First

published in 2015 by Gloucester Publishers Limited, Northburgh House,

10 Northburgh Street, London EC1V 0AT

Copyright © 2015 Jovanka Houska

The right of Jovanka Houska to be identified as the author of this work has been
asserted in accordance with the Copyrights, Designs and Patents Act 1988.

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.

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

ISBN Kindle: 978-1-78194-211-6

ISBN epub: 978-1-78194-212-3

Distributed in North America by National Book Network,

15200 NBN Way, Blue Ridge Summit, PA 17214. Ph: 717.794.3800.

Distributed in Europe by Central Books Ltd.,

99 Wallis Road, London E9 5LN. Ph 44(0)845 458 9911.

All other sales enquiries should be directed to Everyman Chess,

Northburgh House, 10 Northburgh Street, London EC1V 0AT
email: info@everymanchess.com; website: www.everymanchess.com

Everyman is the registered trade mark of Random House Inc. and is used in this work under licence from
Random House Inc.

Everyman Chess Series

Chief advisor: Byron Jacobs
Commissioning editor: John Emms
Assistant editor: Richard Palliser

Typeset and edited by First Rank Publishing, Brighton.

Cover design by Horatio Monteverde.
Printed by TJ International Limited, Padstow, Cornwall.
 About the Author
Jovanka Houska was the 2013 Commonwealth Ladies Champion and is a many-time
winner of the British Women’s Championship. She has represented the English team, on top board, at
numerous Chess Olympiads, and qualified for the 2006 Women’s World Championship. In 2006 she won
the English Chess Federation Player of the Year award,
the first time a female chess player had received this prestigious recognition.

Also by the Author:

Play the Caro-Kann
Starting Out: The Scandinavian
 Contents
About the Author
Bibliography
Introduction

1 Main Line: Introduction and 11 Bf4

2 Main Line: 11 Bd2
3 Classical Variation: The f4 Plan
4 The Challenging Nf4
5 Panov-Botvinnik Attack: Introduction and 6 Nf3
6 Panov-Botvinnik Attack: 6 Nf3 g6 without 7 cxd5
7 Panov-Botvinnik Attack: 6 Bg5
8 The Exchange Variation
9 Advance Variation: Introduction to 3 ... c5
10 Advance Variation: White Returns the c5-pawn
11 Advance Variation without 4 dxc5
12 The Fantasy Variation
13 The Pseudo-Panov: 2 c4
14 Two Knights Variation
15 The King’s Indian Attack
16 Rare Lines
 Bibliography
Beating the Caro-Kann, Vasilios Kotronias (Batsford 1994)
Caro-Kann Defence: Advance Variation and Gambit Systems,
Anatoly Karpov and Mikhail Podgaets (Batsford 2006)
Caro-Kann Defence: Panov Attack, Anatoly Karpov and Mikhail Podgaets (Batsford 2006)
Dangerous Weapons: The Caro-Kann,
John Emms, Richard Palliser and Jovanka Houska (Everyman Chess 2010)
Easy Guide to the Panov-Botvinnik Attack, Jacob Aagaard (Everyman Chess 1998)
Grandmaster Repertoire: 1.e4 vs the French, Caro-Kann & Philidor,
Parimarjan Negi (Quality Chess 2013)
Grandmaster Repertoire: The Caro-Kann, Lars Schandorff (Quality Chess 2010)
Grandmaster Secrets: The Caro-Kann, Peter Wells (Gambit 2007)
How to Play the King’s Indian Attack, Angus Dunnington (Batsford 1993)
Main Line Caro-Kann, Neil McDonald (Everyman Chess 2001)
Play the Caro-Kann, Jovanka Houska (Everyman Chess 2007)
Play the Caro-Kann (2nd edition), Egon Varnusz (Maxwell MacMillan Chess 1991)
Starting Out: The Caro-Kann, Joe Gallagher (Everyman Chess 2002)
The Caro-Kann: Move by Move, Cyrus Lakdawala (Everyman Chess 2012)
The Classical Caro-Kann, Gary Kasparov and Aleksander Shakarov (Batsford 1984)
The Extreme Caro-Kann, Alexei Bezgodov (New in Chess 2014)
The Safest Grünfeld, Alexander Delchev and Evgenij Agrest (Chess Stars 2011)
Winning with the Caro-Kann Defence, Jeremy Silman (Chess Digest 1993)

 Introduction
Welcome back, dear reader. It has been shockingly over seven (!) years since I wrote my first chess book,
Play the Caro-Kann - a long time in the chess world. A few years ago I developed a niggling feeling that
I hadn’t quite written all I had to say on my favourite opening and it was a marvel at how the opening had
changed so much. I have been pretty much loyal to the Caro-Kann, using it as a weapon against all types
of opponents from beginners to 2700s, so I consider it a very diverse weapon. Most of my resulting
positions ended in the better player winning - I have yet to find an opening that has given me such a
positive results. In fact I have converted a husband and a few of my students to the Caro-Kann cause!

History of the Opening

It was two German players, Horatio Caro and Marcus Kann, who introduced this defence during the
second half of the nineteenth century. At the time, chess was very much in its ‘Romantic’ stage, a time
when it was rude not to accept sacrifices, and gambits were all the rage. Unsurprisingly, the Caro-Kann
did not find too many adherents, but as the concepts of positional chess developed, people began
appreciating the qualities of the opening: Black gets easy development of the pieces, clear plans with a
solid pawn structure, and safe but active play.
Today, the Caro has a grand heritage with many great players, including world champions Botvinnik,
Petrosian and, most significantly, Anatoly Karpov using it with great success. Modern advocates of the
Caro include distinguished grandmasters such as Riazantsev, Laznicka, Andreikin, Dreev, Lalith and
Gunina to name just a few. I fully advocate following the games of these players and shamelessly copying
what they play. After all, why not have a top grandmaster play your opening for you?

Aims of the Caro-Kann

The Caro-Kann has often been maligned for being dry and boring, played by those wishing to bore their
opponents to death. However, the fundamental aim of the opening is to achieve full development,
particularly of the light-squared bishop, and then begin active play to challenge White’s space advantage.
Our basic pawn structure with pawns on c6-e6 (and d5 - if it is still there) is flexible and easy to play.
There are, however, many sides to the Caro and one should not lose sight that at times it can be very
sharp as well as positional. Don’t be swayed by assumptions and stereotypes. Some of the lines I have
recommended are solid, others are sharp, and yet others are based on obtaining a more comfortable
endgame. We must simply get used to playing what the position demands.
The whole premise of 1 ... c6 is very much based on the logic that is also behind the French Defence:
with the first move Black gets ready to support the ... d7-d5 advance. The advantage of our move is that
Black does not block the light-squared bishop on c8, but there is also a disadvantage in that the important
move ... c6-c5 is not available to Black in most lines, at least not immediately.
The centre is usually filled with too much tension for Black to be able to spend two tempi achieving ...
c5, and Black must usually wait until he is fully developed before he can afford to expend another tempo
in pushing the c-pawn one square further. There is one exception to this and that is in the Advance
Variation (Chapters 7-8), the justification being that White has already eased the tension in the centre by
advancing the e-pawn.
As I stated earlier, Black’s main aim is to complete development before beginning any active
operations (although of course there are some exceptions). However, when playing the opening, it is very
useful to keep in mind the following six factors. Some of them are clearly self-explanatory; others, I will
explain how they should be applied to the Caro-Kann directly.
1) The Centre
The centre plays a very important part in most openings and the Caro-Kann is no different. The nature of
the centre is flexible and dependent on the variation chosen by White: at times it can be full of high
tension as in the Panov; relatively static, as in the Exchange Variation; or dynamic, as in an isolated
queen’s pawn (IQP) position. Black should always value the importance of the centre, especially the four
centre squares e4, e5, d4 and d5.

2) Open Lines
Again this is dependent on which variation is played, but as a very general point Black should be looking
to contest the central files, in particular if there is a half-open d-line. Naturally we must pay attention to
the pawn breaks ... c5 and ... e5, placing pressure on the central square d4. As a very general formula, the
... e5 break becomes more attractive if (a) the white king is in the centre, and/or (b) White has made no
effort to control any squares in the centre.

3) Piece Activity
Both sides use this concept a lot in the Caro-Kann. White primarily uses this dynamic concept to speed up
his opening development by attacking the light-squared bishop to gain more space and attacking potential.
Black, on the other hand, depends on easy development and his solid pawn structure to withstand White’s
early initiative. My recommendation in the Advance Variation, however, is based first on early pawn
challenges to the centre and then the resulting dynamism of the pieces.

4) Pawn Structure
Generally speaking Black has a good structure with pawns usually occupying light squares, the only
disadvantage being that to begin with Black’s structure does not control much space. One point to
understand is that if Black exchanges his light-squared bishop for a knight, he should very much play on
the dark squares and use the pawns as a light-square blockade.

5) Strong and Weak Squares

As a very rough guide, Black’s weakest point is the e5-square and this is often utilized by a white knight
to commence an attack. A strong square for Black, if the pawn has been exchanged, is the d5-square, but if
the d5-pawn remains then the c4- and e4-squares are potential strengths.

6) King Safety - Take care of your King!

This is pretty much explanatory one would think? Not the case! It’s surprising to me how many people
don’t follow this advice. In some lines in the Classical Main Line, Black’s play takes the form of
preventing any g2-g4 pawn storms. In other lines, it means not opening the position when his Majesty
hasn’t got cover. In a few lines when White has gone for the jugular very early on, sacrificing all kinds of
material to get at our king, I would recommend sitting down and memorizing the defences. Take care of
the elderly monarch, he needs physical help.

The Modern Approach to the Caro-Kann

Since the introduction of computers and the rise of opening theory, Black openings in general have taken
some serious poundings and the Caro-Kann is no exception. Lines that were perceived to be simple have
become amazingly complex or White has simply become more confrontational. There is no denying
things: opening theory has become more complicated which accounts for the increase in size of this book.
White players, no longer content with simply playing a position as it is, have adopted three approaches:
the first is to try to completely blast Black off the board, the second an attempt to force Black to ‘grovel’
for the draw in a chanceless position, and the last to catch Black in unknown territory.
Unfortunately, it has become more and more necessary for players to have a deeper knowledge of the
theory, so I would suggest that readers make a real effort to understand the moves and ideas in the sharp
lines (Advance, Panov) and the Classical main lines, whilst simply learning some basic principles for the
lines which are known to be less dangerous for Black. In the latter case, having a decent positional
understanding will count for much more than simply memorizing variations.

My Approach to this Book

I have very much tried to write this book exactly how I like books to be written - with plenty of
explanations of ideas and basic principles, along with some new theory. I have also added some ‘Houska’
formulas and taglines, which are designed to help the reader memorize/understand all the lines. I am
famous amongst my students for carrying notebooks and jotting down piece formations, rules and ‘special’
moves when I prepare and I have continued this trend here.
On the whole I have tended to avoid lines where I feel Black only has the chance of playing for ‘two
results’ (a loss or a draw), and instead chosen lines with ‘three results’ on offer. Sometimes this has led to
me, following in my dad’s footsteps, and like the last edition, adding little Houska twists. However, I
would like to add a practical word of advice - an opening repertoire is most effective when one has a few
weapons of choice; it is good to have within an opening the option of a few choices. I would recommend
all readers to include as many twists and turns in their repertoire as possible - a good player is a slippery
fish to catch!
So to all Caro players out there, good luck!

Jovanka Houska,
Bergen,
December 2014
Chapter One
Main Line: Introduction and 11 Bf4
1 e4 c6 2 d4 d5 3 Nc3

One of the advantages of 3 Nc3 is that Black has no useful move other than relinquishing the centre
with 3 ... dxe4, allowing the white knight to reach the central square e4 (3 ... g6 is a possibility, but is
more in the realm of the Modern Defence than this book). It is because of this factor that the ‘Main Line’
(3 Nc3) is probably the most principled test of the Caro-Kann. Structurally White neither goes all out for
an attack, nor does he directly attempt to refute the whole opening. Instead, he tries to seize as much space
as possible, using the light-squared bishop as a target whilst concentrating on his central superiority.
White can also play the similar 3 Nd2. For the purposes of our repertoire, there is no difference
between this and 3 Nc3 as Black can simply exchange pawns on e4. However, I do feel that the move 3
Nd2 allows Black some extra options: the main one is 3 ... g6 and, among others, there’s the little-known
move 3 ... Qb6.
3 ... dxe4
Opening the d-line and setting the scene for a great strategic battle.
4 Nxe4 Bf5
 The Classical (or Capablanca) Variation is the simplest and most logical way to contest the centre.
Black develops his bishop to a very natural square and at the same time challenges White’s control of the
e4-square. The downside to this move is that it gives White time to seize the initiative and grab some
kingside space.
5 Ng3
Moves such as 5 Nc3 simply allow Black to adopt the basic plan of easy development and placing
pressure on the d-pawn.
5 ... Bg6

White has used various methods over the years here in search for an advantage, and there has been a
steady evolution of this line. To begin with White attempted to obtain advantages through the greater
space obtained by the presence of the d4-pawn, winning the bishop-pair, and freer development. It was
not until later that more aggressive plans were formulated, in which White actively played for more space
and greater central control.
It was the former world champion Boris Spassky who devised a new idea of advancing the h-pawn to
h5, cramping the black kingside, and for a while the Classical Variation became synonymous with an
inferior endgame. However, the evolution was not at all one-sided and there were many advances for
Black; as White became more aggressive, so did the second player.

The Centre
As with many openings, controlling the centre is a very important part of both sides’ strategies.

Pawn moves are highly committal - after all, once played we can never take them back! Our third
move has already set the strategic scene: Black intends to develop naturally and apply pressure along the
semi-open d-line, opening up the position with a well-timed ... c5 or ... e5 pawn break. White in turn has
a harder job of breaking open the e-line and tends to rely on his natural space advantage and freer piece
co-ordination to make life unpleasant for Black. That is why White cannot simply play natural moves and
expect to be better.
In the main line of the Classical Variation there is a continuous battle for all four central squares:
Black should aim to control two of the squares, d5 and e4. The d4-square is our target and the e5-square
something that Black should protect or destabilize. In terms of protection the e5-square is the perfect post
for a white knight, which is why Black plays the defensive ... Nbd7 early on in the game. We can also
negate the impact of a white knight on e5 by striking out with ... c6-c5 or exchanging knights and
controlling the open d-line.
One of the most important squares for Black to control is d5, and he should be prepared to fight for it
in radical ways: for example, by playing ... b7-b5. The d4-pawn is usually targeted with a well-timed ...
c5 advance, but this is normally played only once Black has developed his army completely.

Advancing the h-pawn: a Curse or Blessing?

The most common way for White to search for an advantage is by gaining extra space with a quick h4-h5
followed by queenside castling, but there are also more aggressive plans solely directed at attacking the
black king. The advanced h-pawn is a double-edged sword for White. At times it is man of the match,
giving White a winning position when the game may otherwise have been a draw; at other times it is the
curse of White’s position.
 The h-pawn is used as a clamp on Black’s kingside pawn structure to restrain the advance of his
pawns in an endgame or to control his defensive pawn moves in a middlegame. White’s idea is that
should the pawn be supported by its brothers on the g- and f-files in a middlegame or even in an endgame,
then the situation can become critical very quickly for Black. We should also steer clear of any dark-
squared bishop endgames, as the kingside pawns - g7 and h6 in particular - will just be weak.
It is Black’s recurrent aim to prevent White from making massive pawn advances on the kingside. He
does this in several ways:
a) Distraction. The main method here is to strike in the centre very quickly; White cannot concentrate
all his forces on a kingside pawn attack if his own king is under pressure.
b) Elimination. Swapping off minor pieces - the fewer minor pieces on the board, the fewer pieces
there are for attack.
c) Defence. Blockading the g-pawn.
d) Prevention. Keep the white knight on g3, which leads us nicely to the subject of the ‘bad’ g3-knight ...

Good piece/Bad piece



The knight on g3 is White’s only bad piece. If White begins to drift or does not play Ne4 (or Nf1/e2),
then it may become doomed to passivity for the rest of the game, chained to the defence of the h5-pawn.
Black, on the other hand, does not have any ‘bad’ pieces. The knight on g3 does, however, have one
tactical function: it can hop into f5. This can be seen especially in lines when Black breaks too early with
... c6-c5, allowing White to play the pawn push d4-d5.
This g3-knight plays a big part in whether Black should take the game into an endgame.

The Classical Middle/Endgame

It is of vital importance that a Caro player can spot whether resulting endgames after ... c6-c5 are good or
bad. The Houska formula as to whether the endgame is good works like this:
Do you have (some) control of the d-file? Yes!
Is the h5-pawn defended by a knight on g3? Yes!
Then we should have a good Caro-Kann endgame. The reason? Well, the blame for that lies wholly
with that h5-pawn. If White has to use some of his resources defending the pawn with his pieces, it means
he cannot use his energy to advance his queenside pawns and control the d-file.
If, instead, the formula works like this:
Do you have control over the d-file? Yes, well it’s contested.
Now the key to whether the endgame is good lies in the next question:
Is the h5-pawn defended by an enemy piece? Well, no, there is a pawn chain: f3-g4-h5.
Here the chances are that we might have a bad endgame particularly if White can advance his pawns
on the queenside easily and we can’t push our kingside pawns. Of course, there are exceptions, but this is
a powerful formula.
Here are some practical examples from my games.
 This position is from V.Klotz-J.Houska, German League 2014. Black has the better chances, although
please note though that it’s still a long way from winning. Because of the pressure on the h5-pawn, White
cannot challenge Black on the d-line. It is for precisely this reason White cannot dedicate herself to her
goal of gaining control on the d-line and swapping off pieces for the queenside pawn majority to make its
presence felt. Black, on the other hand, simply needs to maximize the placing of the knights.
The next example is from P.Hommerson-Houska, Gibraltar 2013, and shows scenario two in full
force.

Here we can see that White is contesting control over the d-file and although he is supporting h5 with
the queen, he can easily play g2-g4, defending its h-brother. Black is slightly worse here and in danger of
drifting into realms of danger. Play continued 22 Nb3 Rxd2 23 Rxd2 Nxb3 24 axb3 Qc7?! when White
has the advantage and, indeed, won the game because of these mobile queenside pawns.
Let’s go back to the position after 5 ... Bg6:
6 h4 h6 7 Nf3
 An important move: White induces Black to play the defensive ... Nd7 to prevent the white knight on
f3 landing on the powerful square e5.
Instead, 7 f4 is rarely seen and should be met by logical moves, as demonstrated by Grandmaster Igor
Rausis in the following game: 7 ... e6 (to prevent the immediate f4-f5) 8 Nf3 Nd7 9 h5 Bh7 10 Bd3 Bxd3
11 Qxd3 Bd6 12 Bd2 Qc7 13 Ne5 Rd8 (13 ... Nxe5?! is a bit dodgy as the bishop has to return to f8 after
13 fxe5) 14 Qb3 Ngf6 15 0-0-0 0-0 16 Kb1 c5 and with the centre opening up, Black is doing well in a
sharp position, R.Mamedov-I.Rausis, Abu Dhabi 2006.
7 ... Nd7
Both 7 ... Nf6 and 7 ... e6 allow 8 Ne5. In my original book I was quite dismissive of this variation,
but if this line has the stamp of approval from World Champion Magnus Carlsen then one must take note!
In fact I would strongly recommend to the reader to include this line in your repertoire, particularly if you
have, like me, exclusively dedicated your games to 7 ... Nd7.
I have been experimenting with this line in the recent months and play has continued along these lines:
8 h5 (8 Ne5!? is of course critical, but White is trying to transpose into the 7 ... Nd7 lines) 8 ... Bh7 9 Bd3
Bxd3 10 Qxd3 e6 11 Bd2 c5 12 0-0-0 Nc6! with an interesting game ahead.
8 h5 Bh7 9 Bd3
White aims to castle queenside, so it is wise to swap off Black’s light-squared bishop.
9 ... Bxd3 10 Qxd3
 10 ... e6!
This is the most flexible move here and the only one I will be recommending. Why do I like this move
much more than alternatives? It is because it opens the diagonal for the dark-squared bishop, which will
prove to be very important.
We now come to an important cross-roads for the whole variation - where will White choose to place
his dark squared bishop: d2 or f4? As we will see later on, this choice will be very important.
The other main moves available to us are 10 ... Nf6 and 10 ... Qc7. At first it was assumed that it was
irrelevant which move Black chose as they would all transpose to the same thing: Black would castle
queenside and then undermine the white centre with a well-timed ... c7-c5 break. After a while, though, it
was discovered that Black need not castle queenside, but could steer the game into sharper territory by
castling short, whereupon each side would begin a race to hack the other side to pieces.
To begin with, Caro players were obsessed with ... a7-a5 ideas (the great Danish Grandmaster Bent
Larsen was one of the proponents), but as White developed new and effective attacking ideas, the focus
again shifted on to plans involving ... c7-c5, attacking the d4-pawn and aiming to utilize the c-file as
Black does in the Sicilian. It is this plan of castling kingside that I recommended in my original book and I
have again continued with this idea.

Kingside Castling
Generally, in response to the plan of castling kingside, there have been three White plans:
a) Playing Nf3-e5, securing a strong post for the f3-knight.
b) Centralizing the queen and king’s rook on the e-file.
c) Playing the g3-knight to f5 upon which it would sacrifice itself if left there or, if captured, Black
would have to play with doubled f-pawns.
However, first White must make a choice as to where to place the dark-squared bishop and this
choice sets the scene for Black’s replies.

Introduction to 11 Bf4
 White places his bishop on the most active square possible and at the same time gets ready to support
a knight jump to e5. This move also has the advantage of not blocking the d-file, so that Black’s move ...
c6-c5 loses some strength (compare this with 11 Bd2, discussed in Chapter 2).
For a long time 11 Bf4 has been the most critical move against the Classical Variation. Black has to
tread very carefully in this line to stay in the game: if he plays too passively then he faces suffering a long
and joyless endgame in a cramped position, but play too recklessly and it can be all over before it has
even started.
Step 1: provoke the advance of the c-pawn.
11 ... Qa5+!

Firstly, if White does not drive the queen back then Black will have achieved his first aim which is to
discourage castling queenside. When the h-pawn has been pushed all the way to h5, normally the last thing
White wishes to do is house his monarch on that side of the board, so to continue with the most aggressive
plan he must chase the lady away.
Our second aim is that by encouraging White to drive our queen away we have provoked the advance
of the c-pawn. As I encourage us to castle kingside, this means that a pawn on c3 or c4 is within touching
distance of our pawn on b7. It only takes one move for us to either open lines or to close the position
completely.
12 Bd2
12 c3 really does not challenge Black at all. White voluntarily abandons his ideas of castling
queenside - something an experienced Caro player will absolutely relish because there is just no kingside
attack anymore. After 12 ... Ngf6 13 a4!? (attempting to take advantage of the exposed black queen on a5,
but the problem with this is that should White advance the c-pawn he will make a big hole on b4) 13 ...
Nd5! 14 Bd2 Qc7 White has:
a) 15 Rh4?! makes no sense here. Indeed, after 15 ... Be7 16 Rg4? N7f6! (Black chases away the
white rook) 17 Rh4 (17 Rxg7 loses the exchange to 17 ... Bf8) 17 ... Bd6 White has problems in
completing his development.
b) 15 0-0 Bd6 16 Ne4 N5f6 17 Nxd6+ Qxd6 18 Rfe1 Nxh5 when Black is a pawn up and without
much danger, S.Kasperski-W.Sapis, correspondence 1993.
12 ... Bb4!?

The most provocative approach. To retain any advantage White should avoid any simplification of the
position, so he must either consent to Black’s wish and push the c-pawn forward one square, which is by
far the main move, or play 13 Ne4.

A) 13 Ne4
B) 13 c3

I once faced 13 Qb3, but this strange move is easily met with 13 ... Bxd2+ 14 Nxd2 Ngf6 15 0-0-0 0-0 16
Nge4 c5 with no prospect of a white kingside attack. Moreover, having already opened the c- and d-lines,
the position was already good for Black in L.Varnam-J.Houska, British Championship, Torquay 2009.

A) 13 Ne4
This idea, whilst tempting, is not really very attractive for White.
13 ... Ngf6!?
Provoking White to play his planned 14 Nd6+, so why does Black encourage this check? There are
many answers. First of all, there are not enough white pieces to break open the position and Black’s
pieces, with the obvious exception of the monarch, are very harmoniously placed. I must stress that
although it may appear daunting to play with a king on e7, Black has the very easy and straightforward
plan of centralization. When the king, like the coward that he is, finally scuttles back to g8 then Black
should stand very well.
14 Nd6+ Ke7

15 Nc4
White refrains from taking the b7-pawn in order to keep the middlegame tension.
Perhaps the most interesting try in this variation is to play 15 Nxb7!? Bxd2+! (the immediate 15 ...
Qb6 is not as good because of 16 c3 and White is a touch better) 16 Nxd2 (16 Qxd2 is not so great as
after 16 ... Qb5 White must return the pawn and reach an inferior position; after 17 Nc5 Nxc5 18 dxc5
Rhd8! Black is slightly better due to the better activity of his rooks) 16 ... Qb4:
a) Should White choose to hold on to the pawn with the dubious 17 Qa3?! Qxa3 18 bxa3 Rab8! Black
will follow through with ... c6-c5 to open lines against the weak c2-pawn. Although White remains
material ahead, it is a hollow victory. Not only is White’s queenside pawn structure shattered, but he has
weak pawns on c2, a2, a3 and, of course, not forgetting our old friend on h5. Moreover, he is also behind
in development - not a great start by White. After 19 Rb1 (if 19 Na5 Rhc8! 20 Nab3 and 20 ... c5 anyway)
19 ... c5 20 dxc5 Rhc8 21 Ke2 Rc7 22 Na5 Rxb1 23 Rxb1 Nxc5 the h-pawn will have to fall should
White wish to keep his rook active.
b) With 17 Qb3 White dangles the central d-pawn in a bid to open lines: 17 ... Qxd4 (challenge
accepted!) 18 0-0-0 Rab8! and White’s king is also not entirely comfortable.
 Following 19 Qa3+ c5! (the lifesaving move; our position would be completely untenable if that c-
pawn didn’t exist) 20 Na5 Rhc8 (the point; Black’s rooks are sitting nicely on the b- and c-files, while
covering against the family fork) 21 Nf3 Qb4 (one of the easiest ways to deal with White’s ‘initiative’;
Black can perhaps even get away with such blatant materialism as 21 ... Qxf2 22 Rd2 Qg3 23 Nc4 Nd5
and White still has some way to go to prove his attack) 22 Qxb4?! cxb4 23 Nd4 Rb6 24 Nab3 Rd6 25
Rhe1 Rd5 Black was slightly better in T.Radjabov-V.Ivanchuk, Wijk aan Zee 2008.
15 ... Bxd2+ 16 Nfxd2 Qc7

17 0-0-0
Alternatively:
a) 17 Ne4 Nxe4 18 Qxe4 Rhd8! (the concepts of centralization and king safety are rolled into one) 19
0-0-0 Nf6 20 Qe2 c5 and Black is doing well.
b) 17 Nf3 is a strange move. Black, of course, should play normally: 17 ... Rhd8! 18 0-0-0 c5 (this is
standard play from Black; in fact we should play like this against most of White’s moves) 19 Rh4 Kf8 20
dxc5 Nxc5 21 Qe2 Rxd1+ 22 Qxd1 Rd8 23 Qe2 Rd5 24 Nce5 Qd6! is good, although even after 24 ...
Nfd7 Black had the slightly better position in P.Vachtfeidl-J.Houska, London 2013. The weak h5-pawn is
a liability for White.
 17 ... Rhd8!
Centralization! Black prepares to challenge for control of the d-file once it has been opened.
18 Qa3+
18 Kb1 was played in R.Ekstroem-R.Dautov, Swiss League 2004, and according to Dautov the move
poses no threat at all to Black: 18 ... c5! 19 Nf3 Kf8 (the king artificially castles and heads to g8) 20 Qe2
Nb6!? 21 dxc5 Qxc5 22 Nxb6 axb6!? (the idea behind this ambitious move is to gain control of the d-file)
23 Ne5 Rd5! (the point; Black obtains control of the d-file as 24 Rxd5? drops a pawn to 24 ... Qxd5 25 a3
Qxg2) 24 f4 Rad8 25 a3?! (giving Black a target to attack) 25 ... b5! is slightly better for Black.
18 ... c5

The only way Black can meet the check.

19 Nb3 b6
Supporting the c5-pawn.
20 dxc5 bxc5!?
A new, less drawish idea aimed at keeping the knight on b3 bad. After 20 ... Nxc5 21 Nxc5 bxc5 the
position is firmly equal. Both sides have one exposed pawn and neither can really break open the
position. There is a danger for Black that White may be able to exploit the queenside majority, but as long
as Black doesn’t concede the d-file and continues with the theme of centralization we should be more than
fine: 22 g3 Rab8 (22 ... Kf8 is not quite right ironically; the king is safe on e7 and of greater importance is
the idea of moving the rooks to their ideal squares and the centralization of forces on the d-file) 23 Qf3
(White should probably accept that he has no advantage and play for equality with 23 Rxd8 Rxd8 24 Qf3
Rd5 25 Ne3) 23 ... Rd5! (again we see the theme of centralization) 24 Qe2 Kf8 25 f4 (25 Rxd5 exd5 26
Qe5 Qb7 27 Nd6 Qb6 28 Re1 looks distinctly better than the game) 25 ... Qb7 26 b3 Rbd8. This is
slightly better for Black as possession of the d-file gives him the better game, D.Jakovenko-I.Khenkin,
Moscow 2004.
21 Nd6 Kf8 22 Na5 Kg8 23 Rd2 Nb6 24 Rhd1
 24 ... Nbd5!
Improving over the 24 ... Nfd5 of A.Petrushin-A.Lastin, Voronezh 2005.
25 Ndb7 Rdb8 26 Nxc5 Rb5 27 Nab3 Rc8 28 Nd3
Alternatively, 28 c4 Rxc5.
28 ... Nb6 29 Qa6 Rxh5
Black has excellent play.

B) 13 c3 Be7 14 c4
The main line. Instead, 14 0-0 Ngf6 (attacking the h5-pawn straight away) 15 c4 is dealt with in Line B2,
below.
14 ... Qc7

White has three options here:

B1) 15 d5?!
B2) 15 0-0
B3) 15 0-0-0
B1) 15 d5?!
This line does looks incredibly sharp, but it quickly steers play into an endgame with good chances for
Black.
15 ... cxd5!
It had been previously thought impossible for Black to open the position with the king in the centre,
but we can and should.
16 cxd5 Qd6!

The critical move. The d-pawn is pinned so White is unable to take the pawn on e6 immediately,
while our next move will be 17 ... Ngf6.
17 0-0-0
Alternatively:
a) 17 Qb3 Ngf6! (Black should just continue with his development as the b7-pawn is poisoned) 18
dxe6 (after 18 Qxb7? 0-0 Black will recover at least one pawn, since the h5-pawn is difficult to defend
especially as White cannot castle queenside: for example, 19 0-0 Nc5 20 Qb5 Qxd5 21 Bc3 Rfb8 22 Qe2
Nxh5 and Black picks up a pawn) 18 ... Qxe6+ 19 Qxe6 fxe6.
 I would describe this position as misleading in the sense that upon first appearance it looks like White
is doing quite well, especially with that weak pawn on e6. However, things aren’t so clear-cut: the poor
placement of the knight on g3 and the weak pawn on h5 hang around White’s neck. In fact I once had this
position on New Year’s Day when my opponent thought it would be an easy draw, which is not the case.
All that Black needs to remember is to use those open central lines and maximize the potential of his
knights:
a1) 20 0-0-0 Bc5 (targeting the f2-pawn) 21 Rhf1 Rc8 22 Kb1 0-0 23 Bc3 Nd5 (Black improves his
position is as much as he can; after all, small improvements usually turn into big ones) 24 Bd4 N7f6 kept
an eye on the h5-pawn in E.Van Haastert-D.Fridman, Dutch League 2005.
a2) 20 Rc1 0-0 21 Nd4 is well met with 21 ... Bc5 22 Be3?! (the bishop is too passive here; 22 Bc3
Nd5 23 0-0 Nf4 24 Rcd1 Rad8 is equal, although Black has active play) 22 ... Nd5! 23 0-0 (23 Nxe6 is
not possible due to 23 ... Bxe3 24 Nxf8 Bxc1 and Black is better) 23 ... Nxe3 24 fxe3 Bb6 25 Kh2 Ne5!.
Black relentlessly targets White’s weaknesses, in this case the e3-pawn and exposed king.
a3) 20 Nd4 Kf7 21 0-0-0 Nc5 22 Kb1 Rhd8 23 Nb5 Nce4 24 Nxe4 Nxe4 25 Be3 a6 26 Nc3 Nxc3+
27 bxc3 Rac8 28 Bd4 Rd5 and Black was well on the way to establishing an advantage in D.Spence-
J.Houska, Hastings 2008/09.
b) 17 Qb5 Ngf6! (again Black concentrates on developing his pieces) 18 dxe6 (the b-pawn is still
poisoned as above) 18 ... Qxe6+ 19 Kf1 0-0 20 Re1 Qd5 21 Qxd5 Nxd5 22 Nf5 Bf6 23 Rh3 Rfe8 24
Rxe8+ and a draw was agreed, although it looks like Black has the better of things here with his more
active pieces, A.Berelovich-I.Khenkin, Belgian League 2005.
c) 17 Bc3 Ngf6 18 0-0-0 transposes to the main line.
17 ... Ngf6
Black now threatens to win a pawn with ... Nxd5, so the following play is practically forced.
18 Bc3 Qxd5! 19 Qxd5 Nxd5

Black trades off a wing pawn for a central pawn. Normally in the endgame a wing pawn is worth
more than a central pawn whilst in the opening the reverse is true. However, all rules have exceptions and
this is one of these.
20 Bxg7
With the rather rash-looking 20 Rxd5? White sacrifices the exchange in order to obtain active piece
play, but the sacrifice proves unsound: 20 ... exd5 21 Bxg7 Rh7! (the only move; Black must defend the
important h-pawn) 22 Nf5 Bf8! (the exchange of bishops will effectively release the black rook from its
cage) 23 Re1+ Kd8 24 Bxf8 Nxf8 25 Re7 Rc8+ 26 Kd2 Rc7 27 Rxc7 Kxc7 28 Ne5 Ne6 29 Ng4 Kc6 and
Black was much better in J.Smeets-R.Ovetchkin, Moscow 2005.
20 ... Rh7
Black must defend the h6-pawn even if it means that the rook is out of play for a short while.
21 Bd4
We have reached an endgame where both sides have some advantages. White has a very strong bishop
on d4 and technically speaking a better pawn formation. To counteract this Black must act quickly.
21 ... N7f6!

The knight restricts the scope of the bishop on d4 and the knight on g3, but more importantly also
allows the rook to join the game via g7.
22 Rh4?!
White wishes to prevent the knight hopping into f4 harassing the g2-pawn, but I suspect that this move
is inaccurate. Better is to get the pieces into the game via 22 Kb1, making way for the rook to go to c1 and
then c7. After 22 ... Rg7 23 Rc1 things should be equal.
Instead, 22 Ne5 Rg7 23 Kb1 a6 gets ready to move the black rook from a8.
22 ... Rg7 23 Kb1 Rd8!
Bringing the rook into play whilst at the same time preventing the bishop from taking the a-pawn (24
Bxa7?? loses to 24 ... Nc3+).
24 Rd2?
Black’s position now gathers momentum. 24 Rc1 looks a much better move, trying to get some
counterplay, although even here the kingside pawn weaknesses give Black a slight advantage: 24 ... Rg4!
25 Rxg4 Nxg4 26 Ne5 Nb4 27 Nxg4 Rxd4 28 f3 Bg5 and Black is much better; mating ideas around the
white king prevent White’s rook from becoming active.
24 ... a6 25 a3 Rg4!
 After the exchange of rooks, Black’s knights gain significant entry squares.
26 Rxg4 Nxg4 27 Ne4 Nf4!
Forking g2 and h5.
28 Rc2 Nxg2
Black is much better, E.Rodriguez Guerrero-I.Khenkin, Bajada de la Virgen 2005.

B2) 15 0-0

I still find this idea of castling kingside very dubious. It seems to me on a practical level that the pawn
on h5 will give White headaches and not just in the middlegame, but in the endgame too.
15 ... Ngf6 16 Rfe1
After 16 Qe2 0-0 17 Ne5 Nxe5 (my original recommendation of 17 ... c5 is also good enough,
although more complicated) 18 dxe5 Nd7 19 Bc3 Rfd8 20 Rfd1 a5 21 Rd2 Nc5 22 Rad1 Rxd2 23 Qxd2
Rd8 24 Qc2 Rxd1+ 25 Qxd1 b5 Black has a small edge and easy play following 26 cxb5 cxb5 27 Ne2
Nd7.
Instead, 16 Bc3 0-0 17 Rac1 Bd6! 18 Nd2 Bxg3 19 fxg3 Nxh5 saw Black win a pawn in
B.Hjartarson-H.Thorsteinsdottir, Reykjavik 2010, and 16 Qc2 0-0 17 Rfe1 Bd6 has the same concept of
hitting the weak h5-pawn.
16 ... 0-0 17 Nf5

The whole idea behind this line. White brings his bad knight into an attacking role.
Instead, 17 Qe2 is a little slow and 17 ... Rad8 (17 ... Bd6 is an interesting alternative) 18 Bc3 Bd6!
19 Ne5 c5! undermining White’s control of the e5-square:
a) 20 Nxd7?! was a dubious choice by Michael Adams: 20 ... Rxd7 (White has problems with his
knight on g3) 21 d5 (or 21 dxc5 Bxg3 22 fxg3 Qxc5+ 23 Qe3 Qxe3+ 24 Rxe3 Nxh5 and Black is a pawn
up with excellent winning prospects) 21 ... Bxg3 22 Bxf6 Bh2+ 23 Kh1 gxf6 24 g3 (24 dxe6 is bad for
White after 24 ... Rd4 25 e7 Re8 26 g3 Bxg3! 27 fxg3 Qxg3 and Black threatens mate on h4) 24 ... exd5
25 Kxh2 d4 and Black was slightly better in M.Adams-A.Riazantsev, Bastia (rapid) 2005.
b) 20 dxc5! Bxe5 21 Bxe5 Nxe5 22 Qxe5 Qxe5 23 Rxe5 Nd7 24 Re3 (or 24 Ree1 Nxc5 25 Rad1 a5!
26 Kf1 f5, which is a fine little positional move, shutting the g3-knight out from the game and making the
h5-pawn forever weak; after 27 Ke2 Kf7 28 Ke3? f4+! 29 Kxf4 Nd3+ 30 Ke3 Nxe1 31 Rxe1 Black later
won in D.Baramidze-S.Ernst, Dieren 2006) 24 ... Nxc5 was J.Lopez Martinez-N.Borne, Banyoles 2006.
The position is equal, although Black has the better long-term prospects because of his control of the d-
file and the weak h5-pawn.
17 ... Bd6 18 Nxd6
Let’s also check out 18 Nxh6+ gxh6 19 Bxh6. Black should be safe after the calm 19 ... Rfe8 20 Qe3
(20 Ne5 Nxe5 covers all the mate threats) 20 ... Nh7 21 c5 Bf8 22 Bf4 Qd8!, but the position would have
been a lot more dangerous had the white king been castled queenside.
18 ... Qxd6 19 Qb3
19 Ne5? is already a blunder after 19 ... Nxe5 20 Rxe5 Ng4 21 Qg3 Nxe5 22 Bf4 Qe7 and Black won
the exchange in D.Omarsson-G.Kjartansson, Reykjavik 2007.
19 ... a5
Black wants to control the light squares with this move.
20 Rad1 b5!?
 In typical fashion Black plays to undermine the c-pawn to secure the d5-square for a knight.
21 Ne5
Taking the pawn 21 cxb5 cxb5 22 Ne5 (if 22 Qxb5 Rfb8 23 Qc4 Rxb2 and Black wins back the pawn
with the superior position) 22 ... b4 23 Qf3 Rfc8 24 Nxd7 Nxd7 25 Bf4 Qd5 26 Qxd5 exd5 27 Re7 Nf6
results in equality.
21 ... bxc4 22 Qxc4 Rfc8 23 Qe2 Qd5 24 Bf4 Rd8 25 Rd3 Nxe5 26 Bxe5 Nd7 27 Rg3 Nxe5 28
dxe5
A draw was agreed here in L.Dominguez-A.Dreev, World Team Championship, Beersheba 2005.

B3) 15 0-0-0
The main line.
15 ... Ngf6

This is the critical test of the whole variation beginning with 11 ... Qa5+. White is not dissuaded from
castling queenside and instead continues with his standard plan of development. Before we consider the
precise move orders, let’s look at some important themes.
Which pawn to push: ... b5, ... a5 or ... c5?
Well, technically it can be all three. However, the real answer to this question is entirely dependent on
White’s approach; it is not a universal given that Black will merely challenge the centre with ... b7-b5 and
play along the d-line. In fact I have devised a rule:
Rule: Play 16 ... b7-b5 against any white 16th rook move, but if White plays 16 Kb1 (designed to
discourage our ... b5 idea) and starts huffing and puffing with a rook on the e-line, Black usually plays
for ... a7-a5, securing the b4-square for the bishop.
Note too that ... c6-c5 is usually only seen when White has moved the knight from g3 and there is no
danger of a White knight jumping to f5.

Example 1) ... b7-b5

To reach this position, which arises in Line B31, White played 16 Rhe1 and Black has just answered
with 16 ... b5!. This is a very powerful idea - Black aims to induce a positional weakness (the d5-square)
in White’s camp, whether White pushes with c4-c5 or allows an exchange on c4. It’s true that this
weakness is a bit superficial for the time being, but it is a permanent positional weakening - pawns
cannot move backwards! Despite appearances, Black will not be concentrating his forces on the
queenside, but instead aims to initiate play in the centre, targeting the weak d4-pawn.

Example 2) ... a7-a5

White has just played 16 Kb1 0-0 17 Rhe1 to which Black responds 17 ... a5!, securing the b4-square for
the bishop in case white plays Ng3-f5.
 Example 3) the rare ... c6-c5

Here White has played 16 Kb1 0-0 17 Qc2 Rfd8 18 Ne4, letting go control over the f5-square. Black
replies with 18 ... c5 as now 19 d5 is ineffective.

Dealing with White’s g2-g4 Attack

I offer my apologies in advance for this, but I will be discussing this pawn push a lot. This is White’s
most dangerous threat to the safety of our king, so it is particularly important for Caro players to know
when this move is a threat and what to do about it.
If White offers to sacrifice his g-pawn after g2-g4 when is it safe for Black to capture? The answer
lies in whether Black can defend a knight on g4 with ... f7-f5 without losing the e6-pawn.
Here’s an example where it’s safe to capture:
 White has just played the move 19 g2-g4. Here it is safe to capture with 19 ... Nxg4, as following 20
Rg1, 20 ... f5 is possible (see Line B31).
In certain situations Black can even prevent g2-g4:

White has just played 20 Qd3-e2, planning g4. Black counters with 20 ... Ng4!. This is a very
important idea in preventing the g2-g4 pawn advance - it is not easy for White to chase the g4-knight away
and, if a pair of knights is still present, allows a knight on d7 to go to d5.

‘Improving’ Moves and Transpositions

White can utilize many move orders and transpositions, but none that can trick Black out of playing his
chosen variation or force him into an inferior line. Nonetheless, you should still be aware of them. In the
middlegame White has some moves, which I call ‘improving’ moves. White simply wants to improve his
position before beginning aggressive play.
These moves are: Qe2, improving the placement of the queen and removing it from any potential
attacks on the d-file; Kb1, getting the king into safety; and Rhe1 or Rde1, centralizing the rooks. More
often than not the order of these moves is insignificant, as they generally transpose into one another. It is
for this reason I have grouped two moves together as will become clearer.
 Returning to the position after 15 ... Ngf6, we will look at the following options for White.

B31) 16 Rhe1 (and 16 Rde1)

B32) 16 Qe2 and Ne5
B33) 16 Ne4
B34) 16 Kb1 and a rook move
B35) 16 Kb1 and Ne4
B36) 16 Kb1 and Qc2
B37) 16 Kb1 and Qe2

B31) 16 Rhe1
16 Rde1 is not as logical as a white rook is needed on the d-file. After 16 ... b5! 17 c5 0-0 18 Ne2 (or 18
Ne4 Rfd8 19 Qe2 a5 when Black is doing very well, as there is no danger of g2-g4, which is the only
point to White’s play; moreover, Black is exerting a lot of pressure on d4 and he also has ... a5-a4-a3,
weakening White’s queenside even more) 18 ... Ng4 19 Ref1 Rfd8 20 Qc2 a5 actually happened during a
recent game of mine.
 Even though I didn’t remember any published analysis, these moves are easy to find over the board.
The game continued: 21 Nh4 a4 22 Bf4 Qb7 23 f3 Ngf6 24 Kb1 b4 25 Qc4 Nd5 26 Bc1 a3! (a firm
favourite of the computer’s; put a black pawn on a3 and it will automatically give you an advantage!) 27
b3 e5! (standard; Black wants to break the d-file open, which should only be done only at the correct
time, but in this situation White’s counterplay doesn’t arrive in time) 28 Nf5 Bf6 29 f4 exd4 30 g4 d3 31
Qxd3 Nxc5 32 Qc4 Ne4 33 Kc2 Ndc3 34 g5 Qd7 35 Ne3 Qd2+ 36 Bxd2 Rxd2+ 37 Kc1 Nxe2+ 38 Qxe2
Bb2+ 0-1 T.Qiu-J.Houska, British League 2014.
16 ... b5!

I firmly believe this move to be best. Black asks White to make a decision: to either close the centre
with 17 c5 or open lines against his own monarch with 17 cxb5. Instead, 16 ... 0-0 allows 17 Nf5 which
is a completely different ball game.

B311) 17 c5
B312) 17 cxb5

Once when I was preparing against an opponent who had zero games versus the Caro-Kann, I flicked
through this variation testing out random moves. In those thirty minutes of prep I thought what happens if
one plays 17 Ne5? To my surprise, lo and behold my opponent played it and I got a chance to test out my
analysis. After 17 ... 0-0 (as prepared that morning!) 18 Kb1 Nxe5 19 dxe5 bxc4! (now the knight gets to
the central d5-square) 20 Qxc4 Nd5 21 Rc1 Rab8! (Black needs to begin counterplay along the b-line - a
key motif in this line is to swap off the dark-squared bishops) 22 Qxc6 (or 22 Ka1 Rfc8 23 Qg4 Kh8 24
f4 Bb4! and Black has the initiative in this dynamic position) 22 ... Ba3 (the beautiful point) 23 Qxc7
Rxb2+ 24 Ka1 Nxc7 25 Rxc7 Rxd2 Black is doing extremely well.
Instead, 17 Kb1 keeps the central tension and shifts the king to the centre. The problem with these
approaches is that Black gets to utilize the open lines more than White does after 17 ... bxc4 18 Qxc4 Rc8,
with the aim of playing ... c5 and opening the c-file to attack:
 a) 19 Bb4!? Bxb4 20 Qxb4 c5 21 dxc5 0-0! (now that some lines have opened, Black should get his
king to safety as quickly as possible) 22 Rc1 Nxc5 (it is perfectly safe to capture the pawn) 23 Rc4 Rfd8
24 Rec1 Rd5 defends from the side.
b) 19 Ne4 (getting ready to push with g2-g4 and start a kingside attack) 19 ... 0-0 20 Nxf6+ Nxf6 21
g4 c5! (striking back in the centre; 21 ... Nxg4 taking is too dangerous as after 22 Rg1 f5 23 Qxe6+ White
has a very dangerous attack: 23 ... Kh7 24 Rde1 Bd6 25 Qg6+ Kh8 26 Re6) 22 g5 hxg5 23 Bxg5 Qb7
(attacking the knight on f3 and handing control of the c-file to the rook) 24 Ne5? cxd4 25 Qxd4 Rfd8 (now
Black contests the d-file) 26 Qf4 (or 26 Qh4 Rxd1+ 27 Rxd1 Ne4 and Black wins material with 28 Bxe7
Nc3+! 29 Kc1 Nxd1+ 30 Kxd1 Qd5+) 26 ... Rxd1+ 27 Rxd1 Ba3 28 b3 Nd5 29 Qf3 Nc3+ and wins.

B311) 17 c5

White ‘locks down’ the queenside. First of all, I must stress that this is a huge decision by White.
Black will now be banking on the long-term pressure against the d4-pawn. Indeed, just imagine the
scenario where we swap a set of minor pieces and then triple on the d-line - delicious! In order to avoid
this nightmare situation White needs to rush his kingside play and open up lines.
17 ... 0-0
 There is now a series of knight moves.
18 Ne4
18 Nf5 is tactically tempting, but flawed: 18 ... exf5 19 Rxe7 Nxc5! (Black wins a pawn) 20 Qe2
Ncd7 21 Nh4 Nd5! (picking up an exchange; instead, after 21 ... Qd6 22 Nxf5 Qd5 the game ended in a
draw in P.Leko-A.Karpov, 1st rapid matchgame, Miskolc 2006; it is interesting to note that in the rest of
this match between Anatoly Karpov and Peter Leko, the Hungarian GM never once opened with 1 e4
again) 22 Nxf5 Nxe7 23 Qxe7 (or 23 Nxe7+ Kh8 24 Nf5 Nf6 and Black is doing very well) 23 ... Qd8 24
Qe4 Nf6 and taking control of one of the key central squares leaves the c5-pawn irreparably weak.
Likewise, after 18 Ne5 Nxe5 19 dxe5 Nd7! 20 Ne4 (20 Bxh6 blunders a piece to 20 ... Nxc5!) 20 ...
Nxc5 21 Nxc5 Bxc5 22 g4 (trying to begin an attack) 22 ... Bxf2 23 Rf1 Qxe5 24 Bxh6 (24 Rxf2 is a
massive blunder as 24 ... Qc5+ wins on the spot) 24 ... Rfd8 25 Qc2 Rxd1+ 26 Rxd1 Bc5 Black is
winning.
18 ... Rfd8!
I wish to preserve the rook on a8 for some potential attacking ideas on the queenside. Black in no way
should help the kingside attack by swapping knights - let White do that job himself.
19 g4
Instead, 19 Nxf6+ Nxf6 20 Qe2 Ng4! (20 ... Nxh5 21 Ne5 Nf6 22 Bxh6 gxh6 23 Qf3 could be
awkward for Black to defend) is a key concept to remember.

Black blocks the extremely dangerous g2-g4 pawn storm in one sweep. In a similar position with the
queen’s rook on e1, the Scottish GM Jonathan Rowson agrees with this idea: prevention is better than the
cure. Now 21 g3 (or 21 Kb1 Rd5! when Black’s plan is very positional, namely to double on the d-file
and pile on the pressure on the backward d-pawn; instead, chasing away the knight with 21 Ng1? fails to
21 ... Rxd4, while 21 Qe4? loses to 21 ... Nxf2) 21 ... Rd5 22 Bf4 Qd7 23 Nd2 e5 24 dxe5 Bxc5 looks
unpleasant for White. Note too that 19 Qe2 Nxe4 20 Qxe4 Nf6 21 Qe2 Ng4 would be a transposition.
19 ... Nxg4
 Safest and the best. Otherwise g4-g5 looks unpleasant for Black.
20 Rg1
After 20 Qb3 Ngf6 21 Nxf6+ Bxf6 Black is okay.
20 ... f5 21 Qb3 Nf8
Of course the e6-pawn needs to be defended.
22 Ng3 Kh8
White was threatening Nxf5.
23 Ne2
White sacrifices another pawn to bring the bishop into play and avoids 23 Ne5 Nxe5 24 dxe5 Bxc5 25
Ne2 Bxf2 26 Rgf1 Qxe5.
23 ... Bf6 24 Bf4 Qf7 25 Ne5 Bxe5 26 Bxe5 Nxe5 27 dxe5 Qxh5
Black has won a pawn and White will find it difficult to commence an effective enough attack to
justify the material deficit.

B312) 17 cxb5!

I believe this to be the absolutely critical try. Indeed, nothing else is dangerous.
 17 ... cxb5+ 18 Kb1 0-0
Maintaining the dynamism in Black’s set-up. Alternatives are:
a) 18 ... Qc4!? 19 Qxc4 bxc4 leads to a complex endgame. The c4-pawn is in a very strange and
unique situation; Black either reinforces the defenders of it or counterattacks against the d4-pawn.
b) 18 ... Qb7 is unsatisfactory due to 19 Nh4 0-0 20 Ngf5!! and then:
b1) 20 ... Bd8 21 Bxh6! exf5 (21 ... Nxh5 is the only try, but things become unpleasant for Black after
22 Bxg7 Nxg7 23 Nh6+ Kh8 24 Rh1 and White’s attack is irresistible) 22 Nxf5 Nxh5 (or 22 ... Qd5 23
Bxg7 and White is winning) 23 Bxg7 Nxg7 24 Nxg7 Kxg7 25 Qg3+ Kf6 26 Re3 Bc7 27 Qh4+ Kg6 28 f4
and Black will not survive with his king so exposed, such as after 28 ... Qxg2 29 Re2!.
b2) 20 ... exf5 21 Nxf5 Bd8 22 Bxh6 and Black is completely uncoordinated against White’s attack.
19 Nf5

19 ... exf5!
Necessary. 19 ... Rfe8 20 Nxe7+ Rxe7 21 Qxb5 Rb8 was better for White in the important game
V.Akopian-D.Solak, Dresden Olympiad 2008.
20 Rxe7 Qd6 21 Re2
After 21 Re3 f4 22 Re2 Qd5 23 Ne5 Rfe8 24 Nxd7 Nxd7 25 Rxe8+ Rxe8 26 Bxf4 Qxg2 Black’s
active pieces enable him to comfortably maintain equality.
21 ... Qd5 22 Nh4 Rae8
 Defending the pawn in an unusual way. The point is that 23 Rxe8 Rxe8 24 Nxf5 Ne4 regains the pawn.
23 Rde1 Re4 24 Nxf5 Qxf5 25 Rxe4 Nxe4 26 Qxe4 Qxf2 27 Re2 Qf1+ 28 Re1 Qf2 29 Re2
Here a draw was agreed in V.Rojicek-B.Lalith, Pardubice 2012.

B32) 16 Qe2

White begins measures to pre-empt the opening of the d-line whilst keeping open options of both Nf3-
e5 and Ng3-e4 with g2-g4. Black, for our part, should carry on as normal.
16 ... 0-0 17 Ne5
If 17 Ne4 we can follow the rule - there is no danger of a white knight jumping to f5 so we can play
17 ... c5!. Following 18 Nxf6+ Bxf6! (putting pressure on the centre) 19 d5 (or 19 g4 cxd4 20 g5 hxg5 21
h6 Rfc8 22 Kb1 and at the very least we have 22 ... gxh6 23 Rxh6 Qxc4 when the queens must come off)
19 ... exd5 20 cxd5 Rfe8 21 Be3 c4 White is getting into trouble.
White can also wait with 17 Kb1 - see Line B37 below.
17 ... b5!?
Suggested by GM Lars Schandorff, this move seems more straightforward than my original
recommendation of 17 ... c5, challenging the centre right away. Here 18 Ng6! is, of course, the most direct
try: 18 ... Rfe8 (not 18 ... fxg6? 19 Qxe6+ Kh8 20 hxg6 and suddenly all sort of unpleasantries are
happening on h6) 19 Nxe7+ Rxe7 20 d5! (20 Nf5? fails to 20 ... exf5 21 Qxe7 Re8 and the white queen is
trapped; an important idea to remember) 20 ... Qd6 21 Rhe1 Rae8 and here White should play 22 Qf3! (in
my original book I only considered 22 Be3? which after 22 ... exd5 23 Nf5 Qa6 gives Black a very
dangerous initiative) 22 ... Qa6 23 d6 Qxd6 24 Bxh6 Qa6 25 Bg5 Qxc4+ 26 Kb1, which leads to a very
unclear position.
18 Ng6 Rfe8 19 Nxe7+ Rxe7

20 Kb1!?
Others:
a) 20 c5 hands Black all the long-term chances after 20 ... Ree8! 21 Kb1 a5 22 Ne4 Nxe4 23 Qxe4
Nf6 24 Qf3 Nd5 25 g4 f6, which is more than good enough.
b) 20 cxb5 cxb5+ 21 Kb1 Ree8 22 Qxb5 Rab8 23 Qd3 Qb7 24 Bc3 Rec8 has the decisive threat of ...
Nfd5.
c) 20 Nf5 Ree8 21 Nxh6+ (retreating with 21 Ne3 leads to the opening of the centre with 21 ... e5) 21
... gxh6 22 Bxh6 bxc4 and White is not fast enough on the kingside as moves such as 23 g4 lose to the
counterattacking 23 ... c3.
20 ... bxc4 21 Qxc4 Nb6 22 Qc2 Rd7
Black has a pleasant position with good squares for his knights, open lines against the white king and
pressure towards the weak d4-pawn.

B33) 16 Ne4
 Now that the knight has come directly to e4 and White has not spent time on making little moves to
improve his position, it is not so wise for Black to castle directly into a brewing storm. Instead, I like the
approach used by Dautov in his game against Carmeille, which aims to resolve the issues in the centre
before castling.
16 ... Rd8!
Black leaves his king in the centre and concentrates on the strategic break ... c5.
17 Kb1
Important alternatives are:
a) 17 Qe2 can be even met with direct action: 17 ... b5 (17 ... c5 18 d5 Nxe4 19 Qxe4 Nf6 20 Qe2
exd5 21 Rhe1 dxc4 22 Nh4 is very unclear) 18 Nxf6+ Nxf6 19 c5 Rd5 20 g4 Nh7 21 Kb1 0-0 22 Qe4
Rfd8 23 Bf4 Qc8 with good play on the d-line for Black.
b) 17 g3 was played against me in a rapid game once. I don’t remember how the game exactly
continued, but I know I got into trouble so it is useful to know how to handle this line. We should go 17 ...
Nxe4 18 Qxe4 Nf6 and then:
b1) To 19 Qe2 Black should react with 19 ... b5 20 c5 (20 Kb1 bxc4 21 Rc1 Rd5 22 Qxc4 0-0 23
Qxc6 Qxc6 24 Rxc6 Rf5 recovers the lost pawn) 20 ... a5 21 g4 a4 22 g5 hxg5 23 Bxg5 Nd5 24 Bd2 Bf6
and Black’s king is safe in the centre.
b2) 19 Qc2 0-0 (now that the queen is not looking at the e6-pawn Black can castle safe in the
knowledge that g3-g4 is not dangerous) 20 Bf4 Bd6! and Black is doing well.
c) 17 Nxf6+ Nxf6 with a further divide:
c1) As seen in previous examples, 18 Qc2 is not dangerous for Black: 18 ... 0-0! 19 g4 (this doesn’t
work with the queen on c2; instead, 19 Rh4 plans to play Bf4 and g2-g4, but the strategic breakthrough 19
... c5! gives Black a tiny edge, H.Bohm-Y.Seirawan, Wijk aan Zee 1980) 19 ... Nxg4 20 Rhg1 f5 21 Ne5
Rxd4 and White’s position was falling apart in R.Schnabel-A.Braun, Berlin 2009.
c2) 18 Qe2 is more to the point.
 Black should react in the following way: 18 ... b5 (please don’t castle yet as g2-g4 is really a very
dangerous threat) 19 Kb1 bxc4! 20 Rc1 0-0 21 Qxc4 Rc8 (Black has good squares for his minor pieces)
22 Ne5 c5 23 Qa6 Bd6! with few problems ahead.
17 ... c5

This was my original recommendation and I will be sticking to it. I actually played it in a weekend
tournament in Harstad, remembering the rule that ... c6-c5 is more attractive once White has moved the
knight from g3.
18 Bc3
18 d5?! is not as dangerous as it appears: 18 ... exd5 19 cxd5 Nxe4! 20 Qxe4 Nf6 21 Qa4+ (or 21
Qc4 Rxd5 22 Bc3 Rxd1+ 23 Rxd1 0-0 and Black is a pawn up) 21 ... Qd7 22 Qxd7+ Rxd7 23 Bc3 Rg8
24 Bxf6 Bxf6 25 d6 Kd8 26 Rhe1 Kc8 27 Rc1 b6 28 b4 Rxd6 29 bxc5 bxc5 30 Rxc5+ Kb7 with equality.
18 ... cxd4
A logical choice.
19 Bxd4
After 19 Nxd4 a6 White gets nowhere.
19 ... 0-0 20 b3
 Instead, 20 Nxf6+ Bxf6 21 Qe2 Bxd4 22 Rxd4 Nf6 reaches the typical Caro ending that I talked about
in the endgame section above.

White has to care for two structural weaknesses, c4 and h5. I.Ionica-J.Houska, Harstad 2014,
continued 23 Rxd8 Rxd8 (securing the d-file) 24 Ne5 Rd4 25 f3 Qd6 26 Kc2? b5! (opening lines against
the white king) 27 cxb5 Nd5 (bringing a new piece into the attack) 28 Nc6? Nf4 29 Qe1 (29 Qe5 Rc4+ 30
Kb1 Rxc6 wins material) 29 ... Nxg2 30 Qe2 Nf4 31 Qe1 Rd5 32 a4 Rxb5 33 axb5 Qd3+ 34 Kc1 Ne2+
0-1.
20 ... a6
Getting ready to play ... b7-b5.
21 Qe3 b5!
Black softens up the white queenside.
22 Nxf6+ Bxf6 23 cxb5 axb5 24 g4
White must begin active operations on the kingside otherwise he will simply be crushed.
24 ... Rc8 25 Rc1 Qb7!

Instead, 25 ... Qd6 was played in L.Carmeille-R.Dautov, German League 2005, which continued 26
Rhd1 Rxc1+ 27 Qxc1?! Qd5 28 Bxf6 Qe4+ 29 Qc2 Nxf6 30 Qxe4 Nxe4. Black has the preferable
position and now White made a massive blunder, probably in time trouble: 31 Nd4?? (31 Rd7 leads to a
very complicated ending, although Black probably has good chances) and the game finished on the spot
with 31 ... Nc3+.
26 Rhd1 Rxc1+ 27 Rxc1 Qd5 28 Bxf6
Again, after 28 Rc7 Rd8 White is under pressure in the centre.
28 ... Nxf6 29 Rc5 Qa8
The weak g-pawn combined with the exposed white king means White is in trouble. Remember the six
factors I mentioned in heavy piece endgames - king safety is of the utmost importance.
30 g5 hxg5 31 Rxg5? Rc8!! 32 h6 Ne4
With the deadly threat of ... Nc3+.
33 Rxg7+ Kf8 34 Ka1 Rc2 35 a4 bxa4 36 Rxf7+ Kxf7 37 Qxe4 Qxe4 38 Ng5+ Kg6 39 Nxe4 axb3
Black is winning.

B34) 16 Kb1
An important little move. White prepares his army for the pawn advance ... b7-b5.
16 ... 0-0 17 Rhe1

Here we will look at the combination of Kb1 plus a rook move in the centre. White’s plan is clear: he
wants to jump with the knight to f5, so it is imperative that we make a hiding place for the bishop on b4.
17 Rde1 is inaccurate as Black’s counterattack on the d-line is all the more powerful once White has
shifted a rook from it. After 17 ... a5! White might try:
a) 18 Nf5 Bb4 when the situation is not entirely easy for Black after 19 Nxh6+ (if 19 Bxh6 exf5!) 19
... gxh6 20 Bxh6 going for mate. However, Black should be greedy and grab the available material: 20 ...
Bxe1 21 Qe3 (after 21 Rxe1 Kh8 22 Bxf8 Rxf8 23 Qe3 b5 24 Ne5 Nxe5 25 dxe5 Nh7 Black is defending,
as he is after 21 Ne5 Bxf2 22 Qf3 Kh7) 21 ... Nh7!
 22 Rxe1 (there isn’t even a threat after 22 Ng5 so Black can play 22 ... Bb4 23 Rh3 Bd6, winning) 22
... Rfc8 (the knight on h7 is doing a sterling defensive job) 23 g4 Qd6 24 g5 Nhf8 25 g6 fxg6 26 hxg6 Qe7
27 Rh1 Qf6 28 Bf4 c5 29 g7 Ng6 and it is difficult to see White’s initiative: for instance, 30 Bd6 cxd4 31
Qh6 Qxg7 32 Qh5 Rf8! 33 Rg1 Rf5 would be beautiful defensive play.
b) 18 Ne5 is met with 18 ... Nxe5 19 dxe5 and now there is a key move: 19 ... Rfd8! 20 Qe2 Nd7 21
f4 a4 22 Be3 b5! (22 ... a3 is also satisfactory for Black, although I later lost in R.Britton-J.Houska,
British Championship, Sheffield 2011).
Instead, 17 Bc3 a5 18 Ne5 Bb4 19 f4 Rfd8 20 Ne4 Nxe4 21 Qxe4 Bxc3 22 bxc3 a4 is slightly better
for Black.
17 ... a5!

Giving the bishop a hiding place on b4.

18 Ne5
Black’s point is that 18 Nf5 Bb4! 19 Nxh6+ gxh6 20 Bxh6 Bxe1 21 Qe3 Nh7 covers the threats, while
a model game for Caro-Kann players is the following: 18 Bc1 Rfd8 19 Qc2 a4 20 Ne5 Nxe5 21 dxe5
Rxd1 22 Qxd1 Nd7 23 f4 Nc5 24 Ne4 Rd8 25 Qc2 Nxe4 26 Rxe4 Qb6 27 Qxa4 Qg1 28 Qc2 Rd1 29 Re2
Rf1 30 a3 Bc5 31 Qd2 Rd1 32 Qc2 Bd4 33 c5 Rf1 34 Rd2 Rxf4 35 Ka2 Rf2 and Black later won a very
smooth game in P.Leko-V.Ivanchuk, Morelia/Linares 2008.
18 ... Nxe5!

The easiest move. Instead, 18 ... Bb4 gives White the opportunity to play 19 f4!.
19 dxe5 Rfd8!
This intermediate move is equally important as Black must start fighting for control over the d-file.
20 Qe2
20 Qc2 is inferior due to 20 ... Ng4 21 Bc3 Bb4.
20 ... Nd7 21 f4 Nc5 22 Be3 a4 23 Bd4 a3

Weakening the dark squares for some future counterplay.

24 b3 Qa5 25 Bxc5
Sidestepping 25 Qe3 Nxb3!.
25 ... Bxc5 26 Ne4 Bb4 27 Rxd8+ Rxd8 28 Rd1 Rxd1+ 29 Qxd1 Qb6 30 Qe2 Qd4
Black stood very well in D.Svetushkin-F.Berkes, German League 2009.

B35) 16 Kb1 0-0 17 Ne4

 As there are no threats in the position I have to say I rather like this, Schandorff’s recommended way
of playing.
17 ... Rad8
17 ... b5 is the move I originally recommended. White should probably play 18 Nxf6+ Nxf6 19 Rc1,
which is a new idea by Gerzhoy. Here I like 19 ... Qb7 20 Ne5 bxc4 21 Qxc4 c5 22 dxc5 Ne4 23 Rc2 (23
Be3 Bxc5 maintains equality) 23 ... Nxc5 and Black has no problems.
18 Nxf6+
Ironically the best thing for White’s attack is to swap off a set of knights to (a) make a potential knight
on g4 more unstable, and (b) reduce the number of defenders around the black king. Other moves are:
a) 18 Qe2 Nxe4 19 Qxe4 Nf6 20 Qe2 transposes to the main line.
b) 18 g3 is not as impressive as it looks: 18 ... b5 19 Bf4 Qb7 20 Nxf6+ Nxf6 21 Ne5 c5 22 cxb5
Rxd4 and Black had seized the initiative in Ivanchuk-Anand, Monaco (blindfold) 2009.
c) If 18 Rdg1 Black can stop the action with 18 ... Ng4! and there is also the more enterprising 18 ...
c5!? 19 g4 Ne5 20 Nxe5 Rxd4 21 Nxf6+ Bxf6 22 Qe2 Qxe5 when Black has the advantage.
d) Of course, we also have to consider the ultra-aggressive 18 g4. We should grab the pawn: 18 ...
Nxg4 19 Qe2 (19 Rhg1 is met with the routine defensive move 19 ... f5 20 Nc3 Bf6 and now 21 Qe2 is
countered by 21 ... Rfe8 when there is no clear way for White to continue his attack) 19 ... Kh8.
 This is handy prophylaxis against any capture on e6 with check. Here White can try:
d1) 20 Rdg1 f5 21 Nc3 e5!? 22 dxe5 Rfe8 23 Bf4 Bf6 24 Qc2 (or 24 exf6 Qxf4 25 fxg7+ Kg8 and
White loses material) 24 ... Ndxe5 25 Qxf5 Qb6 26 Rxg4 Nxg4 27 Qxg4 Bxc3 is rather good.
d2) 20 Rhg1 is a better try, although after 20 ... f5 21 Nc3 Rde8 22 Be3 Bf6 23 Qd2 e5 24 d5 e4 25
Nd4 Nde5 26 Nxf5 Nf3 Black is still winning.
18 ... Nxf6 19 Qe2
The immediate 19 g4 again doesn’t quite work as Black can easily protect his knight on g4: 19 ...
Nxg4 20 Rhg1 f5 21 Qe2 Bf6! 22 Qxe6+ Rf7 23 Qxf5 (after 23 Rde1 Qd7 Black has consolidated his
position) 23 ... Nxf2 24 Rde1 (if White throws the contents of the kitchen sink at us, we should remain
calm and grab the material offered: 24 Bxh6 Nxd1 25 Qg4 Qd7!) 24 ... Bxd4 25 Qe6 Bxb2! 26 Kxb2
Nd3+ 27 Kc2 Nxe1+ 28 Rxe1 Qc8 and Black has the advantage; the two minor pieces are insufficient
compensation for the rook plus two pawns.
19 ... c5 20 g4 Qc6!?

This is a rather interesting shot, sacrificing material for play against the white king. Instead, 20 ...
cxd4 21 g5 hxg5 22 h6 is complicated.
21 d5
 The point is that 21 g5 Qe4+ is rather strong: 22 Qxe4 Nxe4 23 gxh6 Nxf2 24 Rhg1 Nxd1 25 hxg7
Rfe8 26 h6 Nf2 and the two pawns are going nowhere.
21 ... exd5 22 g5
After 22 Qxe7? Rfe8 again the white queen is trapped on e7.
22 ... Ne4 23 gxh6 Rfe8! 24 Rdg1
I’ve also analysed:
a) 24 hxg7 Bg5! 25 cxd5 (25 Bxg5 Nc3+ 26 bxc3 Rxe2 27 Bxd8 dxc4 28 Nd2 Rxd2 picks up
material) 25 ... Rxd5 26 Be3 Bxe3 27 Rxd5 (or 27 fxe3 Ng3 28 Qg2 Nxh1 29 Rxh1 Qh6 and Black is
safe) 27 ... Nxf2! 28 Rhd1 Nxd1 29 Rxd1 c4 with a complex position where Black should concentrate on
cracking open the white king position and blockading the kingside pawns.
b) 24 cxd5 Qxd5 25 Bc1 Qf5 26 Ka1 Rxd1 27 Rxd1 Bf6 28 hxg7 Qxh5 29 Qb5 Rb8 30 Qe2 Re8 31
Qb5 Re7 and Black does not stand worse.
24 ... Bf6 25 Be3
25 hxg7? Ng3 nets material, as shown by 26 Qd1 Nxh1 27 h6 Nxf2 28 h7+ Kxh7 29 Rh1+ Kg8 30
Rh8+ Kxg7 and Black is winning.
25 ... d4 26 hxg7 dxe3 27 h6 Kh7

The danger points are covered so Black is fine for the time being.
28 fxe3 Qd7 29 Rh5 Qd3+ 30 Qxd3 Rxd3 31 Ne5 Bxe5 32 Rxe5 Rdd8! 33 g8Q+ Rxg8 34 Rxg8
Nd2+ 35 Kc1 Kxg8
Only Black has chances to win here.

B36) 16 Kb1 0-0

Now we take a look at Kotronias’s invention:
17 Qc2
 White’s play is designed to prevent our ... b7-b5 idea. This new line captured white players’
imagination for a while. After all, it is complex and caries great risk for both sides.
Here are some pointers to guide you:
Step 1: Prepare for the opening of the c- and d-files by first playing ... Rfd8. This is vital as it may be
necessary to play ... c6-c5 before playing ... Rac8.
Step 2: When the g3-knight finally lifts off only then should we strike out with ... c5. This is because
after White plays d4-d5 the last thing Black wants is a white knight to hop into f5.
Step 3: Grab material and where possible start pushing the c-pawn - the further the better.
17 ... Rfd8
Now that ... b7-b5 has been prevented Black changes tack and plays for ... c6-c5. White can continue
in various ways: the knight on g3 can hop out of the way in order to prepare the standard kingside pawn
push, or White can play the tricky looking Rh1-h4. This has the idea of beginning a piece attack on the
kingside and with the added benefit that after ... c6-c5 any exchange on d4 can be met with Rhxd4.
17 ... Rfe8 is also possible, but I prefer the text move.

Now White has:

B361) 18 Ne4 et al
B362) 18 Bc1

B361) 18 Ne4
This leap always needs to be taken very seriously in this variation, but here it is not so dangerous. As
such, White might prefer:
a) 18 Rh4 Rac8 (there are no immediate threats so Black should simply build up to ... c6-c5) 19 Ne2
(19 Bf4 Bd6 20 Bxd6 Qxd6 is slightly better for Black; the more pieces that are exchanged, the less
attacking potential White has, and here I like Black’s position) 19 ... c5 20 d5 (after 20 Rc1 Black should
simply wait and see with 20 ... b6 when it’s not clear what White is actually doing) 20 ... exd5 (of course
we should be greedy) 21 Ng3 (if 21 Bf4 Bd6! or 21 g4 Nb6 22 g5 Ne4 23 Bf4 Qd7 when Black has a big
advantage; after 24 cxd5 hxg5 25 Qxe4 gxh4 he should be able to fend off White’s initiative) 21 ... d4 22
Nf5.

When the knight comes to f5 Black needs to defend the h6 and g7 points in anticipation of a piece
sacrifice. As such 22 ... Bf8! is indicated.
b) 18 Rde1 doesn’t really fit in with the main idea behind this variation. Black should as usual
continue in the standard fashion with 18 ... a5 (securing a square for the bishop) 19 Ne5 Bb4 and here we
have:
 b1) 20 Bxh6 is an interesting sacrifice: 20 ... gxh6! (20 ... Bxe1 is too risky in view of 21 Nxf7 Nf8 22
Nxd8 and the position is unclear) 21 Nxf7! Nf8! (defending all the entry points) 22 Nxd8 (after 22 Nxh6+
Kh8 23 Rd1 Qg7 24 Qc1 White is a long way from advancing his kingside pawns) 22 ... Bxe1 23 Nxe6
Nxe6 24 Rxe1 Nxd4 25 Qg6+ Qg7 26 Qxg7+ Kxg7 27 Re7+ Kf8 28 Rxb7 Re8 with a complicated
endgame in prospect.
b2) After 20 f4 Black should play 20 ... Nxe5! (best, although there’s also 20 ... c5 21 Nxf7 Kxf7 22
Qg6+ when Black can play the calm 22 ... Kg8, as 23 f5 leaves the white queen embarrassed after 23 ...
Nf8) 21 fxe5 and now there is the fantastic 21 ... Rxd4! 22 Bxb4 (or 22 Be3 Rg4 23 exf6 Bxe1 and
White’s attack is nothing) 22 ... axb4 23 exf6 Qxg3 when Black’s pieces are extremely active.
c) The knight retreat 18 Nf1 needs to be countered by the forward approach: 18 ... c5! 19 Ne3 (or 19
d5 exd5 20 Ng3 d4 21 Nf5 Bf8 and White is getting nowhere after, say, the aggressive 22 g4 Qc6 23 N3h4
Re8 when Black will exchange the queens) 19 ... cxd4 20 Nxd4 Nc5 21 Bc1 Qe5 22 Ne2 Nce4 23 f4 Qc7
and White’s attack is not really going anywhere.
18 ... c5!

19 g4
Slowing Black down with 19 d5 also needs to be taken into account: 19 ... exd5 20 Nxf6+ Nxf6 21 g4
Nxg4 22 Rhg1 (22 Qf5 Nxf2 23 Rhg1 Qc8 24 Qe5 Bf8 25 Bxh6 Ng4 saves the day) 22 ... Qc8 23 Rg3 d4
24 Rdg1 d3! (it is important to cut off the queen’s access to the kingside) 25 Qb3 f5 26 Ne5 Bf6 (Black
has a good defensive set-up with great counterattacking possibilities) 27 Nxd3 Qc6. Here 28 f3 Nf2 29
Bc3 may look very strong, but 29 ... Nxd3 30 Bxf6 Qxf6 gives Black the better position: for example, 31
Rxg7+ Qxg7 32 Rxg7+ Kxg7 33 Qxb7+ Kf6 34 Qc6+ Kg5 35 Qg6+ Kf4 and Black is winning.
19 ... cxd4 20 g5 d3!

21 Qxd3 Nb6 22 Nxf6+

Black can also be happy after 22 Qe2 Nxe4 23 Qxe4 Qxc4 24 Qxc4 Nxc4 25 Bc1 hxg5.
22 ... gxf6! 23 Qc3 fxg5 24 b3
24 Nxg5 is nothing to be scared off. Black needs to keep a cool head and play 24 ... hxg5 25 Bxg5 (25
h6 e5 blocks everything up) 25 ... Bxg5 26 Rdg1 f6 27 Qxf6 Rd1+ 28 Kc2 and now the neat trick 28 ...
Rc1+ wins the game.
24 ... Rac8 25 Rdg1 Qc5
White’s attack has been halted and Black is much better.

B362) 18 Bc1
 This is White’s main continuation. He still has the plan of advancing the kingside pawns, but first of
all prepares for Black’s plan of ... c6-c5.
18 ... Rac8 19 Ne2
If 19 Ne4 c5!, but White can also play another knight move in 19 Nf1 c5! 20 d5! (White usually tries
to slow down Black’s central counterplay by donating the d-pawn; after 20 g4 cxd4! 21 g5 hxg5 22 h6
Ne5 Black is doing well) 20 ... exd5 21 g4 and here:
a) When I faced this position I wished to neutralize White’s activity by playing 21 ... Ne5, but White
can keep his initiative if he plays accurately: 22 Nxe5 Qxe5 23 Ng3! (23 Ne3? dxc4 24 Nf5?! Rxd1 25
Rxd1 Rd8 and Black was already much better in R.Schnabel-J.Houska, French Team Championship 2011)
23 ... dxc4 24 Rde1 Qc7 25 g5 and the position is still very dangerous for Black.
b) 21 ... Nxg4 is probably the most accurate approach: 22 Rg1 (if 22 Rxd5 Ndf6 23 Rf5 Qd6 24 Ne3
Nxe3 25 Bxe3 Qd3 when White is material down and the queens will come off) 22 ... Ndf6 23 cxd5 c4 24
Ng3 and here 24 ... c3! is very strong.
19 ... c5 20 d5 exd5 21 g4 Nxg4 22 Rdg1 Ndf6 23 Nh4 d4

24 Nf4
24 Nf5? loses to 24 ... d3 25 Qd2 dxe2 26 Qxe2 and now Black must sidestep the danger with 26 ...
Kf8 27 f3 Qe5 28 Qc2 Nf2 when sacrificing with 29 Rxg7 simply allows Black to capture material on h1.
24 ... Kh8!
Instead, 24 ... Qd7 25 f3 d3 26 Qd2 Ne5 27 b3 Ne8 28 Nd5 Bxh4 29 Rxh4 Qd6 30 Qg2 was very
unclear in V.Kotronias-A.Braun, Dresden Olympiad 2008.
25 Nf5
25 Qf5 Qd7! is another important point behind Black’s play.
25 ... Bf8 26 f3 Ne3 27 Nxe3 Qxf4 28 Nd5
28 Ng4 Qd6 29 Qf5 Nxg4 30 fxg4 also looks good for Black.
28 ... Qd6 29 Bf4 Qc6 30 Be5
Instead, 30 Bxh6? Nxd5 wins instantly.
30 ... Nxd5 31 cxd5 Rxd5
White has absolutely nothing here.

B37) 17 Qe2
 Here we see another approach. White removes the queen from the potentially tricky d-line and plays
for the advance of the g-pawn. As we know, this is a very dangerous idea and one that every Caro player
must have a concrete answer to.
17 ... b5!?
Sticking to the spirit of my original book. I encountered this line against Women’s World Champion
Hou Yifan at the World Sports Mind Games in 2008. There, as we will see, I simply blockaded White’s
attack. It was a solid approach and despite me blundering twice (!), the black position was still solid
enough to hold and eventually I won on the counterattack.
The formula with the text is to blockade White’s kingside attack, counterattack on the d-line and
establishing a pawn wedge of a4-b5-c6.
18 c5
After 18 cxb5 cxb5 19 Nf5 exf5 20 Qxe7 Rfe8 21 Qa3 Qb7 Black is very active, while 18 Ne5 bxc4
19 Ng6 Rfe8 20 Nxe7+ Rxe7 21 Qxc4 Nb6 22 Qc2 Qd6 again looks balanced.
18 ... Rfd8 19 Ne4
We should discuss too 19 Nf1 Nf8 20 Ne3 Nd5 21 Bc1 a5 22 g4 Nh7, blockading the white pawns.

After 23 Rhe1 (or 23 Ne5 Bf6 24 Nxd5 Rxd5 25 f4 Rad8 and Black’s counterplay down the d-file
offsets White’s attack on the kingside) 23 ... a4 24 Ne5 we have:
a) 24 ... Bf6 25 Nxd5 Rxd5 26 Bf4 Qb7 (26 ... a3! was a better move as there are no good discovered
attacks on the queen) 27 Bg3 Rad8 28 Qe3 b4 (28 ... Bxe5? was my choice in Hou Yifan-J.Houska,
Beijing (rapid) 2008) 29 f4 a3 30 b3 Bxe5 31 fxe5 Qd7 with strong pressure against d4.
b) 24 ... a3 is always a good idea: 25 Nxd5!? (the sacrifice is not sound: 25 Nxf7 Kxf7 26 Nxd5 exd5
27 Qe6+ Kf8 28 Bxh6 and Black can simply play 28 ... Bf6 29 g5 Re8 30 Qf5 gxh6 31 gxf6 Qf7 32 Re7
Rxe7 33 fxe7+ Ke8 34 Qg6 Qxg6+ 35 hxg6 Nf6 36 Re1 Kd7, winning) 25 ... Rxd5 26 b3 Bf6 27 f4 Rad8
and Black is doing fine, such as after 28 Bxa3 Rxd4.
19 ... Nxe4 20 Qxe4 Nf6 21 Qe2

21 ... Rd5!
Solid and best. Unfortunately after 21 ... Ng4 White can begin to dislodge the black knight with 22
Rh3! (exploiting the absence of the rook from e1; instead, 22 Ne5 is answered by 22 ... Nxe5 23 Bf4 f6 24
dxe5 Bxc5 25 Rxd8+ Rxd8 26 Qg4 f5 and Black is just much better) and now:
a) 22 ... e5 is what I played, but after 23 Rg3 f5 24 Nxe5 Rxd4 25 Nxg4 fxg4 26 Rxg4 Rxg4 27 Qxg4
White was better in K.Shiven-J.Houska, Purley 2012.
b) Black has to play the strange 22 ... Kf8! to avoid any Bxh6 tricks. Following 23 Rg3 Nf6 24 Qe3
Rd5 White can launch some tactical attacks, but after 25 Rxg7 Kxg7 26 Ne5!? (or 26 Qxh6+ Kg8 27 Ne5
Rxe5 28 dxe5 Qxe5 and Black defends) 26 ... Rxe5 27 Qxh6+ Kg8 28 Qf4 (28 Bf4 Rxh5 also wins) 28 ...
Nd5 Black is winning.
22 g4 Nh7
 Blockading the attack.
23 Qe4
Likewise, after 23 Ne5 Bxc5 or 23 Rhg1 a5 24 Be3 Bf6 25 Qc2 Rad8 all is well.
23 ... Rad8

24 Bf4 Qc8
Keeping an eye on the e6-pawn.
25 Rdg1
After 25 Ne5 f6 26 Ng6 Bxc5 27 Rc1 Rxd4 28 Qf3 Bf8 29 Rxc6 Qb7 30 Rhc1 Rd1 Black will
emerge a pawn up.
25 ... Ng5 26 Nxg5 Bxg5 27 Bxg5 Rxd4!
Now Black’s counterattack on the d-line will soon start to come into play.
28 Qe3 hxg5 29 h6 R8d5 30 hxg7 Qd7
This has been a good example of how the tables can be turned on White.

Summary
This is a long variation (and chapter!), and the key manoeuvre for Black to remember is to provoke the
advance of the white c-pawn. As we have seen, play develops classically should White not castle
queenside. When he doesn’t, Black should always aim to use the h5-pawn as a blackmailing tool and
strive to attack down the d-line.
When White castles queenside, Black should head the opposite way. The play quickly becomes pretty
dynamic and what is critical is understanding which pawn is going to make that key breakthrough:
a) ... a7-a5 is usually seen when ... b5 and ... c5 are not desirable and when the bishop on e7 needs a
hiding place on b4.
b) ... b7-b5 is usually played on move 16 after a rook move or if White has not made any effort to play
actively, such as by combining Kb1 with Qe2.
c) ... c6-c5 should be played when the white knight has relocated from g3.
Chapter Two
Main Line: 11 Bd2
1 e4 c6 2 d4 d5 3 Nc3 dxe4 4 Nxe4 Bf5 5 Ng3 Bg6 6 h4 h6 7 Nf3 Nd7 8 h5 Bh7 9 Bd3 Bxd3 10 Qxd3
e6 11 Bd2

As I mentioned in the last chapter, it was at one time thought that 11 Bd2 and 11 Bf4 more or less led
to the same thing - but this is only true in lines that involve Black castling queenside. Since we will be
adopting the approach of kingside castling, there will be some fundamental differences which are
probably in White’s favour. Fashions change and chess evolves - if you had asked me in 2007 to choose
which one of the bishop moves was the scariest I would have answered 11 Bf4, but no longer.
Nowadays 11 Bd2 is the dangerous line. The assumption that Black can just castle and then challenge
the centre with ... c6-c5 has been rudely shaken to its core. In fact I consider this line to be so dangerous
that I have entirely changed my approach against it since 2007.
Please note too that at the end of the chapter we’ll consider another Bd2 approach, but with the pawn
not on h5, i.e. 8 Bd3 Bxd3 9 Qxd3 e6 10 Bd2.
11 ... Nf6 12 0-0-0 Be7
 White already faces a critical decision. To summarize briefly, White has two main plans: centralize
the rooks/queen or thrust the g2-g4 pawn forward in the event of Black castling kingside. Please note that
it is also possible for White to combine the two plans.

A) 13 Qe2
B) 13 Ne4
C) 13 Kb1

A) 13 Qe2
This signals the start of White’s plan to first centralize the pieces and then crack open the black kingside.
13 ... 0-0

The various lines here are again best grouped in pairs, of which the first two piece configurations are
harmless.

A1) 14 Kb1
A2) 14 Ne5
A3) 14 Rhe1
A4) 14 Nf1!?

The Big Trick

After 14 Kb1 it is very important for Black to delay the natural advance of the c6-pawn because after 14
... c5? White can play the stunning 15 d5!!.

Now 15 ... exd5 will allow White to reach the attacking square f5 with terrible consequences. Instead,
Black should play 14 ... Qb6 and delay ... c6-c5 until after he has played ... Rad8.

A1) 14 Kb1
Very importantly White keeps his knight on g3, hiding a devious intention. Here I believe Black should
play:
14 ... Qb6!?
Black completes development and prepares to centralize the two rooks. He also has the idea of
playing ... Qb6-b5.
The natural 14 ... c5? was the traditional method of dealing with this system, but Jobava’s stunning
novelty against Bareev has more or less consigned it to the doldrums: 15 d5!! exd5? (but even if 15 ...
Nxd5! 16 Ne4 with the dangerous threat of g2-g4-g5) 16 Bxh6 gxh6 17 Nf5 Re8 18 Nxh6+ Kf8 19 Ng5
and White has a stunning attack, B.Jobava-E.Bareev, European Club Cup, Rethymnon 2003.
 15 Ne5
White has a variety of moves available, none of which I find particularly challenging as I believe
playing Qe2 in conjunction with Kb1 to be a slight waste of time:
a) 15 c4 and now Black can play the natural 15 ... Rad8 16 Ne5 (if 16 Bc3 Bb4! 17 c5 Qa5 or 16 Nf5
exf5 17 Qxe7 Ne4 18 Rhf1 Qa6 19 Qb4 Ndf6 and Black is very comfortable with well-placed knights and
active pieces) 16 ... c5 17 Bc3 (17 d5 is met by 17 ... Nxe5 18 Qxe5 Nxd5!) 17 ... cxd4 18 Bxd4 Bc5
with a level position; the weak h5-pawn is something that White will have to be mindful of for the rest of
the game.
b) Playing a very quick 15 Ne4 is more dangerous. Here 15 ... Nxe4 16 Qxe4 Nf6 17 Qe2 Qb5!? is
probably the safest route to equality. Black accepts the doubled pawns so that he can eliminate any
potential danger of coming under a kingside attack. These doubled pawns are not necessarily bad either as
they can be a useful tool in undermining White’s protection of the d4-pawn, and the b5-pawn secures the
d5-square for Black. After 18 Qxb5 (if 18 c4 Qf5+ 19 Ka1 Rad8 20 Bc3 Bd6 21 Rhg1 Qe4!) 18 ... cxb5
the position is completely level.

