1

Table of Contents

Title page
Bibliography
Introduction

Chapter 1. 1.c4 c5. Symmetrical with ...g6

Main Ideas Step by Step Annotated Games

Chapter 2. 1.c4 c5 2.Nf3 Nf6 3.Nc3 e6

Main Ideas Step by Step Annotated Games

Chapter 3. Symmetrical with ...e6, ...Nxd5

Main Ideas Step by Step Annotated Games

Chapter 4. Symmetrical with ...b6

Main Ideas Step by Step Annotated Games

Chapter 5. Maroczy Bind

Main Ideas Step by Step Annotated Games

Chapter 6. Anti-Grünfeld
Main Ideas Step by Step Annotated Games

Chapter 7. Anti-Nimzo
Main Ideas Step by Step Annotated Games

Chapter 8. 1.c4 Nf6 2.Nc3 e6 3.e4

Main Ideas Step by Step Annotated Games

Chapter 9. Anti-Slav and Chebanenko

Main Ideas Step by Step Annotated Games

Chapter 10. Anti-Meran

Main Ideas Step by Step Annotated Games

Chapter 11. Anti-Queens’s Gambit

Main Ideas Step by Step Annotated Games

Chapter 12. Odds and Ends

Main Ideas Annotated Games

Index of Variations

2
 The Modern English

Volume 2: 1.c4 c5, 1...Nf6, 1...e6

Chess Stars Publishing

www.chess-stars.com

Copyright © 2019 by Kiril Georgiev and Semko Semkov

Cover by Kalojan Nachev, Semko Semkov

3
 Bibliography

Books

The Full English Opening, Carsten Hansen, New in Chess 2018

The English, David Cummings, Everyman Chess 2016
e3 Poison, Axel Smith, Quality Chess 2016
Mastering the Chess Openings, Volume 3, John Watson, Gambit Publications 2008
The Tarrasch Defence, Nikolaos Ntirlis and Jacob Aagaard, Quality Chess 2011
The Hyper Accelerated Dragon, second extended edition, Raja Panjwani, Thinkers Publishing 2018
The Modern Reti: An Anti-Slav Repertoire, Alexander Delchev, Chess Stars 2012
The Hedgehog vs the English/Reti, Igor Lysyj/Roman Ovetchkin, Chess Stars 2017
Beating Minor Openings, Victor Mikhalevski, Quality Chess 2016

Periodicals

Chess Informant
New in Chess Yearbook

Internet resources

Databases
The Week In Chess (www.theweekinchess.com)
Chess Publishing (www.chesspublishing.com)

4
 Introduction

Many grandmasters claim that the best

answer to 1.c4 is 1...e5. If you consider only the
chances this move offers for complex fight for all
three results, that may be true. However, for a
theoretician, the symmetric answers always present
the biggest challenge. They are notoriously not just
“balanced”, but often bluntly drawish.

I devoted two thirds of the book on the

Symmetrical 1.c4 c5. While plans with further ...g6
or ...b6 generously cede space to White and were
fun to explore, the most restraining scheme with
2.Nf3 Nf6 3.Nc3 Nc6 4.e3 e6 5.d4 d5 was the
toughest nut to crack.
5...Bxf3 6.Qxf3 cxd5 7.Bb5+ Nc6 8.Bxc6+
bxc6 – Chapter 9.
White’s task is more complicated in the
Meran set-up. Five years ago the plan with g2-g4
inspired big hopes, but today we know that it is
effective only against ...b6. That should not sound
discouraging, as the quiet development Be2 is not
deprived of venom.

Our Anti-Grünfeld Chapter 6 offers several

very different systems.
In my youth I was strongly influenced by the
instructive game Karpov-Kortchnoi from the world
title match in Merano 1981:
1.c4 Nf6 2.Nc3 d5 3.cxd5 Nxd5 4.Nf3 Nxc3
5.bxc3 g6 6.d4 Bg7 7.e3 c5 8.Bb5+ Nd7 9.0-0 0-0
In Chapter 2 I analyse 6.cxd5 exd5 7.Bb5, 10.a4 a6 11.Bd3 b6 12.Rb1 Bb7 13.e4 Qc7 14.Re1
where we generally learn how to play around an e6
isolated and hanging pawns. In Chapter 3 the roles
are reversed – 6...Nxd5 7.Bd3 puts us on the other
side of the barricade, but the extra tempo does
matter. White has several ways of obtaining a lasting
initiative. The good news is that Black does not have
forced draws and should work strenuously to keep
the balance.

