxe4 161
B12) 15.Nxe4 162
B2) 10.Qb3 Qd7 11.Nc3 Nxc3! 12.Bxf5 Qxf5 164
B21) 13.bxc3 165
B22) 13.Qxb7 166
 
Chapter 9
1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4 5.d4 d5 6.Bd3 Bd6 7.0-0 0-0 8.c4 c6 9.Qc2 Na6!
A) 10.Bxe4 170
B) 10.a3 171
B1) 10...Bg4 171
B2) 10...f5!? 174
B21) 11.c5?! 174
B22) 11.b4 175

651
 B23) 11.Nc3 Nc7 177
B231) 12.c5 178
B232) 12.b4 Bd7! 181
B2321) 13.Bb2 182
B2322) 13.c5 183
 
Chapter 10
1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4 5.d4 d5 6.Bd3 Bd6 7.0-0 0-0 8.c4 c6 9.Nc3 Nxc3
10.bxc3 dxc4 11.Bxc4 Bf5
A) 12.Ne5 187
B) 12.Re1 189
C) 12.Ng5 h6 190
C1) 13.Nxf7N 190
C2) 13.Qf3! 192
D) 12.Bg5 Qa5! 194
D1) 13.Re1 195
D2) 13.d5!?N 196
D3) 13.Nh4 198
  
Chapter 11
1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4 5.Nc3 Nxc3 6.dxc3 Be7
A) 7.Bf4 0-0 8.Qd2 Nd7 9.0-0-0 Nc5! 201
A1) 10.Nd4 202
A2) 10.h4 204
A3) 10.Be3 Ne4! 205
A31) 11.Qe1 206
A32) 11.Qd4 207
B) 7.Be3 0-0 209
B1) 8.Bd3 Nd7 210
B11) 9.Qd2 210

652
 B12) 9.Qe2 211
B2) 8.Qd2 Nd7 9.0-0-0 Nf6 10.Bd3 c5 212
B21) 11.c4?! 216
B22) 11.Bg5 218
  
Chapter 12
1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4 5.Nc3 Nxc3 6.dxc3 Be7 7.Be3 0-0 8.Qd2 Nd7 9.0-0-0
Nf6 10.Bd3 c5
A) 11.Kb1 b5! 12.Rhe1 Rb8 13.Bg5!? Be6 14.Qf4 Qa5 222
A1) 15.b3 224
A2) 15.Rxe6!N 226
B) 11.Rhe1 Be6 227
B1) 12.a3 228
B2) 12.Kb1 Qa5 229
B21) 13.c4 229
B22) 13.a3 231
B3) 12.Bg5 235
 
Chapter 13
1.e4 e5 2.d4 exd4
A) 3.c3 dxc3 241
A1) 4.Nxc3 241
A2) 4.Bc4 243
B) 3.Nf3 Bb4† 246
B1) 4.Bd2 Bxd2† 246
B11) 5.Qxd2 246
B12) 5.Nbxd2 248
B2) 4.c3 dxc3 249
B21) 5.bxc3?! 249
B22) 5.Nxc3 250

653
C) 3.Qxd4 Nc6 252
C1) 4.Qc4 252
C2) 4.Qe3 Nf6 254
C21) 5.e5?! 255
C22) 5.Bd2 255
C23) 5.Nc3 Bb4 6.Bd2 0-0 7.0-0-0 Re8 257
C231) 8.Qf4!? 257
C232) 8.Bc4 258
C233) 8.Qg3 Nxe4 9.Nxe4 Rxe4 260
C2331) 10.c3 260
C2332) 10.Bf4 262
 
Chapter 14
1.e4 e5 2.f4 d5 3.exd5 exf4
A) 4.d4? 265
B) 4.Bc4?! 266
C) 4.Qf3 267
D) 4.Bb5† 269
E) 4.Nf3 Nf6 271
E1) 5.Be2 271
E2) 5.c4 273
E3) 5.Bc4 Nxd5 275
E31) 6.Bxd5 275
E32) 6.0-0 277
E4) 5.Bb5† c6! 6.dxc6 Nxc6 7.d4 Bd6 279
E41) 8.Qe2†?! 280
E42) 8.0-0 281
  
Chapter 15

654
1.e4 e5 2.Nc3 Nf6
A) 3.g3 288
B) 3.f4 d5! 4.fxe5 Nxe4 290
B1) 5.Qf3 290
B2) 5.d3 292
B3) 5.Nf3 294
C) 3.Bc4 Nxe4 4.Qh5 Nd6 5.Bb3 Be7 295
C1) 6.Nf3 297
C2) 6.Qxe5 298
  
Chapter 16
1.e4 e5 2.Bc4 Nf6 3.d3 c6
A) 4.Qe2 302
A1) 4...Bc5 303
A2) 4...d5!? 303
B) 4.Nf3 d5 5.Bb3 Bb4† 305
B1) 6.c3 305
B2) 6.Bd2 Bxd2† 308
B21) 7.Nbxd2 308
B22) 7.Qxd2 0-0 309
B221) 8.exd5 310
B222) 8.Nxe5 313

655
Table of Contents
Title Page 2
Key to Symbols used & Bibliography 4 4
Preface 5 5
1 3.Bc4?! & 3.Nc3 7 8
2 3.d4 21 36
3 6.Nxd7 55 108
4 4th Move Alternatives 75 147
5 5.Bd3 & 5.c4 102 205
6 5.d3 123 249
7 5.Qe2 139 280
8 5.d4 154 310
9 9.Qc2 169 340
10 9.Nc3 186 374
11 5.Nc3 200 403
12 11.Kb1 & 11.Rhe1 221 446
13 Centre Game 240 486
14 King’s Gambit 264 536
15 Vienna Game 287 583
16 Bishop’s Opening 301 611
Appendix: Konguvel Ponnuswamy – Swapnil Dhopade 317 643
Variation Index 320 647

656