The b5-pawn helps support a strong knight on d5 and only Black can take advantage of the semi-open
c- and d-files. Indeed, 19 Ne5 Rfd8 20 Bc1 Rd5 21 f3 was okay for Black in Shen Yang-Xu Yuanyan,
Wuxi 2005. Play might continue along the lines of 21 ... Rad8 22 c3 a6 23 Be3 Bd6 24 Bf4 Bb8 and there
are no dangers for Black.
c) 15 Nf5? would be a silly mistake as Black will play the basic 15 ... Ba3 16 Nxh6+ gxh6 17 b3 Qb5
18 Qe3 Ng4 19 Qf4 Qf5 and again he is not in any danger.
d) 15 Rhe1 can now be met with 15 ... c5, and if 16 d5 Nxd5 17 Ne4 N5f6 18 g4 Nxe4 19 Qxe4 Nf6
20 Qc4 Rfd8 21 g5 hxg5 22 Bxg5 Qb4 when again White’s attack is over before it even started.
15 ... Rad8 16 Nc4
After 16 Ne4 Nxe5 17 dxe5 Nxe4 18 Qxe4 Qxf2 Black picks up a pawn.
16 ... Qa6 17 Na5 b5 18 Ne4 c5!?
Breaking the position open.
19 dxc5 Nxe4 20 Qxe4 Nxc5 21 Qe2 Na4!? 22 Nb3 Bf6

23 Bc1
23 c3? Qc6 24 g4 allows Black to strike with 24 ... b4! when he is somewhat for choice: 25 cxb4
Nxb2 26 Rc1 Qd5 27 Rh3 Nd3 and White is suffering.
23 ... Qc6 24 g4 Qc4 25 Qf3
Alternatively, 25 Qxc4 bxc4 26 Rxd8 (not 26 Na5 Rb8 and White loses) 26 ... Rxd8 27 Na5 Rc8 28
c3 Bd8 29 Nb7 Bc7 30 Kc2 Rb8 31 b3 cxb3+ 32 axb3 Nxc3 and Black has won material.
25 ... Rc8 26 Rd2 Rfd8 27 Rxd8+
This doesn’t help, but if 27 Rhd1 Rxd2 28 Rxd2 Nb6 when the knight is heading to d5 while White
doesn’t have the capability to orchestrate an attack.
27 ... Bxd8 28 Qd1 Bf6 29 f4 Qc3 30 g5 Nxb2 31 Qe2 hxg5 32 fxg5 Be5 33 Re1 Nc4 34 g6 f6 35
Rd1 Nb6 36 h6 gxh6 37 Rd3 Nd5 38 Qf2 Qxd3!
Here White gave up in E.Mendez Fortes-Y.Bacallao Alonso, Santa Clara 2014.

A2) 14 Ne5
This is an old line that is really not seen very much these days primarily because of the following
response.

The Traditional Push ... c6-c5

It is very important for Black to memorize this position:
 This is possible as the discovered attack 16 Bxh6?! is not possible due to 16 ... Qa5! winning the a2-
pawn.
So, after 14 Ne5 we go:
14 ... c5!
Opening an important file near White’s king. Despite appearances the black queen is not in any
danger.
15 dxc5
15 Ng6 was played in E.Korbut-A.Ushenina, Ekaterinburg 2006. After 15 ... Re8 16 Nxe7+ Qxe7 17
dxc5 Qxc5 18 Bc3 Rac8 19 Kb1 I like 19 ... Qc6!?, preventing the bad knight on g3 from reaching e4,
instead of trading off queens with 19 ... Qc4 as happened in the game.
15 ... Nxc5 16 Ng4
As we just saw, the discovered attack 16 Bxh6?! does not work due to 16 ... Qa5, which is more than
good enough for Black: 17 Be3 Qxa2 and White’s king is in danger ( ... Rac8 and ... Nb3+ is one very
unpleasant threat).
Instead, after 16 Kb1 Qc7 (16 ... Qb6 is another possibility, as in I.Ortiz Suarez-B.Macieja, Mexico
City 2012) 17 Rhe1 (17 Bf4 is met too with 17 ... Rfd8) 17 ... Rfd8! 18 f4 Na4 19 Qb5 Nc5 20 Qe2 Rac8!
(Black repeats moves either because he was short on time or maybe just to tease White a little and find
out his frame of mind) 21 Bc1 Na4 22 Rxd8+ Rxd8 23 Ne4 Qb6 24 Ka1 Nxe4 25 Qxe4 Bb4 26 Rf1 Rd4!
Black is much better. Here his play was very instructive, combining threats to White’s king with pressure
in the centre, F.Sanz Alonso-J.Magem Badals, Seville 1993.
16 ... Qc7 17 Kb1
Black is already slightly better here because of the open c-file. The sacrifices 17 Nxh6+?! gxh6 18
Bxh6 Rfd8 19 Qe3 Ng4 and 17 Bxh6?! gxh6 18 Nxh6+ Kh7 19 Ng4 Qf4+ just do not work.
17 ... Rfd8 18 Bc3
 18 ... Nxg4
Even stronger would have been 18 ... Nd5!, since White isn’t about to land a blow on the kingside.
19 Qxg4 Bf8 20 a3?!
Making a weakness on the queenside that Black can target. This move simply makes White’s
defensive task harder. That said, after 20 Rxd8 Rxd8 21 Rd1 Na4! White must either concede the d-file or
accept a very weak pawn structure. Black is slightly better after both 22 Bd2 and 22 Rxd8 Nxc3+ 23 bxc3
Qxd8.
20 ... b5 21 Qe2 Qc6 22 f3?!
White does not challenge Black at all and plays in a purely defensive manner.
22 ... Rxd1+ 23 Rxd1 a5!‚
White can’t prevent the opening of files on the queenside.
24 Qd2
24 Qe1 is probably more stubborn mainly because the rook can defend the bishop from d3, although
after 24 ... b4 25 axb4 axb4 26 Bxb4 Qa6 27 Ba3 Na4 28 Rd3 (this possibility shows the difference
between this line and the text move) 28 ... Nb6! Black is ready to go ... Nc4.
24 ... b4! 25 axb4 axb4 26 Bxb4 Qa6 27 Ba3 Na4! 28 Qd4 Bxa3 29 bxa3 Qb5+ 30 Qb4 Nc3+
0-1 Xie Jun-L.Brunner, 4th matchgame, Bern 1995.

A3) 14 Rhe1
 A natural centralization. White wants to get Nf5 in.
14 ... Qb6!
An important move to remember. I believe this is the easiest way of dealing with this dangerous
system. The ‘safe’ move 14 ... Qb6 neutralizes White’s dangerous 15 Nf5 idea and plans either ... Qb5 or
the traditional ... c6-c5 break.
Instead, 14 ... a5!? was what Bent Larsen recommended. However, I believe here the advance is
incorrect. It is true that Black can advance his pawn quite far up the board, but the problem is that White
has a very harmonious position: for example, 15 Ne5 a4 16 a3 c5 17 Ng6 Re8 18 Nxe7+ Qxe7 19 d5 Qd6
20 dxe6 Rxe6 21 Qf3 and White is doing rather well, P.Negi-P.H.Nielsen, Copenhagen 2009.
I must also mention 14 ... Qc7!? 15 Ne5 c5, which you may wish to explore.
15 Ne5
Another logical choice. Instead:
a) 15 Nf5 is not so dangerous due to 15 ... exf5 16 Qxe7 Rfe8 17 Qd6 Re4!.
b) 15 c4 (preventing any ... Qb5 ideas) 15 ... Rfd8 16 Kb1 c5 and after the dubious 17 d5 Black can
simply capture the pawn with 17 ... exd5 18 Nf5 Bf8!.
c) 15 Kb1 should be met with the thematic 15 ... c5.
15 ... Rfe8 16 Nc4
More attacking options are:
16 f4 should be met with 16 ... c5! and here I have chosen to focus my efforts on the aggressive 17 f5
(17 Be3 is normal and after 17 ... Rad8 18 dxc5 Nxc5 and Black can be satisfied with his position).
 Black can defend with 17 ... Nf8! 18 fxe6 (18 Be3 Rac8 is equal; note that a move like 19 Kb1 would
be a blunder after 19 ... Nd5! threatening both the bishop and a fork on c3) 18 ... Qxe6 (the safest move,
avoiding any sacrifices on f7) 19 Kb1 cxd4! 20 Qf1 (after 20 Qd3 Black can play the tricky 20 ... N8d7
21 Ng6 Qg4 22 Nxe7+ Rxe7! and he is better) 20 ... Qa6, which is level.
b) 16 Nxf7 also has to be considered, but after 16 ... Kxf7 17 Qxe6+ Kf8 18 Nf5 (18 Bxh6 can be
neutralized by some careful moves: 18 ... Bb4 19 Bxg7+ Kxg7 20 h6+ Kf8 21 h7 Rxe6 22 h8Q+ Kf7 23
Qxa8 Rxe1 and Black has too much extra material) 18 ... Bb4 19 Bxb4+ Qxb4 20 Qd6+ Qxd6 21 Nxd6
Rxe1 22 Rxe1 b6 White doesn’t have anything like enough compensation for the material deficit.
16 ... Qa6

17 Na5
Likewise, after 17 Kb1 c5 18 dxc5 Bxc5 I don’t see any problems for Black.
17 ... b5!
Black should only swap queens when it is in his interests to do so. Here with White’s pieces a little
awkwardly situated, it is best to keep the queens on and continue the attack.
18 Kb1 c5!
Opening the c-file for the attack.
 19 dxc5 Nxc5 20 Nb3 Nxb3 21 axb3 Qc6
Black can claim a slight advantage based on the weak h5-pawn.

A4) 14 Nf1

This modern line has in mind the critical g2-g4. It is probably the most challenging system for Black
to face and we need to play a very strange-looking manoeuvre in 14 ... Nb6! with the plan of starting a
piece attack with ... Na4 and ... Qb6 or ... Qd5.
14 ... Nb6!
The key move, with the aim of playing the knight to a4. Instead, 14 ... c5 15 g4 is considered very
dangerous.
15 Kb1
An important point is 15 g4 Qd5!.

After 16 Kb1 Qe4 17 Qxe4 Nxe4 18 Be3 Nd5 19 Re1 Rfd8 20 N1d2 Nxd2+ 21 Nxd2 c5 Black is
fine.
Instead, 15 c4 c5 (15 ... a5 is also possible with the idea of moving the pawn all the way up the
board) 16 dxc5 Nbd7 17 g4 Nxc5 18 Bc3 (or 18 g5 Qd3! 19 Qxd3 Nxd3+ 20 Kc2 Ne4 and Black is
better) 18 ... Qc7 19 N1d2 (if 19 g5? Nfe4! 20 gxh6 Nxc3 21 bxc3 Qf4+ 22 Qe3 Qxc4 23 hxg7 Rfd8!
when Black is winning) 19 ... Nh7 20 Kb1 Rad8 reaches a pretty complicated position.
15 ... Na4!

16 Ka1!?
We should also look at:
a) 16 b3 is quite a critical test, and if 16 ... Nb6 17 g4. However, after 17 ... Nh7 (I have to confess
this is my favourite move, preventing White’s ideal pawn push) 18 Ng3 a5! 19 a4 Nd5 20 Ne5 Bb4 21 f4
(21 c4 is also incredibly complicated: for instance, 21 ... Bxd2 22 Qxd2 Nb4 23 f4 b5 24 g5 hxg5 25 h6
g6 26 Ne4 bxc4 27 Nxc4 Qd5 28 Nxg5 f6 29 Nh3 Rfb8 and it’s all a big mess) 21 ... Bxd2 22 Qxd2 b5
(of course, Black must play this move, otherwise we might as well just sign the scoresheet and walk
away) 23 g5 f6! 24 Ng6 (24 gxf6 Nhxf6 is an easier position for Black to play, but to be honest it remains
incredibly sharp) 24 ... fxg5 25 Nxf8 Qxf8 White still has to fend off a very dangerous attack.
b) The careless 16 g4? is convincingly met by 16 ... Ba3!.

After 17 b3 Nd5! 18 Qe1 Nb2! 19 Rc1 Nc4! Black definitely has all the attacking chances: for
example, 20 Rd1 Nxd2+ 21 Qxd2 Bb4 22 Qd3 Nc3+ 23 Kb2 Qa5 24 a4 Nxd1+ and wins.
c) 16 Qc4 Nb6 17 Qb3 (17 Qe2 also leads to a repetition) 17 ... Qd5 18 Qd3 Qe4 is a route to an
early draw.
d) 16 Qd3 was played against me once, but after 16 ... c5 17 Qb5 Qe8 18 Qxe8 Rfxe8 Black has the
better endgame:
d1) 19 dxc5 Ng4 20 Be3 was A.Rudakov-C.Gu, Maribor 2012, and after 20 ... Bxc5! Black has a
very pleasant endgame.
d2) 19 Ng3 cxd4 20 Nxd4 Red8 is a good endgame for Black; note the bad knight on g3. After 21 Nb3
Rac8 22 Rhe1 b6 23 Be3 Nd5 24 Bd4 is forced (24 Nd2?? Bf6 and Black is winning), but Black will
simply double the rooks and attack the weak h5-pawn.
16 ... c5
The most direct counterattacking attempt.
17 Bc1
Again, we should discuss some alternatives too:
a) After 17 g4 Black is pretty quick with his counterplay in the centre: 17 ... cxd4! 18 g5 hxg5 19 h6
g6 20 Bxg5 Qb6 21 Bc1! (or 21 Rb1 Nc5 threatening ... d4-d3, and if 22 Nxd4 Nce4 23 Be3 Kh7 24 Nf5
Bc5!) 21 ... Rad8 22 h7+ Kh8!.
b) 17 b3 Nb6 18 dxc5 Bxc5 19 Bxh6 Nbd5 20 c4? gxh6 21 cxd5 Nxd5 and White’s king is in big
trouble.
c) 17 Qb5 Qe8! 18 Qxe8 Rfxe8 19 Ng3 cxd4 20 Nxd4 Bc5 21 Nge2 Red8 22 Rh4 Nb6 23 Rf1 Bxd4
24 Nxd4 Rd5 25 b3 e5 0-1, Y.Sari-O.Kanmazalp, Kemer 2014.
17 ... Qc7 18 g4

18 ... Nd5!?
Another piece heads to the queenside.
I also like the solid 18 ... Nh7, whereas after the brave 18 ... Rad8?! it gets pretty scary: 19 g5 hxg5
20 h6! g6 21 Nxg5 cxd4 22 Nh2 (every move in this line can mean the difference between victory and
defeat, so great energy must be expended on each one) 22 ... Qb6! (or 22 ... Nc5 23 h7+ Kh8 24 Nhf3 with
the idea of Nh4 followed by a knight sacrifice) 23 h7+ Kh8 24 Nhf3 and I believe Black has a difficult
position. There are too many minor pieces on the board for Black to rest comfortably.
19 Ne3
One of the pitfalls is that the immediate 19 g5 runs into 19 ... Nac3!! 20 bxc3 Nxc3 21 Qc4 Nxd1 22
gxh6 Rfc8 23 Rg1 (or 23 N1d2 Nxf2 24 Rg1 Bf6! 25 Bb2 cxd4 26 Qf1 Qf4 27 Qxf2 Rxc2 and White is
struggling) 23 ... Bf6 24 Bg5 cxd4 25 Qxc7 Rxc7 26 Bxf6 Rxc2 when White has to give up a lot of
material to prevent mate.
Instead, after 19 Ng3 Rac8 20 g5 Ndc3 21 bxc3 Nxc3 22 Qc4 Nxd1 23 Rxd1 cxd4 24 Qxc7 Rxc7 25
gxh6 Rxc2 26 hxg7 Rfc8 27 Kb1 Rxf2 28 Nxd4 White’s two minor pieces do not compensate for the
missing pawns.
19 ... Nf4 20 Qb5 Nb6 21 g5 a6! 22 Qb3 cxd4 23 gxh6 dxe3 24 hxg7 Rfd8 25 h6 Rxd1 26 Rxd1
exf2
Black’s king position is pretty shaky, so we need to tread with care, but with an extra piece in the bag
we should manage to do this with ease.

B) 13 Ne4

The concept of g2-g4 is by far the scariest idea that Black will encounter against my recommended
lines. At times I will be actively avoiding this idea with an exchange of queens, as seen in this example:
13 ... Nxe4 14 Qxe4 Nf6 15 Qe2 Qd5 16 Kb1 and here I recommend 16 ... Qb5, offering to reach a
queenless middlegame where the threat of g2-g4 is neutralized completely.
Another way of dealing with this pawn push is to play ... c6-c5 before castling: for instance, 13 ...
Nxe4 14 Qxe4 Nf6 15 Qd3 c5.
13 ... Nxe4 14 Qxe4 Nf6
 Now we have:

B1) 15 Qe2
B2) 15 Qd3

B1) 15 Qe2 Qd5! 16 Kb1

After 16 c4?! Qe4! 17 Qxe4 (after 17 Be3 b5! 18 Ne5 bxc4 19 Qxc4 0-0 20 Rhe1 c5 21 f3 Qd5 I went on
to win in B.Anhchimeg-J.Houska, Calvia Olympiad 2004) 17 ... Nxe4 18 Be3 f5 19 g3 Bg5 20 Nxg5 hxg5
21 Rh2 Kf7 22 Rg1, as in Z.Hracek-D.Navara, Ostrava 2014, I like the computer’s suggestion of fixing
the pawn structure with 22 ... g4!? and then following up with 23 ... b5.
16 ... Qb5!?

The only move that avoids the well-known Caro-Kann ending, 16 ... Qe4 17 Qxe4 Nxe4 18 Be3,
which I won’t be covering in this book. Coincidentally our position also arises after 13 Kb1 Qb6 14 Ne4
Nxe4 15 Qxe4 Nf6 16 Qe2 Qb5.
17 Qe5
a) 17 c4 Qf5+ 18 Ka1 and now:
 a1) I recently got to test out my initial recommendation, 18 ... Rd8, and here my opponent played 19
Bc1 (19 Be3 Ng4! covers the e5-square and after 20 Nh4 Qxh5 White is getting nowhere). Black has
wasted a lot of time provoking pawn advances and the danger is that White may play Ne5 followed by g2-
g4. 19 ... 0-0 was my original intention, which I followed: 20 Ne5 c5 and White now played 21 f3!
sacrificing a pawn for his attack. 21 ... Rxd4 22 g4 and here I blundered with 22 ... Rxd1? 23 Rxd1 Qh7
and my queen was horribly cut out of the game. Oops! Instead of 19 ... 0-0 Black should enter a
manoeuvring game, but 19 ... Qe4! 20 Be3 Ng4 21 Nd2 Qf5 22 Nf1 Nf6 23 f3 is still a dangerous attack
for Black to navigate.
a2) 18 ... 0-0-0!? 19 Bc3 (19 Be3 Bd6 covers the e5-square; 19 Bc1 allows the normal breakthrough
19 ... c5 20 d5 Rhe8 21 dxe6 Rxd1 22 Rxd1 Qxe6 and White has problems with the h5-pawn) 19 ... Bd6
20 Ne5 Bxe5 21 dxe5 Qe4 22 Qxe4 Nxe4 23 Bd4 c5 24 Be3 b6 (the black knight is safe trapped in the
centre for the time being) 25 Kb1 f5 26 exf6 Rxd1+ 27 Rxd1 Nxf6 leaves the white h5-pawn vulnerable
again.
b) 17 Qe3 c5!? 18 Qe5 0-0 19 g4 Nxg4 20 Qg3 f5 21 Ne5 cxd4 22 Nxg4 fxg4 23 Qxg4 Bf6 sees the
bishop reach its perfect defensive diagonal while the black queen can take up an active role. I think Black
is already better.
17 ... Rd8 18 Bc1 0-0
White doesn’t have a clear-cut plan of attack. What happens next is that Black plays in a concerted
way while White just drifts.
19 Rd3 Rd5 20 Qe2 Rxh5 21 Re1 c5 22 Ne5 cxd4 23 f4 Bb4 24 Rf1 Rc8 25 g4 Rh3 26 Nf3 Qc4
27 Qd1 Nd5 28 Rxd4 Qb5 29 Qd3 Qb6 30 Rc4 Rd8 31 Ka1 Bc5 32 Qe4 Be3 33 f5 exf5 34 Qxf5
Bxc1
This spelt the end in E.Paehtz-V.Gunina, Tromso Olympiad 2014.

B2) 15 Qd3

A retreat clearly designed to prevent Black’s plan of ... Qd5.

15 ... c5!
Making use of the queen sitting on d3. I much prefer this to 15 ... Qd5 16 c4 Qe4 17 Qb3, and if, say,
17 ... b5 18 Rde1 Qh7 19 Ne5 Rc8 20 Qh3 0-0 21 c5 Nd5 22 Ng6.
16 Be3
Important alternatives are:
 a) 16 Qb5+ is another critical move, but the resulting endgame is pretty comfortable for Black as
White will not manage to restrain the kingside pawn majority: 16 ... Qd7 17 Qxd7+ Nxd7 18 Be3 (or 18
Bc3 Bf6 19 Rhe1 0-0-0 20 Re4 cxd4 21 Bxd4 Bxd4 22 Rexd4 f5 23 Rc4+ Kb8 24 Rcd4 ½-½, Yu
Ruiyuan-M.Parligras, Fujairah City 2012) and here I like the simple 18 ... cxd4 19 Bxd4 Bf6 20 Be3 Ke7
21 Rd3 Rhd8 22 Rhd1 b6 when I don’t think White can claim even a minimal edge.
b) 16 Bc3 0-0! 17 dxc5 Bxc5 18 Qxd8 Rfxd8 19 Bxf6 gxf6 20 Rhf1 f5 21 c3 Kg7 and Black’s
monarch will quickly enter the game.
c) Should White play the safe 16 Kb1 I prefer to delay castling with 16 ... Rc8 (16 ... 0-0 17 g4 is
unclear and 17 ... Nxg4 18 Qe2 very complicated):

c1) 17 g4 cxd4! (17 ... Nxg4 18 Qe2 is again very unclear) 18 g5 (18 Rhg1 Qd5 19 g5 hxg5 20 Rxg5
Qc4 forces the exchange of queens and neutralizes any initiative for White) 18 ... hxg5 19 Bxg5 Qd5 20 h6
gxh6 and with ... Qe4 imminent, Black has no problems.
c2) 17 Ne5 0-0 18 dxc5 Qc7 19 f4 Rfd8 20 Qe2 Qxc5 and we already have a position where we can
outplay White, as indeed Black did in V.Rjanova-A.Gvanceladze, Loo 2013.
d) 16 Qb3 doesn’t really threaten anything so Black can play in all calmness 16 ... 0-0, and if 17 dxc5
Qc7 18 g4 Ne4! stopping any g4-g5 ideas: for example, 19 Be3 Bxc5 20 Nd4 e5 21 Nf5 Rfd8 or 19 g5
Nxd2 20 Rxd2 hxg5 21 h6 gxh6 22 Rxh6 Kg7 (White is way too uncoordinated to have anything remotely
threatening) 23 Rh5 Qxc5 24 Qxb7 Bf6 and Black is safe, but White really isn’t.
16 ... Qa5
There is also the solid 16 ... 0-0 17 Ne5 (17 g4 doesn’t work because of 17 ... Qd5! and something
big drops) 17 ... Qa5 18 Kb1 Rad8! (18 ... Rfd8 looks more natural, but I guess Schandorff didn’t want to
allow 19 Bxh6 gxh6 20 Qg3+ Kf8 21 Qf4 Kg7 with a repetition) 19 g4 (not 19 Bxh6 gxh6 20 Qg3+ Kh8
21 Qe3 Ng8 and Black emerges with extra material) 19 ... cxd4 20 Bxd4 and then:
a) 20 ... Nd5?! 21 c3 (21 Qe4! would have given White the more comfortable game as it is not clear
how Black will make progress) 21 ... Bf6 22 Nc4 Qc7 23 Bxf6 Nxf6 with equality in L.Rasmussen-
L.Schandorff, Helsingor 2014.
b) More correct would have been to simply play 20 ... Bc5 21 c3 Bxd4 22 cxd4 Nd7 23 f4 Nxe5 24
fxe5 f6 with a completely equal if drawish position.
17 Kb1 Rd8!
17 ... 0-0 is still risky and requires accurate handling, but is playable. After 18 g4 Black must, though,
be precise:
 a) 18 ... cxd4? makes life easier for White. Now his bishop obtains a great diagonal and what
followed was a great attacking game: 19 Bxd4 Nxg4 20 Rhg1 f5 21 Bxg7 Kxg7 22 Rxg4+ fxg4 23 Qg6+
Kh8 24 Qxh6+ Kg8 25 Qxe6+ Rf7? 26 Ne5 Raf8 27 h6 Qa6 28 h7+ Kxh7 29 Rh1+ Bh4 30 Rxh4+ Kg7 31
Rxg4+ Kh7 32 Qxa6 (32 Rh4+ Kg8 33 Qh3 Qf6 34 Nxf7 Rxf7 would have been a more convincing
continuation) 32 ... bxa6 33 Nxf7 Rxf7 34 f4 Kh6 35 b3 Kh5 36 Rg5+ Kh4 37 Rg6 1-0, E.Safarli-
A.Kovchan, Turkish Team Championship 2011.
b) Instead, Black should play 18 ... Nxg4 19 Rdg1 f5 20 Ne5 cxd4! 21 Nxg4 dxe3 22 Nxh6+ Kh8! 23
Qxe3 Bf6! (once the bishop gets to f6 Black tends to establish a fortress which is extremely hard for
White to break down) 24 Rg6 Qe5 25 Qxe5 Bxe5 26 Rxe6 Bf4 and the white knight is trapped on the
edge.
18 Qb3 0-0

19 g4?!
Thanks to Black’s clever move order, the key pawn thrust comes at the wrong time.
19 ... Nxg4 20 Rdg1 cxd4! 21 Bxd4 Qf5 22 Bxg7 Kxg7 23 Nh2 Rd4 24 f3
We’ve followed K.Abdulla-O.Girya, Plovdiv 2014, and now the best approach was to go on to the
counterattack:
24 ... Rb4! 25 Rxg4+ Qxg4 26 Nxg4 Rxb3 27 cxb3
White is in trouble as Black has the superior pawn majority and the better minor piece.

C) 13 Kb1
A noncommittal choice. White protects the a2-pawn and sidesteps any forthcoming danger on the c-line.
When I wrote my original book I believed that Black could simply play 13 ... 0-0 following through with
the traditional ... c6-c5, with the conclusion that Black has solved his problems. The line has since moved
on substantially and is now far too complicated, so I now prefer 13 ... Qb6.

C1) 13 ... 0-0
C2) 13 ... Qb6

C1) 13 ... 0-0

As we shall see next, 13 ... Qb6 will form the basis of my main recommendation.
14 Ne4! c5!?
Play after this move becomes super-hot. Instead, 14 ... Qb6?! 15 Nxf6+ Nxf6 16 g4 looks great for
White, but 14 ... Nxe4 15 Qxe4 Nf6 16 Qe2 Qd5 is a possibility.
15 g4!

Unsurprisingly 15 Nxf6+ isn’t as dangerous.

15 ... Nxg4 16 Qe2!
The trickiest move, as 16 Qb3 allows 16 ... Qb6.
16 ... Qb6 17 Ne5!?
I like this move best. Instead, after 17 Rhg1 f5 18 Nc3 Bf6 there is no clear way to break down
Black’s blockade.
 17 ... Ndxe5 18 dxe5 f5! 19 exf6
The best move. 19 Nc3 only leads to a draw.
19 ... Nxf6

This position is the problem. White has too much pressure for a human to withstand without computer
assistance. It does all work out for Black, but it is incredibly complicated, sharp and requires an amazing
memory.
20 Rdg1
20 Rhg1 is possible too, but is not as good. Instead, Negi recommends 20 Nc3 to which Black must
try and neutralize some of the danger with 20 ... Qa6! (being a pawn up, Black can afford to cripple his
pawn structure) 21 Qe5 (21 Nb5 isn’t too promising after 21 ... Rad8 22 Rhg1 Rf7 and Black can start
wriggling, while 21 Qf3 Qc6 22 Qe2 Qa6 offers to repeat) 21 ... Ng4 22 Qg3 Nxf2 23 Bxh6 Rf7 24 Rhf1.
Here you have to know 24 ... Bh4!! when Black may survive.
20 ... Nxe4 21 Qxe4 Bf6 22 c3
Likewise, after 22 b3 c4 23 Bxh6 cxb3 24 axb3 Rf7 25 Bg5 Qd4 Black might well survive.
22 ... Kf7!
The best move according to the computer, but this is just crazy and I can’t in all reality recommend it as
my main line. Instead, 22 ... Rf7 23 Bxh6 Bxc3 24 Qc2 Bd4 25 Be3 Kf8 26 Bxd4 cxd4 27 h6 gxh6 28
Rxh6 Ke7 is another shifty position for Black to have to defend.

The Universal 13 ... Qb6



I am primarily recommending this queen move against 13 Kb1, but it is in fact a universal system that
is be used against all White’s 13th moves, including 13 Qe2.
Our aim is to delay castling kingside for as long as possible and to prepare to strike out with ... c6-c5.
In a perfect world this would lead us to obtain our favourable and favourite endgame. In reply White can
employ four strategies:
a) To wait and force Black to commit his king.
b) To play Ne4 and Nxf6+ gaining control over the e5-square.
c) To attack the black kingside with Ng3-f5, looking to sacrifice.
d) Pretend that Black has castled kingside and start directly attacking on the kingside with Ne4
accompanied with g2-g4.
I will look at all these plans in the context of the following moves.

C2) 13 ... Qb6

Generally I would now prefer the king’s rook on d8 and the a8-rook on c8. I am reluctant to make this
as a rule as there are many exceptions, but that is the guide!
We have:

C21) 14 Rhe1
C22) 14 c4

There’s also 14 Ne4 and then:

a) 14 ... Nxe4 15 Qxe4 Nf6 16 Qe2 (16 Qf4! is more annoying) 16 ... Qb5! transposes to my
recommendation in Line B1, above.
b) I really like 14 ... Rd8!? which was my recommendation and then GM Negi gave it his seal of
approval too. After 15 g4 c5! Negi 16 Be3 (if 16 Bc3 Nxg4 17 Qd2 Qc6 18 Qe2 Qa6! or 16 g5 hxg5 17
h6 gxh6 18 Bxg5 cxd4 and Black is perfectly safe) 16 ... Nxg4! 17 Rhg1 Nxe3 18 Qxe3 (or 18 fxe3 0-0)
18 ... cxd4 19 Qf4 Nc5 20 Rxg7 Rf8 21 Ne5 Nxe4 22 Qxe4 Bf6 23 Rh7 Bxe5 24 Qxe5 Rd5 Black is
holding the position.

C21) 14 Rhe1 0-0

14 ... c5!? has also been played successfully in U.Eliseev-A.Riazantsev, Yekaterinburg 2013.
15 Nf5!?

Karjakin in his 2013 game against Mamedyarov demonstrated a very interesting idea: sacrificing this
knight for an attack.
15 ... exf5!
Best. Instead, 15 ... Bb4 16 Nxh6+! gives White long-term pressure against the exposed king and with
the black queen a little offside, this position contains some serious venom: 16 ... gxh6 17 c3 Ba3 18 Bc1
Rfe8 (to bring the bishop back to f8, from where it can defend the king) 19 g4! (Karjakin aims to open the
g-file, which will enable his rooks to join the attack) 19 ... Nh7 20 c4 Bf8 21 g5 hxg5 22 Nxg5 Nxg5 23
Bxg5 Kh8 (in a later game Black tried 23 ... Be7 24 Qg3 and although it ended in a draw Black was still
under a lot of pressure in K.Szabo-B.Korpa, Budapest 2014) 24 Rg1. Just by playing very natural moves,
Karjakin had developed a very strong attack and won a fine game in S.Karjakin-S.Mamedyarov, Zug
2013.
16 Rxe7
At first sight White has active play and Black a poor pawn structure, but the doubled f-pawns are
actually pretty useful.
16 ... Qd8 17 Re2
17 Qa3!? is another idea from Negi: 17 ... Ne4! 18 Nh4 Nb6 19 Nxf5 (or 19 Ba5 Nd6 20 Rde1 Nbc4
21 Bxd8 Nxa3+ 22 bxa3 Rfxd8 23 Rc7 Re8 24 Rxe8+ Rxe8 25 Rd7 Nc4 and the position is level) 19 ...
Nc4 20 Qh3 Ncxd2+ 21 Ka1 when according to Negi, Black’s best is 21 ... Kh8 22 Qg4 Rg8. Carrying on
his analysis, White has nothing after 23 Qf4 Qb8 24 Qe3 Re8 25 Rxe8+ Qxe8 26 f3 Nxf3 27 gxf3 Ng5.
17 ... Ne4!

This is forced in order to protect the f5-pawn.

18 g3
18 Bf4 was a move my husband suggested to me as an improvement. Black should still go 18 ... Ndf6!,
preventing the undermining pawn push g2-g4:
a) After 19 Be5!? Black must tread carefully to secure equality: 19 ... Ng4 20 Bg3 Re8 21 Rde1 Re6
22 Ne5 Qg5 23 Nxf7 (23 Nxg4 Qxg4 24 Be5 Rd8 25 f3 Qh4 26 fxe4? Rxe5 is a tricky position for White
to handle) 23 ... Kxf7 24 f3 Ngf6 and Black is comfortable.
b) 19 Ne5 Nd5 20 Bc1 Re8 21 f3 Ng3 22 Ree1 Nf6 (eyeing up the weak h5-pawn, although I must
state that we may not play to win it immediately; after all, sometimes the threat is more promising than the
execution) 23 c4 Rc8 (preparation for the pawn push d4-d5) 24 d5 cxd5 25 cxd5 Rc5 26 Qb3 Rxd5 and
Black wins a more important pawn than the dead one on h5.
18 ... Ndf6 19 Nh4 Re8!
 20 Bc1
20 Nxf5 is possible, but Black obtains perfectly satisfactory play after 20 ... Qd5 21 Nh4 c5 22 dxc5
Qxh5 23 Be1 Qxe2 24 Qxe2 Nc3+ 25 Bxc3 Rxe2.
Instead, 20 Be1 has the interesting idea of chasing away the black knight on e4 with f2-f3. Here 20 ...
Qe7 threatens ... Nc3+ and after 21 Ka1 Qd7 22 f3 Nd6 23 Re5 Nxh5 24 Nxf5 (if 24 g4 Nf4) 24 ... Nxf5
25 Rxf5 g6 26 Re5 Rad8 White might claim to be a touch better, but as long as Black keeps his pieces
active he should be OK.
20 ... Qd7
20 ... Qa5 is also attractive.
21 Qf3 Qd5 22 Ka1
22 Nxf5 fails to 22 ... Nd2+.
22 ... Nd6 23 Qxd5 Nxd5 24 Re5

In R.Kasimdzhanov-R.Ruck, German League 2013, White preferred 24 Rde1 and now I like 24 ...
Rxe2 25 Rxe2 b5, securing the knight on d5. After 26 Re5 f6 27 Re1 Kf7 28 b3 a5 29 Kb2 b4 30 Ng2
Nb5 it should be a draw, although we can certainly try to press.
24 ... Nc4 25 Ree1
 25 Rxf5 allows 25 ... Re2 26 Kb1 Na3+!! 27 bxa3 Nc3+ 28 Kb2 Nxd1+.
25 ... Rxe1 26 Rxe1 Nd6
I will give the rest of the moves so you can admire Vitiugov’s technique.
27 f3 Rd8 28 Re5 Nf6 29 Nxf5 Nxf5 30 Rxf5 Rxd4 31 b3 Rd5 32 Rxd5 Nxd5 33 c4 Nf6 34 g4 Nd7
35 Be3 a6 36 Bd4 f5!!

37 Kb2 Kf7 38 gxf5 Nf6 39 Kc3 Nxh5 40 Be5 Nf6 41 Kd3 g6 42 fxg6+ Kxg6 43 Ke3 Kf5 44 Bd6
h5 45 Kf2 Nh7 46 Kg3 Ng5 47 Bc5 Nf7 48 Bd4 Ne5 49 Be3 Nd3 50 Bd2 Ne5 51 Be3 h4+ 52 Kxh4
Nxf3+ 53 Kh5 Ne5 54 Bc1 Nd3 55 Bd2 Nc5 56 Bc1 Ne4 57 Bb2 Nf2 58 Kh6 Nd3 59 Ba3 c5 60 Kg7
Ke6
Here White finally gave up in the model game S.Ganguly-N.Vitiugov, Gibraltar 2014.

C22) 14 c4
As recommended by GM Parimarjan Negi in his recent huge and impressive repertoire book for White.
One advantage of being late with my deadline was that I could check out Negi’s lines and prepare the
reader accordingly.
14 ... Rd8
14 ... Qa6 is too committal, but it is a very important idea in general, discouraging Ne4.
15 Qc2
Played in Negi’s style. We should also explore:
a) 15 Qe2 0-0 and then:
a1) 16 Nf1 c5! (once the g3-knight starts manoeuvring Black needs to strike out in the centre as soon
as possible) 17 g4? (this doesn’t work, but neither does 17 d5? exd5 18 Qxe7 Rfe8 and the queen is
trapped) 17 ... cxd4! 18 g5 d3! 19 Qe1 Ng4 20 gxh6 Bf6 21 Bc3 d2 22 Qxd2 Nc5 23 Qc2 Rxd1+ 24 Qxd1
Bxc3 and Black wins.
a2) 16 Ne4 Nxe4 17 Qxe4 Nf6 18 Qe2 c5! is a critical position to memorize.
 Black can effectively defend against the dangerous-looking 19 g4 with 19 ... cxd4 20 g5 d3 21 Qe1
Ng4! 22 Rg1 Qc6!. Black survives due to this counterattacking possibility: for example, 23 Rg3 (or 23
Rxg4 Qxf3 24 Rg3 Qxh5 25 gxh6 g6) 23 ... Bc5 24 Rxg4 Qxf3 25 Rg3 Qxh5 26 gxh6 g6 27 Rg5 Bxf2 28
h7+ Kh8 29 Qe5+ f6 30 Rxh5 fxe5 31 Rxe5 Kxh7 32 Rxe6 Bh4 and the position is preferable for Black.
b) The dubious 15 Bc3?! was also played very recently. It is true that the bishop changes diagonal, but
this is not a good idea: 15 ... Qa6! (this concept works perfectly here) 16 Qe2 0-0 17 b3 (after 17 Nf1
Rfe8 18 g4 Nxg4 19 Rg1 Ngf6 the bishop is on the wrong diagonal, while after 17 Ne5 c5! 18 b3 cxd4 19
Bxd4 Nxe5 20 Qxe5 Qc6 White can’t really claim an advantage) 17 ... Rfe8 18 Ne4 b5 19 c5 Nxe4 20
Qxe4 b4 21 Bxb4 Rb8 22 Bc3 Nf6 23 Qc2 Nd5 24 g4 Rb7 25 Rhg1 Qe2 26 Ne5 Bf6 27 Nxc6 saw White
give up in D.Causo-O.Barbosa, Bangkok 2014, as 27 ... Nxc3+ 28 Qxc3 Qe4+ is winning.
c) 15 Bf4 0-0 16 Qc2 simply transposes to the main line.
d) 15 Ne4 0-0! 16 Nxf6+ (or 16 g4 Nxg4 17 Qe2 f5 18 Nc3 c5 and Black is defending nicely) 16 ...
Nxf6 17 g4 (if 17 Qe2 c5!) 17 ... Nxg4 18 Qe2 c5 and again Black is defending all the potential weak
points. Moreover, after 19 Rhg1 f5! 20 Bc3 cxd4 21 Bxd4 Qa6 22 Ne5 Bf6! it is White who should be
worried.
e) 15 Rhe1 0-0 16 Qe2 Rfe8 17 Ne5 c5 18 Bc3 cxd4 19 Bxd4 Bc5 20 Nxd7 Nxd7 21 Bc3 Bb4 with
equality.
15 ... 0-0
 16 Bf4
I also checked both 16 Nf1 c5 17 g4 cxd4 18 g5 d3 and 16 Ne4 Nxe4 17 Qxe4 Nf6 18 Qe2 c5 19 g4
Nxg4! and was happy with Black’s position.
16 ... Rfe8!
Both rooks stay close to the black king, which is very important as there will come a time when Black
will need two defensive resources: the first one is ... Be7-f8 and the second point is that once ... Bf8 has
been played, the rook will be required to play ... Re8-e7.
Instead, 16 ... c5 doesn’t quite cut it because of 17 d5!, although actually this isn’t too terrible after 17
... Rfe8! 18 dxe6 (or 18 d6 Bf8) 18 ... Qxe6!.
17 Ne5
Again, there are alternatives:
a) 17 Ne4 Nxe4 18 Qxe4 Nf6 19 Qe2 c5 20 g4 Nxg4 21 Rhg1 e5! 22 Nxe5 (or 22 Rxg4 exf4 and
White has no attack) 22 ... Nxe5 23 dxe5 Qe6 and Black can look to the future with some confidence: for
instance, after the dubious 24 Qe4 Black can simply go 24 ... Rxd1+ 25 Rxd1 Qg4 with some advantage.
b) 17 Nf1 is a move Negi often likes, but here 17 ... c5! 18 g4 cxd4! 19 g5 e5! 20 Bxe5 (20 gxf6 Qxf6
wins back the piece) 20 ... Nxe5 21 Nxe5 hxg5 22 h6 gxh6 23 Rxh6 Bf8! shows one of the advantages to
playing ... Rfe8 - the bishop is on hand to help out its king.
c) 17 Rhe1 is another waiting move, but this one minimizes the impact of g2-g4 and after 17 ... a5 18
Ne4 Nxe4 19 Rxe4 Nf6 20 Ree1 Nxh5 21 Be3 Nf6 22 d5 c5 23 dxe6 Rxd1+ 24 Rxd1 Qxe6 Black is a
pawn to the good.
17 ... c5
 Brave, and good.
18 Nxf7!?
18 Bxh6!? is another sacrifice that we need to consider. Our defence runs 18 ... gxh6 19 Nxf7 Kxf7 20
Qg6+ Kf8 when White’s best bet is to try 21 Ne4! (after 21 Qxh6+ Kg8 22 Qg6+ Kh8 the black king is
safely tucked away in the corner) 21 ... Ng8! (Black needs to protect the h6-pawn) 22 Rh3.

Now Black should play the fantastic 22 ... e5!!, connecting the queen to her Majesty and at the same
time threatening to swap queens. After 23 Rf3+ it is absolutely imperative that Black plays 23 ... Bf6
(anything else loses, as shown by 23 ... Ndf6?? 24 Rg3 and Black cannot stop mate without suffering huge
material losses) 24 dxc5 Qe6, which defends: for instance, 25 Rfd3 (25 Rd6 doesn’t work because of 25
... Qf7 26 Qg3 Re6 27 Rfd3 Rxd6 28 Nxd6 Qe6 29 Nxb7 Rb8 and White has trouble justifying his two-
piece deficit) 25 ... Re7 26 Nxf6 (or 26 Rd6 Qf7 27 Nxf6 Qxg6+ 28 hxg6 Ngxf6 29 Rxf6+ Nxf6 30
Rxd8+ Kg7 and the knight is superior to the pawns) 26 ... Ngxf6 27 Qxh6+ Ke8 28 Qh8+ Ng8 29 Rg3
Ndf6 30 Rxd8+ Kxd8 31 h6 Rd7 and two pieces to the good, Black is winning.
Instead, again after 18 Nxd7 Rxd7 19 dxc5 Rxd1+ 20 Rxd1 Bxc5 White has nothing.
18 ... Kxf7 19 Qg6+ Kg8 20 Bxh6 Bf8! 21 Bxg7 Bxg7 22 h6 Re7!
 This is a more or less forced sequence of moves. It looks a bit hairy at times, but Black has too many
pieces around his king for it to get too critical.
23 hxg7 Rxg7 24 Qh6 Kf7!
Threatening to liquidate with ... Rh7.
25 d5! Re8!
The position is not easy to play for either side, but White has to somehow justify his sacrifice.
26 Nh5?! Rg6 27 dxe6+
Likewise, after 27 Qf4 exd5 28 Rxd5 Kg8!! (calmly sidestepping all the threats) 29 Rg5 Ne4!! 30
Nf6+ Qxf6 31 Qxf6 Ndxf6 32 Rxg6+ Kf7 33 Rgh6 Nxf2 Black is significantly better.
27 ... Rxe6 28 Rxd7+ Nxd7 29 Qh7+ Ke8 30 Nf4 Rh6 31 Qg8+ Nf8 32 Rd1 Re7 33 Nd5 Qg6+
White doesn’t have enough pawns to compensate for the extra piece.

White Leaves the Pawn on h4

1 e4 c6 2 d4 d5 3 Nc3 dxe4 4 Nxe4 Bf5 5 Ng3 Bg6 6 h4 h6 7 Nf3 Nd7 8 Bd3

This line wasn’t so common in 2007. Fast forward a few years, and it’s still not so popular. So what
does White hope to achieve by omitting h4-h5? Well, he hopes to obtain a strong kingside attack with g2-
g4-g5 without the possible long-term weakness of the h5-pawn. The downside to this strategy is that
Black is no longer bothered by the clamping effect of the h-pawn, which makes castling kingside all the
more attractive. The plan I will be advocating is very much on a level with the ‘safe’ plan recommended
against the main line.
8 ... Bxd3 9 Qxd3 e6

10 Bd2
After 10 Bf4 Black should follow the standard pattern with 10 ... Qa5+:
a) I don’t really understand White’s motives for not chasing back the bothersome black queen, i.e. 11
c3 Be7 12 0-0 Ngf6 13 Rfe1 (13 b4 Qa3! followed by ... Nd5 is perfectly acceptable for Black) 13 ...
Qd8 (eliminating the possibility of any Nf5 ideas; the position has no immediate dynamic possibilities, so
Black can afford such luxuries) 14 c4 0-0 15 Rad1 Re8 16 Re2 Qa5 17 b3 Rad8 and Black had no
problems in A.Gurkin-R.Ovetchkin, Pavlodar 2012.
b) 11 Bd2 Qc7 12 0-0-0 Ngf6 13 Kb1 Bd6 14 Ne4 Nxe4 15 Qxe4 Nf6 16 Qe2 0-0 17 g4 Nxg4 18
Rhg1 h5 is a lesson for White to note.
10 ... Ngf6 11 0-0-0 Be7
11 ... Qc7, guarding the e5-square, was my original choice, but this idea of castling queenside goes a
little against my normal philosophy. Moreover, 12 Ne4 0-0-0 (or 12 ... Nxe4 13 Qxe4 Nf6 14 Qe2 Bd6 15
Kb1 0-0-0 16 c4 Rhe8 17 Ne5 c5 18 Bc3 and I would say this position is slightly better for White) 13 g3
Nxe4 14 Qxe4 Bd6 15 Rhe1 Nf6 16 Qe2 c5 (16 ... Kb8 17 Kb1 Rhe8 18 c4 c5 19 dxc5 Qxc5 20 Bc3 is a
touch annoying too) 17 dxc5! leads to an easier position for White to play on account of Black’s exposed
king, rather than 17 c4 cxd4 18 Nxd4 a6, which equalized in A.Skripchenko-V.Solovjova, Sochi 2006.
12 Kb1 Qb6
 Why not use the same strategy employed in our favourite safe variation and wait for White to commit?
13 c4
After 13 Rhe1 Black should wait further with 13 ... Rd8, waiting to strike with ... c6-c5 at the right
time: 14 h5 (or 14 c4 0-0 15 Nf5 exf5 16 Rxe7 Ne4 17 Qa3 c5 and the white rook is uncomfortably
placed in the heart of the enemy camp, although White can maintain equality with 18 Ba5 Qf6 19 Bxd8
Rxd8 20 Rxd7 Rxd7 21 Qxa7 Qd8 22 Qa4 cxd4) 14 ... c5 15 d5? Nxd5 16 Nf5 Bf6 and White’s attack is
going nowhere.
13 ... Rd8 14 Qe2 0-0 15 Bc3 Bb4! 16 Rc1 Bxc3 17 Rxc3 c5 18 Rb3 Qc7 19 Rd1 cxd4 20 Nxd4
Qf4

This is a dream position for Caro players. With play along the d-file and knights with good jobs to do,
Black is a lot better.
21 Rxb7 Nc5 22 Rb5 Rxd4 23 Rxc5 Rxd1+ 24 Qxd1 Qxf2 25 Qd6 e5
With the loss of material inevitable, White already gave up in H.Fleuch-A.Gavrilov, Sibenik 2014.

Summary
11 Bd2 has become the sharpest line in the Main Line Caro primarily because of the complications
involved with the g2-g4 push. Again the critical nature of the play is determined by White’s choice on
move 13 after 11 ... Ngf6 12 0-0-0 Be7.
After 13 Qe2 Black can play in the style of my first book and castle kingside planning ... c6-c5. The
trickiest line there sees 14 Nf1 when Black should undertake something radical: we begin to attack the
white king with ... Nb6-Na4 and only after break with ... c6-c5.
After 13 Ne4 Black should halt g2-g4 by exchanging knights and if 15 Qe2 initiate ... Qd5 followed by
... Qb5!? or if 15 Qd3 strike out immediately with 15 ... c5. The general vibe is that counterplay starts
before kingside castling.
13 Kb1 is the most flexible approach and here I advocate a complete change of approach with 13 ... Qb6.
Negi’s recommended approach for White requires careful handling and I recommend meeting 14 c4 with
14 ... Rd8 followed by 15 ... 0-0 when Black is well placed to meet any g2-g4 pushes. A key move is ...
Rfe8, allowing the dark-squared bishop to drop back to f8 to join in the defence. The positions are very
complex, but White needs to sacrifice a lot of pieces in order to keep any initiative going.
Chapter Three
Classical Variation: the f4 Plan
1 e4 c6 2 d4 d5 3 Nc3 dxe4 4 Nxe4 Bf5 5 Ng3 Bg6

One of the downsides of developing the bishop to f5 is that it becomes a target for attack. We saw in
the previous chapters White targeting the bishop with h2-h4-h5. However, there are two alternative forms
of attack: the not-so-subtle idea of playing f2-f4 and catching the king in the centre, and the subtle Ne2/h3-
f4, inconveniencing the bishop.
There are two forms of the f2-f4 attack. The first sees White choose to simply march the pawn on
move 6 with a direct show of aggression. The second version usually arises once Black has committed his
king’s knight to f6. A word of caution: once White has played f2-f4, we must always anticipate the move
f4-f5 even if it costs White material. Should Black gain an iron grip over the f5-square then White will
soon come to regret the weakening of his kingside and leaving the e4-square so vulnerable.
We will first look at the immediate 6 f4 and then at a manipulation of this basic idea when White plays
6 Bc4 e6 7 N1e2 (or 7 Nh3) 7 ... Nf6 8 0-0 Nbd7 9 f4.

A) 6 f4
B) 6 Bc4
C) 6 Nh3

The Direct Attack: 6 f4

 One of the disadvantages of attacking early in such a forthright manner is that it allows Black to make
preparations and develop with the following in mind:

a) Take control over the f5-square. This means that Black may develop the g8-knight to e7 to reinforce
his control: for instance, if the white bishop develops to d3.
b) Break open the position in the classical way with ... c6-c5 when White’s dark squares will become
even more vulnerable.

A) 6 f4
This move is a speciality of IM Kofidis, but has even been played by GM Judit Polgar on occasion. The
main advantages to this aggressive system are that White has a massive grip on the important e5-square
and can play for the dangerous pawn thrust f4-f5 when the knight on g3 will spring into life. However,
there are also some drawbacks: the first is that Black’s natural break ... c6-c5 will be more powerful than
normal as the g1-a7 diagonal has been weakened. Secondly, from a positional perspective, White has
voluntarily given himself a bad bishop on c1.
6 ... e6
 7 Nf3
I have also faced 7 h4 h6 8 Nf3. Here I played the perfectly acceptable 8 ... Nd7 9 h5 Bh7 10 Bd3
Bxd3 (10 ... Bb4+ is also an interesting idea, forcing White to make an additional queenside weakness)
11 Qxd3 Ngf6 12 Bd2 c5!? (a thematic reaction) 13 0-0-0 cxd4 14 f5 e5 15 Rhe1 Be7 16 Nxe5 Nxe5 17
Rxe5 0-0 18 Ne4 (if 18 Bf4 Bd6) 18 ... Nxe4 19 Rxe4 Bf6 20 g4 (already I think my position is
preferable as the attack on the c-file should swing the game in Black’s favour) 20 ... Rc8 21 Kb1 Qb6 22
g5 hxg5 23 Rh1 Rfe8 24 Rxe8+ Rxe8 25 h6 Qc6 26 Qh3 and now in N.Khoudgarian-J.Houska, Dresden
Olympiad 2008, I would have had a clear advantage if I had played 26 ... d3! 27 c3 gxh6 28 Qxh6 Re2.
Instead, after 7 Bc4 Nf6 8 c3 Nbd7 9 Qe2 Bd6! (preventing White from playing f4-f5 by attacking the
knight on g3) 10 Nf3 Nd5!? Black doesn’t have any problems, M.Demirkesen-Y.Bozdemir, Antalya 2011.
7 ... Bd6!

This is the problem with White showing his cards so early. Black can develop whilst guarding the f5-
square at the same time.
8 Bd3
Exchanging Black’s most active piece. Other possibilities are:
a) Should White wish to preserve the light-squared bishop with 8 Be2, as in J.Polgar-A.Khalifman,
Las Vegas 1999, Black has a choice:
a1) I played 8 ... Nf6 9 0-0 Nbd7 10 Ne5 0-0 11 h4 c5! 12 c3?! (12 h5 has to be played, although after
12 ... Bf5 13 Nxf5 exf5 14 h6 g6 15 Bf3 Qc7 16 c3 Ne4 I don’t think Black has any troubles) 12 ... cxd4
13 cxd4 Rc8 14 Be3 Nd5 15 Rf3 Nxe3 16 Rxe3 Nxe5 17 fxe5 in A.Nikolova-J.Houska, Plovdiv 2008,
but here I missed the chance for 17 ... Bc2 18 Qe1 Be7 with a decisive advantage.
a2) 8 ... Ne7 is another pleasant knight move: for example, 9 Ne5 Bxe5 10 fxe5 c5! (White’s centre
must be broken up, otherwise he will have a very nice space advantage along with the two bishops) 11
dxc5 Qxd1+ 12 Bxd1 Na6! (preventing b2-b4) 13 0-0 (13 Be3 Nb4 also regains the pawn) 13 ... Nxc5 14
Bf3 Nc6! targeting the weak e5-pawn. Black has already equalized here, as he is not lagging behind in
development, while his two knights are doing an excellent job of controlling the key squares and attacking
White’s isolated pawn on e5.
b) 8 Ne5 has hardly been played. It’s not so convincing here, though, primarily because it is simply
too early. Indeed, after 8 ... Bxe5 9 fxe5 c5 10 Bb5+ Nc6 11 Bxc6+ bxc6 12 Be3 Ne7! 13 dxc5 Nd5 14
Bf2 Rb8 15 b3 Nb4 White must return the extra material.
c) After 8 c3 again it is up to the reader to choose where to develop the g8-knight. In both cases the
plan is the same: to break with a well-timed ... c6-c5. For example, 8 ... Ne7 (8 ... Nf6 9 Be2 0-0 10 Ne5
Qc7 11 h4 c5 12 h5 Bf5 13 Nxf5 exf5 is a different flavour of how to play with Black) 9 Bc4 Nd7 10 h4
h6 11 h5 Bh7 12 Bd3 Bxd3 13 Qxd3 Qc7 14 Ne5 Rd8 15 Qe2 c5 16 0-0 cxd4 17 cxd4 0-0 is slightly
better for Black, A.Papastavropoulos-P.Prohaszka, Achaea 2012.
d) Black can meet 8 Bc4 in the standard way, completing development and then pushing ... c6-c5: for
example, 8 ... Ne7 9 0-0 0-0 10 Qe2 c5 11 Be3 Qc7 12 c3 Nd7 13 Ng5 cxd4 14 cxd4 Nf6 15 Rac1 Qa5
16 Kh1 h6 17 Nf3 Nf5 and Black has a pleasant game. White has not managed to launch an attack and the
hole left on e4 will soon start to tell in, D.Vargic-I.Saric, Bizovac 2005.
8 ... Ne7!
Now because White has played Bd3 it is vital to play the knight to e7. Remember to never let White
play f4-f5 easily.
9 Qe2
Instead, after 9 0-0 Nd7 10 Kh1 Qc7 11 Ne5 Rd8 12 Qg4 (12 Qe2 Bxd3 13 Nxd3 0-0 14 Bd2 c5! is
very thematic; Black does not have to think too hard about opening the position and already has a very
satisfactory set-up) 12 ... Nf6 13 Qh3 Bxd3 14 Nxd3 0-0 15 c4 (Y.Kim-D.Kononenko, Taganrog 2013) 15
... Qa5 White is struggling to find a plan while Black plans to undermine the with ... b7-b5.
9 ... Nd7
Black should develop calmly.
10 Bd2 Bxd3 11 Qxd3 Qc7 12 Ne5 Rd8!

Generating X-ray threats against the white queen.

13 Qe2 c5!
The battle for the e5-square has begun. Black begins to weaken White’s control of this critical square
by loosening up the white centre, while White, on the other hand, must do everything possible to maintain
it, otherwise his dark squares will become very weak.
14 Ne4 0-0
Preventing any tactics on f7. White must now do something about the pawn on d4 and so initiates a
series of exchanges, but from now on the initiative lies with Black.
15 Nxd6
After 15 c3 cxd4 16 cxd4 Nf5 the d-pawn is nothing more than a nuisance for White.
15 ... Qxd6 16 Nxd7 Rxd7 17 dxc5 Qxc5
 The position is already slightly better for Black.
18 0-0-0 Nd5!
Centralizing the knight.
19 Be3 Qa5 20 a3 Rc8
Black has a very pleasant position. He will double rooks on the c-file and play ... b7-b5-b4, attacking
the white king. Indeed, White is simply struggling to hold it together and didn’t survive in A.Kofidis-
S.Atalik, Panormo 1998.

B) 6 Bc4 e6 7 N1e2
Instead, 7 Nh3 would transpose to Line C, below.
7 ... Nf6
7 ... Bd6 is a viable alternative.
8 0-0 Nbd7!
This is safer than 8 ... Bd6 which leads to massive complications after 9 f4 Qd7 10 Bd3 Bxd3 11
Qxd3 g6 and now White can play the complicated 12 f5!.
9 f4
 One of the points behind delaying f2-f4 is that Black has developed the knight to f6, thereby lessening
his control over the important f5-square. This is a very enterprising move. White has the obvious and
aggressive idea of playing f4-f5, aiming to open the f-file and get the knight to the dangerous f5-square.
Black’s task naturally is to prevent or dissuade this from happening.
The move 9 f4 has from a positional standpoint some flaws: it makes White’s dark-squared bishop
bad (Black can even play a little bit on this idea by going ... Bf5), and it also gives Black the hope that
one day there just might be one of his minor pieces sitting pretty on e4.
9 ... Nb6!
Chasing the bishop back.
9 ... Bd6 allows White’s main idea, namely to open up the position for his bishops: 10 f5! exf5 11
Nxf5 Bxf5 12 Rxf5 Qc7 13 Bf4 0-0 14 Bxd6 Qxd6 15 Ng3 is slightly better for White and even has some
similarities to a King’s Gambit position.
Here White has two main moves:

B1) 10 Bd3
B2) 10 Bb3

B1) 10 Bd3

10 ... Bxd3 11 Qxd3 g6!

Covering the critical f5-square. In order to crash through White must now take tremendous risks.
12 f5
This does not give White enough compensation for the piece, but to be honest it is still quite scary. For
those who are too scared by it all, I would recommend playing 7 ... Bd6, allowing the knight on g8 to
develop to e7.
The more slowly White plays the position, the more risky the sacrifice f4-f5 becomes. As general
advice I would recommend that Black should clear a path for the king to flee to the queenside if needs be:
for instance, after 12 b3 I like 12 ... Qc7 (12 ... Qd7 is also possible) 13 Bb2?! (better is 13 c4 Bg7! 14
Ba3 0-0-0 with a complex position) 13 ... Bg7 14 c4 0-0 and Black’s position is rock solid.
Instead, 12 c4 was a recent try by White. After 12 ... Qd7 13 b3 Bg7 14 Ba3 Black should play 14 ...
h5 followed by queenside castling. In D.Bisby-P.Sowray, British League 2013, 14 ... Nc8?! was
premature as White had the sacrifice 15 f5 gxf5 16 Nxf5! exf5 17 Ng3 and it was already becoming
dangerous for Black.
12 ... gxf5
This is where White plays something scarily crazy.
13 Nxf5!? exf5

A critical position. Black must defend accurately here with different moves, so I will provide some
triggers on how to remember the right defences.
14 Bg5
Alternatively:
a) 14 Ng3 is another scary move when Black must play the accurate 14 ... Rg8!. The idea is to threaten
mate with ... Qd5: for example, 15 Re1+ (we can easily conclude that if White plays 15 Nxf5 we follow
our plans with 15 ... Qd5, threatening mate) 15 ... Kd7 (the king must now run) 16 Bf4 (if 16 Qxf5+ Kc7
and the black king is merrily on its way to safety) 16 ... Bd6 17 Bxd6 Kxd6 18 Nxf5+ Kc7 19 Re7+ Kb8
20 c4 a6. Black has started consolidating his king position, so, for instance, if 21 Rxf7 Ka7 22 Re1 Rf8
23 Rfe7 Nc8 and Black will soon start to mobilize his extra piece.
b) After 14 Qxf5 we must begin the defence with the theme cover, defend and counterattack: 14 ...
Be7! (this bishop is required to do two things: break any potential pin and cover the exposed king on the
e-file) 15 Bg5 Nbd5 (the only move to defend the other knight) 16 c4 h6! (a necessary counterattack) 17
Bh4 Ne3 18 Qf4 Nxf1 19 Rxf1.
 Now Black needs to play the deep defensive concept 19 ... 0-0! 20 Rf3 (if 20 Qxh6 Nh7! and 20 Ng3
Nh7 also saves the game) 20 ... Ne4!!, which is actually winning for us.
14 ... Be7
This useful covering move also defends the knight.
15 Ng3
Instead, 15 Rae1!? Nfd5 16 Qxf5 f6 17 Bh6 Qd7 parries all the threats.
15 ... Rg8!
Utilizing all the pieces in the defence of the black king.
16 Rxf5
16 Nxf5 Rxg5 17 Nxe7 is strongly met by 17 ... Kxe7! 18 Qe3+ Ne4!! 19 Qxe4+ Kf8 and the king has
escaped.
16 ... Rg6!

This is the idea - the rook protects the knight and builds a shelter for the black king.
17 Re1 Kf8 18 Nh5
After 18 Rxe7 Qxe7 19 Nh5 Rxg5! White loses on the spot.
18 ... Nbd5! 19 c4 Nxh5 20 Rxd5 Bxg5 21 Rxd8+ Rxd8 22 Qa3+ Kg8 23 Qxa7
 This eventually led to a draw in M.Khachian-K.Asrian, Yerevan 1999, but here Black should play:
23 ... Bf6!
White has problems with his own king’s safety.

B2) 10 Bb3

The bishop is out of play here and now Black is given the time to play a handy, little move.
10 ... Qd7!
As we have seen, the pawn push f4-f5 carries a lot less danger if Black can make a quick getaway
with his king.
11 c4
White can play several other moves:
a) 11 a4 allows Black to play 11 ... Nbd5! fixing some weaknesses. After 12 Bd2 Be7 13 Qe1 0-0 14
c4 Nc7 Black will secure the b4-square and target the d4-pawn, I.Bugarcic-P.Antanaskovic, Petrovac
2004.
b) 11 f5 is a critical pawn sacrifice, as well as the very move we have tried to prevent.

After 11 ... exf5 12 Nf4 0-0-0 13 c3 Bd6 14 h4 h6 15 Bc2 (A.Faizulaev-D.Khamrakulov, Tashkent

2008) 15 ... Rhe8 (a solid choice) 16 a4 Kb8 17 a5 Nbd5 18 Bxf5 Bxf5 19 Nxd5 Bxg3 20 Nxf6 gxf6 21
Qf3 Bxh4 22 Qxf5 Qxf5 23 Rxf5 Re2 Black has the preferable position.
c) 11 Qe1 has the sneaky if obvious threat of f4-f5. Following 11 ... 0-0-0! 12 Qa5 (12 a4 should be
met with the mirror pawn push 12 ... h5: for example, 13 a5 Nbd5 14 a6 h4 15 axb7+ Qxb7 16 Nh1 h3 17
g3 c5 and it is White’s king which will be the more vulnerable) 12 ... Kb8 13 a3 Bd6 14 c4 Bc7 15 Qc3
Nc8 16 Be3 h5 Black has the advantage and the easy attacking plan of ... h5-h4, H.Vedder-M.Becker,
Belgian League 2011.
d) 11 c3 doesn’t do too much apart from reinforcing the centre. After 11 ... Bd6 12 f5 Bxf5 13 Nxf5
exf5 14 Bc2 g6 15 Bg5 Ne4 16 Bxe4 fxe4 17 Bf6 0-0 Black’s king is safe for the time being as we will
quickly challenge the nasty bishop on f6. Indeed, 18 Qd2 Rfe8 leaves White with nothing.
11 ... Be7

The immediate 11 ... h5!? also looks promising: 12 f5 Bxf5 13 Nxf5 exf5 14 Qd3 g6 15 a4 Rd8 16 a5
Nc8 17 Bg5 Bg7 18 Bc2 Nd6 19 Nc3 0-0 and Black was a pawn up for not much in E.Gullaksen-
G.Østmoe, Oslo 2011.
12 Kh1?!
This move is too cautious.
12 ... h5!?
Black begins active operations to evict the knight from the g3-square.
13 f5 exf5
Preferable to 13 ... Bxf5, after which 14 Nxf5 exf5 15 Bg5 g6 16 Nc3 gave both sides equal chances
in M.Crosa-G.Milos, Sao Paulo 2003.
14 Bc2 Ne4!
 Guarding the f5-pawn.
15 Nxe4 fxe4 16 Nf4 Bf5 17 Nxh5?
A mistake, although likewise after 17 Qe2 Qxd4 18 Nxh5 Be6 19 Bxe4 0-0-0 Black is just winning.
17 ... 0-0-0!
White has big problems with his king.

C) 6 Nh3

6 Bc4 e6 7 Nh3 is another move order.

6 ... e6 7 Bc4
After 7 f4 Nd7 8 Bc4 Bd6 the king’s knight will soon emerge on e7.
7 ... Bd6
Delaying the development of the knights for a while and leaving the white knight out there on h3. Very
recently I tried a similar approach to our plan against the Bc4 and Ne2 variation, but the white knight is
actually better placed on h3 than on e2. Indeed, 7 ... Nf6 8 f4 Nbd7 9 0-0 Nb6 10 Bb3 Qd7 11 Qf3!?
revealed one of the points behind putting it on h3.
8 0-0
 White again plays a waiting game before advancing the f-pawn. Instead, 8 Nf4 Nf6 would transpose to
Line B.
8 ... Nd7

Black joins in and delays making a decision over the g8-knight.

9 Re1
Now that the rook has moved from the f-line, it will be safe to develop the knight to f6. However, 9 f4
is met by 9 ... Ne7!, 9 Nf4 Ngf6 10 c3 Nd5 11 Nxg6 hxg6 12 h3 N7f6 13 Bb3 Qc7 14 Qf3 0-0-0 is fine for
Black, and 9 Nh5 Bxh5 10 Qxh5 Ngf6 11 Qh4 0-0 12 Nf4 Nd5 leaves him with an excellent knight on d5.
9 ... Ngf6 10 Nf4 0-0 11 Nxg6
In the end White changes plan to winning the bishop-pair, but Black still stands well.
11 ... hxg6 12 Ne4 Nxe4 13 Rxe4 Nf6 14 Re1 Qc7 15 g3 c5 16 dxc5 Bxc5 17 Qe2 Rac8 18 Bb3
Rfd8 19 Bf4 Qc6
Black has fully equalized, S.Azarov-A.Riazantsev, European Club Cup, Rhodes 2013.

Summary
The f2-f4 push can be played directly or in a more controlled manner. I believe that playing f4 directly,
i.e. on move 6, is not subtle enough and Black simply needs to develop with the idea of preventing f4-f5.
Black’s counterplay is based on achieving ... c6-c5 to exploit White’s weakened dark squares.
Positionally speaking, Black may also be able to establish a base on e4.
Should White play f2-f4 after Black has played ... Nf6 then Black needs to take measures against f4-
f5. He should not make the break too easy for White and so ... Nbd7-b6 is a good plan. Eventually White
may still play f4-f5, but he will have to sacrifice material in order to do so.
Finally, combining Bc4 with Nh3 is a less effective method of playing f2-f4. Here Black should delay
developing the g8-knight until it is clear that White will not be playing for f2-f4-f5.
Chapter Four
The Challenging Nf4
1 e4 c6 2 d4 d5 3 Nc3 dxe4 4 Nxe4 Bf5 5 Ng3 Bg6

One of the most important ideas we shall see in this chapter is that of harassing the bishop, with
particular focus on the knight manoeuvre Ng1-e2/h3-f4, with or without the pawn push h2-h4. What makes
these systems so bothersome is that there are various move-orders and ideas. In light of this I will be
breaking down the variations into the following piece formations: Bc4, h4 and Nf4, as well as h4, Bc4
and Nf4. As we will see, White can’t go Bc4, Nf4 and then h4.
I have nicknamed this variation the full works, because White implements all of his dangerous piece
formations in one go, i.e. he throws the full works at us.
By move 9 we should see a position similar to this one.
 The danger behind this system can be manifested in two ways:
a) A piece sacrifice on e6.
b) The knight jump Ngh5 with the idea of pawn storming the black king.
We will then move on to what happens should White play Bc4 and Nf4, omitting an early h4.

It is now time for me to formulate another rule: Nf4 = Bd6.

Quite simply, when a knight arrives early on f4, as in the diagram above, Black must play ... Bd6.
This ensures that the following position should arise.
 White has just played 9 h4?! and we need to be answer this move with the counterattacking 9 ... Qc7!.

The Second Face of Nf4

The concept of Ne2/h3-f4 is not tied inextricably to the move Bc4. Indeed, Nf4 in its second form aims to
do one of two things:
a) Harass the bishop by playing h2-h4.
b) Exchange itself for the bishop on g6.
Thus we will see White going 6 N1e2 e6 7 Nf4 Bd6 8 h4, as well as a more dangerous version where
he first plays the prophylactic 8 c3 Qc7 and only then 9 h4!?.

Please note that in both situations Black has prioritized bringing the bishop to d6 over developing
the knight from g8.
In both scenarios Black must counterattack against the knight on f4 with ... Qc7, forcing White to
overprotect the f4-knight as h4-h5 is not playable, as we will see.

Swapping the Knight for a Bishop

White expends a lot of time and energy if he wishes to nurture a long-term advantage of the bishop-pair.
The game becomes a lot quieter than in previous lines, but Black can nullify this advantage by securing
strong posts for his knights and by playing for a swap of dark-squared bishops with a trade on f4.
Please note that this swap can be done in two ways: with the immediate exchange on g6 after 6 N1e2
e6 7 Nf4 Bd6, which gives us a useful half-open h-file, and with 8 Ngh5, which Black should simply meet
with 8 ... Bxh5 9 Nxh5 g6!.

Now let’s look at the themes and the moves in greater detail:

A) 6 Bc4
B) 6 N1e2
C) 6 Nf3

Instead, 6 h4 h6 7 Bc4 (7 N1e2 e6 8 Nf4 Bh7 9 Bc4 is another possible transposition) 7 ... e6 8 N1e2
transposes to Line A, while 6 Nh3 (of course it could also be 6 N1e2) 6 ... e6 7 Nf4 Bd6 8 Bc4 Nf6 9 h4
is an incorrect move order, as we will see at the beginning of Line A2.

The Full Works System

A) 6 Bc4 e6

We now have:

A1) 7 h4
A2) 7 N1e2

Note that in Line A2 White will meet 7 ... Nf6 with 8 Nf4, rather than 8 h4, which would take play back into Line A1.

A1) 7 h4
Let’s face it, 7 Nf3 is simply not very problematic to Black. My advice would be to develop normally and
get on with things. Indeed, after 7 ... Nd7 8 Qe2 (if 8 0-0 Ngf6 9 Re1 Qc7 10 c3 Bd6 11 Bg5 0-0 and
Black will play for a well-timed ... c6-c5 break) 8 ... Ngf6 9 0-0 Be7 10 Re1 0-0 11 Bb3 (the sacrifice 11
Bxe6 is not quite correct, in view of 11 ... fxe6 12 Qxe6+ Rf7 13 Ng5 Bf8 14 c3 Qa5 15 Nxf7 Bxf7 16
Qf5 Qb6) 11 ... a5! 12 a4 Re8 13 Bd2 Nd5 14 Rad1 Qb6 Black had a good position in B.Nikitinyh-
A.Zontakh, St Petersburg 2005.
7 ... h6 8 N1e2 Bd6 9 Nf4 Bh7
 We have now reached a critical position for this chapter. It is my opinion that because White has
committed himself by advancing the h-pawn, he cannot simply continue with natural moves. Indeed, I
believe that White must be prepared to continue aggressively. These ‘attacks’ can take form two different
forms:
a) White can try to use the tactically soft point on e6 to catch the king in the middle or to exploit the
loose bishop on d6. In its most direct form, White will castle quickly and strike whilst the black king is
still in the centre.
b) The second plan is a more sophisticated form of attack, namely to target the g7-pawn with Ngh5. In
the most aggressive version of this plan, White seeks to castle queenside and open lines with a pawn
push.
We have:

A11) 10 0-0
A12) 10 Ngh5
A13) 10 c3
A14) 10 Qe2

A11) 10 0-0
 White is planning to use his advantage in development by blasting the position open with a piece
sacrifice on e6. It’s true our monarch gets caught in the centre, but the attack isn’t so dangerous.
10 ... Bd6
The natural 10 ... Nd5? would be a big mistake after 11 Bxd5! cxd5 12 Qh5! and suddenly it is very
difficult to meet the threat of Nxe6.
11 Nxe6
Going straight ahead. Otherwise:
a) 11 Ngh5?! leads to nothing once White has castled kingside: 11 ... 0-0 12 Re1 Re8 (Black’s idea is
to play ... e6-e5 at the right time, while avoiding 12 ... Nbd7? 13 Nxe6! fxe6 14 Rxe6, as in P.Simacek-
S.Mazur, Sibenik 2012) 13 c3 (13 Nxf6+ was another recent try, but after the natural 13 ... Qxf6 14 Nh5
Qxh4 15 g3 Qd8 16 Qg4 Bf8 17 Bxh6 Bf5 18 Qf4 gxh6 19 g4 Bg6 20 Nf6+ Kg7 21 Nxe8+ Qxe8 22 Kg2
Nd7 23 Rh1 Kg8 Black was already winning in T.Manhardt-R.Kreisl, Austrian League 2014) 13 ... Nbd7
14 Bd2 e5 15 Qb3 Rf8 and Black was much better because of the h4 weakness in J.Bellon-Y.Seirawan,
Las Palmas 1981.
b) 11 Re1 trains yet another piece on e6.

Please don’t waste time on this position with moves such as ... Qc7. The black monarch needs a safe
home and fast. Castling also has the advantage of taking the sting out of any form of danger from Ngh5, so
11 ... 0-0 and then:
b1) If 12 Bd3 Black can play in a very natural manner: 12 ... Bxd3 13 Qxd3 Nbd7 14 Bd2 Re8 15 Qf3
Qb6 16 Ne4 Nxe4 17 Rxe4 Nf6 18 Nh5 Nxe4 19 Bxh6 Qxd4 and now after 20 Bxg7 (U.Hartig-F.Levin,
Oberhausen 2011) 20 ... Qc5 White does not have an attack and 21 Qg4 Qxf2+ 22 Kh1 f5 wins on the
spot.
b2) 12 c3 Re8! (I must mention that the developing 12 ... Nbd7? again fails to 13 Nxe6 fxe6 14 Rxe6,
M.Lagarde-L.Handler, Herceg Novi 2008) 13 Qf3 Nbd7 14 Be3 (after 14 Bd3 e5! 15 Bxh7+ Kxh7 16
Nfh5 exd4 17 Rxe8 Qxe8 Black is doing well) 14 ... Nd5.

This is a model set-up for Black. Moreover, after 15 h5 Qf6 16 Nge2 N5b6 17 Bb3 e5 Black was
already much better in A.Longson-B.Lalic, British League 2000.
c) After 11 Nfh5 0-0 White doesn’t have any attack.
11 ... fxe6 12 Bxe6 Qc7 13 Re1 Nbd7! 14 Bg8+

The last moves were fairly forced and now I am still recommending:
14 ... Kd8
14 ... Kf8 is also well-known, but is more complicated as Black has to remember Botvinnik’s stunning
16 ... g6!!. That arises after 15 Bxh7 Rxh7 16 Nf5:
a) 16 ... Kg8 is the computer’s suggestion, and if 17 Re6 (or 17 Qf3 Bf8) 17 ... Bf8 18 c4 g6 19 Ng3
Rf7.
b) However, it is Botvinnik’s wonderful 16 ... g6!! that gives Black a good game. After 17 Bxh6+ Kg8
18 Nxd6 Qxd6 19 Bg5 Re7 20 Qd3 Kg7! the piece was superior to the three pawns in M.Tal-
M.Botvinnik, World Championship (Game 9), Moscow 1960.
15 Bxh7 Rxh7

16 Qf3
Alternatively:
a) 16 Nf5 g6 17 Nxd6 (17 Nxh6 Bh2+ 18 Kf1 Bf4 19 Ng4 Bxc1 20 Qxc1 Nxg4 is winning) 17 ...
Qxd6 18 Qf3 Qd5 19 Qg3 Kc8 20 Qxg6 b6 21 Bxh6 Kb7 and now that the king is tucked away safely on
b7 Black can set about winning back some of the pawns.
b) 16 Qd3! Kc8 17 Nf5 g6 18 Nxd6+ Qxd6 19 Qxg6 b6 and Black is safe.
16 ... Bxg3!
Eliminating the strong knight on g3 which threatened to leap into f5.
17 fxg3 Qd6 18 Bf4 Qd5!
I much prefer this to 18 ... Qxd4+ when, despite what the computer suggests, Black is playing with
fire.
19 Qe3 g5 20 hxg5 hxg5 21 Bxg5 Kc7
Black is prepared to give back the piece in order to obtain a very powerful initiative, while avoiding
21 ... Rg7 22 Bh4.
22 c4 Qxc4 23 Bxf6 Nxf6 24 Qe5+ Kb6
 25 b3
25 Qxf6 Rah8 26 Qd6 is White’s best defence and after 26 ... Rh1+ 27 Kf2 Qc2+ 28 Kf3 Qd3+ 29
Kf2 Qd2+ 30 Re2 Rf8+! (deflecting the queen from the d4-square) 31 Qxf8 Qxd4+ 32 Re3 Qxb2+ 33 Kf3
Rxa1 34 Qd8+ Ka6 35 Qd3+ b5 Black is better.
25 ... Qc3 26 Rac1 Qb4 27 Rc4
Not 27 Qxf6 Rah8 when White has problems.
27 ... Qa5 28 Rc5? Qd2 29 Re2 Qd3 30 Rc1 Nd5 31 a4 Re7 32 Qxd5 cxd5 33 Rxe7 Qxd4+
Here I won in Z.Hagarova-J.Houska, Chisinau 2005.

A12) 10 Ngh5!?

White plans to tie Black dark-squared pieces to the defence of the g7-pawn. However, this is quite
premature as Black hasn’t even castled kingside and we can counter along simple lines.
10 ... Nbd7 11 c3 Nxh5! 12 Nxh5 Nf6!
Eliminating the troublesome knight.
13 Bf4
13 Rh3 is unusual and really too unconventional to be good. After 13 ... Nxh5 14 Qxh5 Bf5 15 Rh1
(and that was the end of that experiment in E.Both-G.Schebler, Belgian League 2009) 15 ... Bd6 16 Be3
b5! 17 Be2 0-0! now that White can’t castle kingside, the perfect strategy is to make the queenside as
insecure as possible.
Another take is 13 Be3 when 13 ... Nxh5 14 Qxh5 Bd6 is now possible as after the greedy 15 Bxe6
Black wins a piece with 15 ... g6 16 Qxh6 Bf8 17 Bxf7+ Kxf7 18 Qf4+ Kg8 19 0-0-0 Qd5.
13 ... Nxh5 14 Qxh5 Bd6 15 Be3
15 Bxe6? does not work due to 15 ... Bg6! and Black wins a piece. Instead, 15 Qg4 Bxf4 16 Qxf4 0-0
17 0-0 a5 18 a3 a4 19 Rad1 Ra5 gave Black the preferable position in J.Wegerer-B.Franciskovic,
Schwarzach 2009.
15 ... Qf6
Black had no problems with his excellent bishop on h7 and enjoyed easy development in R.Zelcic-
A.Dreev, Neum (blitz) 2000.

A13) 10 c3

GM Matthew Turner once chose this move against me. The idea is to support the d4-pawn and at the
same time protect the c2-pawn from that bishop devilishly hiding on h7.
Remember that a formula to follow is that if the white king is in the centre, we want to strike with ...
e6-e5 rather than ... c6-c5.
10 ... Bd6!
The natural choice. In my original book I liked 10 ... Nd5. By closing the diagonal for the bishop,
Black prevents any sacrifices happening on e6 as 11 Qh5 Nd7 12 Nxe6 fails to 12 ... Qe7. However, 11
Qe2 Bd6 12 Nxd5 Bxg3 13 Nf4 Bxf4 14 Bxf4 is preferable for White.
11 Nfh5!?
Despite appearances, a knight on this square is not as scary as it looks. The alternatives are:
a) Should White play quietly with 11 Qf3 0-0 12 Be3 I like 12 ... Qa5 13 Nfh5 (White gets into
trouble after the reckless placement of the king with 13 0-0-0: 13 ... Nbd7 14 Nfh5 b5 15 Bb3 b4 and
Black has the initiative) 13 ... Nxh5 14 Nxh5 Nd7 15 0-0 Qf5 16 Qxf5 Bxf5 17 Be2 when Black has the
straightforward plan of utilizing the half-open d-file.
b) Funnily enough 11 Ngh5?! is the less menacing knight move primarily because it makes the pawn
break ... e6-e5 all the more attractive, particularly as no white knight can now jump into f5. After 11 ... 0-
0 12 Qf3 (or 12 0-0 Nbd7 and Black is planning the counter-strike ... e6-e5) 12 ... Nbd7 13 Bd2
(likewise, if 13 g4 e5!) 13 ... Re8 Black is completely safe and is planning the to break: for example, 14
0-0-0 e5 15 Qg3 Bf8 16 dxe5 Ne4 and Black either regains his material or wins the queen.
c) 11 Bxe6? loses a piece to 11 ... Qe7!.
11 ... 0-0!

Black is forced to play this committal move, but we have enough defenders.
12 Bg5
Obviously one of the scariest moves on the board. Other moves allow Black to prepare the counter-
strike ... e6-e5:
a) 12 Qf3 Nbd7 13 Bf4 (after 13 Bg5 and Black can get away with taking: 13 ... hxg5 14 hxg5 Nd5 15
Qg4 Bg6 16 Qh4 f5 and Black’s king is relatively safe) 13 ... e5 14 Be3 exd4 15 Bxd4 Re8+ 16 Kf1 Be5
and Black has a fantastic position, especially with the white king caught in a difficult position, M.Vaculik-
V.Kulakov, Pardubice 2008.
b) 12 Bxh6 doesn’t work: 12 ... gxh6 13 Qd2 Bxg3! forces White to recapture with 14 Nxg3 Kg7 with
no problems for us.
c) After 12 Nxg7 Kxg7 13 Qd2 Black has the very calm 13 ... Ng8 14 Nh5+ Kh8 15 Bd3 Bg6, again
with little trouble ahead.
12 ... Nbd7!
A very natural move and an important one. To understand the defensive nature of the move, let’s look
at what happens should Black capture on g5: 12 ... hxg5 13 hxg5 Nd5 14 Bxd5 exd5 15 Nf6+. The
situation is far from clear, but this is the point: Black needs a knight on d7.
13 Qf3
13 Nxf6+ makes life easier for Black: 13 ... Nxf6 14 Qf3 b5! (a nice move that also discourages
White from castling queenside) 15 Bb3 (after 15 Bd3 Bxd3 16 Qxd3 Re8 now 17 0-0-0 is not possible as
Black can simply capture with 17 ... hxg5 18 hxg5 and follow through with 18 ... Bf4+, winning the g5-
pawn) 15 ... a5 and I don’t see how White can easily continue his attack:
a) Capturing the pawn is not possible as the queen get side-tracked by 16 Qxc6 hxg5!.
b) 16 Nh5 hxg5! (critical and best, although 16 ... Be7 17 Bxf6 Bxf6 18 Nxf6+ Qxf6 19 Qxf6 gxf6 is
the simplest route out of danger) 17 hxg5 Ne4 and now White must play the incredible 18 g6!! Bxg6 19
Qh3 (this is the only move) 19 ... Bh7! 20 Nxg7! Ng5! 21 Qh6 Qe7 22 Nh5 f5 23 f4 Ne4 24 Bxe6+ Kh8
with a complicated position.
13 ... hxg5
 13 ... Bxg3 is a calm alternative: 14 Nxg3 Qa5 15 Bxf6 Nxf6 16 0-0 was M.Uribe-P.Lafuente, Cali
2007, and here Black should play 16 ... Rfd8, planning to break with ... c6-c5 at the right moment.
14 hxg5 Nd5

The f6-square is covered so there are no dangerous knight checks.

15 Qg4
15 Nxg7 fails to 15 ... Qxg5.
15 ... Re8
Vacating the f8-square for a knight to defend the bishop on h7. The alternative is 15 ... Bg6, moving
the bishop off the vulnerable h-file:
a) 16 Qh4 (or 16 Qh3) 16 ... f5! 17 0-0-0 (not 17 gxf6 N7xf6 18 Qg5 Nxh5 19 Qxg6 Nxg3 20 Qh7+
Kf7 21 fxg3 Bxg3+ 22 Ke2 Qg5 23 Bxd5 exd5 24 Rhf1+ Bf4 and Black is winning) 17 ... N7b6 sees
Black keeping things under control.
b) 16 0-0-0! b5 17 Bd3 Qe7 18 Qh4 is complicated.
16 Qh4
After 16 Nxg7 Kxg7 17 Qh3 the black king needs to run away from the action with 17 ... Kf8 18 Qxh7
Bxg3 19 g6 Bxf2+ 20 Kxf2 Qf6+ 21 Kg1 Qxg6 and Black should be OK, although the game is still
incredibly complicated: for example, 22 Bxd5 exd5! 23 Qxg6 fxg6 24 Rh7 Nf6 25 Rf1 Kg8 26 Rxb7 Re6
and Black is by now winning.
16 ... Bxg3 17 Nxg3 Nf8
 As Bent Larsen used to say, with a knight on f8 you don’t get mated.
18 Ne4
18 Bd3 is neutralized by 18 ... f5!.
18 ... Ne7
Heading to g6 to cover the h8-square.
19 0-0-0 Neg6 20 Qh2 b5 21 Bd3 b4 22 c4
After 22 cxb4 a5 23 bxa5 Qxa5 24 a3 Red8 25 Bc2 Rdb8 Black has the better chances.
22 ... Re7

23 f4 Rd7
Black is tangled up, but undoubtedly has the superior position thanks to his extra piece.

A14) 10 Qe2
 A developing move that targets the e6-pawn. Please bear in mind that White may wish to castle
queenside and continue the middlegame battle from there.
10 ... Bd6
Now we have two main moves to explore:

A141) 11 c3
A142) 11 Ngh5

Lesser options are:

a) 11 Bxe6? is a big blunder: 11 ... 0-0! and White cannot retreat the bishop without an unpleasant pin
happening on the e-file.
b) Should White try and delay the kingside attack with 11 Be3 then 11 ... b5 discourages him from
castling queenside. With 12 Bb3 Nbd7 13 a4 a6 14 0-0 0-0 we see both sides going short and now Black
can start to exploit White’s clumsy pieces and pawn on h4.
c) This time 11 Nfh5 is not as dangerous because there are no pawn thrusts to open the kingside: 11 ...
0-0 12 Be3 (if 12 Bg5 hxg5 13 hxg5 Qa5+) 12 ... Nbd7 13 0-0-0 Qa5 14 Kb1 Rac8 15 Bd3 Bxd3 16
Rxd3 Rfd8 and Black has the plan of playing for ... c6-c5.