The other symmetric in spirit system is the

Slav. One of the most attractive features of the
English Opening is that it assures White of an
initiative after 1...c6 2.Nf3 d5 3.e3 Nf6 4.Nc3 Bg4
5.cxd5!?

15.e5 h6 16.h4 with a nice clear plan. I even

followed in Karpov’s footsteps 20 years later against
5
Krasenkow. However, since then Black has them! They often offer examples and explanations
discovered that he should play ...Bd7 instead of which you would not find in the “Step by Step”
...Nbd7, with decent chances to equalise. sections.

Therefore, our main dish is the hot line 1.c4 Kiril Georgiev
Nf6 2.Nc3 d5 3.cxd5 Nxd5 4.Nf3 c5 5.e4!? (instead March 2019
of the above-mentioned 5.e3 g6 6.Bb5+)

We have discovered several interesting ideas

to make Black work hard over the board.

The Anti-Nimzo approach again gives you a

choice. If you are a hardcore fan of the Carlsbad
structure in the QGD, you should choose 1.c4 Nf6
2.Nc3 e6 3.e4 – Chapter 8. Although we hide some
trumps up our sleeve, I believe that 3...d5 should be
balanced.
AlphaZero’s choice is 3.Nf3! Bb4 4.Qc2,
and it is the subject of Chapter 7.

You might ask, what about 3...d5? Well, my

personal opinion is that the best you can do is to
enter the main lines with 4.d4. However, throughout
the chapters on the Symmetrical English we often
had to deal with e3, followed by a queen’s
fianchetto. So we added Chapter 11, where we
discuss the Anti-QGD approach.

The final chapter called “Odds end Ends”

briefly cover our strategy against the King’s Indian,
the Modern Defence and the Dutch. In short, my
advice is: go into the main lines! I only explain how
to reach them through the English move order 1.c4
Nf6 2.Nc3.

When reading the book, please do not skip

the “Main Ideas” sections, but always start with

6
 Chapter 1. 1.c4 c5. Symmetrical with ...g6
Main Ideas

1.c4 c5

The symmetrical openings have proved to be

the most reliable choice against 1.e4 and 1.d4. 1.c4
c5 is no exception – White has been struggling to
get even the slightest edge. Black’s position is
incredibly flexible, and we should think about at
least five possible set-ups: with ...d5, with ...g6, with
...e6, with ...e5, and with ...b6.
Any cunning move order from us, devised This position has been tested in thousands of
against a particular variation, could stumble into a games, and Black has established long equalising
counter-exploitative strategy from Black. lines where memory is the decisive factor. The main
line for decades had been:
Here are a couple of simple examples. If a) 6.g3 Qb6! 7.Nb3
White played 2.g3 to discourage the hedgehog with 7.Ndb5 Ne5 8.Bg2 a6 9.Qa4 (9.Na4!?)
...b6, he should surrender the centre after 2...Nf6 and 9...Rb8 10.Na3!? (10.Be3 Bc5 11.Bxc5 Qxc5
3...d5. Or if he put in d4 first, then Black could take 12.Qa3 b6 13.Nd6+ Ke7 14.Qxc5 bxc5 15.Nxc8+
and follow up with ...e6. The situation is even more Rhxc8 16.b3 d5 17.cxd5 c4 18.b4 Rxb4=) 10...Bxa3
complex, as Black could delay ...c5 and return to it 11.Qxa3 Nxc4 12.Qb3 Qc7 13.0-0 b5 14.a4 0-0
later. For instance: 1.c4 Nf6 2.Nc3 c5 – we have 15.axb5 axb5 16.Bf4 d6 17.Qb4 e5 18.Bg5 Qc5
already committed our knight to c3. Or 1.c4 Nf6 19.Qxc5 dxc5 20.b3=.
2.Nc3 d5 3.cxd5 Nxd5 4.Nf3 c5!? 5.e3 Nxc3!? 7...Ne5 8.e4 Bb4 9.Qe2 d6÷
6.bxc3 g6, when we get a timid version of the
Grünfeld with e3.
Whatever we settle for should consider all
those transpositions.

A little background

At first I was not sure on what system to

build my repertoire. Then I started from the opposite
– I made a list which lines I wanted to avoid. For
instance:
1.c4 c5 2.Nf3 Nc6 3.d4 cxd4 4.Nxd4 Nf6
5.Nc3 e6
This may not be too clear, but Black scores
Critical position 1 more than 50% here, and the best I could offer for
White is the following draw: 10.Bd2 0-0 11.0-0-0 a5
12.f4 a4 13.fxe5 axb3 14.a3 dxe5 15.axb4 Ra1+
16.Nb1 Qa6 17.Qd3 Qa2 18.Bg2 Re8 19.Bc3 Re7
20.Kd2 Rd7 21.Na3 Rxd1+ 22.Rxd1 b5 (22...Rxd3+
23.Kxd3) 23.c5 Rxd3+ 24.Kxd3=.
Having also toyed with 6.Bf4 d5 and 6.Ndb5
d5, White eventually looked at:
7

b) 6.a3.
Now apart from 6...d5 7.cxd5 exd5 8.Bg5
(8.g3 Bc5; 8.Bf4 Bc5) 8...Bc5 9.e3 Bxd4=, Black
also employs:
6...Be7 (6...Bc5!? 7.Nb3 Be7) 7.e4 0-0 8.Nf3
Qa5 9.Bd2 Qh5 10.Be2 d5. This sucks all the life
out of the position, e.g. 11.exd5 exd5 12.cxd5
Nxd5=.
Would you like to have something like that
as a main White repertoire?!
We face a similar situation after 1.c4 c5
2.Nf3 Nf6 3.Nc3 Nc6 4.g3 d5 5.cxd5 Nxd5
Critical position 2
Again, White cannot even reach 50%. Of
course it is a sharp, unbalanced position, where
White plays for all three results, but the second
players have learnt which models to follow. Besides,
it is easier to hold the balance with more space.
8
The latest trend for White is to seek new
ways, and the main direction is to fianchetto the
queen’s bishop, while developing the other one on
e2 or b5. That is also perfectly consistent with our
repertoire against 1...e5, which I proposed in the
first volume.
We play e3
The best move order is: 2.Nf3!, passing the
ball into Black’s court.
His main plans are:
with ...e6, ...d5 – see Chapter 2/3.
with ...b6 – see Chapter 4;
with 2...g6 3.d4, going for the Maróczy
Bind;
with ...g6 and ...d5, which is the Panov
Attack.
The latter set-up is the subject of this
chapter.
2...Nf6 (2...Nc6 3.e3!) 3.Nc3!
It is still premature to commit ourselves with
3.e3 as Black could then opt for an excellent version
of the KI by 3...g6!.
3...Nc6 4.e3! This is already okay as we’d
have d2-d4-d5 against the KI set-up.
4...g6 5.d4 cxd4 6.exd4 d5 7.cxd5 Nxd5
8.Qb3!
 It is vital to provoke ...e6. Black cannot 12.Qc2 is more popular, but we need the
avoid it since 8...Nxc3 is strongly met by 9.Bc4! queen on the kingside. We intend to trade
Nd5 10.Bxd5 e6 11.Bxc6+ bxc6 12.0-0 Qd5 dark-squared bishops with 12...Bg7 13.Ba3 Rc8
14.0-0 Bf8 15.Bxf8 and organise an attack with h4.

13.Qe3! White’s play in this line is entirely

based on a long-term dark square strategy which
should ultimately give White an attack on the enemy
king. The character of play is similar in the
alternative line:
8...e6 9.Bb5 Bg7 10.0-0 0-0 11.Bxc6 bxc6
12.Re1!? – see Game 1 Giri-Mamedyarov, Wijk aan
Zee 2018.
Perhaps the most topical line is 8...Nxc3
9.Bc4! e6 10.bxc3 Bd7 11.Bd3 Na5, where I
recommend 12.Qd1!