A141) 11 c3
A quiet move. Black should now concentrate on completing development and at the right time go ... Nf6-
d5, dislodging the white knight from f4.
11 ... 0-0
 12 Be3
Freeing the queen’s rook. Others:
a) 12 Nfh5 should be met with 12 ... Nbd7! (don’t be tempted to swap a pair of knights just yet) and
now 13 Bg5? doesn’t work in view of 13 ... hxg5 14 hxg5 Nd5 (again covering all the important squares)
15 Nxg7 Qxg5! 16 Nxe6 fxe6 17 Qxe6+ Kh8 18 Rxh7+ Kxh7 19 Bd3+ Kg7 20 Qxd7+ Qe7+, winning.
b) 12 0-0 Re8 and then:
b1) 13 Be3 Nbd7 14 Rad1 Nd5! (a model idea; 14 ... Nb6 is also acceptable, but only to put a knight
on d5) 15 Bxd5 exd5 and Black already had the advantage on account of the weak h4-pawn and the
bishop-pair in M.Darmarakis-C.Goritsas, Greek Team Championship 2013.
b2) 13 Qf3 Nbd7 14 Nfh5 Nxh5 15 Qxh5 Nf6! 16 Qf3 Nd5 17 Bxd5 cxd5 (17 ... exd5 was perhaps a
little bit symmetrical in J.Eric-V.Gunina, Khanty-Mansiysk Olympiad 2010) 18 Bf4 Bf8 and Black will
use the half-open c-file to play for a long-term advantage.
12 ... Nbd7 13 Rd1
Instead, after 13 Nfh5 Re8 14 Bg5 hxg5 15 hxg5 Nd5 16 Nxg7 Nf4 White is too late with his attack.
13 ... Nd5
I really like this thematic leap. 13 ... Qc7 14 Nfh5 Nd5! is another way of playing.
14 Qf3
Alternatively, 14 Nxd5 cxd5 15 Bd3 Bxd3 16 Qxd3 Bxg3! 17 fxg3 Nf6 and Black stands very well
indeed.
14 ... Nxe3 15 fxe3 Kh8! 16 Ngh5 e5
Black is powerfully placed.
17 dxe5 Nxe5 18 Qg3 g6
Already I would say that Black is winning.

A142) 11 Ngh5
 I am not going to lie to you dear reader, this is a dangerous move. White, however, is playing a
dangerous game by committing himself to such an early attack. Should Black manage an early and
effective counter-strike (here ... e6-e5 springs to mind), then White will soon find himself in lots of
trouble.
11 ... 0-0
Again, Black should defend coolly by housing his king thus.
12 g4!?
This is indeed scary, but Black can counter in the centre.
12 Nxf6+ Qxf6 presents no danger to Black and 12 c3 is too slow, as Black demonstrated perfectly
with 12 ... Nbd7 13 0-0 e5! 14 Nh3 Re8 15 Qf3 exd4 16 cxd4 b5 17 Bb3 Be4 18 Qd1 Nxh5 19 Qxh5 Qf6
and White soon lost in A.Gurel-E.Ozkan, Antalya 2011.
12 ... Re8!

Preparing the counter-strike ... e6-e5, but 12 ... Nxh5 13 Nxh5 e5 14 g5 is just too crazy for me. After
14 ... exd4 15 Kf1 a very complicated position ensues.
13 g5!?
There is nothing for it. White has to roll with the consequences of his 12th move and push on, as 13
Be3 Nd5 (again we sees this move) 14 g5 Bxf4 15 Nxf4 (or 15 Bxf4 Nxf4 16 Nxf4 Qxd4 and White’s
attack has broken down) 15 ... Nxf4 16 Bxf4 Qxd4 17 Be5 Qe4 18 0-0-0 Nd7! should be better for Black.
13 ... Nxh5 14 Nxh5 Bg6!?
The easy (or some might say lazy) way out.
Of course, 14 ... e5!? is the most principled move, but the position becomes very wild. It’s hard to
fully analyse everything here, but may main line runs 15 Qf3 (15 0-0 Qd7! prevents the queen from
coming to g4 and after, say, 16 Qf3 hxg5 17 Bxg5 Bg6 18 Bf6 Bxh5 19 Qxh5 gxf6 20 Rae1 Na6 21 Bd3
Kf8 22 Bf5 Qd8 23 f4 the position is unclear) 15 ... exd4+ 16 Kf1 Qe7. To be honest I am not sure I can
give this position an accurate assessment. One line runs 17 gxh6 Qe1+ 18 Kg2 Be4 19 Bxf7+ Kh7 20
Rxe1 Bxf3+ 21 Kxf3 Rxe1 22 hxg7 Nd7 23 g8Q+ Rxg8 24 Bxg8+ Kxg8 25 b3 and after all this Black has
a small edge because of 25 ... d3 26 cxd3 Be5 27 Rb1 Nc5 when White must start to return some of his
extra material.
15 Kf1
15 gxh6 would be a huge blunder due to 15 ... Qa5+, winning the knight, while 15 Nf4 Bxf4 16 Bxf4
Qxd4 17 Be5 Qe4 18 0-0-0 Nd7 again should be slightly preferable for Black: for instance, 19 Bg3 Nb6
20 Bd3 Qxe2 21 Bxe2 hxg5 22 hxg5 Rad8 and Black can begin to grind.
15 ... Bxh5 16 Qxh5 g6 17 Qf3 h5 18 Kg2 Nd7

19 Bb3!
White has to take action against the threatened ... e6-e5.
19 ... a5 20 a4!
More or less forced as otherwise the bishop gets chased away.
20 ... Nb6 21 c4 c5 22 Be3 Qc7 23 Rhd1 Rac8 24 Rac1 cxd4 25 Bxd4 Nd7
Black has everything under control.

A2) 7 N1e2
White holds back on h2-h4, at least for the time being.
7 ... Nf6
7 ... Bd6 is also a valid possibility with the idea of avoiding any f2-f4 plans: for instance, 8 0-0 (the
main difference to our main line is that after 8 h4 h6 9 Nf4 Black must play 9 ... Bxf4 10 Bxf4 Nf6 with a
complex position) 8 ... Qc7 9 f4 and now the point is that Black has 9 ... Ne7!.
8 Nf4
 Of course, 8 h4 would take play back into Line A1.
8 ... Bd6

White has three main options here:

A21) 9 h4?!
A22) 9 Bb3
A23) 9 c3

9 0-0 is another calm move. After 9 ... Qc7 10 Qf3 0-0 11 Nxg6 hxg6 12 Bg5 Nbd7 13 Rad1 Nd5 14 Bb3
b5! Black will reroute the knight on d7 to f6 and then react to White’s plans, J.Radulski-A.Anastasian,
Antalya 2004.

The Wrong Way to Achieve the Full Works

A21) 9 h4?!

This is inaccurate as Black can apply pressure on the f4-knight.

9 ... Qc7!
 After all this is the whole point of ... Bd6.
10 Nxg6
Not an optimal exchange, but neither are the alternatives particularly rosy:
a) 10 Qf3? throws a pawn after 10 ... Bxc2!. In my old analysis I stopped here, but when I reached this
position over the board I became a little bit unsure. It’s good to know that the capture is viable, which it
is, as we’ll see after 11 Nfh5 Nxh5 12 Nxh5:

a1) 12 ... 0-0 13 Nxg7! is what concerned me. The position is somewhat murky after 13 ... c5.
a2) 12 ... Rg8! 13 0-0 Nd7 14 Re1 0-0-0! is the cool way to proceed. The point is 15 Qxf7 Bg6 16
Qf3 Rgf8 17 Qe2 Nb6 18 Bxe6+ Kb8 19 Bg5 Rde8 and while Black may be a pawn down, he will attack
and have all the fun. Here 20 Qg4 Na4 21 Re2 Nxb2 22 Rxb2 Bxh5 23 Qxh5 Rxe6 24 Rab1 b6 25 Qxh7
Rfe8 26 Be3 Bh2+ 27 Kh1 Bf4 is just an example of how things might go.
b) 10 h5 is met with the standard 10 ... Bxc2! 11 Nxe6 fxe6 12 Qxc2 Bxg3 13 fxg3 Qxg3+ 14 Qf2
Qxf2+ 15 Kxf2 Nd5 and although White has the bishop-pair in compensation for the pawn deficit, Black’s
healthy knight on d5 gives us a good game.
c) 10 Nge2 Nbd7 11 f3 (White starts to weave a net in order to trap the g6-bishop; after 11 h5 Bf5 the
black bishop is safe from harm) 11 ... e5 (the principled strike; White has neglected his development and
must be punished energetically) 12 dxe5 Nxe5 13 Bb3 0-0-0 14 Qd4 Rhe8.
 This position is a dream for Black. Moreover, 15 Qxa7 Bb4+ 16 c3 Nd3+ 17 Nxd3 Bxd3 18 Qa8+?
Kd7 wins on the spot.
10 ... hxg6
Now White is lumbered with a weak pawn on h4.
11 Qf3 Nbd7 12 Bg5 c5!
Making it difficult for White to castle queenside. There are no hard and fast rules as to which pawn
break Black should aim for, so I would judge each situation on merit. Here the decision is taken to play ...
c5 to target White’s bishop on c4 and discourage him from castling queenside.
13 dxc5
13 0-0-0 does not quite work: 13 ... cxd4 14 Bb5 (after 14 Rxd4 Be5 Black is much better, as shown
by 15 Bf4 0-0-0 16 Bb5 Bxf4+ 17 Rxf4 Ne5) 14 ... Rc8 15 Kb1 Kf8! (castling would play into White’s
hands and 15 ... Qxc2+? is just downright bad because the pin on the d7-knight makes the f6-knight a little
loose: 16 Ka1 Qc7 17 Rc1 Qd8 18 Ne4 Be7 19 Rxc8 Qxc8 20 Rc1 and White has an irresistible attack)
16 Rxd4? (White must defend the f2-pawn) 16 ... Qxc2+ 17 Ka1 Be5! 18 Qa3+ Kg8 19 Rxd7 Nxd7 20
Bxd7 Rxh4!! (this deflection tactic is simply crushing) 21 Rb1 Rd4 0-1, N.McDonald-S.Nagle, Budapest
2004. Mate follows after 22 Bxc8 Rd1.
13 ... Qxc5
 The problem for White is that Black is threatening to play 14 ... Qe5+, picking up the b2-pawn and the
c2-pawn is a little loose.
14 Bxf6
After 14 Bb3 Qe5+ 15 Be3 Qxb2 16 0-0 Qc3 17 Qxb7 Rb8 18 Qxa7 Rxh4 19 Rad1 there is the nifty
19 ... Qc8 20 Rxd6 Ra8, trapping the queen.
14 ... Nxf6 15 Bb3 0-0-0 16 c3
Not ideal, but after 16 0-0-0 Rxh4 17 Rxh4 Qg5+ 18 Kb1 Qxh4 Black is just a pawn up.
16 ... Bxg3 17 Qxg3 Rh5

The situation for the white king is fast becoming dire.

18 0-0 Rdh8
Now that the h4-pawn cannot be defended Black is completely winning.
19 Qf3 Rxh4 20 g3 Rh2 21 Bc2 g5 22 Be4 g4 23 Qf4 Qh5
Mate was unstoppable in S.Tiviakov-A.Dreev, Ubeda 1999. Almost never does one see Tiviakov
routed so.

A22) 9 Bb3
 At first sight this move seems a little incomprehensible, but the point is to defend the c2-pawn so that
White can support the knight on f4 with Qf3. This line became a speciality of the Dutch Grandmaster
Sergei Tiviakov after his debacle against Dreev, which we’ve just seen.
9 ... a5!
Black plans a very easy game based on establishing a knight on d5 and mobilizing the b8-knight to f6.
Then we need to react in the appropriate way to anything that White does. With a strong knight on d5,
Black is not afraid of letting White have the bishop-pair. The move ... a5 is in fact designed to keep a
knight on d5 by making White play either a4, a3 or c3. The point is that all those moves weaken the
queenside and so make a later kick-back with c3-c4 all the more risky.
10 a4
Should White choose not to give Black the b4-square and plays the more solid 10 c3, we should put
pressure on White to resolve the situation with the f4-knight. This can be done with 10 ... Qc7 11 Qf3 a4!
(seizing space and securing the b4-square anyway) 12 Nxg6 hxg6 13 Bc2 Nbd7 14 0-0 Nd5. This
centralizes the knight and threatens ... Bf4, exchanging the dark-squared bishops. Play might continue 15
c4 Nb4 16 Bb1 c5! 17 d5 (17 Be3 cxd4 18 Bxd4 Ne5! attacks the c-pawn) 17 ... exd5 18 Re1+ Kf8 19 a3
(if 19 cxd5 Nf6, while 19 Bxg6 fails to 19 ... Ne5) 19 ... Nc6 20 cxd5 Nd4.
 Even though Black has not castled, he has an excellent position. His knight is situated on the extremely
strong square d4, he has attacking potential down the h-file, and all the while White is behind in
development.
Instead, 10 a3 was the recent choice of expert Tiviakov when faced with our system. However, after
10 ... Qc7 11 Qf3 Nbd7 12 c3 0-0 13 0-0 Bxf4 14 Qxf4 (14 Bxf4 Qb6! 15 Bc4 Nd5 is all very standard
too) 14 ... Qxf4 15 Bxf4 a4 16 Ba2 b5 (the light-square pawn-chain does a magnificent job of shutting out
the a2-bishop) 17 Bd6 Rfc8 18 f3 Nb6 19 Rf2 Nc4 20 Bb4 Nd7 21 Rd1 c5 22 Bxc5 he could find nothing
better than to offer a draw in S.Tiviakov-B.Macieja, Nakhchivan 2013.
10 ... Qc7!

Putting pressure on White over the f4-knight.

11 Nxg6
If 11 Qf3 Nbd7 12 h4 e5 13 h5 exf4 14 hxg6 fxg3 15 gxf7+ Kf8 Black has won material.
11 ... hxg6 12 Qf3
Preventing the dark-squared bishop swap, which 12 Qd3 Nbd7 13 c3 Bf4! would have carried out.
12 ... Nbd7 13 0-0 c5 14 dxc5 Nxc5 15 Bc4 0-0 16 Rd1 Rad8 17 Bg5 Be7
Black is active enough to compensate for the bishop-pair.

A23) 9 c3
 Another solid choice.
9 ... Qc7!
Again we see Black putting pressure on the knight on f4.
10 Qf3 Nbd7 11 0-0
The direct 11 h4 is met with 11 ... c5 as the black bishop has a little hole on c2 it can escape to if
needs be. Now:
a) 12 dxc5 Bxc5 13 Bb5 0-0-0 14 0-0 (after 14 h5 Ne5 White finds himself very quickly in a lot of
trouble) 14 ... Ne5! leaves Black very actively placed.
b) 12 Bb3 cxd4! 13 h5 Ne5 14 Qd1 (after 14 Ba4+ Ke7 15 Qd1 Nc4 16 hxg6 Bxf4 17 Bxf4 Qxf4
Black is actually much better) 14 ... Bf5 15 Nxf5 exf5 16 cxd4 (or 16 Qxd4 0-0) 16 ... Nc6 17 g3 Qe7+
18 Ne2 0-0 19 0-0 Rad8 with play against the d4-pawn.
11 ... Nb6
Playing towards d5.
12 Bb3
12 Bxe6 is a variation which I once tried out on my iPad. After 12 ... fxe6 13 Nxe6 Qf7 14 Nf5 Bxf5
15 Qxf5 Qg6 16 Re1 Qxf5 17 Nxg7+ Kd7 18 Nxf5 Rae8 it is easier to play the side with the extra piece.
Moreover, after 19 Bh6 Rhg8 20 f3 Nfd5 21 Re4 Nc4 22 Rae1 Bc7 I think Black has good chances to
play for a win.
12 ... 0-0-0!
 Black quite rightly adopts an aggressive approach to the position as White’s bishop is in the way of
his pawns, meaning that he can’t easily orchestrate a quick attack. Black has time to play ... Nbd5 forcing
White to decide what to do with the knight on f4, which in turn allows Black to play on the h-file. Note
that the kingside square that is most vulnerable is the pawn on h2.
13 a4
If 13 c4 we are quite happy to counter and break out with 13 ... c5, but White might centralize with 13
Rd1 Nbd5! 14 Nxg6 (14 Nxd5 exd5! 15 Bg5 h5 16 h4 Bxg3 17 Qxg3 Qxg3 18 fxg3 Rd6 is slightly better
for Black; if allowed he will play ... Ne4 with a beautiful outpost for the knight, as well as the e-file for
the rooks) 14 ... hxg6 and now:

a) After 15 Bg5 Bf4 Black will have a strong initiative down the h-file.
b) 15 h3?! was played by transposition in P.Popovic-L.Nisipeanu, French League 2000, in which
Black had a superb position after 15 ... Nf4! 16 Ne4 Nxe4 17 Qxe4 g5 18 Bxf4 Bxf4 19 Rd3 Rh4.
c) Chasing the knight away is possible and likely best, but the knight doesn’t mind going to f5. Indeed,
after 15 c4! Ne7 16 c5 Bxg3 17 fxg3 Ned5 18 Bg5 Rh5 19 h4 Rd7 Black should be fine.
13 ... Nbd5!
 14 Nxg6
14 Nxd5 cxd5 15 Bg5 h6 16 Bxf6 gxf6 17 Qxf6 is a rather risky pawn grab, since 17 ... h5 leaves
Black with a dangerous looking attack.
14 ... hxg6 15 a5 a6 16 Bxd5 exd5 17 Bg5 Rh7 18 Bxf6 gxf6 19 Qxf6 Rdh8 20 h3 Bxg3 21 fxg3
Rxh3!! 22 Rf4 R3h7 23 Re1 Qxa5
Black is better.

B) 6 N1e2 e6

Black gets ready to follow the rule: we must meet Nf4 with ... Bd6. Here White has:

B1) 7 Nf4
B2) 7 h4

B1) 7 Nf4 Bd6!

 We are now ready to meet 8 h4 with 8 ... Qc7, but that isn’t White’s only option:

B11) 8 h4
B12) 8 c3
B13) 8 Nxg6

Note that 8 Bc4 would take play back into Line A2.

B11) 8 h4
Very direct and Black should immediately counter with our favourite idea:
8 ... Qc7! 9 Qf3
Leaving c2 en prise, but Black has a good game in any case:
a) 9 h5 simply goes into an endgame: 9 ... Bxc2! 10 Qg4 Ne7! (Black ignores the threat to g7, which
cannot really be taken) 11 Bc4 (11 Qxg7?! Rg8! 12 Qh6 Rg4 chases the white pieces back and after 13
Nfe2 Na6 Black is ready to castle queenside and threatens to play ... Nb4; after 14 Qd2 Nb4 15 Nc3 Bxg3
16 fxg3 Bf5 he is already winning) 11 ... Nd7 and Black is slightly for choice.
b) 9 Nxg6 hxg6 10 Ne4 Bf4! (Black gets rid of White’s bishop-pair; 10 ... Nf6 also worked out well
for Black: 11 Nxd6+ Qxd6 12 c3 Nbd7 13 Qb3 0-0-0 14 g3 Rde8 15 Be2 e5 16 Be3 Nd5 17 Kd2 and
Black soon won in F.Alcer-A.Braun, Berlin 2014) 11 g3 Bxc1 12 Qxc1 Nd7 13 Qe3 Ngf6 14 0-0-0 Nxe4
15 Qxe4 Nf6 16 Qe3 0-0-0 17 Bg2 Qa5 and the knight was just as good as the bishop on g2 in
A.Delorme-P.Bailet, Sautron 2014. Black will put pressure on the d-line and wait for the action to kick
off.
9 ... Bxc2!
The best thing to do is to grab that pawn.
10 Bd2
After 10 Nfh5 Black should simply check: 10 ... Bb4+ 11 Bd2 Qd6 12 Nxg7+ Kf8 13 a3 (or 13 N7h5
Qxd4 14 Bxb4+ Qxb4+ 15 Qc3 Qxc3+ 16 bxc3) 13 ... Bxd2+ 14 Kxd2 Ba4 with nothing to complain
about.
10 ... Nf6 11 Nfh5 Nxh5 12 Nxh5 f5! 13 Qc3 Be4
 Black is already a touch better.
14 Bc4 Bd5! 15 Bxd5 cxd5 16 Qxc7 Bxc7 17 Nxg7+? Kf7 18 Nh5
Now Black starts targeting the weak d4-pawn.
18 ... Nc6 19 Bc3 Rhg8 20 g3 b5 21 a3 a5
The d-pawn will fall.

B12) 8 c3 Qc7
The most accurate move order.
9 Qf3
There are the usual two other options:
a) 9 Nxg6 hxg6 10 Ne4 Bf4 11 g3 Bxc1 12 Rxc1 Nf6 13 Bg2 Nbd7 14 0-0 Nxe4 15 Bxe4 Nf6 16 Bg2
Qa5 and Black has a very solid position.
b) 9 Ngh5 (it doesn’t matter which knight arrives on h5 - the answer is the same) 9 ... Bxh5 10 Nxh5
g6 11 Ng3 Nd7 (exchanging bishops is also very satisfactory for Black: 11 ... Bf4 12 Ne4 Nd7 13 g3
Bxc1 14 Rxc1 Ngf6 15 Nxf6+ Nxf6 16 Bg2 0-0! 17 0-0 Rfe8 18 Re1 a5 19 Qe2 a4 20 Qf3 Kg7 and Black
has an extremely solid position, albeit one which requires careful manoeuvring before either side can
open the position, B.Firat-A.Sumets, Zadar 2013) 12 Qf3 (12 Bd3 doesn’t place the bishop on a great
square and allows 12 ... Bf4 13 Bxf4 Qxf4 14 0-0 Ngf6; likewise, if 12 Bg5 Bf4) 12 ... e5 (12 ... f5! is
even more accurate as ... e6-e5 will soon follow) 13 Bc4 0-0-0! 14 Qxf7 (self-pinning with 14 Bxf7 isn’t
attractive to the human eye, but despite appearances it is tricky to win the bishop; Black should play 14 ...
exd4 15 cxd4 Bb4+ 16 Kf1 with a complicated position) 14 ... exd4 15 0-0 Ngf6 16 Bd2 (16 cxd4!) 16 ...
Rdf8 17 Qg7 h5 18 f4? Qb6 19 Kh1 h4 20 Ne2 h3 21 Qxg6 Rhg8 22 Bxg8 Rxg8 0-1 was I.Madl-
V.Gunina, Tromso Olympiad 2014.
9 ... Nd7
 10 h4
We meet 10 Nfh5 with 10 ... 0-0-0! as the g7-pawn is not as important as development is. After 11
Nxg7 h6 (preventing a bishop coming to g5) 12 N7h5 (or 12 Be2 Ngf6 13 N7h5 Nxh5 14 Nxh5 e5 and the
danger is not over for White) 12 ... e5 13 Bc4 exd4 14 cxd4 Bb4+ 15 Kf1 Nb6 16 Bb3 Kb8 17 Be3 Ne7
with the white king lumbering around on f1 and a weak pawn on d4, Black has full compensation for the
pawn.
10 ... 0-0-0!?
10 ... Ne7 was my original idea, but perhaps White has an advantage after the tricky 11 Nxg6 hxg6 12
Ne4 e5 13 Bc4 f6 14 Ng5.
11 Nxg6
Memory trigger: the bishop is perfectly safe on c2. Thus if 11 h5 Bc2 when White is in danger of
being hit by the counter-strike ... e6-e5: for example, 12 Bd3 (or 12 Qe2 Bxf4 13 Bxf4 Qxf4 14 Qxc2
Ngf6, while both 12 Bd2 and 12 Be3 run into 12 ... e5!) 12 ... Bxd3 13 Nxd3 Bxg3 14 fxg3 e5 15 dxe5
Nxe5 16 Qf5+ Qd7 17 Qxd7+ Nxd7 18 Be3 Ngf6 19 0-0-0 Rhe8 20 Bd4 b6 21 h6 c5 22 Bxf6 gxf6 and
Black is fine.
11 ... fxg6 12 Ne4 e5! 13 Nxd6+ Qxd6 14 Be3 Ne7 15 0-0-0 Qe6 16 b3 Nd5 17 Bc4 N7b6
 Black has good counterplay and White can no longer hold on to the bishop-pair.

B13) 8 Nxg6
White tries to simplify the position in his favour.
8 ... hxg6 9 Ne4 Nf6

Black is unafraid of playing with two knights versus two bishops. He has plenty of activity (threats
against d4 and h2), so it is not very easy for White to coordinate his pieces in this semi-closed position.
10 Nxf6+
Alternatively:
a) 10 Nxd6+ Qxd6 11 h3 Nbd7 12 c3 0-0-0 13 Be3 (13 Bd2?! is too passive: 13 ... Nb6 14 Be2 e5
and with the white king in the centre, Black correctly opens up lines, leaving White with big problems) 13
... Nd5! 14 Qf3 Nxe3 15 Qxe3 Qd5 16 Be2 e5 and Black had the more comfortable development in
T.Fogarasi-R.Zelcic, Austrian League 2005.
b) 10 Ng5 Nbd7 11 g3 Qa5+ 12 c3 Rd8 13 Bg2 e5 14 0-0 0-0 15 Re1 exd4 16 Qxd4 Bc5 17 Qf4 Rfe8
18 Bd2 Rxe1+ 19 Bxe1 and the two white bishops are not working terribly well together.
10 ... gxf6 11 c3!
11 g3? is not possible because of 11 ... Qa5+ 12 c3 Qd5! and White loses material, while 11 Be3 is
also dubious: 11 ... Qb6! 12 Qc1 Nd7 13 Be2 f5 and Black has the initiative.
11 ... Nd7 12 g3 f5 13 Bg2 Qf6!
 Preparing ... f5-f4.
14 Qb3
Forcing Black to castle queenside, but the queen will get in the way of a future pawn advance.
Instead, 14 0-0 is really provocative. Black can’t castle queenside comfortably, but we can simply
charge: 14 ... g5! 15 b4 Qg7!? (I’m less keen on 15 ... g4 16 b5! and 15 ... Qh6 16 h3 f4 17 g4 might be a
tough nut to crack, but 15 ... a6!? is interesting) 16 b5 (after 16 a4 a6 now ... g5-g4 is in the air) 16 ...
cxb5 17 Bxb7 Qh7! and the h-file will have to open.
14 ... 0-0-0 15 Bd2
After 15 f4 Qg7 16 0-0 Qh7 17 h3 Nf6 18 c4 Ne4 the initiative is with Black, as it is after 15 0-0 g5!.
15 ... e5 16 0-0-0 f4 17 Kb1 exd4 18 cxd4 Nb6
Black has few if any problems.

B2) 7 h4 h6 8 Nf4

Instead, 8 Bc4 would take us back into Line A1.

8 ... Bh7 9 Bd3?!
This move of Volokitin’s is connected with a sharp, but not completely correct sacrifice.
 We would be back in the main line of A1 after 9 Bc4 Bd6, while 9 c3 is a less aggressive set-up.
Here 9 ... Nf6 10 Bd3 Bd6 11 Ngh5 Bxf4 12 Nxf4 Bxd3 13 Nxd3 Qd5 is a typical way to activate and
after 14 f3 Nbd7 15 Bf4 0-0 Black was fine in V.Kupreichik-V.Korchnoi, Val Maubuee 1990.
9 ... Qxd4 10 Be3 Qd6 11 Nxe6!?
White opens up the d- and e-files to attack the black king. Instead, 11 Bxh7 Rxh7 is already slightly
better for Black.
11 ... Qxe6!
Correct. After 11 ... Bxd3?! 12 Nxf8 Qb4+ 13 Qd2 Qxd2+ (or 13 ... Qxb2 14 Rd1 Qxc2 15 Qxd3
Qxd3 16 Rxd3 Kxf8 17 Bc5+ Ke8 18 Nf5 Nd7 19 Nxg7+ Kd8 20 Bd4 with compensation) 14 Kxd2 Bxc2
15 Rhe1 is slightly better for White.
12 Bxh7 Rxh7

13 Qd3!
The rook might be on h7, but after 13 0-0 Nd7 14 Re1 0-0-0 15 Bxa7 Qf6 Black is much better.
13 ... g6!
Improving over the stem game, which saw 13 ... Nf6?! (Black is forced to give up the exchange after
this move) 14 0-0-0 Nbd7 15 Bd4 0-0-0 16 Bxf6! Qxf6 17 Qxh7 Ba3 (17 ... g6 was later recommended as
a better approach) 18 c3 g6 19 Kb1 Rh8 20 Ne4 Qe5 21 Qxf7 Qxe4+ 22 Ka1 Qe7 23 Qxg6 Bc5 24 Rhe1
Qf6 25 Qg4 Qf7 26 Re5 Bb6 27 h5 Kc7 28 Rf5 Qe6 29 Qg6 Qe2 30 Qd6+ and 1-0 in A.Volokitin-
I.Khenkin, German League 2003.
14 0-0-0 Be7 15 h5
I’ve also analysed:
a) 15 Rhe1 Qxa2! 16 c3 Qa1+ 17 Kc2 Qa6 18 Qd4 (if 18 Qd2 Qa4+ 19 Kb1 Na6 20 Qd7+ Kf8 or 18
c4 Qa5, covering the d8-square) 18 ... c5 (but not 18 ... Nf6? 19 Bg5) 19 Qd5 (19 Qxc5 Qa4+ 20 Kc1
Qxd1+! 21 Rxd1 Bxc5 22 Bxc5 Nf6 wins) 19 ... Qa4+ 20 Kb1 Nc6 21 Qd7+ Kf8 22 Qxb7 Rb8 and Black
is winning.
b) 15 Bc5 Qd5! 16 Qe2 Qxc5 17 Ne4 Qe5 and White has nothing.
15 ... Nd7 16 hxg6 fxg6
 17 Kb1
Likewise, after 17 Rhe1 0-0-0 18 Bb6 Qxa2 19 Bxd8 Bxd8 Black keeps the advantage.
17 ... 0-0-0 18 Bxa7 Nb6 19 Qc3 Nc4
Black is simply better.

C) 6 Nf3

Finally, it’s time for something a bit different. We should take a look at what happens should White opt
for a ‘normal’ game with no fancy manoeuvres. In all honesty, these set-ups are often played by lower-
rated white players looking to draw and in reality they promise White absolutely nothing.
Instead, 6 Nh3 e6 7 Nf4 would transpose to Line B1.
6 ... Nd7!
Preventing the white knight from going to e5.
7 Bd3
White sets his hopes of getting an advantage on the greater freedom of his pieces, but he will lack
sufficient coordination and strength to breach Black’s position.
Instead, 7 h4 h6 transposes to the main line, as seen in our first two chapters, while after 7 Bc4 Black
can safely reply with natural moves: 7 ... e6 8 c3 Bd6 9 Qe2 Qc7 10 0-0 Ngf6 11 Re1 0-0 12 Ne5. We’ve
followed M.Haubro-A.Rudolf, Copenhagen 2012, and now Black should play 12 ... Bxe5! 13 dxe5 Nd5
14 a4 a5 when she will start occupying the d-line.

7 ... Ngf6
It is clear that the bishop on g6 will have to be exchanged, the only question is where this will take
place. Whilst I have to admit that play is relatively quiet, the good thing is that Black’s ideas are
relatively straightforward and logical.
Another approach is to play 7 ... Bxd3 8 Qxd3 Ngf6. The following game impressed me so much I
will give it in its entirety: 9 0-0 e6 10 Re1 Be7 11 c4 0-0 12 Bf4 Re8 13 Rad1 a5 (an alternative idea
now that the white rook has moved from a1) 14 Ne2 Bb4 15 Rf1 a4 16 Qc2 a3 17 b3 h6 18 Nc1 Qa5 19
Nd3 Bf8 20 h3 Qf5 21 Rfe1 c5 22 dxc5 Bxc5 23 Qc1 Bb6 24 Nde5 Red8 25 Nxd7 Rxd7 26 Rxd7 Nxd7
27 b4 Nf6 28 Re5 Qd3 29 Rb5 Rc8 30 Rxb6 Rxc4 31 Qd2 Qb1+ 32 Kh2 Nd5 33 Rxb7 Rxf4 34 Rb8+
Kh7 35 b5 Qb2 36 Qd3+ f5 37 Ra8 Qxf2 38 Rxa3 Re4 39 Qd1 Nf4 40 Qg1 Qb2 41 Rb3 Qxa2 42 Rb1
Re2 43 Nh4 Qa3 0-1, M.Godena-E.Danielian, Cento 2011. Enjoyable stuff!
8 0-0
8 h4 is illogical as the bishop is in no longer in any danger on g6: 8 ... e6 9 c3?! (White is now guilty
of playing too quietly) 9 ... Bd6 10 Ne2 Qc7 11 Qc2 0-0-0 (Black is planning ... e6-e5, opening lines
against White’s king) 12 Bd2 Bxd3! 13 Qxd3 e5!? 14 dxe5? (White makes a big error and opens up the
centre for Black; better was to keep the position closed with 14 Ng5 Nb6 15 0-0-0 Kb8, although Black
has a pleasant position) 14 ... Bxe5 15 Nxe5? Nxe5 16 Qf5+ (White was counting on this move, but ... )
16 ... Qd7! 17 Qxd7+ Rxd7 18 Nf4? Neg4! 19 Rd1 Re8+ 20 Be3 Nxe3 21 fxe3 Rxe3+ 22 Ne2 Rde7 23
Rd2 Ne4 and Black was winning in S.Khoo-A.Hagesaether, London (rapid) 2006.
Should White swap bishops with 8 Bxg6 in order to castle queenside, there is again nothing to fear: 8
... hxg6 9 Bg5 e6 10 Qd2 Be7 11 0-0-0 a5 12 Kb1 a4 13 h4 a3 14 b3 Bd6 and Black stands well as
White’s attack is going nowhere.
8 ... e6 9 Re1 Be7 10 c4
Preventing any plan with ... Nd5. We welcome the small imbalance which occurs after 10 Bf4 Nd5 11
Bxg6 hxg6.
10 ... 0-0 11 Bxg6 hxg6 12 Bf4
12 Bd2?! with the idea of placing the bishop on c3 is very passive: 12 ... b5! (an excellent move;
Black secures the d5-square for his pieces) 13 Rc1 bxc4 14 Rxc4 c5 15 Ne5? (White should really have
exchanged on c5) 15 ... Rc8! 16 Bg5 (now after 16 dxc5 Bxc5 17 Nxd7 Qxd7 Black is slightly better due
to his better piece placement) 16 ... Nb6 17 Rc1 c4! 18 Re2 Qd5 19 Rec2 Rfd8 was J.Vrchotka-
E.Dizdarevic, Pula 2005. Black has really improved his position: he has more space, the pawn on c4 is a
giant clamp, and the pawn on d4 is an easy target.
12 ... b5!?

A useful way to try to unbalance the position.

13 b3
After 13 Rc1 bxc4 14 Rxc4 Rc8 15 Qc2 Nd5 Black gets the use of the d5-square and ... c6-c5 is
coming.
13 ... Rc8 14 Qe2 Qa5 15 Rac1 bxc4 16 bxc4 Bb4 17 Red1 c5 18 d5 exd5 19 cxd5 Rfe8 20 Qb2
Nb6 21 d6 Ba3 22 Qc3 Qa4
White’s knights are doing nothing and Black is better.

Summary
The plan of Nf4 is versatile, but so long as we twin the move Nf4 with ... Bd6 then we should be immune
from the dangerous h-pawn thrust. By going for the full works with Bc4, h4 and Nf4 White can try to
sacrifice on e6 or attack with Nh5, but Black can calmly handle the complications. White is really
dancing on hot coals with these early attacks as he will either be subjected to a deadly counterattack in the
centre or his h4-pawn will become a sitting target.
By going Bc4 and Nf4, but delaying h4, White keeps his options open: he can swap on g6 and/or
castle kingside, although neither approach is too dangerous. White might first defend c2 and then go h2-
h4, but even here we are fine and should secure the d5-square for a knight.
Instead, 7 Nf4 and then 8 h4 is simply a trap which Black can easily repel by counterattacking against the
vulnerable white knight on f4. Indeed, if White wishes to play Nf4 along with h4 then he must delay h4
and play c2-c3 first, but Black is fine after ... Qc7 preparing queenside castling. As always Nh5 is met
with ... Bxh5 and then a classical game arises with Black’s plan being to eliminate the bishop-pair.
Chapter Five
The Panov-Botvinnik Attack: Introduction and 6 Nf3
1 e4 c6 2 d4 d5 3 exd5 cxd5 4 c4

Activity versus Structure

The Panov-Botvinnik Attack is one of the most popular and sharpest lines in the Caro-Kann. Right
from the beginning, White plays for a direct attack against the black king, with a healthy disregard for
positional elements. With his fourth move, White places Black under pressure by challenging the d5-
pawn, and this is indicative of how he will continue. He is often prepared to accept long-term structural
weaknesses in return for free and easy development, active piece play and open files, which can all
combine to give Black a tough time in the middlegame.
The aim of this book is to steer you safely through all the potential problems Black may face, so that
we can exploit our long-term positional advantage. White’s aims are pretty direct, while Black, on the
other hand, usually adopts a temporary defensive attitude: think in terms of ‘you punch and I block’. This
is not meant to sound negative, for when Black successfully beats back the initiative, at the risk of
sounding like a preacher, he will reap the rewards until kingdom comes.
The ruined pawn structure that opened files of attack for White in the middlegame is simply weak and
ineffectual in the latter stages of the game. It’s exactly because of this high risk/high gain strategy that
White will try to avoid exchanging pieces, which will alleviate any space issues Black may have, or take
the position into an endgame. White’s goal is simply to mate or win material, nothing less.
4 ... Nf6 5 Nc3 Nc6
 Right from the fourth move, we can see that White has conducted himself aggressively by placing
pressure on the d5-pawn. However, rather than supporting the pawn by playing 5 ... e6, with my choice 5
... Nc6 Black first concerns himself with developing his pieces. This is ambitious, but also quite a risky
approach since the centre is unsupported by pawns here.
White has two main approaches:
a) To complete development with 6 Nf3 (the Four Knights Variation).
b) To place pressure on the d5-pawn with 6 Bg5.
In this chapter and the next we will look at 6 Nf3, while Chapter Seven is devoted to 6 Bg5. There are
a whole host of pawn structures that arise in each variation; some are interlinked, others completely
independent.

A New Approach
In my original work I recommended the following line:
6 Nf3 Bg4 7 cxd5 Nxd5 8 Qb3 Bxf3 9 gxf3 Nb6
However, after ...
10 d5!
 ... I was unable to solve all the problems which White has managed to create in the past seven years,
so it was back to the drawing board.
My next idea was to recommend the Panov Endgame after 9 ... e6. However, I soon realized I couldn’t
add anything new to existing theory, although it is a perfectly acceptable way to play as Black. Then I
spotted a line that has suddenly swung back into fashion and I was quickly hooked.

6 Nf3 g6

6 ... g6 is an ambitious move and one that actually can arise too from the Symmetrical English via, for
instance, 1 Nf3 c5 2 c4 Nc6 3 Nc3 g6 4 e3 Nf6 5 d4 cxd4 6 exd4 d5. Black does not support the centre
with a pawn and instead elects to develop the dark-squared bishop to the perfect long diagonal. Indeed, if
White does nothing we will simply have a perfect anti-Isolated Queen’s Pawn position.
This means that White from the outset must act rapidly and there are three typical ways he can
respond:
a) Pressure the centre, either through the exchange of pawns on d5 or by attacking the defender of d5,
our knight on f6.
b) Determine the pawn structure and obtain a queenside pawn majority where funnily enough the long-
term prospects are with White.
c) Chose to develop and try to make any important structural decisions later on.

6 Nf3
Should White try to avoid mainstream lines with 6 cxd5 Nxd5 after 7 Bb5 Black can go 7 ... Nxc3 (or
7 ... e6 8 Nf3, transposing to a 2 c4 variation and one we’ll consider in Chapter Thirteen) 8 bxc3 (after 8
Bxc6+ bxc6 9 bxc3 Qd5 10 Qf3 e6 11 Qg3 Ba6 Black was already doing well in H.Gold-N.Croad,
Dunedin 2008) 8 ... Qd5 9 Be2 Qxg2 10 Bf3 Qg6 11 Ne2 e5 12 d5 Nd8 13 Rg1 Qa6 14 Qb3 Bd7 15 Be3
Rc8 and White didn’t have enough for his pawn in L.Shamkovich-U.Adianto, New York 1994.
6 ... g6 7 cxd5
A standard move in the Panov and one that most white players will play if unsure of what to do.
Very important alternatives are 7 Bg5 and 7 Be2, as we’ll see in Chapter Six.
7 ... Nxd5
 Now White must challenge our finely placed knight on d5:

A) 8 Qb3
B) 8 Bc4

A) 8 Qb3
The critical choice.
8 ... Nxc3

Now recapturing may look obvious, but there’s actually a better move for White. We have:

A1) 9 bxc3
A2) 9 Bc4!

Instead, 9 Qxc3 looks daft as Black can play 9 ... Bg7 when the d-pawn is in trouble already.

A1) 9 bxc3
 9 Bc4 is the main move, but White may not find/play it if he is acting on autopilot.
9 ... Bg7
Black’s idea here is simple and will be based on Nimzowitsch’s method of playing against weak
pawns: control, restrain then destroy. This means that after developing Black must set up a light-square
blockade primarily on the open c-line.
10 Be2
Alternatively:
a) 10 Bc4 0-0 11 Bxf7+?! Rxf7 12 Ng5 e6 13 Nxf7 Kxf7 14 0-0 Na5 15 Qb4 b6 and Black is already
doing extremely well.
b) 10 Ng5 is also dealt with rather easily: 10 ... e6! 11 Ne4 0-0 12 Ba3 (F.Beltz-J.Gilbert, London
2000) 12 ... Nxd4! 13 cxd4 Qxd4 14 Rd1 Qxe4+ with a beautiful position for Black: for example, 15 Be2
Qxg2 16 Bf3 Qg5 17 Bxf8 Bxf8 and White has too many things to worry about.
c) 10 Bd3 doesn’t cover the knight on f3, a factor Black may exploit in a few more moves: 10 ... 0-0
11 0-0 Be6 12 Qxb7 Bd5 13 Qb5 e5 14 dxe5 (or 14 Nxe5 Nxe5 15 dxe5 a6 16 Qa4 Bxe5 and Black is
well on the way to regaining the pawn) 14 ... a6 15 Qa4 Bxf3! with the advantage.
10 ... 0-0 11 0-0
Black has no problems after 11 Bf4 e5! 12 dxe5 Qa5 13 0-0 Nxe5 14 Nd4 b6 15 Bxe5 Bxe5 16 Nc6
(16 Bf3 Bxd4 17 cxd4 Be6 18 d5 Bd7 is also equal) 16 ... Qxc3.
11 ... Qc7
 As John Emms has mentioned, this is a Grünfeld position and one Black should welcome as he can
exploit the position of White’s queen to gain time for his queenside counterplay.
12 Rd1 Na5 13 Qa3 Be6 14 d5
Or 14 Be3 Nc4 with an easy game for Black.
14 ... Bg4
Black has the straightforward plan of attacking the c- and d-pawns.

A2) 9 Bc4!

9 ... Nd5
This is my recommendation. Black can also play 9 ... e6 10 bxc3 Bg7, but White has the awkward
reply 11 Ba3.
10 Bxd5 e6 11 Bxc6+
Abandoning the bishop-pair, but if 11 Bc4 Bg7 12 d5 (or 12 0-0 0-0 and White does not have a
satisfactory answer to the threat of ... Na5 and ... Nxd5) 12 ... exd5 13 Bxd5 Qe7+ 14 Be3 Qb4+! (after
this move the initiative passes to Black on account of the superior placement of his bishops) 15 Ke2 Qxb3
16 axb3 (or 16 Bxb3 Bf5 when Black has the better-placed bishops and threats against the b2-pawn) 16 ...
0-0 17 Bxc6 bxc6 18 Rxa7 Rxa7 19 Bxa7 (A.Torrecillas Martinez-X.Vila Gazquez, Barcelona 2011) 19
... Re8+ 20 Kd2 Rd8+ 21 Kc1 Be6 and Black will regain his material with the superior chances.
11 ... bxc6 12 0-0 Qd5!

We have arrived at the tabiya position for the 7 cxd5 variation and this chapter. Black has the bishop-
pair in a relatively open position, which can translate to a long-term advantage. Structurally speaking our
weak c6-pawn is offset by White’s weak d4-pawn, but Black’s main problem is that he is lagging in
development so may soon come under pressure from White. This may mean that he is not able to do
normal things like castling kingside. Indeed, we may have to castle by hand.
The first step for Black was to centralize the queen, with two ambitions. Firstly, we want White to
swap queens - remember that long-term advantage? Secondly, the queen is an extremely powerful
aggressor and defender on d5, even allowing the black kingside pawns to advance if need be.
After the powerful move 12 ... Qd5 White must decide whether or not to move his queen:

A21) 13 Qxd5
A22) 13 Qc2?!
A23) 13 Qa4
A24) 13 Qc3
A25) 13 Bg5
A26) 13 Bf4

Instead, 13 Qe3 looks clumsy and is well met by 13 ... f6 (13 ... Bg7 is also possible, but White can
repeat with 14 Qa3 Bf8 15 Qe3).

A21) 13 Qxd5 cxd5

 14 Bf4
What can we say about this endgame? Well, at first sight it appears that White has a very promising
position since he will gain control over the c-file. First of all, we have to acknowledge that the c-file has
to be neutralized effectively and quickly. Here are some guidelines:
a) Take control over the e5-square and do not allow a minor piece to aid a white rook on the seventh
rank, which can be lethal for us.
b) Gain space on the kingside.
c) Chase away any white rook on the seventh rank.
d) Utilize horizontal motifs - don’t forget the rook on h8 can challenge a c7-rook by mobilizing itself
via h7.
Overall, we should actually embrace this endgame: Black can cover the key squares and his bishops
will gradually come into their own.
14 Bg5 could also be played, but it doesn’t make so much sense because the move has no relevance to
the open c-file: 14 ... Be7! (of course, we must not let the bishop entrench itself on f6) 15 Bxe7 Kxe7 16
Rfc1 f6! (controlling the e5-square) 17 Rc7+ Bd7 and Black will soon play ... Kd6 followed by ... Rhc8
with a good endgame.
14 ... f6!
Taking control over the important e5-square.
15 Rfc1 h5!
15 ... g5!? is also another good possibility: 16 Bg3 h5 17 h4 gxh4!? (a little unnatural, but it is
important to open lines and not get bogged down by structural issues; the position is dynamic and we must
play accordingly) 18 Bxh4 Be7 19 Rc3 (19 Rc7 Bd8 chased away the rook in R.Ruck-V.Bologan, Warsaw
(blitz) 2011) 19 ... Rg8 20 Rac1 Bd7 21 Rc7 Bd8 22 Rb7 Rg7 23 Bg3 Rf7 24 Nd2 Bb6 25 Nb3 Bb5 26
Rxf7 Kxf7 and Black had no problems in F.El Debs-V.Bologan, Gibraltar 2012.
16 h4
After 16 Rc7 g5 17 Be3 Bd6 White will be kicked back.
16 ... Rh7!
 Model play and horizontal defence along the rank.
17 Ne1 Rb7!? 18 Nd3 Bd7 19 Rc2 a5 20 Rac1 a4 21 Rc7
21 Nc5 looks good, but Black can simply play 21 ... Bxc5 22 dxc5 Bc6! when it will be Black
advancing those pawns.
21 ... Raa7! 22 Rxb7 Rxb7 23 Rc7?! Rxc7 24 Bxc7 Bb5 25 Nf4 Kf7 26 a3
Only Black can play for a win here and, indeed, he eventually won in J.Kraai-G.Kacheishvili, US
Internet League 2012.

A22) 13 Qc2?!
Playing against the weakness on c6, but this is one of the weakest queen moves. White abandons any
potential counterplay on the kingside and solely aims to prevent the liberating move ... c6-c5. Naturally,
when confronted with such single mindedness, we too should stubbornly pursue our goal. Resistance
should be met with determination.
Please make a careful note over how carefully Black now prepares his dream, the ... c6-c5 break.
13 ... Bg7 14 Bf4
If White is even more single-minded and plays 14 Be3 0-0 15 Rfc1 Bb7 16 Qe2 (16 a3 Rfd8 17 b4
a5! weakens the grip) 16 ... Qb5! 17 Qe1 Rfd8 18 Rc5 Qb6 19 Qc3 Rd5 20 Rc1 Rad8! he finds himself
heading nowhere.
14 ... 0-0 15 Be5
 Forcing Black to block in his own bishop, but he is quite happy to do so.
15 ... f6!?
Black must keep the bishop-pair on the board, as this is the basis for our long-term advantage.
16 Bg3 Bb7 17 Rfe1
If 17 Qc5, as pointed out by Ribli, Black can shut out the white bishop on g3 with 17 ... Qxc5 18 dxc5
e5, which is slightly better for Black.
17 ... Rfe8 18 Re3 Rac8 19 Rc3 Bf8!
Black must prepare the active ... c6-c5 - the bishops need to come alive.
20 a3
And not 20 Rc1? Qxa2.
20 ... c5

The dream has been achieved and there is no way that White can punish Black for it. From now on
Black will have a slight plus playing with the bishop-pair and a mobile pawn majority.
21 dxc5 Rxc5 22 Rxc5 Qxc5 23 Qxc5 Bxc5
Black went on to grind out his edge in S.Giddins-J.Emms, Port Erin 1999.
A23) 13 Qa4

A slightly better option than 13 Qc2, but Black can doggedly pursue the white queen.
13 ... Qb5
The safest approach.
Preventing Ne5 with 13 ... f6!? is more risky, but should be safe enough on account of Black’s
centrally placed queen and the lack of minor pieces on the board. After 14 Bf4 Kf7 15 Rac1 (15 Rfe1 Be7
16 Rac1 Bd7 17 h4 is designed to prevent the space-gaining ... g5, but after 17 ... Rhc8 I don’t think Black
has any problems) 15 ... g5 (15 ... Be7 would avoid any sacrifice on g5) 16 Bxg5!? (a complicated
sacrifice; 16 Bg3 Bb7 is a little passive) 16 ... fxg5 17 Ne5+ Kg8 18 f4 we reach a very complicated
position which may not be to everyone’s liking.
14 Qc2
White’s reverts to the conventional plan of preventing the freeing pawn break ... c6-c5. The difference
between this position and that resulting from the dubious 13 Qc2 is that the black queen is on b5,
something Black needs to make use of.
14 ... Bg7 15 Bf4 0-0 16 Rac1
Alternatively, 16 Rfc1 Bb7! 17 Ne5 (17 a4!? Qd5 18 Qe2 Rfd8 19 Rc5 Qb3 20 Be5 Rd5 21 Bxg7
Kxg7 22 Rxd5 Qxd5 23 Rc1 is perhaps a slight plus for White, although Black should hold) 17 ... Rfd8
(counterattacking against d4) 18 a4 Qd5 19 Nf3 is equal.
16 ... Ba6!?
 Developing the bishop with tempo as it is not possible for White to capture the c6-pawn for the time
being.
17 Rfd1 Rac8 18 Ne5 Rfd8! 19 Nxc6
White may have won the weak c6-pawn, but he is now in a complete bind and Black will play to win
the d4-pawn.
19 ... Bf8! 20 a4 Qb6
White will have to give back the material to extricate himself from this unpleasant pin, J.Nunn-
R.Dzindzichashvili, Hastings 1977/78.

A24) 13 Qc3!

This position was long commonly held to be slightly better for White, but in modern day chess one
needs concrete reasons before reaching assessments.
13 ... f6!?
This move was suggested as an interesting option by John Emms and I like it. Unlike the 13 Qc2
variation where Black had easy play by fianchettoing his bishop and castling quickly, in this variation in
order to win Black must keep his bishop on the a3-f8 diagonal.
 Like in the previous lines, Black if given the opportunity will break with ..c6-c5. However, a word of
warning: this should not happen until he is well prepared for the opening of the c- and d-lines. It is for this
reason that the bishop is needed on e7, providing some horizontal cover for our king.
Note too that it may not be in our best interests to cling possessively on to the weak c6-pawn. We may
feel sorry for it at times, but there are situations when the pawn must be abandoned to its own fate and so
allow the light-squared bishop a life of its own. Usually in this situation Black should be
counterattacking or have pressure against the weak d4-pawn (or even pressure on the b-line). The c6-
pawn will only be given away in exchange for activity. With the presence of a weak white d-pawn and
plenty of open lines and diagonals, material does not matter so much.
Finally, should everything come to a deadlock - for instance, if White has prevented the ... c6-c5 pawn
break or piled lots of pressure on the e-line - Black can seek to make use of the king on f7 by advancing
on the kingside with ... g6-g5.
13 ... Be7 is the other move with more or less the same aims:

a) This allows 14 Bh6, but Black can connect the rooks by playing 14 ... f6! and 15 Qe3 (if 15 Rfe1
Kf7 16 h4 Bd7 17 a3 Rhb8 and as John Emms wrote in Dangerous Weapons: The Caro-Kann, Black is
well coordinated) 15 ... Bd7 16 h4!? Rb8 17 Qf4 Rc8 18 Rac1 (or 18 Rfe1 Qf5 19 Qe3 Kf7) 18 ... Qf5
19 Qe3 Kf7 20 Rc4 c5 equalized in I.Khamrakulov-A.Fier, Navalmoral de la Mata 2011.
b) 14 Bf4 f6 (controlling the e5-square) 15 Rac1 Bb7 16 Rfe1 g5! seizes space and Black doesn’t
have any problems at all.
14 Re1
Forcing the black monarch to shuffle out of the firing line. There is no need to worry about the king
situated on an unconventional square, though: White does not yet have the firepower to cause any trouble.
14 Bf4 is likely to transpose.
14 ... Kf7 15 Bf4 Be7
 Please memorize this crucial position. Defensively Black’s main priority is to still to cover the e5-
and c5-squares. All three plans discussed earlier are still very relevant.
16 Re2
We should consider too:
a) 16 Bc7 makes no play for the centre, so Black should go 16 ... Bb7 17 Ba5?! c5! 18 Rad1 Rhe8 19
dxc5 Qxc5 20 Qd2 e5 and the unopposed bishop is coming into its own.
b) 16 Rac1 Rd8! ignores the c6-pawn and sees Black begin his plan of liberating the c8-bishop.
Following 17 Qxc6 Qxc6 18 Rxc6 g5! 19 Bc7 (or 19 Be3 Bb7 20 Rc7 Bd5 21 a3 a5 and Black’s
excellent piece coordination makes this an incredibly hard position for White to play) 19 ... Rd5 20 h3 h5
21 Rd1 Bd7 22 Rcc1 Rc8 Black will soon regain the pawn and with it the advantage.
c) 16 h4 Rd8 17 Rac1 a5 18 Qxc6 Bb7 19 Qxd5 Bxd5 20 b3 Bxf3 21 gxf3 Rxd4 and Black stands
fine.
d) 16 Re3 forces Black to undertake some defensive measures over e6 - unlike the pawn on c6, this is
not one we can let go of. After 16 ... Re8 17 Rae1 Bd7 18 a3 Kg7 suddenly ... c6-c5 is in the air and if 19
b4? Black can simply play 19 ... a5! with a promising position.
16 ... a5 17 Rc1
17 a3 Ba6 18 Rc2 Rhc8 19 Rac1 g5 20 Be3 Rab8 is just the kind of chess we should play. After 21 h3
h6 22 Nd2 Bd6 White is at a loss over what to do: for instance, 23 f3 preparing Ne4 is met by 23 ... f5!.
17 ... Ba6 18 Rec2 Rhc8 19 Qe3 a4 20 Nd2 g5!
 White has Black in a bind with regards to the ... c6-c5 push, so Black carries out his alternate plan:
seizing space on the kingside.
21 Bg3 Rd8 22 Ne4
22 Rxc6? loses to 22 ... Bb7.
22 ... Qxd4 23 Qf3
We’ve followed M.Neubauer-J.Bernasek, Graz 2013, and now Black can win material:
23 ... g4! 24 Qxg4 Bd3 25 Qh5+ Kg8 26 Re2 Qd5 27 Qg4+ Kh8 28 Ree1 f5 29 Qh5 f4!

A neat double attack. 29 ... Bxe4 30 Red1 Qb5 31 Qf7 Re8 32 f3 f4 33 Bf2 Bd3 is trickier.
30 Qf7 Bxe4 31 Qxe7 fxg3 32 hxg3 Rf8 33 Rcd1 Qf5
The extra piece should soon begin to tell.

White Moves the Bishop

 What if White is happy to allow the exchange of queens and simply develops? Here I have formulated
a rule to help Black:
If White allows us to exchange queens on move 13, we should only immediately exchange queens if
he has played 13 Bg5.
Black then counts on the three weak queenside pawns to slow down White’s activity. In order to do
this, the first emphasis for Black is to defend solidly and consolidate his position.
The reason for the rule is that 13 Bf4 carries no dynamic threat so Black can delay the queen exchange
and instead play for ... f7-f6 and ... g6-g5 quickly. He seeks to do this as quickly as possible to ensure that
White does not have an effective piece sacrifice on g5.
Instead, 13 Be3?! is not the most challenging move as the bishop has a purely defensive role on e3: 13
... Be7 14 Rfc1 f6 15 Nd2 0-0 16 Qxd5 (or 16 Qc3 Bd7 17 Bf4 Rfd8 18 Nb3 Be8 19 h3 Rac8 20 Re1 g5
21 Bg3 Bg6 22 Rac1 Kf7 23 Re3 h5 and Black already had the advantage in P.Vishnu-H.Melkumyan,
Albena 2013) 16 ... cxd5 17 Rc7 (likewise, if 17 Bf4 a5 18 Rc7 Bd8!) 17 ... Bd6 and the rook is driven
off the seventh rank.

A25) 13 Bg5
 White is threatening a middlegame initiative with 14 Bf6. This means that it favours Black to swap the
queens in order to defuse the potential attack.
13 ... Qxb3!
Instead, 13 ... Bg7 allows White a comfortable game: 14 Rfc1 f6 15 Be3 Bf8 16 Nd2 Be7 and Black
had to waste time manoeuvring the bishop to e7 in D.Jakovenko-V.Bologan, Poikovsky 2013.
14 axb3
Black now has two weaknesses to be concerned about, but then again White has three pawn islands in
return for his greater piece activity - it’s a fair balance.
14 ... Bg7
14 ... Be7 15 Bxe7 Kxe7 16 Ne5 is too passive for Black. Remember the situation is different from
the original endgame after 13 Qxd5 cxd5 because White has not wasted time and has simply developed
his pieces.
15 Be3
Black should not panic if White plays 15 Rfc1. He should simply play 15 ... Bb7 as he cannot let go of
the c6-pawn just yet and instead must defend solidly:

a) 16 Rc4 f6! (controlling the e5-square) 17 Bf4 Kf7 (the king hides itself safely away on f7 where it
is not in any danger from forks or bishop checks) 18 Rac1 Rhd8! 19 Nd2 a5 20 Ne4 Bf8 21 Be3 (with 21
Nc5 the knight gets to its destination, but this is actually a little meaningless after 21 ... Bxc5 22 dxc5 e5!
when the initiative passes to Black with his extra pawn on the kingside) 21 ... Rd5 22 Nc5 Bxc5 23 Rxc5
Rxc5 24 Rxc5 a4! 25 bxa4 Rxa4 and Black is fine.
b) 16 Ne5? (forcing matters too early) 16 ... f6! 17 Nc4 fxg5 18 Nd6+ Ke7 19 Nxb7 Bxd4 20 Rxc6
Rhc8 and Black has the edge on account of the combination of rook and bishop.
15 ... 0-0 16 Rfc1
Luckily for us, 16 Ne5 c5! releases the bishop.
16 ... Rd8
 17 Ne5
Likewise, 17 Rxc6 Bb7 18 Rc4 Bd5 19 Rca4 Rab8 20 Nd2 Bxb3 21 Nxb3 Rxb3 22 Rxa7 Rxb2 leads
to nothing for White.
17 ... Bd7! 18 h3
After 18 Nxd7 Rxd7 19 Rxc6 (or 19 Ra4 Rad8 20 g3 Bxd4 21 Rxd4 Rxd4 22 Bxd4 Rxd4 23 Rxc6
Rd2 which is equal) 19 ... Bxd4 20 Rd1 Rad8 21 Bxd4 Rxd4 22 Rxd4 Rxd4 23 g3 Rd2 and it’s a draw,
while 18 Ra6 looks menacing, but Black can calmly retreat with 18 ... Be8 19 Rca1 c5.
18 ... Be8 19 Nxc6 Bxc6 20 Rxc6 Bxd4 21 Bxd4 Rxd4 22 Rc7 a6 23 Ra4 Rd1+ 24 Kh2 Rd2 25
Rf4 Rf8 26 Ra7 e5 27 Rf3 e4 28 Rf6 Rxb2 29 Rfxa6 Rxf2
Black was slightly for choice in P.H.Nielsen-J.Emms, Esbjerg 2000.

A26) 13 Bf4

It’s best not to capture the queen on b3 immediately as Black will have to deal with two pressing
issues: the opening of the a-line and guarding the e5-square. The standard strategy in this position is to
develop the bishop to e7 (it is needed on this diagonal in order to keep an eye on the c5-square), play ...
f6, and hide the black monarch on f7 (if we are not allowed to castle).
 13 ... f6!?
The most interesting try, defending the e5-square and preparing the immediate advance ... g6-g5.
Instead, 13 ... Qxb3 14 axb3 f6 (Black must take time out to defend against an invasion on e5; 14 ...
Bg7 15 Bd6 is awkward) 15 Rfc1 Bb7 16 Nd2 g5 17 Bg3 h5 18 Ne4 is slightly better for White.
14 Rfc1
White has tried several moves here:
a) 14 Qc3 Be7 15 Rfe1 Kf7 takes us back into Line A24.
b) 14 Rac1 allows Black to exchange queens with 14 ... Qxb3 15 axb3 g5! (seizing space) 16 Bg3
Ba6 17 Rfe1 Kd7 and I don’t see why he should have any problems.
c) 14 Qxd5 cxd5 15 Rfc1 h5 16 h4 (or 16 Rc7 g5 17 Bd2 Bd6 and the intruder gets kicked off the
seventh rank) 16 ... Rh7 shows some great lateral thinking by Black.

Following 17 Ne1 Rb7 18 Nd3 a5 19 b3 Bd7 20 Rab1 Be7 21 Rc3 Black was again doing well in
S.Pavlov-M.Panarin, Moscow 2013. For example, he might play 21 ... Kf7 22 g3 Bb4 23 Rc2 Ba3 24
Re1 Rc8 25 Rxc8 Bxc8 and try to open lines with the pawn thrust ... a5-a4.
d) One advantage with our move order is that after the restraining 14 h4 Qxb3 15 axb3 Bb7 16 Rfc1
(if 16 Nd2 c5!) 16 ... Kf7 17 Nd2 Be7 18 Ne4 Rhd8 19 Nc5 Rxd4! 20 Nxb7 Rxf4 21 Rxc6 Rxh4 22 Rc7
Rd4 Black has a slight pull.
e) 14 Rfe1 g5!? 15 Bg3 Qxb3 16 axb3 Kf7 17 Rec1 Bb7 18 Nd2 Bb4 19 Nc4 Rhd8 and again the
pressure on the d4-pawn in conjunction with the bishop-pair offsets any problems that Black may have.
14 ... g5
With the king still on e8, I am not so worried about any sacrifices on g5.
15 Bg3
Alternatively, 15 Bc7 Kf7 16 Nd2 Ba6 17 Qxd5 cxd5 18 Bg3 Be7 19 Rc7 Rhc8 20 Rac1 Rxc7 21
Rxc7 Rc8 again leaves Black with few problems and the bishop-pair, while after 15 Nxg5? fxg5 16 Be5
Qxb3 17 axb3 Rg8 18 Rxc6 Be7 19 Rc7 a6 Black can outplay White with his extra piece.
15 ... Be7
 16 h4
If 16 Nd2 Black should play actively: 16 ... h5 17 h3 h4 18 Bh2 Rg8 19 Qxd5 cxd5 20 f3 a5! (now b2
will be a counter-target) 21 Nb3 (or 21 Rc7 a4 22 Rac1 Bd7 23 Rb7 Bd8 24 b3 Bc8 25 Rh7 Bb6!, which
is simply masterful) 21 ... Bd7 22 Nc5 Ra7 23 Rc2 Bxc5 24 Rxc5 a4 25 Rac1 Rb7 (further instructive
play by Black) 26 R1c2 Ke7 27 Kf2 Ra8 28 R5c3 Ra5 29 Rc7 Rb6 30 Ke3 a3! and Black was winning in
L.Gonda-B.Predojevic, Dunaharaszti 2012.
16 ... gxh4 17 Nxh4
It’s also hard to complain about 17 Bxh4 Qxb3 18 axb3 Bb7 19 Nd2 Kf7 20 Ne4 Rhd8 21 Ra4 a5 22
Nc5 Ba6 23 Rxa5 Bb5 24 b4 Rxa5 25 bxa5 Rxd4.
17 ... Qxb3!
Now this is well-timed and preferable to the 17 ... Bd7 of Z.Arsovic-A.Kovacevic, Banja Koviljaca
2013.
18 axb3 Bb7

This endgame is deceptive. At first it looks as though White is doing well with the black bishop
locked in on b7. However, Black has the greater ability to improve his position: for instance, there is an
open fifth rank which the h8-rook can use; in some situations the rook can manoeuvre itself via g4 to e4
and then invade on e2. Black’s bishop-pair also works incredibly well to hold his position together for
the time being.
19 Rc4
After 19 d5 exd5 20 Nf5 Kd7 21 Nxe7 Kxe7, although far from winning, Black is material up and can
be happy.
19 ... Kf7 20 b4 Rhd8 21 Nf3 Rd5 22 Ne1 a5 23 bxa5 Raxa5 24 Rxa5 Rxa5 25 Nc2 h5
The two bishops will soon come into play.

B) 8 Bc4
This is an alternative way to pressure the black knight. Here Black should respond with a retreat and be
happy to reach a decent version of the Grünfeld.
8 ... Nb6! 9 Bb3
9 Bb5 doesn’t trouble Black after the natural 9 ... Bg7 10 0-0 0-0 11 Bxc6 bxc6 12 Bf4 Bg4 13 Re1
Nd5 14 Be5 Bxe5 15 Rxe5 e6 16 Re4 Bxf3 17 Qxf3 Qb6 18 b3 Rad8 when he has the easier plan of
attacking the d4-pawn.
9 ... Bg7

10 d5
The most critical line. Black stands fine after 10 Be3 0-0 11 h3 (preventing the light-squared bishop
developing to g4) 11 ... Na5!:
a) 12 0-0 allows Black to double White’s pawns and remove one of the controllers of d5: 12 ... Nxb3
13 axb3 (or 13 Qxb3 Be6 14 Qc2 Nd5 15 Bg5 Rc8 and Black was the more active in O.Moilanen-
J.Kortelainen, Finnish League 1991) 13 ... Nd5! 14 Qd2 Re8 15 Rfc1 Bf5 is slightly better for Black.
b) 12 Bc2 Nac4 13 Bc1 Nd5 14 Nxd5 Qxd5 15 0-0 with equality.
10 ... Na5
Of course, we should attack the bishop.
11 0-0 0-0 12 Re1
Targeting the weak e7-pawn. If 12 Bg5 the best approach is to play 12 ... Nxb3 13 axb3 Re8, while
after 12 Bc2 Nac4! the knight is not only an annoyance for White, but will also be a fine blockader on d6.
12 ... Bg4!
Regardless of whether White pushes the pawn to d5, this is our favoured mode of development.
Those looking for a natural game can also play 12 ... Nxb3 13 axb3 Bg4 14 Bg5 Re8 15 h3 Bxf3 16
Qxf3 Nc8! (blockading the d5-pawn) 17 Re2 h6 18 Bf4 Nd6 with equality.
13 Bg5
13 h3 Bxf3 14 Qxf3 Nxb3! 15 axb3 Re8 16 Be3 Nc8 17 h4 h5 18 Ra4 Nd6 saw Black carry out the
ideal manoeuvre in G.Kuzmin-R.Keene, Rovij/Zagreb 1975.
13 ... Re8 14 h3
It is important to note that whilst 14 d6?! looks tempting, it gives Black too many pieces for the queen
after 14 ... Qxd6! 15 Bxf7+ Kxf7 16 Rxe7+ Qxe7! 17 Bxe7 Kxe7! (but not 17 ... Rxe7? which loses to 18
Ng5+ Kg8 19 Qxg4):

a) 18 Nb5? Kf8! (the king runs back to safety) 19 Qd6+ (19 Nc7 Rad8 20 Qc1 Re7 keeps the extra
material and after 21 Qf4+ Kg8 22 Qxg4 Rxc7 Black is much better) 19 ... Kg8 20 Nc7 Red8 21 Qb4
Rac8 22 Qxg4 Rxc7 23 Qe6+ Kh8 24 Ng5 Rf8 25 Rd1 h6 26 Qh3 Re7 and the pieces were clearly
superior to the queen in P.Tregubov-I.Kurnosov, Moscow 2008.
b) 18 Nd5+ Kf8 19 Qd2 Bxf3 20 Nc7 Rad8 21 Qf4+ Kg8 22 Nxe8 Rxe8 23 Qxf3 Bxb2 is only
slightly better for Black.
14 ... Bxf3 15 Qxf3 Nxb3 16 axb3 Bxc3 17 bxc3 Qxd5 18 Qxd5 Nxd5 19 c4 f6! 20 cxd5 fxg5

21 Rac1 a5
 21 ... Rad8 also ended in a draw in D.Fridman-M.Rodshtein, Beersheba 2013.
22 Rc7 b5 23 Rcxe7 Rxe7 24 Rxe7 a4 25 bxa4 bxa4 26 d6 a3 27 d7 Kf8 28 Re2
A draw seems rather likely.

Summary
The essence of 6 ... g6 is for Black to obtain a perfect anti-IQP position. After the exchange on d5, 8 Qb3
is the most critical response. White needs to play actively otherwise Black obtains everything he wants
from the opening. Indeed, if Black is left untroubled, he will play for ... c6-c5 or seize space on the
kingside. The e5-square is important defensively and should be controlled at all times, particularly if
there is danger on the c-line.
Overall, Black should play dynamically, not fear developing his bishop to b7 or king’s rook to h7, and
counterattack against d4, aiming to gradually open and take over the position.
Chapter Six
The Panov-Botvinnik Attack: 6 Nf3 g6 without 7 cxd5
1 e4 c6 2 d4 d5 3 exd5 cxd5 4 c4 Nf6 5 Nc3 Nc6 6 Nf3 g6

In this chapter we turn our attention to what happens if White doesn’t capture on d5 and instead plays
7 Bg5 or a normal developing move such as 7 Be2.
Let’s consider the most menacing move first where White applies pressure on the centre with 7 Bg5.
Here Black’s best reply is 7 ... Be6, supporting the centre with a piece rather a than pawn and threatening
a future ... dxc4 pawn grab.

White now has to make a decision: whether to eliminate the knight on f6 followed by c4-c5, or to play
c4-c5 immediately.

 Returning to 6 ... g6, in this chapter we will focus on:

A) 7 Bg5
B) 7 Be2

White’s other moves are either toothless or plain misguided:

a) 7 Bd3 doesn’t make much sense in view of the potentially weak d4-pawn. Indeed, after 7 ... Bg7
White must already go 8 h3 to prevent 8 ... Bg4.
b) If 7 c5 Black can respond thematically with 7 ... Bg7 8 Bb5 0-0 9 0-0 (or 9 h3 Ne4) 9 ... Bg4 10
Be3 Ne4! (a thematic idea we’ll see much more of below) 11 Be2 Nxc3 12 bxc3 b6 13 Qa4 Qc7 14 h3
Bxf3 15 Bxf3 e6 16 Rac1 bxc5 17 dxc5 (Huang Qian-I.Krush, Khanty-Mansiysk 2012), and here I like 17
... Rab8 18 Rfd1 Rb2 when Black’s pieces are beautifully placed.
c) 7 Qb3 is met with the standard reply 7 ... Na5!? 8 Qa4+ (or 8 Qb4 dxc4 9 Bxc4 Nxc4 10 Qxc4 Bg7
11 0-0 0-0 12 Re1 when Black is doing well after 12 ... b6 followed by ... Bb7) 8 ... Bd7 9 Qd1 dxc4 10
Ne5 Bg7.

A) 7 Bg5 Be6!
This quickly leads to another important tabiya:

A1) 8 Bxf6
A2) 8 c5
A3) 8 Qb3!?
A4) 8 Be2

Lesser common are:

a) 8 Qa4?! is the start of a suspect plan: 8 ... Bg7 9 Ne5?! 0-0! (Black should let the c6-pawn fall as it
does not make sense for White to spend so much time pawn hunting instead of developing his forces) 10
Nxc6 bxc6 11 c5 (11 Qxc6 dxc4 is already preferable for Black) 11 ... Bd7 12 f3 Qb8 (12 ... Qc7 was
seen in G.Fant-P.Roder, Copenhagen 2001, and 12 ... Re8!? is also rather tempting) 13 0-0-0 e5 with a
good position for Black.
b) 8 cxd5 Nxd5 (now Black’s pressure on d4 gives him an easy game) 9 Bd3?! (9 Be2 Bg7 10 0-0 0-0
11 Qd2 Rc8 12 Rac1 Qb6 is very pleasant and straightforward for Black) 9 ... Bg7 10 0-0 0-0 11 Re1
Re8 (the immediate 11 ... Nxd4 is possible too, and if 12 Nxd4 Bxd4 13 Bxg6 Bxf2+ 14 Kxf2 Qb6+ 15
Kg3 Qc7+ 16 Kf2 Qc5+, winning) 12 a3? Nxd4! netted a pawn in E.Sandberg-A.Gupta, Berkeley 2011.

White Commits and Exchanges on f6

 Taking on f6 is a pretty double-edged decision. White abandons the bishop-pair, but usually
establishes a queenside pawn majority after pushing through with c4-c5. I must say that White puts a lot of
trust in the fact that Black’s weak pawn structure will limit the scope of his own bishop-pair.

A1) 8 Bxf6 exf6

As we have just seen, White tends to advance his c-pawn here, but it isn’t forced:

A11) 9 Qb3
A12) 9 c5

In reality 9 cxd5 doesn’t promise too much for White: 9 ... Bxd5 10 Bb5 Bh6 (a little bit unnatural, but
this controls the c-file and gives the king some breathing space) 11 Nxd5 Qxd5 12 Qe2+ Kf8 (castling by
hand) 13 0-0 Kg7 14 Bc4 Qd6 15 Rfd1 Rhe8 16 Qd3 Na5 17 Bb5 Rec8 18 d5 a6 19 Ba4 Nc4 20 Rab1
b5 21 Bb3 Rc5 and Black doesn’t have any problems at all.
Note too that it doesn’t make sense to play 9 Be2 as Black can simply reply 9 ... Bb4!, as
recommended by John Emms. After 10 0-0 0-0 we have:

a) 11 c5 Bxc3 12 bxc3 b6! 13 Qa4 Qc7 14 Qa3 Na5 15 Rab1 Rfe8 (I believe it is a touch better to
play the multi-faceted 15 ... Bf5, attacking the rook on b1 and so preventing White from ganging up on the
weak b6-pawn; in addition Black prepares to take control over the e-file) 16 Rb2 Bg4 17 Qb4 Rab8 and
Black was fine in A.Kislinsky-V.Malakhatko, Kiev 2003.
b) 11 Rc1? loses a pawn to 11 ... dxc4 12 d5 Bxc3.
c) 11 cxd5 Bxd5 12 a3 Bxc3 13 bxc3 Rc8 14 Qa4? (or 14 Nd2 Na5 15 Rc1 Re8 16 c4 Be6 17 Qa4 b6
and White’s central pawns will be more of a liability than a strength) 14 ... Qa5! (exchanging the queens
only increases the grip Black has on the light squares) 15 Qxa5 Nxa5 16 Rac1 Nb3! won material in
A.Berke-E.Dizdarevic, Bosnjaci 2013.

A11) 9 Qb3

This is designed to bully Black into a panicked response.

9 ... Bb4!
Covering all the attacks.
The sly 9 ... a6 is also possible: 10 c5 (10 Qxb7? loses on the spot to 10 ... Na5 and 10 cxd5 Na5 11
Qa4+ b5 12 Qd1 Bxd5 is nothing to fear either) 10 ... Qc7 11 Be2 (11 Nxd5 leaves White in big trouble
after 11 ... Qa5+ 12 Nd2 0-0-0) 11 ... Rd8 12 0-0 Bg7 13 Rfd1 0-0 14 Rd2 Rfe8 15 Rad1 Na5 16 Qb4
Bh6 17 Rc2 Bf5 18 Bd3 Bg4 and Black slowly began to outplay his opponent in R.Riegler-M.Sekyra,
Graz 2010.
10 Be2
Alternatively:
a) After 10 c5 Bg4 11 0-0-0! (not 11 Be2 Bxf3 12 Bxf3 Nxd4 13 Qa4+ Kf8 14 0-0 Bxc3 15 bxc3
Nxf3+ 16 gxf3 and White’s opening has been a complete failure) 11 ... Bxf3 12 gxf3 Bxc3 13 Qxc3 (13
bxc3 0-0 14 Qxb7? Qa5 15 Qb2 Rab8 16 Qc2 Qa3+ already forced resignation in M.Mikadze-
N.Batsiashvili, Batumi 2012) 13 ... 0-0 Black’s position is completely safe.

White’s pawn structure is very ineffective for attacking and to make matters worse he has too few
minor pieces to secure an effective breakthrough. Unsurprisingly 14 h4 h5! (blocking White’s counterplay;
now all that is needed is to build up the rooks along the open e-file and pressure the weak d4-pawn) 15 f4
Qc7 16 Qd2 f5 17 Kb1 Rae8 18 Bg2 Re7 19 Rhe1 Rxe1 20 Rxe1 Qd8 21 Qe3 Kg7 22 Qg3 Qf6 23 Bxd5
Rd8 24 Bg2 Rxd4 leaves Black doing well.
b) After 10 a3?! Black will simply capture: 10 ... Bxc3+ 11 bxc3 0-0 12 Be2 Na5! with the
advantage.
10 ... 0-0 11 c5
11 0-0 Bxc3 12 Qxc3 dxc4 13 Bxc4 Bd5 is another position slightly more preferable for Black.
11 ... Bxc5!?
 12 dxc5 d4 13 Qxb7?!
Greedy, although we would hardly mind 13 Qc2 Qa5!.
13 ... dxc3 14 0-0
As played in J.Adamski-M.Malmberg, Copenhagen 2003. White loses by force after 14 Qxc6? cxb2
15 Rb1 Qa5+ 16 Nd2 (or 16 Kf1 Qxa2 17 Bd3 Bc4) 16 ... Rad8 17 Rxb2 Qc3.
14 ... Nd4! 15 Nxd4 Qxd4 16 bxc3 Qxc5
Once again Black is slightly better due to White’s weak queenside pawns.

A12) 9 c5

As we are fighting against the slow advance of the queenside pawns, Black’s first goal is to give both
bishops a role in the game. It is no good if one has an active role and the other does absolutely nothing. It
is for this reason that I am suggesting the interesting:
9 ... Bh6!?
There are several ideas behind this relatively untested move (9 ... Bg7 is the more mainstream
choice). Firstly, we have the plan of rerouting the bishop to the more active h2-b8 diagonal via f4.
Secondly, Black plans to aggressively open the position by advancing his kingside pawns. Note too that
just because the bishop has moved to h6, this doesn’t mean that it cannot retreat to g7 if it becomes
convenient to simply pile pressure on the weak d4-pawn.
10 Be2!
I’ve also spent a long time analysing the alternatives:
a) 10 h4 can simply be ignored: 10 ... 0-0 11 h5 g5! (a strange move, but one that can’t be punished;
now ... g5-g4 is in the pipeline) 12 Bb5 (or 12 Bd3 f5 and Black has no problems) 12 ... g4 13 Ng1 f5 and
Black has been able to play things the way he wanted to.
b) 10 g3 is a little slow: 10 ... 0-0 11 Bg2 b6! (undermining the queenside when c5 can’t be supported
by b2-b4) 12 0-0 Rb8 13 Qa4 Qd7 14 Rfd1 Rfc8 15 b4 bxc5 16 bxc5 Rb4 17 Qa3 a5 18 Na4 Bf8
(preventing Nb6, as if 19 Nb6? Rxb6!) 19 Rab1 Rcb8 20 Rb3 Bg4! and White is being outplayed. The
weakness of d4 and his lack of control over the b-file will start to tell.
c) After 10 Bd3 the e6-bishop gets a job and 10 ... 0-0 11 0-0 Bg4! is indicated.

In order to avoid the loss of a pawn White must play 12 Be2 (12 h3 Bxf3 13 Qxf3 f5 14 Qxd5 Bg7
sees the bishop return to capture the d4-pawn) 12 ... f5! 13 h3 Bxf3 14 Bxf3 Bg7 15 Nxd5 (or 15 Ne2 Qh4
16 Bxd5 Rad8 17 Bxc6 bxc6 18 Qa4 Rfe8 19 Rfe1 Re4 20 Qxc6 Qe7 21 Kf1 Re8 22 Qb5 Bxd4 and
Black has a pleasant bind) 15 ... Nxd4 when Black has a superb bishop on g7.
d) 10 Bb5 (trying to control any possible break on the queenside) 10 ... 0-0 11 0-0 (11 Bxc6 bxc6 12
0-0 Rb8 13 b3 Qa5 14 Na4 Rfe8 15 h3 Kg7 16 Re1 Bf4 is an example of how the two bishops can out
manoeuvre the two knights) 11 ... Rc8 (also good enough is 11 ... Bg4 12 Be2 f5 13 h3 Bxf3 14 Bxf3 Bg7
15 Bxd5 Bxd4 16 Qb3 Qe7 and the c-pawn is vulnerable) 12 h3 (prophylaxis against ... Bg4, but all this
does is create a target) 12 ... a6 13 Bd3 Bf4!? 14 a3 Qd7 and now:
 d1) 15 Na4 allows the wonderful 15 ... Bxh3 16 gxh3 (or 16 Nb6 Qg4 17 Ne1 Rcd8) 16 ... Qxh3 ( ...
Nxd4 is threatened) 17 Be2 Qg4+ 18 Kh1 Rfe8 19 Re1 Re3 20 fxe3 Qh3+ 21 Kg1 Bxe3#.
d2) 15 b4 Rfe8 (these may look like harmless developing moves, but Black is planning a devious
blow) 17 gxh3? (or 17 Ne2 Bc7 18 gxh3 Qxh3 19 Ng3 h5 20 Qd2 h4 21 Qh6 Nxd4 22 Nxd4 Bxg3 23 fxg3
Qxg3+ 24 Kh1 Qh3+ 25 Kg1 Qxd3 26 Nf3 Re4 27 Qd2 Qxa3 with big compensation for the piece) 17 ...
Qxh3 18 Be2 (if 18 Ne2 Qxf3 19 Nxf4 Qxf4 or 18 Re1 Nxd4!) 18 ... Re5! 19 dxe5 Nxe5 20 Re1 Nxf3+
21 Bxf3 Bh2+ 22 Kh1 Bg3+ 23 Kg1 Qh2+ 24 Kf1 Qxf2#. Simply beautiful.
10 ... 0-0 11 0-0 a6 12 a3

12 ... f5
I really like this move as White can’t lock in the e6-bishop. Instead, 12 ... g5 was a less accurate
move order in R.Bilek-J.Boguszlavszkij, Zdar nad Sazavou 2010.
13 b4!?
Otherwise:
a) 13 Re1 g5 14 b4 g4 15 Ne5 Bg7 16 Bf1 Qg5 17 Nxc6 bxc6 18 Ne2 f4 19 Qd2 Bh6 and Black is
certainly doing okay.
b) 13 Ne1?! Bg7 and the d4-pawn is very vulnerable.
 c) 13 Qb3 Bg7 14 Qxb7 (or 14 Rfd1 Qd7 15 Na4 Rad8 and it is not entirely clear what White wants
to do) 14 ... Nxd4 15 Nxd4 Bxd4 16 Rfd1 Bxc5 17 Nxd5 Rb8 18 Qc6 Rc8 19 Qb7 a5 when the dark-
squared bishop is just incredibly powerful.
13 ... g5
13 ... Qf6 is also tempting.
14 b5?!
Now the initiative passes to Black.
14 ... axb5 15 Nxb5
Again, if 15 Bxb5 then 15 ... Bg7! is good.
15 ... g4 16 Ne1 b6 17 Nc2 bxc5 18 dxc5 Qg5

19 Ncd4
After 19 f4 gxf3 20 Bxf3 Qe7 Black wins a pawn in view of 21 Nd6 Qa7.
19 ... Nxd4 20 Nxd4 Bg7 21 Nxe6 fxe6 22 Ra2 Rfc8 23 Rc2 Rxa3 24 Bb5 Rc7 25 Qe2 Kf7 26 Re1
Qe7
Black is holding the white attack at bay and can look forward to soon enjoying a material advantage.

The Queenside Pawn Majority after 8 c5

 This is a pawn structure that Black must become well acquainted with in the Panov. First of all, what
are White’s aims in determining the position like this?
The answer is that White hopes to benefit from the long-term advantages of having such a clamp - this
pawn structure is well-known for being highly dangerous in the endgame. Not only that, but White’s plan
for improvement is relatively straightforward: to push the queenside pawns. In order to avoid suffering
from this clamping effect, Black typically relies on three ways to alleviate his slight cramp:
a) The thematic ... Nf6-Ne4 hop.
b) Undermining the pawn majority with ... b7-b6.
c) Attacking the defenders of the d4-pawn.
Note that after 8 c5 it is too early for Black to jump in immediately with the knight. It is usually better
to wait for ... Ne4 until White has determined where his light-squared bishop is going.
Thus Black plays 8 ... Bg7 when it is up to White to develop the light-squared bishop with careful
consideration given to Black’s three strategic plans. Should White play a move without consideration to
our themes, for example 9 Be2, Black can now play 9 ... Ne4.

This has the positional threat of 10 ... Nxc3 11 bxc3 b6, breaking up the clamp.
One of the advantages of having our bishop sitting on the a1-h8 diagonal is that it becomes a natural
attacker against the unsupported d4-pawn. White simply cannot defend against all three of Black’s
strategic plans and should the light-squared bishop deploy itself to d3 or b5, then the Black bishop sitting
on e6 may be given a job to do on g4: for example, we meet 9 Bd3 with 9 ... Bg4.
Another example of ... Bg4 can be seen here:

A2) 8 c5 Bg7

9 Bb5
If 9 Be2 unsurprisingly I am tempted by the thematic 9 ... Ne4! 10 Bf4 0-0 11 0-0 Bg4 12 Rc1 e6 13
Re1 Bxf3 14 Bxf3 Ng5, which is slightly better for Black.
Instead, after 9 Bd3 Bg4 (thematically attacking the defender of the d4-pawn) 10 Be3 0-0 11 0-0 (or
11 h3 Bxf3 12 Qxf3? e5!) 11 ... Bxf3 12 Qxf3 e5! Black is already much better, while 9 Qc2 doesn’t work
because of 9 ... 0-0 10 Be2 Ne4! (anyhow) 11 Nxe4 Bf5 and Black has the advantage.
9 ... 0-0 10 Bxc6
If 10 0-0 again Black should play the thematic 10 ... Ne4! 11 Be3 Rc8, just as he does with 10 Qc2
Ne4!.
10 ... bxc6
Now Black cannot undermine the queenside with ... b6.
 11 Ne5
Instead, 11 0-0 Ne4 12 Bf4 (or 12 Be3 Nxc3 13 bxc3 Qa5 14 Qd2 Rab8) 12 ... Bg4! puts one of our
main plans into operation.
11 ... Qc7

12 Rc1?!
12 Bf4 is not satisfactory either because of 12 ... Nh5 and after 12 0-0 Black should continue with 12
... Ne4! 13 Nxe4 dxe4 14 Bf4 g5! 15 Bg3 Qb7 16 Qc2 f6 17 Nc4 Qb4 18 b3 Bxc4 19 bxc4 f5 20 Be5
Bxe5 21 dxe5 Rad8 when White will not be able to take advantage of Black’s advanced kingside pawns.
12 ... Ne4 13 Bf4 g5!
Energetic and thematic.
14 Bg3 f5!

Even stronger than the 14 ... Nxg3 15 hxg3 Bxe5 16 Qh5 Rfb8 17 Qxh7+ Kf8 18 dxe5 Qxe5+ of
F.Velikhanli-S.Nadir, Abu Dhabi 2003.
15 f4 Bxe5 16 dxe5 Rab8 17 Rc2 Rfd8 18 0-0 Nxg3 19 hxg3 d4 20 Nb1 d3 21 Rd2 Qa5 22 Rxd3
Rxd3 23 Qxd3 Qxc5+ 24 Kh2 Rxb2
Black’s pieces dominate the board.
A3) 8 Qb3!?