9
 Chapter 1. 1.c4 c5. Symmetrical with ...g6
Step by Step
1.c4 c5 2.Nf3
I do not see any advantages in playing 2.Nc3
first.
a) After 2...Nf6 you should transpose with
3.Nf3 anyway.
I was tempted by 3.e3, planning to meet
3...g6 by 4.g3!? Bg7 5.Bg2 0-0
6.Nge2 Nc6 7.d4 cxd4 8.exd4 d6 9.0-0.
White has prevented ...d5 and enjoys a space
advantage.
However, 3.e3 defines too early White’s
plan. Black could push 3...d5, restricting White’s
choice to 4.cxd5 Nxd5 5.Nf3. That is balanced after
5...Nxc3 6.bxc3 g6.
b) One of the biggest drawbacks of 2.Nc3 is
2...g6. Then the tempo spent on Nc3 would be
important if we chose to push d4:
3.Nf3
3.e3 is always premature in KI set-ups when
Black has not committed his knight to c6. The
reason is that 3...Bg7 4.Nf3 Nf6 5.d4 0-0 gives
White the “poor man’s Benoni” as Axel Smith calls
it – some may find it playable, but I’m on the
opposite opinion, and practice confirms my distrust.
3...Bg7 4.d4 cxd4 5.Nxd4 Nc6 6.e3 Nf6
7.Nc2 d6 8.Be2 Bf5!, as mentioned in the comments
to Black’s 6th move of the main line.
The other popular approach with g3 is also
10
ineffective when Black has not spent a tempo on an
early ...d6:
3.g3 Bg7 4.Bg2 Nc6 5.e3 e5
The point is that White’s most challenging
plan with 6.a3 Nge7 7.b4
Or 7.Rb1 a5! (7...d6 8.b4 0-0 9.d3 would
transpose to the famous game Kasparov-Karpov,
Seville 1987, which I analysed in volume 1 – see
Game 23, page 186.) 8.Nge2 0-0 9.d3 d6 10.0-0
Rb8=.
7...cxb4 8.axb4 Nxb4 9.Ba3 Nec6=, e.g.
10.Qb1 a5 11.Nd5 Ra6.
c) After 2...Nc6 White does not achieve
anything with 3.e3 e5 4.Nge2?! (4.Nf3! transposes
to 2.Nf3 Nc6 – see below, while 4.g3 transposes to
the above-mentioned line b), owing to 4...Nf6! 5.d4
exd4 6.exd4 d5.
2...Nf6
After 2...Nc6 we can already play 3.e3!
3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 e6 throws us
out of our repertoire.
3...e5
3...g6 4.d4 Bg7 (4...cxd4 5.exd4 Bg7 6.d5
Ne5 7.Nxe5 Bxe5 8.Bd3 d6 9.0-0 Nf6 10.Nc3 0-0
11.Re1 Bd7 12.h3 a6 13.Bf1 b5 14.Bh6²) 5.d5 Na5
6.e4² with more space.
3...Nf6 4.Nc3! transposes.
4.Nc3

a) 4...Nf6! 5.d4 e4 6.d5 offers White the
better pawn structure and a modest, but stable plus.
See more details in Game 3 Mamedyarov-Vachier
Lagrave, Biel 2018.
b) 4...f5 5.d4 (5.Be2 is also interesting) 5...e4
Now both 6.Ng1!? and 6.d5 are pleasant for
White. See Game 2 Buhmann-Lupulescu, Kings
Medias 2015.
2...g6 3.d4!, aiming for the Maróczy Bind, is
covered in Chapter 5. In this move order the plan
with 3.e3 is innocuous as Black simply plays the
Benoni and we do not have d4-d5 with a tempo.
2...b6 is covered in Chapter 4
2...e6 is the subject of Chapter 2/3.
3.Nc3
It is essential to await Black to define its plan
11
before committing ourselves with e3. 3.e3 could
face 3...d5 4.cxd5 Nxd5, when 5.Bb5+ Bd7 is
balanced, and 5.b3 Nc6 6.Bb2 Ndb4! is even worse
for White. 5.Nc3 Nxc3! 6.bxc3 g6 is currently fine
for Black.
Besides, 3.e3 g6 is a perfect version of the
Benoni for Black.
3...Nc6
We deal with 3...g6 4.d4 (or 4.e4, followed
by 5.d4) 4...cxd4 in Chapter 5. White also has 4.e4!?
3...e6 4.e3 transposes to Chapter 2/3.
3...d5 4.cxd5 Nxd5 5.e4 is another critical
lines. I cover it in Chapter 6 /line B.
4.e3
The older and more popular plans with 4.d4
cxd4 5.Nxd4 e6 and 4.g3 d5 have been deeply
analysed, and Black has pinpointed well known
equalisers. For instance, the latter line might
continue:
5.cxd5 Nxd5 6.Bg2 Nc7 7.a3 g6 8.d3 Bg7
9.Be3 Ne6 10.h4 h6 11.Rc1 Bd7 12.0-0 Rc8= or:
5.d4 e6! 6.cxd5 Nxd5 7.Bg2 cxd4 8.Nxd4
Nxc3 9.bxc3 Nxd4 10.Qxd4 Qxd4 11.cxd4 Bd6
12.0-0 Rb8 13.e4 b6=.
4...g6
The key of our design is the stab 4...d5
5.cxd5 Nxd5 6.Bb5. It practically forces 6...Nxc3
(6...g6 7.Qa4 Nxc3 8.Bxc6+²) 7.bxc3 Bd7 8.0-0 e6
9.d4 Be7, when we get a mobile pawn centre
without allowing exchanges. All that said, White
must play energetically, with the break d4-d5 in
mind. Otherwise Black will get counterplay on the
queenside. A good example is the game
Ganguly-Xiong, Saint Louis 2018. Note that Xiong
is the highest rated fan of this variation, and it was
his third try in less than an year. No doubt that both
opponents were heavily prepared at home:
a) 10.Rb1!? (to prevent ...Ne5) 10...0-0
11.Re1 Qc7 12.e4 Rfd8 13.g3 Bf6 14.Bf4 Qa5
Now the bold 15.d5! exd5 16.exd5 Bf5
17.Bc4 Qxc3 18.Rc1 Qb2 19.d6!! would have put
Black to the wall.
The immediate threats are Bb3 or Qd5. For
instance: 19...h6 20.Bb3 Bd4 21.Nxd4 Qxd4 22.Qh5
c4 23.Qxf5 or:
19...Re8 20.Qd5 Be6 21.d7 Re7 22.Rxe6
Rxe6 23.Rd1!
12
Now 23...b6? loses to 24.Ne5! Bxe5
25.d8+=Q, while 23...Qc2 24.Ne5! Nxe5 25.Bxe5
Rd8 26.Bc7 also favours White. Best defence is:
23...h6! 24.Qf5! (or 24.Qxc5 Qc3 25.Rd3
Qc2 26.Kg2 with an initiative) 24...g6 25.Qd3 Nb4
26.Qf1 Re4 27.Be5! regaining the material with
interest.
b) The earlier game L’Ami-Xiong, Wijk aan
Zee 2018, went:
10.e4 0-0 11.Be2
The set-up with Bd3 looks more natural, but
in some lines Black might have ...Nb4.
11...Qc7 12.Be3 Rfd8 13.Rc1 Rac8. Now
White’s plan is d4-d5, but perhaps he could
postpone it and play 14.g3 first, having in mind:
14...Be8 15.d5 or 14...Na5?! 15.Bf4 Bd6
16.e5 Be7 17.d5. Strongest reply is:
14...cxd4 15.cxd4 Qa5 16.Qb3 Qb4 17.Qd3.
We keep the queens and watch for d4-d5.
4...e6 5.d4 is covered in Chapter 2 and 3.
5.d4 cxd4
6.exd4
This recapture leads to an isolated pawn,
which is strategically a double-edged asset.
However, 6.Nxd4 Bg7 7.Nc2 is less promising.
This line aims to hinder ...d5, which was
possible after 7.Be2. The idea is to reach the
Maróczy Bind even at the cost of another tempo:
7...0-0 8.Be2 d6 9.0-0
 a) 9...Bf5! 10.Nd4 (10.f3 Bxc2) 10...Nxd4 We have considerably more space. More
11.exd4 d5 12.c5 Ne4 13.Nxe4=. importantly, every exchange will underline the
weakness of the central pawn island. The computer
b) 9...Nfd7 10.Bd2 (preventing doubled may be confident in Black’s resources, but I would
pawns after ...Bxc3) 10...a5 is passive. We could certainly prefer White here. An illustrative line is:
connect the knights with Nc2-a3-b5 and establish a 16...Bxc2 17.Rxc2 Nd4 18.Rc1 e6 19.Be3
firm grip on the centre with f3. Nf5 20.Bf2 Bh6 21.Rb1 d5 (21...Ne3 22.Qxd6 Rc6
23.Qd3 Nxf1 24.Kxf1²) 22.Qb3 Be3 23.cxd5 Nd4
c) 9...Be6 10.e4 24.Qa2 Bxf2+ 25.Rxf2 e5 26.Qb2 Nf6 27.Bd3 Nxd5
If Black does not generate counterplay 28.Nxd5 Qxd5 29.Be4². The d3-knight looks great,
quickly, we’ll enjoy a nice space advantage. For but its stand could be undermined with f4.
instance, the plan with ...b5 is ineffective: 10...Rc8
11.Bd2 a6 (11...Ne5 12.b3) 12.Rc1 Ne5 13.b3, when 6...d5 7.cxd5 Nxd5
13...b5 is a bluff – 14.cxb5 axb5 15.Bxb5 Qb6
16.Be2². (Mind the trap 16.Qe2? Rxc3! 17.Bxc3
Bg4.) More energetic is the idea with ...f5:
10...Nd7 11.Bd2 Nc5
A typical manoeuvre to attack e4. 11...Rc8
could be met by 12.Kh1.
12.f3 f5 13.b4 Nd7 14.Rc1 Rc8
The tempting 14...f4 leaves us with a stable
centre. After 15.Kh1 we could organise the break g3
with the help of Rg1 or Qe1.
15.exf5 Bxf5 16.a3