One of the trickiest sidelines to meet. White puts pressure on the centre and attacks the b7-pawn. As
can be expected we need to deal with this ‘trick’ rather directly.
8 ... Na5 9 Qa3
White is up to mischief again.
Instead, 9 Qa4+ allows Black to play 9 ... Bd7! 10 Qb4 dxc4!, as does 9 Qb5+ Bd7 10 Qb4 dxc4 11
d5 Qb6 12 Qxb6 axb6 13 d6 h6 14 Be3 exd6, which also looks good for him.
Should White play 9 Qb4 Bg7 10 Bxf6 (10 cxd5 is not very challenging: 10 ... Nxd5 11 Qa4+ Nc6 12
Bb5 0-0 and Black has a perfect position) 10 ... Bxf6 11 cxd5 Bxd5 12 Nxd5 Qxd5 13 Qb5+ then after 13
... Qxb5 14 Bxb5+ Kf8 and any Caro player can begin to outplay him on account of the weak d4-pawn and
the open c-line.
9 ... Bg7!?
Probably the simplest. The point behind White’s troublesome ninth move is that after 9 ... dxc4 the
position becomes incredibly complicated: for example, 10 d5! (10 Rd1 is no good because 10 ... Nd5!
blockades everything) 10 ... Nxd5 11 Rd1 f6! (11 ... Bg7? loses material after 12 Nxd5 Bxd5 13 Rxd5,
and if 13 ... Qxd5? 14 Qxe7#) 12 Be2 Bg7 13 0-0 0-0 14 Bd2 Nc6 15 Bxc4 Bf7 16 Nxd5 Bxd5 17 Bc3
Bxc4 18 Rxd8 Rfxd8 and by now Black is probably okay.
10 cxd5
Black has no problems after 10 Bxf6 Bxf6 11 cxd5 Bxd5 12 Bb5+ Nc6.
10 ... Nxd5 11 Bb5+ Nc6
 12 Bxc6+
If 12 Ne5 Rc8!, while 12 0-0 0-0 13 Bxc6 bxc6 14 Rac1 Nxc3 15 Rxc3 f6 16 Bf4 Bd5 is fine for
Black.
12 ... bxc6 13 0-0 h6!
This is a must. 13 ... 0-0? is not the best choice because of 14 Nxd5 Qxd5 15 Bxe7.
14 Bh4 Qd6! 15 Nxd5
After 15 Qb3 Rb8 16 Qc2 Bf5 17 Qa4 0-0 Black should be better on account of the bishop-pair, as
shown by 18 Bg3 Nf4 19 d5 Rb4!.
15 ... Bxd5
I think Black stands well here with his good bishops.
16 Qxd6? exd6 17 Rfe1+ Be6!
By now White is definitely worse.

A4) 8 Be2

White delays any decisions in the centre by playing normally for the time being. Black should respond
in kind because there will come a time (move 10!) when he will threaten to play ... dxc4 and simply win a
pawn. The big risk for White is that the longer he postpones making a decision in the centre, the better
placed Black will be for the time when White finally decides to act.
8 ... Bg7
As John Emms points out, 8 ... dxc4? would lose a piece after 9 Bxf6 exf6 10 d5.
9 0-0 0-0 10 cxd5
After 10 c5 Ne4 (this thematic knight move is still very pertinent here) 11 Be3 Nxc5! 12 dxc5 d4 13
Nxd4 Nxd4 Black does not have any problems and 10 h3 Rc8 forces White to make a decision over the
c4-pawn.
10 ... Nxd5

11 Qd2
Alternatively, 11 Rc1 Rc8 12 Qd2 Qb6! 13 Na4 (instead, 13 Rfd1 Rfd8 14 Na4 Qc7 15 Nc5 Bf5 is
equal and 13 Be3 Rfd8 14 Ng5 Nxe3 15 fxe3 Nxd4 16 exd4 Bxd4+ 17 Kh1 Be3 a good example of
Black’s tactical possibilities) 13 ... Qb4! leaves White under pressure.
11 ... Qb6!
Heaping pressure on the d4-pawn.
12 Rfd1 Rfd8 13 Rac1 Rac8 14 Na4 Qb4 15 Qxb4 Ndxb4 16 Nc5 Nxa2 17 Ra1 Bd5 18 Nxb7 Re8
19 Nc5 Nab4
Black’s active minor pieces give him the advantage.

B) 7 Be2
 We have now transposed into another line of the Grünfeld (perfect, as I don’t think many Panov
players will be familiar with this), and one considered unpromising for White. Black should simply play
naturally until the ‘happy moves’ run out.
7 ... Bg7 8 0-0 0-0 9 c5
As soon as White plays this committal move, we can leap with our knight. Thus he might prefer one
of:
a) If White plays the aimless 9 Bf4 Bf5 10 Re1 then 10 ... dxc4 11 Bxc4 Rc8 12 Be2 (after 12 Bb3
Na5 Black gains the bishop-pair and a very satisfactory position) 12 ... Nd5! is very pleasant for Black.
b) 9 Re1 Be6! (threatening the c-pawn) 10 c5 (or 10 Ng5 Bf5!) 10 ... Ne4!:
b1) 11 Bf4 Bg4 12 Be3 is well met by 11 ... f5!.
b2) 11 Be3 and now Black has the little tactic 11 ... Nxc5! 12 dxc5 d4 13 Nxd4 Nxd4 14 Bf3 Nxf3+
15 Qxf3 Qc7, which is slightly better for hi,.
b3) 11 h3! Qc7 12 Bb5 Rad8 13 Qa4 Bf5 14 Ne2 e5 15 Bxc6 bxc6 16 Be3 f6 is again slightly better
for Black.
c) 9 h3 deprives Black of the g4-square and now we can follow the recommendation of Delchev in his
book The Safest Grünfeld: 9 ... dxc4 10 Bxc4 Na5 11 Be2 (11 Bd3 is also met in the same way - with a
blockade of the d5-square) 11 ... Be6 12 Ne5 (or 12 Ng5 Bc4!) 12 ... Rc8 13 Bf3 Nd5 14 Ne4 Nb4 and
now the recommended move 15 Nc3 (15 Nc5? is met by 15 ... Rxc5) 15 ... Qb6 16 Be3 Nc4 17 Nxc4
Bxc4 18 Be2, although even here Black stands slightly better on account of the weak d4-pawn.
d) 9 b3 looks rather strange and Black should simply proceed as normal with 9 ... Bf5:
 d1) I must add that should White play 10 Nh4 then Black has the tactical trick 10 ... Nxd4!? (10 ... Ne4
is also very promising) 11 Qxd4 Ne4 12 Qd3 Nxc3 13 Nxf5 dxc4 14 Qxc4 b5 15 Qd3 gxf5 with a
pleasant position.
d2) 10 Be3 dxc4 11 bxc4 Ng4 is also pleasant for Black.
e) After 9 Be3 we can follow the plan of taking on c4 and blockading the d5-square: 9 ... dxc4! 10
Bxc4 Na5 (driving the bishop off its best diagonal) 11 Be2 Be6 12 Rc1 Nd5 13 Qd2 Nxe3 14 fxe3 Nc4
15 Bxc4 Bxc4 16 Rf2 Rc8 when Black should connect the rooks and play for the pawn break ... e7-e5.
f) 9 cxd5 Nxd5 with an easy game.
g) 9 Bg5 should be met with 9 ... Be6 (following the pattern with 9 ... dxc4 is also possible) 10 c5
Ne4 11 Bf4 Nxc3 12 bxc3 b6 13 Bb5 Na5 14 cxb6 axb6 and Black stands well with the open c-file and
backward pawn as a target.
9 ... Ne4

Highly thematic of course.

10 Be3
10 Bf4 neglects the defence of the d4-pawn and after 10 ... Bg4! 11 Rc1 (or 11 h3 Bxf3 12 Bxf3 Bxd4
13 Nxe4 dxe4 14 Bxe4 Bxc5 and Black is better on account of his material advantage) 11 ... e6 Black has
a very satisfactory position. The bishop retreat to e3 can be met with ... f7-f5, whilst 12 Nd2 is met by the
crafty 12 ... Qf6 13 Bxg4 (or 13 Be5 Bxe2 14 Bxf6 Bxd1 15 Bxg7 Nxc3 16 bxc3 Kxg7 17 Rfxd1 b6 and
Black is doing well) 13 ... Qxf4!.
10 ... b6!?
An unusual choice. More thematic would be 10 ... Nxc3 11 bxc3 b6 12 cxb6 axb6 followed by ... Ba6
and play along the open queenside lines.
11 cxb6
After 11 Qa4 Nxc3 12 bxc3 Bd7 13 Bb5 (if 13 Qb3 bxc5 14 Qxd5 cxd4 15 cxd4 Nb4 16 Qb3 Rb8
and Black has the preferable position) 13 ... Qc7 White cannot maintain the central pawn tension for too
long.
11 ... axb6 12 Rc1 Bb7 13 Qb3
Instead, 13 Qc2? Nxc3 14 Qxc3 Rxa2 nets a pawn, 13 a3 Nd6 14 Bd3 e5 15 dxe5 Nxe5 16 Nxe5
Bxe5 17 Be2 Nf5 favours Black, and non-moves such as 13 Re1 Nd6 14 b3 (or 14 Nxd5 Nf5 15 Bc4 Na5
16 Nb4 Qd6 17 a3 Nxc4 18 Rxc4 e6 19 Nd3 Rfd8 20 h3 Rac8 21 Rxc8 Rxc8 22 Qe2 Qd7 when White
can hardly make a move) 14 ... Nf5 give Black a very comfortable position.
13 ... Nxc3 14 Rxc3 Na5 15 Qc2 Nc4! 16 b3
Likewise if 16 a3 Nxe3 17 fxe3 Qd6.
16 ... Nxe3 17 fxe3 Qd6
Black is fine.

Summary
This chapter has seen those lines where White doesn’t exchange on d5 and either raises the tension in the
centre with 7 Bg5 or develops naturally with 7 Be2. The former should be met with the enterprising 7 ...
Be6 which will threaten the c4-pawn in the future. After 8 Bxf6 exf6 9 c5 Black’s goal is to make use of
the bishop-pair, perhaps by advancing the kingside pawns and targeting the weak d4-pawn. Should White
decide not to give Black the bishop-pair then Black must remember the strategically important ... Ne4!
when again White is vulnerable to a kingside expansion.
The tricky lines are only tricky if we can’t remember the antidote: 8 Bxf6 exf6 9 Qb3 allows 9 ... Bb4!
and the immediate 8 Qb3 should be chased away with 8 ... Na5!. Instead, the developing move 7 Be2 is
likely to be played by someone who is unfamiliar with the positional motifs - it doesn’t trouble Black at
all and he can simply concentrate on development.
Chapter Seven
The Panov-Botvinnik Attack: 6 Bg5
1 e4 c6 2 d4 d5 3 exd5 cxd5 4 c4 Nf6 5 Nc3 Nc6 6 Bg5

6 Bg5, piling pressure on d5, is an idea which stems from Botvinnik himself. With this move White
seeks complications and a sharp middlegame without allowing any long endgames. Here I am sticking
with my original recommendation and it is a different style of position to that which we’ve seen in the last
two chapters, being very solid and so perfect against those looking for a wild middlegame.
6 ... e6
In the Dangerous Weapons book I advocated a more dynamic try and one I can still recommend for
those looking for a sharper game, 6 ... dxc4 7 Bxc4 h6!?.
7 Nf3
The main line.
Instead, with 7 c5? White commits his pawn structure too early: 7 ... Bd7! (now Black is threatening
the strategic undermining ... b7-b6) 8 Bb5 (consistent; White wishes to capture on c6 and then begin the
advance of the b-pawn, but here this idea is deeply flawed) 8 ... b6! 9 Bxc6 Bxc6 10 b4 bxc5 11 bxc5
Bxc5! 12 dxc5 d4! 13 Nf3 dxc3 14 Qxd8+ Rxd8 15 Rc1 Ba4 16 0-0 c2 and Black was already winning in
A.Kovacevic-K.Kachiani, Interlaken 2001.
The most that White can hope for after the simple 7 cxd5 is equality: 7 ... exd5 8 Nf3 (winning a pawn
with 8 Bxf6 is really playing with fire: 8 ... Qxf6 9 Nxd5 Qe6+ 10 Qe2 Bd6 11 0-0-0 0-0 12 Qxe6 Bxe6
13 Nc3 Nb4! 14 a3 Nd5 and Black has very fast development and active play) 8 ... Be7! 9 Be2 0-0 10 0-0
h6 11 Bh4 Re8 12 Ne5 Ne4 and I don’t see any problems for Black in this symmetrical position.
7 ... Be7
 Now White can handle the position in two ways. By far the most popular and the most critical is the
early push with c4-c5. The alternative is to reach an IQP position which I believe is simply harmless.

A) 8 c5
B) 8 Bd3

Key Themes after 8 c5

Black should be aware of some positional considerations which underlie this variation. White has
firm control over the e5-square and will often aim to expand rapidly on the queenside by marching his
pawns onwards and upwards. It is important to recognize that White has the better pawn structure, so
Black should avoid blindly exchanging pieces into an endgame while leaving the queenside majority
intact.
As we have seen in related previous positions, it is absolutely crucial for Black to break free from the
bind with pawn breaks and find a way to give the light-squared bishop a role. This can be done in two
ways:
a) Playing ... b7-b6, particularly if White cannot play b2-b4.
 b) Playing the more risky and less usual ... e6-e5.
Undermining the c5-pawn with ... b7-b6 is very thematic and one of the most important concepts in
this variation. It is especially strong if White has mishandled the opening. So, for instance, after 9 Bd3
Black can simply reply 9 ... b6! when White’s queenside majority disappears. Other examples can be
found in the main line after 11 Rc1 when again Black plays 11 ... b6!?.

Another version of the ... b7-b6 idea comes in conjunction with a Stonewall set-up. In order to offset
the positional weakness of the e5-square and White’s advantage of the queenside majority, Black
unbalances the position by opening lines for his rooks to exert their power.

Here Black has used the pawn breaks ... b7-b6 and ... a7-a5 to generate activity to complement the
strong knight on e4. Structurally Black is worse, but dynamically he has terrific play on the queenside.
With regards to the concept of piece play, Black has a tougher job in hand than in previous Panov
lines and this is primarily due to the bad bishop trapped on c8. In this variation it becomes even more
crucial for Black to play the freeing knight hop ... Nf6-e4.
 Indeed, 9 ... Ne4 is probably our most important move in this variation. This important jump allows
Black to:
a) Relieve his spatial problems by exchanging a set of minor pieces.
b) Add power to a ... b7-b6 pawn break as the knight attacks the c5-square.
c) Make it harder for White to play b2-b4.
d) Use the knight on e4 as a stronghold and support it with ... f7-f5.
We can thus formulate another rule: please twin the move c4-c5 with the knight hop ... Ne4.
The ultimate aim of the knight hop is to achieve a position like the next diagram, but please note this
will only arise if White has mishandled the opening.

Here White did not cover the c3-knight with a piece and so allowed Black to ruin his stronghold on
c5.
One point that I cannot stress enough is that Black must undertake active measures to avoid suffering
under the cramping bind. The queenside pawn majority makes its advantage felt in the endgame - please
don’t allow White an easy hand getting there.
 This position from P.H.Nielsen-S.Milliet, French League 2004, is an example of how things can go
wrong. The simple reason is that Black did not sense the danger and did not play the active ... Nf6-e4 hop.
As a result White has a fantastic pawn structure on the queenside, one that threatens to overrun Black
completely, and Black has left it too late to start any central counterplay.

A) 8 c5 0-0!

Keeping the flexibility in the position and waiting for White to commit his light-squared bishop.
Recently I experimented with 8 ... Ne4 9 Bxe7 Nxe7, but now White can play 10 Bd3 f5 when he has
gained time to push the b-pawn.
9 Bb5!
By far the best move. White aims to control the e5-square by attacking the knight on c6 and to prevent
any undermining ... b7-b6 pawn break. At the same time he may try to saddle Black with a backward c-
pawn by exchanging the bishop for the knight and then attacking the resulting pawn on c6.
Less challenging are:
a) If 9 Bd3 Black should begin undermining White’s queenside structure immediately as White will
have to make a second move with his light-squared bishop to protect the c-pawn, i.e. 9 ... b6! 10 Bb5 (or
10 cxb6 Qxb6 11 0-0 Nb4 12 Be2 Ba6 and I don’t think Black has any problems) 10 ... Bd7 11 Bxc6
Bxc6 and now:

a1) If 12 b4 Ne4! 13 Bxe7 Qxe7 14 Rc1 (14 Qc2 is vulnerable to the undermining 14 ... a5!) 14 ... a5
15 a3 (15 Ne5 Nxc3! 16 Nxc6 Qc7 17 Rxc3 Qxc6 18 b5 Qxb5 is also good for Black) 15 ... Nxc3 16
Rxc3 axb4 17 axb4 Bb5 and Black catches the white king in the middle.
a2) 12 Ne5 Qe8 13 b4 h6 14 Be3 Ne4 15 Nxe4 dxe4 16 Nxc6 Qxc6 and Black has an easy target: the
d4-pawn.
b) 9 Rc1 also allows Black to begin the normal process with 9 ... Ne4 10 Bxe7 Qxe7 11 Bb5!
(otherwise ... b6 was threatened) 11 ... Nxc3 12 Rxc3 e5!? 13 dxe5 Nxe5 14 0-0 (M.Kobalia-A.Galkin,
Novokuznetsk 2008) 14 ... Bg4! 15 Be2 Nc6 16 Re1 Rfe8 and Black’s activity offsets the weak d5-pawn.
In fact should White play the catastrophic 17 Qxd5 Rad8 18 Qc4 Bxf3 19 gxf3 Black is already winning.
c) 9 a3 is too slow as Black will play the thematic 9 ... Ne4 10 Bxe7 Qxe7 11 Qc2 Nxc3 12 Qxc3 and
now he even has time for the drastic 12 ... e5 13 dxe5 Bg4 14 Be2 Bxf3 15 Bxf3 d4 16 Qb3 Qxe5+ which
wins material.
9 ... Ne4!

Again I cannot stress enough the importance of this move - quite simply it is the move that underpins
the whole variation. As we saw in Nielsen-Milliet, Black does not have enough space to develop
satisfactorily if he avoids this move. He must strike, as the proverbial says, when the iron is hot. First of
all Black grabs some space for his own pieces and then will make some exchanges.
10 Bxe7
10 Nxe4 is not promising for White after 10 ... dxe4 11 Bxc6 bxc6 12 Bxe7 Qxe7 13 Ne5 and now the
key point 13 ... Ba6!. Indeed, Black should be prepared to gambit a pawn to prevent White from short
castling (although White would be quite foolhardy to accept), D.Zagorskis-D.Fridman, German League
1998.
10 ... Nxe7!

The only way to recapture if we want to force through our ... b7-b6 pawn break. It is also the most
popular continuation and reaches something of a tabiya:

A1) 11 0-0
A2) 11 Rc1
A3) 11 Qc2

A1) 11 0-0
White absolutely needs to protect the c3-knight, but the text can easily be seen if he is not familiar with the
key motifs.
11 ... Nxc3!
Please remember this is our basic plan, to ruin the white structure and then undermine the c5 strong
point.
12 bxc3 b6 13 cxb6 axb6
 14 Qe2
Otherwise Black will play with ... Ba6 with ease.
14 ... Qc7! 15 Rfc1 Qa7
Best. Black is determined to trade the light-squared bishops. He wishes to play the numbers game and
prove that White’s two pawn weaknesses (on a2 and c3) outweigh the weak b6-pawn.
15 ... Bd7 is also satisfactory.
16 Ne5
Funnily enough I had this position this year, but when I reached move 15 and was on the verge of
playing 15 ... Qa7 I got worried about 16 Bd3. However, this was silly as Black can play 16 ... Ba6 and
the sacrifice 17 Bxh7+? Kxh7 18 Ng5+ Kg8 19 Qh5 completely fails to 19 ... Bd3.
Instead, after 16 a4 Ba6 17 Rcb1 Rfc8 Black is a touch better.
16 ... Ba6 17 a4

17 ... Bxb5!?
Black preferred 17 ... Rfc8 in A.Hunt-D.King, British League 2008, when 18 Nxf7 could have led to a
very dangerous position for Black (18 Qb2 Qc7 was fine for Black in the game), as shown by 18 ... Kxf7
19 Re1 Bxb5 20 Qxe6+ Kf8 21 axb5 (if 21 Re3 Qd7 22 Rf3+ Ke8 23 Qf7+ Kd8 24 Re1 Bxa4 25 Rfe3
Nf5 and Black is escaping the attack) 21 ... Qc7 (21 ... Qxa1 22 Qxe7+ Kg8 23 Qe6+ Kf8 24 Qxc8+ Rxc8
25 Rxa1, which is preferable for White.
18 Qxb5 Rfc8 19 Nd7 Rc6 20 Ne5 Rc7 21 Nd3 h6 22 h3 Nf5
Black will improve the placing of his knight to reach a more comfortable position.

A2) 11 Rc1

Defending against Black’s dream of getting White to recapture on c3 with a pawn. After this move
Black has two interesting tries, one undermining White’s strong point, the other aiming for a dynamic
Stonewall set-up:

A21) 11 ... b6!?

A22) 11 ... Bd7

A21) 11 ... b6!?

In my opinion White absolutely now has to play the most challenging move:
12 c6!
The logical 12 b4 is, of course, met with 12 ... a5! 13 a3 axb4 14 axb4 bxc5 15 dxc5 (15 bxc5? would
be very bad: 15 ... Nxc3 16 Rxc3 Qa5 and, oops, White has chucked a piece) 15 ... Nxc3 16 Rxc3 Rb8.
This has been assessed as being slightly better for Black and by continuing the analysis I would have to
agree: 17 Qe2 Ng6 18 0-0 Qf6 19 Rfc1 e5 is, indeed, slightly better for Black.
Instead, after 12 Nxe4 dxe4 13 Ne5 bxc5 14 Rxc5 Qa5+ 15 Qd2 Qxa2, according to GM Nigel
Davies, it is not clear what compensation White has for the pawn, while 12 Bd3 f5!? is an interesting idea
of Khenkin’s; Black aims to support the knight with a Stonewall pawn fortress.
12 ... Qd6
 The play does now become rather forced, but Black is completely fine.
13 Qa4
This was the choice of Naiditsch against Leko, but there it looked as though Black had it all under
control.
However, 13 0-0!? is met by the kick-back 13 ... a6 14 Bd3 Nxc6 15 Nxe4 dxe4 16 Bxe4 Bb7 17 Ne5
(of course, 17 Qc2 demands Black to play the forced 17 ... Nb4! and now 18 Bxh7+?? is a terrible
mistake because of 18 ... Kh8 19 Qb1 Bxf3 20 gxf3 f5 and the white bishop is lost) 17 ... Rac8 18 Qd3! (a
better try than the 18 Rc3 of L.Ootes-T.Burg, Belgian League 2013) 18 ... Rfd8! 19 Nxc6 (after 19 Bxh7+
Kf8 White has to give back the material) 19 ... Bxc6 20 Bxc6 (if 20 Qxa6 Ra8) 20 ... Rxc6 21 Rxc6 Qxc6
22 Qxa6 Ra8, which also gives Black a slight advantage.
13 ... a6 14 0-0 Bd7!

The point.
15 cxd7 axb5 16 Qxb5 Nxc3 17 Rxc3 Rxa2 18 Ne5
After 18 Rfc1 Ra5 19 Qb3 Ra7 20 Ne5 f6 21 Rc8 Ra8! 22 Rxa8 Rxa8 23 Qb5 Rd8 24 Nc6 Rxd7 25
Qxb6 Kf7 26 b4 e5 the position is roughly equal.
18 ... f6 19 Nc6 Qxd7 20 Nxe7+ Qxe7 21 Qxb6 Ra7 22 Rc6 Rb7 23 Rxe6 Rxb6 24 Rxe7 Rxb2
 The players soon agreed a draw in A.Naiditsch-P.Leko, Dortmund 2009.

A22) 11 ... Bd7!?

My original choice, offering the bad bishop for exchange. Very generally, to decide whether you
should exchange a set of pieces or not, I recommend that you play what IM Robert Bellin calls “the game
of equivalents” - just assess the value of your piece in comparison to the value of its enemy counterpart.
12 Bd3

Unsurprisingly White is keen to retain the bishops.

From a positional perspective, 12 Bxd7 does not make much sense as White trades off one Black’s
worst pieces for one of White’s more active ones. On the plus side, this move saves time and in the
openings time is often crucial. After 12 ... Qxd7 we should consider:
a) After 13 Ne5 here I like 13 ... Qc7 14 Nxe4 dxe4 15 0-0 Rad8 16 Re1 f5 17 Nc4 Rd5 18 Ne3 Rd7
with a level game.
b) 13 Qd3 Nc6! 14 Nxe4 Nb4! (this little zwischenzug saves the day) 15 Nf6+ (if 15 Qc3 Nxa2 is
possible and after 16 Ne5 Qd8 17 Qc2 dxe4 18 Rd1 Qd5 19 Qa4 f6 20 Nc4 b5! 21 Qxb5 Rfb8 22 Qa4
Nb4 23 Ne3 Qc6 24 Qxc6 Nxc6 25 Rd2 Rb4 26 d5 exd5 27 Nxd5 Rc4 Black has the better chances;
instead, 15 Qe2 dxe4 16 Qxe4 Nxa2 recovers the pawn and after 17 Ra1 Nb4 Black has a nice outpost on
d5 and a target on d4) 15 ... gxf6 is most certainly not a disaster for Black. Indeed, the doubled f-pawns
protect the important e5-square and open up the g-file, which will benefit Black.
12 ... f5!?
With this move Black establishes a stronghold in the centre, reinforcing the e4-square and setting the
stage at some point for a kingside offensive. However, I have some words of caution: Black should
recognize that it is White who holds all the long-term strategic cards in the shape of this queenside pawn
majority and the e5-square.
To offset this, Black needs to keep his position alive with dynamic possibilities, even if it means
opening lines on the queenside. We also need to combine this dynamic approach with targeting White’s
weaknesses: for instance, we can apply pressure to the soft spot on d4. Where possible Black should play
to maintain the knight on e4 - mindless retreats are not allowed and at times it may even be possible to
sacrifice the knight for a mighty pawn phalanx. Our third ambition is to begin play where we are the
strongest, which is on the kingside. Here we can utilize some common ideas in the Stonewall: the rook lift
from f6-h6 and bringing the light-squared bishop into play via e8-h5.

There are good points to Black’s play, but I must stress that ... f7-f5 is highly committal, leaving a big
hole on e5 for a white knight. If this happens and Black has not generated any counterplay then he is
effectively in a bad Stonewall and positionally lost. Now:

A221) 13 Ne5?!
A222) 13 0-0

Using the Stonewall analogy, if White plays 13 Bxe4 fxe4 in order to jump in with the knight with 14 Ne5,
Black should play 14 ... Nc6!, and if 15 Nxc6 bxc6! followed by ... Qg5 and ... e6-e5.
Instead, after 13 b4 b6 14 0-0 a5 15 a3 axb4 16 axb4 bxc5 17 bxc5 Ng6 18 Nb5 Bc6 Black will plan
to use the open lines on the queenside for his two rooks.

A221) 13 Ne5?!
 Jumping into this square when White has no back-up control is positionally dubious.
13 ... Nc6!
Of course, Black needs to challenge the white knight.
14 Nxd7
Following 14 Nxc6 bxc6 (14 ... Bxc6 was tried in T.Hirneise-J.Van den Bersselaar, Gausdal 2008,
but I prefer the text) 15 0-0 (15 f3? is too hasty and loses to the old trick 15 ... Qh4+ 16 g3 Nxg3) Black
should play 15 ... e5! 16 dxe5 (or 16 f3 Nxc3 17 bxc3 e4 and Black has a very pleasant position) 16 ...
Qe7 17 Ne2 (White loses a pawn after 17 Bxe4 fxe4 18 Qd4 Rae8, while 17 Re1 Rab8 18 b3 Nxc5 19
Bb1 Ne6 20 Qd3 g6 21 Ne2 Qb4 22 Red1 a5 23 g3 a4 results in dynamic equality) 17 ... Rab8! Black
will regain his pawn and is fine after 18 Qc2 Qxe5 19 b3 Rfe8 20 f3 Nf6.
14 ... Qxd7 15 Ne2
15 f3 Nxc3 16 bxc3 e5 gives Black an advantage.
15 ... Qe7 16 0-0 Qg5!?
Now the bothersome f2-f3 won’t be possible because of ... Qe3+.
Instead, after 16 ... e5 17 Bb5! (17 f3 allows 17 ... exd4!, which is a thematic piece sacrifice; in return
for the material Black obtains a dangerous pawn mass and a tremendous amount of space) 17 ... Rac8 18
f3 Nf6 19 Bxc6 bxc6 20 dxe5 Qxe5 21 Nd4 White has a very small plus.
17 a3
After 17 Kh1 Qh4 18 Kg1 Rf6 19 f3 Rh6 20 h3 Ng3 21 Nxg3 Qxg3 22 Kh1 Qf4 the d4-pawn will fall.
17 ... Rf6
 The rook swings into the attack.
18 Qb3
Taking care and wisely avoiding both 18 b4 f4 19 f3 Raf8! 20 fxe4? f3 and 18 f3? Qe3+.
18 ... Rb8 19 Bxe4 fxe4 20 Qe3 Qh5 21 f3 exf3 22 Rxf3 Rxf3 23 Qxf3 Qxf3 24 gxf3 a5 25 Kf2 b6
26 Ke3 Kf7
The position is pretty balanced, but Black can be happy. His rook will soon target the weak b2-pawn
whilst his knight is perfectly placed blockading the passed c5-pawn.

A222) 13 0-0
White chooses to develop in a simple way. He wants to exploit the e5-square and advance on the
queenside.
13 ... b6!?

Inspired by John Nunn’s advice on queenside pawn majorities, I tested his theory of opening up the
queenside to generate active play for the black pieces. Naturally we have to worry about the protected
passed pawn on c5, but we are a long way from a king and pawn ending just yet.
Instead, in my original book I recommended 13 ... Nc6, putting pressure on the d4-pawn, defending the
hole on e5 and very importantly delaying the advance of the queenside pawn majority.
14 b4
Others:
14 cxb6 gives Black an easy game: 14 ... axb6 15 Ne1 Qb8 16 f3 Nxc3 17 bxc3 Ng6 and White has
made too many awkward moves.
b) 14 Ne5 bxc5 15 dxc5 Be8 16 b4 Qb8! saves the day.
14 ... Ng6!
14 ... a5 was my original line, but the computer pointed out that 15 b5 bxc5 16 dxc5 Nxc5 17 Nxd5
Nxd5 18 Rxc5 Qb6 19 Ne5 Qxc5 20 Nxd7 Qc3 21 Nxf8 Rxf8 is better for White.
15 g3
Here too I considered a number of possibilities:
a) 15 b5? would be a mistake after 15 ... bxc5 16 dxc5 Nxc5 17 Nxd5 Nxd3 when Black is much
better.
b) 15 Ne2 Qe8! (guarding the b5-square) 16 Qb3 Ba4 17 Qb2 Rb8! 18 Qa3 (or 18 a3 bxc5 19 dxc5
e5 20 h4 h5) 18 ... Bc6 19 Qxa7 bxc5 20 bxc5 Ra8 21 Qb6 Rf7 22 Qb3 Rfa7 and suddenly Black has a lot
of counterplay on the queenside.
c) 15 Ba6 prevents our desired ... a7-a5 and after 15 ... Rb8! White might try:

c1) 16 c6 Bxc6 17 Nxe4 fxe4 gives White less than nothing.

c2) 16 a3 and now we have a clever manoeuvre aimed at getting the liberating ... e6-e5 into place: 16
... Rf7 17 Re1 bxc5 18 dxc5 Qf6 19 c6 Bxc6 20 Nxe4 fxe4 21 Rxc6 Ne7 22 Rd6 Nf5 23 Rc6 Ne7 with a
strange draw by repetition.
c3) 16 Re1 already gets White into trouble after 16 ... bxc5 17 bxc5 Qa5 18 Bd3 Rb2 when Black is
very active.
c4) 16 Ne2 bxc5 17 bxc5 Rb2!?, with a further divide:
c41) 18 Bd3 Bc6 19 Rb1 Rxa2 20 Qc1 Qe7 21 Ne1 (not 21 Nf4? Nd2 and Black is doing very well)
21 ... e5 22 f3 Ng5! (the knight needs to manoeuvre itself to e6; the position is still complicated, but Black
has good chances) 23 Nc2 e4 24 fxe4 fxe4 25 Rxf8+ Qxf8 26 Ba6!! Ne6 27 Nb4 Rxe2 28 Bxe2 Nxd4 29
Ba6 Nf4 30 Qe3 Qxc5 31 Nxc6 Nf3+ 32 Kf2 Qc2+ 33 Be2 Qxb1 34 Ne7+ Kh8!? is a long, fun and
instructive line, and one which should end in perpetual check according to the silicon beast.
c42) 18 c6?! Qb6! 19 cxd7 Qxa6 20 Rc7 Qa5 (Black will win the d7-pawn) 21 Qc1 leads to an equal
position after 21 ... Rxe2 22 Rc8 Nd6! 23 h4! (if 23 Qc6 Nf7) 23 ... Qb6 (this move is absolutely
necessary to move the queen in range of the d7-pawn; after 23 ... Nf7? 24 h5 Ne7 25 Re8! Black can’t
defend the knight on e7) 24 a4 Nxc8 25 dxc8Q Rxc8 26 Qxc8+ Nf8.
d) 15 Re1 a5! 16 a3 axb4 17 axb4 bxc5 18 dxc5 Rb8 19 Na2 Qc7 20 c6? Bxc6 21 b5 Qa7! wins.
15 ... a5!

16 a3
Best. After 16 b5 bxc5 17 dxc5 Nxc5 18 Nxd5 Nxd3 Black is much better.
16 ... axb4 17 axb4 bxc5 18 bxc5 Qa5 19 Qc2 Bc6
Black has enough activity to compensate for the weak e5-square and passed pawn.
20 h4 Rfb8 21 h5
Further precision. Instead, after 21 Kg2 Qa3 ... Rb2 is coming.
21 ... Ne7 22 Ne5 Be8!

23 Bxe4
After 23 Be2 Qb4 Black will follow up with ... Nc6, pressing on d4.
23 ... fxe4 24 Qe2
Now Black needs to shift his queen over to the kingside to avert any potential danger there:
24 ... Qd8 25 Qg4 Nf5 26 Rfd1
 If 26 Ne2 Bb5 27 h6 Qf6 and Black is much better.
26 ... Qf6 27 Qf4 Ne7 28 Qg4 Rb2
This has shown how to play actively as Black and he is doing pretty well here.

A3) 11 Qc2 Qa5

Protecting the knight on e4 through indirect means.
12 Bd3 b6!
Undermining the c5-pawn is the most accurate move order since it limits White’s possibilities. If 12
... Nxc3 13 bxc3 (13 Qxc3 Qxc3+ 14 bxc3 b6 15 cxb6 axb6 16 Kd2 Ba6! was fine for Black in I.Madurga
Lopez-L.Solaesa Navalpotro, Madrid 2011) 13 ... h6 then 14 Rb1! prevents Black from getting in an
effective ... b7-b6 break. With the move order I am recommending, this possibility is eliminated.
13 0-0
After 13 Bxe4? dxe4 14 Qxe4 Ba6! the white king is caught in the middle:

a) 15 Ng5 Ng6 16 h4 bxc5 and then:

a1) If 17 h5 cxd4 18 hxg6 (tactics like 18 Nxh7 do not work after 18 ... dxc3 19 b4 Qd8 20 hxg6
Qd2#) 18 ... Qxg5! (Black should eliminate the dangerous knight) 19 gxf7+ Kxf7 and Black is safe from
harm.
a2) Castling queenside does not solve White’s problems either: 17 0-0-0 cxd4! 18 Rxd4 Rac8 19 Rh3
and now instead of 19 ... h6 as played in E.Virnik-E.Chasovnikova, Mureck 1998, Black could have
improved her position with 19 ... e5!.
b) 15 0-0-0 bxc5 16 Ng5 Ng6 17 dxc5 Rab8! 18 Qd4 Rfc8 19 Nge4 f5 20 Nd2 Qxc5 and again Black
is better.
Instead, 13 a3 bxc5 14 dxc5 Qxc5 15 Nxe4 Qxc2 16 Bxc2 dxe4 17 Bxe4 Rb8 18 b4 Ba6 is equal.
13 ... Nxc3 14 bxc3 h6
 15 Rfb1
Even after 15 Nd2 Black should play 15 ... Ba6! as White has no threats: for example, 16 Nb3 Qa4 17
Qe2 Bxd3 18 Qxd3 Rfb8 19 Rfe1 Qc4 (Black craftily offers the exchange of queens to chase the knight
away from its protection of the c5-square) 20 Qf3 (if 20 Qxc4 dxc4 21 Nd2 bxc5 22 dxc5 Rb5 and Black
is better on account of White’s weak c-pawn) 20 ... bxc5 21 Nxc5 Nf5 22 h3 Rb2 23 Nd3 Rc2 and Black
is doing very well.
15 ... Ba6

We will now follow the game K.Klundt-R.Dautov, Bad Weissee 2001, where Grandmaster Dautov
showed exactly how to play against White’s pawn weaknesses:
16 Bxa6
16 Ne5! is maybe a better try for the advantage. That said, 16 ... Bxd3! (16 ... bxc5 17 Bxa6 Qxa6 18
Nd7 cxd4 19 Nxf8 is better for White) 17 Nxd3 is okay for Black:
a) 17 ... bxc5 18 Nxc5 Rab8 19 Na4 Nf5 (19 ... Rxb1+ is too accommodating, although Black held on
for the draw in L.Nisipeanu-B.Lalith, Pardubice 2012) 20 Nc5 Nd6.
b) 17 ... Rfb8! 18 Rb3 bxc5 19 Nxc5 Rxb3 20 Qxb3 Nf5 21 Rb1 Nd6 is level.
16 ... Qxa6 17 cxb6
 Again 17 Ne5 is possible too:
a) 17 ... bxc5 18 Nd7 Rfe8 19 Nxc5 Qc6 20 Rb7 Nf5 21 Rab1 Nd6 22 R7b4 (if 22 Rd7 Red8) 22 ...
Nc4 23 Qa4 Qxa4 24 Rxa4 a5 looked interesting for Black in L.Klundt-E.Krueger, Velden 2009.
b) 17 ... Rfc8 18 Nd3 Nc6 19 Rb2 Na5! 20 a4 (not 20 Rab1? Nc4 and Black nets an exchange) 20 ...
Nc4 21 Rb5 bxc5 22 Nxc5 Qd6 23 Qd1 Rab8 when Black should play to control the b-file or undermine
the centre with ... e6-e5.
17 ... axb6 18 Qb3 Rfc8 19 a4 Qa5 20 Ra3 Nc6 21 h3 Qa7 22 Raa1 Rab8 23 Qb5 Qc7 24 Rc1
Na5 25 Ne5 f6 26 Nf3 Qc4 27 Rab1 Qxb5 28 Rxb5 Nc4 29 Kf1 Ra8
The a-pawn must fall.

Some General Rules in an IQP

An IQP position arises when White, probably not knowing the main plan with 8 c5, simply develops his
bishop from f1. Of course this move allows us to play 8 ... dxc4 9 Bxc4.

Black’s Goals:
a) To control the advance of the d4-pawn.
b) To swap minor pieces and control open lines.
The main strategy for Black is the blockade of the d5-square, which is usually achieved with a knight
sitting directly on the square or controlling it from f6. The ‘bad’ light-squared should be developed on the
long diagonal, either on b7 or, if that is not possible, via d7 and c6. If White begins an assault on h7, in
general Black should try to avoid pushing the pawn to h6 which may simply become a target. A much
better defensive move is ... g6. I personally call this formation of f7-g6-h7 ‘the wall’ as it blocks the b1-
h7 diagonal. The weakness of the dark squares is not such a problem: Black’s dark-squared bishop is
located close to the king and can go to its aid much more easily than its light-squared partner.

B) 8 Bd3
While we discuss the IQP, I should mention too 8 a3!?, as played by Grandmasters Vescovi and Suba.
This move is not bad in itself as it is a preventative move and one often seen in IQP positions; White
prevents the c6-knight from going to b4. After 8 ... 0-0 9 Rc1 dxc4 10 Bxc4 Nd5 White might try:
 a) 11 Bxe7 Ncxe7 12 0-0 b6 13 Ne5!? (after 13 Nxd5 Nxd5 14 Qa4 Bb7 15 Ba6 Qe7 16 g3 h5 Black
had absolutely no problems in Y.Balashov-A.Kharlov, St Petersburg Zonal 1993) 13 ... Bb7 14 Qf3 Rc8
15 Bd3 Nxc3 16 Qxb7 Rc7 17 Qa6 Ncd5 is fine for Black.
b) With 11 h4?! White tries to play the position in a very aggressive manner, but the young Indian
Grandmaster Harikrishna showed up the defect in White’s strategy: 11 ... h6! 12 Qd3 hxg5! 13 hxg5 g6 14
Ne5 Nf4! (protecting against Qh3) 15 Qg3 Qxd4 16 Nf3 Qxc4 17 b3 Nd3+ 18 Kd2 Qf4+ 0-1, G.Vescovi-
P.Harikrishna, Khanty-Mansiysk (rapid) 2006.
8 ... dxc4 9 Bxc4

White is now a tempo down on a line of the Queen’s Gambit Accepted. Black’s chief plan, as outlined
above, is to blockade the advance of the d-pawn, which will be done by playing ... Nf6-d5 or ... Nc6-b4-
d5. He must also neutralize White’s strong grip on the squares e5 and c5, which is usually achieved by
swapping off into an endgame.
9 ... 0-0 10 0-0 b6 11 a3
A common idea, White stops any pesky ... Nc6-b4 jumps while getting ready to line his queen and
bishop up on the b1-h7 diagonal.
White can try to eliminate the weak pawn with 11 d5, but after 11 ... Nxd5 12 Bxd5 exd5 13 Bxe7
Nxe7 Black simply cannot be worse.
Likewise, 11 Re1 Bb7 12 a3 Rc8 13 Ba2 Nd5 14 Bd2 Nxc3 15 Bxc3 Bf6 16 Ne5 g6! (but not 16 ...
Bxe5 17 dx5 as played in D.Rensch-B.Finegold, US Internet League 2010) 17 Nxf7? (instead, 17 Qf3
Nxe5 18 Qxb7 Nc6 19 d5 Bxc3 20 bxc3 exd5 21 Rad1 d4 22 cxd4 Rc7 23 Qa6 Nxd4 gives Black the
advantage, while he has no problems after 17 Qa4 Rc7 18 Rad1 Ne7 19 Bc4 Nd5) is not so scary, as after
17 ... Rxf7 18 Bxe6 Nxd4 19 Bxf7+ Kxf7 20 Qa4 b5 21 Qxa7 Qd5 Black wins on the spot.
11 ... Bb7 12 Qd3!
White gets ready to build up an attack by first centralizing all his pieces.
12 ... Nd5!?

A standard technique by Black to eliminate a set of minor pieces.

13 Bxd5
We would begin to take over the light squares and queenside in the event of 13 Bd2 Nxc3 14 Bxc3
Na5 15 Ba2 Bd5.
13 ... Bxg5 14 Be4 f5!
Slightly unusual, but good. Instead, 14 ... h6 15 d5 exd5 16 Nxd5 Ne7 17 Nxe7+ Qxe7 18 Nxg5 hxg5
is slightly awkward.
15 Nxg5
After 15 Bxc6 Bxc6 16 Nxg5 Qxg5 17 f3 the weakness of the backward e-pawn is counterbalanced by
the weakness of the d4-pawn: for example, 17 ... Rf6 (Qc4 was threatened) 18 Rae1 Rd8 is about equal.
15 ... Qxg5 16 Bf3 Rad8 17 Rfe1 Kh8 18 Qf1! Rxd4
 19 Qb5!
19 Rxe6? would be a mistake because of 19 ... Nd8 20 Re3 f4 21 Rd3 Ba6!, which wins material.
19 ... Na5 20 Bxb7 Nxb7 21 Rxe6 Nc5 22 Re8 Rd8 23 Rae1 Qd2 24 h3 Ne4 25 Rxf8+ Rxf8
The position is completely equal.

Summary
The aggressive 6 Bg5 in the Panov should be met with the solid 6 ... e6. Naturally White’s best way of
reacting to this is 7 Nf3 Be7 8 c5. The alternatives are not very promising for White and I must say not
particularly difficult for Black to handle.
Black must deal with this intruder on c5 with the important knight jump on move 9 to e4, with the plan of
exchanging on c3 and undermining with ... b6. White can and should prevent this plan with 11 Rc1. Black
can then play 11 ... b6, which seems to lead to a draw, or if wishes for a more energetic game he can opt
for 11 ... Bd7 followed by ... f7-f5.
 Chapter Eight
The Exchange Variation
1 e4 c6 2 d4 d5 3 exd5 cxd5 4 Bd3 Nc6 5 c3

The Exchange Variation was thought to be one of the most harmless lines against the Caro until Bobby
Fischer utilized it to beat former world champion Tigran Petrosian in 1970. Although not quite as
harmless as it appears, the line is popular (particularly on the Internet) because of its easy and direct
strategy based on very classical ideas of development.

The Asymmetrical Pawn Structure

The asymmetrical pawn structure to a certain extent determines the play. White’s plan in the opening
stages is certainly very simple: to develop the bishops to d3 and f4, the queen follows to c2 or maybe b3,
and then the f3-knight will launch itself into e5, right where those central pawns are pointing. After this,
White can merrily begin a kingside attack.
The long-term plan for Black is to focus his middlegame efforts on the queenside using what is known
as the minority attack, where the two queenside pawns advance up the board hoping to saddle White with
a backward pawn and less space. This advance is particularly dangerous if Black has deposited a knight
on the c4-square. The idea is usually seen when the centre has reached a standstill - either when there is
maximum tension (i.e. both sets of pieces have reached their potential), or minimum tension (the majority
of the minor pieces have been swapped off).
That summarizes the basic aim of the opening for both sides, but to begin with Black has to fight two
opening battles: the first is to gain control over the e5-square and the second is to find a role for the light-
squared bishop. The latter can indeed be a fundamental problem, for if White plays Bd3 and h2-h3 then
where indeed does the problematic bishop go?

The First Battle: the e5-square

Both sides have a strong square which they will seek to control and occupy, often before beginning their
respective attacks. For White it is the important e5-square and for Black it is c4. For this reason it is vital
that Black gets his move order right. His first task is to prevent the dark-squared bishop developing to f4,
so we should see the following move order: 4 Bd3 Nc6! 5 c3 Qc7!? (5 ... Nf6 is another traditional line,
but here White will play 6 Bf4! controlling e5 - don’t do what I did at the Olympiad this year when I
paused, remembered that ... Nf6 is possible and suddenly my hand had made the move for me).

With 5 ... Qc7 Black signals to White that is prepared to contest the e5-square. Now the opening battle
may revolve around White kicking the queen off this important diagonal or he may simply ignore this
slight hiccup and continue as normal. The e5-square will become a hotbed of tension and can become
tremendously important should White fail to control e5 at all.
In fact we can even formulate a very important rule: if White has neglected his development and
doesn’t make any effort to control the centre with Ng1-f3, we should strike out with the dynamic pawn
push ... e7-e5.
Indeed, just take a look at the following examples:

Here White has just played 8 Bf4 (note the placing of the white knight on e2) when the correct move is
8 ... e5!. Do not forget this rule! I once did to my regret and in a game against GM Stuart Conquest.
 Stuart has just played 7 Ne2. I didn’t appreciate the strength of the ... e7-e5 idea and tried the
inaccurate 7 ... g6?!. Of course I should have played 7 ... e5! with an excellent position.
Finally, we should consider 6 Bg5 Nf6 7 Bxf6 gxf6 8 Qf3.

Now Black has the surprising 8 ... e5!, something, as you will see later, which gives Black a
marvellous position. It is an important move to remember as it forms the tactical basis for going 5 ... Qc7.

The Second Battle: the c8-bishop

Rather like in the Queen’s Gambit Declined, we should consider ourselves very happy if we can develop
the light-squared bishop outside the pawn chain, especially to the g4-square. However, Black still has to
be careful even when the bishop arrives at its desired square. The first danger is that a bishop sitting on
g4 can become vulnerable to a Ne2-Ng3 manoeuvre, as in a situation like the following.
 White has just played 7 Qb3, escaping from the pin, and now the bishop is in danger of being trapped out there on g4 after 7 ... e6 8 Ng3,
so Black should simply play 7 ... Bxe2.

Neutralizing the Bd3 and Qc2 Diagonal Attack

White can often exert uncomfortable pressure along the b1-h7 diagonal. However, this pressure can be
easily neutralized with a bishop retreat to g6.

Here Black should simply play 10 ... Bh5! preparing ... Bg6 which should neutralize any potential
danger, while copying White’s own bishop manoeuvre.

White Prevents ... Bg4

Sometimes White plays an early h2-h3 so that Black cannot develop the bishop to the desired g4-square.
Usually in such situations, Black will react by fianchettoing with ... g7-g6 and forcing the exchange of the
bishops on f5. So, for instance, we might see:
 White might, however, be smarter and prevent this concept with Qc2.

Now the bishop can’t go to g4 or f5, but there is a downside to White’s play, one we can exploit with
10 ... Nh5, aiming for ... Nf4, after which White will not be able to stop us from releasing the bishop.

1 e4 c6 2 d4 d5 3 exd5 cxd5
 4 Bd3
Standard, if not the only move available to White:
a) 4 Nf3 Nf6 5 Bf4!? (preventing ... Qc7; after 5 Bd3 Black plays 5 ... Bg4 and 5 Ne5 g6 transposes to
Line E in our final chapter) 5 ... Nc6 6 c3 Bf5 7 Nbd2 (or 7 Ne5 Nd7 8 Bd3 Bxd3 9 Nxd3 e6 10 0-0 Be7
11 Nd2 0-0 12 Nf3 Nf6 and Black doesn’t really have any serious problems; he has the long-term plan of
staging a minority attack on the queenside while White is lacking a light-squared bishop to damage our
kingside with) 7 ... e6 8 Qb3 Qc8 9 Nh4 Be4 10 f3 Bg6 11 g4 Be7 12 Nxg6 hxg6 13 Bd3 Nd7 14 0-0
(E.Sveshnikov-D.Arngrimsson, Albena 2013) 14 ... Bd8 and Black will reroute the bishop to c7.
b) 4 Bf4 Nc6 5 c3 Nf6 6 Bd3 Bg4 7 Qb3 Qc8 8 Nd2 e6 9 Ngf3 Be7 10 0-0 Bh5 (getting ready to
challenge the key diagonal) 11 Ne5 Nxe5 12 Bxe5 0-0 with easy development.
4 ... Nc6 5 c3 Qc7

I am sticking with my original recommendation. White now has three major plans. He will try to:
a) Stop our c8-bishop developing outside the pawn chain even at the cost of neglecting development.
b) Disrupt our development with a timely Bg5.
c) Gain control over the f4-square.
Thus we will focus on:
A) 6 h3!?
B) 6 Bg5
C) 6 Ne2

Other unusual options are:

a) 6 f4 and here we should follow Schandorff’s recommendation with 6 ... Nh6 7 Nf3 Bf5 8 0-0 e6 9
Ne5 Bd6 10 Qe2 0-0 11 Be3 Rae8 12 Nd2 f6 13 Nxc6 bxc6, with a pleasant position in M.Rusanov-
A.Shimanov, Peterhof 2007.
b) 6 Nd2 Nf6 7 Ngf3 Bg4 8 0-0 e6 9 Re1 Bd6 10 Nf1 (with the aim of playing Ng3 and winning the
bishop-pair) 10 ... 0-0 11 Ng3 (after 11 Bg5 Nd7 12 Bh4 Bh5 13 Bg3 Bg6 Black is completely equal;
there is no white attack on the kingside so Black can look forward to the long-term prospects of a minority
attack on the queenside) 11 ... Rae8 12 h3 Bxf3 13 Qxf3 e5 (the common strike for Black once White has
done something funny in the opening) 14 Nh5 Nxh5 15 Qxh5 g6 16 Qd1 exd4 with advantage to Black.

A) 6 h3!?

White’s aim is clear: to prevent Black from developing his light-squared bishop to g4. Of all the
variations in this chapter, this is probably the most annoying for the unprepared; it is annoying if you have
no idea where to put the light-squared bishop. There are, however, some drawbacks to White’s sixth
move: the first is that White takes time out in the opening to make a little pawn move, so Black might
strike out with the enterprising ... e7-e5. The second benefit for Black is that as the white kingside has
been slightly weakened, the plan of ... g7-g6 followed by ... Bf5 becomes a lot more attractive.
6 ... Nf6!
The most flexible move and now White must decide what to do with his knight on g1:

A1) 7 Ne2
A2) 7 Nf3

With 7 Nd2 White does not control any central squares so our reaction is standard: 7 ... e5!. Black should
not worry about the resulting IQP since he will have easy piece play. We should also bear in mind that
White’s weaknesses are on the b1-h7 diagonal and note that we have strong squares on e4 and c4. These
may become important later on. Following 8 dxe5 Nxe5 9 Bc2 Bc5 10 Nb3 (E.Geller-Y.Razuvaev, USSR
Championship, Minsk 1979) 10 ... Bd6 11 Nf3 0-0 12 0-0 Re8 13 Be3 (or 13 Bg5 Ne4 14 Be3 Nc4 15
Qxd5 Nf6 16 Qd4 Nxe3 17 fxe3 Be6 when Black’s two bishops along with the heavy pressure against the
weak e3-pawn give him good compensation for the pawn) 13 ... Bd7 14 Nxe5 Bxe5 15 Qd3 a6 is equal.
Instead, 7 Qe2? Nxd4 nets a pawn.

A1) 7 Ne2
This is aimed at making ... g6 less favourable as Black will not be able to swap off the bad light-squared
bishop. However, White is relinquishing control over the e5-square, so I believe Black should abandon
his strategic ideas and opt for active play in the centre.
Remember: No control over the centre? Play ... e7-e5!
7 ... e5!

8 dxe5 Nxe5 9 Bc2

White here has a variety of other moves to choose from:
If 9 Be3?! Bd6 10 Nd4 a6! (there is no way we will ever allow White to play Nb5, exchanging the
bishop) 11 0-0 0-0 12 Bc2 (if 12 Nf5 Bxf5 13 Bxf5 Nc4) 12 ... Nc4! (as mentioned before, Black makes
use of the strong squares) 13 Bc1 Re8 14 Nd2 b5 15 N2f3 Ne4 16 Qd3 Rb8! (with the beautiful idea of a
rook lift along the sixth rank to aid a potential kingside attack; instead, 16 ... Bb7 was played in B.Giulio-
A.Rasmussen, Dresden Olympiad 2008) 17 Rb1 Rb6 18 b3 Ne5 19 Nxe5 Bxe5 and Black has tremendous
piece coordination.
b) 9 Bb5+ can be met with the natural 9 ... Bd7 10 Bxd7+ Qxd7 11 Ng3 Bd6 12 0-0 0-0 13 Bg5 Ne4!
14 Nxe4 dxe4 15 Nd2 f5 and Black has not only the superior pawn structure, but an outpost on d3 for his
knight, A.Bilgin-R.Bigaliev, Istanbul 2010.
c) 9 Bf4 is, of course, met with 9 ... Bd6 10 Bxe5 Bxe5 11 Nd2 0-0 12 0-0 Bh2+ (standard technique
to push the king one square away) 13 Kh1 Bd6 14 Nd4 a6 15 Re1 Bd7.
 Jobava now handled the black side of the position very impressively: 16 Qf3 Rae8 17 Nf5 Be5 (17 ...
Bxf5 18 Bxf5 Qb6 would be a safer method) 18 Nb3 g6 19 Nfd4 Qb6 20 Re2 Bb8 21 Rxe8 Rxe8 22 Bf1
h5 23 Nc5 Bxh3 24 Na4 Qd6 25 Qxh3 Ne4 26 Kg1 Qf4 27 Nf3 Ba7 0-1, E.Krivoborodov-B.Jobava,
Moscow 2008.
d) 9 Nd4 doesn’t really promise anything. Black may have an isolated pawn, but he has a strong
square for his knight on e4 and the bishops: 9 ... Nxd3+ 10 Qxd3 Bc5 11 0-0 0-0 12 Be3 a6 13 Nd2 Re8
14 Rfe1 Bd7 15 Nc2 Bd6 16 Bd4 (16 Bg5 Bb5 17 Qf3 Ne4 is slightly better for Black too) 16 ... Nh5 and
although the game ended in a draw in A.Semeniuk-P.Maletin, Vladivostok 2012, Black stands a bit better
on account of his superior piece activity.
9 ... Bd6 10 Bg5
After 10 Nd4 a6 (of course, self-explanatory) 11 0-0 0-0 12 Nf5 Bxf5 13 Bxf5 Rfe8 14 Nd2 Rad8 15
Nf3 h6 16 Nd4 Nc4 Black has no problems; his knights have good posts and the dark-squared bishop is
on an excellent diagonal.
10 ... Ne4

11 Be3 0-0
A cool retort. 11 ... Nf6 and 11 ... Be6 are more boring ways of playing the position.
 12 Qxd5
If 12 Nf4 Bf5 13 Nxd5 Qc6 14 Nf4 Nc4 with advantage to Black.
12 ... Nf6 13 Qd4 Nc4 14 Bg5.
After 14 Nd2 Nxe3 15 Qxe3 Bd7 16 0-0 Bb5 17 Rfe1 Rfe8 18 Qf3 Qe7 19 Bd3 Bc6 20 Qf5 Rad8
Black has huge compensation for the pawn as White can hardly move a piece: for instance, 21 Nc4 Bb8
and Black is lining up an attack on the dark squares.
14 ... Be5 15 Qh4 h6 16 Bxf6 Bxf6 17 Qe4 g6 18 Bb3
Not 18 0-0? which drops a piece to 18 ... Bf5.
18 ... b5!

19 Qxa8 Bb7 20 Qxa7

Alternatively, 20 Qxf8+ Kxf8 21 0-0 Nxb2 22 Nd2 Qd7 and the knight on d2 doesn’t have a good
home.
20 ... Ra8 21 Qxa8+ Bxa8
Again Black’s lead in development gives him a very pleasant position.
22 0-0 Nxb2 23 Nd2 Qd7
Black wins back his material with the upper hand.

A2) 7 Nf3
 White’s main move.
7 ... g6!
Of course, the bishop must be developed.
8 0-0
8 Qc2 looks like good prophylaxis against Black’s planned ... Bf5, but we have a crafty trick up our
sleeve: after castling we can play ... Nh5-f4 and chase away the bishop. 8 ... Bg7 9 0-0 0-0 leads to:
a) I like that even after 10 Bg5 that Black can stick to the plan with 10 ... Nh5.

After 11 Qd2 f6 12 Be3 e5 13 g4 e4! 14 Be2 (if 14 gxh5 exd3 15 h6 Bh8 16 Kg2 Qb6 17 Ne1 Ne7 18
Nxd3 Qe6 19 Rh1 Nf5 and White is struggling with his open king) 14 ... exf3 15 Bxf3 f5 16 gxh5 f4 17
Bxd5+ Kh8 18 hxg6 Bxh3 19 Bg2 Bxg2 20 Kxg2 f3+ White was being killed in D.Pershin-M.Matlakov, St
Petersburg 2013. Impressive play.
b) If 10 Be3 Black needs to be a bit smarter and play 10 ... Be6! (waiting for White to develop the
queen’s knight) 11 Nbd2 (if 11 Ng5 Bd7 and White’s best move is to retreat the knight) 11 ... Rac8! 12
Rac1 Nh5 and now White has no way of preventing ... Nf4.
c) 10 Re1 Nh5 11 Be3 Nf4 12 Bf1 Bf5 with equality.
8 ... Bf5!?
 At first sight it looks very strange to double the pawns in this manner. There is a great deal of logic
behind this idea, though: Black secures the central e4-square, opens the g-file for his rooks and leaves
White lumbered with a bad bishop. I also happen to think that this is the best move order as White is not
given time to play Re1 and hide the bishop on f1.
9 Re1
9 Bxf5?! gxf5 only helps Black:
a) 10 c4 (directly challenging Black’s set-up) 10 ... Bg7 11 Nc3 Rd8 12 Nb5 (or 12 Qb3 dxc4 13
Qxc4 e6 and I don’t see any problems for Black) 12 ... Qb8 13 cxd5 (13 g3 is the kind of move my
students face all the time; after 13 ... dxc4 14 Bf4 Qc8 15 Nc7+ Kf8 Black is threatening ... Nh5, winning
material) 13 ... Nxd5 14 Nh4 e6 15 Bg5 Rd7 16 Nf3 0-0 (it might have been worth throwing in the move
16 ... h6 first) 17 Qd2 Nf6 18 Bh6 was agreed drawn in Z.Efimenko-A.Riazantsev, Olginka 2011.
b) 10 Bg5 Ne4! 11 Qc1 f6 12 Bf4 e5 13 Bh2 0-0-0 14 b4 f4 and Black is much better, L.Donovan-
Q.Ducarmon, Belgian League 2012.
c) 10 Qe2 might be the best try with the idea of a quick c3-c4: 10 ... e6 11 c4 dxc4 12 Qxc4 0-0-0 13
Nc3 Rg8 14 Bg5 (14 Rd1 Nd5 15 Bg5 Be7 16 Nb5 Qb6 17 Bxe7 Ndxe7 is a double-edged position
where Black should be okay) 14 ... Be7 15 Rac1 Nd5 16 Bxe7 Ndxe7 17 Nb5 Qb6 18 Ne5 a6 19 Nc3
Qxd4 20 Qxd4 Rxd4 21 Nxf7 Ng6 and Black has the better coordination with ... Nf4 on the way.
9 ... Bxd3!
This is an important move since after the natural 9 ... Bg7? 10 Bxf5 gxf5 11 Nh4! it is difficult for
Black to defend the f5-pawn.
10 Qxd3 Bg7
 11 Nbd2
Black has no problems after 11 Bg5 0-0 12 Nbd2 Rac8 13 Qe3 (both 13 Nf1 and 13 Nb3 are well met
by 13 ... Ne4) 13 ... Rfe8. His long-term aim is to begin a minority attack on the queenside, exploiting the
fact that he can make use of the semi-opened c-file. White, on the other hand, will try and make use of the
important e5-square while trying to chip away at the black king.
Likewise, after 11 Ne5 0-0 12 Bf4 Qb6 13 Nd2 e6 Black has a solid position.
11 ... 0-0 12 Ne5
This can be taken. Instead, after 12 Nb3 Black should make some useful developing moves: for
example, 12 ... e6 13 a4 Rfe8 14 g3 Ne4! 15 Bf4 Qd8 16 h4 h6 17 Kg2 g5! 18 hxg5 hxg5 19 Be3 f5
(Black’s position is a beautiful picture) 20 Ng1 Qf6 21 Ne2 Rad8 22 f3 Nd6 and Black had the upper
hand in R.Ponomariov-Le Quang Liem, Dortmund 2010.
12 ... Nxe5 13 dxe5 Nd7 14 Nf3
14 e6 is not promising as after 14 ... fxe6 15 Rxe6 Nc5 16 Qxd5 Nxe6 Black has the advantage.
14 ... e6 15 Bf4 Rfc8 16 Re2 Rab8 17 Qe3 b5
Black has handled the position perfectly. White has no play in the centre and now the play on the
queenside gives Black the easier position, E.Jonsson-E.Thingstad, Stockholm 2012.

B) 6 Bg5
 I would describe this as the classical approach. White is primarily interested in achieving good
development and securing harmonious squares for his knights.
6 ... Nf6!
This move may be surprising since Black voluntarily offers to give himself doubled pawns, but should
White choose to exchange he will be giving Black a helping hand in establishing a nice pawn centre.
White has played several moves here:

B1) 7 Bxf6?!
B2) 7 Nd2

There’s also:
a) 7 Nf3 Bg4 8 Nbd2 transposes to Line B2, as, indeed, would 6 Nf3 Nf6 7 Bg5 Bg4 8 Nbd2.
b) It dawned on me that White might also play 7 h3. Now I think that Black should be bold and strike
out: 7 ... e5!? (7 ... h6 followed by ... e7-e5 is also acceptable) 8 dxe5 Qxe5+ 9 Qe2 Bc5 10 Qxe5+ Nxe5
11 Bb5+ Bd7, and if 12 Bxf6 gxf6 13 Be2 Rg8 Black will follow up with ... d5-d4.
c) 7 Qb3 is also met with the pawn push 7 ... e5! (the rule, the rule!) 8 Nd2 (8 dxe5 hands Black the
advantage after 8 ... Qxe5+ 9 Be3 Bc5! and 8 Ne2?! leaves the white bishop on d3 with no good squares;
after 8 ... e4 9 Bb5 a6 10 Bxc6+ bxc6 Black has a nibble based on the bishop-pair) 8 ... e4! 9 Bc2 Be6
and White has problems with his queen.
d) If 7 Bh4 I like 7 ... e5 as 8 Bg3 doesn’t really pose a problem because of 8 ... Bd6.
e) 7 Ne2 allows Black to develop the bishop, so we should play 7 ... Bg4 8 0-0 e6 9 Qc2 Bh5! 10
Re1 (10 Nd2 is a mistake because of 10 ... Bxe2 11 Bxe2 Nxd4 12 Qa4+ Nc6) 10 ... Bg6 11 Nd2 Be7
with no problems.

B1) 7 Bxf6?! gxf6 8 Qf3

White fails to prevent Black’s planned break, but the damage has already been done. Likewise, after 8
Nd2 e5 9 Qh5 Be6 10 Ne2 0-0-0 Black is doing well.
8 ... e5! 9 Qxf6
This doesn’t work, but 9 Qxd5 Be6 10 Qf3 0-0-0 11 Ne2 exd4! also gives Black a tremendous
position.
9 ... Rg8
 Hitting the g-pawn.
10 g3
Alternatively:
10 dxe5 does not help White either: 10 ... Qxe5+ 11 Qxe5+ Nxe5 12 Bf1 (White is forced backwards)
12 ... Bf5 13 Nd2 0-0-0 14 0-0-0 Bc5 15 f3 Nd3+ 16 Bxd3 Bxd3 17 g3 Be3 and it is very difficult for
White to develop without returning the extra pawn. For example, 18 Nh3 Be2 19 Rde1 Bxf3 and Black is
much better.
b) 10 Bxh7 is equally horrendous after 10 ... Be7 11 Qf3 e4, trapping the white bishop.
10 ... Nxd4!

This is the tactical point.

11 cxd4 Qc1+ 12 Ke2 Qxb2+ 13 Nd2 e4 14 Rb1 exd3+ 15 Kf1 Qa3 16 Re1+ Be7
Black has a great position.

B2) 7 Nd2 Bg4

8 Ngf3
White plays in a very classical manner, but Black too has a very simple objective: neutralize the b1-
h7 diagonal and play for a minority attack.
We should also consider:
a) 8 Qb1?! has the aim of swapping the g4-bishop for a knight, but it’s not so logical. I believe the
best way to play against this is to follow the rule: 8 ... e5! 9 h3 Bd7 10 dxe5 Qxe5+ 11 Be3 Bc5 12 Ngf3
Qe7 13 0-0 (White erred with 13 Kd1 Bxe3 14 Re1 0-0 15 Rxe3 Qd6 16 Ke1 d4 when Black was doing
very well in G.Sarakauskas-R.Ovetchkin, Plovdiv 2008) 13 ... Bxe3 14 Re1 Bxf2+ 15 Kxf2 Be6 and
Black is a pawn up with a safe position.
b) 8 Qa4!? was played against me by IM Andrew Greet in a rapid game back in 2006. However,
Black is doing fine after 8 ... e6 9 Ngf3 Nd7 (directed against Ne5 ideas) 10 0-0 Bd6 11 Bh4 Bh5 12 Bg3
Bg6 13 Bb5 Bxg3 14 hxg3 0-0 15 Rfe1 Rfc8 16 Re3 a6 17 Bxc6 Qxc6 18 Qb3 h6 19 Ne5 Nxe5 20 dxe5
b5 by when he has the standard plan of attacking the white queenside pawns.
c) 8 Qb3 e6 (8 ... e5!? is also possible as 9 Bxf6 gxf6 gives Black a slight advantage and 10 Qxd5? is
definitely not possible because of 10 ... Rd8, regaining the pawn with some advantage) 9 Ngf3 Bd6 and
Black continues in much the same way with 9 ... Be7 10 0-0 Nd7 11 Bh4 Bh5 12 Bg3 Bg6 13 Bxd6 Qxd6
14 Be2 0-0.
d) 8 Be2 is really not a problem. Black simply retreats and goes 8 ... Bf5 9 Ngf3 e6 10 Bh4 Bd6 (10
... Be7 is also possible) 11 Bg3 0-0 12 Bxd6 Qxd6 13 0-0 Rab8 14 Qa4 (M.Narciso Dublan-F.Peralta,
Badalona 2009) 14 ... a6, planning a minority attack on the queenside.
e) 8 Qc2 Bh5 9 Bh4 e6 10 Ngf3 Bd6 11 0-0 Bg6, again with a reasonable position.
8 ... e6

Black has not faced any problems developing his pieces - this means we have a very normal position.
The first concrete idea is to neutralize any potential kingside attack.
9 Bh4
9 h3 Bh5 nudges Black to follow the correct plan of ... Bg4-h5-g6. Indeed, after 10 0-0 Bd6 11 Qa4
(or 11 Qb3 0-0 12 Rfe1 Rab8 13 a4 Nd7 14 g4 Bg6 15 Bxg6 hxg6 16 Bh4 a6 17 Nf1 b5 18 Bg3 b4 with
the advantage for Black in L.Donovan-N.Eliet, Belgian League 2011) 11 ... Bg6 12 Bxg6 hxg6 13 Rfe1 0-
0 14 Re2 b5! (a common strike) 15 Qc2 Rac8 16 Qd3 b4 17 Rc1 bxc3 18 Rxc3 Qb6 Black’s control over
the open c-line and White’s weak d4-pawn hand Black an easy game, P.Tvarijonas-D.Fridman, Riga
2012.
Instead, after 9 0-0 we can see the neutralization plan in action: 9 ... Bh5 10 Re1 Bg6 11 Be2 Bd6 12
Nf1 Ne4 13 Be3 (A.Bitalzadeh-S.Ernst, Groningen 2006) 13 ... 0-0 14 Nh4 Na5 15 Nxg6 hxg6 16 Bd3
Nc4 17 Bxe4 dxe4 18 b3 Nb2 19 Qc2 Nd3 and Black is much better.
9 ... Bd6 10 Bg3 Bh5!

This manoeuvre is an important defensive idea. The idea is to swap off White’s menacing light-
squared bishop on d3. However, it is usually only a good idea to play ... Bg6 when (a) Black has not
castled short and intends to go queenside, or (b) White has played h2-h3, thereby preventing ideas of Rh3
and Qh4. This is about the only special manoeuvre Black has to remember; everything else is
straightforward.
11 Bxd6
With 11 0-0 White intends Re1 followed by Ne5, but after 11 ... Bxg3 12 hxg3 Bg6! 13 Qe2 (13 Qc2?!
fails to 13 ... Nb4 and 13 Bxg6?! hxg6 plays into Black’s hands; he will castle queenside and attack down
the h-file) 13 ... Bxd3 14 Qxd3 0-0 15 Rfe1 Rac8 Black is fine.
11 ... Qxd6 12 0-0 0-0 13 Re1 Rab8
Intending to begin the queenside offensive.
14 a4 Qc7 15 Qb1 a6 16 Ne5 Rfe8 17 h3 Bg6 18 Bxg6! hxg6 19 Qd3 Nxe5 20 dxe5
Alternatively, 20 Rxe5 Rec8 with the idea of ... Ne8-d6-c4.
20 ... Nd7 21 Qd4 Rec8 22 Re3 Qb6! 23 Qxb6 Nxb6
Black’s position was the easier to play in M.Illescas Cordoba-V.Anand, Linares 1994.

C) 6 Ne2
 With the aim of playing 7 Bf4.
6 ... Bg4!
Solving the problem of the light-squared bishop.
7 f3
The most popular continuation. White does not want to see Black exchange the perceived ‘bad’ bishop
for a knight that can in the future have influence over the e5-square.
Let’s look at what happens if White avoids f2-f3. There is one rule here: when Ng3 is threatened,
capture the knight on e2:
a) 7 Qb3 Bxe2! (following the rule; Black does not want his bishop back on d7 and with one of the
knights gone, he can rest safe in the knowledge that there is one less piece eyeing up the critical e5-
square) 8 Bxe2 Nxd4 (a good route to equality, if a little drawish; 8 ... e6 9 Na3 Rc8 is an alternative
method of playing) 9 Qxd5 and now instead of the original recommendation 9 ... Nxe2, I now like 9 ...
Rd8 (the tempting 9 ... Nc2+? would lose immediately after 10 Kd1 Rd8 11 Bb5+) 10 Bb5+ Nxb5 11
Qxb5+ Rd7 12 Be3 a6 13 Qa4 Nf6 14 Nd2 Nd5 when Black had easy development in J.Groz-S.Kojima,
Cappelle-la-Grande 2011.
b) If White acts normally with 7 0-0 we go 7 ... e6 8 Bf4 (8 f3 transposes to a position we’ll see
below, while after 8 Qe1 Bxe2! 9 Qxe2 Nf6 10 Nd2 Bd6 11 Nf3 0-0 12 Bg5 Nd7 it’s not very clear what
White’s plan will be, S.Vidovic-E.Dizdarevic, Sibenik 2009; Black, on the other hand, will play for the
pawn break ... e6-e5 in order to activate his knights) 8 ... Qxf4 9 Nxf4 Bxd1 10 Rxd1 Bd6 11 Ne2 Nge7
12 Nd2 Rb8 13 a4 0-0 14 Nf3 Rfc8 15 g3 a6 16 Kg2 Na5 17 Nd2 Bc7 18 h4 h6 19 Kf1 g6 20 Kg2 Kg7,
as played in R.Gallego Martinez-J.Becerra Rivero, Andorra 1995.
 Black eventually won the game by playing his knight to c4 and targeting White’s Achilles’ heel, the
weak b2- and c3-pawns.
c) 7 h3 Bxe2! (7 ... Bh5 is also possible, but after 8 Qb3 Black must capture the knight because of the
threatened Ng3) 8 Qxe2 Nxd4 gave Black an easy game in Fan Yun-A.Ostrovsky, Philadelphia 2011.
d) 7 Qa4 Bxe2 (following the formula) 8 Bxe2 e6 9 Bg5 Be7 10 Bxe7 Ngxe7 11 Nd2 0-0 12 Qd1 e5
13 dxe5 d4 14 0-0 dxc3 15 bxc3 Qxe5 16 Rc1 Rad8 17 Qe1 Rfe8 and Black already had the upper hand
in T.Heinemann-L.Nisipeanu, Hamburg 2012.
e) 7 f4 looks strange. Black should simply control the f5-square and get on with developing his
pieces: 7 ... e6 8 0-0 Nh6 9 Nd2 Bd6 10 Nf3 Bf5 11 Ne5 Bxd3 12 Qxd3 Nf5 13 Nxc6 (or 13 g4 Nfe7 14
Ng3 0-0 15 g5 f6, breaking out) 13 ... Qxc6 14 Bd2 h5 15 Rae1 0-0 16 Rf3 b5 17 Rh3 g6 18 Qf3 Kg7 19
Nc1 Qc4 20 Kh1 b4 and Black’s initiative on the queenside was more dangerous than White’s stunted
attack in D.Hiermann-G.Neff, Lienz 2011.
f) We should all know by now how to react to 7 Qc2?: 7 ... Bxe2 and neither reply is good due to the
potential fork on d4.
g) 7 Na3 was played against me once. Obviously Black should deal with the threat in question and
play 7 ... a6! 8 Nc2 e6 9 Bf4 Bd6 10 Bxd6 Qxd6 11 f3 Bh5 12 Ng3 Bg6 13 0-0 Nge7 14 f4 Bxd3 15 Qxd3
g6 (following the rule of keeping f4-f5 under check) 16 Ne3 and now in M.Larsen-J.Houska, Harstad
2013, I was afraid that after 16 ... h5 17 f5 gxf5 18 Ngxf5 Nxf5 19 Nxf5 exf5 20 Qxf5 the position was not
entirely clear, but simply 16 ... 0-0 is fine for Black.
h) 7 Bf4 Qxf4 8 Nxf4 Bxd1 9 Kxd1 e6 10 Nd2 Bd6 11 Nh3 Nf6 and Black has no problems in the
endgame.
i) 7 Nd2 e6 8 Nf1 with the idea of manoeuvring the queen’s knight to g3 was an interesting try by GM
Rozentalis. As White has, though, wasted some time with the knight, Black should simply retreat the
bishop: 8 ... Bh5 9 Nfg3 Bg6 10 Nf4 Nf6 11 Nfh5 Nxh5 12 Nxh5 0-0-0 and Black has few problems with
his development.
Finally, we can return to 7 f3:
7 ... Bd7
 Now that Black has provoked a weakness, he retreats to safety. The critical question for both sides is
when can Black strike out with the thematic and aggressive ... e7-e5.
Instead, 7 ... Bh5?! gives White the better position after 8 Bf4!? Qb6 9 Qb3! (threatening to exchange
queens and leave Black with weak doubled pawns) 9 ... Qxb3 10 axb3.
After the prudent text, we come to the main lines:

C1) 8 Na3
C2) 8 Bf4

Otherwise:
a) 8 f4 is strange and Khenkin expertly demonstrated how to take advantage of the weak e4-square: 8
... g6 9 Nd2 Nf6 10 Nf3 Bf5! 11 Ne5 Bxd3 12 Qxd3 e6 (preventing the pawn push f4-f5) 13 Be3 Bd6 14
0-0 Ne7 15 Qb5+ Kf8 16 Rae1 h5 17 Bc1 a6 18 Qd3 b5 19 Kh1 Kg7 and White had no active plan in
A.Panjkovic-I.Khenkin, Ortisei 2013.
b) 8 0-0 is perhaps a more serious try. I believe Black should respond in the most energetic way: 8 ...
e5! 9 dxe5 Nxe5 10 Nd4 Nxd3 11 Qxd3 Bd6 (the downside to playing an early f2-f3 begins to show) 12
f4 Ne7 13 Be3 0-0 14 Nd2 Rfe8 15 g3 Bc5 16 N2f3 Nf5 17 Bf2 Nd6 and Black was much better in
J.Blokhuis-F.Kroeze, Dutch League 2007.
c) 8 Bg5 e5! 9 Nd2 Bd6 10 Nb3 Nge7 11 Qd2 f6 12 Bh4 0-0 13 Bg3 a5 was slightly better for Black
in C.Cozianu-J.Benjamin, US Internet League 2010.
d) 8 Nd2 is met with 8 ... e5!.

C1) 8 Na3
 This has the rather obvious idea of playing Nb5 as well as of transferring the knight to e3 via c2.
8 ... a6 9 Nc2
Alternatively:
a) If 9 Qb3 Black should strike immediately with 9 ... e5! 10 Qxd5 Be6 11 Qe4 Nf6 12 Qh4 exd4! 13
cxd4 (13 Nxd4 is inferior on account of 13 ... Nxd4 14 Qxd4 Rd8! and White has problems defending his
pieces; here 15 Qa4+ is harmless because of 15 ... b5 16 Nxb5 axb5 17 Bxb5+ Bd7 18 Bxd7+ Rxd7 when
the pawns are not worth the piece) 13 ... Bb4+ 14 Kf2 Qd8 (with the threat of winning the queen with ...
Ng4+) 15 Qg3 Nxd4 and now 16 Qxg7 is met with 16 ... Rg8.
b) 9 0-0 is the only move that dissuades Black from playing his planned ... e7-e5. Now I like 9 ... Nf6
10 Re1 (10 Bg5 should be met with normal development and 10 ... e6, and 10 Bf4 with the standard 10 ...
e5!) 10 ... e6 11 Bg5 Bd6 12 g3 Rc8. It may seem that White has gained a huge concession from Black in
trapping the c8-bishop behind the e6-pawn, but White has weakened his king and has no natural piece
coordination or plan. Indeed, 13 Nc2 (if 13 Bxf6 gxf6 14 Nc2 h5 and Black will soon start an attack on
the kingside) 13 ... h6 14 Bxf6 gxf6 is slightly better for Black. His monarch is safe in the centre and he
has a ready-made attack with ... h5-h4.
9 ... e5!
 Energetic and effective.
10 dxe5
After 10 0-0 Bd6 11 g3 Nge7 12 Ne3 0-0 Black had a nicely centralized position in A.Kiriakopoulos-
I.Simeonidis, Kallithea 2006. He will put the two rooks on the d- and e-files and continue putting
pressure on the d4-pawn.
10 ... Nxe5 11 Ncd4 Bd6 12 Nf5
This looks very tempting, but fails to trouble Black. That said, 12 Bf5 Ne7 13 Bxd7+ Qxd7 14 0-0 0-0
15 h3?! (weakening the dark squares) 15 ... Bc5 16 Kh1 Rfe8 17 f4?! Nc4 is good for Black, J.West-
J.Fang, Parsippany 2008.
12 ... Nxd3+ 13 Qxd3 Bxf5 14 Qxf5 Ne7 15 Qg4

15 ... Ng6!
This is designed to prevent White from exchanging pieces with Bf4. Instead, 15 ... 0-0 16 Bf4 was
equal in S.Khabarov-D.Svetushkin, Paleochora 2011.
16 f4 0-0 17 0-0 Rae8 18 Ng3 f5!
Striking out as after 19 Nxf5?? Qc8 20 Ne3 Qxg4 21 Nxg4 Bc5+ 22 Kh1 h5! White loses material.

C2) 8 Bf4
 Obvious, but Black’s play is now very energetic and, indeed, White must be careful not to emerge
from the opening somewhat worse.
8 ... e5!
A good answer. This line has been played by several grandmasters, amongst them world-class players
Alexei Dreev and Liviu-Dieter Nisipeanu. Now we have:

C21) 9 Bg3?!
C22) 9 dxe5

C21) 9 Bg3?!
A little too timid.
9 ... Bd6

10 Na3
After 10 Nd2 Nf6 11 Qc2 exd4! already Black is better because of White having played f2-f3.
10 ... a6! 11 Nc2 Nge7 12 0-0 h5!?
With the threat of a possible ... h5-h4, exposing a weakness on the h2-b8 diagonal.
 12 ... 0-0 is a safer method of handling the position when Black does not need to fear an IQP because
of the weaknesses created by f2-f3 and the open access down the e- and c-files for his rooks.
13 dxe5 Nxe5 14 Bxe5 Bxe5 15 f4 Bd6 16 Ncd4 h4 17 Kh1 Rh6 18 b4?
Positionally very suspect as this will allow Black to secure the c4-square with a well-timed ... b7-b5.
18 ... Kf8!

Getting the king to safety. Black now has plans of controlling the e-file.
19 Qd2 Re8 20 f5 Ng8 21 Nf4 Nf6 22 Rac1 b5!
Black soon won in B.Gonzalez-K.Sasikiran, Mallorca Olympiad 2004.

C22) 9 dxe5 Nxe5 10 0-0

With 10 Bc2 White takes away the c2-square from the knight on b1. This means it is safe to develop the
g8-knight to f6, as Na3-c2-e3-f5 is no longer an issue. After 10 ... Nf6 White cannot just play on the usual
IQP rules of swapping minor pieces because he has weakened his kingside and Black has very active
pieces. These two factors trump the weak d5-pawn. Here White might try:
a) 11 Ba4 Bd6 12 Bxd7+ Qxd7 13 0-0 0-0 and now the developing move 14 Nd2? is bad after the
obvious 14 ... Bc5+ 15 Nd4 Nd3 when Black wins material.
b) 11 Bb3 is a futile attempt at winning the d5-pawn: 11 ... Bc5 12 Bxd5 0-0-0 and Black is much
better.
c) 11 0-0 Bd6 12 Ba4 Bc6 13 Nd2 0-0 14 Bg3 Nh5 and Black is much better.
d) 11 Nd2 Bd6 12 0-0 0-0 13 Nd4 Nh5! 14 Bg5? (the h2-pawn now becomes incredibly vulnerable)
14 ... h6 15 Bh4 Ng6 16 Bxg6 Bxh2+ 17 Kh1 fxg6 and Black had won material in M.Passchyn-G.De
Schampheleire, Geraardsbergen 2013.
10 ... Bd6
 Now we have a final divide:

C221) 11 Re1
C222) 11 Kh1

Other alternatives are:

a) After 11 Nd4 Black must cover the f5-square with 11 ... Ne7! 12 Na3 (if 12 Nb5 Qb6+!) 12 ... a6
13 Nac2 0-0 14 Ne3 Rfe8! (all very natural moves so far) 15 Qc2 h6 16 Kh1 Qb6 17 Rad1 Nxd3 18
Bxd6 Nxb2 19 Bxe7 Nxd1 20 Qxd1 Rxe7 21 Nxd5 Qd8 by when he has all the positional trumps, with the
better pawn structure and minor piece.
b) 11 Na3 a6 and then:
b1) 12 c4? immediately challenges the centre, but this is just plain wrong: 12 ... dxc4! 13 Bxe5 Bxe5
14 Nxc4 Bxh2+! (14 ... Nf6 15 Nxe5 wasn’t so accurate in M.Ragger-A.Rasmussen, Beijing (rapid)
2008) 15 Kh1 b5 16 Ne3 Ne7 17 f4 Qb6 and Black swings the queen over to h6 to defend the imprisoned
bishop.
b2) 12 Bxe5 Bxe5 13 f4 Bd6 14 Nc2 Nf6 15 Ncd4 0-0! (15 ... 0-0-0 was more double-edged in
T.Galinsky-J.Houska, Oslo 2012) 16 Nf5 Bxf5 17 Bxf5 g6 18 Bc2 Qb6+ and Black is much better.
b3) 12 Nc2 Ne7! (the knight is well placed here and very importantly covers the f5-square) 13 Ncd4
0-0 14 Re1 Rfe8 15 Bg3 Rad8 16 Qc2 N7g6 17 b3 Nxd3 18 Qxd3 h5 19 Bxd6 Qxd6 and again Black
doesn’t have any problems; he has good pieces to offset the weakness of the d-pawn.

C221) 11 Re1 Nf6!? 12 Bxe5 Bxe5

 13 f4 Bd6 14 Nd4+ Kf8 15 g3 h5!
From apparently nowhere Black begins an aggressive operation, but if we investigate the position
deeper we can see that this is not so surprising as the barrier of pawns protecting the king have advanced
to vulnerable squares.
16 Nd2 h4 17 Nf1 hxg3 18 hxg3 g6 19 Qf3 Rh3 20 f5 Kg7 21 Kg2 g5 22 Qe3 Rh5
Black had a winning attack in I.Saric-T.Polak, Split 2005.

C222) 11 Kh1

This is a relatively normal position. Black should simply respond to White’s threats and prioritize
getting his king to safety.
11 ... Ne7 12 Bc2
Instead, 12 Na3 a6 13 Nc2 0-0 14 Ned4 Rac8 15 Ne3 Nxd3 16 Bxd6 Nxb2 17 Qb3 Qxd6 18 Qxb2 b5
was equal in A.Semeniuk-A.Dreev, Russian Team Championship 1999, and after 12 Bxe5 Bxe5 13 Nd2
0-0 14 f4 Bf6 15 Nf3 Rad8 16 Nfd4 Rfe8 17 Qc2 h6 18 Rae1 Nc6 there was too much pressure on the d4
point in A.Turzo-L.Eperjesi, Budapest 2010.
12 ... 0-0
 Black should keep things simple.
13 Na3
If 13 Nd2 Black should simply put the rooks on the open files and look for opportunities to place a
knight on a good post or push ... h5-h4: for example, 13 ... Rfe8 14 Re1 N7g6 15 Bg3 Rad8 16 Nd4 (16
Nb3 allows 16 ... Nc4) 16 ... a6 17 Bf5 h5 18 Bxg6 Nxg6 19 Bxd6 Qxd6 20 N2b3 Ne5 and a knight is
bound for c4.
13 ... N7g6 14 Bg3
14 Qxd5 Bc6 15 Qb3 is rebuffed by 15 ... Nxf4 16 Nxf4 Nxf3!.
14 ... Rfe8!?

14 ... Qa5 is also possible.

15 Bb3
White really had to try the greedy 15 Qxd5.
15 ... Rad8 16 Re1 h5 17 Bxd5? h4 18 Bf2 Bxa3 19 bxa3 Bc6
Black won material in P.Dukaczewski-S.Ansell, London 2010.

Summary
The Exchange Variation is characterized by a simple pawn structure, but the piece play is a little bit more
complex. With 5 ... Qc7 we thwart one of White’s main goals, namely controlling the e5-square. White
can ignore the problem, deprive the c8-bishop of a square, or try to evict the queen from c7.
Ignoring the problem simply gives Black an easy game. The only important features to remember are 6
Bg5 Nf6 7 Bxf6 gxf6 8 Qf3 e5! and how to neutralize a queen and bishop combination on the b1-h7
diagonal with ... Bg4-h5-g6. Black should then play for a minority attack and mobilize his pieces to their
best squares.
Depriving the light-squared bishop of a square with h2-h3 is only a problem if you don’t know you
should fianchetto in order to support ... Bf5. Remember that White must try and control e5 otherwise we
will play ... e7-e5.
Kicking the queen away is the most precarious strategy. It is risky because it may allow Black to engage
in active operations with our favourite break ... e7-e5.
Chapter Nine
The Advance Variation: Introduction to 3 ... c5
1 e4 c6 2 d4 d5 3 e5 c5!?

In the past the Advance Variation was considered to be a timid variation that allowed Black to obtain
a ‘good’ version of the French Advance, a good version because the light-squared bishop was able to
develop outside the pawn chain. This assessment has dramatically changed in recent years and now the
Advance is one of the most popular lines against the Caro-Kann due to White’s natural space advantage.
New ideas keep sprouting at a fast and furious rate and in the 3 ... Bf5 line Black must be prepared to
meet many lines, from the hyper-aggressive 4 Nc3 and 4 h4 to the more sophisticated Short system with 4
Nf3.
All these lines are highly theoretical and in order to escape the theoretical battles I will again focus
on my original recommendation 3 ... c5!?. Leading practitioners of this line are GMs Igor Khenkin, Keith
Arkell and Jonathan Hawkins. Our approach is designed to challenge White immediately and take him out
of the mentality of building up a slow advantage with a stable centre.
How does White usually react?
a) By taking the pawn and accepting the challenge (4 dxc5 followed by 5 Be3), as covered in this
chapter.
b) By taking the pawn and then simply developing, as we’ll see in the next chapter.
c) With a quick counter-strike in the centre with c2-c4, as we’ll come across in Chapter Eleven.
d) By shoring up d4 with 4 c3, which is also covered in Chapter Eleven.
Wait one second, I hear you shout, hasn’t Black just wasted a tempo by first playing ... c7-c6 and then
gone ... c6-c5, so surely this move cannot be good?
Yes, it is true that Black has ‘wasted’ a move, but in playing 3 e5 White has determined the pawn
structure and with 3 ... c5, a move that Black would naturally play anyway, all Black has done is speed
things up.
The advantage of playing ... c6-c5 as early as move three is that White must take critical decisions
very early on, which is not something an unprepared white player will be ready for. Accepting the offered
pawn will swing the game into complications where Black has the superior pawn structure and White’s
weakened e5-pawn can become a target.

The Challenge Accepted

4 dxc5

Taking the pawn is pretty complicated variation, in particular when Black is happy to play a pawn
down for an attack. One thing I want to stress in this particular variation is that we have a recurring theme
of undermining White’s pawn structure, rather than attacking the weakened pawns with pieces.
4 ... e6
I would also recommend the reader investigate the complex 4 ... Nc6!?. The resulting play is again
very double-edged, but Black can develop his bishop to g4.
5 Be3

The major line and the most logical choice. White protects the c5-pawn and develops a piece at the
same time.
Note that 5 b4 doesn’t hold on to the pawn after 5 ... Nc6 6 c3 Nxe5! and that White’s other fifth
moves will be seen in our next chapter.
 5 ... Nh6
Why does Black develop a knight to the edge of the board? Well, the knight is heading to f5 from
where it will exert pressure on the bishop on e3. White now has only a limited choice of good moves:

A) 6 Bxh6
B) 6 c3
C) 6 Nf3

Instead, with 6 f4? White is simply not careful enough and 6 ... Nd7! hits the c5-pawn:

a) 7 b4 looks natural, but White is rudely awakened after 7 ... a5! 8 c3 axb4 9 cxb4 b6! 10 Bb5 bxc5
and taking the pawn is impossible because of the loose bishop on b5. After 11 Bc6 Rb8 12 bxc5 Bxc5 13
Bxc5 Qa5+! Black recovers the piece with an excellent position to boot.
b) 7 Bb5 Qa5+ 8 Nc3 a6! 9 Bxd7+ Bxd7 10 Nge2 (or 10 a3 Qc7 11 b4 b6! and Black has a significant
advantage) 10 ... Nf5 11 Bf2 Bxc5 and Black already had the upper hand in M.Nimtz-I.Khenkin, Bad
Wiessee 2000.
Meanwhile 6 Bd3 was played against me very recently, but it is not well thought out: 6 ... Nc6 7 c4?
(7 Bxh6 gxh6 had to be played; Black’s point is that 7 Nf3 Ng4 regains the material and even more) 7 ...
dxc4 8 Be4 Qxd1+ 9 Kxd1 Nxe5 10 Na3 f5 11 Bd4 fxe4 12 Bxe5 Ng4 13 Nxc4 Nxf2+ 14 Ke2 Nxh1 15
b4 and now in E.Kvisvik-J.Houska, Fagernes 2014, 15 ... b6 would have the won game in an easy manner.

A) 6 Bxh6
Our first critical try.
6 ... gxh6
Technically speaking this is the principled line, but it is hardly ever played. In compensation for the
doubled h-pawns Black has the two bishops and will almost certainly recover the lost pawn. In the battle
for activity the doubled h-pawns are not a weakness; in fact they help Black as he now has the half-open
g-file to use.

 7 Nd2
Alternatively:
a) With 7 b4?! White holds on to the extra pawn for dear life, but with no regard for the consequences.
After 7 ... a5!? (Black should commence active operations to open lines and diagonals without delay) 8
c3 axb4 9 cxb4 b6! Black begins to break up the white pawns and his two bishops can be happy:
a1) 10 Nf3 bxc5 11 bxc5 Bxc5 (Black intends to apply heavy pressure on the white king as quickly as
possible) 12 Bd3 Qa5+ 13 Nbd2 Ba6! (now in order to castle White will have to allow some exchanges
to take place, which is something that will benefit Black as the a-pawn is more of a liability than a
strength) 14 Qe2 Qc3! 15 Bb5+ Nd7 16 0-0 Bxf2+!? 17 Qxf2 Bxb5 18 Rfc1 Qa5 19 Qg3 Qb6+ 20 Kh1
Bd3! 21 Qg7 Ke7 22 Rc3 and Black has no problems.
a2) 10 Bb5+ Bd7 11 Bxd7+ Nxd7! (hitting the two pawns, c5 and e5) 12 Nf3 bxc5 13 bxc5 Qc7 14 0-
0 Bg7! (a good location for the bishop, eyeing the e5-pawn and at the same time giving some shelter to the
black king for when it goes short) 15 Nbd2 Nxe5 when Black has a great position and it is apparent that
the doubled pawns hardly matter at all.
b) 7 Nf3 Nc6 is a safe line and now if White tries to place Black under pressure with 8 c4, as played
in J.Kaufeld-F.Doettling, Dortmund 2001, we can go 8 ... Bxc5! 9 cxd5 exd5 10 Qc2 (if 10 Nc3 Black
should seize space with 10 ... d4 11 Ne4 Bb4+ 12 Nfd2 0-0 13 f4 Bf5 14 Nf6+ Kh8 15 a3 Be7 16 Nde4
Bxf6 17 Nxf6 d3, preventing White’s bishop from developing) 10 ... Qa5+ 11 Nbd2?! Nb4 12 Qb3 Bf5!.
Black develops his bishop to the most active square and suddenly White must figure out an adequate way
of dealing with the threat of 13 ... Nc2+. After 13 Rc1 Qb6! (now the f2-pawn is a target) 14 Be2 Bxf2+
15 Kf1 Rd8 Black has the better chances.
7 ... Nc6!?
 This is not a normal idea as usually Black rushes to win the c5-pawn with an immediate ... Bxc5.
However, we are after bigger fish on e5. Moreover, taking with 7 ... Bxc5 did not prove so successful for
Black in K.Hansen-J.Andersen, correspondence 2003: 8 Nb3 Bf8 9 f4 Bg7 10 Bd3 0-0 11 Qg4 f6 12 Nf3
Nc6 13 0-0-0 Kh8 and White had the more pleasant position.
8 Ngf3 Bd7
Preventing any Bb5 ideas that may hamper Black’s ideas of piling pressure on the e5-pawn.
9 Bd3 Qc7 10 Qe2 Bg7
Now the e-pawn can no longer be defended.
11 c3 Nxe5 12 Nxe5 Bxe5 13 0-0-0 0-0-0 14 Nb3 Ba4
Black is fine here.

B) 6 c3
Preparation for White’s next.
6 ... Nf5 7 Bd4

Generally speaking Black has a little bit of a formula going on here: play ... Nfxd4 followed by ... b7-
b6, putting pressure on the b2- and d4-pawns.
 When the white bishop is on d3 this formula is especially effective, but I have to stress that we do not
use this formula in the main lines, i.e. when White has prepared the defence of the c5-pawn with a future
b2-b4.
Instead, after 7 Qd2 Nd7 8 Bb5 a6 9 Ba4 Qc7 White cannot hold on to the pawn for too long.
7 ... Bd7!
Let us see what happens if Black employ the formula too early: 7 ... Nxd4?! 8 cxd4 b6 9 Bb5+! Bd7
10 Bxd7+ Qxd7 11 cxb6 axb6 12 Ne2 Bb4+ 13 Nbc3 Nc6 14 0-0 0-0 15 Qd3 and Black has nothing for
the pawn.
8 Nf3
The main line, but let’s first look at those scenarios where the formula works like a charm:
a) 8 g4 Nxd4! 9 cxd4 b6! 10 cxb6 Qxb6 11 Qd2 Nc6 12 Nf3 h5 13 gxh5 Rb8 14 Nc3 (or 14 b3 Bb4
15 Nc3 Rc8) 14 ... Qxb2 15 Rb1 Qxd2+ 16 Kxd2 Rxb1 17 Nxb1 Rxh5 and Black will be playing on for a
long time, grinding out a win.
b) Neither is 8 Bd3 really challenging. Indeed, you can guess the reply: 8 ... Nxd4 9 cxd4 b6! 10 cxb6
Qxb6.

This is our dream position. Black will regain the pawn with the long-term advantage of the two
bishops: for example, 11 Ne2 (if 11 Nf3 Black will simply reply 11 ... Nc6) 11 ... Nc6 12 Nbc3 Nxd4 13
0-0 Nc6 14 Bb5 Nxe5 (14 ... Rc8 15 Nf4 Be7 16 Qg4 g6 is also good enough) 15 Nxd5 Qb7 16 Bxd7+
Nxd7 17 Ndc3 Be7 18 Qd2 and Black was better, but kindly gave White a draw in Y.Zhou-J.Speelman,
British League 2014.
8 ... Nc6
 Now White has:

B1) 9 Be2
B2) 9 Qd2
B3) 9 c3

I must strongly stress that against 9 Qd2 and 9 a3 we are not following the formula. However, for the rest
of White’s 9th moves, we can:
a) 9 Bd3 Nfxd4 10 cxd4 b6 follows the formula.

Now:
a1) After 11 cxb6 Qxb6 the d-pawn must fall: 12 Nc3 Nxd4 13 0-0 Be7 14 Rc1 Nxf3+ 15 Qxf3 was
J.Radlovacki-R.Fontaine, Pancevo 2002, and now Black should play 15 ... Rd8 overprotecting the
bishop, and if 16 Qh3 (or 16 Qg3 0-0 and Black has no problems) 16 ... Qxb2 17 Rc2 Qb8 18 Nb5 Bxb5
19 Rb2 a6 20 a4 Qxe5, which keeps the extra piece.
a2) 11 a3 bxc5 12 dxc5 Bxc5 13 0-0 Nd4 is also promising for Black: for example, 14 Nbd2 a5
(preventing White from expanding on the Queenside) 15 Rc1 Ba7!? (keeping the queen on d8 to defend
the g5-square) 16 Re1 Nxf3+ 17 Nxf3 0-0 with an easy game for Black.
a3) 11 0-0 bxc5 12 dxc5 Bxc5 13 Nc3! (13 Nbd2 allows Black the active 13 ... Qb6 14 Nb3 Be7 15
Rc1 Rc8 16 Qd2 0-0 when he is doing well) 13 ... Nd4 14 Rc1 Nxf3+ 15 Qxf3 Qe7 16 Qg4 (or 16 Bb5
Bxb5 17 Nxb5 0-0 18 Rc2 f6 19 exf6 Rxf6 20 Qd3 Raf8 21 Nd4 Bb6 22 Nf3 g5 23 h3 Qg7 and Black has
a good position with pressure against the f2-pawn) 16 ... 0-0 17 a3 f5 18 exf6 Rxf6 19 b4? Bb6 20 Qh5
Rh6 21 Qe5 Qf7 22 Rc2 Rf8 23 Nd1 Ba4 24 Rd2 Bc7 was an easy win for Black in K.Kisonova-
V.Mitura, High Tatras 2001.
We will now consider two examples where the formula is inverted, which is designed to prevent Rc1
from defending the weak c5-pawn:
b) 9 Na3?! attempts to solidify the centre with Nc2 and is well met by 9 ... b6! (unfortunately the
formula doesn’t work here because 9 ... Nfxd4 10 cxd4 b6 allows 11 Rc1!).

Thus we invert the move order and after 10 cxb6 Nfxd4 11 Nxd4 Nxe5 12 bxa7 (Black will simply
round up the pawn if 12 b7) 12 ... Bxa3 13 bxa3 Qa5 14 Qb3 0-0 the white pawns will drop.
c) Likewise, if 9 Nbd2 b6! 10 cxb6 (or 10 Nb3 Qb8 11 cxb6 Nfxd4 12 Nbxd4 Nxe5 13 bxa7? Nxf3+
14 Nxf3 Qxb2! and suddenly Black is much better) 10 ... Nfxd4 11 Nxd4 Nxd4 12 b7 Rb8 13 cxd4 Qb6
14 Be2 Qxd4 when Black was much better in S.Buckley-B.Thipsay, British Championship, Scarborough
1999.

B1) 9 Be2 Nfxd4 10 cxd4 b6

Once again we can follow our favourite formula.
11 cxb6 Qxb6 12 b3
After 12 Nc3 Qxb2 13 Nb5 Qb4+ 14 Nd2 Qa5 15 0-0 Nxe5 Black is fine and 12 Qd2 doesn’t solve
White’s problems: 12 ... Rb8! 13 b3 Bb4 14 Nc3 Qa5 15 Rc1 and now after 15 ... Ne7 White must return
the material.
12 ... Bb4+ 13 Kf1 0-0 14 g3 f6!
 Thematically opening lines when the white king’s safety is questionable.
15 exf6 Rxf6 16 Kg2 Raf8 17 Rf1
This was A.Abreu-G.Siegel, Havana 1998, and now Black missed:
17 ... Rxf3! 18 Bxf3 Nxd4 19 Bg4 Bb5 20 Nd2!
20 Re1? Rxf2+ 21 Kxf2 Nc2+ 22 Kg2 Nxe1+ wins on the spot.
20 ... Bd3
White is completely tied down.

B2) 9 Qd2

A very important variation particularly in light of the fact that we cannot follow the formula. This is
primarily because White is preparing the defensive pawn push b2-b4. Black should instead play in
another active way:
9 ... f6!
Unfortunately White is well prepared for the formula, as shown by 9 ... Nfxd4 10 cxd4 a5 (10 ... b6 11
b4! is the whole point of 9 Qd2) 11 Nc3 b6 12 cxb6 Qxb6 13 Be2 Rb8 14 Rb1 and Black had very little
for the pawn in J.Cubas-J.Molina, Santos 2003.
 After the text, White must decide what to do about the centre. We consider:

B21) 10 exf6?!
B22) 10 b4

Instead, 10 Bd3 is simply no good and forces us to play the good 10 ... Nfxd4 11 cxd4 fxe5 12 Nxe5 Nxe5
13 dxe5 Bxc5 14 0-0 Qh4 15 Qe2 0-0 by when Black was clearly better since his dark-squared bishop
was very strong in K.Bryzgalin-J.Prizant, Russia 2000.

B21) 10 exf6?!
White is too cautious and gives up the pawn centre.
10 ... gxf6

Black can now look forward to pushing the centre pawns.

11 Be3
Alternatively:
a) 11 Be2?! is rather tame: 11 ... e5! 12 Be3 Nxe3 13 fxe3 (13 Qxe3 had to be played even though
Black wins material with 13 ... Qe7! 14 b4 a5 15 Na3 e4! as 16 Nd4 axb4 17 Nab5 Qe5 18 0-0 bxc3 19
Qxc3 Nxd4 20 Nxd4 Rc8 21 c6 bxc6 22 f4 Qd6 23 Bh5+ Kd8 24 Rab1 Qc5 25 Qe3 Bd6 26 Rb7 Qa3 27
Nb3 remains incredibly complicated) 13 ... Bxc5 14 Qxd5 Qe7 15 Nbd2 0-0-0 16 Ne4 Bxe3 was a big
success for Black in C.Tesik-D.Berczes, Budapest 2012.
b) 11 Bb5 is more purposeful and after 11 ... Qe7! White has:
b1) 12 b4 should be played, although White is vulnerable to attacks on the g-file: 12 ... Rg8 13 0-0 (if
13 g3 e5! when 14 Be3? would be a huge blunder because of 14 ... d4 15 cxd4 exd4! and after and 14 0-0
Bh6 15 Be3 Nxe3 16 fxe3 d4 17 cxd4 exd4 18 Kh1 Bxe3 Black is much better) 13 ... Bh6 14 Be3 and
now Black can play the rather splendid 14 ... Ne5! 15 Bxd7+ Qxd7 16 Nxe5 Qg7! 17 g3 (or 17 Bg5 Bxg5
18 f4 Bxf4 19 Rxf4 fxe5 when Black’s chances are better) 17 ... Bxe3! 18 fxe3 Nxg3 with a deadly attack.
b2) 12 0-0?! e5 13 Re1?! (losing by force; of course, 13 Be3 Nxe3 14 Qxe3 Qxc5 is clearly better for
Black since he has a strong pawn centre and the bishop-pair, but this line was relatively best for White)
13 ... 0-0-0! 14 b4 Qg7 15 Be3 d4! was a decisive central breakthrough. Black won 10 moves later in
P.Smirnov-J.Prizant, Russia 2000.
c) 11 Bd3 e5! 12 Bxf5 Bxf5 13 Be3 Rg8 14 0-0 Qd7 15 g3 Bh3 (White’s light squares are a disaster)
16 Re1 0-0-0 17 b4 d4 18 cxd4 exd4 19 Bf4 Qg4! and Black is winning.
11 ... Qe7
Preparing to attack the c5-pawn after the exchanges.
12 b4
More or less forced if White wishes to keep his material advantage. If 12 Na3 Nxe3! 13 fxe3 Qxc5 14
Bd3 0-0-0 15 0-0 Ne5!.
12 ... a5 13 a3 Nxe3 14 fxe3
14 Qxe3 loses the extra material to 14 ... axb4.
14 ... Bh6 15 Bb5 0-0 16 0-0 axb4

17 cxb4
Now Black has a tactical strike:
17 ... d4! 18 Nxd4 Nxd4 19 Bxd7 Nb3 20 Qc3 Nxa1 21 Bb5 Qc7!?
Or 21 ... b6 22 cxb6 Qd6 and White is struggling here.
22 Qxa1 Bxe3+ 23 Kh1 Qe5 24 Qa2
Although the game ended in a draw, White was battling to fight for equality here in O.Hartig-
P.Dittmar, Gibraltar 2007.

B22) 10 b4
 White chooses to overprotect the extra c-pawn, leaving the central tension intact.
10 ... Nfxd4!
Rightly eliminating the tension in the centre before swapping on d4.
11 cxd4 fxe5 12 dxe5
After 12 Nxe5 Nxe5 13 dxe5 Black can undermine the c5-pawn with the simple 13 ... b6 14 cxb6
Qxb6 when 15 a3 a5 16 b5 Bc5 17 Nc3 0-0 18 Nd1 Bd4 would be a disaster for White.
12 ... Qc7

13 Qb2
Alternatively:
a) 13 Nc3 Nxe5 14 Nd4 Be7 15 Ncb5 Qb8 16 f4 a6! 17 Nd6+ Bxd6 18 cxd6 Nc6 and Black stands
well.
b) 13 Qc3 a5 14 b5 d4 and now to stay in the game White must play 15 b6 Qxb6 16 cxb6 dxc3 17
Nxc3 Bb4 18 Rc1 Rc8 when Black has a lot of pressure for the pawn.
13 ... a6!
An excellent defensive move, preventing the pawn push b4-b5.
14 Be2
 In light of what now happens White may be tempted to play 14 h4 g6 15 h5 g5! 16 Nxg5, but here
Black has the tactic 16 ... Nxb4! 17 Qxb4 Qxe5+ 18 Be2 Qxg5 19 Qxb7 Rc8 20 0-0 Rg8 21 Bf3 Bg7 and
wins.
14 ... g6!

Attacking the vulnerable e5-pawn from another angle.

15 Nbd2 Bg7 16 Qb3
Again, after 16 0-0 Nxe5 again we can be satisfied with the black position.
16 ... 0-0 17 0-0 Nxe5
Black has an excellent position here with the possibility of pushing hard in the centre. An example of
things could go is the following:
18 Rac1 Nc6 19 a4?!
After 19 a3 Rf7 20 Rfd1 Raf8 when ... Ne5 or ... Nd4 is coming, and if 21 Nf1 Ne5 22 Nxe5 Bxe5 23
Bf3 Qd8 followed by ... Qg5 and ... Bb5.
19 ... Rf4 20 b5 Rb4 21 Qe3 axb5 22 axb5 Nd4 23 Nxd4 Bxd4 24 Qg5 Bxc5 25 Nf3 b6 26 Ne5
Bxb5
Black is winning.

B3) 9 a3
 As we have seen, when White prepares the defensive pawn move b2-b4 Black cannot play the
immediate formula of ... Nfxd4 followed by ... b7-b6. As such, I believe Black should renew the threat
with:
9 ... a5!
Preventing b4.
Black can try his luck with 9 ... f6!? 10 exf6 gxf6 11 Ng5 which is possible as White threatens a check
on h5 Black is forced to play 11 ... h5 12 Be2 Ncxd4 13 Bxh5+ Ke7 14 Nf7 Qa5 15 Nxh8, but this may be
a position that only computers are prepared to play.
10 Be2
Important alternatives are:
a) Should White play the positional 10 a4!?, securing the b5-square for the bishop, Black should
respond carefully with 10 ... Ncxd4! (the problem is that after the natural 10 ... Nfxd4 11 cxd4 b6 12 cxb6
Qxb6 13 Bb5 there is no clear-cut way of winning the pawn back; indeed, after 13 ... Na7? 14 Bxd7+
Kxd7 15 0-0 Qxb2 16 Nbd2 Qb6 Black was in serious trouble in T.Gelashvili-T.Woodward, Guernsey
2002) 11 cxd4 b6. Now that the b5-square is covered, Black can open lines:

a1) Holding on to the pawn with 12 cxb6?! allows 12 ... Qxb6 13 Qd2 Rb8! (the key to the position is
not to let White consolidate his extra pawn) 14 Ra2 Bb4 15 Nc3 Bxc3 16 Qxc3 0-0 17 Qd2 Qb3 and the
pawn falls anyhow.
a2) 12 Qc2 doesn’t hold on to the extra pawn because of 12 ... bxc5 13 dxc5 Rc8, which looks great
for Black.
a3) The counterattacking 12 g4 is met with 12 ... Nh6 13 cxb6 Qxb6 and again White has to defend
two weak pawns.
a4) 12 Bd3!? g6 13 Bxf5 (13 g4 again weakens the kingside and after 13 ... Nh4 14 Nxh4 Qxh4 15
cxb6 Rb8 16 Nc3 Rxb6 White has a serious problem with where to put his king) 13 ... gxf5 14 cxb6 Qxb6
and Black has massive compensation for the pawn on the open b- and c-lines: for example, 15 Qe2!?
Bb4+ 16 Nc3 Bc8 17 Qb5+ Qxb5 18 axb5 Bd7 19 0-0 Bxc3 20 bxc3 Bxb5 21 Rfb1 Bd3 22 Rb7 0-0 and
Black is better on account of the passed a-pawn.
b) 10 Bd3 is another option to which we should reply with the formulaic 10 ... Nfxd4 11 cxd4 b6.
Again we see the formula in action, but with the insertion of the moves ... a7-a5 and a2-a3. It is not
entirely clear who this benefits, but I see little to fear:
b1) 12 cxb6 Qxb6 13 Nc3 Nxd4 14 0-0 Be7 and Black has recovered the pawn.
b2) 12 0-0 bxc5 13 dxc5 Bxc5 14 Nc3 Nd4 is also very principled and after 15 Nxd4 Bxd4 16 Qe2
Qc7 17 Rfe1 Rb8 18 a4 (securing the b5-square as compensation for the bishop-pair and not 18 Nb5?
Bxf2+!) 18 ... Bxc3 19 bxc3 0-0 again Black stands well.
b3) 12 Qc2 bxc5 13 dxc5 Qe7 (playing to win back the pawn) 14 Nc3 Qxc5 15 0-0 Rb8 (otherwise
b2-b4 is annoying) 16 Rac1 Be7 17 Qe2 Qb6 18 Na4 and now Black has the tricky 18 ... Nd4 19 Nxd4
Qxd4 with zero problems.
10 ... Be7 11 0-0
11 g4 simply leaves White with a gaping hole on the kingside: 11 ... Nfxd4 12 cxd4 b6 13 cxb6 Qxb6
14 Nc3 0-0 15 0-0 f6 and Black can be happy with the long-term play on the f-line.
11 ... 0-0 12 Qd3
If 12 Bd3 now is the time to push the a4-pawn with 12 ... a4.

Black aims to obtain long-term compensation for when we play ... b7-b6. Play may develop 13 Bxf5
exf5 14 Qd3 f4 (a multi-dimensional move, releasing the d7-bishop and also tying the bishop to d4; from
now on it becomes tricky for White to think up a plan) 15 Nbd2 (or 15 Qd2 Na5 16 Qxf4 Nb3 17 Ra2
Bb5 18 Rd1 Be2 19 Re1 Bc4 20 Be3 and the white rook is trapped on a2) 15 ... Qc8 16 Rfe1 Bf5 17 Qe2
Re8 18 c4 (say, White plays 18 Rac1 Qe6 then I am not sure what he is doing, whereas Black has several
plans of expansion: for instance, ... h7-h5 may come to mind or ... Ra5, winning back the c5-pawn) 18 ...
Qd7 19 Rad1 Rad8 20 cxd5 Qxd5 21 Nc4 Bg4 and White is struggling.
White also has to be careful that Black can’t able to slip into old ways. Indeed, after 12 Nbd2 the
defence of the d4-pawn is interrupted so Black can play: 12 ... b6! 13 cxb6 Ncxd4 14 Nxd4 Nxd4 15 cxd4
Qxb6 16 Nf3 Qxb2 regaining the pawn.
Instead, 12 b3 is designed to prevent ... a5-a4, but allows Black to play 12 ... f6! 13 Bd3 fxe5 14
Nxe5 Nfxd4 15 cxd4 Nxe5 16 dxe5 g6 (don’t fall for 16 ... Bxc5? 17 Bxh7+! Kxh7 18 Qc2+ Rf5 19 Qxc5
and Black hasn’t equalized) 17 Qc2 Kg7 18 Nd2 Rc8 19 b4 b6 20 Ba6 Rc7 and the c-pawn will fall.
12 ... a4! 13 Nbd2

Here Black should play:

13 ... b6!
A less accurate approach is 13 ... Na5 14 h4 Rc8 15 h5 Bxc5 16 Bxc5 Rxc5 17 g4 Nh6 18 Nd4
(F.Berkes-P.Bodiroga, Zadar 2010) 18 ... f5! 19 exf6 Qxf6 20 Qg3 e5 21 g5 Nf5 with a mess on the board.
14 cxb6
14 b4 is another possibility and after 14 ... axb3 15 Nxb3 Na5! 16 Nxa5 Rxa5 17 cxb6 Nxd4 18 cxd4
Qxb6 19 Rfb1 Qa7 it will only be a matter of time before the a-pawn is rounded up.
14 ... Nfxd4 15 cxd4 Qxb6 16 Rab1 Rfc8 17 Rfc1
After 17 b4 axb3 18 Rxb3 Qd8 19 Rc3 Qf8 the pressure on a3 is too much for White to bear.
17 ... Na5 18 Rxc8+ Rxc8 19 g3
White’s extra pawn is not worth too much as Black is simply too active.
19 ... Qc7 20 Bd1 Qb6 21 Be2 Qc7
With an interesting repetition.

C) 6 Nf3
 A tricky move. White returns the c-pawn and lets Black get in ... Nh6-f5 for free, but says to Black,
what are you doing with your bishop on c8? This approach does, however, look suspicious to me. Indeed,
I think if White wants to play with this reasoning then he should play 5 Nf3 directly.
6 ... Nf5
A less popular move than 6 ... Nd7.
7 Bg5!
Of course, the best. Let’s just have a quick look at the alternatives:
a) 7 Bd4?! just looks dubious to me: 7 ... Bd7 (7 ... Nxd4 is, of course, fully playable still) 8 c3 Nc6
and now White must take precautions to prepare b2-b4 and not go 9 Bd3?! Nfxd4 10 cxd4 b6 11 0-0 bxc5
12 dxc5 Bxc5 13 Nc3 Be7 14 a3 a5 15 Rc1 0-0 16 Bb1 Qb8 17 Qd3 g6 18 Qe2 Rd8 19 Bd3 Bf8 20 h4
Bg7 21 Rfe1 h5 with an instructive and good game for Black in R.Polzin-F.Vallejo Pons, Hamburg 2010.
b) 7 Bb5+ does not make so much sense since White exchanges Black’s worst piece on the board: 7 ...
Bd7 8 Bxd7+ Nxd7 9 Bg5 Qc7 and Black will win back one of his pawns with equality.
c) 7 Qd2 Nxe3 8 Qxe3 Qa5+ 9 c3 Bxc5 10 Qd2 Nc6 11 Bd3 Qc7 12 Qe2 0-0 and the Greek gift
doesn’t work for White and if 13 0-0 (not 13 Bxh7+? Kxh7 14 Ng5+ Kg8 15 Qh5 Qxe5+) 13 ... f6! White
doesn’t really have enough control over the e5-square, as shown by 14 exf6 Rxf6 15 Nbd2 Bd7 16 Nb3
Bd6 17 Nbd4 a6. Indeed, now 18 Nxc6? would be a big mistake as shown by 18 ... Bxc6 19 Qc2 (19
Rfe1 fails to 19 ... e5 20 Nxe5 Re8 21 Bxh7+ Kxh7 22 Qh5+ Rh6 and Black is winning) when Black can
play the wonderful 19 ... Qf7!, ignoring h7 as he is threatening to sacrifice on f3.
7 ... Qc7!?
7 ... Qa5+ was my original suggestion, but I have changed my mind as I now believe we should
capture on c5 with the bishop.
 8 c3
White can also play:
a) 8 Bd3 is met by the bold 8 ... Bxc5! 9 Bxf5 exf5:
a1) 10 Qxd5 is just too risky. Black obtains lots of play and an open position for the pawn with 10 ...
Be6 11 Qd3 0-0 12 Nc3 (12 0-0 Bc4 is a nice little trick that nets an exchange) 12 ... Nc6, winning back
the material.
a2) A better path for White is to develop with 10 0-0 Be6 11 Nc3 h6 12 Bf4 Qd7 13 Qe2 Nc6 14
Rad1 0-0 15 Be3 Be7 when Black will continue with the straightforward plan of putting his rooks on the
c- and d-files.
b) 8 g4 is crazy, but let’s check it anyway: 8 ... Ne7 9 b4 (or 9 c4 Ng6 10 cxd5 Nxe5 11 Bf4 Nd3+ 12
Qxd3 Qxf4 13 Nbd2 exd5 14 Qxd5 Nc6 15 0-0-0 Be6 16 Re1 Be7 when Black has magnificent
compensation, in fact so much so that it would be prudent for White to play 17 Rxe6 fxe6 18 Qxe6 Qf6 19
Qe4 0-0) 9 ... Nd7! 10 Bb5 a5! (undermining) 11 c3 Nc6 12 Bxc6 bxc6 13 Bf4 g5 14 Bg3 h5 and ... g5-g4
is in the air.
c) 8 b4?! a5! 9 c3 axb4 10 cxb4 b6! would be an idyllic undermining.
8 ... Bxc5 9 Bd3 h6!
I am advocating such a radical approach because of the time White has lost in moving his dark-
squared bishop around the board.
10 Bc1
White has undeveloped his bishop which can only be good news for Black. We can now revert back
to the traditional plans of attacking the e5-pawn and undermining it with ... f7-f6.
My point is that after 10 Bf4 I want to play the radical 10 ... g5!.
 After 11 Bc1 (or 11 Bg3 Nc6 12 0-0 Nxg3 13 hxg3 g4 14 Nh4 Nxe5 15 Bb5+ Bd7 16 Bxd7+ Qxd7
and Black will castle queenside and to safety) 11 ... Nc6 12 0-0 (if 12 Bxf5 exf5 13 Qxd5 Be7 14 0-0
Be6 15 Qd3 Rd8 16 Qe2 g4 17 Ne1 Qxe5 18 Qc2 Bc4 when Black is incredibly active and White is,
well, not) 12 ... g4! 13 Nfd2 (13 Bxf5? gxf3 wins) 13 ... Qxe5 14 Qxg4 Ne3 15 fxe3 Qxe3+ 16 Kh1 Qxd3
17 Qg7? Rh7 (excellent defence along the seventh rank) 18 Qg8+ Bf8 not only is Black completely safe,
but it is White who will have to worry.
10 ... Nc6 11 Bxf5
One of three important options:
a) When White chooses to play the defensive 11 Qe2 Black should respond with 11 ... 0-0 12 0-0 (12
g4 is met with 12 ... Nfe7 and the knight will reroute itself to g6) 12 ... f6!:
a1) 13 exf6 Rxf6 14 b4 (14 Re1? e5! 15 Nxe5 Ng3! wins) 14 ... Bd6 15 Bb2 Rf8 16 b5 Ne5 17 Nxe5
Bxe5 18 g3 Bd7 19 Nd2 Bf6 and ... e6-e5 is coming.
a2) 13 Bxf5 exf5 14 exf6 Rxf6 15 Be3 Bd6 16 Nbd2 Bd7 17 Qd3 f4 18 Bd4 Re6 19 Rfe1 Rae8 and
Black is holding the balance. The bishop on d4 cannot move and the knight on d2 has no role in the game,
as shown by 20 Nb3 b6!.
b) 11 0-0 allows the greedy 11 ... Nxe5 12 Bxf5 Nxf3+ 13 Qxf3 exf5 14 Bf4 Qc6 15 Re1+ Be6 16
Nd2 0-0 17 Qg3 Kh7. White, of course, has some compensation as the extra pawn is slightly meaningless
for the moment, but Black has a solid position and the easy plan of centralizing the two rooks.
11 ... exf5
 12 0-0
If White grabs with 12 Qxd5 then after 12 ... b6 13 0-0 0-0 14 Re1 Rd8 15 Qb3 (15 Qc4 falls foul of
the stinker 15 ... Nb4!) 15 ... Be6 16 Qc2 b5 17 a4 (17 Nbd2 and it’s the same trick again: 17 ... Nb4;
instead, if 17 Bf4 g5 18 Bc1 Rac8 and White can’t develop or 17 Na3 b4 18 Nb5 Qb6 19 Nbd4 bxc3 20
bxc3 Nxd4 21 cxd4 Bxd4 22 Nxd4 Rac8 23 Qb2 Qxd4 and Black has the upper hand in the endgame) 17
... b4 18 Be3 Bxe3 19 Rxe3 Rac8 he is struggling to develop. Now 20 Nbd2 f4 21 Rd3 (or 21 Ree1 bxc3
22 Qxc3 Qb7 and the white queen is caught in the open) 21 ... Nxe5 22 Nxe5 Qxe5 regains the pawn.
12 ... Be6!
For the time being the bishop can settle here, defending both f5 and d5. In the future we will looking to
push ... d5-d4.
13 Re1
Blockading with 13 Nd4 can be met by 12 ... 0-0 14 Be3 Nxe5 15 Bf4 g5 16 Bxe5 Qxe5 17 Re1 Qd6
18 Qd3 Qa6 19 Qxa6 bxa6 20 Nxe6 Rfe8 21 Nd2 fxe6! with a wonderful endgame, and if 13 b4 Black
can simply retreat the bishop: 13 ... Be7 14 Re1 0-0 15 a4 Rac8 as 16 Na3? allows 16 ... d4! undermining
the b4-pawn.
13 ... 0-0
 14 Be3
Alternatively, 14 Nbd2 Qb6 15 Re2 d4 16 Nb3 dxc3 17 Nxc5 Qxc5 and the position is favourable for
Black, as is 14 Nd4 Qb6 15 Be3 f4! 16 Bxf4 Qxb2.
14 ... Bxe3 15 Rxe3 Rfd8 16 Na3 f4 17 Re1 Qb6 18 Qd2?! d4
All our weaknesses have now been eliminated.
19 cxd4 Nxd4 20 Qxf4 Nxf3+ 21 Qxf3 Qxb2
Black is doing very well as the bishop rules over the white knight on a3.

Summary
This chapter dealt with the scenario after 3 ... c5 where White captures the pawn and defends it. The
resulting positions are tense and the first point to remember is the concept of undermining White’s pawn
centre, breaking open the position so that Black’s pieces can come into play.
The formula of ... Nxd4; cxd4 b6 comes into play if White makes no effort to defend the weak c5-pawn.
Please note too that the formula may be inverted to avoid Rc1 ideas. However, Black must be aware that
should White play 9 Qd2 or 9 a3 then the formula does not work, and we must change approach, such as
with ... f7-f6.
Chapter Ten
The Advance Variation: White Returns the c5-pawn
1 e4 c6 2 d4 d5 3 e5 c5 4 dxc5 e6

Black’s fourth move, 4 ... e6, signalled his intention to win back the c5-pawn, but unfortunately it also
did one other thing - it locked out the bishop on c8. For French players who are used to this it’s not such a
huge problem, but for me it was something I initially struggled with. If only the bishop could move through
pawns ...
White’s big plan, regardless of which formation he chooses, is primarily to defend the e5-pawn and
focus on development. After ... f7-f6 has challenged e5 then his plan is to grip the e5-square tightly,
leaving Black in an uncomfortable hold. A more active plan, which is usually seen in the a2-a3 and b2-b4
formations, is to challenge d5 with c2-c4. White may play this a3 and b4 plan as early as move 5, but also
he can choose to play it slightly later.
To make the material easily digestible, I have again come up with some white piece formations that
we will focus on after each of:

A) 5 Nf3
B) 5 a3
C) 5 Qg4
D) 5 Nc3

Instead, 5 Bf4 can easily transpose into Line A after 5 ... Bxc5: for example, 6 Bd3 Nc6 7 Nf3 Nge7 8 0-0
is Line A3.
Likewise, 5 Bd3 Nc6 6 Nf3 transposes to Line A.

A) 5 Nf3
 5 ... Bxc5 6 Bd3
6 Be2 simply interferes with the future defence of the e5-pawn and is well met by 6 ... Nc6 7 0-0
Nge7.
6 ... Nc6 7 0-0 Nge7
White now has to take a decision on which plan to follow.

The Battle over the e5-square

The move 4 dxc5 created one major weakness, namely the e5-pawn, which is out there alone on the edge
of enemy territory. It is Black’s fundamental plan to:
a) Win the pawn, or
b) Play ... f7-f6, challenging the pawn, upon which the next stage is to release the light-squared bishop
from its cage with ... e6-e5.
I believe that White must play accurately to deal with Black’s plan.

We shall now consider what happens should White play:

A1) 8 c3
A2) 8 Nbd2
A3) 8 Bf4!?

A1) 8 c3 Ng6!

The bishop on c1 is denied the opportunity to arrive on f4 which means that when the pawn break ...
f7-f6 arrives White will have less control over the critical e5-square. The general theme in this variation
is simply to develop as much as possible and then strike with ... f7-f6.
9 Qe2?!
Not the best choice as now White won’t be able to develop the knight to d2 because ... Nf4 would be
an unpleasant shock.
Instead, after 9 Re1 0-0 10 Be3 Be7 11 Bd2 Bd7 12 a4 f6! (as in the main line, ... f7-f6 comes once
Black has completed his development) 13 exf6 Bxf6 14 Na3 e5 15 Qb3 Bg4 16 Be4 (T.Thorhallsson-
H.Olafsson, Reykjavik 2014) 16 ... Nge7! 17 Rad1 Qd7 18 Be3 Rad8 19 c4 d4 Black has a very
comfortable position.
9 ... 0-0

10 Bg5
 One of several possibilities:
a) 10 b4 Bb6 11 a4 and now Black should play 11 ... f6! (11 ... Bc7 12 Bxg6 fxg6! is also a
possibility, but the text is better). As mentioned, White doesn’t have too much control over the e5-square
which makes the pawn push ideal here: for example, 12 Bxg6 (or 12 exf6 Qxf6 13 Bg5 Nf4! 14 Bxf4 Qxf4
and should White take measures to seize control over the e5-square with 15 b5 Ne7 16 Ne5? Bc7 17 Nf3
e5 then Black is much better) 12 ... hxg6 13 exf6 (13 Be3 g5 14 Nbd2 Bxe3 15 Qxe3 g4 wins a pawn) 13
... gxf6! 14 Be3 (after 14 b5 Na5 15 Bh6 Re8 Black’s kingside is perfectly safe and his pawn centre is
wonderfully mobile) 14 ... e5! 15 Nbd2 Kg7 16 Rfd1 Be6 17 Bxb6 axb6 18 Nf1 Qc7 19 h3 Rfe8 20 Ne3
Rad8 and Black had few problems in M.Savic-T.Nabaty, Belgrade 2012. He can attack down the h-line or
target the weak c3-pawn.
b) 10 c4 dxc4 11 Bxc4 Qc7 12 Re1 Nd4! and the e5-pawn is incredibly weak.
c) Should White wait with a few ‘improving’ moves before deciding on a plan, then Black should do
the same: 10 Re1 Bd7! 11 Na3 a6 (preventing a knight hopping into b5) 12 Nc2 f6! 13 exf6 Qxf6 14 Bg5
Qf7 15 Rad1 h6 16 Be3 Bd6 and Black will play for either ... Nf4 or ... Nge5.
10 ... Qc7 11 b4
After 11 Re1 Bd7 12 b4 (if 12 Bxg6 fxg6! or 12 Nbd2 h6!) 12 ... Be7 13 Bxe7 Ncxe7 either the e-
pawn or the backward c-pawn will give White a headache.
11 ... Bb6

12 Re1?
An instructive mistake. Also worth studying is the line 12 a4 a6 13 Re1 Ncxe5 14 Nxe5 Nxe5 15
Qxe5 Qxe5 16 Rxe5 f6! 17 Be3 d4 18 Bxd4 Bxd4 19 cxd4 fxe5 20 dxe5 Rf4 and Black is winning.
12 ... Ncxe5!! 13 Nxe5 Qxe5 14 Qxe5 Nxe5 15 Rxe5 f6 16 Bxf6 Rxf6 17 Re2 e5 18 Rxe5 Rxf2
19 Kh1 Bd7 20 Rxd5 Bc6
Black picked up an exchange and won quickly won in E.Muratovic-E.Dizdarevic, Zenica 2012.

A2) 8 Nbd2
 An interesting reply, but White must be super careful that his e5-pawn doesn’t simply drop off,
particularly in light of the fact that Black’s plan will be to attack it.
8 ... Ng6!
The vulnerable e5-pawn is attacked and f4 taken away from the white bishop.
9 Nb3
White develops the knight with tempo.
Instead, 9 Bxg6 gives up the light-squared bishop too easily: 9 ... hxg6 10 Nb3 Be7 11 Qe2 b6 and
Black’s bad bishop will soon be released.
9 ... Bb6
With the intention of routing the bishop to c7, to attack the weak e5-pawn.
10 Re1
If White chooses to defend the e5-pawn with 10 Qe2 0-0 11 Re1 then after 11 ... Bd7 12 c3 Bc7 (now
the pawn can be defended no more) 13 Bxg6 fxg6! (13 ... hxg6 leaves Black vulnerable on the kingside)
14 Nc5 Qe7 15 b4 Rf5 16 Nd3 I think the position is level. Black will aim to expand gradually on the
kingside with a well-timed ... g6-g5-g4.
Meanwhile, after 10 Bg5!? Qc7 11 Re1 Black should play 11 ... Ncxe5 12 Nxe5 Nxe5:
 a) 13 Bf4? would be a huge mistake because of the tactical strike 13 ... Bxf2+! (the bishop cannot be
captured) 14 Kh1 Bxe1 15 Qh5 g6 16 Qg5 0-0 17 Bxe5 (17 Rxe1 is forced, but 17 ... f6 allows Black to
consolidate) 17 ... Bh4! and wins.
b) This means that White should play 13 Bb5+ Nc6 14 Qxd5 (14 Qg4 doesn’t work after 14 ... e5 15
Qg3 f6) 14 ... 0-0 15 Qf3 e5 and the position is level. White can’t hope for much after 16 Bd3 Be6 (16 ...
Nb4!? is the most ambitious try: for example, 17 Be4 f5 18 Bd5+ Kh8 19 Rad1 Nxc2 20 Re4 Qd6! 21
Rc4 Nd4 22 Nxd4 Bxd4 23 Qa3 Qxa3 24 bxa3 h6 25 Be3 Bxe3 26 fxe3 Rd8 and Black has an edge) 17
Bf6 gxf6 18 Qg3+ Kh8 19 Qh4 f5 20 Qf6+ Kg8 with a repetition.
10 ... Bd7
Castling is also good, i.e. 10 ... 0-0 11 c3 (after 11 Be3 a draw was agreed in R.Van Kam​pen-
B.Grachev, Basel 2013, which was a pity as Black could have played 11 ... Bc7 12 Bd4 Nf4! 13 Bf1 f6
14 exf6 Nxd4 15 Qxd4 gxf6 16 g3 e5 17 Qb4 a5 18 Qd2 Nh3+ with the superior position) 11 ... Bc7!
when Black’s play is very straightforward; he wants the e5-pawn.

Now:
a) Giving up the pawn is not troubling as after 12 Nbd4 Ncxe5 13 Nxe5 Nxe5 14 Bc2 Ng6 15 Qh5 e5!
16 Nf5 e4 Black has released the ‘French’ bishop and is a pawn up.
b) 12 Bxg6 fxg6 is rather comfortable for Black on account of the open f-file.
c) 12 Qc2! is White’s best try, but Black can connect the rooks and apply pressure on e5 with a well-
timed ... Qb8. Indeed, after 12 ... Bd7 13 Nbd4 Qb8 14 Nxc6 bxc6 the e5-pawn is under pressure and
now Black can also play ... c6-c5.
11 c3 Bc7
Following the plan of attacking the e5-pawn.
12 Bg5 Qc8 13 Bxg6 fxg6 14 Bh4 0-0 15 Bg3
If 15 Nbd4 Nxd4 16 Nxd4 (or 16 cxd4 Qe8 and Black is well placed for play along the c-file) 16 ...
Qe8 17 f3 h6, again with a comfortable position.
15 ... Ne7 16 Nfd4 h6 17 f4 a5 18 a4 Bb6 19 Kh1 Qc4 20 Qg4 Nf5!
 21 Qxg6 Bxd4 22 Nxd4 Nxd4 23 cxd4 Qxd4 24 Rf1
As played in N.Kosintseva-N.Batsiashvili, Gaziantep 2012. Instead, Rad1 Qxa4 snaffles a pawn.
24 ... Qxb2!
Bold, but good.
25 Rab1
After 25 Rfb1 Qd4 26 Rd1 Qa7 27 Bh4 Qe3 28 Bg3 Rac8 White is struggling to show compensation
for the missing pawn.
25 ... Qd4 26 Rxb7 Qxa4 27 Rc1 Rac8
Black is the side with a strong counterattack.

A3) 8 Bf4!?

The most challenging continuation to meet. White begins to overprotect the e5-pawn and the bishop
will be better placed on f4 or g3 to deal with the ... f7-f6 pawn break. In reply, I propose something a
little bit unconventional.
8 ... h6!?
This radical idea is derived from the game D.Bisby-T.Chapman, British League 2013.
 My original recommendation was 8 ... Ng6 9 Bg3 0-0 10 Nbd2 f5 11 exf6 Qxf6 and after 12 c4! Bd7
13 cxd5 exd5 14 Qb3 Be6 15 Bxg6 Qxg6 16 Qxb7 Nd4 17 Nxd4 Bxd4 18 Nf3 Bc5 we have a
complicated position.
9 Nbd2
The mirror move 9 h3 was my husband’s suggestion when I tested this position against him. After 9 ...
Ng6 Black may play for ... f7-f5:
a) 10 Bh2 leaves the f2 point sensitive to attack: 10 ... Qb6 11 Nc3 (the knight isn’t so well placed
here, getting in the way of the c2-pawn, but if 11 Nbd2 Qxb2) 11 ... Be7 12 Qe2 0-0 and Black will play
for ... Bd7, ... a6, ... Rac8 and counterplay in the centre.
b) 10 Bg3 0-0 11 Nbd2 f5 12 exf6 Qxf6 13 Nb3 (13 Qe2 would just be a mistake after 13 ... e5 14
Bxg6 Qxg6 15 Bxe5 Bxh3! and wins) 13 ... Bb6 14 c4 Nf4 15 Bxf4 Qxf4 16 Rc1 (now Black has a
wonderful resource) 16 ... e5! 17 cxd5 Bxh3! 18 gxh3 (18 dxc6 loses immediately to 18 ... Qg3) 18 ... e4
19 Nc5 (19 dxc6? loses on the spot to 19 ... exf3 20 Kh1 Bc7) 19 ... exf3 20 Rc4 Qe5 21 Re1! (21 dxc6?
is again impossible because of 21 ... Bc7) 21 ... Qh5 22 dxc6 (only now that the queen has been driven off
the dangerous h2-b8 diagonal is it safe to take the knight) 22 dxc6 Bxc5 23 Bf1 bxc6 with an equal
position.
9 ... g5!

I love this idea which is both aggressive and unconventional. Black’s plan is to deposit a knight on f5
and advance the kingside pawns using the white bishop on g3 as a target.
10 Nb3
Instead, 10 Be3 Bxe3 11 fxe3 g4 12 Nd4 Nxe5 13 Be2 f5 14 c4 Bd7 15 Qe1 0-0 16 Qh4 Kg7 is
slightly better for Black and 10 Bg3 is met with the same idea as in the main line: 10 ... Nf5.
10 ... Bb6 11 Bg3
If 11 Be3!? Black should simply capture with 11 ... Bxe3 12 fxe3 g4!? (flouting all the rules of
development, Black plays to win a pawn) 13 Nfd4 Nxe5:
 a) 14 e4 dxe4 15 Bxe4 f5 16 Qe2 (now the bishop is taboo and what is left is get the black king to
safety as quickly as possible) 16 ... N7g6 17 Bd3 (if 17 Qe3 Qg5! or 17 h3 fxe4 18 Qxe4 gxh3 19 Rad1
Qg5 20 Rd2 Rf8 21 Rxf8+ Kxf8 22 Rf2+ Kg8 23 Nc5 Ng4! and Black is winning) 17 ... 0-0 (Black gets
away with his provocative play because there are no pawn levers for White; indeed, none of White’s
pieces can challenge the black defence) 18 Rad1 Nxd3 19 Rxd3 Qb6 20 Re3 Bd7 21 c4 e5 22 c5 Qf6 and
Black is defending against White’s initiative.
b) 14 Qe1 Nxd3 15 cxd3 e5 16 Ne2 and now Black plays the calm 16 ... 0-0 17 Qh4 and then the
surprising 17 ... Nf5! 18 Qxg4+ Qg5 19 Qxg5+ hxg5 20 Kf2 (20 e4 is not tricky at all: 20 ... Ne3 21 Rfc1
d4 22 Rc7 and after 22 ... b6 Black will target the d3-pawn) 20 ... b6 21 Rfc1 (or 21 e4 dxe4 22 dxe4
Nd6! followed by ... Ba6 when Black stands well) 21 ... Ba6. Here White is simply not coordinated
enough.
Instead, 11 Bd2 g4 12 Ne1 Nxe5 13 Qe2 Qc7 14 Kh1 h5 15 Bf4 f6 16 f3 Bd7 17 fxg4 hxg4 18 a4 a6
reaches a fascinating position and after 19 a5 Ba7 20 Rc1 0-0-0 21 c4 N7c6 22 cxd5 exd5 White is not
getting anywhere with his queenside counterplay.
11 ... Nf5!

12 Re1
 If 12 Bxf5 exf5 13 h3 (13 h4 is not scary for Black: 13 ... f4 14 Bh2 Bg4 15 Re1 0-0 and despite
appearances, the black king is safe mainly due to the white bishop being encased on h2: for example, 16
Qd3 Re8 17 Nbd4 Bxd4 18 Nxd4 Nxe5 19 Qb5 Qd7 20 f3 Bf5 21 hxg5 hxg5 and I believe Black has the
upper hand) 13 ... f4 14 Bh2 Bf5 15 Nfd4 Bg6 and then:
a) 16 e6 Qf6 17 exf7+ Kxf7 18 c3 (or 18 Nb5 Kg7 19 Qxd5 Rad8) 18 ... Kg7 and White is not going
to break through without taking big risks.
b) 16 Re1 0-0 17 e6 Qf6 18 exf7+ Bxf7 looks great for Black on account of the shut-in bishop on g2.
Likewise, after 12 a4 a6 13 a5 Ba7 14 Re1 Nxg3 15 hxg3 g4 16 Nh4 Qg5 17 Qe2 Bb8 18 Nc5 Bxe5
19 c4 0-0 Black is better.
12 ... h5! 13 Bxf5 exf5 14 h4 f4! 15 Bh2 g4 16 Nfd4 g3! 17 fxg3 Rg8 18 gxf4 Qxh4 19 Kh1?
White should play 19 Qd2 Qg4 20 c3 Bf5 21 Re3 Nxd4! 22 cxd4 h4 23 Rc1 Be4 24 Re2 Rc8 25
Rxc8+ Qxc8, although this also looks promising for Black, as does 19 c3 Bf5 20 Qf3 0-0-0 21 Kh1 Nxd4
22 cxd4 Kb8.
19 ... Nxd4 20 Nxd4 Qf2 21 Qf3

21 ... Bxd4!
In this messy position it is not surprising that Black erred on the side of caution and played 21 ... Qxd4
22 e6! Bxe6? (22 ... fxe6! 23 Rad1 Qf6 would still have been good) 23 f5 0-0-0 when a draw was agreed
in Bisby-Chapman.
22 Qxd5 Be6 23 Qb5+
23 Qxb7 runs into 23 ... Rd8 24 Qc6+ Kf8 25 Rf1 Qe2 26 Rae1 Qg4.
23 ... Bd7 24 Qxb7 Rd8 25 Rad1 Rg6
The threat of ... Bd7-c6 means that White is in serious trouble.
26 Rf1 Qxc2 27 Rxd4 Bc6!
Black is winning.

B) 5 a3
 White doesn’t even bother to defend the c5-pawn and instead concentrates on kicking back the bishop
on c5 with b2-b4. White will then continue with easy development: the bishop on b2 will control the e5-
square in full preparation for the ... f7-f6 break. In addition, if Black is too slow then c2-c4 will arrive
and the opening of the d-file could be critical.
For Caro players this is a tricky system to face, firstly because Black doesn’t really have an initiative
to play with. The second issue is how on earth will we develop the light-squared bishop on c8,
particularly if ... f7-f6 is not good? This is in fact a very pertinent question. Black is looking to achieve
two things: manoeuvring a knight to f4 or, more dynamically, undermining the e5-pawn with ... f7-f6. To
begin with Black should make White defend the e5-pawn with 6 Nf3.
5 ... Nc6 6 Nf3
6 Bb5 is perhaps confusing systems: 6 ... Bxc5 7 Nf3 Nge7 8 b4 Bb6 9 Bb2 and now Black can play 9
... a5!.
Instead, after 6 b4 a5 7 Bb2 axb4 8 axb4 Rxa1 9 Bxa1 Nxb4 the c5-pawn is incredibly vulnerable.
6 ... Bxc5
This position can also arise via 5 Nf3 Bxc5 6 a3 Nc6.
I would also like to mention an ambitious try which is not at all easy to refute over the board, namely
6 ... f6!?. A brief run through of some of the obvious lines sees:
a) 7 b4 fxe5 8 b5 (or 8 Bb5 Qc7 9 Bb2 e4 10 Ne5 Nf6 11 0-0 Be7 12 f3 0-0 and I don’t see the
problem for Black) 8 ... Nb8 9 Nxe5 Nf6 10 Be3 Qc7 11 Qd4 Nbd7 12 Nxd7 Bxd7 and the c5-pawn will
fall.
b) 7 exf6 Qxf6 8 c4 Bxc5 9 cxd5 exd5 10 Nc3 Nge7 with an open position.
c) A recent rapid game continued 7 Bb5 Bxc5 8 0-0 Bd7 9 b4 Bb6 10 Bb2 fxe5 11 Bxc6 bxc6 12
Nxe5 Nf6 13 Nd2 0-0 14 Nb3 Ne4 15 Bd4 Rf5 16 Nd3 Qh4 17 g3? Qh3 18 Nf4 Rxf4 19 gxf4 Rf8 and
Black had a strong attack in J.Adair-J.Hawkins, Leeds 2014.
7 b4 Bb6
 There are two piece configurations associated with the move a2-a3: White either chooses to develop
the bishop straight away to b2, aimed at preventing an early ... f7-f6, or he delays the development of this
bishop for a while.
8 Bd3
The aim of this move is very simple. White intends to castle and guards against Black’s possible plan
of playing ... Ng6-f4. White’s usual idea is to provide a bolt hole for the d3-bishop to escape to and then
play directly for c2-c4 or wait around for Black to determine the central situation.
A less logical idea is 8 b5. Okay the knight has to retreat to its starting square, but White has lost the
punch behind the pawn push c2-c4. After 8 ... Nb8 9 Bd3 Nd7 10 0-0 Ne7 11 Bf4 Ng6 12 Bg3 Nc5 13
Nbd2 Nxd3 14 cxd3 a6 15 a4 Bd7 16 Qb3 0-0 17 d4 axb5 18 axb5 Rxa1 19 Rxa1 Qc7 20 Qa3 Rc8 Black
had the upper hand on account of the weak b5-pawn and his control over the c-file in I.Glek-R.Dautov,
Katernberg 2005.
Instead, 8 Bb2 will actually transpose to Line B1 below after 8 ... Nge7 9 Bd3 Ng6 10 0-0 (10 Nbd2
doesn’t really make any sense because of 10 ... Bc7 11 Qe2 Nf4 12 Qf1 Bd7 and Black can’t claim to
have any problems).
8 c4 is something that was once suggested to me, but is easily dealt with by the natural and strong 8 ...
Nge7 9 c5 (regardless of whether White plays this pawn push, Black should be attacking the e5-pawn) 9
... Bc7 10 Bb2 Ng6 11 Qe2 Bd7 12 h4 Qb8 when Black wins the e5-pawn.
8 ... Nge7 9 0-0 Ng6
 We now come to White’s main choices:

B1) 10 Bb2
B2) 10 Re1

B1) 10 Bb2 0-0!

Black delays the idea of ... Nf4 for one further move - this is to increase its impact. The problem with
the immediate 10 ... Nf4 is that it is well met with 11 c4 Nxd3 12 Qxd3 dxc4 13 Qxc4 0-0 14 Nc3 and
White is too active in the centre.
11 Re1!
In my opinion White’s best bet. The light-squared bishop obtains a desired escape square and Black
now has to find a way to unravel. Others:
a) 11 c4 is probably too impatient: 11 ... dxc4 12 Bxc4 Qxd1 13 Rxd1 Bc7! (this is the point; the e5-
pawn is threatened and White must play Rd1-e1) 14 Re1 (White can’t get away with sacrificing the pawn
with 14 Nbd2 Ncxe5 15 Rac1 Nxc4 16 Rxc4 Bb8 17 Ne4 b5 18 Rc5 Bb7 when Black escapes from the
cage) 14 ... Rd8 15 Bf1 (15 Nbd2 allows Black to grab the e5-pawn) 15 ... a6 16 g3 b5 17 Bg2 Bb7! (17
... Rb8 was played in M.Sebag-O.Girya, Khanty-Mansiysk 2012, which finished in a draw) 18 Nc3 Rac8
19 Rad1 is equal.
b) 11 Nbd2?! is already somewhat dubious, allowing us to play the ideal 11 ... Nf4! and if 12 Re1 (12
Nb3 is again answered with the thematic 12 ... f6 13 exf6 gxf6 14 c4 e5 and Black is doing well) 12 ... f6!
13 exf6 gxf6 14 Nb3 (or 14 Bf1 e5 15 c4 Kh8! 16 b5 Ne7 when Black has good counterplay) 14 ... e5 (the
central pawn mass is very strong) 15 Bf1 Bg4 16 Bc1 Nxg2 17 Bxg2 e4 with a glorious mess in which
only Black, with the more active pieces, can have the preferable chances.
c) 11 g3 looks strange, but White reasons that if Black’s sole plan is to play ... Ng6-f4 then why
should he not take the liberty to cover this square? However, after 11 ... d4! 12 Re1 f6! 13 exf6 Qxf6 14
Nbd2 e5 15 Kg2 Bg4 16 h3 Bh5 17 Be4! Rad8 (all Black’s pieces are working well; the main thing is that
he should simply focus on improving his pieces) 18 Qe2 Kh8 19 Rad1 (the knight cannot leave the
defence of f3 as shown by 19 Nc4? d3! 20 cxd3 Bd4! and Black suddenly has an amazing bind; the
weakness of g2-g3 is telling) 19 ... Nce7 Black has a great position.
11 ... f6
I recently got the chance to test out my analysis at the Tradewise Gibraltar Chess Festival, versus
Canadian Grandmaster Eric Hansen.

12 exf6
This is forced as otherwise White’s f2-pawn is simply too weak.
12 ... gxf6 13 c4!?
This has to be correct as I don’t think White can allow Black to strengthen the pawn centre even more
with ... e6-e5.
After 13 Bxg6 hxg6 14 c4 dxc4 15 Nbd2 e5 16 Nxc4 I would be done with any troublesome queen
moves and play 16 ... Qxd1 17 Raxd1 Bc7 18 Nd6 Rd8 19 Ne4 Rxd1 20 Rxd1 Kf7 21 b5 Ne7 when
Black can slowly unravel and look forward to using the active king.
Instead, 13 Nc3 is an awkward looking move, if one designed tactically to prevent ... e6-e5.
However, after 13 ... Nce7 the break will happen anyhow.
During my game against Hansen it dawned on me what would I do if I faced the inventive 13 Nh4!?.
Black is forced to play 13 ... Nxh4 14 Qg4+ and can now return the offered knight with 14 ... Ng6 15
Bxg6 e5 16 Bf5+ Kh8 with a good position.
13 ... Nf4
Black’s best bet.
 13 ... dxc4 is also possible and after 14 Bxc4 Qxd1 15 Rxd1 Nge5 White has to play very precisely to
preserve an advantage: for example, 16 Nxe5 (if 16 Nbd2 Nxc4 17 Nxc4 e5 or 16 Be2 Nxf3+ 17 Bxf3 e5
18 Nc3 Be6 19 Nd5 Rad8 20 Nxb6 axb6) 16 ... fxe5 17 Rf1 e4 18 Nc3 Ne5 19 Be2 Nd3 20 Bxd3 exd3
21 Rad1 Rd8 22 Ne4 e5 (it’s more important to release the bishop than hold on to the pawn) 23 Bxe5 Bf5
when his active pieces and the passed d3-pawn give Black a lot of compensation for the material deficit.
14 cxd5
After 14 Bf1 dxc4 I feel the endgame might be a very slight edge for White, but an edge that Black
should be able to eat away at:
a) 15 Qc2 allows 15 ... Nd3 16 Rd1 Qe7 17 Bxd3 cxd3 18 Qxd3 Rd8 19 Qc2 Rxd1+ 20 Qxd1 e5 21
Nc3 Be6 22 Nd5 Bxd5 23 Qxd5+ Qf7 24 Qe4 Rd8 and again the activity of Black’s pieces offsets his airy
king.
b) 15 Bxc4 Qxd1 16 Rxd1 Kg7 17 Nc3 e5 18 Na4 Bg4 19 Nxb6 axb6 20 Bc1 (or 20 h3 Bxf3 21 Rd7+
Kh6 22 gxf3 Rfd8 23 Rad1 Nd4) 20 ... Bxf3 21 Rd7+ Kg6 22 gxf3 Rad8 with the easier position to play.
Say, for instance, White plays 23 Rxd8 Rxd8 24 Be3 then 24 ... Nd4! ensures that one of the bishops has
to exchange itself for a knight.
c) Hansen played a move I hadn’t given too much thought to: 15 Qxd8. Black can now play 15 ...
Rxd8! (I wasn't entirely satisfied with my game after 15 ... Nxd8 16 Bxc4 Bd7 17 Nbd2 Rc8; I thought
plans involving Ne4 would give me the most headaches). Now the greedy 16 Bxf6 (if 16 Bxc4 Kg7
followed by ... e5) is met by 16 ... Rf8 17 Bg5 Nh3+ (this is the point I overlooked in my over-the-board
analysis) 18 gxh3 Rxf3 19 Be3 Nd4 with an equal game.
Instead, if 14 c5 Nxd3 15 Qxd3 Bc7 16 Nc3 d4 17 Nb5 e5 and Black doesn’t have any problems at all
as White cannot break down the black centre.
14 ... Qxd5 15 Be4 Qxd1 16 Rxd1 e5! 17 Nc3 Nd4!

18 Nd5
18 g3? is not possible because of the weakness of f2: 18 ... Nxf3+ 19 Bxf3 Nh3+ 20 Kg2 Nxf2 21 Rd2
Bh3+ and White is in serious trouble.
18 ... Nxf3+ 19 Bxf3 Nxd5 20 Bxd5+ Kg7 21 Be4
If 21 Bc4 Bf5!.
21 ... Rf7 22 Rd3
Instead, 22 Bd5 Re7! and 22 Rd6 Bg4 are level, while 22 a4 Bg4! 23 Rd3 Rd8! 24 Rg3?! is
inaccurate due to 24 ... h5 25 h3 Rd2 and Black is better.
 22 ... Be6 23 Rd6 Re7
This sets some nice traps.
24 Bxb7?
Likewise, 24 Rad1 Rc8! 25 Bxb7? is terrible because of 25 ... Rc2.
24 ... Rxb7 25 Rxe6 Rd8 26 Bc3 Rc7
Threatening to trap the white rook and the imminent invasion of the black rooks into the white camp
means that Black has a virtually won game.

B2) 10 Re1
White makes a hole for the d3-bishop to escape to before playing Bb2. Now Black should throw common
sense out of the window and strike with:
10 ... f6!

This is mainly possible due to the fact that White has not played Bb2, controlling the a1-h8 diagonal.
11 Bxg6+!
Otherwise Black will simply play ... Ngxe5: for example, 11 Be3 Ngxe5 12 Nxe5 Nxe5 13 Bxb6 axb6
14 Qh5+ g6 15 Qh6 Kf7 16 Be2 Kg8 17 Nc3 Nf7 18 Qd2 Kg7 and Black has consolidated.
11 exf6 is another idea: 11 ... Qxf6 (taking advantage of the exposed diagonal) 12 Bxg6+ (12 Ra2 is
rather tame; now Black should break the rules and play one more knight move: 12 ... Nh4! 13 Nxh4 Bxf2+
14 Kh1 Qxh4 15 Rf1 Bg3 16 h3 e5 and White is in terrible trouble) 12 ... hxg6 13 Ra2 and now I like
simple development with 13 ... 0-0. My idea is that after 14 Bb2 (or 14 c4 dxc4 15 Nbd2 Nd4 16 Nxc4
Nxf3+ 17 Qxf3 Qxf3 18 gxf3 Bc7 19 Ne5 Bxe5 20 Rxe5 Bd7 and Black is not worse) we can choose
between 14 ... Qf4!, with ideas of playing ... g5-g4 or swinging the queen over to c4, and the tactical 14 ...
e5 when 15 Qxd5+? (15 Nc3 is more troubling) 15 ... Be6 drops a rook.
11 ... hxg6 12 Bb2
Alternatively, 12 exf6 gxf6 13 Bb2 (13 Qd3 Kf7 14 Bb2 e5 15 Qb3 Be6 16 Nbd2 e4 looks crushing
for Black) 13 ... e5 14 b5 Ne7 keeping the minor pieces around the black king.
12 ... fxe5
12 ... Nxe5 is too risky as after 13 Nxe5 fxe5 14 Bxe5 Qh4 15 Bg3 Bxf2+ 16 Kxf2 Qf6+ 17 Kg1 Qxa1
18 Be5 Qa2 19 Qd3 Black could be in big trouble.
13 Nxe5
13 b5?! is dubious on account of 13 ... e4 14 bxc6 exf3 and now White must go 15 Qxf3!. Instead, 15
Bxg7 is a horrible mistake: 15 ... Qh4 16 Qxf3 Qxh2+ 17 Kf1 bxc6 18 Nd2 Rg8! 19 Be5 Ba6+ 20 c4
Qh1+ 21 Ke2 Qh5 and Black is doing well thanks to the bishop-pair.
13 ... Qf6!?
It seems anti-positional to put the queen on the same diagonal as the white bishop, but we have some
counter-threats. 13 ... Qh4 and 13 ... 0-0 are good alternatives.

14 Qf3
Alternatively:
a) 14 Nd3 Bxf2+ 15 Nxf2 Qxb2 16 Qd3 (16 Nd2 0-0 17 Nf3 a5 18 Rb1 Qxa3 19 Rb3 Qa2 keeps the
queen safe) 16 ... Qxa1 17 Qxg6+ Kf8 18 c3 (18 Rf1 Qd4 19 Kh1 Qf4 defends easily) 18 ... Rh6 and
White has nothing.
b) 14 Qd2 Bc7! is one point and after 15 Qe3 Rh5 16 Nd3 d4 17 Qf3 Bxh2+ 18 Kf1 Qxf3 19 gxf3 Kf7
20 f4 e5! 21 fxe5 Nxe5 22 Bxd4 Bh3+ 23 Ke2 Re8 Black is winning.
14 ... Nxe5 15 Bxe5 Qxf3 16 gxf3 Bd7
Black must have the advantage on account of the open position and his bishop-pair.

C) 5 Qg4
 White aims to prevent easy development of the kingside and this aggressive move should be dealt
with as actively as possible.
5 ... h5!?
This was my earlier recommendation and it is still the most interesting choice, but it does have one
big positional drawback: the weakening of the kingside means that Black has committed himself to active
play. I should stress that Black should at all times play purposeful moves and active ones at that. Step one
is to harass the queen.
6 Qg3
6 Qf4 allows Black to simply capture the pawn and after 6 ... Bxc5 7 Bd3 we have:
a) In my earlier book I liked 7 ... Qb6 (preventing White from enjoying harmonious development) 8
Nc3 Nc6 and now:
a1) 9 Na4 is neutralized by 9 ... Qb4+ 10 c3 Qxf4 11 Bxf4 Be7 12 Nf3 Bd7 13 b3 Rc8 14 Kd2 Nh6
15 h3 f6 (Black must do something otherwise he will drift into an inferior position) 16 Rhe1 (or 16 exf6
gxf6 17 Bg6+ Nf7 18 Bd3 e5) 16 ... 0-0 with a good position.
a2) However, White has 9 a3! renewing the threat of Na4 and obtaining a slight advantage.
b) I now prefer the move 7 ... Nc6!? with the idea of playing ... Nb4.

Here:
b1) 8 c3 Nh6 9 Nf3 and now Black has the tricky 9 ... Ng4 10 0-0 Qc7 11 Bb5 Qb6 12 Nd4 0-0 13
Bxc6 f6! 14 b4 (if 14 Bxd5 fxe5) 14 ... Bxd4 15 Qxd4 Qxc6 16 exf6 e5 17 Qd2 Nxf6 when the open lines
and centre give him the initiative.
b2) The natural 8 Nf3 Nb4 9 Bb5+ Bd7 10 Bxd7+ Qxd7 11 Na3 Ne7 12 0-0 Nbc6 13 Be3 Ng6 14
Qg3 Bxa3 15 bxa3 Qc7 sees the e5-pawn fall.
Instead, 6 Bb5+ helps Black get rid of his bad bishop: 6 ... Bd7 7 Bxd7+ Nxd7 8 Qe2 Bxc5 9 Nf3 and
now Dautov recommends 9 ... Qa5+!? 10 c3 Qa6! 11 Qxa6 bxa6 12 Nbd2 Ne7 13 Nb3 Bb6 14 0-0 Ng6
15 Re1 Bc7 when it is not clear how White can defend the weak e5-pawn.
6 ... Ne7
In my earlier book I recommended 6 ... h4 which is still very interesting, but this is a simpler move.
7 Bd3
After 7 Bg5 Nbc6 8 Bd3 b6 9 Nd2 (if 9 cxb6 Qxb6!) 9 ... bxc5 10 b3 Qa5 11 Ne2 h4 12 Bxh4 Nxe5!
13 Qxe5 Rxh4 14 Rd1 Ba6! Black’s king will be safe. He can castle by hand should the need arrive and
after 15 0-0 Bxd3 16 cxd3 Nf5 17 g3 Rh8 now ... Bd6 is in the air.
 7 ... Na6

With the aim of chasing away the bishop on b3 to secure the f5-square for the knight.
8 Be3
After 8 Bxa6 bxa6 9 Ne2 Qa5+ 10 Bd2 Qxc5 11 c3 h4 the bishop-pair and strong pawn structure are
in Black’s favour.
8 ... Nb4 9 Nf3 Nxd3+ 10 cxd3 Nf5 11 Qf4 Nxe3 12 fxe3 Bxc5 13 0-0 0-0
Black’s king is completely safe.
14 Nc3
14 Ng5 fails to 14 ... f6, and if 14 d4 Be7 15 Nc3 Bd7 16 e4 dxe4 17 Nxe4 Rc8 18 g4 f5 19 exf6 Bxf6
with a complicated game.
14 ... f6! 15 d4 fxe5 16 Qxe5 Rf5! 17 Qg3 Bd6 18 Ne5
After this move the game fizzles out to a draw.
18 Qe1 is the only safe place for the white queen, but after 18 ... Bd7 19 Qe2 Rc8 20 Rac1 a6 21 e4
dxe4 22 Nxe4 Rxc1 23 Rxc1 Bf4 24 Rc3 Bb5 the two bishops are a formidable force. Likewise, 18 Qh3
Bd7 19 Nd2 Rxf1+ 20 Rxf1 Qg5 gives Black a nice position.
18 ... Rg5 19 Qf4 Rf5 20 Qg3 Rg5 21 Qf4 Rf5
½-½, M.Vachier Lagrave-M.Turov, Basel 2013.

D) 5 Nc3?!
This is a rare line largely because it’s not the most challenging set-up with the white knight in front of the
c-pawn. Black should get on with things and undermine the e5-pawn with 7 ... f6.
5 ... Nc6
It is a good idea to prevent the queen coming to g4 and now White must defend e5.
6 Nf3 Bxc5 7 Bd3
Instead, 7 Be2 f6! 8 exf6 Nxf6 is slightly better for Black and 7 Bb5 Bd7 8 Qe2 a6 9 Bd3 Nd4 10
Nxd4 Bxd4 11 0-0 Qc7 12 Bf4 Ne7 is equal.
7 ... f6 8 exf6
After 8 Qe2 fxe5 9 Nxe5 Nxe5 10 Qxe5 Nf6 11 0-0 0-0 White blundered here with 12 Bf4? (12 Qg3
Bd7 13 Bd2 Qb6 is the kind of play we can expect for Black, while after 12 Be3 Bd6 13 Qg5 e5 14 Nb5
Bb8 Black has equalized) 12 ... Ng4 13 Qh5 Rxf4 14 Qxh7+ Kf8 15 Qh8+ Ke7 16 Qxg7+ Kd6 17 Nb5+
Kc6 18 b4 Qe7 19 Qc3 Rxb4, which left him in huge trouble in I.Tsigelnitskiy-D.Yevseev, St Petersburg
2001.
8 ... Nxf6 9 0-0

9 ... 0-0
Black can even play 9 ... e5:
a) 10 Bg5 e4 11 Bb5 and now 11 ... 0-0 is the safe way to play. 11 ... exf3 is also possible, but it’s
complicated after 12 Re1+ Kf8 13 Bxf6 gxf6 14 Qxf3 Rg8.
b) 10 Nxe5 Nxe5 11 Re1 Bxf2+ 12 Kxf2 Nfg4+ 13 Kg1 0-0 and Black is doing very well on the dark
squares. The point is that 14 Rxe5 Nxe5 15 Bxh7+ Kh8! 16 Qh5 Qb6+ wins.
10 Bg5
Both 10 Re1 Ng4 11 Bxh7+ Kxh7 12 Ng5+ Kg8 13 Qxg4 Bxf2+ and 10 Qe1 Bd6 11 Qd1 a6 12 h3
Bc7 look quite pleasant too.
10 ... a6 11 Qe2 Bd7
Now the white knight on c3 is really in the way and Black will play ... Bd6 and ... Qc7, with ideas of
... Ne5 and preparing a pawn advance.
12 Na4 Bd6 13 c4 e5 14 cxd5 Nd4 15 Nxd4 exd4 16 b3 b5 17 Nb2 Qc7 18 g3 Rae8 19 Qd1 Bb4 20
Rc1 Qd6
The d5-pawn will fall.

Summary
After 5 Nf3 Bxc5 6 Bd3 Nc6 matters hinges on when the c1-bishop will be developed. If White is not
quick in determining the bishop’s position and prefers to play moves such as c3, Qe2, Nbd2 then Black
should play ... Ng8-e7-g6, attacking the e5-pawn and importantly depriving White of the f4-square for the
bishop. The most dangerous approach is if White puts his bishop on f4 early when I suggest a rather
radical solution based on making the bishop a target.
Alternatively, White can opt to locate the bishop to b2 by playing a2-a3 and b2-b4. The distinguishing
feature here is that if White chooses to develop with Bb2, Bd3 and 0-0, Black should react with ... f7-f6
only after Re1. However, should White delay Bb2 with 10 Re1, Black should hit out with 10 ... f6!.
Sidelines such as 5 Qg4 and 5 Nc3 don’t promise anything for White. In the case of the former the queen
can simply be chased away, while the latter is dubious and well met by an early ... f7-f6.
Chapter Eleven
The Advance Variation without 4 dxc5
1 e4 c6 2 d4 d5 3 e5 c5

It has become quite trendy these days to delay the capture on c5 and this can be done for two reasons.
Those not wanting to play complex lines will react with the solid 4 c3, keeping the pawn chain intact,
while those who are seeking complications may play either 4 Nf3 or 4 c4!?. We shall thus look at the
following moves:

A) 4 Nf3
B) 4 c4!?
C) 4 c3?!

Otherwise:
a) 4 Ne2 cxd4 5 Nxd4 Nc6! 6 c4 transposes to Line A2, below.
b) 4 Nc3 misplaces the knight and blocks the c-pawn’s advance to either c3 and c4. After 4 ... Nc6 5
Bb5 e6 6 Be3 cxd4 7 Bxd4 Nh6 (rerouting the knight to the better f5-square) 8 Nf3 Nf5 9 0-0 Be7 10 Ne2
0-0 (Black has no problems here) 11 Bd3 Nfxd4 12 Nexd4 f6 13 Ng5? (White begins to lunge wildly) 13
... fxg5 14 Qh5 Rf5! (picking up two pieces for the rook) 15 Nxc6 (15 g4 does not work because of 15 ...
g6!) 15 ... bxc6 16 Bxf5 exf5 17 Rfe1 Be6 Black was easily winning in J.Pareja Perez-D.Campora, Ceuta
1992.

A) 4 Nf3
 This is a tricky little move. White delays making the critical decision in the centre for a move.
4 ... Nc6!
A good and flexible move. Now White cannot put off the decision any longer.
I would also like to make the reader aware of another promising line against those hoping to use this
move order to play c2-c4: 4 ... Bg4 5 c4!? (the problem is that 5 dxc5 Nc6 transposes to 4 dxc5 Nc6 5
Nf3 Bg4 - a variation not covered in this book) 5 ... cxd4! 6 cxd5 Qxd5 7 Nc3 Qd7 8 Bb5 Nc6 9 Qxd4
Nxd4 10 Nxd4 Rc8 11 Bxd7+ Bxd7 with a reasonable endgame for Black.
Now:

A1) 5 Bb5
A2) 5 c4!?

Of course, not 5 Be2 cxd4! and should White attempt to trick us into a transposition with 5 dxc5, I believe
Black should play 5 ... e6 (the ideal 5 ... Bg4 unfortunately transposes into the 4 dxc5 Nc6 lines; 6 Be3 e6
7 Bb5 is perfectly acceptable for Black, but not ideal for this book) 6 Be3 Nh6 7 c3 (if 7 Bd3 Ng4!) 7 ...
Nf5 8 Bd4 Bd7, taking play back into the main line and Chapter Nine.

A1) 5 Bb5 cxd4 6 Nxd4 Bd7

 7 Nxc6
7 e6 was once tried by Stuart Conquest, but it’s really not very good: 7 ... fxe6 8 Bxc6 bxc6 9 0-0 Qc7
preparing e5 10 f4 c5 11 Nf3 Nf6 12 c4 g6 13 Nc3 d4 14 Ne2 Bg7 15 Bd2 Ne4 and although the game
eventually ended in a draw, White had nothing to make up for the material deficit in S.Conquest-K.Arkell,
British Championship, North Shields 2012.
7 ... Bxc6
7 ... bxc6 is also possible: 8 Bd3 e6 9 0-0 Ne7 10 Nc3 Ng6 11 Qh5 Bc5 12 Re1 Rb8 and Black
doesn’t have any problems.
8 Bxc6+ bxc6 9 c4 e6 10 0-0 Ne7 11 Qa4 Qd7 12 Nd2 Ng6 13 Nf3 Be7 14 Rd1 Rd8 15 Be3 c5!
Black has no problems and, in fact, I think he will have the easier game on account of the strong d5-
pawn and the open b-line.

A2) 5 c4!?
Immediately throwing down the gauntlet.
5 ... cxd4!
5 ... e6 is another possibility, but this allows 6 cxd5 exd5 7 Bb5! cxd4! 8 Nxd4 Nge7 9 0-0 a6 10
Bxc6+ bxc6 and White has a small edge.
6 Nxd4
6 cxd5 is another try, but Black can play 6 ... Qxd5 7 Nc3 Qd8! 8 Nb5 a6 9 Nbxd4 Nxe5 10 Nxe5
Qa5+ 11 b4 Qxe5+ 12 Be2 e6 13 a3 Nf6 14 Bb2 Be7 15 Nb5 Qb8 followed by castling into safety.
6 ... e6
 The safest response for Black.
7 Nc3
Invariably played. The alternatives don’t really trouble Black:
a) With 7 Nxc6 bxc6 White simplifies the position hoping that the potentially weak c6-pawn will
trouble Black. However, this approach didn’t create any problems in the game V.Kotrotsos-V.Kotronias,
Salonica 2006, where Black adopted a very French-like solution to the problem, by undermining the e5-
pawn with a well-timed ... f7-f6: 7 ... bxc6 8 Be2 Ne7 9 0-0 Ng6 10 f4 Bc5+ 11 Kh1 0-0 12 Qc2 Bb6 13
Nd2 f6!? (now that Black’s pieces are fully developed, it is the time to begin operations) 14 Nf3 fxe5 15
fxe5 Ba6 16 Bg5 Qe8 17 Bd3 dxc4 18 Bxc4 Bxc4 19 Qxc4 h6 20 Bd2 Rd8 21 Rad1 Rd5!. Black’s pieces
are very active, his rook is entrenched on d5, attacking the weak e5 point, and his knight has the nice
square f4 to hop into. Quite simply Black has no problems here and, indeed, went on to win.
b) 7 Be3 simply allows Black to play 7 ... Bb4+ 8 Nc3 Nge7 9 Be2 and after 9 ... Bxc3+ 10 bxc3
Nxe5 he has picked up a pawn.

I will now give two options for Black; one dynamic and one solid:

A21) 7 ... Bb4!?

A22) 7 ... Bc5

A21) 7 ... Bb4!?

The sharpest choice. This move requires memorization of certain attacking sequences. It is not so much
about patterns and themes, it is simply a case of striking when the iron is hot.
8 Nxc6!
White’s best move:
a) 8 Qg4 is possible, but not so poisonous after 8 ... Nge7 9 Qxg7 Rg8 10 Nxc6 bxc6 11 Qxh7 d4 12
a3 Qa5 and Black is incredibly active: for example, 13 axb4 Qxa1 14 Qb1 Qxb1 15 Nxb1 a5 and Black
has the slightly better chances as long as he can utilize the two rooks.
b) 8 Qa4 Nge7 9 Bg5 Qa5! again with no problems.
8 ... bxc6

9 Qa4
Alternatively:
a) 9 Qg4!? is still possible, but is not quite the best choice because of 9 ... d4! 10 a3 (10 Qxg7? just
loses to 10 ... dxc3) 10 ... Bf8! (unfortunately Black must retreat the bishop all the way back to protect the
pawn) 11 Ne2 (11 Ne4 Qa5+ picks up a pawn) 11 ... c5 12 Nf4 Qc7! (a common theme in this whole
variation; Black continues to apply pressure to the vulnerable e5-pawn) 13 Nd3 Ba6 14 Qd1 (now White
threatens 15 Qa5+; after 14 b3 Rb8 15 Qd1 he has to retreat anyway to defend the pawn) 14 ... Bb7 15
Bd2 Ne7 16 b4 Ng6 and Black is fine.
b) 9 Bd3 d4 10 a3 Ba5 11 b4 dxc3 12 bxa5 Qxa5 13 0-0 Ne7 14 Be3 0-0 15 Qh5 Nf5 and with ... Rd8
coming Black has little problems.
c) Only 9 a3 should Black capture the knight on c3: 9 ... Bxc3+ 10 bxc3 Qa5 11 Bd2 dxc4 12 Bxc4
Qxe5+ 13 Qe2 Qxe2+ 14 Bxe2 Ne7 15 Rb1 0-0 16 0-0 e5 17 Bf3 Ba6 is slightly better for Black.
d) 9 cxd5 is well met by the active 9 ... Qxd5!.
9 ... Rb8!
This move is only possible because of the direct threat of ... d4. A word of warning for the battle
ahead: don’t be overly concerned with material, instead focus on obtaining good compensation for the
pawn. Black’s compensation lies in the fact that it is not easy for White to develop his pieces and castle.
I must also mention Black’s other tries:
a) White gets an easy game after 9 ... a5 10 cxd5 Qxd5 11 Qxb4 axb4 12 Nxd5 cxd5 13 Bd2 b3 14 a3
Ba6 15 Rc1 Kd7 16 Bxa6 Rxa6 17 Rc3 Rb6 18 Ke2 Ne7 19 Rhc1 when Black has to grovel for a draw.
b) 9 ... Qb6 looks as though it neutralizes the threats, but after 10 Be3 d4 11 Bxd4 Qxd4 12 Qxb4!
Qxe5+ 13 Be2 c5 14 Qa3 (Iordachescu’s 14 Qa5 Ne7 15 Na4 also leaves Black on the defensive) 14 ...
Ne7 I wasn’t too happy in A.Tari-J.Houska, Fagernes 2014. The game eventually ended in a draw, but
after 15 0-0! Bb7 16 Rfe1 0-0 17 Bf3 Qc7 18 Bxb7 Qxb7 19 Ne4 I would have suffered due to the weak
c5-pawn.
After 9 ... Rb8! White has three main moves:

A211) 10 a3
A212) 10 Qxa7
A213) 10 cxd5

10 Qxc6+? is definitely inadvisable on account of 10 ... Bd7 11 Qa6 Qc7! (hitting the e-pawn and
threatening to trap the queen) 12 Bf4 Rb6 and the white queen is trapped.
Instead, with 10 Be3 White tries to develop his pieces, ignoring the offered pawns:
a) 10 ... Ne7 11 a3 (after 11 Qxa7 Rb7 12 Qa4 Qc7! - by now Black should be familiar with this
move - 13 f4 Nf5 14 Bf2 d4! 15 a3 Be7 16 Ne4 Rxb2 Black has recovered his pawn and has the better
position) 11 ... Ba5 12 b4 Bb6 13 c5 Bc7 14 f4 0-0 15 Be2 Nf5 16 Bf2 f6 is slightly better for Black. He
undermines the centre and intends to exploit both the long dark-square diagonal and the open f-file.
b) 10 ... d4!? 11 Rd1 Qa5 12 Qxa5 Bxa5 13 Bxd4 Rxb2 14 Rc1 Ne7 15 Bd3 Rxa2 16 0-0 Ra3 17
Rfd1! c5 18 Nb5 Rxd3 19 Rxd3 cxd4 20 Nxa7 0-0 21 Rxd4 Bc7 and Black shouldn’t be worse. Note that
now 22 Nxc8 Rxc8 23 f4 weakens White’s position considerably after 23 ... g5.

A211) 10 a3
 This was played against me by Romain Edouard back in 2007.
10 ... Bxc3+ 11 bxc3
Our aim in these types of positions is to prevent White from making use of the bishop-pair. At the
moment the bishops aren’t up to much, so we should retain the tension and develop.
11 ... Qc7 12 f4
12 Bf4 plans to make the dark-squared bishop active, but after 12 ... Nf6 13 Be2! Nd7 14 0-0 Ba6! 15
Be3 (the queen finds itself trapped after 15 Qxa6? Nc5) 15 ... Bxc4 16 Bxc4 dxc4 17 Rfb1 0-0 18 f4 Nb6
Black has no problems.
Instead, 12 cxd5? fails to the obvious 12 ... Qxe5+.
12 ... Bd7 13 Bd3 c5 14 Qc2 dxc4 15 Bxc4 Ne7 16 0-0 0-0 17 Be3 Bb5!

The key to this position is to swap the light-squared bishops and prevent the white kingside pawns
from advancing.
18 Bxb5 Rxb5 19 c4 Nf5 20 Bf2 Rbb8 21 Rfd1 Rfd8
Black will soon go ... h7-h5, securing the strong knight on f5.