This position closely resembles the Panov

Attack in the Caro-Kann, but the inclusion of Nf3
Nc6 significantly changes the subsequent play.

8.Qb3 (8.Bc4 Nb6∞) 8...Nxc3

After 8...e6 9.Bb5 Bg7 10.0-0 (10.Bg5 Qb6

11.Qa3 f6=) 10...0-0 11.Bxc6 bxc6 Black plays for
13
two results only. Although White achieves more
than 60% with 12.Na4 and 12.Ne4, I believe that
White’s chances to win are slim after them:
12.Na4 Ba6! 13.Re1 Nb6! 14.Nc5 Bc4
15.Qc3 Bd5. Black has activated his only passive
piece and should hold the draw.
12.Ne4 prevents 12...Ba6 13.Re1 Nb6
(13...Qb6 14.Qd1) in view of 14.Bg5, but allows
12...Qb6! 13.Re1 Qxb3 with nice drawing chances. I
came to the conclusion that White should anticipate
...Ba6 with:
12.Re1!?, renewing the positional threat of
Nc3-a4-c5.
The point of this move order is that in the
event of 12...Ba6 we’ll send the knight rightwards –
13.Ne4, denying the manoeuvre ...Ba6-c4-d5. See
more details in Game 1 Giri-Mamedyarov, Wijk aan
Zee 2018.
9.Bc4! e6
9...Nd5 10.Bxd5 e6 11.Bxc6+ bxc6 12.0-0
Qd5 has been tried lately.
14
Four factors work in White’s favour here:
Black’s split pawns; his lag in development; his
weak dark squares; the poor prospects before the
c8-bishop. Our first task should be to discourage
...c5 and keep the queens at the same time. So we
go:
13.Qe3!. The idea is that:
a) 13...c5 14.dxc5 Qxc5 opens dangerously
the position and Black cannot find time to castle
after 15.Qf4 Qf5
15...Bg7 16.Bd2 a5 (16...0-0 17.Bb4)
17.Rac1.
16.Qc4 Qc5 17.Qh4 Bg7 18.Bg5 Bb7
19.Rac1 Qb5 20.a4 (20.Be7 Bxf3 21.Bc5) 20...Qa5
(20...Qf5 21.Qb4) 21.b4 Qf5 22.Rc5 Qe4 23.Qxe4
Bxe4 24.Rc7 0-0 25.b5±. Black has finally
connected his rooks, but at a high price – White will
soon have a passed pawn on b6, supported by a
bishop from e3.
b) 13...Bb7 pursues the same goal. 14.Bd2 c5
15.dxc5 Bxc5 16.Qc3, when 16...0-0 drops an
exchange.
c) The attempt to complete development
with 13...Bg7 14.Bd2 0-0 was a positional disaster
after 15.Rfc1 Rd8 16.Rc5 Qd6 17.Ba5±, Wang
Hao-Zuo Yifan, Hangzhou 2018.
d) 13...f6 is the most popular move. Black
plugs the dark-squared holes, planing to lead out his
bishop to d6 or e7. We have 14.Nd2!
 19.Rab1 Kc6 20.Bc5 Rcb8 21.Kd3 Kd5 22.Nd2.

11.Bd3

The game Svidler-Shankland, Hoogeveen

2018, saw 11.Bb5!? a6
Perhaps 11...Bg7 12.Ba3 Ne7 holds on –
13.Rb1 b6 14.Be2 Bc6 15.0-0 0-0.
12.Be2 Na5 13.Qb1 Rc8 14.0-0 f6
14...Bg7 15.Ba3 Rxc3 (15...Bf8 16.Bb4)
16.Qb2;
14...Qc7 15.Ne5 Bg7 16.Bf4.
15.c4 b6 16.c5! with an attack.
The knight heads for e4 to prevent ...c5.
After 14...Bg7 we develop our bishop with 11...Na5 12.Qd1
15.b3, while:
14...Be7 15.Ne4 0-0 16.Re1 Bb7 17.Nc3
defines a target on e6. The queen protects the f3-knight against
...Bd7-c6xf3.
10.bxc3 12.Qc2 is more frequent, but Black obtains
good play by killing the f3-knight:
12...Bg7 13.Ba3 Bc6
13...Bf8 14.Bxf8 Kxf8 15.0-0 Rc8 16.Rfe1
Kg7 17.Qb2 Re8 18.Ne5 is slightly better for White.
Then 18...f6 would be a serious concession –
19.Nxd7 Qxd7 20.h4 Nc4 21.Qb4 a5 22.Qb3 Qc6
23.Rab1 Re7 24.Be4².
14.Qe2 Rc8 15.0-0 Bxf3!? 16.Qxf3 Bf8
17.Bb5+ Nc6

10...Bd7

Threatening ...Na5. The alternative is

10...Bg7 11.Ba3 Na5
11...Bf8 is premature as Black deprives
himself of the possibility to eat by the bishop the
white knight should it leap to e5. White gets an edge
with 12.Bxf8 Kxf8 13.h4 Na5 (13...h6 14.Qb2) Black is lagging behind in development, but
14.Qb4+ Kg8 15.h5. I did not find anything concrete:
12.Bb5+ Bd7 13.Qb2 Qb6 14.Bxd7+ Kxd7
15.Qe2 Qa6 16.Qxa6 bxa6 17.Ke2. White has an a) 18.c4 Bxa3 19.Qxa3 Qxd4! (19...a6
extra “half-pawn” and a clear idea to trade the 20.Bxc6+ Rxc6 21.Rab1²) 20.Rab1 Qd7 21.Rfd1
enemy knight with Nd2, e.g. 17...Rhc8 18.Bb4 Nc4 Qe7 22.Qxa7 0-0 23.Bxc6 Rxc6 24.Rxb7 Qf6.