A212) 10 Qxa7
 Accepting the gambited pawn, but will Black obtain sufficient counterplay.
10 ... Rb7!
Attacking the queen. In this variation timing is everything, so we cannot afford to play a single
wasteful move.
11 Qa4!
The best retreat as 11 Qd4 allows Black to play 11 ... c5 12 Qd1 d4 13 a3 Ba5 14 b4 dxc3 15 bxa5
Qxa5! (Black should speed up the plan of putting a knight on d4) 16 Bd3 Ne7 17 0-0 Rd7 18 Be3 Nf5 19
Qe2 0-0 20 Bxf5 exf5 21 Rac1 (or 21 Rfc1 Re8 22 Qc2 Bb7 and the c3-pawn is poisoned because of 23
Qxc3? Rd1+) 21 ... Rfd8 22 Bg5 Re8 when he is doing well.
11 ... Qc7
We should attack the e5-pawn. This is the second key move to remember.
12 f4!
After 12 Bf4?! Ba5! Black threatens the b2-pawn and to win a piece with ... d4: 13 0-0-0 Bxc3 14
bxc3 Ne7 15 Bd3 0-0 and with the open b-file, Black is sure to stand better.
12 ... Ne7!
Developing.
13 a3
13 Be2 Qb6! prevents the king from finding a safe home and after 13 Bd2 d4 14 Ne4 Bxd2+ White has
to play 15 Nxd2, as after 15 Kxd2 0-0 his king would be out there in the open.
13 ... Bc5!
White isn’t allowed to develop easily.
14 Bd3 0-0
 The most natural continuation. Black’s king is safe unlike White’s.
15 Qc2 h6 16 b4
Likewise, if 16 Bd2 Bd4 17 b4 dxc4 18 Bxc4 Qb6 19 Rd1 Rd8 20 Ne4 Rbd7 21 Nd6 Nf5 and
White’s king is suffering or 16 Qe2 f6 17 exf6 Rxf6 and the position remains difficult for White.
16 ... Bd4 17 Rb1 dxc4 18 Bxc4 Rd8 19 Ne4 c5 20 Nd6?
20 b5 is not too comfortable either for White: 20 ... Qa5+ 21 Qd2 (or 21 Bd2 Qxa3 22 Nd6 Rbd7 23
Qd3 Qa8 24 Qf3 Bb7! with a dangerous attack) 21 ... Qa8 22 Nd6 Rbd7 followed by an exchange
sacrifice on d6.
20 ... cxb4!

21 Nxb7? Bc3+ 22 Kf2 Qxc4 23 Qe2

23 Nxd8? Bd4+ nets the queen.
23 ... Qxe2+ 24 Kxe2 Bxb7
Black is winning.

A213) 10 cxd5
 White’s best chance and the main continuation.
10 ... exd5
10 ... Qxd5?! leads to an endgame where White has the long-term advantage due to the bishop-pair
after 11 Qxb4 Rxb4 12 Nxd5:
a) 12 ... exd5 13 b3! Ne7 14 Bb2 retains a pull.
b) Likewise, 12 ... cxd5 might be met by 13 b3 and 13 f3 Ne7 14 b3 Bd7 15 Ba3 also gave White an
edge in R.Alarcon Bahamondes-D.Romero Araya, Santiago 2010.
There are three major moves here:

A2131) 11 Qxa7
A2132) 11 Qxc6
A2133) 11 a3!?

A2131) 11 Qxa7

Possible, but now Black will chase the white queen again. However, here instead of easy
development, Black should play for the more pressing ... d5-d4. Please be aware though that ... d5-d4 is
not an immediate threat as it can be met with the counterattacking a2-a3, chasing the bishop away.
11 ... Rb7! 12 Qa4 Ne7!
Renewing the possibility of ... d5-d4 and the e5-pawn will become a target later on.
12 ... Qc7 was my old recommendation and is still viable: 13 Bf4 Ne7! 14 a3 Bc5 15 Bd3 Rxb2 16 0-
0 0-0 and Black is okay.
13 a3
After 13 Bd2 d4 14 Ne4 Bxd2+ 15 Nxd2 Qd5 a pawn will fall, while 13 Be2 d4 14 a3 Ba5 15 b4
Bc7 16 Ne4 d3 17 Bf3 Bxe5 demonstrates our basic plan.
13 ... Ba5 14 b4 Bc7!

This time it is the bishop that attacks the e5-pawn.

15 Bf4
Holding on to the pawn. After 15 f4 0-0 16 Be3 Bb6! the white king is too exposed.
15 ... f6!? 16 Bd3 0-0 17 Qc2 Kh8!
Now ... fxe5 is coming and there is not much that White can do.
18 exf6 Bxf4 19 fxe7 Rxe7+ 20 Ne2 Qd6! 21 h3 Be5 22 Rc1 Ra7!
Again, White has to give up material to deal with his positional woes.
23 Qxc6
Alternatively, 23 0-0 Bd7 24 Qb3 Rfa8 and the a-pawn will fall.
23 ... Rxa3! 24 Qxd6 Bxd6
Black is much better on account of the bishop-pair.

A2132) 11 Qxc6+
 White’s most dangerous try.
11 ... Kf8!
A key move. The aim is to preserve the d5-pawn. I have to admit that this looks like outrageous play
by Black, compromising our king when behind on material. However, the position is much more complex
than just this. Our rook is well placed on b8 and White has the pressing matter of unravelling his c3-knight
from the pin. He can attack the b4-bishop or block the e1-a5 diagonal:

A21321) 12 a3
A21322) 12 Bd2

Please note this variation requires memorization and a constant need by Black to attack.
Instead, if 12 Be2 Ne7 13 Qa4 d4 14 a3 Ba5 15 b4 Bb6 16 Nd1 (the f2-pawn must be defended) 16 ...
d3! 17 Bf3 Bd7 18 Qb3 Be6 19 Qa4 Ng6 and now 20 Bb2?? would be a terrible blunder due to 20 ... d2+
21 Kf1 Bc4+ 22 Kg1 Qd3 23 h3 Bb5 and the queen is trapped.

A21321) 12 a3 Ne7!
Naturally we cannot let White win a second pawn. Indeed, the white queen needs to be kicked back.
13 Qa4 Ba5!
Preserving the pin.
14 b4 Qc7!
 A key move.
15 Bb2
If 15 Qb3 Qxe5+ 16 Be2 Bb6 17 Bb2 Bg4 18 f3 Bd7 19 Ne4 Qe6, but 15 Ne2 is White’s other
serious defence. Now Black should ignore the e5-pawn (this is important!) and play 15 ... Bb6! (15 ...
Qxe5? unfortunately fails to 16 Qxa5 Qxa1 17 Qd8#) 16 Bb2 (or 16 Qb3 Qxe5 17 Bb2 Qd6 18 g3 d4 19
Bg2 Ba6) 16 ... d4, seriously disturbing White’s hope of easy development. After 17 Qd1 (if 17 Rd1 Nf5
18 Qb5 a6 19 Qd3 Qxe5 and White is completely stuck) 17 ... Ba6 Black develops at the same time
threatening ... d4-d3, leaving White to sort out his development:

a) 18 Bxd4 Rd8 pinning the bishop to the queen.

b) 18 a4 d3! 19 Nf4 Bxf2+ 20 Kxf2 Qb6+ 21 Kg3 Nf5+ 22 Kh3 Qh6+ 23 Kg4 Qh4+ 24 Kxf5 Rxb4
and Black will force checkmate.
c) 18 Qa4 Bc4 19 Rd1 Rd8 20 Bxd4 Bxd4 21 Nxd4 Rxd4 22 Bxc4 Qxc4 23 Rxd4 Qxd4 24 0-0 g6
and Black has a material advantage.
d) 15 Bd2 Qxe5+ 16 Be2 Bc7 17 g3 g5! (forced) 18 h4 h6 19 hxg5 hxg5 20 Rxh8+ Qxh8 and it is
White’s king that will be weak: for example, 21 Qxa7 Rb7! 22 Qa4 Bb6 23 Rc1 d4 24 Nb1 Qh1+ 25 Bf1
Qe4+ and White will suffer in the endgame.
 15 ... Qxe5+ 16 Be2 Bc7!
Preventing White from castling kingside.
17 Nd1
White can also try to hold the position together with 17 Qd1 when a crazy line would run 17 ... a5 18
b5 d4 19 Na4 Nf5 20 Rc1 Bd6 21 g4 Nh4 22 Kf1 Bb7 23 Bxd4 Bg2+ 24 Kg1 Qe7.
17 Qb3 d4 18 Na4 Bd7 is also pretty complicated.
17 ... d4! 18 Ne3
This is White’s tactical idea, interfering with Black’s coordination.
18 ... Qf4
Keeping the mating threat alive so that White cannot castle into safety.
19 Nc2
After 19 Nc4 Qe4 20 Qb3 Qxg2 21 Bf3 Qh3 22 Bxd4 Be6 23 Rg1? Nf5 Black has defended against
the immediate threats, wrecked White’s king safety and regained the sacrificed material, while 19 g3?
fails miserably to 19 ... Qe4.
19 ... Bb6! 20 0-0 Bb7

Now the threat is ... Bc7.

21 Ne1
If 21 Bc1 Qd6 22 Rd1 Bc7 23 f4 Bb6 24 Bd3 h5 and suddenly the rook on h8 will come into play
with devastating effect after ... h4-h3, and the thoughtless 21 Rad1? loses to 21 ... Bc7 22 g3 Qe4.
21 ... g6 22 Bf3
22 Nf3 Nf5 23 Rad1 Bxf3 24 Bxf3 Bc7 is a nice trick to be aware of.
22 ... Kg7 23 Rd1 Nf5
Protecting the d-pawn.
24 g3 Qd6 25 Bxb7 Rxb7 26 Qb3 h5 27 Nf3
I think Black is doing okay. It is true the black king is on the same diagonal as the white bishop, but
Black’s pieces are rather actively placed to tie White down.
27 ... Re8 28 Rd3
28 Nd2 Rbe7 29 Nc4 Qd5 30 a4 Kg8 31 a5 Bc7 wouldn’t be a great strategy for White.
28 ... Kg8
Removing the king from the diagonal.
29 Rfd1 Re4 30 Nd2 Re2 31 Nc4 Qe6 32 Nxb6 Re1+ 33 Rxe1 Qxe1+ 34 Kg2 Qe4+ 35 Kg1 Rxb6
 This looks good for Black.

A21322) 12 Bd2

This is the main line and one often encountered over the board, but it is not hard for Black to handle.
12 ... Ne7!
Attacking the queen.
13 Qa4 Qc7
And now attacking the e5-pawn.
14 f4
Should White try to simplify the position and take advantage of the strange black king position with 14
Be2 play might go 14 ... Bd7! (chasing away the queen one more time; if 14 ... Qxe5 I am worried that
White can get away with 15 0-0 Bxc3 16 Bf4 Qxe2 17 Bxb8 Bf6) 15 Qa6! (the ultra-cautious 15 Qd1 is
met with the direct 15 ... d4 16 Nb1 d3!! 17 Bxd3 Qxe5+ 18 Qe2 Qxb2 19 Bxb4 Qc1+ 20 Qd1 Qxd1+ 21
Kxd1 Rxb4 and the irony of it all is that Black is better on account of the position of the white king) 15 ...
d4 16 Ne4 Qxe5 17 Bxb4 Rxb4 18 Qxa7 Bb5! 19 Qa8+ Be8 20 f3 Rxb2 21 Qa3 d3 22 Bxd3 Rxg2 23
Rb1 h5 with a mess all over the board.
14 ... Bd7 15 Qd1 Ba5!
 Attacking the b2-pawn. White is so uncoordinated for the time being that he cannot react to this in a
sensible way.
16 Qc1
If 16 Rb1 Bf5!.
16 ... Qb7 17 Bd3
Alternatively, if 17 b4 Bxb4 18 Rb1 d4! 19 a3 a5 20 axb4 dxc3 21 Bxc3 Qe4+ 22 Kf2 Nd5! 23 Qd2
axb4 with a complicated position and 17 Be2? doesn’t work on account of the thematic 17 ... d4 18 Bf3
dxc3!.
17 ... d4!
Carrying on attacking.
Also reasonable is 17 ... Rc8 18 Qd1 Qxb2 19 Rb1 Qa3 20 Rb3 Qc5, keeping control over the weak
dark squares and now:
a) 21 Ba6 Rd8 22 Qc1 Bc6? (after 22 ... Bb6! 23 Rf1 Nf5 I don’t think White can claim an advantage,
particularly should he allow Black to play ... h7-h5, consolidating the outpost) 23 Qa3 was slightly better
for White in R.Edouard-N.Stewart, Hastings 2011/12.
b) 21 Qe2 h5 22 Ba6 Rc7 23 Qf2 d4 24 Nd1 Bxd2+ 25 Qxd2 Rh6 26 Bd3 Rb6 and Black has
successfully developed his pieces, which means he has solved all his problems.
18 Ne2
If 18 Ne4 Bf5 19 Bxa5 Bxe4 20 Qd1 Bxd3 21 Qxd3 Qxg2 when Black has recovered the pawn,
standing better after 22 Rf1 Qd5! 23 Bd2 Rxb2.
18 ... Qxg2 19 Rg1 Qf3 20 Rg3 Qh1+ 21 Rg1 Qf3 22 Bxa5 Qxd3 23 Qd2 Qxd2+ 24 Bxd2 Nc6 25
b3 f6 26 Rd1
Both sides have equal chances in this complex late middlegame.

A2133) 11 a3!?
 This was Richard Palliser’s recommendation in Dangerous Weapons: The Caro-Kann. The idea is
not to go for a direct refutation, but to obtain a stable edge.
11 ... Bxc3+
Forced, as if 11 ... Ba5 White can now play 12 Qxc6+ Bd7! (or 12 ... Kf8 13 b4 Bc7 14 Nxd5 Bxe5
15 Be3 and White is in time to repel Black’s activity and begin his own deadly counterattack) 13 Qxd5
Ne7 14 Qd3 Qc7 15 b4 Bb6 16 Bb2 Qxe5+ 17 Qe2 Qf5 18 g4 Qg5 19 Ne4.
12 bxc3 Qc7! 13 Bf4
Rather unusually the material is level, but White has the positional advantage of the bishop-pair which
Black must neutralize in order to claim equality.
Instead, 13 Qd4 Ne7 14 e6 Nf5 is at least okay for Black.
13 ... Be6 14 Bd3 Ne7 15 0-0 0-0 16 Qc2

Visually this looks like an easy game for White. After all he has the bishop-pair and if things become
too quiet for his liking, he can push f4-f5 or even c3-c4. In order to prevent this from being the case Black
needs to swap a set of bishops or find a good post for his knight.
16 ... Kh8!
The key move and one that ensures Black of equality.
 16 ... h6 was seen in M.Erdogdu-G.Timoshenko, Kalamaria 2008, but this weakens the kingside and
after 17 Be3! c5 (otherwise the bishop would land on c5) 18 f4 g6 (now the h6-pawn is weak, but 18 ...
c4?! 19 Bh7+! Kh8 20 f5 is dangerous) 19 Qf2 White’s dark-square pressure should not be
underestimated.
Black might also have met the threat to h7 with 16 ... Ng6!?, but then comes 17 Bg3 c5 18 c4! and the
problem is that Black can’t use the knight hop to d4.
17 Bg3
If 17 Be3 c5!. White’s bishop must not be allowed to entrench itself on c5 and Black plans the
thematic ... c5-c4, chasing away the d3-bishop with the aim of empowering our own:
a) After 18 c4 Black should play the surprising 18 ... dxc4 19 Bxc4 Nf5 and the e3-bishop has to
move.
b) 18 Qd2 Bf5 19 Bxf5 Nxf5 20 Qxd5 Nxe3 21 fxe3 Rbd8 22 Qe4 Rde8 23 Rab1 Re7! (not 23 ...
Rxe5 24 Rb7!) 24 Rfd1 Rfe8 25 Rd5 h6 and Black will regain the pawn on e5.
c) 18 f4 c4 19 Be2 Bf5 20 Qd2 Be4! when Black should play to double on the b-file.
17 ... c5

18 c4
After 18 f4 we should, of course, chase away the light-squared bishop: 18 ... c4 19 Be2 Bf5 and
Black will obtain the desired good bishop and good knight.
18 Qd2 is another way of neutralizing the ... c5-c4 pawn push, but this means that White abandons
some control over the b1-h7 diagonal. Black should play 18 ... Qa5 19 f4 c4 20 Bc2 Bf5 21 Bxf5 Nxf5 22
Bf2 Rb3 23 g4 Ne7 24 Rfc1 Nc6 when he is holding the position and more thanks to the active b3-rook.
18 ... dxc4!
With the idea to bring a knight round to d4.
19 Bxc4 Nf5!
The knight now gets the wonderful d4-square to sit on, whereas the white bishop is left out of play on
g3.
20 Bxe6 Nd4 21 Qc4 fxe6 22 Rac1 Qb7 23 Rfe1 h6 24 Qc3 Qa6
Black is better. White’s bishop is shut in by his e5-pawn while all of Black’s pieces are active and,
importantly, can be improved further.

A22) 7 ... Bc5

 A solid, less theoretical alternative to 7 ... Bb4.
8 Nxc6
Black practically forces White to exchange a set of knights, as if 8 Be3 Qb6 9 Na4 Qa5+ 10 Bd2 Bb4.
8 ... bxc6
We now come to a parting of the ways. White can try to punish Black by attacking the weak g7-pawn
or he can relinquish the central tension:

A221) 9 Qg4
A222) 9 cxd5

A221) 9 Qg4 Bf8

A very solid option. Black has the superior pawn structure and his plan is simple, to play ... Ne7-g6. Our
model expert in this line is the British Grandmaster Keith Arkell.
I must also mention that the pawn sacrifice 9 ... Ne7!? is possible.

10 Bd3
Also critical is 10 cxd5 when Black must tread carefully and play 10 ... cxd5 (10 ... exd5 gives White
a slight pull on account of his being able to blockade the c6- and d5-pawns) 11 Bg5! h5! (11 ... Ne7 fails
to 12 Nb5!). Now:
a) Again 12 Qg3 requires some accuracy from Black: 12 ... h4! 13 Qg4 Be7 14 Bxe7 Nxe7! 15 Qxg7
Rg8 16 Qf6 Rb8! (forcing White to seek refuge on the queenside) 17 0-0-0 Qc7! 18 Be2 Qc5! with a
messy game ahead. A sample line might run 19 Bh5! (after 19 Qxh4 Rxg2 20 Rhf1 Bd7 21 Rd2 Ng6 22
Qf6 Rb4! 23 Rc2 Rf4 White is in serious trouble) 19 ... Rf8 20 Qxh4 Rb4 21 Qg3 d4 22 a3 Ra4 23 Qd3
Ba6 24 Qc2 Rc4 25 Be2 dxc3 26 Bxc4 cxb2+ 27 Qxb2 Bxc4 28 Kb1 Bd5 29 Rc1 Qb6 30 Qxb6 axb6 31
f3 Nc6 with an unbalanced endgame.
b) 12 Qh4 Be7 13 Bxe7 Nxe7 14 Nb5 Nf5 is the point and here Black has a promising position.
c) 12 Qa4+ Bd7! 13 Bb5 (the endgame doesn’t promise White too much after 13 Bxd8 Bxa4 14 Ba5
Bc6 15 Nb5 Rb8 16 a4 Kd7 17 Nxa7 Bb4+ 18 Bxb4 Rxb4 19 Nxc6 Kxc6 20 Bd3 Ne7 21 Kd2 Rxb2+ 22
Ke3 Rhb8, with an equal game) 13 ... Be7! 14 Be3 Nh6 15 0-0 a6 16 Bxd7+ Qxd7 17 Qxd7+ Kxd7 18
Na4 Rab8 19 Nb6+ Ke8 20 Rac1 Nf5 21 Bc5 f6! (getting ready to mobilize the central pawns) 22 Bxe7
Nxe7 and Black is doing fine.
10 ... Ne7
Black continues with the natural plan of putting a knight on g6.
11 0-0
After 11 h4 Ng6 12 Qe2 Qc7 13 f4 Bc5 14 h5 Ne7 15 cxd5 exd5 Black should be fine, although with
both kings slightly vulnerable there is a lot of risk involved for both sides.
11 ... Ng6

12 f4
The most aggressive move, if far from forced:
a) 12 Qe2 is a more considered choice for White, but the attraction for Black in this system lies in its
simplicity - Black has the same plan throughout. Thus we develop with 12 ... Be7 13 f4 0-0 14 cxd5 (14
Bd2 Ba6!? pressures the c4-pawn and another plan is to play ... d5-d4 followed by ... c6-c5 and ... Bb7,
while after 14 Kh1 d4 15 Ne4 c5 16 b3 a5 17 Ba3 Qb6 Black has no problems) 14 ... cxd5:
 a1) 15 Be3 looks flashy, but Black can play that game too: 15 ... d4! (not the 15 ... Qa5 of Y.Zhou-
K.Arkell, Brighton 2011, as after 16 a3 White was able to secure a grip on the dark squares) 16 Bxg6 (or
16 Rad1 dxe3 17 Bxg6 Qb6 18 Bd3 Bb7 19 Na4 Qb4 20 b3 Bc6 21 Qxe3 Bxa4 22 Qh3 g6 23 bxa4 Qxa4
and Black’s queen keeps a watchful eye on all the relevant parts of the board) 16 ... dxe3! 17 Be4 Rb8 18
Rfd1 Qb6 19 b3 f5 20 exf6 Bxf6 and the open position will favour the two bishops.
a2) 15 Nb5 d4! (not allowing White to set up a blockade on the d4-square) 16 Qf2 Bb7 17 Nxd4 Qd5
18 Be3 Nh4 and suddenly the long light-square diagonal will give White a tremendous headache.
a3) 15 Kh1 d4 16 Ne4 Bb7 17 b3 a5 18 Bb2 a4 also gives pleasant play for Black. Indeed, after 19
Rad1 axb3 20 axb3 Ra2 21 Bb1 Ba6 22 Qd2 Rxb2 23 Qxb2 Bxf1 24 Rxf1 Qd5 he has all the chances.
b) 12 Qh5 is another way of defending the e5-pawn to which Black should react with 12 ... Be7,
preventing the white bishop developing to g5 with tempo. Now:
b1) After 13 f4 0-0 (Black should keep calm and let White see what he can do; our expert chose 13 ...
Bc5+ 14 Kh1 Qh4 in L.Trent-K.Arkell, British League 2011, but this wasn’t necessary) 14 Rd1 Qb6+ 15
Kh1 Ba6 16 cxd5 cxd5 and Black has no problems.
b2) 13 Bxg6 fxg6 14 Qf3 Qc7 15 Re1 (or 15 cxd5 exd5 16 Re1 Bf5 and Black will castle) 15 ... Rf8
16 Qg4 Rf5 and Black can castle by hand without any real problems: for example, 17 Bf4 Rb8 18 b3 Kf8
is equal.
12 ... Bc5+ 13 Kh1 0-0 14 Na4 Be7 15 Be3 d4 16 Bd2 c5 17 Be4 Rb8 18 b3 Bb7 19 Bxb7 Rxb7 20
Nb2 Qd7
20 ... Nh4!? preventing f4-f5 also comes into consideration.
21 Nd3 Qc6
 22 f5
22 Rf2 is too single-minded. Black should now begin immediate counterplay with 22 ... a5 23 Nb2 (if
23 Raf1 a4, while 23 Bxa5 Ra8! is the point) 23 ... Rfb8! 24 Qe2 Bd8 (Black will play to manoeuvre a
knight to f5) 25 f5? exf5 26 Rxf5 Re7 27 Raf1 f6 when he has the advantage.
22 ... exf5 23 Rxf5 a5! 24 Raf1 a4 25 bxa4 Qxa4
25 ... Ra7 keeps better control over the e6-square.
26 e6 fxe6 27 Rxf8+ Nxf8 28 Qf3 Qa8! 29 Ne5
After 29 Qf7+ Kh8 30 a4 Bf6 31 Qh5 Qxa4 32 Qxc5 Kg8 33 Bb4 Rf7 White is getting nowhere with
his initiative.
29 ... Qc8
29 ... Bf6! is the most accurate try and after 30 Ng4 Nd7 31 Qc6 unbelievably Black should play 31 ...
Kf7!. A sample line runs 32 Re1 e5 33 Qd5+ Kf8 34 Qd6+ Kg8 35 Nxf6+ Nxf6 36 Qxe5 Rb2, winning.
30 a4?
Correct was 30 Qf7+ Kh8 31 Qh5 Rb2 32 Nf7+ Kg8 33 Nh6+ Kh8 34 Nf7+ with a repetition (34
Qf7? loses to 34 ... Ng6).
30 ... Bf6 31 Ng4 Rf7 32 a5 Qb7 33 Qe2 e5 34 Nf2 Bh4 35 Nd3 Qe4 36 Qd1 Qxd3 37 Rxf7 Kxf7
38 a6 Nd7 39 a7 Nb6 40 g3 Bg5 41 Qh5+ Kf8 42 Kg1 Bxd2
0-1, A.Hunt-K.Arkell, London 2011.

A222) 9 cxd5 cxd5!?

 A brave continuation. Black battles to keep the pawn structure intact. 9 ... exd5 is another possibility,
but Black may have to face White blockading the hanging pawns, which is something that should give him
a small edge.
10 b4
After 10 Bd3 Black continues with the normal plan of developing his knight: 10 ... Ne7 11 0-0 Ng6 12
Na4 Be7 13 f4 0-0 14 b3 Bd7 and White can’t claim to have any advantage.
Instead, 10 Qg4 Bf8 transposes to the notes to White’s 10th move in Line A221. Here I would also
like to draw the reader’s attention to the dangerous gambit 10 ... Ne7!? 11 Qxg7 Rg8 12 Qxh7 Rb8! 13
Qc2 Nc6 14 Bf4 Rg4 with powerful compensation in S.Haslinger-J.Houska, Liverpool 2008.
10 ... Bf8
Obviously not 10 ... Bxb4? 11 Qa4+, while 10 ... Be7 interferes with the development of the knight,
but it may also be possible.
11 a3
A recent try, rather than:
a) 11 Bd3 Qc7! 12 Bb2 (or 12 Bb5+ Bd7 13 Bxd7+ Qxd7 14 a3 Ne7 15 0-0 Ng6 16 f4 Be7 and Black
will take control over the open c-line) 12 ... Qxe5+ 13 Be2 Nf6 14 Nb5 Qxb2 15 Nc7+ Kd8 16 Nxa8
Bxb4+ 17 Kf1 Ne4 and White has nothing to play for.
b) 11 Bd2 and now very importantly Black should now play 11 ... Ne7!, which means that White
doesn’t get the chance to entrench a bishop on a6:
b1) 12 Bd3 Nc6 is fine for Black.
b2) 12 Nb5 Nc6 is an only move, as Black needs to cover the important c-file (12 ... Ng6? loses on
the spot to 13 Qc2). After 13 Rc1 Bb7 14 Qa4 Qd7 15 Bd3 (15 Be2 a6 16 Nd6+ Bxd6 17 exd6 Qxd6 18
b5 axb5 19 Qxb5 Rb8 20 0-0 0-0 leads to nothing) 15 ... a6 16 Nd6+ Bxd6 17 exd6 Ne5 18 Qxd7+ Kxd7
19 Be2 Rac8 the d6-pawn is doomed and Black is much better.
b3) 12 Rc1 threatens the devastating Nb5. Black must intercept the white attack down the c-file with
12 ... Bd7!:
 b31) 13 Ba6 Qb6 14 Qe2 Nc6 15 Na4 Qb8 16 Nc5 (of course, not 16 f4? Nxb4 when there are too
many loose pieces) 16 ... Bxc5 17 bxc5 Qxe5 18 Qxe5 Nxe5 19 0-0 Bc6! 20 Bc3 f6 21 Rfe1 Kf7 22 Bxe5
fxe5 23 Rxe5 Rhe8 with equality, although Black can look forward to rounding up that weak c5-pawn.
b32) 13 Bb5 puts the wrong piece on the b5-square and now 13 ... Ng6! 14 Bg5 Be7 is a completely
safe way of playing things.
b33) 13 Na4 is again the wrong plan. Black continues as normal with 13 ... Ng6 14 Nc5 Bxc5 15
Rxc5 Nxe5 16 Bc3 f6 17 Qh5+ g6 18 Qh6 Kf7 when he will soon play ... Rac8 and stands well.
b34) 13 Bd3 Ng6 14 Nb5 Rc8! 15 f4 a6 16 Nd6+ Bxd6 17 exd6 0-0 18 Bxa6 Rxc1 19 Qxc1 Qb6 and
White has quite a few significant weaknesses in the camp. Here 20 Bd3 Rc8 21 Qb1 Qd4 22 g3 e5 is an
example of how quickly things can become unpleasant for him.
b35) 13 Nb5 Nc6! 14 a3 a6 15 Nd6+ (if 15 Nc3 Nxe5!) 15 ... Bxd6 16 exd6 0-0 17 Bd3 Qf6 18 0-0
e5 and White will have some compensation, but Black has plenty of trumps too in the connected central
pawns and good piece play.
11 ... Ne7

12 Bg5
12 Nb5!? is a tricky move that requires careful handling. The threat is Qg4 followed by Nd6+, so we
go 12 ... Nc6! and then:
a) 13 Qc2 Bb7 14 Bb2 Be7 15 Nd6+ Bxd6 16 exd6 f6 blunts a key diagonal.
b) 13 Bb2 a5 (13 ... a6!? is also possible) 14 Rc1 Bd7 15 Nd6+ Bxd6 16 exd6 axb4 17 axb4 0-0 and
White is struggling.
12 ... a6!
This is forced as otherwise Nb5 is incredibly unpleasant.
13 Rc1
We have been following G.Jones-K.Arkell, Douglas 2014, and here I believe Black should play:
13 ... h6!? 14 Qh5
This has to be the most critical move. After 14 Bh4 g5 15 Bg3 Bg7 Black will simply play ... Bd7
followed by ... 0-0 and neither is 14 Bxe7 particularly challenging: 14 ... Bxe7 15 Na4 0-0 16 Nc5 a5 and
Black is doing very well. Finally, 14 Bd2 Ng6 15 f4 Be7 16 g3 0-0 17 Na4 Bd7 18 Nc5 Bxc5 19 Rxc5
Bb5 is also okay for Black.
14 ... Bd7 15 Bd3
After 15 Be2 Qb8 16 Bf4 (16 f4? Ng6 wins) 16 ... g6 17 Qh3 Bg7 White must give up the pawn for an
initiative.
15 ... g6!

Black should force the white queen to somewhere less threatening on the h-line.
16 Qh4
Alternatively, 16 Qh3 Bg7 17 Bf6 Bxf6 18 exf6 Ng8 19 0-0 Nxf6 20 Rfe1 Kf8 with the better
position.
16 ... Bg7 17 Bf6 g5 18 Qh5
Sacrificing the queen isn’t too good either because of 18 Bxg7 gxh4 19 Bxh8 Ng6 20 Bf6 Qb6 21 0-0
Nf4 followed by ... h4-h3 and ... Rc8.
18 ... Bxf6 19 exf6 Ng8 20 0-0 Nxf6 21 Qe2 Kf8
I believe this is key. Black needs to castle by hand to alleviate the danger.
22 f4
Black defends after 22 g3 Kg7 23 f4 Qb6+ 24 Kh1 g4 25 Qe5 Rhc8 26 f5 Qb8.
22 ... Qb6+ 23 Kh1 Qd4 24 f5
24 fxg5 hxg5 leaves White to worry about the counterattack along the h-file.
24 ... Kg7
 Black’s position should be very solid and White still has to justify the pawn deficit.

B) 4 c4!?

In my opinion this move order is inaccurate if White intends to follow up with 5 Nf3.
4 ... cxd4
Now we have:

B1) 5 Nf3?!
B2) 5 Qxd4

Instead, 5 Ne2! forces 5 ... Nc6 6 Nxd4 e6, transposing to Line A2.

B1) 5 Nf3?! Bg4!

The key idea. The bishop is developed to a good square and White has to start resolving the tension in the
centre.
6 cxd5 Qxd5 7 Nc3 Qa5!?
 An uncommon choice, but a good one nonetheless.
8 Bb5+
8 Qxd4 is possible, but allows 8 ... Bxf3 9 gxf3 Nc6 10 Bb5! (if 10 Qe3 e6! 11 Bd2 0-0-0 12 f4 Nh6
13 Bg2 Nf5 and Black has no problems) 10 ... Qxb5 11 Nxb5 Nxd4 12 Nxd4 e6! (12 ... 0-0-0 is not good
enough: 13 Be3 e6 14 Nb5 Bb4+ 15 Kf1 a6 16 Rc1+ and although Black later won, he was in trouble in
P.Reinwald-I.Pejic, Rijeka 2010) 13 Be3 a6 14 Rc1 Ne7 (an incredibly important move; Black needs to
solidify the position by establishing a knight on d5) 15 Rc7 Rb8 followed by ... Nd5 cementing Black’s
position together.
8 ... Nc6 9 Bxc6+ bxc6 10 Qxd4 Bxf3 11 gxf3 e6 12 Be3
Alternatively, 12 Bd2 Rd8 13 Ne4 Rxd4 14 Bxa5 Rd5! (Black needs to force the bishop to retreat to
c3 to prevent White from playing Nc3, dislodging our powerful rook on d5) 15 Bc3 Nh6 16 Ke2 Be7 17
Rac1 Nf5 18 f4 0-0 with a solid position.
12 ... Ne7 13 0-0-0 Qc7!
Black does not fully equalize after 13 ... Nd5 14 Qc4 Rc8 15 Nxd5 exd5 16 Qg4 Rb8 17 e6 f6 18 Kb1
Ba3 19 b3 0-0 20 e7.
14 Qc4 Nf5!

Covering the d6-square.

15 Nb5 Qb7 16 Nd4 Rb8 17 Qc2
17 Rd2? is already a mistake because of 17 ... Bb4 18 Rc2 Nxe3 19 fxe3 c5 and now White loses
material: 20 Nxe6 fxe6 21 a3 (21 Qxe6+ Qe7 22 Qc6+ Qd7 23 Qa6 0-0 is the safest route to a small plus
for Black: for example, 24 a3 Qh3 25 axb4 Qxf3 26 Re1 cxb4 and White will suffer with such an exposed
king) 21 ... 0-0 22 axb4 Qxf3 23 Re1 cxb4 24 Qxe6+ Kh8 25 Qc6 Qf5 26 e6 b3 27 Rc3 Qf2 (27 ... Qa5!
would have won on the spot) 28 Rd1 Rbd8 29 Rxd8 Rxd8 30 Qd7 Qe1+ 0-1, Z.Jovanovic-R.Zelcic,
Marija Bistrica 2011.
17 ... Nxe3 18 fxe3 Bc5 19 Rd3 Bxd4 20 exd4 0-0 21 Rhd1 Qd7 22 Rc3 Rb4 23 Qe4
Here I believe Black should let White stew a little bit in his own mess.
23 ... Rc8!
23 ... Rfb8 24 Qxc6 Qxc6 25 Rxc6 Rxb2 26 Rc8+ eventually led to a draw, but was too simple for
White to defend in F.Vallejo Pons-A.Delchev, Plovdiv 2010.
24 Rc5 g6 25 Rd3 Qd8 26 b3 a5 27 a3 Rb5
White doesn’t have a simple plan to prove equality.
B2) 5 Qxd4 e6
This is rather simple.
Also fully viable is 5 ... Nc6!? 6 Qxd5 (White must capture as something like 6 Qd3?! d4 would only
see Black seizing the initiative at no material cost) 6 ... Bf5!? 7 Nf3 (or 7 Qxd8+ Rxd8 8 Nf3 e6 9 Be3
Bb4+ 10 Nbd2 Nge7 11 0-0-0 Ba5 and Black should win back the pawn on e5) 7 ... e6 and then:
a) 8 Qb5? Bc2! 9 Nfd2 (9 Nc3? a6 10 Qxb7 Na5 traps the queen) 9 ... Qc7.
b) 8 Qxd8+ Rxd8 9 Be3 Nge7 10 Nc3 Ng6 11 h3 Bb4 12 a3 Bxc3+ 13 bxc3 Bd3 and Black will win
back the pawn on e5.
6 cxd5 Nc6!

This is the point, developing with tempo.

7 Qa4
7 Bb5 is natural, but Black has nothing to fear after 7 ... Qxd5 8 Qxd5 exd5 9 Nc3 Bb4 10 Bd2 Nge7
11 a3 Ba5 12 Nf3 (or 12 Nge2 0-0 13 0-0 Bg4) 12 ... Bg4, which is the easiest way for him to achieve
comfortable development.
7 Qf4 doesn’t pose problems either after 7 ... exd5 8 Nf3 Bc5:
a) 9 Bd3 Nge7 10 0-0 0-0 11 Nc3 and now Black can remove the e5-pawn clamp with 11 ... f6! 12
exf6 (12 Qh4 is met with 12 ... Bf5 13 Bxf5 Nxf5 14 Qh3 fxe5 and Black is better) 12 ... Rxf6 13 Qh4 h6
14 Bd2 (attacks such as 14 Ng5 don’t work due to 14 ... Ng6 and 14 Bg5 also looks scary, but is met
easily enough: 14 ... Rxf3! 15 gxf3 Ne5 and now White either has to keep the threat of mate alive or
abandon the extra material as after 16 Bxe7? Bxe7 Black wins) 14 ... Be6, reaching a strange position.
White looks to have a lot of trumps, but the awkward position of the knight on c3 and the bishop on d2
means that Black is doing okay: for example, 15 Rac1 Nf5! 16 Bxf5 Bxf5 and should White try to offset
the pressure along the f-file, Black can play 17 Be3 d4 18 Ne4 Bxe4 19 Qxe4 Bb6 20 Rfd1 Qd7 and now
21 Nxd4? would be a terrible mistake because of 21 ... Rd6.
b) White cannot play the tricky 9 Qg3? as Black can simply ignore the threat: 9 ... Nge7! 10 Qxg7 Rg8
11 Qxh7 Bf5 12 Qh4 Qb6 13 Nc3 0-0-0 and White cannot develop his pieces.
7 ... exd5 8 Nf3
8 Bb5 is also natural, but Black will simply respond with 8 ... Bd7 9 Nf3 a6 10 Bxc6 bxc6 when he
doesn’t have any serious problems. A knight will come to f5 and should White not do anything, Black can
use the b-line and the mobility of his central pawns to powerful effect.
 8 ... Bc5 9 Nc3 Nge7 10 Bd3 d4!

Now is the time for this pawn push, preventing the white queen swinging over to the kingside.
11 Ne4 Bb4+ 12 Bd2 Bxd2+ 13 Nexd2 0-0 14 0-0 Bg4 15 Rac1 Ng6 16 Rfe1 Re8 17 Bb5 Bxf3 18
Nxf3 Ngxe5 19 Nxe5 Rxe5 20 Red1 Qb6
The position is completely level and play may continue like this:
21 Bxc6 bxc6 22 Qxc6 Qxc6 23 Rxc6 Re2 24 b4 Rxa2 25 g3 g6 26 Rxd4 a5 27 b5 Rb2 28 b6 a4
29 Rd7 Rb8 30 Rcc7 a3 31 Rxf7 a2
White is forced to take the repetition.

C) 4 c3?!

White tries to maintain the pure pawn structure normally achieved in the Advance Variation of both
the French and the Caro-Kann. The problem with this approach here is that the placing of Black’s light-
squared bishop has not been decided yet. Unfortunately for White, there are no lines in this variation
which promise him a small positional advantage and, indeed, Black will obtain promising counterplay.
4 ... Nc6
Placing the d4-pawn under a little bit of pressure. In order to develop his pieces, White must first
resolve this issue.
5 Nf3
If White tries to be more sophisticated and to prevent the bishop developing to its optimum g4-square
with 5 Be3 then Black should continue 5 ... Qb6! 6 Qd2 cxd4 (the c5-pawn was under threat) 7 cxd4 Bf5
8 Nc3 e6 9 Nf3 Nge7 10 Be2 Bg4!.

In these closed positions it is useful to know that the knight is worth more than the bishop. In this
specific position we can see that both the black knight and bishop are fighting for the f5-square, but it is
clear that the knight will do a better job on that square. From f5 the knight will put more pressure on d4
and crucially unblock the way for the dark-squared bishop on f8. After 11 0-0 Nf5 12 Rfd1 Be7 13 Na4
Qd8 14 a3? (not understanding the positional subtleties I create an important weakness on the queenside)
14 ... Bxf3 15 Bxf3 Na5! 16 Qd3 0-0 17 Nc3 Nc4 Black soon won in J.Houska-H.Hunt, British League
1998.
5 ... cxd4
This could very well be the simplest move and the easiest one for Black to play. The only downside
to this move is that White gets to put a knight on c3 in one go, but we can’t have everything.
5 ... Bg4 was my old line, but then 6 dxc5 transposes to quite a critical line of the 4 dxc5 Nc6
variation.
6 cxd4 Bg4!
Of course, all the same rules still apply.
7 Bb5
Natural play results too from:
a) 7 Nc3 e6 8 Be2 (8 Qb3 is just bluff: 8 ... Bxf3 9 Qxb7 Qc8 10 Qxc8+ Rxc8 11 gxf3 Nxd4 12 Ba6
Rc6 13 Bd3 Bb4 and Black stands well) 8 ... Qb6 9 Be3 (9 0-0 Bxf3 10 Bxf3 Qxd4 11 Qb3 Qb4 wins a
pawn without too much danger) 9 ... Bxf3 10 gxf3 (10 Bxf3 allows Black to play the risky-looking 10 ...
Qxb2 11 Rc1 Bb4 12 Qd3 Nge7 13 0-0 0-0 14 Na4 Qa3; however, not only does the Black queen escape,
but Black remains a comfortable pawn up) 10 ... Nge7 11 Bd3 g6 12 Na4 Qa5+ 13 Nc3 Nf5 14 Bxf5 gxf5
and this an excellent Advance Caro. Black has no weaknesses, can play for control over the c- and g-
files, and will attack the d4-pawn.
b) 7 Be2 e6 (Black should simply continue with normal development) 8 0-0 Qb6 9 Nbd2 Nh6 10 Nb3
Nf5 11 Be3 Be7 12 Ne1?! (12 h3 Bxf3 13 Bxf3 a5 leads to a perfectly good Caro-Kann; Black has strong
pressure on d4 combined with play along the c-file) 12 ... Bxe2 13 Qxe2 0-0 14 Nf3 (14 g4 Nxe3 15 fxe3
a5 is again comfortable for Black) 14 ... Rfc8 15 g4 Nxe3 16 fxe3 Nb4 17 Nbd2 Rc2 18 a3 Nc6 19 b4
Nd8 20 Rfc1 Rac8 21 Rxc2 Rxc2 with the superior position for Black in A.Shaaban-I.Rausis, Cairo
2002.
c) 7 Be3 e6 8 a3 Nge7 9 h3 and now Black plays 9 ... Bxf3! 10 Qxf3 Qb6 when White does not have
an easy task defending d4 and b2.
7 ... Qb6 8 Nc3 e6 9 0-0
9 Qa4 is not satisfactory after 9 ... Bxf3 10 gxf3 Nge7 11 Be3 Rc8 12 Rb1 Nf5 13 Ke2 a6 14 Bxc6+
Rxc6 15 Rhc1 Be7 16 Nd1 0-0 17 Rxc6 bxc6 18 Rc1 Rc8 19 Rc3 a5 20 a3 h6 21 f4 g6 22 Rb3 Qa6+ 23
Rd3 c5 with a perfect Caro-Kann, V.Mrva-R.Briestensky, Tatranske Zruby 2011.
9 ... Nge7 10 Be3 Nf5

11 h3 Bh5
11 ... Bxf3 is obviously the ideal move, but after 12 Qxf3 Black has to take measures to protect the d5-
pawn because of the loose knight on f5.
12 g4 Nxe3 13 fxe3 Bg6 14 Nh4 Bb4!
After 14 ... Be7 White has the dastardly 15 Nf5.
15 Bxc6+ Qxc6 16 Rc1 0-0 17 Qb3 Bxc3 18 Rxc3 Qa6 19 Rc7
Or 19 Nxg6 fxg6 and White’s kingside weaknesses will begin to tell.
19 ... Be4 20 Qxb7 Qe2 21 Rcxf7 Qxe3+ 22 R1f2 Rfc8
Now it is White who will have to be careful.
23 Nf3
Not 23 Rxg7+?? Kh8 when White is helpless against Black’s mating threats on the back rank.
23 ... Kh8 24 Qe7 Rc1+ 25 Kg2 Rg8 26 Rf8 h6
White’s attack gets nowhere.

Summary
4 Nf3 has become one of the ‘hot’ moves. Play can become incredibly complicated and it is for this
reason I am recommending two systems. After 4 ... Nc6 5 c4 cxd4 6 Nxd4 e6 7 Nc3 my original 7 ... Bb4
is still viable, but it leads to complicated and messy variations. Be warned: memorization and accuracy is
required. The most challenging lines involve accepting the c6-pawn, but only after White has captured on
d5 first.
The safe alternative of 7 ... Bc5 aims to keep to a formulaic approach, relying on the ‘good’ structure
to hold the position. As Black will be lagging in development for a few moves, his priority is to
neutralize White’s initiative.
Instead, 4 c4 cxd4! 5 Nf3 is simply inaccurate as Black can play 5 ... Bg4! targeting the weak e5-
pawn. Neither is 5 Qxd4 advantageous, provided Black accepts playing with an IQP.
 Chapter Twelve
The Fantasy Variation
1 e4 c6 2 d4 d5 3 f3

A few years ago, in a quandary about what to play against Grandmaster Keith Arkell, I decided to try
out the Fantasy Variation in the hope that I could draw him away from his beloved endgames and into a
tactical mess. My approach didn’t work at all: not only did I not get into a tactical position, I lost horribly
in an endgame. However, what made the biggest impact on me was that during the post-mortem Keith said
he felt that playing the pawn to f3 was a serious positional concession: White hampers the development of
his kingside pieces and simply makes the dark squares around his king very weak. I have to say that I
agree with the gist of this simple view with, although I acknowledge that the reality is complicated. There
are an awful amount of open lines which can make for a menacing game.

The Pros and Cons of f2-f3

Despite first appearances, 3 f3 does have some positional basis to it. After all, top grandmasters have
played the Fantasy Variation and one can safely say that a strong grandmaster rarely plays positionally
suspect openings. With f2-f3, White bolsters the centre with pawns and gets ready to open the f-file for his
rooks to attack à la King’s Gambit. The Fantasy Variation has an extraordinarily sharp character and is
often accompanied by sacrifices and dangerous traps. It does, however, have a clear downside: the pawn
on f3 hampers White’s development and the Achilles’ heel of White’s position is the f2-square and the
weakened g1-a7 diagonal. As such, Black should adopt a dark-square strategy.

My Recommendation
Continuing with the spirit of the first edition, I will again be recommending:
3 ... dxe4 4 fxe4 e5 5 Nf3 Bg4
 The position is dynamic, open and critical. In some ways Black gives White exactly what he wants
from the opening - the open f-line, active pieces and an insane amount of tension in the centre. However,
in return White has given Black active pieces, easy development and targets in the centre. This
provocative choice asks White: well, how are you going to come and attack me?
As things stand after 5 ... Bg4, as soon as White advances a pawn or swaps in the centre then the
structural advantage will transfer to Black. On the other hand, it is usually preferable for White if he can
force Black to exchange on d4. This means that both sides will try to bring the central tension to breaking
point - each forcing the other to commit to an exchange or to advance a pawn.
White’s attacking plans generally focus on these particular points:
a) The open f-line - White will aim to transfer the queen to the kingside or land a knight on the
dangerous f5-square.
b) Pressure the centre so much with moves such as Nd2-c4 that Black will cave in and play ... exd4.
In fact the recommendation of a recent repertoire book on the Fantasy Variation solely focuses on this
plan.
Black will also need to be alert to the following piece manoeuvres:
a) Nd2-Nc4, pressuring the bishop on d6 and the pawn on e5.
b) Bc1-Bg5 followed by Qe1-Qg3/Qh4, or Nf3-Nh4, aiming to transfer the knight to f5.
Before we continue with 5 ... Bg4, please note that 5 ... exd4? would be a very bad mistake. I cannot
stress this enough - train yourself not to think of this move. Suddenly after 6 Bc4 Nf6 7 0-0 White has a
raging attack on the f7-square and Black is practically helpless. Indeed, 7 ... Be7 8 Ng5 0-0 9 Nxf7 Rxf7
(9 ... Qb6 would lose by force to 10 Rxf6!! Bxf6 11 Qh5! d3+ 12 Kh1 Qf2 13 Ng5+ Kh8 14 Qxh7#) 10
Bxf7+ Kxf7 11 e5 Kg8 12 exf6 Bxf6 is not a position I would ever wish to play.
After 5 ... Bg4 we will focus on:

A) 6 c3
B) 6 Bc4

The latter is the most popular move, aiming the bishop at the vulnerable f7 point, while the former has the
idea of attacking the e5 point with Nd2-Nc4.
Joe Gallagher recommended a set-up with 6 Be3 in his book Starting Out: The Caro-Kann. It is a
very unusual idea, but nonetheless Black has a straightforward approach to it: 6 ... Nd7! 7 Nc3 (if 7 dxe5
Qa5+ 8 Bd2 Qb6 and Black is better) 7 ... exd4 8 Bxd4 Qa5, getting ready to castle queenside.

After 9 Bc4 0-0-0 (with the threat of ... Ne5) 10 Qe2 (10 0-0? Bc5! wins material) 10 ... Nb6 11 0-0-
0 Nxc4 12 Qxc4 Be6 and I much prefer Black’s position.

A) 6 c3

This move is recommended by Alexei Bezgodov in his book The Extreme Caro-Kann. The aim is
quick Nbd2-c4 putting immediate and uncomfortable pressure on the centre. This idea is trickier than it at
first appears, particularly as I am reluctant to play the weakening ... b7-b5.
6 ... Nd7! 7 Bd3
This bishop move actually makes a lot of sense. As we will see later, on c4 the bishop might be
vulnerable and sometimes White actually retreats it from there in order to make way for a knight to
manoeuvre its way to c4. So, why should White not implement the plan in one go?
It’s also interesting to play 7 Be2 when suddenly the e5-pawn is under attack. However, after 7 ...
Ngf6 8 0-0 Bd6 9 Nbd2 0-0 10 Nc4 Nxe4 11 Bd3 b5 12 Bxe4 bxc4 13 Qc2 (not 13 Bxh7+? Kxh7 14
Ng5+ Qxg5!) 13 ... h6 14 Bxc6 Rb8 Black shouldn’t have any problems.
7 ... Ngf6 8 0-0
 One of the downsides to playing Bd3 is that the bishop can be vulnerable and after 8 Nbd2?! exd4 9
cxd4 Nc5 10 Bc2 Black pressurizes the white centre with 10 ... Ne6 11 Nb3 a5! before destroying it with
12 a4 Bb4+ 13 Bd2 Bxf3 14 gxf3 Nxd4.
8 ... Be7!?

A very rare move, but a valid one. If the whole point of the variation is to play Nc4 harassing the
bishop on d6 then why should Black be so accommodating?

Advantages of 8 ... Be7

The d-line is clear which means that Black has the tactical strike of ... Nc5, attacking the potentially loose
bishop on d3. Also as our bishop is no longer on d6, should a white knight arrive on c4 it will not be
placing unbearable pressure on the centre. Another point is that there is no danger of e4-e5 forking two
pieces which makes the pawn exchange ... exd4 a lot more attractive. In fact this is a common theme in the
following variations - Black will exchange on d4 and then either break up the centre with ... c6-c5 or
manoeuvre with ... Nc5-e6, attacking the d4-pawn. Finally, the bishop is well-placed defensively on e7,
covering the g5-square: any white bishop sortie to g5 might be tactically loose and tactics against h7 are
less likely to be effective.

Disadvantages to 8 ... Be7

Black’s bishop move does not support the centre which means that at some stage we will be obliged to
exchange pawns on d4.

Black’s Plan
What is the plan if White does nothing? This is often a question I have asked my trainers, so it is only fair
that you should demand it from me.
Well, the answer is that because of the pressure on the d-file it is actually very hard for White to do
nothing and play a normal game. For one thing the d4-pawn is tactically vulnerable, so for instance,
should White play 9 Qc2?! Black can reply 9 ... Bxf3 and after the natural 10 Rxf3 go 10 ... Qb6! as 11
Be3 doesn’t work because of 11 ... Ng4.
If we look at our key position again, we can see an abundance of open lines, diagonals and threats,
meaning that the position is incredibly dynamic.
 To maintain this activity, one thing is clear - Black, as in the days of old, must continually attack the
centre, which will often mean exchanging on d4 and taking the attack from there.

Playing Against the Two-Pawn Centre

It looks very dangerous to give White a pawn centre that looks like this:

Once we have given this centre, the worst thing in the world would be to allow White to consolidate
this advantage so, of course, Black needs to challenge directly with ... c6-c5. Then d4-d5 allows Black to
blockade on d6 and begin targeting the backward e4-pawn.
We also have tactical possibilities, so should White leave his bishop unguarded we might see a
position like this:
 Of course, 11 dxc5 is strategically undesirable, while after the natural retreat 11 Bc2 to maintain the
initiative Black needs to play 11 ... Ne6!. The central pawns are then restrained in the sense that it is not
particularly beneficial for White to advance either of them and Black should play 12 ... Bh5, with the idea
of an eventual ... Bg6, attacking these hanging pawns with gusto.
After 8 ... Be7 we will consider:

A1) 9 Nbd2
A2) 9 Be3
A3) 9 h3

I should also mention:

a) After 9 Qe1 Black simply ignores this attack - it gets nowhere. Indeed, 9 ... 0-0 10 Qg3 doesn’t really
threaten anything so Black should play the pawn sacrifice 10 ... Bh5! 11 Nxe5 Nxe5 12 Qxe5 Bg6. Now
the tables are turning and it is White whose centre will feel the heat: 13 Qf4 c5! (a recurring theme; the
bishop on d3 will come under X-ray pressure down the open d-line) 14 Be3 Nd5 15 exd5 Bxd3 16 Rd1
c4 17 Qf3 Bd6 18 Nd2 Qc7 19 h3 f5 and White’s extra pawn means nothing.
 b) 9 Qc2 leaves the d4-pawn vulnerable after 9 ... Bxf3! (again Black relinquishes central control in
order to begin attacking the d4-pawn; remember should White advance either centre pawn, the structural
advantage passes to Black) 10 Rxf3 exd4 11 cxd4 Qb6 12 Kh1! (sidestepping 12 Be3? Ng4!) 12 ... Qxd4
13 Be3 Qd6 14 Nd2 Ng4 15 Bf4 Qc5 when the strong knights offset White’s temporary initiative.
c) 9 Bg5 makes the bishop very vulnerable here so, for instance, after 9 ... 0-0 10 Qe2 c5! 11 d5?
Black has the tactic 11 ... Nxd5! 12 exd5 Bxg5 13 Qe4? f5 and wins.
d) 9 Na3 should be met with 9 ... Bxa3 10 bxa3 0-0 11 Bg5 c5 12 d5 c4 13 Bc2 h6 14 Bh4 Qa5 15
Rb1 Qxa3 16 Rxb7 with a complicated struggle ahead, but one in which I don’t think Black is worse.
e) 9 Kh1 0-0 10 Be3 transposes to Line A2.

A1) 9 Nbd2
Allowing Black a tactical jump.
9 ... exd4 10 cxd4 Nc5

This isn’t really a tactical trick as the move is pretty obvious, but it is strategically very important.
11 Bc2
After 11 Qc2 Ne6 12 Qc3 (or 12 e5 Nxd4!) 12 ... a5 13 a3 0-0 14 Nc4 Bxf3 the d4-pawn drops.
11 ... Ne6!
Pressuring the centre without delay.
12 Nb3 Bh5!
Black has the idea of a future ... Bg6, provoking the advance of the e4-pawn. Once this pawn
advances Black will obtain a magnificent knight entrenched on d5. Please do not play ... Bg6 quickly,
though, as unless it threatens the capture of the e4-pawn, White will simply play Nh4 eliminating this
dangerous piece.
13 Be3
If 13 e5 Nd5 and ... Bg6 is coming.
13 ... 0-0
 14 Qe2
After 14 e5 Nd5! 15 Qe2 a5 16 a4 Qb6 17 Rae1 Nxe3 18 Qxe3 Rfd8 Black’s pieces are perfectly
placed for defending the kingside and counterattacking against White’s vulnerable centre.
14 ... a5!
The central tension is at maximum here, so Black takes measures on the wing in order to dislodge the
b3-knight.
15 Rad1
Alternatively, 15 a4 Bg6 16 Bf2 and here Black should attack further with 16 ... Ng4! 17 Bg3 Qb6 18
Rfe1 c5! 19 h3 (19 d5? is, of course, impossible because of 19 ... c4+) 19 ... cxd4 20 hxg4 d3+ 21 Bf2
Qxb3 22 Bxb3 dxe2 when he has a pull on account of the ruined white structure.
15 ... Bg6
Finally the move comes.
16 Nbd2 Bb4!
Attacking the defender of e4 once again.
17 Bf2 Re8

It is not clear what White will do to maintain his centre and, indeed, it can easily collapse.
 18 a3? Bxd2 19 Rxd2 Nxe4! 20 Bxe4 Nf4
Suddenly Black is winning.

A2) 9 Be3
A natural choice.
9 ... 0-0

10 Kh1
Another waiting move and a sensible one to boot, removing the king from the open diagonal.
Instead, 10 h3 will transpose to the notes to White’s 10th move in Line A3, which leaves:
a) 10 Qc2 allows 10 ... Bxf3 and after 11 gxf3!? (or 11 Rxf3 Ng4 and White can’t hold on to the
bishop as the awful 12 Bc1? loses to 12 ... exd4 13 cxd4 Qb6) 11 ... Nh5! 12 Qd2 h6 now ... Bg5 is
coming and Black stands well.
b) 10 Nbd2 Bh5 11 Kh1 (11 h3 again transposes to a position we’ll see in Line A3) 11 ... Nc5 12 Bc2
Ne6 13 Bb3 exd4 14 cxd4 Bb4 15 Bxe6 fxe6 16 Qb3 Bxd2 17 Bxd2 Qb6 18 Ng5 (or 18 Qxe6+ Bf7 19
Qf5 Bg6 20 Qe6+ Bf7 with a repetition) 18 ... h6 19 Nxe6 Qxb3 20 axb3 Rf7 21 e5 Ne4 22 Bb4 Rxf1+
23 Rxf1 Bf7 24 Nc5 Nxc5 25 Bxc5 Bxb3 with a draw being the most likely result.
10 ... exd4 11 cxd4 c5!
 Breaking up the centre.
12 d5
This allows Black to set up a dark-squared blockade, but 12 e5 is not very good due to the simple 12
... Nd5 when White’s centre is left in tatters. There is no tactical solution through 13 Bxh7+ (or 13 Bg1
cxd4) 13 ... Kxh7 as everything is covered.
12 ... Bd6 13 Qc2
13 Nbd2 looks the right move, but here Black can play 13 ... Qc7 14 h3 Bxf3 15 Nxf3 Rae8! when his
position is favourable on account of his impressive control over the dark squares and White’s weak e4-
pawn.
If the white knight goes the other way to c4 then after 13 Na3 Ne5 14 Qc2 Nxd3 15 Qxd3 Qe7 16 Nc4
Qxe4 17 Qxe4 Nxe4 18 Nxd6 Nxd6 19 Bxc5 Rfd8 20 Ne5 Be2 21 Rf2 Bh5 we find the white rook and
bishop vulnerable to a family fork, while the black knight sits on its perfect square blockading the passed
d-pawn.
13 ... c4!

A deep positional sacrifice, but one seen fairly often in Benoni-type structures.
14 Qxc4
 After 14 Bxc4 Rc8! the bishop is not comfortable out there on c4: for example, 15 a4 Qe7 16 Nbd2
Nxd5! (regaining the pawn) 17 Bxa7 (or 17 Bg5 N5f6!) 17 ... Nb4 18 Qb3 b6 19 a5 bxa5 20 Rxa5 Nc6 21
Raa1 Nxa7 22 Rxa7 Rb8 23 Qc3 Bc5 and the bishop-pair gives Black a lot of compensation.
14 ... Rc8 15 Qa4 Bxf3 16 gxf3 Nh5
Just look at the queen stranded on the wrong side of the board. Indeed, Black has an almost decisive
attack.
17 Nc3 Qh4 18 Rf2 Ne5 19 Qc2 a6 20 a4 f5 21 Rg1 Bc5! 22 Bxc5 Rxc5
White must drop some material.

A3) 9 h3

9 ... Bh5 10 Qc2

With the idea of protecting the d3-bishop.
As we have seen, 10 Nbd2 doesn’t work so well because of 10 ... Nc5!, while after 10 Be3 0-0 White
has tried:
a) 11 Qc2 can be met by 11 ... exd4 12 cxd4 c5! 13 d5 and now Black has the tactical idea of 13 ...
c4!.
 Once again, White lacks a particularly good way to take the pawn: 14 Qxc4 (or 14 Bxc4 Rc8! and
White is in difficulties due to threat of ... Bxf3, his weak dark squares and, of course, the unpleasant pin
on the c-file; after 15 Qb3 Nc5 the e-pawn falls) 14 ... Bxf3 15 Rxf3 (if 15 gxf3 Rc8 16 Qb3 Bc5! and
White is going to suffer for the rest of the game due to those dark squares) 15 ... Ne5 16 Qc2 Nxf3+ 17
gxf3 Rc8 18 Nc3 Bc5 and I think Black has the better chances in this unbalanced endgame.
b) 11 Nbd2 exd4 12 cxd4 Ne5! 13 Bc2 Nxf3+ 14 Nxf3 Bg6 15 Qb1 Bd6 16 Ne5 Bxe5 17 dxe5 Nd7
18 e6 fxe6 19 Rxf8+ Qxf8 20 Qd1 Qe7 and Black is a pawn up with good chances of outplaying the
opponent.
10 ... 0-0 11 Be3
Should White do nothing and simply develop with 11 Nbd2 then Black’s plan should be 11 ... Bg6 12
Nc4 exd4 13 cxd4 c5!. Again, this is all possible because the bishop is nestled in safety on e7. White now
has to give up his perfect centre and play 14 e5 Bxd3 15 Qxd3 Nd5 16 Qe4 N7b6 17 Na5 cxd4 18 Nxd4
Qd7 19 Rxf7 Rxf7 20 e6 Qe8 21 exf7+ Qxf7, with equality.

11 ... exd4
One of two satisfactory approaches for Black, both of which involve some form of elimination in the
centre and then challenging immediately with ... c6-c5. The other is 11 ... Bxf3 12 Rxf3 exd4 13 cxd4 c5!,
indeed targeting the white pawn centre immediately. Please note that Black does this in the absolute
knowledge that the kingside is absolutely secure. Here:
a) 14 Nc3 cxd4 15 Bxd4 Bc5 is pleasant for Black.
b) 14 e5 is tempting, but absolutely wrong because of 14 ... Nd5 15 Bxh7+ Kh8 16 Qf5 Nxe3 17 Rxe3
(should White play for mate with 17 Qh5 then 17 ... Bg5 saves the day and wins for Black) 17 ... Bg5 18
Re4 cxd4 19 Na3 Qe7 when Black is better.
c) 14 d5 c4! is another strong possibility.
12 cxd4 c5!
 With the bishop on e7 and no danger of e4-e5 winning material, this pawn push becomes very
attractive.
13 e5
We would also be happy to see 13 d5 c4!.
13 ... Nd5! 14 Bxh7+ Kh8 15 Bd2
15 Qf5? g6! collects the bishop on h7.
15 ... cxd4 16 Be4
If 16 Qe4 and after 16 ... Nc5 the white queen has to beat a hasty retreat to Qc2, while 16 Nxd4? loses
to 16 ... Bc5. Note too the marvellous line 16 g4 Rc8! 17 Qe4 Nc5 18 Qxd4 Nb3! 19 axb3 Bc5.
16 ... Rc8 17 Qb3 Ne3! 18 Bxe3 Nc5

Another fantastic idea.

19 Qc4
Or 19 Qd5 dxe3 20 Nc3 Qb6 and Black has the advantage.
19 ... dxe3 20 Nc3 Ne6 21 Qa4 Bc5 22 Ne2 Qb6
Black has a wonderful position with a nicely protected pawn on e3, the bishop-pair and very active
pieces.
B) 6 Bc4 Nd7

This leads to a further and important parting of the ways:

B1) 7 c3
B2) 7 0-0

B1) 7 c3
White supports his centre and threatens active operations on the queenside, perhaps with Qd1-b3 hoping
to quickly utilize the open f-line. I must advise that c2-c3 is a flexible move - White is not obliged to play
Qb3, but may revert to lines with Qe1-Qg3 or Bg5 and Nh4.
7 ... Bh5!

Recommended by Schandorff and I must say I like this move. It overprotects f7 and gives Black some
spicy options including even castling queenside. Generally speaking we will be playing for a piece
formation with the queen on c7 and the dark-squared bishop on d6.
My original recommendation was 7 ... b5, but to be honest I have lost my enthusiasm for it as it rules
out the option of Black castling queenside - something which can be very dangerous for White. After this
move White can play 8 Be2 Ngf6 9 a4! a6 10 dxe5 Nxe4 11 axb5 cxb5 12 Qd5 with the easier position to
play. As such, I have come to the conclusion that Black needs to focus on the centre and not give White an
extra target.
White now has three main moves:

B11) 8 Be3
B12) 8 Qb3
B13) 8 0-0

Instead, if 8 Nbd2 Ngf6 9 0-0 (or 9 dxe5 Nxe5) 9 ... Bd6 10 Qb3 Qc7 11 Ng5 0-0 and White is getting
nowhere, while 8 Bg5 Ngf6 9 dxe5 doesn’t trouble Black in view of 9 ... Nxe5! 10 Qxd8+ Rxd8 and
White can’t capture the knight on e5.

B11) 8 Be3

This commits the bishop to e3 a tad early and, furthermore, it ends up being tactically vulnerable in a
rather surprising way. Black should now simply play the formation: ... Ngf6 and ... Bd6.
8 ... Ngf6 9 Nbd2
Another line runs 9 0-0 Bd6! 10 Qe1 (protecting the bishop; 10 Qb3 Qc7! 11 Ng5 leads to nothing as
Black may simply 11 ... 0-0 when the h2-pawn is vulnerable and the g5-knight will, of course, be driven
away) 10 ... 0-0 11 Nbd2 Re8 (obstinately carrying on with the plan of piling pressure on the e4-pawn)
12 Qh4 Bg6 13 Bg5 (this is the Black’s point - White has wasted time bringing the bishop to g5; instead,
13 Rae1 exd4! 14 cxd4 Nxe4 15 Nxe4 Rxe4 16 Bg5 Rxh4 17 Bxd8 Rh5 18 g4 Rxd8 19 gxh5 Bxh5 is a
complicated endgame, but with two pawns for the exchange Black is by no means worse) 13 ... Qc7 14
dxe5 Nxe5.
 Now 15 Bxf6? Nxc4 16 Nxc4 Rxe4 is a rather nice variation.
9 ... Bd6
There is no real point in playing 9 ... Ng4 as White will simply reply 10 Bg5.
10 0-0
If 10 Qe2 0-0 11 Bb3 Re8, developing as normal, while 10 h3?! fails to 10 ... exd4 11 cxd4 Bg3+.
10 ... exd4! 11 cxd4 Bxh2+!

Revealing the tactical problem with developing the bishop to e3 so early. White has sacrificed (or
blundered!) a pawn to obtain this centre, but I believe Black should be able to defend against his resulting
initiative.
12 Kh1
Of course, if 12 Kxh2 Ng4+!.
12 ... Bc7
Wisely removing the bishop from danger.
13 e5
Probably the best choice, as 13 Qe1 0-0 14 Qh4 Bg6 15 Bg5 h6 16 Bxh6 Nxe4! neutralizes everything
and 13 Bg5 h6! 14 Bh4 (or 14 Bxf6 Nxf6 15 Qb3 b5 when all the critical points are covered) 14 ... g5! 15
e5 Ng4 16 Bg3 Nb6! (16 ... Ne3 17 Qb3 Nxf1 18 Rxf1 Nb6 19 Ne4 is not very clear) 17 Qb3 0-0 18
Rae1 Kh8 19 Bd3 f5 20 e6 Bxg3 21 e7 Qb8 22 exf8Q+ Qxf8 23 Re6 Bf4 gives Black fantastic
compensation for the exchange.
13 ... Nd5 14 Bxd5!
Following 14 Bg5 Qxg5 15 Nxg5 Bxd1 Black is simply a pawn up. Even after the tricky 16 e6 he can
calmly play 16 ... fxe6 17 Nxe6 Bg4! 18 Bxd5 Bxe6! 19 Bxe6 Rf8 20 Ne4 Ke7 and White has nothing.
14 ... cxd5 15 Qb3! Nb6 16 Bg5 Qd7

Black should be okay here.

17 Qb4 h6 18 Bh4 g5 19 Bg3 Qe7 20 Qb5+
Now Black has a repetition if he should desire it.
20 ... Kf8!?
The most ambitious try and I see no reason why Black should stand worse.

B12) 8 Qb3 Qc7!

I like this move keeping options open as regards queenside castling.
I once played 8 ... b5, but wasn’t completely comfortable with my winning chances after 9 Be2 Ngf6
10 a4 exd4?! (10 ... b4! is better) 11 cxd4 bxa4 12 Qxa4 Nxe4 13 Qxc6 Bb4+ 14 Nc3 Nef6 15 0-0 0-0 16
Ba6 and the game eventually ended in a draw in I.Snape-J.Houska, British Championship, Canterbury
2010.
9 0-0
Another natural move.
9 Ng5 looks scary, but White is pushed back with 9 ... Nb6! 10 Bd3 (10 0-0 Nxc4 11 Qxc4 Nf6 12
dxe5 Ng4 is better for Black as the threat of ... Qb6 will give White a serious headache) 10 ... Rd8 11 0-0
Nf6 12 a4 (12 Be3 h6 13 Nf3 Ng4 is very satisfactory for Black too) 12 ... Bd6 13 a5 Nbd7 14 Bc4 0-0.
By now Black has consolidated and can start taking advantage of his harmoniously placed pieces.
9 ... Bd6 10 Ng5
Now Black has only the one move, but it also happens to be quite good.
10 ... Ndf6!
 Please note that everything else loses. We are now in virgin territory, so let’s experiment a little.
11 Bf4!?
If 11 a4 I think it is best for Black to chase away the white knight with 11 ... h6 12 Nf3 0-0-0!?. Now:
a) 13 dxe5 Bxe5! 14 Nxe5 Qxe5 and White has problems with threats such as ... Qc5+ followed by ...
Nxe4.
b) 13 Qc2 g5 14 a5 Kb8 15 a6 b6 16 g3 (if 16 dxe5 Bxe5 17 Nxe5 Qxe5 18 Be3 Ng4 or 16 Be3 Ng4)
16 ... Bg6 17 Nbd2 Ng4 18 Qb3 N8f6 and White is behind in development.
c) 13 Nbd2 Ng4 14 h3 Ne3 15 Rf2 Nf6 16 Bd3 Rhe8 17 Nb1 exd4 18 cxd4 and now 18 ... Nf5!! 19
exf5 Bxf3 is a lovely idea.
Instead, after 11 dxe5 Bxe5 only Black can be better, while 11 Be3 b5 12 Bd3 h6 13 Nf3 Ng4 14 Bd2
exd4 15 cxd4 Bxh2+ 16 Kh1 Bf4 is a big mess.
11 ... 0-0-0 12 dxe5?
A mistake, but White might well be worried by 12 Bg3 Nh6 13 Nd2 exd4!.
12 ... Bxe5 13 Bxe5 Qxe5 14 Nxf7? Bxf7 15 Bxf7 Kb8!

Funnily enough, White is practically lost. The dark squares around his king are just too weak.
16 h3
 Or 16 Qe6 Qc5+ 17 Kh1 Nh6 18 b4 Qg5 and ... Nhg4 is coming.
16 ... Nh6 17 Bc4 Nhg4
It will be checkmate soon.

B13) 8 0-0
White’s main move.
8 ... Ngf6

There are a lot of differing plans here, so it is important for us to categorize the different attacks after:

B131) 9 Bg5
B132) 9 Nbd2

If 9 Qe1 to remain consistent with my coverage, I am going to recommend 9 ... Bd6 10 Bg5 (10 Nh4 is the
only move that prevents our desired queenside castling, but 10 ... Bg6 11 Nf5 Bxf5 12 Rxf5 0-0 13 Bg5
exd4 14 cxd4 Qb6 15 Bxf6 Nxf6 16 Rxf6 Qxd4+ 17 Rf2 Qxc4 is annoying for White, as is 10 Qh4 Qb6
with uncomfortable pressure against the d4-pawn) 10 ... Qc7 11 Qh4 0-0-0.
Instead, 9 Qb3 is neutralized pretty effectively by 9 ... Qb6!.

B131) 9 Bg5
The critical continuation against which we follow our natural formation:
9 ... Bd6
 Now White can look to attack without delay or first develop:

B1311) 10 Qe1
B1312) 10 Nbd2

10 Qd3 isn’t so good. Firstly, because the queen is vulnerable on d3 and, secondly, the idea of Nd2-c4
can be dealt with relatively easily: for instance, 10 ... 0-0 11 Nbd2 Re8 12 Rae1 b5 13 Bb3 h6 14 Bh4 a5
15 a4 Nc5.

B1311) 10 Qe1

This is highly committal, but has the idea of launching a dangerous kingside attack. My policy for this
is: step to the right. That is my way of saying we should go ... Qc7 followed by queenside castling. If
White intends to attack on the kingside, then why not castle queenside?
10 ... Qc7! 11 Qh4
If 11 Nbd2 Black can still opt for the bold 11 ... 0-0-0!? 12 b4 (or 12 a4 h6 13 Bxf6 Nxf6 14 a5 Bg6
15 a6 b5 16 Ba2 Rhe8! and Black’s play in the centre offsets White’s queenside play) 12 ... h6 13 Bxf6
Nxf6 14 b5 cxb5 15 Bxb5 Kb8 16 Rb1 and now he can even go pawn grabbing with 16 ... Qxc3 17 Ba6
b6, reaching a complicated position in which he doesn’t stand worse.
11 ... 0-0-0!?

12 dxe5
12 Nbd2?! h6 13 Be3 g5 is better for Black and 12 b4 h6 13 Bxf6 Nxf6 14 dxe5 Bxe5 15 Nxe5 Qxe5
dxe5 must count as a positional victory.
12 ... Bxe5!
Surprising, but good.
13 Nxe5
Alternatively:
a) 13 Bxf6 Nxf6 14 Nxe5 Qxe5 15 Rf5 Rd1+ 16 Bf1 Qc7 would be losing for White.
b) 13 Bb3 Rhe8 14 Nbd2 h6 15 Bxf6 Nxf6 16 Nc4 Bxf3 is better for Black; the weakness of the e4-
pawn and the dark squares will begin to tell in the long run.
13 ... Nxe5 14 Bxf6 gxf6 15 Qxh5
15 Bb3 Be2 16 Rxf6 (16 Rf2? Qb6! is even worse) 16 ... Ng4 wins the exchange as Black threatens
the deadly ... Qb6+.
15 ... Nxc4 16 Qc5 Ne5
 17 Qxa7 Ng4 18 g3
18 Qa8+ doesn’t help either because of 18 ... Kd7 19 Rd1+ Ke6 and wins.
18 ... Qe5
White’s pieces are completely offside and unable to deal with the threats of ... Qxe4 and ... Qh5.

B1312) 10 Nbd2

A more flexible move. White keeps all his options open as to which plan to follow. It doesn’t matter
though as I believe Black’s best is to mimic White and keep all our options open.
10 ... Qc7!?
This also carries the threat of a capture on d4 followed by ... Bxh2+ now that the queen cannot dash
over to h4 to protect the h-pawn. I prefer this to the standard 10 ... 0-0 which should be met with 11 Qe1!
Qc7 (or 11 ... Re8 12 Qh4 and the position is favourable for White) 12 Qh4 Bg6 13 Bb3.
11 Qc2
The queen is offside here and I believe Black can now safely castle kingside.
Instead, 11 Qe1 transposes to the note to White’s 11th move in Line B1311. Independent tries are:
a) 11 Bb3 exd4 12 cxd4 Bxh2+ 13 Kh1 Bf4 14 Bxf6 (after 14 Bxf4 Qxf4 15 e5 Qh6 16 Kg1 Ng4 17
Nc4 0-0 White has some compensation for the missing pawn, but Black shouldn’t have too much trouble)
14 ... Nxf6 15 e5 Nd5 16 Ne4 0-0 and nothing bad happens to Black.
b) 11 Bd3 exd4 12 cxd4 Bxh2+ 13 Kh1 Bf4 14 Bxf6 Nxf6 15 e5 Nd5 16 Ne4 0-0 and White has little
compensation for the sacrificed pawn.
c) 11 h3 makes an important kingside weakness on g3. Now Black can implement the traditional plan
of castling and applying pressure on e4: 11 ... 0-0! 12 Qe1 Rae8!? and now 13 Qh4? is a mistake due to
13 ... Bg6 when the pressure on the centre is overwhelming and forces White into unfavourable trades.
d) After 11 Kh1 there is no kingside weakness so I am tempted to play 11 ... h6 12 Bh4 and then the
enterprising 12 ... 0-0-0 13 a4 g5 14 Bg3 exd4 15 Bxd6 Qxd6 16 cxd4 Bxf3 17 Nxf3 Nxe4 18 Bxf7 Kb8
when Black really has few problems.
e) 11 dxe5?! Nxe5! breaks up White’s centre far too easily.
f) 11 Bxf6 Nxf6 12 Qe1 0-0-0 13 dxe5 Bxe5 14 Nxe5 Qxe5 15 Rf5 Qc7 and again there is little
trouble for Black.
Returning to 11 Qc2:

11 ... h6!
A little accuracy is required here in order to ensure that White doesn’t get the time to play Nf3-h4 in
one go, so we avoid 11 ... 0-0 12 Kh1!? Rfe8 13 Nh4!.
12 Bh4
12 Be3 runs into 12 ... Ng4! and 12 Bxf6 Nxf6 13 Nh4? (if 13 dxe5 Bc5+ or 13 Bb3 Bxf3 14 Nxf3 0-0
15 dxe5 Bxe5 16 Nxe5 Qxe5 and Black really cannot have any problems especially considering White’s
weak e4-pawn; finally, if 13 Bd3 Bxf3 14 Nxf3 Rd8 15 Kh1 0-0 and Black stands well once again) 13 ...
exd4 14 cxd4 (or 14 Nf5 Bxh2+ 15 Kh1 0-0-0 and Black is better) 14 ... Bxh2+ 15 Kh1 0-0-0! illustrates
some of our main aims.
12 ... 0-0!
 Switching sides, but we have the simple plan of pressuring e4.
13 Rae1
After 13 Kh1 b5 14 Bd3 Bg6 15 a4 a6 Black has the plan of ... Rfe8 and ... Nf6-h5 with a good
position. Moreover, White must be a little careful as moves such as 13 Bd3 allow Black to play messy
lines like 13 ... Ng4!? 14 Rae1 exd4 15 cxd4 Bxh2+ 16 Kh1 Bf4 17 e5 Ne3 18 Rxe3 Bxe3 19 Nc4 Bf4 20
g3 b5.
13 ... Bg6!? 14 Bd3
If 14 Bf2 exd4 15 cxd4 Ng4 16 g3 c5 and the perfect pawn centre cannot stay intact.
14 ... b5 15 Bxf6 Nxf6 16 Nh4
Black is fine after 16 dxe5 Bxe5 17 Nxe5 Qxe5 18 Nf3 Qc5+, which may be why 16 c4 is the
computer’s recommendation, but this is difficult to believe in after 16 ... Qe7!.
16 ... exd4 17 Nxg6 dxc3 18 Nxf8 Bxh2+ 19 Kh1 cxd2 20 Qxd2 Be5
Black will win the knight and have an excellent position due to White’s weak dark squares.

B132) 9 Nbd2

Developing the knight in front of the bishop isn’t too challenging for Black.
 9 ... Bd6 10 Bd3
Overprotecting the e4-pawn and threatening 11 Nc4, but White is wasting time here.
10 Bb3 is another possibility and one well met by 10 ... 0-0 11 Nc4 Bc7 12 dxe5 Nxe4 13 Bc2 Bg6
14 Qe2 Ndc5. Likewise, 10 Qe1 0-0 is fine, and if 11 Nh4 (11 Qh4 isn’t dangerous without a bishop on
g5; Black can simply retreat and 11 ... Bg6 12 Bd3 exd4 13 cxd4 Nc5! 14 Bc2 Ncxe4 15 Nxe4 Nxe4 is
one cute tactical point) 11 ... exd4 12 cxd4 Qb6 13 Nf5 Bg6 14 Nb3 Bxf5 15 Rxf5 Rae8 16 Bd3 c5,
breaking up the white centre.
10 ... b5!?
Only now do we play this, preventing the desired knight hop.
11 a4 a6 12 Qe1 0-0

13 axb5
Alternatively:
a) 13 Nh4 Nc5 14 Bc2 exd4 15 cxd4 Ne6 16 Nf5 Nxd4 17 Nxd4 Bc5 and Black is winning a pawn.
b) 13 Kh1 Bc7 (13 ... Re8!?) 14 Nb3 (14 Nh4? is too loose: 14 ... exd4 15 cxd4 Nc5 and Black wins
a pawn) 14 ... Re8 15 Bg5 h6 16 Bh4 Bg6 17 Qb1 Qe7 18 Bg3 Rac8 19 Nh4 Bh7 20 axb5 axb5 is very
solid for Black.
13 ... axb5 14 Rxa8 Qxa8 15 Nh4 Qb8 16 Nf5 Bg6 17 Nxd6 Qxd6 18 Nf3 Re8!
White is in a bind.

B2) 7 0-0 Ngf6 8 c3

 The major difference between the two variations 7 c3 and 7 0-0 is that versus the former Black
throws in the move ... Bh5, whereas here he has played an early ... Ngf6. One of the things to note is that
as the bishop still stands on g4 an Qe1 attack is more attractive, but White should take care as there are
still a few tricks involving ... Bxh2.
White usually, but not always shores up d4 so:
a) 8 Be3 allows Black to liquidate: 8 ... exd4 (handing most of the long-term prospects to Black; 8 ...
Bd6 is also good) 9 Bxd4 (if 9 c3? dxe3 10 Bxf7+ Kxf7 11 Ng5+ Kg6) 9 ... Bc5! 10 e5 (or 10 c3 Qe7 11
Nbd2 0-0 12 Qc2 Ne5! 13 Nxe5 Qxe5 and Black is better) 10 ... Bxf3 11 gxf3 Qb6! and White’s early
aggression is backfiring.
b) 8 Bg5 Bd6 9 c3 would transpose to Line B21.
c) 8 Qe1 is always an attacking idea to consider, but here it is not so effective due to 8 ... Bd6 9 c3 (9
Qg3 is easily met by 9 ... Qc7 and 9 Qh4 by 9 ... Bxf3 10 Rxf3 exd4) 9 ... 0-0 10 Bg5 and now it is
important to play 10 ... Qb6! when Black is much better.
This is an important move to remember and now 11 Nbd2 Bxf3! (not 11 ... Rae8 12 Bd3 and Black
was struggling in M.Calzetta-A.Hagesaether, Gibraltar 2009) 12 Nxf3 exd4 13 cxd4 Rae8 nets a pawn.
d) 8 Nbd2 Bd6 9 h3 Bh5 10 c3 b5 11 Bb3 0-0 12 Qe1 Qb6 13 Kh1 Rae8 (standard development from
Black) 14 Bc2 Bg6 15 Qh4 exd4! (rather than the 15 ... Qc7 of V.Bachin-T.Cheremnova, Sochi 2012) 16
cxd4 (or 16 Nxd4 Bb8) 16 ... c5! leaves Black with a fantastic position.
8 ... Bd6
 I like this principled choice, although if you want to, you can transpose to our lines after 7 c3 with 8 ...
Bh5 9 Bg5 Bd6 10 Nbd2 Qc7 taking play back into Line B1312.
After the text White has a few approaches, depending on how aggressive he is feeling:

B21) 9 Bg5
B22) 9 Be3
B23) 9 Qe1
B24) 9 Nbd2
B25) 9 Qb3

9 Qc2 doesn’t quite make sense as the white queen should be somewhere eyeing the black king. Here 9 ...
0-0 10 Nbd2 b5 11 Bd3 Qb6 12 a4 a6 13 a5 Qa7 14 Kh1 c5 15 d5 c4 is better for Black.
We must, of course, also test the rather radical 9 Bxf7+ Kxf7 10 Qb3+, but with a series of careful
moves in 10 ... Ke7 11 Ng5 Rf8 12 h3 h6! 13 hxg4 hxg5 14 Bxg5 Qb6! 15 Nd2 Qxb3 16 axb3 Ke6 17
dxe5 Bc5+ 18 Kh2 Nxg4+ Black had emerged a piece ahead in J.Knudsen-N.Pedersen, Ballerup 2012.

B21) 9 Bg5
 9 ... 0-0 10 Nbd2
10 Qe1 is the most dangerous move in general, since if a queen starts attacking our king you can rest
assured this is the line you have to have covered. However, again we can be fairly happy after 10 ...
Qb6!? followed by ... exd4 and ... Rae8, playing to win the e4-pawn: for example, 11 Nbd2 Bxf3 12 Nxf3
exd4 13 cxd4 Rae8 14 Bxf6 Nxf6 15 e5 Bxe5.
Instead, 10 Nbd2 h6! 11 Bh4 should be met by 11 ... Qc7 and ... Bh5-g6.
10 ... h6!
Schandorff’s suggestion designed to take the h4-square away from any knight or queen, thereby
neutralizing White’s attack.
11 Bh4
11 Be3 Bh5! was recommended by Schandorff and after 12 Kh1 Re8 13 Qc2 Qc7 14 Bg1 b5 15 Bd3
a6 it is unclear what White is playing for.
11 ... Qc7 12 h3!?
Alternatively, 12 Qc2 b5 13 Bd3 and then:
a) 13 ... a6 14 h3 Be6 15 Rae1 Nh5 and White has no plans for improvement.
b) 13 ... Nh5 is also a promising possibility and after 14 Rfe1 (D.Howell-J.Richardson, British
League 2010) 14 ... c5! 15 h3 Bxf3 16 Nxf3 Nf4 17 Bxb5 cxd4 18 Bxd7 Qxd7 19 cxd4 Rfc8 20 Qd1 exd4
Black is better.
12 ... Bh5 13 Qc2 Bg6!
 The standard plan for pressuring the centre.
14 Bd3 b5 15 dxe5 Bxe5! 16 Nxe5 Qxe5
Now the weak e4-pawn will offset the bishop-pair.
17 Nf3 Qf4 18 Rae1 Nc5 19 Bxf6 Qxf6 20 b4 Nxd3 21 Qxd3 Rfe8 22 Re3 a5
I quite like the black position with play along the open d- and e-files, Y.Zherebukh-E.Bareev, Moscow
2009.

B22) 9 Be3

White aims to solidify the centre before beginning active operations.

9 ... 0-0 10 Nbd2
Alternatively:
a) 10 h3 Bh5! (as there are no immediate benefits to exchanging on f3, it is best to keep the bishop on
the board) 11 Nbd2 Bg6 (11 ... exd4?! is no longer possible because after 12 cxd4 c5 13 g4! Bxg4 14
hxg4 Nxg4 15 Ng5 Nxe3 16 Qh5 h6 17 Nxf7 White is winning) 12 Bd3 and now 12 ... exd4 13 cxd4 c5 is
very effective.
b) 10 Qc2 Bxf3 11 gxf3 (11 Rxf3 Ng4 12 Bd2 exd4 13 cxd4 Bxh2+ is the big problem for White) 11
... Nh5 12 Nd2 exd4 13 cxd4 Bf4 and Black has the advantage on the dark squares.
c) 10 Qe1? is not well timed because of the simple 10 ... Nxe4!.
10 ... exd4
Black only trades on d4 under specific conditions. There must be resulting pressure on the white
centre or a tactical resource - or both, as here.
11 cxd4 c5!

12 dxc5
Practice has also seen:
a) 12 Qb3 cxd4! 13 Bxd4 Bc5! 14 Qe3 Bxd4 15 Nxd4 Ne5 16 h3 Qb6! 17 Nf5?! (even after 17 Rxf6
Qxf6 18 hxg4 Nxg4 19 e5 Qxe5 20 Qxe5 Nxe5 21 Bb3 Rad8 the open position favours the pair of rooks)
17 ... Rfd8 18 Qxb6 axb6 19 Ne7+ Kf8 20 Nd5 Nxd5 21 Bxd5 Be6 22 Bxb7 Rxa2 23 Rxa2 Bxa2 24 Nf3
Nd3 and Black had the better position in H.Ziska-R.Kasimdzhanov, Khanty-Mansiysk 2010.
b) 12 d5 was played in M.Muzychuk-T.Coleman, Gibraltar 2012, and here I believe Black should
focus on putting as much pressure as possible on the weak e4-pawn after 12 ... Re8:
b1) If 13 Qc2 Bxf3! (the point is that White cannot recapture with a knight) 14 Rxf3 (14 gxf3 is
interesting, but White cannot play an early f3-f4; now 14 ... Qb8 is a multi-dimensional move with the
idea of meeting 15 f4 with 15 ... b5 16 Be2 c4 when White can’t advance in the centre and his bishop is
vulnerable on e3) 14 ... Ne5 15 Rh3 Neg4 16 Bg5 Qc7 17 g3 Be5 when Black has excellent control over
the dark squares.
b2) 13 Qb3 Rb8 14 h3 Bxf3 15 Rxf3 b5 16 Bf1 c4 17 Qc3 Nxe4 18 Nxe4 Rxe4 is better for Black.
c) 12 e5 cxd4 13 exf6 dxe3 14 fxg7 Kxg7 15 Ne4 Be5 protects the exposed king and if White goes 16
Qe2 Black can play 16 ... Qb6 17 h3 Bh5 18 Rad1 Nf6 19 Nxf6 Bxf6 when he is much better.
12 ... Bxc5 13 Bxc5 Nxc5
With the white centre demolished all that is left is to neutralize any form of attack along the f-line and
Black will easily stand better on account of the weak e-pawn.
14 e5 Nfd7
 15 b4?!
Now Black has the tactical shot:
15 ... Nxe5! 16 bxc5 Bxf3 17 Bxf7+ Rxf7 18 Nxf3 Nxf3+ 19 Rxf3 Rxf3 20 gxf3 Kf8 21 Qxd8+
Rxd8 22 Rb1 Rd7 23 Kf2 Ke7 24 Ke3 Kd8
With the white king cut off from the queenside Black soon won in Y.Kolotilin-A.Belozerov,
Novokuznetsk 2008.

B23) 9 Qe1
A dangerous idea. White wishes to relocate the queen to the kingside in anticipation for an attack there.
9 ... 0-0
9 ... Qb6!? could also be a tricky try.
Now we have:

B231) 10 Bg5
B232) 10 Nbd2

Instead, 10 Qg3 is easily repelled by 10 ... Qc7 and 10 Qh4 Qb6! is a cool counterattacking move. Here a
serious mistake is 11 Bg5? (11 Nbd2 Bxf3! 12 Nxf3 exd4 13 cxd4 Ne5! is another point and after 14 Bb3
Rae8 15 Kh1 Ng6 the e-pawn falls) 11 ... Bxf3 12 Rxf3 Qxb2 13 Rxf6 Nxf6 14 Bxf6 gxf6 15 Qxf6 Qc1+
16 Kf2 exd4 when there is no perpetual.

B231) 10 Bg5 Qb6

 Let’s try to refute this line by counterattacking on the queenside and trusting that our kingside will
hold.
11 Nbd2
Giving up material, but after 11 Qh4 Bxf3! 12 Rxf3 Qxb2 13 Rxf6 Nxf6 14 Bxf6 gxf6 15 Qxf6 Qc1+
16 Bf1 Qe3+ 17 Kh1 Rad8 there is nothing. Likewise, if 11 Kh1 exd4 12 cxd4 Bxf3 13 gxf3 Qxd4 14 Qe2
Qe5 15 Be3 b5 16 Bd3 Nc5 17 f4 Qe6 and White has to give up one of his bishops.
11 ... Bxf3! 12 Nxf3
After 12 gxf3 exd4 13 cxd4 Qxd4+ 14 Be3 Qxe3+ 15 Qxe3 Bc5 Black is a pawn to the good.
12 ... exd4 13 Nxd4
13 cxd4 Rae8 wins the e4-pawn and after 14 Bd3 Nxe4 15 Bxe4 f5 16 Nd2 fxe4 17 Rxf8+ Nxf8
Black has the advantage, while I would meet 13 e5!? with 13 ... Rae8 (13 ... dxc3+ 14 Kh1 cxb2 15 Rd1
Bb4 16 Qe2 Nd5 17 e6 is a complicated position) 14 Bxf6 (calm moves such as 14 Kh1 completely fail
to 14 ... Nxe5) 14 ... Nxf6.
13 ... Rae8 14 Bd3 c5 15 Nf5 c4+ 16 Be3 Bc5 17 Bc2 Qxb2 18 Qg3 Bxe3+
White is in big trouble.

B232) 10 Nbd2
 I am unsure about this move as White blocks in his bishop on c1.
10 ... Re8
10 ... Qc7 is also possible.
11 Bd3
11 Bb3 is perhaps a better bet, although now 11 ... Bc7! is useful prophylaxis against Nc4. Indeed,
now the knight cannot move from d2 as it has to defend the e4-pawn.
11 ... exd4 12 cxd4 Nc5!

12 ... c5 is also possible but more complicated than my recommendation.

13 Bc4?
Or 13 Bc2 Bxf3! 14 Rxf3 Ne6 15 Nb3 Nxd4 16 Nxd4 Bc5 and White finds his centre has crumbled
rather dramatically.
13 ... Ncxe4 14 Bxf7+ Kxf7 15 Nxe4 Kg8!
White has serious problems.

B24) 9 Nbd2
 9 ... 0-0 10 Qc2
Instead, if 10 a4 Rc8 11 Bb3 Qe7, while 10 Qe1 takes play back into Line B232.
10 ... b5!?
White is hoping for something like 10 ... Qe7 11 a4 Rae8 12 Bd3 h6 13 Nh4.
11 Bd3 Qb6 12 a4 a5 13 Kh1 Rfe8 14 h3 Bh5 15 Qb3 Rab8
White is struggling with how to continue his development.

B25) 9 Qb3

It is well known that White gains nothing from this queen sortie.
9 ... 0-0!
Black ignores the threat completely and concentrates on completing his development.
10 Qxb7
The critical test, but not everyone is brave enough to snatch the pawn:
a) After 10 Bg5 Qc7 11 Nbd2 b5 12 Be2 a draw was agreed in M.Wiander-P.Grochowalski,
Stockholm 2009, but Black has the more pleasant position.
b) 10 Ng5 Bh5 11 g4? is just a terrible weakening: 11 ... Bg6 12 h4 b5 13 Be6 Nc5 14 dxc5 Bxc5+ 15
Kg2 Qd3 and wins.
c) 10 Bxf7+? fails to 10 ... Rxf7 11 Ng5 Bh5 12 Nxf7 Bxf7 13 Qxb7 (M.Jensen-N.Pedersen,
Copenhagen 2010) 13 ... exd4 14 cxd4 Bc4 15 Qxc6 Bxh2+ 16 Kxh2 Rc8 17 Qb7 Bxf1 and Black is
winning.
10 ... exd4
The usual reaction to the capture on b7.
11 cxd4
With 11 Qxc6 White tries to be clever, but this move order does not help him in view of 11 ... Ne5! 12
Nxe5 Bxe5 (Black now threatens ... Rc8) 13 h3 and then:

a) 13 ... Rc8 14 Qb5 Qc7 15 hxg4 dxc3 16 Nxc3 Nxg4 and White needs to give back material in order
to survive the onslaught.
b) Even stronger is 13 ... dxc3! 14 bxc3 (14 Nxc3 Rc8 15 Qa6 Qd4+ 16 Kh1 Qxc4 wins a piece) 14
... Rc8 15 Qa6 Nxe4 16 hxg4 Rxc4! 17 Bf4 (or 17 Qxc4 Qb6+ 18 Kh1 Ng3+ and Black wins) 17 ... Bxf4
18 Rxf4 Qc7 and Black has a devastating position.
11 ... Nb6
Blocking in the queen is superior to 11 ... Nxe4 12 Nc3.
12 Bb3
12 Bd3 is strongly met by the unexpected 12 ... Bxf3 13 Rxf3? (even after 13 gxf3! c5! 14 Bg5 cxd4
15 f4 Qd7 16 Qxd7 Nfxd7 17 e5 f6! Black has great pieces) with a very pleasant choice:
a) After 13 ... Bf4! Black wins the pawn back while obtaining the better chances: 14 Rxf4 (ignoring
the problem does not help White: 14 Nc3 Qxd4+ 15 Kh1 Bxc1 16 Rxc1 Ng4 and Black is better or 14
Bxf4 Qxd4+ 15 Be3 Qxb2 16 e5 Nfd5 17 Rh3 g6 18 Bh6 Rfc8 19 Rf3 Rc7 and wins) 14 ... Qxd4+ 15
Kh1 Qxd3 and White’s position is horrible.
b) 13 ... Ng4!? might even be stronger: for example, 14 e5 (or 14 h3 Bh2+ 15 Kh1 Qxd4) 14 ... Bb8!
15 Rf4 Nxe5 16 dxe5 Qxd3 17 Nc3 Bxe5 18 Rf3 Qd6 with a large advantage for Black.
12 ... Nxe4
 13 Nc3!
Best and now a complex endgame arises.
Instead 13 Qxc6? was considered by Volzhin as “over-optimistic and bad”. After 13 ... Bxf3! 14 gxf3
(or 14 Rxf3 Rc8! 15 Qxe4 Rxc1+ 16 Kf2 Bxh2 and Black has a winning attack) 14 ... Bxh2+!! (a great
demolition sacrifice) 15 Kxh2 (15 Kg2 Qh4 16 Qxe4 Qg3+ 17 Kh1 Qh3 18 Qg4 Qxf1+ 19 Kxh2 Qxc1
wins too) 15 ... Qh4+ 16 Kg1 Qg3+ 17 Kh1 Qh3+ 18 Kg1 Ng3 19 Re1 Qh1+ 20 Kf2 Qh2+ 21 Ke3 Rae8+
Black picks up the rook as well as the game.
13 ... Nxc3 14 bxc3 Qc8 15 Qxc8 Raxc8 16 Ng5 Nd5 17 c4 Nf6 18 c5 Bb8
In this complicated, roughly equal endgame, both sides have weak pawns they can target.

Summary
The Fantasy Variation has amassed an attacking but slightly dubious reputation. If White plays the
‘extreme’ 6 c3 Nd7 7 Bd3, I believe Black can be a little bit sneaky and develop the bishop to e7 - that
should confuse them! The aim is to pressure the centre and/or play a well-timed ... c6-c5.
More mainstream ideas are 6 Bc4 Nd7 and then 7 0-0 and 7 c3. Although White’s plans are very similar,
the arising positions are slightly different. The plan that Black absolutely must memorize is how to react
to the idea of Bg5 followed by Qe1. With the flexible 10 ... Qc7 we wait for White to commit before
sending our king to either wing.
Chapter Thirteen
The Pseudo-Panov: 2 c4
1 e4 c6 2 c4

This variation has for a long time lived in the shadow of the Panov-Botvinnik Attack. Some have
dismissed 2 c4 as “rarely having a character of its own” and left it at that. However, this move is fast
becoming recognized as a valid independent system with some dangerous points. Like the Panov, the
positions are characterized by an Isolated Queens Pawn (IQP), but unlike the Panov, the nature of the
game is determined very early on, with Black even voluntarily allowing White the option of playing with
an extra pawn, albeit with two doubled, isolated pawns. There are then two scenarios:
a) White holds on to the material advantage.
b) White allows Black to win back the pawn.
2 ... d5 3 cxd5
Note that 3 exd5 usually leads to the same thing, although 3 ... Nf6!? is an additional option for Black.
Now there is nothing better than for White to take the game back into the Panov proper with 4 d4, or play
4 Nc3, as the pawn sacrifice is too dangerous to accept: 4 dxc6 Nxc6 5 Nf3 e5 6 d3 e4 7 dxe4 Qxd1+ 8
Kxd1 Nxe4 9 Be3 Bf5 10 Nbd2 0-0-0 11 Kc1 Nxd2 12 Nxd2 Bb4 and Black has very active piece play
for the pawn.
3 ... cxd5 4 exd5
4 e5 takes the game into the realm of a good Advance Caro for Black: 4 ... Nc6 5 d4 Bf5 6 Nc3 (or 6
Bd3 Bxd3 7 Qxd3 e6 8 Nf3 Nge7 and Black has no problems at all) 6 ... e6 7 Nf3 Nge7 8 Be2 Bg4
followed by ... Nf5.
4 ... Nf6
 The most principled reply, to which White has three main options:

A) 5 Bb5+
B) 5 Qa4+
C) 5 Nc3

With the first two variations, White is looking at holding on to the d5-pawn (strategy 1), with the idea
behind the checks being to obstruct Black’s coordination on the d-line. His third choice, however, returns
the pawn and heads for a normal IQP situation.

The Doubled d-pawns: A Weakness or Strength?

This is actually a very complicated question. Objectively speaking, yes, the two pawns by themselves are
very weak, but the practical problem is that they can be defended by almost the entire white army. This
little doubled pawn line-up virtually dominates most of the squares on the c- and e-files, which can give
Black a problem with space: he does not have an easy way to manoeuvre his pieces around the two
pawns.
However, and here is White’s problem, he has contorted and twisted the pieces to defend d5 so that
should Black win back the pawn, Black can be exceedingly satisfied and can then start concentrating on
the remaining d-pawn. This can be particularly seen in the 5 Qa4+ lines, where White is willing to place
some members of his army on awkward squares, as long as they contribute to defending the d5-pawn.
White has two methods of protecting the pawns: he can do it directly, but also indirectly by either
attacking the pawn on e7 or making threats against f7.

The Endgame Line: 5 Bb5+

The variation with 5 Bb5+ more often than not leads to an endgame - one that will be very pleasant for
Black should he manage to win the d5-pawn and prevent White from penetrating down the c-file. The
general formula for handling this check and reaching the said endgame works is: chase the bishop away
with ... a7-a6 and ... Rb8, which forces White to swap the bishop on b5 for the knight on d7.
White can avoid the exchange of queens, but strategically Black will have the superior minor pieces
and the better structure. After the queens are exchanged, the position will look like this:
 Here Black can win back the pawn very easily by playing 10 ... b5, followed by rerouting the bishop
to b7 to target d5.
The game D.Drollinger-F.Doettling, German League 2004, illustrates Black’s Ideal Position.

This is a fantastic position for Black. By playing quietly White has allowed Black to suppress any
queenside activity and to secure two long-term advantages: the weak d-pawn and the two bishops.

A) 5 Bb5+
 White wants to make it as awkward for Black as possible to capture the d-pawn so he forces Black to
place a piece on d7. I call this the ‘Endgame Line’ just to make it clear in my mind that the game will
transpose to a tame endgame.
5 ... Nbd7 6 Nc3
6 Nf3 really doesn’t trouble Black. One should simply get on with winning the d5-pawn and playing
against the IQP: for example, 6 ... a6 7 Ba4 b5 8 Bb3 Bb7 9 0-0 Bxd5 10 d4 e6 11 Nc3 Bxb3 12 axb3 b4
13 Nb1 and Black already had the upper hand in J.Vojinovic-M.Sebag, Gibraltar 2014.
6 ... a6 7 Qa4
The most popular continuation. White makes it impossible to capture the bishop.
If 7 Bxd7+ Black can simply recapture with 7 ... Qxd7 and after 8 Nf3 (8 Qb3 Qg4! is an active
continuation; after 9 Kf1 g6! Black will develop logically and attack the d5-pawn, L.Ljubojevic-A.Dreev,
Moscow (blitz) 1993) 8 ... Nxd5 9 0-0 Nxc3 10 bxc3 e6 11 a4 (preventing Black from playing ... b7-b5)
11 ... Bd6 (the light-squared bishop will come into the game via d7 and c6, so Black should simply
concentrate on castling) 12 Ba3 0-0 13 Qb3 Rb8 14 Qb6 (after 14 Bxd6 Qxd6 15 a5 Bd7 Black doesn’t
have too many problems; he has a weak b-pawn, but the counterplay on the c-line and the light squares
more than offsets this) 14 ... Bc7 15 Qc5 Re8 16 Ne5 Qd8 17 Rfe1 f6 18 Nf3 Bd7 the light-squared
bishop is out, giving Black an advantage, M.Berkovich-S.Erenburg, Israeli League 2011.
Instead, if 7 Ba4 b5 8 Bc2 Nb6 and Black will regain the pawn.
7 ... Rb8!
 If you don’t fancy playing an endgame then 7 ... g6 is also good: 8 Nf3 Bg7 9 0-0 0-0 10 Bxd7! Nxd7!
11 Re1 (targeting the e7-pawn; if 11 Qh4 Nf6 12 d4 Nxd5 and White has to play very accurately to
maintain his activity) 11 ... Nb6 12 Qb3 e6! 13 dxe6 Bxe6 (the point is that the white queen is awkwardly
placed) 14 Rxe6 (or 14 Qd1 Rc8 15 d4 Nd5 16 Bg5 Qa5) 14 ... fxe6 15 Qxe6+ Kh8 16 d4 Qc7 and Black
has plenty of open lines for his rooks to enjoy.
8 Bxd7+ Qxd7 9 Qxd7+
This leads to an endgame which is perfectly acceptable for Black. Alternatively:
a) Avoiding the endgame with 9 Qf4 does not promise much either. After 9 ... Ra8 10 Nf3 (White
gives the pawn back for some development; if 10 Qd4 g6 the queen will be chased around a bit and 11
Nf3 Bg7 12 0-0 0-0 13 Ne5 Qc7 14 Re1 Ng4!? 15 f4 Bf5 16 d3 Rac8 17 h3 Nf6 18 g4? Bd7 19 Be3 h5
would see White overextending, while 10 Qf3 b5 11 Nge2 Bb7 12 Nf4 Rd8 was played in D.Smerdon-
A.Rasmussen, Beijing (rapid) 2008, and now 13 d4 Nxd5 14 Nfxd5 Bxd5 15 Nxd5 Qxd5 16 Qxd5 Rxd5
17 a4 b4 18 Be3 g6 gives Black the better position) 10 ... Nxd5 11 Nxd5 Qxd5 12 0-0 Black has two
equally valid ways of developing.

I prefer the safer 12 ... e6 (12 ... g6 is the alternative) 13 d4 (13 Qg3 tries to make things awkward for
Black, but White only creates problems for himself after 13 ... f6! 14 d4 g5! 15 Be3 Bd6 16 Qh3 h5,
which is slightly better for Black as White’s pieces are in a pickle) 13 ... Be7 14 Qg3 (not 14 b3?! 0-0 15
Bb2 b5 when Black not only has the two bishops, but the better pawn structure too; in V.Gonzalvo Lara-
S.Erenburg, Las Palmas 2003, 16 Rac1 Bb7 17 Rfd1 a5! saw Black not hurrying to contest the file and
instead concentrating on making provoking weaknesses on the queenside) 14 ... 0-0 15 Bh6 Bf6 when the
weakness of the IQP is beginning to tell.
b) 9 Qb3 has also been tried and after 9 ... Qg4 10 g3 b5 11 d4 Bb7 moves such as 12 a4 are not
possible because of 12 ... Nxd5.
9 ... Bxd7 10 Nge2 b5!?

This is probably the most interesting try for Black, since if White does nothing then Black will quickly
stand better.
11 a3
After 11 b4 Rd8 12 Nf4 Bf5 13 a4 g5 14 axb5 axb5 (14 ... gxf4 15 bxa6 Bd3 16 Ra5 Bg7 is too
complicated) 15 Nfe2 Nxd5 already in practical terms Black is a little better; he has the better pawn
structure and two bishops that will roam freely across the board.
11 ... Bc8!?
The aim is to relocate the bishop to b7 and capture the d5-pawn. White must stir up counterplay
quickly because if Black is able to consolidate then he will stand much better.
11 ... Rd8 is also a good option: 12 h4 (if 12 d3 Bf5, while 12 b4 Bf5 13 a4 is probably White’s best
chance and after 13 ... Nxd5 14 Nxd5 Rxd5 15 axb5 axb5 16 Nc3 Re5+ 17 Kd1 f6 it’s level) 12 ... Bc8
13 Nf4 Bb7 14 d4 Nxd5 15 Ncxd5 Bxd5 16 Nxd5 Rxd5 17 Be3 e6 was slightly better for Black in
A.Wuerdinger-M.Bintakis, German League 1999.
12 d4
If White tries to hold on to the pawn with 12 h4 Bb7 13 Nf4 then 13 ... Rd8 wins it back.
12 ... Bb7 13 Nf4
13 0-0 is too slow: 13 ... Nxd5 14 Nxd5 Bxd5 15 Nf4 Bc4 16 Rd1 e6 17 Nd3 Bd6 18 b4 Bxd3 19
Rxd3 Kd7 20 Bd2 Rhc8 21 Kf1 Rc4 and Black later won easily in D.Drollinger-F.Doettling, German
League 2004.
13 ... Rd8 14 Be3 g5 15 Nd3
 Now instead of the slow 15 ... h6, in the game P.Taylor-J.Houska, British League 2005, I should have
continued:
15 ... Nxd5! 16 Nc5 Bc8 17 Bxg5
17 a4 b4 18 Nxd5 Rxd5 is also very good for Black, as shown by 19 Rc1 a5 20 Ne4 Bd7 21 Rc5
(alternatively, 21 Bxg5 Rg8 recovers the pawn, a pawn falls after 21 Nxg5 Bg7, and if 22 Nf3 Bxa4!, and
after 21 Nc5 Bg7 Black will gang up on the cursed d-pawn) 21 ... Rxc5 22 Nxc5 Bc6.
17 ... Rg8 18 h4 f6 19 Bd2 Rxg2 20 Kf1
After 20 Ke2 Rg4 21 N3e4 Nf4+ 22 Kf3 Ng6 23 h5 Nh4+ 24 Ke2 Nf5 the loss of the d-pawn is
unavoidable.
20 ... Rg8 21 N3e4 Kf7 22 Ke2 h6 23 Rhg1 Rxg1 24 Rxg1 f5! 25 Nc3 Nf6
White still has unpleasant issues to face regarding the d-pawn.
26 Be3 f4 27 Bxf4 Rxd4 28 Bg3 Bg7
This position is ideal for Black’s bishop-pair.

The Deadlock after 5 Qa4+

5 Qa4+ is a highly risky choice. White banks on the material strength of the doubled pawns over
harmonious development, which leads to a game that develops naturally for both sides: Black attacks the
pawn and White defends.
In order to understand things better, let’s take a sneak peek at one of the resulting middlegame
positions from the main line:

The diagram shows something of a tabiya position. We see that Black has to all intents and purposes
developed in a natural and logical way - the only thing left is to win the d5-pawn. After ... b6-b5, Black
will try and win the pawn in one of two ways: either through getting a knight to b6 or, if White has
prepared for this eventuality by preparing d5-d6, playing ... Rc8 followed by ... Nc5. White, on the other
hand, has doubled pawns and a queen at risk of being trapped on a3, although he does have a goal: to
defend d5, attack the e7 square and slowly solidify the position.
White’s plans can take the form of:
a) Pressuring the e7-pawn and then switching attention to the f7-pawn.
b) Preparing and then pushing d5-d6 (usually with Bc1-f4).
c) Prophylaxis against Black’s planned d7-knight moves.

B) 5 Qa4+
Regardless of whether it is a knight or a bishop that gives the check, our reply is always the same:
5 ... Nbd7 6 Nc3 g6 7 Nf3 Bg7
 We now arrive at a crossroads between the old and new:

B1) 8 Qb3
B2) 8 Bc4

B1) 8 Qb3
White hopes to catch Black in a rare line, but Black is fine if he remembers the key move 9 ... Nc5!. We
want to blockade the d5-pawn with ... Nc5-e4-d6, not first go ... a6 and ... b5, which doesn’t really work
here.
8 ... 0-0 9 Bc4
Holding on to the pawn. Instead, 9 d4 prevents our desired manoeuvre, but that doesn’t matter because
we can now play to win the d5-pawn: 9 ... Nb6 10 Bc4 Bf5 11 0-0 Rc8 12 Be2 Nfxd5 and Black can be
tremendously satisfied with this position.
9 ... Nc5!

The key move. White’s point is that after 9 ... a6 10 a4 Nc5 11 Qa3 Nce4 (now this idea comes one
move too late) 12 0-0 Nxc3 13 dxc3 Nxd5 14 Rd1 his position is preferable on account of his easy
development, D.Hua-A.Lenderman, Las Vegas 2012.
10 Qa3!
With 10 Qd1 White is taking too many liberties in the opening. Black should open lines quickly with
10 ... e6! 11 dxe6 (11 d4 exd5 12 dxc5 dxc4 is equal) 11 ... Bxe6. Now should White take then 12 Bxe6
Nd3+ gives Black very good compensation for the pawn.
10 ... Nce4! 11 0-0
After 11 Nxe4 Nxe4 12 0-0 Nd6 13 Bb3 b5 14 d3 a5 White is in a tangle.
11 ... Nd6

Black uses his knights to harass the white bishop while at the same time pre-empting any pressure
White may place on the e7-pawn. The plan is to get in ... b7-b5, develop the bishop to b7 and so bring
another piece to attack the d5-pawn.
12 d3
Perhaps the best way of holding on to the pawn, but if 12 Bb3 then 12 ... a5 is rather tempting:
a) 13 Ne5 Nd7 14 Nxd7 (instead, 14 d4 Nxe5 15 dxe5 Bxe5 recovers the pawn and after 16 Re1 Bg7
17 Bg5 Re8 18 Qc5 Bf5 19 Rac1 Rc8 20 Qa7 h6 21 Be3 b5 22 Bb6 Ra8 23 Bxd8 Rxa7 Black is better)
14 ... Bxd7 15 Bd1 b5 16 Qb3 Qb6 and White’s position is a disaster; it’s only a matter of time before the
d5-pawn falls.
b) 13 Nd4 Bd7 and the effective pawn push ... b7-b5 is in the air.
12 ... a6 13 Re1 b5
 Direct. Instead, 13 ... Rb8 was my old recommendation.
14 Bxb5
After 14 Nxb5 axb5 15 Qxa8 bxc4 16 Ne5 Qc7 17 dxc4 Nxc4 18 Bf4 Nd6 19 Rac1 Qb6 20 Nc6
Nxd5 21 Nxe7+ Nxe7 22 Rxe7 Be6 the two minor pieces are not worse than the rook and pawn: for
instance, if 23 Qc6 Qxc6 24 Rxc6 Nf5 and White must give up further material.
14 ... Nxb5 15 Nxb5 Nxd5 16 Nbd4 Qb6 17 Nb3 Bf5 18 Bg5 Rac8 19 Bxe7 Nxe7 20 Rxe7 Bxd3
Black’s bishops are dominant.

B2) 8 Bc4

The most fashionable and the most challenging line. White uses all his resources to defend the central
pawn whilst Black concentrates all his army on attacking it. If Black is successful in winning back the
pawn then he will easily be better.
8 ... a6! 9 d3 0-0 10 Qa3
A little trick is that White cannot play 10 0-0? because of 10 ... b5 11 Bxb5 Nb6! and Black picks up a
piece.
10 ... b6!
 The plan is pretty clear: to develop the bishop to b7.
11 0-0 Bb7

We’ve returned to the tabiya. White’s main choices are:

B21) 12 Bf4
B22) 12 Be3
B23) 12 Re1

After 12 Bg5 Black should reply 12 ... b5 13 Bb3 Nb6! (a straightforward attack on the d5-pawn) 14 Rfe1
Re8 Bxf6 Bxf6 16 Ne4 Nxd5, which sees Black regain the pawn with the better piece coordination.

B21) 12 Bf4
White supports the d5-d6 pawn push, so Black should change tack and play accordingly.
12 ... Rc8!
A key move. Instead, after 12 ... b5 13 Bb3 Nb6 14 d6 e6 it is not easy for Black to win back the
pawn.
13 d6
After 13 Rfe1 b5 14 Bb3 Nc5! 15 d4 Nxb3 16 axb3 Nxd5 17 Nxd5 Bxd5 18 Ne5 Bb7 with the
bishop-pair and weak d4-pawn to target, Black stands very well, while after 13 Rad1 b5 14 Bb3 Nc5
Black will regain the pawn after an exchange on b3.
13 ... e6! 14 Nd2 b5 15 Bb3 Nc5!
 An excellent move, interrupting White’s defence of the d6-pawn.
16 Nce4
If 16 d4 Nd3! 17 Ne2 b4 18 Qa4 Nxf4 19 Nxf4 Qxd6 and Black is much better.
16 ... Nfxe4 17 Nxe4 Nxd3
White is in trouble with ... Bxb2 threatened, winning the queen.
18 Bg5 Qd7 19 Bc2
After 19 Nf6+ Bxf6 20 Bxf6 Qc6 21 f3 Qb6+ 22 Kh1 b4 the poor queen is trapped.
19 ... Rxc2 20 Qxd3 Bxe4 21 Qxe4 Rc4 22 Qd3 Rd4 23 Qe2 Rxd6
Black was a pawn to the good in G.Kacheishvili-V.Bhat, Berkeley 2008.

B22) 12 Be3
Now Black can set in motion a series of forced moves leading to a better if complicated endgame.
12 ... b5! 13 Bb3 Nb6!

Natural. The following moves may be less so, but the key is to keep on attacking.
14 Bxb6
14 Bc5 is not so dangerous: 14 ... Nfxd5 15 Nxd5 Nxd5 16 Bxd5 Qxd5 17 Bxe7 Rfc8 and ... Rc2 is
coming.
14 ... Qxb6 15 Qxe7
Forced.
15 ... a5!

This is the big point: one of White’s minor pieces has to give up its protection of the d5-pawn.
16 Qe3
Likewise, 16 d6 Bxf3 17 gxf3 a4 18 Bd5 Rad8 is good for Black, while after 16 a4? Rfe8 17 Qa3 b4
Black wins the knight.
16 ... Qxe3
Accepting the endgame.
17 fxe3 a4 18 Bd1
White gets into an awful muddle after 18 Bc2 Rfc8 19 a3 (otherwise Black will undermine the c3-
knight with ... a4-a3) 19 a3! Nxd5 20 Nxd5 Rxc2 when Black’s bishops and rooks are simply
overwhelming: for example, 21 Nb4 Rxb2 22 d4 Re8 23 Rfe1 Bf6 24 Nd3 Rc2 25 Nb4 Rc3 26 Kf2 Bxf3
27 Kxf3 Bxd4 and Black is winning.
18 ... Nxd5!
Black is still a pawn down, so needs to have the bishop-pair.
19 Nxd5 Bxd5
 Now b2 is under threat and White needs to defend accordingly.
20 d4
Alternatively:
a) 20 Rf2 is surprisingly insufficient as Black will simply play 20 ... a3 21 d4 Bh6 when a pawn will
fall.
b) After 20 e4 Bb7! (keeping an eye on the e4-pawn) 21 Rb1 Black should open the f-line with 21 ...
f5! 22 exf5 (22 e5 Rad8! again regains the missing pawn) 22 ... Rxf5 23 b3 a3 24 Rc1 Rf7 25 Re1 Rd8
when Black is wonderfully active with a great pawn on a3.
20 ... Rfe8
This second attack is designed to break White’s harmony.
21 Kf2 a3!

Another key move.

22 Bb3
After 22 b3 Rac8 Black will play ... Rc3 next and the position looks awful for White.
22 ... axb2 23 Rab1 Bxb3 24 axb3 Ra2
The b2-pawn must be defended with absolutely everything.
 25 Rfe1 Bf8 26 Re2
26 Ke2 is not so good as Black will play 26 ... Rc8 27 Kd3 Bb4 28 Re2 Rc3+ 29 Ke4 Ba3 30 Ree1
Rc1 31 Kd3 Ra1 and the blockade is in tatters.
26 ... Ba3! 27 Rc2
And not 27 Ke1? Ra1!.
27 ... Ra1 28 Nd2
We’ve followed A.Stripunsky-Y.Seirawan, US Championship, Saint Louis 2012, and now Black
should play:
28 ... f5!

Preventing the white king from transferring itself to the queenside and fixing the e3-pawn on a dark
square. Instead, 28 ... Kf8 29 Ke2 Ke7 30 e4 was already reasonable for White who went on to win in the
game.
29 Kf3
After 29 g3 g5! (remember the king must not cross the e-file) 30 b4 Kf7 31 Rc5 Bxb4 32 Rc7+ Re7 33
Rxe7+ Kxe7 34 Rxb2 Bc3 35 Rc2 b4 again it is only Black playing for a win.
29 ... Kf7 30 Rc7+ Re7 31 Rxe7+ Kxe7 32 Ke2 Rxb1 33 Nxb1 Bd6 34 h3 b4
Shutting out the knight.
35 Kd3 Bg3 36 Kc2
36 Nd2 doesn’t give White any joy either as Black simply plays 36 ... Be1 37 Nb1 Kd6 38 Kc2 Bc3
39 Nxc3 bxc3 when he is winning the king and pawn ending.
36 ... Bf2 37 Kxb2 Bxe3 38 Kc2 Bxd4 39 Kd3 Bf2
A pawn up, Black retains good chances to play for the win.

B23) 12 Re1
 The most accurate move in my opinion.
12 ... Re8 13 Ng5
White relentlessly focuses on a new spot: the f7-pawn. Practice has also seen:
a) 13 Be3 is a quieter move when 13 ... b5 14 Bb3 Nb6! (this is very much like the previous line
where we sacrificed the e7-pawn) 15 d6 (not 15 Bxb6 Qxb6 16 Rxe7? Bf8 and the white queen is
trapped) 15 ... exd6 16 Ng5 d5! (if 16 ... Nbd5 White has the ambitious sacrifice 17 Nxf7!? Kxf7 18 Bg5)
17 Bd4 Qc7 sets a crafty trap: 18 Rac1? would be a big mistake because of 18 ... Ng4!.
b) After 13 Bg5 White cannot hold on to the pawn after the straightforward 13 ... b5! 14 Bb3 Nb6 15
Re2 Nbxd5 16 Nxd5 Bxd5 17 Bxd5 Nxd5 and Black has no problems whatsoever.
c) The enterprising 13 d6 was tried very recently, although I have the feeling that it might be because
White got his lines confused. After 13 ... e6 14 Ne4 b5 15 Bb3 a5! 16 Bg5 h6 17 Bxf6 Nxf6 18 Rac1
Nxe4 19 dxe4 b4 20 Qa4 Qxd6 Black was already much better in N.Getz-J.Hammer, Fagernes 2014.
13 ... Rc8!?

The most challenging and the most provocative idea: Black plans to play ... b5 followed by ... Nc5,
interrupting White’s attack on e7 and threatening the d5-pawn. White has a whole variety of moves
available to him, of which the most important are:
B231) 14 Be3
B232) 14 d6

Otherwise:
a) 14 Bd2 is not incisive enough and Black can now begin his fight-back: 14 ... b5 15 Bb3 Nc5!
(please don’t play the natural 15 ... Nb6 as here we would get into a bit of trouble after 16 d6 e6 17 Nge4
and Black doesn’t win back the pawn) 16 d6 e6 17 Nce4 Nfxe4 18 Nxe4 Nxd3 19 Ba5 (this looks very
impressive for White, but the offside queen and Black’s threat of ... Bxb2 lead to serious problems for
White) 19 ... Qd7 20 Bxe6 and now in O.Perez Mitjans-J.Avila Jimenez, Sant Marti 2010, 20 ... Qxe6! 21
Qxd3 Bxe4 22 d7 Bxd3 would have won the game.
b) 14 Bf4 b5! 15 Bb3 and now Black should counter attack with 15 ... Nc5! (15 ... Nb6 doesn’t deal
with the threat of 16 d6) 16 d6 e6 17 Rad1 (if 17 d4 Nd3!) 17 ... Nh5 18 Be3 Nxb3 19 axb3 f5!, depriving
the white knights of the e4-square so that Black can get on with the plan of winning the d6-pawn.
c) If 14 Nge4 the solution is simply to capture: 14 ... Nxe4 15 Nxe4 and now Black targets the bishop
on c4 with 15 ... Ne5!.

Once again White’s position makes a strange impression, but we must watch out for that d5-d6 move.
After 16 Bf4 (16 Bb3?, of course, loses to 16 ... Nxd3) 16 ... b5! (16 ... Nxc4 17 dxc4 Rxc4 18 Rad1
Bxd5 19 Qxa6 is more complicated but better for Black too, R.Yankovsky-S.Erenburg, Las Vegas 2012)
17 Bxe5 Bxe5 18 Bb3 Qb6 19 Rac1 Rxc1 20 Rxc1 a5 and White must abandon the defence of the d5-
pawn as his queen is in deep trouble: 21 Qc5? Rc8 22 Qe3 Rxc1+ 23 Qxc1 a4 when Black has won the
pawn back and has everything he could ask for.

B231) 14 Be3 Ne5

 Now that ... Nc5 has been prevented, Black instead takes advantage of the weak position of the bishop
on c4.
15 Rad1
15 d6 is another attempt, but after 15 ... Nxc4 16 dxc4 Qxd6 17 Qxd6 exd6 18 Bxb6 Rxe1+ 19 Rxe1
Rxc4 Black has no problems, while he has the advantage in the event of 15 Bd4 Nxc4 16 dxc4 Rxc4 17
Rad1 b5 18 Qb3 Nxd5 19 Bxg7 Kxg7.
15 ... h6!
Again we see Black counterattacking.
16 Nge4
Black intends to swap off a white knight should one arrive on e4.
16 ... Nxe4 17 Nxe4 Nxc4 18 dxc4 Rxc4 19 f3
19 Qd3? falls for a cheap trick: 19 ... Qxd5! 20 Qxd5 Bxd5 21 Rxd5 Rxe4 and although White
managed to hold on for a draw, Black was much better in K.Rasmussen-J.Aagaard, Aalborg 2006.
19 ... f5!

A big improvement from the Georgian Grandmaster Giorgi Kacheishvili, who is particularly skilled
in handling these IQP positions.
 20 Nf2 Qd6 21 Qd3 Rec8 22 Bxh6 Bxh6 23 Re6 Qxd5 24 Rxg6+ Bg7 25 Qxd5+ Bxd5 26 Rxd5
Kf7 27 Rg5
27 Rxb6 Bd4 costs White an exchange.
27 ... Rc1+ 28 Nd1 Bf6 29 Rh5 e6 30 Rd6 Ke7 31 Rd3 Rd8 32 Rh7+ Ke8
Losing a piece, White threw in the towel in G.Antal-G.Kacheishvili, Lubbock 2009.

B232) 14 d6

An obvious idea and yet White does not get anything out this move. The d6-pawn has been effectively
thrown to the wolves and the threat to f7 can be easily met.
14 ... e6! 15 Bf4
Not the only way to try to cover d6:
a) 15 Qb4 h6! chases away the knight. This is a common theme in this variation: first Black defends
against the threat and then he starts to drive White back. Now:
a1) 16 Nxe6? looks like a great sacrifice, but White has pinned his own knight, so unsurprisingly
Black can refute this tactic: 16 ... a5! 17 Qb5 (or 17 Qb3 Nc5) 17 ... Nc5! and White is in deep trouble.
a2) 16 Nge4 Nd5! (now that the bishop on c4 is protected, it makes no sense to swap knights on e4)
17 Qb3 Nxc3 18 Nxc3 b5 19 Qb4 Ne5 and White is struggling.
b) 15 Nge4 leaves an unprotected bishop on c4. Yes, that only means one thing: the knight on e4 can
be taken. A.Demuth-R.Kreisl, Merlimont 2011, continued 15 ... Nxe4! 16 Nxe4 Bxe4! 17 Rxe4 b5 18 Bb3
Nc5 19 Be3 (or 19 Re3 a5 20 d7 Qxd7 21 Qxa5 Nxd3 and White is in a lot of trouble; Black is simply so
active with multiple targets on the dark squares) 19 ... Qxd6! (Black has to get rid of this pawn) 20 Bxc5
Rxc5 21 Rd1 and now with the dominating 21 ... Bd4 Black could have obtained a practically won game.
15 ... b5!
Pushing White back. Instead, 15 ... h6 results in complications after 16 Nxe6 fxe6 17 Rxe6 Kh8 18
Re7 with a tricky position.
16 Bb3
Sacrificing a piece only gives White a lost position after some accurate moves from Black: 16 Bxe6?!
fxe6 17 Nxe6 Qb6 18 Qb3 Kh8 (getting the king to safety) 19 d4 b4 20 Na4 (if 20 Ng5?? and the threat of
Nf7+ is very easily met by 20 ... h6) 20 ... Qc6 21 f3 Nd5 and now 22 Rac1? loses to 22 ... Rxe6! 23
Rxc6 (or 23 Rxe6 Bxd4+ 24 Kf1 Qb5+ 25 Ke1 Nxf4 26 Rxc8+ Bxc8 27 Re8+ Kg7 28 Rg8+ Kh6 and
Black is winning) 23 ... Rxe1+ 24 Kf2 Bxc6 25 Kxe1 Nxf4 when Black has too many pieces for the
queen.
16 ... Qb6
Threatening 17 ... b6.
17 Qb4 h6 18 Nge4 Bxe4!

This is played for tactical reasons.

19 dxe4
We can see Black’s points after both 19 Be3 a5 20 Qa3 Qb8 21 Nxe4 a4 22 Nxf6+ Bxf6 23 Bd1 Be5
24 Qb4 Bxd6 25 Qh4 h5 and 19 Nxe4 Nxe4 20 dxe4 Bxb2!.
19 ... Ng4 20 Bg3 Nge5 21 Rad1 Nc6!
White’s queen is in trouble, so he is forced to undertake drastic measures.
22 Na4 bxa4 23 Qxb6 Nxb6 24 Bxa4 Bxb2 25 d7 Nxa4 26 dxe8Q+ Rxe8
Black should have good winning chances here.

C) 5 Nc3

This is the most common reply from White in the 2 c4 variation. He avoids some of the critical lines
that can be found in the Panov-Botvinnik Attack and instead concentrates on obtaining a good IQP position
- it is for this reason that this is perhaps the trickiest system in the Pseudo-Panov.
5 ... Nxd5
White now has a choice between allowing Black to fianchetto the king’s bishop and forcing Black into
an ... e6 system:

C1) 6 Bc4
C2) 6 Nf3

Less often White tries:

a) 6 d4 Nc6 reaches a position we discussed at the start of Chapter Five.
b) If White plays the slightly incredulous 6 Qb3, Black should react calmly and aim for a ... g6
system: 6 ... Nb6 7 Nf3 Nc6 8 Bb5 Be6! (developing with tempo) 9 Bxc6+ bxc6 10 Qc2 g6 and the
position is at the very least equal.
c) White usually avoids the immediate 6 Bb5+ because of 6 ... Bd7. Now:
c1) 7 Bc4 Nb6 8 Bb3 Nc6 9 d4 Bf5 10 Nf3 (if 10 d5 Ne5!) 10 ... e6 11 0-0 Be7 with a favourable
IQP position for Black due to the bishop on f5.
c2) 7 Qa4 Nb6 8 Qb3 e6 9 Nf3 Nc6 10 0-0 Be7 11 d4 0-0 12 Bf4 Na5 13 Qc2 Bxb5 14 Nxb5 Nd5
(Black has blockaded the d4-pawn and got rid of the dangerous light-squared bishop, so he is doing pretty
well) 15 Bd2 Nc6 16 Nc3 Bf6 and Black was fine in Z.Korpics-P.Tomcsanyi, Gyor 1992.

C1) 6 Bc4

6 ... Nb6 7 Bb3 g6!?

With White not developing in a critical manner, once again we are able to fianchetto:
8 Nf3
8 Qf3 threatens f7, but is a little dubious. However, after 8 ... e6 9 Nge2 Bg7 10 d4 Nc6 11 0-0 Black
must be careful:
a) 11 ... Nxd4?! is just plain greedy and after 12 Nxd4 Qxd4 13 Rd1! Qe5 14 Bf4 Qf5 15 g4 Qa5 16
Ne4 0-0 17 Bd2 Qb5 18 a4 Qe8 (or 18 ... Qc6 19 Rac1) 19 Bc3! f6 20 Bb4 White picks up an exchange.
b) 11 ... 0-0 12 Rd1 Na5! (Black prevents the advance of the d-pawn) 13 Bc2 Bd7 (more model play,
transferring the bad bishop to the long diagonal) 14 Bf4 Bc6! 15 Qg3 Bd5 16 Bd6!? Re8 17 Nb5 Nac4
saw Black blockade the d5-square in M.Chiburdanidze-S.Lalic, Novi Sad Olympiad 1990.
 8 ... Bg7 9 d4 Nc6

Play has transposed to the Panov proper, specifically Line B of Chapter Five.

C2) 6 Nf3 Nc6 7 Bb5

This move is played in order to discourage Black from a ... g6 system. It virtually forces Black to play
the defensive move:
7 ... e6 8 0-0
8 d4 is possible but has no independent value, while after 8 Ne5 Bd7 9 Bxc6 Bxc6 10 Nxc6 bxc6 11
0-0 Bd6 12 d4 0-0 Black’s strong knight fully compensates for the slightly weak pawn on c6.
8 ... Be7 9 d4 0-0
 Here White can play two moves:

C21) 10 Re1
C22) 10 Qe2

Moves such as 10 Bd3 do not trouble us after 10 ... Nf6! followed by ... b7-b6 with easy development.
Before we look at the various moves I would like to stress some strategic themes in the arising IQP
positions. The isolated pawn provides White with a great deal of attacking and dynamic possibilities:
a) A kingside attack, especially by lining up the queen and light-squared bishop on the b1-h7 diagonal.
b) Good piece play in the centre, which often leads to explosive sacrifices on f7, g6 and even e6.
c) The thematic d4-d5 push.
As we can see, these are all factors pertaining to an initiative; Black, because of the weak d4-pawn,
has the long-term prospects. This means that Black’s simple objective is to obtain a solid position and
where possible exchange minor pieces. If White has a normal position with few attacking prospects then
Black can look forward to an endgame where he will press against that d4-pawn.

The Danger of d4-d5

This is a particularly dangerous threat and something Black needs to bear in mind continuously. Let me
give you the following example:
 In this rather ordinary looking position, it looks as though Black can simply play 15 ... Rac8 with a
good position. This, however, would be a terrible mistake because after 16 d5! exd5 17 Bf5 Ra8 18 Nb5
Qb8 19 g3! White wins material.

Releasing the c8-bishop

Generally we want to adopt the following formation which is designed to put pressure on the d4-pawn.

The black queen tends to move to c7 if attacked by a knight on e4 and the black knight on d5 may also
go back to f6 before the bishop on c8 is sorted. We leave the question of the light-squared bishop last for
one reason - because if White retreats his bishop to d3, it might be possible to play ... b6 and ... Bb7,
developing the bishop to its ideal diagonal. For instance, after 10 Re1 Qd6 11 a3 Rd8 12 Bd3 Nf6 13 Be3
Black can and should play 13 ... b6 followed by ... Bb7, reaching a satisfactory position. If ... b6 is not
possible, the bishop will make a slow journey to freedom from d7 and may even in the short term hide
itself on e8 to cover vital kingside points.

C21) 10 Re1
 10 ... Qd6!?
10 ... Bd7 was my old main line, but I have lost my enthusiasm for it as after 11 Bd3 Nf6!? (11 ... Rc8
is a solid alternative) 12 a3 Rc8 an uncomfortable move for Black to face is 13 Bf4!. Now Black must
play 13 ... Qb6! 14 Rc1 g6 (14 ... Qxb2? doesn’t work on account of 15 Nb5! and after 14 ... Rfd8 15 b4
the queen doesn’t have too many squares available) 15 Na4 Qa5 16 Bd2, but White’s position is the
easier to play: for example, 16 ... Qd5 17 Nc3 Qd6 18 Nb5 Qb8 19 g3 a6 20 Bf4 Qa8 21 Nd6 is slightly
better for White.
After the text, generally speaking White has three major plans: Ne4, a2-a3 covering the b4-square,
and to retreat the bishop to d3. We will focus on:

C211) 11 Bc4
C212) 11 a3
C213) 11 Ne4

Note that 11 Bd3 Rd8 12 a3 will transpose to Line C212, while after the direct 11 Bg5 Black can simply
develop: 11 ... Bxg5 12 Nxg5 Nf6 13 Bxc6 Qxc6 14 Rc1 Qb6 15 Qd3 h6 16 Nge4 Nxe4 17 Nxe4 Bd7 18
Re3 Qd8 19 Rf3 Qh4 20 Qe3 Bc6 and Black’s position is preferable on account of the weak d4-pawn,
B.Heberla-Q.Loiseau, Cappelle-la-Grande 2014.

C211) 11 Bc4
 This is not too dangerous.
11 ... Rd8 12 Bb3 Nf6 13 Be3 b6
With the bishop gone from b5, Black can simply develop his light-squared bishop on the long diagonal
without any problems.
14 Qe2 Bb7 15 Rad1 Na5 16 Bc2 Rac8

These are all standard developing moves. Now White tries to move forward to generate something,
but Black can kick him back.
17 Ne5 Nc6 18 Ne4 Nxe4 19 Bxe4 f5! 20 Bf3 Bf6 21 Bf4 Nxd4 22 Rxd4 Qxd4 23 Bxb7 Qxf4 24
Nf7
This was A.Faizulaev-I.Khenkin, Tashkent 2013, where Black would have had the superior position
after:
24 ... Kxf7! 25 Bxc8 Rd2 26 Qxe6+ Kg6
The black rook will cause problems.

C212) 11 a3
A careful choice. White will now retreat the bishop on b5 to the more pertinent b1-h7 diagonal.
 11 ... Rd8

The point behind Black’s tenth move becomes clear - he gets to put pressure on the d4-pawn
incredibly quickly.
12 Bd3
White needs to be careful, as shown by 12 Qd3 Bd7! 13 Ba4 Nxc3 14 Bc2 (14 bxc3 Nxd4 nets
material) 14 ... Ne2+ 15 Rxe2 g6 16 Bh6 Bf6 17 Rd1 Be8 when Black stands very comfortably. He has
the desired solid position and the requisite pressure on the d4-pawn, B.Jaracz-I.Rausis, Bern 2006.
After 12 Qc2 Black should also develop with 12 ... Bd7. Here if White lines up the queen and bishop
the wrong way round with 13 Bd3, Black can play 13 ... h6! 14 Bd2 (if 14 Ne4 Black retreats the queen
with 14 ... Qc7 and then plays ... Rac8) 14 ... Rac8 15 Nxd5 exd5!? (the pawn structure is symmetrical,
but just look at Black’s pieces!) 16 Bc3 Bg4 17 Ne5 Nxe5 18 dxe5 Qd7! (even better than the 18 ... Qb6
of D.Arutinian-A.Deviatkin, Canberra 2011) 19 Qd2 d4 20 Bb4 Bxb4 21 Qxb4 Bf5! 22 Rad1 Bxd3 23
Rxd3 Rc2 24 h3 Qf5 25 Rf3 Qc8 when he is very active and, indeed, enjoys the initiative.
12 ... Nf6
The general reaction to the bishop retreat. Black moves the knight to put pressure on the d4-pawn.
13 Be3
A sensible choice as White defends the pawn. Otherwise:
a) 13 Nb5 is a waste of time. Black will simply play 13 ... Qd7 14 Bf4 a6 15 Nc3 b5 and to make any
sense of the position White must play 16 Ne5 Nxe5 17 Bxe5 Bb7, but Black is very happy here.
b) 13 Ne4 doesn’t work as Black can make some exchanges: 13 ... Nxe4 14 Rxe4 b6 (getting on with
development) 15 Bf4 Qd5 16 Qe2 Bb7 17 Bc4 Qh5 and now 18 d5? doesn’t work because of 18 ... Na5
19 dxe6 Bxe4 20 exf7+ Kf8 21 Qxe4 Nxc4 22 Qxc4 Qxf7, winning.
13 ... b6
Creating a path for the bishop on c8.
14 Qe2
White can also connect the rooks with 14 Qc2 Bb7 15 Rad1 and now Black must play the careful 15
... g6! (he must be careful not to play 15 ... Rac8 16 d5!, as in C.Matamoros Franco-P.Dias, Campillos
2007, when 16 ... exd5 17 Bf5 Ra8 18 Nb5 Qb8 19 g3 and Bf4 is a problem).
 To make sense of his position White has to position the queen on a more active square: 16 Qe2 Rac8
17 Bb1 and now d4-d5 is threatened, so Black must play 17 ... Nd5 18 Ne4 Qc7 19 Bh6 Bf8 20 Bg5 (or
20 Qd2 Nce7 21 Rc1 Qb8 22 Rxc8 Rxc8 23 Bxf8 Rxf8 24 Ne5 Qd8 25 Ng4 Nf5 and Black has achieved
a solid position) 20 ... Re8!? (20 ... Be7 leads to a repetition) 21 Qb5 Be7 22 Ng3. Here in D.Bojkov-
I.Enchev, Sunny Beach 2009, Black should have played 22 ... Bxg5 23 Nxg5 Na5 24 Ba2 Qe7 when
White’s attack leads nowhere.
Likewise, after 14 Qd2 Bb7 15 Rad1 Rac8 16 Bf4 Qd7 17 Ne5 Nxe5 18 dxe5 Qc6 19 f3 Nd5 Black
stands well.
14 ... Bb7 15 Rad1 h6
Ruling out Bg5.
15 ... Rac8 is also possible, but allows 16 Bg5 h6 17 Bh4 and here Black should probably play 17 ...
Nd5, rather than 17 ... Nxd4?? 18 Nxd4 Qxd4 19 Be4, which was already winning in B.Smith-B.Finegold,
Lubbock 2010.
16 Bb1
After 16 d5 Nxd5 17 Nxd5 exd5 18 Bb1 Bf6 19 Qc2 g6 20 Bxh6 Ne5 21 Nxe5 Bxe5 22 h3 d4 Black
is incredibly active. Just look how the d-pawn is more of a space-gaining asset than a weakness.
16 ... Bf8
 All Black does is to simply coordinate his pieces in preparation for White’s plans. Here Black covers
h6 against any sacrifices or should he need to play ... g7-g6.
17 Nb5 Qb8 18 Qc2 Ne7 19 Ne5 Be4 20 Qc1 Bxb1!
The fewer minor pieces on the board, the less White can attack. In fact what happens next is typical of
positions with an IQP: White gets pushed back, Black activates his forces, and under pressure White
crumbles.
21 Qxb1 Qb7 22 Rc1 Ned5 23 Qd3 Rac8 24 Bd2 a6 25 Rxc8 Rxc8 26 Nc3 Nxc3 27 Bxc3 b5 28
Bd2 Qd5 29 Be3 Bd6 30 Kf1 Qa2 31 Bc1 Nd5 32 g3 Nb6 33 Kg2 Qd5+ 34 f3
White resigned as 34 ... Rxc1 wins material, N.Miezis-A.Dreev, Reykjavik 2004.

C213) 11 Ne4
I think this is the most tempting move.
11 ... Qc7

Once again White must decide whether or not to make a useful little move with his a-pawn:

C2131) 12 Bd2
C2132) 12 a3

Alternatively:
a) 12 h4 looks aggressive enough, but Black can play 12 ... Qb6 13 Bxc6 (if 13 Bd3 Black can simply
go 13 ... Nxd4 14 Nxd4 Qxd4 and White has nothing) 13 ... Qxc6 14 h5 b6 15 h6 g6 16 Ne5 Qe8 17 Bd2
Bb7 18 Rc1 f6 19 Nc4 Qd7 when White’s attack leads nowhere.
b) 12 Bg5 Bxg5 13 Nexg5 h6 14 Ne4 Bd7 15 Rc1 Qb6 16 Bc4 Rfd8 17 Bb3 Be8 kept everything
covered in N.Miezis-S.Shankland, Biel 2011.
c) 12 Bd3 Rd8 (again putting pressure on the d4-pawn) 13 a3 Bd7 14 Bg5 Bxg5 15 Nfxg5 h6 is once
more fairly comfortable for Black.

C2131) 12 Bd2

Black should now carry on as normal:

12 ... Bd7 13 Nc5
Now that the bishop is on d2, if 13 Rc1 the position lends itself to 13 ... Qb6!, preventing Bd3 as the
pawn on d4 will be en prise. After 14 Bc4 Back can play the natural 14 ... Rfd8 15 Nfg5 Be8, protecting
against any possible sacrifices, followed by ... Rac8.
Instead, after 13 Bd3 Ndb4! 14 Bb1 Rac8 15 Nfg5 g6 White’s attack is going nowhere.
13 ... Be8 14 Nb3 Ncb4!
14 ... Bf6 15 Qe2 Qb6 16 Bc4 Nce7 17 Ne5 Rc8 18 Bd3 Ng6 19 g3 Ba4 was also satisfactory for
Black in C.Peptan-I.Kadimova, Plovdiv 2008.
15 Bxe8 Rfxe8 16 Ne5 Bd6 17 Re2 Rad8 18 Nf3 Nc6!
This ties White to the defence of d4. White has no initiative, so Black stands better.

C2132) 12 a3
 A common and useful IQP move, preventing any ... Nb4 ideas.
12 ... Rd8!?
I am sticking to the main idea of pressuring the d4-pawn.
13 Bg5
One of many moves:
a) 13 Qc2 allows Black a tactical shot: 13 ... Ne5! 14 Qb3 Nxf3+ 15 Qxf3 Bd7 16 Bd3 h6 17 b4 a6
18 Bd2 Bc6 19 Rac1 Rac8 20 Bb1 Qb8 21 Qd3 Bb5 22 Qb3 Rxc1 23 Bxc1 Qc7 and Black had no
problems in L.Ortega-I.Khenkin, Frascati 2005.
b) Retreating the bishop with 13 Bf1 allows Black to play the unusual 13 ... e5 14 Nxe5 Nxe5 15 dxe5
Ne3 16 Bxe3 Rxd1 17 Raxd1 Qxe5 18 Nc3 Be6 19 Bxa7 Qa5 when he stood well in E.Bareev-Le Quang
Liem, Moscow 2010.
c) 13 Qd3 allows 13 ... a6 14 Ba4 b5 15 Bc2 g6 (setting up the wall) 16 Bg5 Bb7 17 Rac1 Bxg5 18
Nfxg5 Qe7 19 Nf3 Nf6 and Black has an excellent anti-IQP position; there is no attack to fend off and we
have easy play against the d4-pawn.
d) 13 Bxc6 is useless because of 13 ... Qxc6 14 Ne5 Qe8 15 Qb3 (15 Qf3 b6 is slightly better for
Black too) 15 ... b6 16 Bd2 Bb7 17 Qh3 Rac8 18 Nf3 Nf6 19 Neg5 and now Black should play 19 ...
Bxf3 (instead of 19 ... h6 which allowed the tactic 20 Nxf7!? in A.Hoare-P.Batchelor, British
Championship, Torquay 2013) 20 Nxf3 Nd5 with a deliciously simple game.
e) After 13 Bd3 Nf4 14 Bf1 b6 15 b4 Bb7 16 Bb2 Rac8 17 Rc1 Qb8 White doesn’t have an attack.
f) 13 Bd2 is another waiting move and after 13 ... Bd7 (as White has interrupted the defence of the d4-
pawn, Black can also get away with playing 13 ... Qb6) 14 Bd3 Be8 (protecting the f7-pawn in case of
Neg5) 15 Rc1 Qb6 I don’t see what White is doing.
13 ... Bd7
 14 Bxe7
Again, White has tried a few things:
a) 14 Rc1 Qb6 15 Bxc6 (if 15 Bxe7! Ncxe7 16 Bd3 Nf5, closing off the key diagonal) 15 ... Bxc6 16
Ne5 and now Black should probably play the cautious 16 ... Be8 (16 ... Bxg5 17 Nxg5 was good for
White in Y.Pelletier-A.Dreev, Cap d’Agde (rapid) 2002) 17 Nc4 Qb5 18 Qg4 Bxg5 19 Qxg5 b6 20 Ned6
Qd7 21 Qe5. White’s position looks impressive, but Black can now beat him back with 21 ... Ne7 and ...
b7-b5.
b) 14 Bd3 Bxg5 15 Nexg5 h6 and White must retreat as after 16 Bh7+? Kf8 the bishop gets trapped on
h7: 17 Nxf7 Kxf7 18 Ne5+ Nxe5 19 dxe5 Qb6 20 Qf3+ Ke7 and the black king should be safe enough.
14 ... Ndxe7
The safest possibility, keeping control over the e5-square, and not 14 ... Ncxe7?! 15 Bxd7 Qxd7 (or
15 ... Rxd7 16 Ne5 Rdd8 17 Ng5 and the white knights are rather frightening) 16 Ne5 Qb5 17 Qh5 g6 18
Qh3, which is slightly better for White.
15 Qe2
15 Bd3!? is probably the right way to go (15 Nfg5 simply gets kicked back with 15 ... h6), although if
15 ... Nf5 16 Neg5 h6! 17 Bxf5 exf5 White is more or less obliged to sacrifice a piece with 18 Nxf7 Kxf7
19 d5 (if 19 Qb3+ Kf8 20 Nh4 Qd6) 19 ... Be8 20 Qd3 Kg8 21 Qxf5 Ne7 22 Qe6+ Kf8 when ... Bf7 will
cover most of White’s threats.
15 ... Be8 16 Nc3 a6 17 Bxc6 Bxc6
 From now on the position is better for Black. All that is left is to improve the activity of his pieces.
18 Rad1 Rd6 19 Ne4 Bxe4 20 Qxe4 Qc6 21 Qf4 Rad8 22 Ne5 Qe8 23 Nc4 Rc6 24 Na5 Nd5! 25
Qe5 Rc7 26 Rc1 Rxc1 27 Rxc1 Qb5 28 b4 Qd3
The computer’s choice, working out that after 29 Nxb7 Qxa3 30 Qe1 Rb8 the b-pawn is gone. Instead,
28 ... h6 was played in S.Franklin-J.Speelman, British League 2014, where Black eventually ground out
the win.

C22) 10 Qe2

White has the idea of placing a rook on d1.

10 ... Bd7 11 a3
A good defensive move. White gets ready to line up his queen and light-squared bishop towards the
h7-square.
Instead, after 11 Rd1 Rc8 12 Nxd5 exd5 White’s advantage is minimal. Indeed, following 13 Ne5 a6
14 Bd3 Bf6 15 Bf4 g6 16 Bc2 Re8 17 Bb3 Bf5 White really had no advantage in G.Milos-A.Dreev,
Moscow 2001.
11 ... Rc8 12 Bd2 Bf6 13 Ne4?!
 This natural move was played in the game J.Polgar-A.Dreev, Moscow (rapid) 2002, but now White
has big problems defending the d-pawn. Instead, 13 Qe4 Nce7 would have been about equal.
13 ... a6
Not 13 ... Nxd4??, which loses to 14 Nxf6+ Qxf6 15 Nxd4 thanks to the loose bishop on d7.
14 Bd3!?
14 Bxc6 Bxc6 gives Black a firm control over the key blockading square d5.
14 ... Nxd4 15 Nxd4 Bxd4 16 Nd6 Bxb2!
A strong counterblow. Black sacrifices the exchange for two pawns and a well coordinated position.
17 Rab1 Bxa3 Nxc8 Qxc8 19 Bxh7+ Kxh7 20 Qd3+ Kg8 21 Qxa3 Bc6
With two good pawns for the exchange, Black has no problems and, indeed, Dreev went on to win.

Summary
There are three very different systems within the 2 c4 variation. 5 Bb5+ is not commonly played and as a
result there have not been too many recent advances in this line. 5 Qa4+ is still somewhat popular, but I
believe Black can equalize with relatively few problems. To begin with Black develops with the aim of
winning back his pawn, and does this by developing the light-squared bishop to b7. The key tabiya arises
at move 12 and here:
a) If White plays 12 Bf4, planning d5-d6, then Black develops a rook to c8 and aims for ... b5 and ...
Nc5.
b) 12 Be3 leads to a worse endgame for White after 12 ... Nb6, although this is the only line that
requires some degree of memorization.
c) Attacking the pawn with 12 Re1 forces Black to play 12 ... Re8! and after 13 Ng5 Rc8!. Black has
the added options of playing a knight to e5.
5 Nc3 is a more tricky line in the sense in that the resulting positions are more fluid. Of course, Black
should recapture the pawn immediately and if allowed we should transpose into fianchetto systems.
White’s 7 Bb5, however, forces 7 ... e6 and after 10 Re1 I am recommending 10 ... Qd6, protecting the
knight and allowing the c8-bishop to develop either to d7 or to b7. Black’s approach is provocative; the
fundamental idea is that after 11 Ne4 he will retreat the queen to c7 and play ... Bd7 without worrying
about Nxd5 ideas. As in most IQP positions, our aim is to solidify the position, react to White’s threats
and eventually enjoy a superior endgame.
Chapter Fourteen
The Two Knights Variation
1 e4 c6 2 Nc3 d5 3 Nf3 Bg4

The Two Knights Variation is characterized by White developing his knights instead of pushing the d-
pawn. Although this system does initially not look promising, it has the advantage of keeping the pawn
structure flexible. White can build up with a controlled d2-d4 or play the position like the King’s Indian
Attack with d2-d3 followed by a kingside attack with g2-g4.

Why 3 ... Bg4?

The most challenging move after 3 Nf3 is undoubtedly 3 ... Bg4, although at first sight it may appear to be
a little strange. Whilst researching this variation on the Internet, I happened to come across a chess forum
where players discussed their opinions on openings. It became obvious that when talking about the Two
Knights Variation many people did not understand the reasoning behind this bishop move. People assumed
that handing the bishop-pair to White on a plate was tantamount to conceding a small but lasting
advantage.
However, this is not the case. Firstly, if we take a look at the pawn structure we notice that most of
Black’s pawns will be on light squares. This means he has a light-square bind that can restrict the activity
of White’s light-squared bishop. Secondly, the position is not yet of an open nature, so the advantage of
the bishop-pair is not apparent. Thirdly, we must look at the potential of both White’s f3-knight and
Black’s light-squared bishop, and also some tactical possibilities. If Black wanted to keep the bishop, in
order not to become ‘bad’ it would need to be developed outside the pawn chain, either on f5 or g4.
Let’s see what happens when the bishop is placed on f5:
3 ... dxe4 4 Nxe4 Bf5?! 5 Ng3! Bg6 6 h4 h6
White has not expended a tempo on d2-d4 and now this can be used to good effect:
7 Ne5! Bh7?! 8 Qh5!
 Oh dear. Sadly for Black he is forced to play 8 ... g6 with an absolutely miserable position. In
addition, look at the power of the knight on e5 and how with just two pieces White wrecks the black
position: 9 Qf3 Nf6 10 Qb3 Nd5 11 Qxb7, etc.
Thus the bishop should go to g4 (3 ... Nf6!? was my Dangerous Weapons recommendation; much as I
like it, here I will be sticking with my original choice):

There are two main methods of play for White in the Two Knights Variation: the positional approach
based on the classical school of thought, and the aggressive approach.

A) 4 d4
B) 4 h3

Instead, 4 Be2 is not really very challenging. Black simply needs to develop: 4 ... e6 5 d4 Nd7 6 0-0 dxe4
7 Nxe4 Ngf6 8 Nxf6+ Nxf6 when the backward d-pawn will be an endgame weakness. Moreover, we
have the familiar principles of utilizing the half open d-line and perhaps the undermining ... b7-b5 push,
P.Mulet-I.Hera, Wroclaw 2014.
A) 4 d4!?

A critical move advocated by the Spanish GM Francisco Vallejo Pons.

4 ... Nf6 5 h3
What else? For instance, 5 e5 Nfd7 6 h3 Bxf3 7 Qxf3 e6 and Black has successfully closed the
position.
5 ... Bxf3 6 gxf3
This is White’s point. He obtains the bishop-pair and a strong pawn centre and claims that the doubled
pawns will not matter until an endgame.
Instead, 6 Qxf3 dxe4 transposes to Line B2, below.
6 ... e6 7 Qd3
With the plan of moving the bishop from c1 followed by 0-0-0. Others:
a) After 7 Bf4 Black can make use of the loose bishop and play 7 ... Bb4 8 Qd3 Bxc3+:
a1) Sacrificing a pawn with 9 Qxc3!? dxe4 10 fxe4 (if 10 Rg1 Nd5!) 10 ... Nxe4 11 Qd3 Qa5+ 12 c3
Nf6 13 Rg1 g6 14 Rg5 is possible, although we are very solid after 14 ... Nd5.
a2) 9 bxc3 reaches a strange position.

White has completely ruined his pawn structure, but has obtained the bishop-pair. Here 9 ... 0-0 10
Rg1 Nbd7 11 Bh6 Ne8 12 c4 dxc4 13 Qxc4 Qf6 14 Bd2 Qxf3 15 Qd3 Qxd3 16 Bxd3 e5 17 Bb4 exd4 18
Bxf8 Kxf8 is good for Black.
b) After 7 Be3 I like the same response, but with a different follow-through: 7 ... Bb4 8 Qd3 0-0 (8 ...
Bxc3+ is no longer so good: 9 Qxc3 dxe4 10 Rg1 Nd5 11 Qb3 Nxe3 12 Qxe3 and Black must return the
pawn) 9 Rg1! (after 9 0-0-0 Qa5 10 a3 Bxc3 11 Qxc3 Qxc3 12 bxc3 dxe4 13 fxe4 Nxe4 Black wins a
pawn which should give him a slight advantage) 9 ... Nbd7 10 a3 dxe4 11 fxe4 Bxc3+ 12 bxc3 Re8 13
Be2 e5 14 dxe5 Rxe5 15 f3 Qa5 16 Bd4 Rg5 (White has a big problem with the safety of his king) 17 Kf2
Rg6, clearing the way for the queen to switch sides.
c) If 7 Rg1 I like 7 ... Qb6!. Now 8 a4 would be a mistake as this would mean that there is no safe
place for the white king. After 8 ... a5! a knight will hop into b4 and 9 Qd3 Na6 leaves Black doing well.
d) White needs to maintain the central tension so 7 e5 would be a terrible move strategically. White
closes the position down and it is only Black that can benefit from this. He will simply play 7 ... Nh5
followed by ... Ng7-g6 and then reroute the knight to f5.
e) 7 Bg2 has also been tried, but the bishop is shut in and after 7 ... Be7 (7 ... Bb4 is also possible) 8
0-0 0-0 9 Ne2 Nbd7 10 Ng3 Black should play the energetic 10 ... c5 (10 ... Ne8 wasn’t as convincing in
D.Tambussi-N.Del Cuadro, Villa Ballester 2003) 11 c3 Qb6 12 Be3 cxd4 13 cxd4 Rfc8 when the bishop
is just stuck there on g2.
7 ... Nbd7

8 Rg1!
This is probably the best move as White can’t really hold the centre if he develops the dark-squared
bishop. So, for example, if 8 Bf4 Bb4 9 Rg1 (or 9 0-0-0 Qa5!) 9 ... g6! 10 0-0-0 Qa5 and White has to
determine the pawn centre. It’s the same issue after 8 Be3 Bb4 9 0-0-0 Qa5 10 Rg1 g6!.
8 ... Qc7 9 Be3
9 Bd2 leaves the d4-pawn slightly vulnerable after 9 ... 0-0-0 10 0-0-0 Nh5 11 Ne2 g6 12 Kb1 Kb8.
9 ... 0-0-0 10 Qd2
And here if 10 Ne2 dxe4 11 fxe4 Ne5, while 10 0-0-0 runs into some slight trouble after 10 ... dxe4 11
fxe4 Nc5 12 Qe2 Ncxe4 13 Nxe4 Nxe4 14 Qf3 f5 15 Bd3 Nf6 16 Bc4 Nd5 and although White has some
compensation, Black is a pawn up.
10 ... Nh5! 11 Bd3 g6 12 Ne2 Kb8 13 0-0-0 Nb6 14 Kb1 f6
It is difficult to suggest a plan of improvement for White without taking substantial risks.
Giving up the Bishop-Pair: 4 h3 Bxf3
The bishop-pair is White’s big trump card and he has two approaches to try and turn this slight advantage
into an annoying long-term feature. The first is something we can call the positional approach, which aims
to change the nature of the position from closed into open. The easiest way of doing this is by playing
the move d2-d4, such as in the following position.

White is hoping for ... dxe4, opening up the position for that comfortable edge. However, it is not easy
for White to force Black to do this; Black has resources at hand to keep the position closed. Indeed, in the
above diagram Black will play ... Bb4 and force White to close the centre.
The second approach aims to use space to squash Black and deprive the knights of posts. On a chess
website there are some lectures entitled “Every Russian Schoolboy knows ... ”, something some of my
younger students have delighted in listening to. Well, every Russian schoolboy knows that knights need
posts from which they cannot be kicked back from. If Black’s knights are languishing at the back with no
comfortable post to hop to then the position will be very pleasant for White.
Such strategy can be seen in a position like this:

If Black continues normally with ... Be7 then White will simply play Bg2, 0-0, retreat the queen to e2
and gradually expand on the kingside. Please note that Black has no natural plan in such a position.
Alternatively, White can be more direct about grabbing space and pushing the knights back.

An Aggressive Plan: the Kingside Thrust g2-g4-g5

This approach is designed simply to suffocate Black and make play extremely uncomfortable, particularly
if Black doesn’t have a safe place to castle. As we have seen, White can implement this plan in a slow
manner, but more often than not in this variation he plays the g2-g4 thrust directly.

Please note that although this position is from a main line in the Two Knights, we will be avoiding it.

Fighting Aggression with Positional Themes

In playing the aggressive g2-g4, White leaves a square temporarily vulnerable on f4. One of our more
powerful strategies is to gain control over this square. Another aim as Black is to eliminate White’s
advantage of the bishop-pair by swapping off the dark-squared bishops via ... Bb4 followed by ... d5-d4.
Rule: we can tie the move d3 with the bishop sortie ... Bb4.

B) 4 h3

4 ... Bxf3
 The two knights is a quiet variation with some distinguishing features. Black has willingly given up
the bishop-pair and counts on the closed pawn structure to lock the f1-bishop down. Moreover, it is
because of the closed pawn structure that both sides can afford great liberties with timing and
development. Black, in particular, makes some very provocative moves in order to tempt White into
committing himself with the e4-e5 pawn advance.
4 ... Bh5 is highly provocative, but a perfect choice against those people who don’t like to play
aggressively: 5 exd5 cxd5 6 Bb5+ Nc6 7 g4 Bg6 8 Ne5 Rc8 9 d4 e6 10 h4 f6 11 Nxg6 hxg6 leads to a
complicated position.
5 Qxf3
5 gxf3? is possible if highly unlikely, although it was played by Tal against Botvinnik in a world
championship match for psychological effect. Suffice to say, when White voluntarily makes such a
positional concession as this, Black cannot stand worse. The best thing for Black to do is play normally: 5
... d4 6 Ne2 c5 7 d3 Nc6 8 f4 e6 9 Bg2 Nge7 10 Bd2 Ng6 and White lacks the space to orchestrate an
effective attack.
5 ... Nf6

Theory has fluctuated much between 5 ... e6 and 5 ... Nf6. The former is currently more trendy, but I
have stuck with my original choice of the latter. The only real difference between the moves is if White
plays a very quick 6 d4 or exchanges on d5.
White now has several choices between closing the position, aiming to open it and keeping things
flexible:

B1) 6 e5
B2) 6 d4
B3) 6 g3
B4) 6 d3

Black should not be intimidated by the caveman lunge 6 g4 and should simply initiate a set of exchanges:
6 ... Nxe4 (this was one point of Black playing 5 ... Nf6) 7 Nxe4 dxe4 8 Qxe4 Nd7 9 Bg2 Nf6 10 Qc4 e6
11 0-0 Be7 and Black will simply develop in Caro-Kann style, castling kingside, developing the queen to
the c-file, a rook to d8, a knight to d5 and the remaining bishop on f6. Pretty standard really.
B1) 6 e5

I believe this move is simply premature. White quickly determines the pawn structure without even
keeping us guessing. The move falls right in with Black’s plans of keeping the position closed. What
happens next is pretty straightforward for a Caro-player:
6 ... Nfd7! 7 Qg3
Stopping Black’s development seems to be the main try for White:
a) 7 e6? is tempting, but wrong. However, I can guarantee an opponent will fall for this at some stage,
especially in blitz. After 7 ... fxe6 8 d4 e5! 9 dxe5 Nxe5 10 Qg3 (if 10 Qh5+ Nf7!) 10 ... Nbd7 11 f4 Nf7
12 Be3 Qc7 13 Be2 g6 14 Qf2 Bg7 15 0-0 0-0 16 Rad1 Nd6 White had no compensation for the pawn in
H.Van Gool-K.Arkell, Vlissingen 2003.
b) 7 d4 is met by natural play: solidify the d5-pawn and then ... c6-c5. After 7 ... e6:
b1) Should White continue his development with a move like 8 Bf4 (or any bishop move), Black
undermines the centre as quickly as possible, especially since the white knight is awkwardly placed on
c3: 8 ... c5! 9 dxc5 (9 Nb5 Nc6 is pretty nice too) 9 ... Nc6! 10 Qg3 (after 10 0-0-0 Bxc5 11 Qg3 g6
Black has plans of attacking the e5-pawn with ... Bd4 and playing on the c-file; instead, 10 Nb5 is
answered with the simple 10 ... Bxc5) 10 ... g6! 11 Bb5 Bxc5 12 0-0 0-0 13 a3 Rc8 14 Ne2 Qb6 and
Black has an easy position to play with pressure on the queenside.
b2) 8 Qg3 tries to bind the bishop to the defence of the g-pawn and after 8 ... Qb6!? 9 f4 (if 9 Qd3 c5!
10 Be3 Nc6 11 dxc5 Bxc5 and Black already has the upper hand) 9 ... Qxd4 (9 ... g6 is the absolutely safe
way to play the position) 10 Be3 Qb4 11 0-0-0 White has compensation for the pawn. However, he has
nothing immediate and nothing deadly, so Black has good chances for the advantage: 11 ... g6 12 h4 h5 13
Be2 Qa5 and although a draw was agreed here in S.Djuric-A.Groszpeter, Venice 2003, play could have
continued 14 Kb1 Na6 15 Qf2 Bc5 16 g4 hxg4 17 Bxg4 Qb6 18 Bd4 Bxd4 19 Rxd4 Nac5 and Black is
slightly better.
7 ... e6 8 Be2
8 d4 transposes back to variation ‘b2’ above.
8 ... c5!
 Grabbing some central squares.
9 f4
Note that after 9 0-0 Nc6 10 Bb5 Nd4 11 d3 Nxc2 12 Bg5 we should go 12 ... Qc8!, as 12 ... Qc7?
would be a horrific blunder in view of 13 Nxd5 exd5 14 e6.
9 ... Nc6 10 Nb5
10 d3 does not look too attractive either: 10 ... Nd4 11 Bd1 Nf5 with a pleasant closed position.
Likewise, 10 f5 is not scary after 10 ... exf5 11 Nxd5 Ndxe5 and Black has no difficulties, as shown by 12
c4 Qd6 13 d3 g6 14 Bf4 Bg7.
10 ... g6 11 c3 a6 12 Na3 c4 13 Nc2 Nc5 14 d3 cxd3 15 Bxd3 Nxd3+ 16 Qxd3 Bc5
Black already had a fine position in E.Rozentalis-E.Moldobaev, Calvia Olympiad 2004.

B2) 6 d4

White insists on playing the very move our fifth move was designed to deter.
6 ... dxe4 7 Qe3
White wants to preserve the d-pawn.
After 7 Nxe4 to make sense of his play, Black is obliged to grab the d4-pawn. This pawn sacrifice is
not too well regarded by theory, but I believe it is dangerous. My solution is based on exchanging a set of
knights: 7 ... Qxd4 8 Bd3 (we must also consider the forcing 8 Nxf6+ exf6 9 Be2 Bd6 10 0-0 0-0 11 Rd1
Qe5 12 g3 Qe7, which is fine for Black) 8 ... Nxe4 (the simplest; 8 ... Nbd7!? was my original but more
complicated solution) 9 Bxe4 Nd7 10 0-0 Qf6 11 Qg3 (11 Qb3? falls foul of a fork: 11 ... Nc5) 11 ... e5
(preventing White’s queen from raiding our queenside) 12 Bg5 (if 12 f4 Bd6!) 12 ... Qe6 13 Rad1 (13 f4
f5! keeps the position blocked) 13 ... g6.

At last we will develop our bishop:

a) 14 f4 f5 15 Bd3 e4 16 Be2 h6! 17 Qc3 (or 17 Bh4 Nb6 18 Qc3 Rh7 and ... Be7 is coming next with
the advantage) 17 ... hxg5 18 Qxh8 0-0-0 19 Qc3 gxf4! with very strong counterplay.
b) 14 Qc3 Be7! (the best move; after 14 ... Bg7 15 Qb4 and White is a lot better) 15 Bh6 0-0-0 is
about equal.
7 ... e6 8 Nxe4 Nxe4 9 Qxe4 Nd7 10 c3
10 g3 doesn’t quite work because of 10 ... Nf6 11 Qd3 Qd5 and White will have to abandon plans of
castling kingside: 12 Rg1 0-0-0 13 Be3 Qa5+ 14 c3 Bd6 15 Bg2 Rhe8 16 Qc4 Nd5 17 Bf3 Nxe3 18 fxe3
and Black had a satisfactory position in A.Muzychuk-V.Gunina, Beijing (Basque) 2013.
10 ... Nf6 11 Qf3 Be7
Now that there have been some exchanges, Black must be very careful not to let White improve his
position. White will centralize his rooks on the d- and e-files and push for the break d4-d5. Black has to
counter this idea by playing an effective ... c6-c5, and by effective I mean a position where Black has
control over the d-line.
 12 Bd3 0-0 13 0-0
After 13 Bf4 Nd5 14 Be5 Bd6 15 0-0 Bxe5 16 dxe5 Qb6 Black has successfully neutralized the
bishop-pair.
13 ... Qb6 14 Bg5 Rfd8 15 Rfe1?!
Instead, 15 Rad1 Qa5 16 Bf4 Qxa2 grabs a pawn, while 15 Rfd1 Nd5! 16 Bd2 a5! is an absolute
must.
15 ... Rd5!?
As played in M.Ghinda-W.Watson,Thessaloniki Olympiad 1988. This is a very interesting idea,
although Black must be on guard for some concealed dangers. The position looks reasonably quiet, but
White has tricks involving sacrificing on e6 and also some potential back-rank mating ideas. Instead, 15
... h6?! would be dubious: 16 Bf4 Rd7 17 b4 Nd5 18 Bd2 a5 19 a3 and after defending everything on the
queenside, White now has a real target to attack on the kingside.
16 Bh4

16 ... Rad8?!
16 ... a5 17 b3 Rad8 was more accurate: 18 Rad1 R5d7 19 Re5 h6 and ... c5 is coming.
17 Bc4 R5d7 18 Re2! Nd5
 Not 18 ... c5?! 19 Rae1! (19 dxc5?, as played in the game, hands Black the open d-file and the
initiative on a plate: 19 ... Qxc5 20 Bb3 and Black now had the stunning 20 ... g5! 21 Bg3 Rd3 22 Re3
Rxe3 23 fxe3 Rd2) 19 ... Qc7 20 Bxf6 Bxf6 21 d5! e5 and White is much better.
19 Bxe7
Now if 19 Bg3 c5! as Black does not have to worry about threats on e6 and the back rank.
19 ... Rxe7 20 Rae1 Red7 21 b4
Preventing the ... c6-c5 break.
21 ... a5 22 a3 axb4 23 axb4 Ra8
As it is difficult for White to dislodge the very strong knight from d5, Black has no problems.

B3) 6 g3

White retains some flexibility in his pawn structure by delaying moving the d-pawn, so he can choose
d2-d3 or d2-d4 according to Black’s response.
6 ... e6 7 Bg2 Nbd7
My original recommendation, although I also really like the move 7 ... Bb4.
8 d4
With 8 d3 White adopts a King’s Indian Attack set-up, but this is not dangerous: 8 ... Bb4! (this move
becomes a lot more attractive once White has played d2-d3) 9 0-0 0-0 10 Nd1 (if 10 Qe2 Bxc3 11 bxc3
dxe4 12 dxe4 Qa5 13 Rd1 Rfd8 and Black has plenty of squares for his knights to hope about in, while 10
Bd2 is well met by 10 ... d4! 11 Nb1 Qb6!) 10 ... Re8 11 exd5 (11 Ne3 Qb6 12 exd5 cxd5 13 c3 Bd6 14
d4 Rac8 is equal) 11 ... exd5 and keeping the position symmetrical makes it harder for White to use the
bishop-pair.
8 ... Bb4
This is not without risk, but is critical.
9 e5 Ne4
 These closed positions are actually very tricky for White to handle. Firstly, Black’s position is very
solid; so solid that we can lose time moving our minor pieces backwards and forwards. Secondly, it is
difficult for White to make use of the light-squared bishop on g2 and the knight on c3.
10 0-0 Nxc3!? 11 bxc3 Be7 12 Qg4
Probably the best move. 12 Qd3 0-0 13 f4 was preferred in E.Berg-A.Galkin, Moscow 2004, and
after 13 ... Nb6! 14 f5 (let’s also try the pawn storm: 14 g4 Nc4 15 f5 Bh4! and I don’t see a way for
White to crash through) 14 ... exf5 15 Qxf5 f6! 16 e6 (if 16 exf6 Rxf6!) 16 ... Bd6 17 Bf4 White is getting
nowhere.
12 ... g6
Here we have a very closed position with one open b-line.
13 f4?!
Positionally suspect, although this makes it tricky for Black to castle kingside. We should now plan to
blockade the important a-pawn, unusually enough with the queen ( ... Qa5), and seek to play ... Nb6-c4,
controlling the queenside.
Instead, after 13 Bh6 Bf8 14 Bd2 h5 15 Qf3 Be7 (Black should focus on controlling the queenside and
keeping the position as solid as he can) 16 a4 Nb6 17 Rfb1 Nc4 18 Bc1 b6 19 Bf1 Rc8 Black’s position
is very solid.
13 ... Qa5!
A necessary move to prevent White from advancing the a-pawn.
14 Qf3 Nb6
 15 g4 h5 16 f5 hxg4 17 hxg4 Kd7?!
Instead, 17 ... gxf5! 18 gxf5 0-0-0 gives Black a clear advantage.
18 Rf2 gxf5 19 gxf5 Rag8 20 Qd3 Nc4
With an advantageous position for Black, who indeed won easily in B.Rzayev-D.Svetushkin,
Roquetas de Mar 2012.

B4) 6 d3

By far the most popular continuation. The big idea behind White’s move order, apart from supporting
the e4-pawn, is to set the scene for a fight in a complicated middlegame structure.
6 ... e6
My general recipe here is to develop the bishop to b4 (this is essential once White has pushed the d-
pawn), in order to exchange the dark-squared bishop and leave White with a poor bishop sitting on g2.
Another common move in association with ... Bb4 is to play ... Qb6 or, more rarely, ... Qa5.
White’s main moves are:

B41) 7 g4
B42) 7 a3
B43) 7 Be2
B44) 7 Bd2

Let’s deal with the sidelines first where we will see how powerful Black’s ideas are:
a) 7 g3!? Bb4! (the most aggressive continuation) 8 Bd2 d4 9 Nb1 and now Black gains time by
playing 9 ... Qb6!.

After 10 b3 Bxd2+ 11 Nxd2 Qc5 (hitting the c-pawn) 12 Qd1 (if 12 Rc1 White is more or less
powerless to prevent Black from bringing his knight to b4) 12 ... Nbd7 13 Bg2 h5!? 14 0-0?! (or 14 f4 e5
15 f5 0-0-0 and Black has the straightforward plan to open up lines on the kingside) 14 ... h4 15 b4 Qe7
16 g4 e5! Black secures the f4-square for his knight. Moreover, White’s bishop looks absolutely terrible
on g2 and Black is already much better.
b) 7 Qg3 is a rare move, but prevents the bishop developing to b4: 7 ... Qb6!? (Black copies White’s
example by making it difficult for White to develop his queenside) 8 Be2 (8 e5 gives Black a good French
with no bad bishop - regardless of what the computer may say) 8 ... Nbd7 9 0-0 g6 10 Rb1 Bg7 11 e5 (or
11 Bf4 e5! 12 Bd2 d4 13 Nd1 Qa6, which has a tactical point: Black is threatening ... Nxe4, as well as
a2) 11 ... Nh5 12 Bxh5 Bxe5! 13 Bf4 Bxf4 14 Qxf4 gxh5 and Black has no problems; he is a pawn up and
has a ready-made attack down the g-file.

B41) 7 g4?!
I really don’t like this move. White creates a gaping hole on f4 and once the dark-squared bishops are
removed, White is left with a dud of bishop on g2 locked in by its own pawns. This is exactly what Black
is aiming for: good knight versus bad bishop.
7 ... Qa5
With the idea of putting a bishop on b4. The immediate 7 ... Bb4 is also possible.
8 Bd2 Bb4 9 Bg2
Alternatively:
a) 9 Qd1 d4 10 Ne2 Bxd2+ (swapping the bishops is, of course, a good idea) 11 Qxd2 Qb6! (Black is
not quite ready for the endgame: 11 ... Qxd2+ 12 Kxd2 as after 12 ... e5 White can play 13 f4 when he
stands a bit better) 12 0-0-0 c5! 13 f4 Nc6 14 Bg2 0-0-0. Here the kingside initiative may initially look
promising, but Black has good counter-chances on the queenside: for instance, 15 g5 Ne8 16 Kb1 Nd6 17
h4 c4 and suddenly the position is tricky for White.
b) 9 a3 is met with 9 ... d4 10 Nb1 Bxd2+ 11 Nxd2 Nbd7 12 0-0-0 h5! 13 Qf4 (if 13 Be2 0-0-0 14
Rhg1 hxg4 15 hxg4 Ne5 16 Qg3 g5!; controlling the f4-square is an important strategy for Black) 13 ... g5
14 Qd6 Qe5! (Black has control over the f4-square, so any endgame will be good for him) 15 Qb4 0-0-0
16 Be2 c5 17 Qa4 Kb8 18 Nc4 Qf4+ 19 Kb1 hxg4 with advantage to Black.
9 ... h5!

Striking whilst the white king is awkwardly placed in the centre.

10 gxh5
Hardly great, but neither are:
a) 10 g5 does nothing for White: 10 ... d4 11 gxf6 dxc3 12 bxc3 Bxc3 13 Qe3 gxf6 14 d4 Bxd2+ 15
Qxd2 and Black is material ahead, which gives him good winning chances in the endgame.
b) 10 Qe2 d4 11 Nb1 Bxd2+ 12 Nxd2 hxg4 13 hxg4 Rxh1+ 14 Bxh1 Nbd7 15 f4 g5 obtains a nice grip
on the f4-square.
c) 10 Qe3 is perhaps the best option for White as it leads to a complicated game, although after 10 ...
dxe4 11 dxe4 hxg4 12 hxg4 Rxh1+ 13 Bxh1 Nbd7 14 0-0-0 0-0-0 15 a3 Nxg4 16 Qg3? Nde5 17 axb4
Qa1+ 18 Nb1 Nc4 19 Qc3 Nxf2 20 Bf3 Nxd1 21 Qxc4 Qxb2+ 22 Kxd1 Qxb1+ Black has the advantage.
10 ... d4 11 Nb1 Bxd2+ 12 Nxd2 Nbd7 13 Qg3 Nxh5 14 Qd6
 Here Black played 14 ... Qc5 in A.Khamatgaleev-D.Svetushkin, Patras 2002, but he had something
even stronger:
14 ... e5!
Black has a wonderful position as 15 b4 doesn’t work in view of 15 ... Qd8 16 b5 (16 Nc4 Rh6 wins
the queen) 16 ... cxb5.

B42) 7 a3

The idea behind this move is to prevent the irksome ... Bb4 and make the move ... d5-d4 as difficult as
possible. Next White has a choice between two plans: the aggressive g2-g4 and the more positional g2-
g3.
Be wary of determining the structure with ..d5-d4. This is usually played to gain something: for
example, we previously saw it being used as a mechanism to swap the bishops. Here, however, by
determining the pawn structure Black makes White’s plan of a kingside pawn advance more
straightforward.
7 ... Nbd7
Now we have:
B421) 8 g4
B422) 8 g3

Instead, 8 Be2 is clumsy and gets in the way of the queen. The way to deal with this move is instructive: 8
... Bd6 9 0-0 Qb6 (pinning the f2-pawn and tying the bishop to c1) 10 g3 Be5! 11 Qg2 0-0 12 Kh1 Rad8
13 f4 Bxc3! (this is the point of ... Be5, to ruin the white structure in anticipation of f2-f4) 14 bxc3 dxe4
15 dxe4 Nc5 16 Bf3 Qb5 17 Be3 Qc4 18 Bd4 Nfxe4 19 Bxe4 Nxe4 20 Qxe4 c5 and Black has the
advantage.

B421) 8 g4

Traditionally perceived as the most principled reply to 7 ... Nbd7 as Black can no longer retreat the
knight to d7. Now Black’s key strategy is to use the f4-square to try and swap a pair of bishops.
8 ... h6!?
Preventing the pawn advance g4-g5.
Instead, 8 ... g6 is the main continuation, while 8 ... Bc5!? was my old recommendation. Here 9 g5! is
critical, especially to an unsuspecting customer: 9 ... Ne5 10 Qg3 Nfd7 and now 11 h4! maximizes the
impact of f2-f4 and looks to be somewhat easier to play as White.
With apologies, we now have a further divide:

B4211) 9 Bg2
B4212) 9 Qg3

Instead, 9 h4 is a very interesting pawn sacrifice, played to displace the black knights. Now 9 ... Qb6 is
the safe route to equality.
 a) White can easily go wrong if he is not incisive: 10 Bg2 d4! (the pawn push is now possible
because White has played g2-g4, leaving the f4-square vulnerable) 11 Ne2 (if 11 Nd1 Bd6!) 11 ... Bd6 12
g5 (or 12 Qh3 e5!) 12 ... Ne5 13 Qg3 hxg5 14 hxg5 Nh5 15 Qh4 followed by 15 ... 0-0-0 (15 ... Ke7 is
also very strong) .
b) 10 g5 hxg5 11 hxg5 Ne5 12 Qg2 Rxh1 13 Qxh1 Nfd7! 14 f4 Ng6 15 Qf3 0-0-0 16 Ne2 Bc5 and
now the tempting 17 b4? is a big mistake after 17 ... dxe4 18 dxe4? Nde5! 19 fxe5 Nxe5 20 Qg2 Bf2+ 21
Qxf2 Rd1+ 22 Kxd1 Qxf2 23 Bh3 Qf3 and the white bishop is lost.

B4211) 9 Bg2
The most natural move.
9 ... Bd6 10 Qe2
After 10 Be3 I like 10 ... Qa5, trying to prevent White from castling queenside: for example, 11 0-0 0-
0 12 Qe2 Qc7 13 f4 d4 14 Bxd4 Bxf4 when Black should stand well.
10 ... d4
A good decision: because White has played g2-g4, there is a gaping hole on f4.
10 ... dxe4 also worked out well for Black: 11 Nxe4 Nxe4 12 Qxe4 Nf6 13 Qa4 h5 14 g5 Nd5 15 Bd2
g6 16 0-0-0 0-0 17 f4 Qc7 18 Rhf1 Rfd8 19 h4 Bf8 ½-½, I.Khamrakulov-K.Arkell, European Club Cup,
Kemer 2007.
11 Nb1
Likewise, after 11 Nd1 Qc7 12 c3 c5 13 b4 0-0 14 0-0 Bh2+ 15 Kh1 Bf4 16 bxc5 Bxc1 17 Rxc1
Qxc5 18 f4 Rac8 19 c4 e5 20 f5 Qxa3 21 Nf2 Qe7 22 h4 Nh7 Black should have no problems.
11 ... Qc7
 Please note Black’s control over f4, one of the key factors in playing ... d5-d4.
12 c3 c5 13 Nd2 Bf4 14 Nf3 Bxc1 15 Rxc1 Qb6 16 0-0 e5 17 b4
If 17 Nh4 0-0 18 Nf5 Rfe8 19 h4 Nf8 20 g5 hxg5 21 hxg5 N6h7 22 f4 exf4 23 Rxf4 Ng6 24 Rg4 Nhf8
25 Rf1 Rad8 26 c4 Qc7 27 Qf2 Ne6 28 and White is getting nowhere.
17 ... dxc3 18 Rxc3 cxb4 19 axb4 0-0 20 Rc4 Ne8 21 Qb2 Nd6 22 Rc2 Rfe8 23 Rd1 Rad8 24 Qa1
a6 25 h4 Nb5
Black had a big advantage in I.Paulet-R.Ignacz, Pamplona 2007.

B4212) 9 Qg3

Preventing our plan of playing ... Bd6 and taking control over the f4-square.
9 ... Qb6!?
One of those annoying Caro moves. Now White must think twice about developing the bishop on c1.
10 Bg2
10 e5 is rather brutally met with the counterattack 10 ... d4 11 exf6 dxc3 and now we have two
different captures:
a) 12 bxc3 0-0-0! 13 Bg2 gxf6 14 0-0 when Black should get his counterattack going with 14 ... h5!:
 a1) After 15 Be3 Black can play the daring 15 ... Qb2 16 Rfb1 Qxc3 17 Ra2 (or 17 Bxa7 hxg4 18
hxg4 Rg8 19 Ra2 Ne5 and Black is much better) 17 ... Bc5 18 Rab2 Nb6 19 Rb3 Qxc2. As I like to say,
fortune favours the brave.
a2) 15 g5 Qa5! is very effective and now the opening of the g-file would be a bit disastrous for White:
for instance, 16 gxf6? Qxc3 17 Bf4 when the dark squares are neutralized by 17 ... Bh6! and White will
suffer along that open g-line.
b) 12 fxg7 Bxg7 13 b3 0-0-0 14 Bg2 Be5 15 Qf3 f5 and Black can be happy as White has no safe
place for his king.
10 ... 0-0-0 11 e5
Alternatively, 11 0-0 dxe4 12 dxe4 Qc7 13 Bf4 e5 14 Be3 Bc5 15 b4 Bxe3 16 Qxe3 Kb8 with a
dynamic position for us, whereas White has no obvious entry point and the bishop on g2 is just useless.
11 ... Nh7 12 0-0 g5 13 Ne2 Qc7 14 d4 f6 15 exf6 Bd6 16 Qe3 Nhxf6

17 f4
Of course, 17 Qxe6? drops a piece to 17 ... Rhe8.
17 ... Rde8 18 fxg5 hxg5 19 Qxg5?!
We’ve followed R.Reinaldo Castineira-J.Gomez Esteban, Mondariz 1999. Instead of the game’s 19 ...
Bh2+ Black can play:
19 ... e5!
Black seizes the initiative and is doing pretty well.

B422) 8 g3

This is probably the most accurate plan for White. He intends to develop his pieces and only then
seize the necessary space.
8 ... g6
Now that the bishop has been deprived of the b4-square, Black changes tack to put the bishop on the
long diagonal. I must stress, dear reader, that you will not make the same mistake as I once did where I
simply developed the bishop to e7. The problem then is that once Black has completed his development
he simply has no plan.
The formula here: if White prevents ... Bb4 and fianchettoes, we copy.
9 Bg2 Bg7 10 0-0 0-0

11 Qe2
Otherwise:
 a) After 11 Be3 White finds himself in a bit of trouble if Black plays 11 ... Ne5 12 Qe2? (better is 12
Qd1 d4 13 Bf4 Nh5 14 Bxe5 Bxe5 15 Ne2) 12 ... d4 13 Bf4 (not 13 f4 dxc3 14 fxe5 Nh5! 15 Bf4 cxb2 16
Rab1 Qd4+ and Black is winning) 13 ... Nh5 14 Bxe5 Bxe5 15 Nb1 when Black has no problems. The
plan is to concentrate our efforts on the queenside or to win a rook and two pawns for two pieces. Indeed,
15 ... Bxg3!? 16 fxg3 Nxg3 17 Qf2 Nxf1 18 Qxf1 should be slightly better for Black.
b) 11 d4 Nb6 is quite a handy little move, putting pressure on the d4-pawn. Once again, the closing of
the centre with e4-e5 will favour Black. If, instead, White chooses to keep the tension with 12 Bg5 Black
is forced to respond with 12 ... h6 when White has to give up the bishop-pair because of the pressure on
d4 - if 13 Be3 Nc4!. Instead, if 12 Rd1 Nxe4 13 Nxe4 dxe4 14 Qxe4 Black can counterattack in a rather
unusual way: 14 ... e5 15 Be3 f5 16 Qd3 f4! 17 Qb3+ Nd5!! (an excellent move; Black renews his threat
of capturing the bishop) 18 Bd2 fxg3 19 fxg3 exd4 20 Qxb7 Qd6 and White must tread carefully.
11 ... Qb6!?
Creating a concealed threat.
12 Kh2
The hope is, for example, 12 Rb1? Nh5!.

This attacks both c3 and g3, and after 13 Na4 Nxg3! (not 13 ... Qd4 14 b4! as now 14 ... Nxg3? is not
possible because of 15 Qe1 Nxf1 16 Bb2! when the black queen is trapped in the centre of the board) 14
Qe1 (after 14 Qg4 Qa6 15 exd5? Nxf1 Black soon won in B.Shovunov-A.Lastin, St Petersburg 1997) 14
... Qa6 15 fxg3 Qxa4 Black has won a pawn.
12 ... Ne8!?
Rerouting the knight to the central square d6, which also conveniently also defends the f5-square.
13 Bd2 Nd6 14 f4 Rfd8 15 g4
Another possibility is 15 e5 Nf5 16 g4 Nd4 17 Be3 c5. In these types of positions, he who has the
most space wins we might say. Black has a fine knight on d4 and plans is to steamroller his a- and b-
pawns down the board to victory.
15 ... Nb5
Alternatively, 15 ... Bxc3!? 16 Bxc3 Nb5 17 Bd2 Nd4 18 Qd1 Qxb2 19 Rb1 Qxa3 20 Rxb7 dxe4 21
dxe4 Nc5 and Black has the advantage.
16 Nxb5 Qxb5 17 f5
After 17 Rab1 dxe4 18 Bxe4 Nc5 19 Bf3 Na4 20 Qe1 Qc5 21 Qc1 Rd7 22 Be3 Bd4 White’s attack
has not been at all successful.
 17 ... exf5 18 gxf5 dxe4 19 Bxe4 Qxb2 20 fxg6 hxg6
Black’s position is secure as after 21 Rxf7 he can simply play 21 ... Qxa1.
21 Rab1 Qxc2 22 Rfc1 Qa2
Black is safe and doing well.

B43) 7 Be2

7 ... Nbd7
Black has played a variety of moves here, but the text is still holding up well. Note that we are
delaying the formula for a move as we want White to reveal his plans.
8 Qg3
Alternatives are:
a) 8 g4 Bb4 9 0-0 h6 10 a3 Bxc3 11 bxc3 dxe4 12 dxe4 0-0 13 Bd3 Qa5! prevents g4-g5 and White’s
ruined pawn structure gives Black full compensation for the two bishops.
b) 8 Bd2 Qb6 (8 ... d4 is a double-edged sword; it seizes space, but White now has a straightforward
plan of rolling on the kingside) 9 0-0 Bd6 (in this position it is difficult for White to come up with a plan,
partly as his light-squared bishop is not well placed on e2; meanwhile, our eventual plan is to push for ...
d5-d4 and then open the c-file with ... c5-c4) 10 Rfe1 0-0 11 Rab1 d4 12 Nd1 c5 13 b3 c4! 14 dxc4 Ne5
is a nice trick.
c) 8 0-0 Bd6 9 d4 dxe4 10 Nxe4 Nxe4 11 Qxe4 0-0 12 c3 Nf6 13 Qd3 h6 and Black will play to
exchange the dark squared bishops and use the half-open the d-file to his advantage.
8 ... g6
This was my original recommendation, although I also quite like 8 ... Qb8!?, controlling the e5-
square: for example, 9 Bf4 e5 10 Bd2 (or 10 Bg5 d4 11 Nb1 Be7 12 0-0 Nc5 13 f4 Ne6! and Black has
neutralized the threat of f2-f4) 10 ... d4 (10 ... dxe4 11 Nxe4! is better for White) 11 Nb1 Nc5 12 0-0 Ne6
and Black has a firm grip on the f4-square.
9 0-0 Bg7
 10 Bf4
After 10 f4 0-0 I don’t like 11 e5 for White. Having played the white side of the Closed Sicilian for
many years, I have never managed to make these kingside attacks work, whereas Black has nice play on
the queenside. However, 11 f5 d4 12 Nd1 exf5 13 exf5 Qe7 14 Bg4 Nd5 also looks good for Black, while
after 11 Be3 d4 12 Bxd4 Nxe4 13 dxe4 Bxd4+ 14 Kh1 e5 15 f5 b5 the white bishop will not be allowed
to point at the potentially weak f7-square.
Instead, 10 Rd1 Qb6 11 e5 Nh5 12 Bxh5 Bxe5 13 Qh4 gxh5 is satisfactory for Black, who has a safe
space for his king on the queenside.
10 ... Qb6!
10 ... 0-0 is bad as it allows White to clamp down immediately with 11 Bd6! Re8 12 e5.
11 Rab1 e5!?
I feel this is a little better than the customary 11 ... 0-0 primarily because it cuts out the option of 12
Bc7.
12 Bg5
Alternatively:
a) 12 Bd2 d4 13 Nd1 Qa6 is a crafty little move that sets a trap (14 ... Nxe4) and 14 Re1 Qxa2 15
Rc1 0-0 16 f4 Bh6 leaves Black a clear pawn up.
b) 12 Bxe5? loses immediately to 12 ... Nh5! 13 Qh4 Bxe5.
12 ... d4 13 Nd1
After 13 Na4 Qb5 14 b3 Nxe4! Black is much better, as if 15 dxe4 Qxe2 16 Rfe1 Qa6.
13 ... Qa6!
 A crafty little move creating two threats: 14 ... Nxe4 and 14 ... Qxa2.
14 Bxf6
14 a3 Nxe4! wins at least a pawn, while after 14 Re1 Qxa2 15 Rc1 0-0 Black is much better.
14 ... Bxf6 15 f4 Qxa2 16 Nf2 h5
Preventing the knight from jumping to g4 where it would be highly annoying.
17 Ra1 Qxb2 18 Rfb1 Qxc2 19 Bd1 Qd2 20 Rxb7 exf4 21 Qf3 Qe1+ 22 Kh2 Bh4 23 Nh1 g5
Black has the deadly threat of 24 ... Ne5!.

B44) 7 Bd2

Played with a direct idea in mind after 7 ... Bb4.

7 ... Qb6!?
At first my recommendation was going to be the standard 7 ... Nbd7 8 g4 and then the slightly unusual
8 ... Bb4 (8 ... b5 was the choice in my first book, but after 9 g5! b4 10 Nd1!? Ng8 11 Ne3 h6 12 g6 Qf6
13 gxf7+ Kxf7 14 Bg2 Black has a worse type of equality). However, 9 g5 Ng8 10 d4 bothered me. My
main recommendation is based on the same ... Bb4 idea, but encouraging White to refrain from the idea of
pushing the g-pawn.
 I must also mention that a good short-cut against this main line is to copy Caro-Kann expert Keith
Arkell and play 7 ... Bb4. Yes, Black loses some time and falls for White’s ‘trick’, but the position is
closed and one can get away with a variety of sins. Disclaimer: one must, though, like closed positions.
The main problem with Black’s move order is considered to be 8 e5 Nfd7 9 Qg4 Bf8 as his pieces are
pushed back. However, White has no clear-cut way to obtain an advantage: for example, 10 f4 h5 11 Qf3
h4 12 0-0-0 g6 13 g4 hxg3 14 h4 Rxh4 15 Rxh4 Qxh4 16 Bg2 Qh8 17 Rh1 Qg7 18 Qxg3 Na6 and White
failed to break through in S.Ansell-K.Arkell, London 2011.
Against my main recommendation, White has two main ideas:

B441) 8 g4!?
B442) 8 0-0-0

Minor alternatives are:

a) 8 Be2 Nbd7 9 0-0-0 Bd6 10 g4 d4 11 Nb1 Ne5 12 Qg2 Ng6 13 g5 Nd7 14 Qg4 Qc7 and with
control of the f4-square secured for a knight, Black’s chances are clearly better.
b) 8 Rb1 gives Black a moral victory; not being able to castle queenside could potentially cause
White issues with overextending on the kingside After 8 ... Bd6 9 g4 d4 10 Ne2 c5 11 Bg2 Nc6! 12 0-0 0-
0 13 Ng3 Qc7 Black has control over the all-important f4-square, which means White can’t retreat the
queen and begin a dangerous pawn storm.
c) 8 Qg3 is nothing to fear either: 8 ... d4! 9 Nd1 h5 (with the aim of playing ... h5-h4 to chase the
queen away) 10 Be2 (or 10 h4 c5 11 Be2 Nc6 12 0-0 g6 13 f4 0-0-0 14 Kh1 Kb8 15 f5+ Bd6 16 Bf4
Bxf4 17 Rxf4 e5 18 Rf1 Rdg8 and Black has the advantage) 10 ... Nbd7 11 0-0 h4 12 Qh2 (if 12 Qg5 g6
with ... Be7 soon to follow) 12 ... Nh5 13 Bxh5 Rxh5 14 f4 Bb4 and Black is safe to castle queenside.
d) 8 Na4 Qc7 9 Nc3 Nbd7 10 g4 Bb4 11 g5 Ne5 reveals one advantage of the queen gaining the free
move ... Qc7 and after 12 Qg2 Nfd7 13 f4 Ng6 14 Qf2 Bxc3 15 bxc3 dxe4 16 dxe4 e5 17 f5 Nf4 White
has overextended.

B441) 8 g4 d4!
 The knight on b8 has a more attractive destination, namely c6. Note too that 8 ... Qxb2 9 Rb1 Qxc2 10
e5 is simply way too risky for Black.
9 Nd1
9 Ne2 runs into trouble already after 9 ... Qxb2.
9 ... c5! 10 g5
After 10 Qg3 Nc6 11 f4?! Nb4 12 Bxb4 cxb4 13 Bg2 Be7 14 0-0 Rc8 the c2-pawn will give White a
headache for a long time, whilst the open c-line and the exchange of two sets of minor pieces mean that
the kingside pawn storm is not especially scary.
10 ... Nfd7 11 h4 Nc6 12 a3 Bd6 13 h5 0-0-0!

14 Bh3
As played in J.Sikora Lerch-K.Grycel, Rymarov 1999. Instead, 14 Qxf7?? Nce5 15 Qxe6 Nf3+ 16
Ke2 Nxd2 17 Kxd2 Bf4+ would be an embarrassing turn of events and even 14 h6 gxh6 15 Rxh6 Nde5 16
Qh5 Ng6 looks rather promising for Black.
14 ... c4! 15 dxc4 Nde5 16 Qb3 d3! 17 Qxb6 Nf3+ 18 Kf1 Nxd2+
Netting a bishop.
B442) 8 0-0-0

Defending the pawn in this way is the most natural response, but as White has not managed to get in
g2-g4 he may be on unfamiliar territory.
8 ... Bb4!
Linking in the idea of swapping the dark-squared bishops.
9 Qg3
Attacking the g7-pawn, but White no longer can push his pawns.
After 9 a3 Black can calmly retreat the bishop with 9 ... Bd6 when 10 g4 leads to:
a) 10 ... Nbd7 11 g5 Ng8 12 h4 Ne7 13 h5 d4 14 Ne2 c5 15 Qh3 Rc8 16 Ng3 is preferable for White,
although a tough struggle still lies ahead, G.Papadopoulos-D.Papakonstantinou, Achaia 2013.
b) I prefer 10 ... d4 11 Ne2 c5 12 g5 (or 12 Qg2 Nc6 13 f4 Nd7 14 e5 Be7 when Black will castle
queenside and prepare the pawn advance ... c5-c4) 12 ... Nfd7 13 h4 as now Black has the strike 13 ... c4!
14 dxc4 Ne5 15 c5! (15 Qb3 Qxb3 16 cxb3 Nd3+ 17 Kb1 Nxf2 wins material) 15 ... Bxc5 16 Qg3 Nbc6
17 Nf4 Rc8 18 Nh5?. Here Black can play the spectacular 18 ... 0-0 (the easiest move is 18 ... d3!?) 19
Nf6+ Kh8! and now moves such as 20 f4 are refuted by 20 ... d3 21 fxe5 Bxa3!! (a crushing blow) 22
bxa3 dxc2 23 Bb4 Nxb4 24 axb4 Qxb4 which is very nice.
9 ... 0-0!?
 10 e5
Closing the position is very natural.
After 10 a3 Black is now forced to play 10 ... Bxc3 11 Bxc3 d4 12 Bd2 Nbd7 13 e5 Nd5!, sacrificing
the exchange for the initiative: 14 Bh6 g6 15 Bxf8 Rxf8 16 h4 (ignoring any potential problems on the
queenside, White begins counter measures; after 16 Be2 Nc5 the move ... Na4 will give White a
tremendous headache with his queen cut off from the action) 16 ... Nc5 17 h5 Na4! 18 b4 c5 19 Qh4 and
now Black can play the marvellous 19 ... g5!! 20 Qxg5+ Kh8 21 Rh3 Rg8 22 Qh6 cxb4 23 Rf3 Qc7 24
axb4 Nac3 25 Kd2 Nxd1 26 Kxd1 b6 with the positional threat of ... a7-a5.
10 ... Ne8 11 f4
I think the best move is probably 11 a3 Bxc3 12 Bxc3. However, Black still obtains an initiative by
breaking down the white king with some nifty piece manoeuvres: 12 ... d4 13 Bd2 Na6 14 Kb1 Nc5 15
Ka1 Na4 16 Rb1 a5 17 Be2 Qc5 18 Rhc1 Nc7 (the knight is heading for b5) 19 Bh6 g6 20 h4 (20 Bxf8
Rxf8 leaves White helpless against ... Nb5 and ... Nxa3) 20 ... Nd5! (20 ... Nb5 only leads to a draw after
21 Qf4 Nxb2 22 Rxb2 Qxa3+ 23 Ra2 Qc3+ 24 Kb1 Qb4+) 21 Qf3 (or 21 Bf3 Ndc3) 21 ... Ra6 22 Bxf8
Kxf8 23 h5 Rb6 with the threat of terrible ... Rxb2 in the air and Black is winning.
11 ... Nd7 12 Be2 Bxc3 13 Bxc3 d4 14 Bd2
White’s two bishops are not working particularly efficiently.
14 ... Nc5! 15 b3 a5 16 Rdf1
Likewise, if 16 a4 Nxb3+ 17 cxb3 Qxb3 18 Qf2 c5 19 Bf3 Ra6 and Black’s position is devastating.
16 ... a4 17 b4 Nb3+!!
 A wonderful tactic. Once the queens and rooks come alive, White has no chance.
18 cxb3 axb3 19 axb3 Qa6 20 Kd1 Qa1+ 21 Bc1
Here in Fang Yan-Yu Ruiyuan, Chinese Team Championship 2013, Black sadly missed:
21 ... Ra2! 22 Rf2
22 Bf3 Nc7 23 Rf2 Nb5 24 Rxa2 Nc3+ 25 Ke1 Qxc1+ 26 Kf2 Qe3+ 27 Kf1 Nxa2 wins.
22 ... Nc7 23 Bg4 Nd5 24 Rxa2 Nc3+ 25 Ke1 Qxc1+ 26 Kf2 Qxh1
White will not manage to survive.

Summary
The Two Knights is a flexible line primarily because White does not determine the pawn structure. With 3
... Bg4 we signal we are prepared to swap bishop for knight. 4 d4 is an interesting idea, but I feel that
after 4 ... Nf6 5 h3 Bxf3 6 gxf3 Black can develop easily and make use of the weakened dark squares.
In the main line, if White combines a kingside fianchetto with d2-d4, Black needs to pressure the
centre with ... Bb4, forcing White to do something over his e-pawn. Another important idea is to meet g2-
g4 by controlling the f4-square with ... Bd6 and ... Qc7.
6 d3 e6 7 Bd2 is the main line and here I am very excited about 7 ... Qb6!?, aiming to deter White from
going g2-g4.
Chapter Fifteen
The King’s Indian Attack
1 e4 c6 2 d3 d5 3 Nd2 e5 4 Ngf3 Bd6

The King’s Indian Attack (or KIA) is a solid and somewhat formulaic system. Its practitioners simply
aim to reach a pawn structure they are comfortable in. White’s approach is solid and sensible, so we
should not view this opening as an invitation to see red. I propose instead that our reaction will be
logical, with the simple aim of gaining control over the d-file combined with a queenside pawn initiative.
Usually in the Caro-Kann Black plays with a space deficit, but here White gives him the opportunity
to fight for the centre on a level footing, so the play is really based on who can manoeuvre in the most
effective way. As the stakes in the centre are equal, it is especially of importance that each side prevents
the other from establishing structural dominance on the wings.
There are three main plans for White. To each Black should adapt his method of play.

The Philidor with Colours Reversed

 White is in effect a tempo up on the Philidor Defence (1 e4 e5 2 Nf3 d6). The Philidor is a solid but
defensive set-up for Black, and the same can be said when it is played a tempo up with the white pieces.
In response, Black has several aims:

a) To manoeuvre the knight from d7 via f8 to g6, where it will threaten to jump to f4.
b) To secure the c5-square with ... a5.
c) To take control of the soon-to-be-opened d-file.

A Good IQP for Black

There is only one good central pawn break in the KIA and this is the push d3-d4 which is usually played
immediately on move 5. Black should not be afraid of entering an IQP position. He enjoys free
development and in addition is a tempo ahead on some lines of the French Tarrasch, so we may see a
position like this:

The usual IQP principles apply, but there is one difference: there is no e-pawn, which means Black
can attack the f2 point at some stage, as well as targeting the white king and controlling the open c- and e-
files. Note too that the g8-knight should be developed to e7, as ... Nf6 allows an unpleasant pin with Bg5.

The Main Line KIA

Here Black’s main plan is to play along the classical themes of development: centralize the rooks and
attempt to take control of the d-file. This should be combined with a gradual queenside expansion. An
important but little-known idea is that it is usually good for Black to provoke c2-c3 (if White hasn’t
already played it himself), the reason being to weaken the d3-square, which would make an excellent
outpost for a knight.
Returning to the position after 4 ... Bd6:
 Our decision to take over the centre is a good one. Now we have:

A) 5 Be2
B) 5 d4
C) 5 Qe2
D) 5 g3

A) 5 Be2
White intends to play a Philidor with colours reversed. It’s not very dynamic, but is pretty solid.
5 ... Nf6

6 0-0
6 d4!? is a rare move and one which gives White nothing after 6 ... Nxe4 7 dxe5 (7 Nxe5 takes the
position into a type of Petroff; 7 ... 0-0 8 0-0 Nd7 9 Nxe4 dxe4 10 Nc4 Bc7 is clearly fine for us) 7 ...
Nxd2 8 Qxd2 (if 8 Bxd2 Black should reply 8 ... Be7 9 0-0 0-0 10 c3 c5 11 Bf4 Nc6 12 Qd2 Bf5 13 Rfd1
Qb6 14 Bd3 Bxd3 15 Qxd3 Rad8, with no problems) 8 ... Bc7! (this time the bishop is needed on c7 to
eye the e5-pawn) 9 0-0 0-0 (not 9 ... Bg4 10 Qb4! Qc8? 11 Bg5 and White was already winning in
W.Sandner-R.Aschenbrenner, Austrian League 2010; a vicious trap indeed) 10 Qd4 Nd7 11 Bd3
(targeting the pawn on h7; instead, 11 Bg5? Bxe5! nets a pawn) 11 ... Bb6 12 Qf4 Nc5, removing the
dangerous bishop on d3.
6 ... 0-0 7 c3
Alternatively:
a) 7 d4 goes into an equal ending, but not one which is necessarily drawn: 7 ... Nxe4 8 dxe5 Bc7 9
Nxe4 dxe4 10 Qxd8 Rxd8 11 Ng5 h6 12 Nxe4 Bxe5 13 c3 Bf5 14 Ng3 with complete equality in
G.Welling-A.Mirzoev, Bratto 2012, although the players only agreed a draw after 59 moves.
b) 7 b3 was tried in the high-level game P.Harikrishna-M.Vachier-Lagrave, Gibraltar 2014, and after
7 ... Re8 (7 ... Qe7 was MVL’s choice) 8 Bb2 b6 9 Re1 Nbd7 10 c3 a5 White’s position is not so easy;
Black will put his bishop on a6 and the queen on c7.
7 ... Re8

Black has several aims here: to manoeuvre the knight to g6 from where it will eye the f4-square, to
secure the c5-square, and to prevent any queenside expansion with a4-a5. He would also like to take
control of the soon-to-be-opened d-file.
8 Re1
White gets straight on with his own plan of transferring the knight to g3.
Attempting to seize space immediately with 8 b4 is undermined by 8 ... a5 and 8 a4 a5! 9 Qc2 Nbd7
10 b3 Nf8 11 Re1 Qb6! (Black develops as actively as possible) 12 Nf1 Ng6 13 Ne3 (13 Bg5 was
suggested by Solozhenkin, but 13 ... Ng4 14 Ne3 Bc5 15 Nxg4 Bxg4 16 Rf1 Bxf3 17 Bxf3 d4 gives Black
a small edge) 13 ... Be6 Black has a very comfortable position with the easy plan of centralizing the
rooks. White can play 14 Ng5, but it didn’t achieve anything after 14 ... Bd7 in W.Kobese-A.Dreev,
Hyderabad 2002.
8 ... Nbd7 9 Qc2 Nf8 10 Nf1
After 10 Bf1 Ng6 White still has to come up with an active plan. Here we also have 10 ... h6 11 b4 a5
(remember, we want some queenside action) 12 Rb1 axb4 13 cxb4 Ng6 14 a3 Bd7 15 Bb2 d4! and Black
already stands better with his extra space. The position now looks much like a Reversed Closed Ruy
Lopez, J.Hickl-Y.Seirawan, Zagreb 1987.
10 ... Ng6 11 g3 h6 12 Ne3 Qc7 13 Bd2 a5!
 Now the bishop has moved to d2, White was threatening to expand on the queenside.
14 Rad1 Be6 15 Bc1 a4 16 exd5 cxd5 17 d4 e4 18 Nd2 Rec8 19 a3 h5 20 Nef1 h4
Black had a great position in M.Markovic-D.Svetushkin, Herceg Novi 2001.

B) 5 d4

This line may look scary to the uninitiated, but Black simply has to remember that he has to accept an
IQP and enjoy the free piece development. There is no fixed scheme on how Black should develop his
pieces - it’s mainly down to common sense. The rooks go on open files and the knights aim for the strong
squares, c4 and d4. Our pawn push is usually ... d5-d4, but sometimes Black can advance the a7-pawn to
inconvenience a white knight on b3.
5 ... exd4 6 exd5 cxd5!?
6 ... Nf6 7 dxc6 Nxc6 is the other main alternative.
7 Nxd4
Another approach is to throw in a check: 7 Bb5+ Nc6 8 Nxd4 and now I like 8 ... Bd7! 9 N2f3 (9
N2b3 is a better choice, but Black still has an easy plan of development: ... Nf6, ... 0-0 and ... Qc7
followed by using the open e- and c-lines: for example, 9 ... Nf6 10 0-0 0-0 11 Bg5 h6 12 Bh4 Re8 13
Bd3 Nxd4 14 Nxd4 Be5 15 c3 Qb6 is level) 9 ... Nf6 10 0-0 0-0 11 Bg5 h6 (11 ... Rc8 12 c3 Re8 13 Bd3
Nxd4 14 Nxd4 Be5 also led to a level position for me in A.Stojanovic-J.Houska, Kowalewo Pomorskie
2008) 12 Bh4 Re8 13 c3 Nxd4 14 Nxd4 Bxb5 15 Nxb5 Be5 followed by ... Qb6. The position is
completely level.
7 ... Ne7
7 ... Nc6 is another option and transpose to the 7 Bb5+ lines. Another possibility is 8 N2f3 Nf6 9 Bd3
0-0 10 0-0 Re8 11 h3 Nxd4 12 Nxd4 Ne4 and Black is doing okay.
8 Bd3
8 Be2 may surprisingly be sharper if White delays castling to secure active piece play: 8 ... 0-0 9
N2b3 Nbc6 10 Bg5 h6! (stronger than 10 ... a6 11 Bh4 Nxd4 12 Qxd4 Qc7 13 Bg3 Nf5 14 Bxd6 Qxd6 15
Qd2 Qg6, which led to a draw in S.Tiviakov-L.Nisipeanu, St Vincent 2004) 11 Bh4 (after 11 Be3 Re8 12
c3 a6 13 0-0 Qc7 14 g3 Ne5 15 Re1 Nc4 the knight on c4 gives White a few headaches) 11 ... Re8 (11 ...
Nxd4!? 12 Nxd4 Qb6 13 0-0 Nc6 14 Nb3 a5 15 Qxd5 Be5 16 c3 a4 is another example of how Black can
deal with the position actively) 12 0-0 Nxd4 13 Nxd4 Qb6 14 Re1 Nc6 15 Nb3 Be5 16 c3 a5 17 a4 d4 18
cxd4 Nxd4 and Black is the more active.
8 ... Nbc6

9 N2b3 Ne5
We will now follow a high-level game. As White no longer has any tricks with Qh5 we will simply
capture the bishop on d3.
10 0-0 0-0 11 Bg5 a6
I am not sure about this move, though. Indeed, I would rather develop my pieces 11 ... Re8 12 Re1
Bg4 13 Be2 Bxe2 14 Qxe2 Qd7 15 Rad1 Nc4 16 c3 Nc6 and Black stands well.
12 Re1 Nxd3 13 Qxd3 Qc7 14 h3
After 14 Bxe7 Bxe7 15 c3 g6 16 Qf3 Bd6 17 g3 it’s difficult for Black to develop the light-squared
bishop on c8.
14 ... h6 15 Bd2 Bd7 16 a3 Nc6 17 Nf3 Ne5
 18 Nxe5 Bxe5 19 Ba5
Wisely avoiding 19 Qxd5? Bxb2!.
19 ... b6 20 Bxb6 Bh2+ 21 Kh1 Qxb6 22 Kxh2 Qxf2 23 Qxd5 Qxc2 24 Rac1 Qf5 25 Qxf5
Here a draw was agreed in V.Kramnik-P.Leko, Nice (blindfold) 2008.

C) 5 Qe2
White plans to obtain an IQP position, but this time with the queens exchanged which should favour him.
However, this move is pure bluff because, as we shall see, White doesn’t ever get to achieve his aims.
None of the tactical tricks work for him, so in effect White is forced to play a normal KIA formation, but
with the queen on e2.
5 ... Nf6
This was my most recent choice when facing this line. My original pawn sacrifice still works though:
5 ... Ne7!? 6 exd5 cxd5 7 Nxe5 (a brave decision; should White continue with 7 c4 0-0 8 g3 Nbc6 9 Bg2
Bg4 10 cxd5 Nd4 11 Qd1 Nxd5 12 h3 Bh5 13 0-0 he has an unpleasant position and after 13 ... f6 14 Ne4
Bc7 15 Be3 Nxe3 16 fxe3 Nxf3+ 17 Bxf3 Bf7 Black will focus his play on the dark squares) 7 ... 0-0 8 d4
Re8 9 Ndf3 Nf5! (Black keeps the idea of ... f6 in reserve and instead focuses on fast development) 10
Be3 (the only move as Black was threatening ... f6, winning a piece) 10 ... f6 11 Nd3 Qe7 12 0-0-0 Nxe3
13 fxe3 Qxe3+ 14 Qxe3 Rxe3 and Black has the advantage of the bishop-pair.
6 g3
The best move, as shown by:
a) 6 d4?! exd4 7 exd5+ (7 e5 0-0 8 exd6 Re8 loses material) 7 ... Be7 8 Nxd4 0-0 9 Qf3 Bg4 10 Qb3
Re8 11 Qxb7 Qxd5 12 Ne2 and here in P.Poobalasingam-J.Houska, Brentwood 2008, I could have simply
played 12 ... Bxe2 13 Bxe2 Qxg2 14 Rf1 Nbd7 15 Qb3 Qxh2 and White is in serious trouble. Even the
endgame doesn’t relieve his problems as shown by 16 Qg3 Qxg3 17 fxg3 Bd6 18 Rf3 Re7 with some
advantage.
b) 6 exd5 is not very good either as the following game shows: 6 ... cxd5 7 Nxe5 0-0 8 d4 Re8 9 Ndf3
Nc6 10 Be3 Nxe5 (10 ... Bf5 is also good) 11 dxe5 Bxe5 12 Nxe5 Rxe5 13 h3 Bf5 14 c3 Qa5 15 Qd2
Ne4 16 Qd4 Rae8 17 b4 Qa3 18 Bb5 Nxc3 19 Bxe8 Qb2 20 Bb5 Qxa1+ 21 Kd2 Qb2+ 0-1, L.Carmaciu-
L.Jarmula, Cappelle-la-Grande 2014.
6 ... 0-0 7 Bg2
We now have a King’s Indian Attack set-up but with the queen on e2 which makes Black less
vulnerable to any exd5 ideas.
7 ... Re8 8 0-0 a5

I like the simplicity of this plan. We simply aim to establish a knight on c5 or b4.
9 a4
Alternatively:
a) One of Black’s main ideas is 9 b3 a4!.
b) After 9 Re1 Na6 10 Nf1 dxe4 11 dxe4 Nc5 White doesn’t get to develop in the way he wants to.
c) 9 Rd1 is one way to make some sense of White’s queen move. After 9 ... Be6 10 Nf1 (or 10 Ng5
Bg4 and the knight must retreat) 10 ... Nbd7 11 Ng5 Bg4! 12 Nf3 h6 13 h3 Be6 14 Nh4 Nc5 15 Nf5 Bxf5
16 exf5 Qd7 17 g4 e4 18 dxe4 Ncxe4 19 Be3 a4 Black is close to establishing complete superiority on
the queenside.
9 ... Na6 10 Nb3 h6 11 Re1 Be6 12 Bd2 Nb4

13 Bc3 d4!
This seems even better than the 13 ... Qc7 of I.Smirin-V.Iordachescu, Khanty-Mansiysk Olympiad
2010.
14 Bxb4 axb4
 Black has an easy target: the weak a4-pawn.
15 Nfd2 Qe7 16 Qf1 Ra6 17 Nxd4
17 f4 Ng4! would send the knight off towards a fantastic square on e3.
17 ... exd4 18 e5 Bxe5 19 Rxe5 Re8 20 b3 b5
Black is doing well.

D) 5 g3
The last of the set-ups we must consider and the main continuation for White.
5 ... Nf6 6 Bg2 0-0 7 0-0 Qc7!?

I still like this approach, sidestepping some of the drawing lines White has after 7 ... Re8. The plan
behind this move is to develop the f8-rook to the d-file rather than the customary e-file. Please note that as
long as Black plays logically and constructively, he should stand more than okay, since White’s set-up is
fairly toothless. Note too that when the knight on d2 goes to f1 Black will capture on e4. Our goal: to
achieve superiority on the queenside.
8 Re1
If 8 Nh4 Bg4 9 Qe1 then 9 ... Re8 is now indicated, while after 8 b3 a5 (8 ... Be6 is also possible as
after 9 Ng5 Bg4 10 f3 Bh5 11 g4 Bg6 12 h4 Nbd7 13 h5 h6! White’s dark squares are too weak) 9 a3 Bg4
10 h3 Bh5 11 Bb2 Nbd7 (this is pretty much all standard) 12 Qe1 (or 12 c4 dxe4 13 dxe4 Bc5 when
Black has active pieces and the straightforward plan of controlling the d-file) 12 ... Rfe8 13 Nh4 Bg6 14
Nxg6 hxg6 15 h4 b5 16 Bf3 d4 17 Kg2 a4 18 Bd1 c5 19 Rh1 Nb6 20 Nf3 axb3 21 cxb3 c4 Black was
pressing and went on to win in Zhang Xiaowen-Guo Qi, Wuxi 2013.
8 ... Be6!?
 This is Grandmaster Peter Wells’ idea. There are now plenty of moves available to White, but the
theme is pretty clear for Black: in some order to place a rook on the d-file, secure the c5-square for a
piece with ... a5, swap on e4 and push on the queenside.
Black normally delays developing the light-squared bishop, preferring to secure the location for his
queen’s knight first. The traditional approach also can work well, as we can see in Caro-Kann expert
Riazantsev’s games: 8 ... Nbd7 9 b3 (or 9 Nf1 dxe4! following the rule, and if 10 dxe4 Nc5; instead, 9 c3
Re8 10 d4 exd4 11 Nxd4 Ne5 12 exd5 Nxd5 is equal) 9 ... Re8 10 Bb2 a5 11 a3 dxe4 12 dxe4 b5 13 Nh4
Bf8 14 Qe2 Nb6 15 Rad1 a4 16 b4 Be6 17 Nf5 Rad8 18 Ne3 c5 19 Qxb5 Bd7 20 Qe2 cxb4 21 axb4
Bxb4 with a big advantage for Black in D.Andreikin-A.Riazantsev, Khanty-Mansiysk (blitz) 2013.
9 c3?!
Although this move looks harmless enough, it is a serious positional concession. Black is immediately
given an entry square on d3 for his knight, rook or queen, which is something that should not be
underestimated at all. A knight on d3 can immediately harass the b2-pawn and either force White on to the
defensive or into a concession. Equally a rook or queen on d3 can cut the white position in two and/or
allow Black to set up an Alekhine Cannon (treble on the d-file with the queen behind the rook).
The alternatives are better, if not exactly scary:
a) 9 b3 Nbd7 10 Bb2 and now:
 a1) 10 ... dxe4 11 Nxe4 Nxe4 12 dxe4 f6! looks quite nice and improves over 12 ... Rfd8 13 Ng5 Nf8
14 Nxe6 Nxe6, which was level in T.Taylor-J.Banawa, Los Angeles 2011.
a2) 10 ... Rfe8 (overprotecting the e5-pawn) 11 h3 (11 exd5 cxd5 12 c4 d4 13 Ne4 Nxe4 14 Rxe4 Bf5
15 Re2 Nc5 16 Rd2 a5 secures the c5-square for the knight; after 17 Ng5 Bg6 18 Bd5 Be7 19 Ne4 a4 20
b4 Nxe4 21 dxe4 Bxb4 22 Rd3 and the situation is complicated, but White has to be careful with his
bishops cut off from the defence of their king) 11 ... dxe4 12 dxe4 Rad8 13 Ng5 Nf8 14 Qe2 Bc5 15 Nc4
Bc8 16 Rad1 Rxd1 17 Rxd1 b5 18 Nd2 Rd8 19 a3 a5 and Black is pressing.
b) 9 d4 simply leads to a drawish position after 9 ... Nbd7 10 dxe5 Nxe5 11 Nxe5 Bxe5 12 exd5
Bxd5. That said, I would prefer to play Black here, mainly because of the weakness of the light squares
around the white king and because Black will be first to get his rooks to the centre.
c) 9 Nf1 (planning to route the knight via e3 to f5, but this should remind us of a certain trigger) 9 ...
dxe4 10 dxe4 Rd8 11 Qe2 h6 12 Ne3 Nbd7 13 Nf5 Bf8 14 Bd2 Nc5 15 Bc3 Nfd7 16 b3 g6 17 N5h4 a5
18 Rad1 b5 and it is Black who is pressing for the advantage on the queenside.
d) 9 h3 has been another recent try, but with 9 ... Re8 (after 9 ... h6 10 Nh4 Nbd7 11 Qf3 a5 12 Nf1
dxe4 13 dxe4 Black is okay, but White is probably happy too with the threat of starting a kingside attack,
R.Britton-R.Kane, London 2011) 10 Nh4 Nbd7 11 Qf3 Bf8 12 g4 g6! Black keeps control over the
kingside.
 Indeed, after 13 Nf5 (or 13 Nf1 dxe4 14 dxe4 Rad8 15 Bg5 Be7 16 Nf5 gxf5 17 gxf5 Bc4 18 Qg3 Kh8
and White doesn’t have an attack) 13 ... dxe4 14 dxe4 Rad8 15 Nf1 gxf5 16 gxf5 Bc4 17 Ne3 Ba6 18 Ng4
Nxg4 19 Qxg4+ Kh8 20 Be3 White has nothing.
e) Finally, please note that 9 Ng5 is met by 9 ... Bg4! when the best move for White is probably to
retreat the knight.
9 ... Rd8 10 Qe2 Nbd7 11 h3 h6 12 Nh4 Bf8 13 Qf3 Nc5!
Targeting the d3-square.
14 Nf5?
A natural mistake. White superficially improves the placing of his knight only to fall foul of a tactic.
14 ... Bxf5 15 exf5 e4!
16 dxe4 dxe4 17 Qe2 Nd3 18 Rf1 e3! 19 Qxe3 Bc5 20 Qe2 Qxg3 21 Qf3 Qxf3 22 Nxf3 Re8
Black managed to convert his material advantage in J.Howell-P.Wells, British Championship,
Swansea 1987.

Summary
The KIA is a quiet non-confrontational system primarily because White never gets to play the usual e4-e5
pawn push in the centre. That’s because Black plays 3 ... e5 himself. White can challenge the centre
immediately with 4 Nf3 Bd6 5 d4!? when Black simply needs to accept the IQP and develop.
5 Qe2 sets a trick that doesn’t work (threatening the e5-pawn) and leads to a worse version of the main
line. There Black should meet White’s Re1 and c3 set-up with 7 ... Qc7 usually followed by ... Rd8,
obtaining easy play.
Chapter Sixteen
Rare Lines
1 e4 c6

To be honest, in this final chapter some of the lines I came across are unorthodox but somewhat
logical, while others are plain bizarre. I will mention them all for completeness sake. I must add a few
words on the best way to deal with these opening surprises psychologically. The first thing to do is to
react calmly and logically. Natural developing moves are rarely bad, but you should not fall into the trap
of thinking that Black stands better immediately. Whilst in a minority of cases this is true, in most
variations all White has done is to deflate his own advantage, nothing less.
We start this chapter by returning to the Classical Main Line and then move on to some unusual second
moves:

A) 2 d4 d5 3 Nc3 dxe4 4 f3
B) 2 d4 d5 3 Nc3 dxe4 4 Nxe4 Bf5 5 Nc5
C) 2 Ne2
D) 2 Nc3 d5 3 Qf3
E) 2 Nf3 d5 3 exd5 exd5 4 Ne5
F) 2 f4
G) 2 b3

A) 2 d4 d5 3 Nc3 dxe4 4 f3
 The Blackmar-Diemer Gambit, one of the few openings with a cult following. White sacrifices a
pawn to obtain fast development and attacking prospects down the f-file. However, it appears to me that
Black can get a simple Caro-style position minus the white f-pawn.
4 ... exf3
The best recipe for dealing with these gambits is to accept the material and hold on.
5 Nxf3
5 Qxf3 has also been seen, but 5 ... Qxd4 6 Be3 Qb4! is awkward for White, and if 7 0-0-0 Bg4!.
5 ... Bf5!
The most natural-looking move and the best in my opinion. The light-squared bishop will be excellent
defender of the kingside.
6 Bc4
Alternatively:
a) With 6 Bd3 White wants to get rid of the important f5-bishop, but after 6 ... Bxd3 7 Qxd3 e6 8 0-0
Nf6 9 Bg5 (9 Bf4 does not pose a threat at all: 9 ... Be7 10 Rae1 0-0 11 Ne5 Nbd7 12 Rf3 c5! and by
undermining the e5-square, Black secures himself an advantage) 9 ... Be7 10 Rae1 0-0 11 Ne5 Nbd7 12
Nc4 Nd5 13 Bxe7 Qxe7 and Black had the advantage in M.Hirtreiter-D.Law, correspondence 1993.
b) After 6 Bg5 Black can simply develop his knight with 6 ... Nf6 7 Bc4 (exchanges with 7 Bxf6 are
not what White wants) 7 ... e6 8 Ne5 Be7 9 0-0 and now Black has time for 9 ... 0-0 10 g4 Bxc2. Black
has the advantage after 11 Qd2 h6 12 Bh4 Bh7 since White clearly doesn’t have enough for his pawns.
6 ... Nf6
 We now come to a parting of the ways. White can attack target f7 or continue his development:

A1) 7 Ne5
A2) 7 0-0

A1) 7 Ne5 e6 8 g4!?

8 0-0 is a move I have often faced on the Internet. The calm reaction is: 8 ... Bg6 (better not to grab too
many pawns; moreover, 8 ... Bd6? 9 Nxf7! Kxf7 10 Rxf5 wins) 9 Bg5 and here Schandorff’s 9 ... Nbd7! is
good enough, as shown by 10 Nxd7 Qxd7 and 10 Qe2 Be7.
8 ... Bg6
Simplest, although 8 ... Nfd7!? 9 gxf5 Nxe5 10 Be2 Qh4+ 11 Kf1 Qh3+ 12 Kg1 Be7 13 Bf4 Ned7 14
fxe6 Qxe6 is slightly better for Black.
9 h4
This has been played in a few correspondence games.
9 0-0 is not dangerous because of 9 ... Bd6! 10 Bg5 Nbd7 11 Qe2 Qb6! 12 Nxd7 Kxd7, which is
reminiscent of the Scandinavian Opening as the black monarch takes control over its destiny and starts the
process of castling by hand. Following 13 Bxf6 gxf6 14 Rad1 Qxb2 White has nothing, since everything is
protected after 15 Qe3 Qb4 16 Be2 Bc7 17 Rxf6 Qd6.
Likewise, after 9 g5 Nd5! 10 0-0 Bd6 and 9 Bg5 Nbd7 10 Qe2 Bb4 11 0-0-0 Qa5 White has little to
show for his pawn deficit.
9 ... Nbd7
Perhaps even better is 9 ... Bb4!?.
10 Nxd7
After 10 Qe2 Bb4 11 Rh3 Nb6 12 h5 Be4 again Black is doing well.
10 ... Qxd7 11 h5 Be4!
 12 Rf1
Black will challenge for the initiative after this, but White’s problem is that 12 Nxe4 Nxe4 13 Qd3 f5!
maintains the strong knight on e4. After 14 gxf5 exf5 15 Be3 0-0-0 16 0-0-0 Kb8 (protecting the a7-pawn
against any d4-d5 ideas) 17 Qb3 (or 17 Rhg1 Bd6 18 Rg2 Rhe8 19 Rf1 Ng3 20 Rxg3 Bxg3 21 Bg5 f4 and
it looks all over for White) 17 ... Bd6 18 Be6 Qc7 19 Bxf5 Ng3! Black is winning.
12 ... Bb4 13 h6 Rg8 14 hxg7 Nd5 15 Bd2 Bxc3 16 Bxc3 Nxc3 17 bxc3 Rxg7
Black is a lot better.

A2) 7 0-0
Now Black must blunt the bishop on c4 and play:
7 ... e6 8 Ng5

This seems to me to be the most dangerous try against our set-up. White threatens a big trick with Nxf7
and now has additional possibilities of sacrificing on e6.
8 ... Bg6 9 Bxe6!?
9 Ne2 was tried in B.Go-A.Donchenko, Groningen 2013, but is quite slow:
a) However, after 9 ... c5 10 c3 Black needs to tread very carefully, which he didn’t with 10 ... h6?
(10 ... cxd4!) 11 Nxe6 fxe6 12 Nf4 Bf7 13 Bxe6 Bxe6 14 Nxe6 Qb6 15 Qe2 in the game.
b) 9 ... Bd6 would be a natural reaction: for example, 10 Bxe6 fxe6 11 Nxe6 Qe7 12 Re1 Ne4 and
Black stands well.
Instead, after 9 Qe2 Bd6 10 Bxe6 0-0 11 Bc4 Re8 12 Qd1 Nbd7 13 Bd3 Qb6 White’s play along the
f-file really doesn’t amount to much.
9 ... fxe6 10 Nxe6 Qd7!

An important move. The queen must protect her king.

11 Qe2
11 Re1 Be7! would snuff out any trouble down the e-file.
11 ... Kf7 12 Ng5+ Kg8 13 Qc4+ Nd5 14 Bd2 h6 15 Nf3 Bd6
Black is beginning to unravel and should be able to convert his extra piece.

B) 2 d4 d5 3 Nc3 dxe4 4 Nxe4 Bf5

We now return to the Classical Main Line to look at one offbeat variation which takes the game into new
territory.
5 Nc5 Nd7!?
GM Riazantsev’s choice, giving up the pawn but gaining so much more in return. In my original book I
recommended the solid and simple line 5 ... b6 6 Nb3 e6 7 Nf3 Nf6, and if 8 g3 a5. This line is still
viable and good, but I just found something even better.
6 Nxb7 Qc7
A key move. Now White’s reply is forced.
7 Nc5 Nxc5 8 dxc5 Rd8!
 Displacing the queen and gaining time with the rook. So far these are the critical moves to remember.
Here White has two real options:

B1) 9 Bd3
B2) 9 Qe2

Lesser options:
a) 9 Qf3?! allows Black to pick up the sacrificed pawn immediately with 9 ... Bxc2 when White must
already deal with both ... Qa5+ and Black’s easy development.
b) Likewise, after 9 Bd2?! Nf6 10 Nf3 e5 11 Be2 Bxc5 12 0-0 e4 13 Nh4 e3 White is already in
serious trouble.

B1) 9 Bd3 e6
Black’s play is straightforward. He will exchange on d3 at the right time and then target the weak d3- and
c5-pawn. There is no big rush as White can hardly move.
10 Be3
10 Nf3 was played recently, but after 10 ... Bxc5 11 0-0 Bxd3 12 cxd3 Ne7 13 Qe2 Rd5 14 Bg5 Ng6
15 Rac1 0-0 Black didn’t have any problems at all in D.Paravyan-A.Riazantsev, Moscow (blitz) 2014.
10 ... Nf6
 11 Qe2
Now the time is right for Black to initiate exchanges, but if 11 Nf3 e5 12 0-0 (12 Nh4 Bg4 13 f3 Nd5
14 Bf2 Be6 15 g3 Nb4 gives White all sorts of problems) 12 ... e4 13 Nd4 Rxd4 14 Bxd4 exd3 15 Bxf6
gxf6 16 Qf3 Be6 17 Qxf6 Rg8 18 cxd3 Bxc5 19 Rfe1 Qd6 Black should be better.
11 ... Bxd3 12 cxd3 Qa5+ 13 Qd2 Qb5!
Piling pressure on the weak d3-pawn means that Black can now recapture on c5 should White not
force the pace.
14 a4! Qxd3 15 Qxd3 Rxd3 16 Ne2 Ng4! 17 Nc1 Rd5 18 b4 a5!

19 Ke2
Better was 19 bxa5!, although after 19 ... Nxe3 20 fxe3 Bxc5 21 Ke2 Ke7! 22 Nd3 Bd6 White’s two
a-pawns will fall and he remains worse.
19 ... Nxe3 20 fxe3 axb4 21 Nd3
After 21 a5 Bxc5 22 a6 Kd7 Black will catch that a-pawn.
21 ... Bxc5 22 Rhc1 Bd6 23 Rxc6 Kd7 24 Rc4 Rc8 25 Rxc8 Kxc8 26 Rc1+ Kd7 27 h3
Black was much better in M.Krzyzanowski-D.Fridman, Wroclaw 2010.
B2) 9 Qe2

White’s only serious alternative to 9 Bd3, but Black still regains his pawn, leaving the white queen
stuck on a stupid square.
9 ... e6 10 Be3
Alternatively, 10 Qc4 Qe5+ 11 Be2 Qxc5 12 Qxc5 Bxc5 13 c3 Ne7 14 Nf3 f6 15 0-0 e5 16 b4 Bb6
17 a4 Nd5 18 Bb2 Nf4 is another example of how play might develop.
10 ... Qa5+ 11 c3 Bxc5 12 Qc4
Trying to make sense of his queen move, but it doesn’t work out. That said, 12 b4 fails to 12 ... Qa3
and already it’s all a bit disastrous for White.
12 ... Bxe3!
All this is forced - and good.
13 Qxc6+ Ke7!

The king is safe on e7 and Black’s lead in development give him easy play.
14 fxe3 Nf6 15 Bb5?!
15 Nf3 Be4 16 Qb5 Qxb5 17 Bxb5 Rb8 18 c4 (or 18 a4 a6 19 Bxa6 Rxb2 20 0-0 Rd8 21 Rfd1 Rxd1+
22 Rxd1 Nd5 23 Rd4 Bxf3 24 gxf3 Nxe3 with the advantage) 18 ... a6 19 Bxa6 Rxb2 20 0-0 Ra8 21 Bb5
Raxa2 is also promising for Black on account of his active rooks and White’s weak pawn islands.
15 ... Rd6 16 Qb7+ Nd7 17 Bxd7 Rxd7 18 Qf3
After 18 Qb4+ Black should play 18 ... Qxb4 19 cxb4 Rc8 when his rook will find its way on to the
cherished seventh rank.
18 ... Rb8 19 b4 Qe5 20 Ne2 Rxb4 21 0-0 Rb2! 22 Nd4 Qe4 23 Qg3 Bg6 24 Rf4
This was E.Lavendelis-A.Riazantsev, Warsaw (rapid) 2011, where Black should simply have
retreated:
24 ... Qb7 25 Raf1 e5! 26 Nf5+ Kf8
White lacks an effective blow and finds himself coming up short.

C) 2 Ne2
In 2011 during the European Team Championships I was told by our captain to check out a hot new game
in this 2 Ne2 line. After seeing that game it went straight into my Caro database. It also inspired Richard
Palliser’s chapter in Dangerous Weapons: The Caro-Kann and I will be repeating the line here.
2 ... d5 3 e5 d4!

Seizing some very useful space and asking White just how he will untangle his kingside and hold the
e5 point. His main choices are:

C1) 4 c3
C2) 4 Ng3

White has a variety of other moves here. Black responds to each in different (arguably even seemingly
random) ways, but there is a clear theme running throughout and that is to pressure the e5-pawn:
a) 4 c4?! was mentioned by Jeroen Bosch in the forum section of Secrets of Opening Surprises 8.
Black should simply fianchetto the king’s bishop and get on with development: 4 ... g6 5 d3 Bg7 6 f4 Nh6
7 Ng3 0-0 8 Be2 Nf5 9 Nxf5 Bxf5 10 0-0 f6 (please note too that this undermining move happens right at
the end of our development) 11 exf6 exf6 and Black has a pleasant space advantage.
b) 4 Na3!? aims to apply to pressure the potentially weak d4-pawn:
b1) 4 ... g6!? 5 c3 c5! (the d4-pawn must be supported) 6 cxd4 cxd4 7 Qa4+ Nc6 8 Nb5 Bd7 9 Nexd4
Bg7 10 Nxc6 bxc6 and Black wins the e-pawn. Thus, again our tagline here is: fianchetto.
b2) 4 ... f6 is also an interesting option; after all normal moves are hardly standard in this variation.
Here 5 Ng3 (or 5 f4 fxe5 6 fxe5 Qd5 and the pawn on e5 falls) 5 ... Qd5 6 exf6 Nxf6 7 c4 Qd6 8 d3 e5 9
Be2 Be6 10 Nc2 Be7 11 0-0 0-0 12 f4 Nbd7 seems very good for Black.
c) After the quiet 4 d3 I like adopting a Gurgenidze set-up with 4 ... g6 (again, we fianchetto) 5 Ng3
(or 5 c3 c5 6 b4 cxb4 7 Nxd4 bxc3 and Black cannot stand worse) 5 ... Bg7 6 f4 and now Black can
undermine the e5-pawn immediately with 6 ... f6 7 exf6 Nxf6 8 Be2 Nd5 9 0-0 0-0 10 Na3 b5. Here
Black has an easy plan of expanding on the queenside with moves such as ... Na6-c7, while White’s
pieces are hampered by the pawn on f4.
d) After 4 Nf4 as a guide when White releases the pressure on the d4-pawn, Black can attack the e5-
pawn.

Indeed, we should attack e5 with the queen: 4 ... Qc7! 5 Nd3 (if 5 Qe2 Black should play Palliser’s
suggestion of 5 ... g5! 6 Nd3 Bf5 - suddenly ... Nd7 is a large threat - 7 c3! dxc3 8 dxc3 Nd7 9 Bxg5
Bh6!? 10 Bxh6 Nxh6 11 f4 0-0-0 with promising play for the pawn in view of Black’s superior
development) 5 ... Bf5 6 f4 e6 7 Qf3 Nd7 8 g4 Bxd3 9 Bxd3 g5! 10 c3 dxc3 11 dxc3 gxf4 12 Bxf4 Nxe5
and Black had won a pawn in A.Savage-J.Sarkar, Arlington 2010.
e) 4 b4 is whacky if strangely consistent and we should attack the b4-pawn: 4 ... e6 (for once a
natural move; 4 ... f6 is more complicated but also viable) 5 Bb2 c5! 6 bxc5 Bxc5 7 Nf4 Nc6 (all very
natural moves indeed so far) 8 Nd3 Be7 9 Qg4 g6 10 c3 Nh6 11 Qe4 Nf5 (11 ... f5!? is a good
alternative) 12 Na3 0-0 13 Nc2. This was M.Meinhardt-D.Hausrath, Viernheim 2014, and now after 13 ...
dxc3 14 Bxc3 Qd7 15 g4 Nh4 16 Be2 Rd8 Black will play ... b7-b6 with a wonderful position.
f) After 4 f4 Black copies and adopts a mirror approach: 4 ... c5 (another approach is 4 ... Nh6 5 Ng3
g5!? 6 fxg5 Ng4, breaking up White’s centre) 5 Ng3 Nc6 possibly followed by a Gurgenidze-style ... g6
and ... h5, which seems quite comfortable for Black.
g) Finally, 4 g3? drops a pawn to 4 ... Qd5.

C1) 4 c3 c5
 Of course, the asset on d4 must be supported.
5 b4!?
This vigorous move was proposed by Gawain Jones in his notes to the game which initially inspired
me, Abergel-Edouard. Again, there are several alternatives:
a) 5 Ng3?! (once the pressure is released on d4, Black can attack the e5-pawn) 5 ... Qd5! 6 f4 g5
thematically undermines the white centre and after 7 Bb5+ (7 fxg5?! Qxe5+ 8 Qe2 Qxg5 9 Ne4 Qd5 10
Na3 Nc6 doesn’t give White anywhere near enough for his pawn) 7 ... Nc6 8 Qf3 Qxf3 9 gxf3 gxf4 10
Ne4 Bg7 11 Nxc5 Bxe5 I agree with Palliser that Black’s position is better.
b) After 5 Nf4 Black doesn’t attack the e5-pawn, but instead plays the more solid 5 ... e6!?:
b1) 6 Na3 is met with the logical 6 ... Nc6 7 Bb5 Bd7 and 8 cxd4 doesn’t quite work, since 8 ... cxd4
9 Qa4 fails to 9 ... Qg5!.
b2) 6 Bb5+ was tried very recently when our Caro-Kann guru Riazantsev countered with 6 ... Bd7 7
Na3. Now in A.Motylev-A.Riazantsev, Moscow (blitz) 2014, Black might have called White’s bluff with
7 ... Bxb5! 8 Nxb5 a6 9 Na3 Nc6 when the e5-pawn will fall.
5 ... Qd5
This should be fine for Black.
6 bxc5
6 d3 Qxe5 7 bxc5 Qxc5 8 Nxd4 Nc6 is one plausible line leading to rough equality.
6 ... d3!
 An important move to remember.
7 Qa4+ Nc6 8 Ng3 Qxe5+ 9 Qe4 Nf6
9 ... Qxe4+ is also great for Black, as shown by the sample line 10 Nxe4 f5! 11 Ng5 e5 12 Bxd3 h6 13
Nh3 Bxc5 14 0-0 Be6 15 Bb5 Nf6 16 Re1 Kf7 17 Bxc6 bxc6 18 Rxe5 Bd6 19 Re1 Ng4 20 g3 Ne5.
10 Qxe5 Nxe5 11 f4 Ng6 12 Bxd3
Black is happy to see this, but if 12 f5 Ne5!.
12 ... Nxf4

13 Bb5+?!
Not ideal, but 13 Bf1 e5 (otherwise White will play d2-d4 anyhow) 14 d4 Ne6 15 Bb2 Bd7 16 Bc4
b6! gives Black very easy play.
13 ... Bd7 14 0-0?
14 Bxd7+ had to be played, but after 14 ... Kxd7 15 0-0 Nd3! 16 Ba3 e6 the c-pawn falls.
14 ... Bxb5 15 Rxf4 e5 16 Rf5 Nd7 17 d4 g6 18 Rf2 Bg7 19 Na3 Bc6 20 d5 Bxd5 21 Nb5 0-0 22
Nc7 Bc6 23 Nxa8 Rxa8 24 Bb2 Nxc5 25 Rd1 e4 26 Re2 Nd3
White had already seen enough and gave up in M.Codenotti-D.Lemos, Seville 2011.
C2) 4 Ng3

White has removed the immediate pressure on the d4-pawn, so Black can follow the guide by playing:
4 ... Qd5! 5 Qe2
5 f4 allows the typical undermining advance 5 ... g5!. Following 6 d3 (6 fxg5 Bg7 recovers the pawn
with interest) 6 ... gxf4 7 Bxf4 Bh6! (Black needs to destabilize the e5-pawn) 8 Qd2 Bxf4 9 Qxf4 Nd7 10
Nd2 Qxe5+ (10 ... Nxe5 was less clear in F.Farran Martos-M.Narciso Dublan, El Sauzal 2010) 11 Qxe5
Nxe5 12 Nb3 Nf6 13 Nxd4 h5 14 Be2 h4 15 Ne4 Nxe4 16 dxe4 Rg8 Black has a very playable position,
as shown by 17 Kf2 Bd7 18 Rhf1 c5!.
5 ... Nd7 6 f4
White might like to go 6 e6, but 6 ... fxe6 (6 ... Nc5!? 7 exf7+ Kxf7 followed by expanding in the
centre can’t be bad either) 7 d3 e5 8 c4 Qd6 fails to give him enough compensation.
6 ... g6!?
Also possible is the dramatic 6 ... g5!?, a move that Black should always try to make work:
a) 7 fxg5 is clearly a mistake: 7 ... Nxe5 8 Qe4 and now Black’s best might be the simple 8 ... Qxe4+
9 Nxe4 h6 with an excellent position.
b) 7 c4?! Qe6! prepares to surround the e5-pawn after all with 8 d3 (8 f5?! Qxe5 9 d3 h6 failed to
give White anything for his pawn in G.Smith-M.Broom, British League 2003) 8 ... gxf4 9 Bxf4 Bg7
(Bosch).
c) The most critical try here is 7 e6!? fxe6 8 Qh5+ Kd8 9 fxg5 h6! 10 d3 Qe5+ 11 Be2 hxg5 12 Qxg5
Qc7 when the position is rather messy.
7 d3
Now Black can strike out, so White might prefer 7 c4. Black can, though, counter actively and
effectively: 7 ... Qa5 8 a3 Bh6 9 b4 Qc7 10 Qe4 f6 11 Qxd4 fxe5 12 fxe5 Nxe5.
7 ... f6!
 Undermining the e5-pawn by delicate means.
8 exf6
The alternatives are scarcely ideal for White:
a) 8 Ne4 fxe5 9 g3 Ngf6 10 Bg2 Nxe4 11 Bxe4 (if 11 Nd2 Nxg3) 11 ... Qc5 12 0-0 Bg7 as after the
natural 13 f5 Black can play 13 ... Nf6! 14 fxg6 Nxe4 15 Qxe4 Bh3! 16 Rf2 Qd5 17 Qxd5 cxd5 18 Rf7 (or
18 gxh7? e4! when Black’s passed pawns are incredibly strong) 18 ... Bf6 19 Rxh7 Kd7 and he is better;
he will win g6-pawn and use the two bishops to aid the advance of his centre pawns.
b) 8 c4 Qa5+ 9 Bd2 Qb6 10 b4 fxe5 11 fxe5 Bg7 and White can’t hold on to the e5-pawn.
8 ... Ngxf6 9 Ne4 Bg7 10 c4 Qe6

Already White must be rather careful, trailing somewhat in development.

11 Nbd2
Alternatively, 11 g3 Nxe4 12 dxe4 and now Black has the surprising 12 ... h5 13 Bg2 h4 14 Nd2 hxg3
15 hxg3 Rxh1+ 16 Bxh1 Qh3 17 Qf3 Nc5 when again he is the only one having any fun.
11 ... Nxe4 12 Nxe4
If 12 dxe4 Qf6! followed by ... e7-e5.
12 ... Nf6 13 Nf2
 A complicated position is reached after 13 g3 Nxe4 14 dxe4 0-0 15 Bg2 Qd6 16 e5 d3 17 Qe3 Qb4+
18 Bd2 Qxc4 19 0-0 Bf5, but Black is a pawn up.
13 ... 0-0
Black can be happy. Once the queens come off he will make way for ... e7-e5 and enjoy the resultant
open lines.

D) 2 Nc3 d5 3 Qf3

Bringing the queen out immediately is not without its danger for Black, but if dealt with correctly it
should be harmless.
3 ... dxe4 4 Nxe4 Nf6
Simple and effective development as 5 Nxf6+ exf6 6 Bc4 Nd7 7 d4 Nb6 is in no way a problem for
Black.
5 Bc4 Nbd7 6 Nxf6+
After 6 d4 Nb6 7 Bd3 Qxd4 8 Ne2 we should follow Schandorff’s recommendation of 8 ... Qd7 9 h3
(if 9 0-0 Qg4) 9 ... Nxe4 10 Qxe4 (or 10 Bxe4 f5 11 Bd3 e5) 10 ... Qd5 11 Qh4 Bf5 with easy play.
6 ... Nxf6 7 Ne2 e5 8 Qg3 Qd6 9 0-0
 9 ... Be6
Calm and best. With 9 ... b5 Black punches out on the queenside too early: 10 Bb3 c5 11 d3 a6 12 Re1
Ra7 13 a4 and Black had to weaken further squares on the queenside in L.Galego-R.Swinkels, Torres
Vedras 2011.
10 Bxe6 Qxe6 11 d4 Bd6 12 Qxg7?
Too risky, as is 12 dxe5 Bxe5 13 Nf4 Qf5 14 Qxg7?? (this loses on the spot; instead, 14 Re1 0-0-0!
gives Black a small nibble on account of having the better control over the central lines) 14 ... Ng4 15
Re1 0-0-0 and wins.
12 ... Rg8 13 Qh6 0-0-0 14 Bg5 Rg6 15 Qh4 Rdg8
Black has the superior position.

E) 2 Nf3 d5 3 exd5 cxd5 4 Ne5!?

The idea behind this strange move is that the check on b5 can become rather uncomfortable to meet.
4 ... Nf6 5 d4 g6
A huge amount of moves have been tried here. The only real issue for Black is how to handle White’s
strong knight on e5. Usually Black does it relatively simply by putting his own knights on c6 and d7,
which is why if this line is to have any power to it White must throw in the check on b5:

E1) 6 c3
E2) 6 Bb5+

We should first consider some non-challenging lines:

a) 6 Nd2 is too passive. After 6 ... Bg7 7 Bd3 Nc6 8 Ndf3 0-0 9 c3 Nd7 Black removes the strong
knight and prepares an ... f6 and ... e5 pawn break. Practice has seen 10 Nxd7 Bxd7 and then:
 a1) 11 Qb3 didn’t fare too well for White after 11 ... Bg4! 12 Qxb7 Bxf3 13 gxf3 Qd6 14 f4 and now
in A.Postojev-U.Haag, Staufer 2010, Black should have played 14 ... Bxd4, as after 15 cxd4 Nb4 he is
attacking White’s bishop and threatening to trap the white queen.
a2) 11 Bf4 Bg4 12 0-0 f6! (playing for e5) 13 Be2 e5 14 dxe5 fxe5 15 Bg3 and Black had a very
promising position in M.Sebag-H.Vogel, Dresden 2004.
b) 6 Bd3 Bg7 7 0-0 (7 Bf4 is dubious because of 7 ... Qb6 8 b3 Nfd7 when it is not easy for White to
support d4 and e5) 7 ... 0-0 8 Nd2 Nc6 (8 ... Nfd7 also fits in with our theme of undermining the e5-
knight) 9 Nxc6 bxc6 10 Nb3 e5 11 dxe5 Ng4 12 Bf4 Nxe5 13 c3 Nxd3 14 Qxd3 a5 and Black had a very
pleasant position in D.Zakarian-H.Lazaridis, Greek Team Championship 2008.
c) 6 c4 Bg7 7 Nc3 0-0 8 Qb3 (or 8 Be2 dxc4 9 Bxc4 Nfd7!, exchanging the strong knight, the only
strong feature in White’s position; after 10 Nxd7 Qxd7 Black had no problems in V.Sareen-V.Burmakin,
Balaguer 2006) 8 ... Nc6! gave Black an excellent Grünfeld in O.Bohaddoune-L.Akhrouf, Marrakesh
2010.
d) 6 h4 is just hot air. After 6 ... Bg7 the game should just continue as normal, since there is nothing
for White after 7 h5 Nxh5 8 Rxh5 gxh5 9 Qxh5 Bxe5.
e) 6 Be3 Bg7 7 Nc3 0-0 8 Qd2 Nc6 9 0-0-0 (White is trying to get something like a Barry Attack, but
here it just doesn’t work) 9 ... Be6! 10 h4 Nxe5 11 dxe5 Ne4 12 Nxe4 dxe4 and Black already has the
advantage on account of the weak e5-pawn.

E1) 6 c3 Bg7 7 Bd3

7 Nd2 0-0 8 f4 is an alternative way for White to reinforce the e5-knight, but doing this gives Black the
e4-square, which we exploit with 8 ... Nc6 9 Bd3 Bf5!.
 I love this move, although, of course, it’s not an unfamiliar strategy for Caro-Kann players. Following
10 Bxf5 (or 10 Qe2 Bxd3 11 Qxd3 Nd7 12 0-0 Ndxe5 13 fxe5 f6 14 exf6 exf6 15 Nf3 Qb6! when Black
will play to double his rooks on the e-line and reroute his c6-knight to a better square, M.Weller-
N.Maleki, Differdange 2008) 10 ... gxf5 11 0-0 e6 12 Rf3 Ne7 13 Qe1 Ng6 14 Nxg6 hxg6 15 Rh3 b5 16
Nf3 Ne4 Black didn’t have any problems whatsoever in M.Gajic-M.Pacher, Marianske Lazne 2013.
7 ... 0-0 8 0-0 Nc6 9 Qe2
Once again we would be happy to see 9 f4 Bf5!.
9 ... Nd7 10 f4 Nf6 11 Be3 Bf5

12 Nd2 Bxd3 13 Qxd3 e6 14 Rae1 Rc8 15 Bf2 Ne8 16 Ndf3 Bf6!

Improving over 16 ... Nd6 17 Ng5 Qe8 (taking action against any possible knight sacrifices on the
kingside) 18 Qh3 h6 and here in M.Carlsen-P.Leko, 4th matchgame, Miskolc (rapid) 2008, White should
have played 19 Nexf7!, creating problems.
17 Ng4
17 Ng5 Nd6 18 Qh3 Bxg5 19 fxg5 Nxe5 20 dxe5 Ne4 would be a strategic disaster for White.
17 ... Nd6 18 Nxf6+ Qxf6
A knight is coming to e4.
E2) 6 Bb5+

This is the true point behind 4 Ne5. White gets to keep his knight on e5 without playing the
positionally risky f2-f4.
6 ... Nbd7 7 0-0
White’s attempts to scare Black after 7 Qf3 fail: 7 ... Bg7 8 c4 0-0 9 Bxd7 Nxd7! and already White is
on the back foot. Likewise, 7 Nc3 Bg7 8 Qf3 0-0 9 Bxd7 Bxd7 10 Bg5 Bf5 11 Bxf6 exf6 12 Ng4 Bxg4 13
Qxg4 Re8+ 14 Kf1 Qb6 15 Rb1 f5 was a big disaster for White in J.Broekmeulen-D.Arutinian, Kallithea
2008.
7 ... Bg7

8 Re1
White chooses to reinforce the e5-knight in the most natural way. Otherwise:
a) Retreating the knight is an admission of failure: 8 Nf3 0-0 9 Nc3 Nb8 10 h3 a6 11 Bd3 Nc6 12 a3
Bf5 13 Bxf5 gxf5 14 Ne2 e6 15 Bg5 Qb6 was absolutely fine for Black in M.Carlsen-A.Morozevich,
Nice (rapid) 2008.
b) 8 Nd2 0-0 9 Ndf3 Ne4 10 Bd3 and here Black should make things awkward for White by playing
10 ... Qb6 (instead of the 10 ... Ndf6 of G.Jones-N.Mohota, Canberra 2009) 11 c3 Nxe5 12 Nxe5 Bf5,
with the idea of playing ... Nd6. Note too that 13 f3 is not possible because of 13 ... Bxe5.
8 ... 0-0 9 Nc3 a6 10 Bf1 e6 11 a4
After 11 Nd3 b6 12 Ne2 Bb7 13 c3 although Black’s bishop is somewhat walled in behind the d5-
pawn, the position is easy to play because there are no dangerous pawn breaks for White. Indeed, the
players agreed a draw here in Yu Shaoteng-U.Adianto, Cebu City 2007.
11 ... Nxe5 12 dxe5 Nd7

13 f4
13 Bf4 might be a better shot, although Black now has the time to manoeuvre the knight: 13 ... b6 14
Qd4 Bb7 15 Rad1 Rc8 16 g3 Nb8 17 Qe3 Nc6 18 Bg2 Na5 and Black has the advantage.
13 ... b6 14 Be3 Bb7 15 Qg4 Qe7 16 Bf2?! Qb4!
Now White blundered his e-pawn with 17 Nd1? in T.Schwab-M.Helbig, Bad Godesberg 2011.

F) 2 f4
A rather bizarre move if you ask me.
2 ... d5 3 e5
A different approach is 3 Nf3 dxe4 4 Ng5 Nf6 5 Bc4.
 White is actually willing to sacrifice his queen to obtain a massive initiative, but after 5 ... e6 (the
situation after 5 ... Bg4 is very complicated and I would recommend everyone to steer clear of it: 6 Bxf7+
Kd7 7 Qxg4+! Nxg4 8 Be6+ Kc7 9 Bxg4 gives White too much play) 6 Nc3 b5! 7 Bb3 Be7 8 Ngxe4
(Black should not be afraid of White winning back the pawn as he has saddled himself with a giant
weakness; indeed, the pawn on f4 is much more of a nuisance than an asset) 8 ... 0-0 9 d4? (natural, but
bad) 9 ... b4 10 Nxf6+ Bxf6 11 Ne2 Ba6 12 0-0 c5 Black has a clear edge with White having weakened
his king quite significantly.
Instead, should White try to play like the Grand Prix Attack with 3 Nc3 dxe4 4 Nxe4 Nf6 5 Nf2 (5
Nxf6+ exf6 really poses no problems), I like the Caro move 5 ... Bf5 (it’s not a Sicilian!) 6 Nf3 e6 7 g3 (7
Nh4 weakens the dark squares and Black can reply 7 ... Bc5 8 Nxf5 Bxf2+ 9 Kxf2 exf5 when the white
king will soon find itself under pressure) 7 ... Qb6 8 c3 h6 9 d4 Nbd7 10 Bg2 Be7 11 Ne5 Rd8 12 0-0
Nd5 13 Nfg4 0-0 14 Qb3 Nxe5 15 Nxe5 c5 16 Qxb6 axb6. Black has achieved the desired break and
more or less all we can hope for from the Caro-Kann.
3 ... Bf5 4 Nf3 e6

5 c3
White makes no kingside weaknesses and reinforces the centre. He has also tried:
 a) 5 d4 c5 6 c3 Nc6 is a good version of the Advance French for us: 7 Be3 Qb6 8 Qd2 cxd4 9 Nxd4
Nxd4 10 Bxd4 Bc5 11 Na3 a6 12 Bxc5 Qxc5 13 b4 Qc7 14 c4 Ne7! (Black secures the d5-square for his
knight) 15 cxd5 Nxd5 16 Nc4 0-0 and the game is equal, T.Duran-J.Sykora, Czech League 1998.
b) 5 g3 h5!? (securing the f5-square) 6 h3 gives us a pleasant choice:
b1) 6 ... Be4 7 Bg2 Ne7 8 d3 Bxf3! (knights are better than bishops in closed positions) 9 Qxf3 Nf5
10 0-0 Qb6+ 11 Kh2 Bc5 12 c3 Be3! (Black has the foresight to see that his bishop will become bad
should he retreat, so instead he exchanges it, leaving a closed position with two good knights) 13 b3 Nd7
14 Na3 Bxc1 15 Rfxc1 g5! (breaking up White’s centre) 16 Qe2 gxf4 17 gxf4 0-0-0! and Black has a
stable advantage. His pieces are the better located, the white king is decidedly shaky and, indeed, Black
won quickly in V.Akopian-S.Shevelev, Las Vegas 1993.
b2) 6 ... Nh6 is the same plan, but a different execution: 7 Bg2 (or 7 d4 Be4! 8 Bg2 Nf5) 7 ... Be7 8 0-
0 Bg6 9 d4 Nf5 10 Kh2 c5 11 c3 Nc6 is another fantastic position for Black.
5 ... Nd7 6 h3
6 d4 is a better try, but Black still obtains a very comfortable Advance Caro-Kann. The white pawn
on f4 is really more of a hindrance than a strength and our next move will prevent any further kingside
advances: 6 ... h5! 7 Be3 Nh6 8 Be2 Be7 9 0-0 Qb6 10 Qb3 h4 11 Nbd2 Ng4 12 Bf2 Nxf2 13 Rxf2 h3 14
g3 c5. Black has all the fun in this position now that the ... c6-c5 pawn break has been achieved.
6 ... h5
Of course, g2-g4 must be stopped.
7 d3
After 7 d4 Be7 8 Be2 Bh4+ 9 Nxh4 Qxh4+ 10 Kf1 Bg4 a black knight will arrive on f5 with great
ease and be followed up by a long-term plan of ... c6-c5. To coin a phrase for White, no pawn break, no
plan.
7 ... Be7 8 g3 Nh6 9 Be2 Qb6 10 Na3 f6!

White has played too many pawn moves, so Black can afford to make a pawn break without first
castling.
11 exf6 Bxf6
11 ... gxf6!? was bolder and perhaps better. Here White’s g3-pawn is very weak and after 12 Nd4 0-
0-0 he cannot go pawn hunting: 13 Bxh5? Rdg8 14 Qf3 Nc5 and White’s position quickly becomes
disastrous.
12 Nc2 0-0-0 13 a4 a6 14 a5 Qc7 15 Be3 e5! 16 fxe5?
 White opens up the centre with his king stuck there. This is a bad strategy and one that is punished
ruthlessly.
16 ... Nxe5 17 Bf4
17 Bb6? is not possible in view of 17 ... Nxf3+ 18 Kf2 Qe5 19 Bxd8 h4 and Black is winning.
17 ... g5 18 Bxe5 Bxe5 19 Nxe5 Qxe5 20 Kf2 h4 21 Qg1 Rhf8 22 Ke1 Bxd3 23 Nd4 hxg3 24 Kd2
Bxe2 25 Nxe2 Rf2 26 Re1 Re8
0-1, I.Glek-V.Gunina, Moscow 2013.

G) 2 b3

Garry Kasparov, annotating former world champion Max Euwe’s game against Richard Reti at
Amsterdam 1920, called this line “imaginative but incorrect”. There is a certain truth to this; by not
establishing a strong pawn centre and concentrating on the flank, White does not set Black any opening
challenges.
2 ... d5 3 Bb2
This is the main idea behind the line. White gambits a pawn to obtain fast development and attacking
prospects on the long diagonal.
3 exd5?! concedes the centre to Black and after 3 ... cxd5 4 Bb2 Nc6 5 g3?! (5 Nf3 preventing ... e7-
e5 is a much better move, although White only gets an equal position for his efforts after 5 ... Nf6) 5 ... e5!
(Black logically occupies the centre) 6 Bg2 Be6 7 Qe2? was Euwe’s reckless choice and now Black can
play 7 ... Rc8!, targeting the c2-pawn with a great position.
3 ... dxe4 4 Qe2
Attacking the pawn directly. Should White prefer 4 Nc3 Nf6 Black should hold on to the pawn for as
long as he can, since White must put his pieces on awkward squares in order to regain it: 5 Nge2 Bf5 6
Ng3 e6 (there is no need to retreat the bishop) 7 Qe2 Nbd7 8 Ngxe4 Be7 9 0-0-0 a5!. White’s strategy has
been very committal, so it makes sense for Black to commence a queenside attack. His pieces all point in
that direction and the black king is perfectly happy in the centre for the time being, A.Reprintsev-A.Dreev,
Internet (blitz) 2003.
Instead, with 4 Ne2 the knight heads to g3 and after 4 ... Nf6! 5 Ng3 Nbd7 6 Nc3 Nc5 (Black protects
the pawn with as many pieces as possible as it restricts White’s natural development) 7 Qe2 Bg4 (forcing
White to gambit) 8 f3 exf3 9 gxf3 Bd7 10 0-0-0 Ne6 11 Bh3? (this allows Black to make some exchanges,
although if 11 d4 Nf4! 12 Qe3 Qc7 and Black is ready to castle queenside) 11 ... Nd4! 12 Bxd7+ Qxd7 13
Qf2 0-0-0 14 Nce4 e6 Black has much the better position, A.Bokuchava-B.Gurgenidze, Georgian
Championship 1974.
4 ... Nf6 5 Nc3
The disadvantage with this is that Black can now attack the queen.
5 ... Bg4 6 Qe3 Bf5 7 h3 Nbd7 8 g4 Bg6 9 Bg2 Nd5 10 Nxd5 cxd5 11 f3 e5!?
An interesting try. 11 ... Qc7 is also a candidate move.
12 fxe4 Qh4+ 13 Qf2 Qxf2+ 14 Kxf2 dxe4 15 Re1 Bb4 16 Bc3 Bxc3 17 dxc3 Rc8 18 c4 b5!
Black continues to undermine White’s pawn structure, not allowing him any breathing space, and went
on to win in Tu Hoang Thai-Ni Hua Ho Chi, Minh City 2014.

Summary
I don’t quite understand the point of the Blackmar-Diemer Gambit as Black obtains good development,
although of course we should be mindful of tricks on f7. 4 Nxe4 simply must be a better move, although 4
... Bf5 5 Nc5 is nothing to fear. Here we follow in a strong Russian Grandmaster’s footsteps with 5 ...
Nd7! 6 Nxb7 Qc7 7 Nc5 Nxc5 8 dxc5 Rd8 when we should recover the pawn with good development.
Instead, 2 Ne2 is probably the most random variation in the whole book with the weirdness beginning
when Black plays 3 ... d4!. There are a few patterns to memorize: should White take no action in the
centre then Black should fianchetto on the kingside, moves like Nf4 release the black queen, and 4 b4
allows 4 ... e6 with tempo. A further reminder: Black must always maintain/support the pawn on d4 - do
not let it go!
2 Nc3 d5 3 Qf3 doesn’t really trouble Black as the white queen will stand somewhat awkwardly after
the exchange on e4. Likewise, 2 Nf3 d5 3 exd5 cxd5 4 Ne5 is nothing to fear, although we should
challenge the knight on e5, ideally with ... Nc6 and ... Nfd7, trying to force f2-f4 out of White.
Finally, 2 f4 d5 3 e5 is just a pleasant Advance Caro for Black and 2 b3 looks pretty suspect.