15
White has won a pawn, but the game should be
drawn – 25.Qd4 Qxd4 26.Rxd4 e5 27.Re4 Ra8.
b) 18.Bb2 Bg7 19.Rfb1 0-0 20.Ba3 Re8
21.Bc5 e5=.
c) 18.Bc1 Bg7 19.a4 0-0 20.Ba3 Re8
21.Rfb1 Qf6 22.Qxf6 Bxf6 23.Bc5 Red8 (23...b6)
24.Bxc6 b6!=.
12.Qb1 Bg7 13.h4 Rc8 14.h5 seems
promising at a glance, but the “dumb” 14...Qb6 is a
cooler.
12...Bg7
12...Rc8 13.0-0 Bg7 transposes, while
13...Rxc3? would be suicidal owing to 14.Bb2 Rc8
15.d5+–.
13.Ba3 Rc8
It is dangerous to trade bishops before White
castled – 13...Bf8 14.Bxf8 Kxf8 15.h4!, provoking
15...h5. Then White’s knight obtains the g5-square
for an attack. Also the g6-pawn could be a target –
16.Ng5!².
13...Bc6 is based on the correct
understanding that White’s most dangerous piece is
the f3-knight, and it should be killed – 14.0-0 Bf8
15.Bxf8 Bxf3 16.Qxf3 Kxf8
Our best chance here is to play for an attack
with 17.Qf4.
Instead 17.Be4 Rc8 18.Bxb7 Nxb7 19.Qxb7
16
Kg7 20.Rfc1 Rc7 21.Qf3 Qc8 wins a pawn, but it
would be difficult to convert it. I’d rather sacrifice
one:
17...Kg7 18.Qe5+ Kg8 19.h4 h5 20.Qg3 Qb8
21.f4 Rh7 22.d5!? exd5 23.Qe3 Nc4 24.Bxc4 dxc4
25.f5ƒ.
14.0-0!
14...Bf8
14...Rxc3?! 15.Bb4 Rc8 16.d5 rips the centre
wide open – 16...Bf8 17.a3 Bxb4 18.axb4 Nc4
19.Qe2 Nb6 20.dxe6 Bxe6 21.Bb5+ Kf8 22.Qe5.
14...Nc4 is also dubious – 15.Bxc4 Rxc4
16.Qb3 Qc8 17.Bc5.
14...Bc6!? 15.Ne5 Bxe5 16.dxe5 gives White
an initiative, although it might be impossible to
extract something tangible.
15.Bxf8 Kxf8 16.Rc1
16.h4 without the rook support is not a threat
anymore – Black could simply ignore it.
16...Kg7 17.Re1
White’s pawn centre is a good base for a
kingside activity, e.g. 17...Re8 18.Re3 Nc4 19.Bxc4
Rxc4 20.Ne5 Ra4 21.Qb3 f6 22.Nxd7 Qxd7 23.h4ƒ.
 Chapter 1. 1.c4 c5. Symmetrical with ...g6
Annotated Games

1. Giri – Mamedyarov
Wijk aan Zee 21.01.2018

1.c4 c5 2.Nf3 Nc6 3.Nc3 g6 4.e3 Nf6 5.d4

cxd4 6.exd4 d5 7.cxd5 Nxd5 8.Qb3 e6 9.Bb5 Bg7
10.0-0 0-0 11.Bxc6 bxc6 12.Re1

You get the idea. White is not rushing, as his

strategic pluses are stable – Black cannot find an
employment for his c4-bishop. 24...Bd3 25.Rbc1
Re8 26.Ne5 Bxe5 27.Rxe5 Qb8 28.Qe3 Bc4 29.Re1
Qc7 30.Bc5 Rb8 31.h4 h5 32.Qh6 a5 33.Rg5±.
14.Qd1 Ba6 15.Na4 Qb5 16.Nc5

12...Qd6?!

An amazing positional mistake for a player

of the rank of Mamedyarov. I guess he did not pay
enough attention to White’s last move and followed
the home brewed scheme devised against 12.Na4,
when Black often answers 12...Qd6. The only
drawback of 12.Re1 is that it allows:
12...Qb6 13.Qc2 (13.Ne4 Qxb3 is close to a
draw) 13...Nb4
13...Nxc3 14.bxc3 Qa5 15.Bf4 will most
likely lead to opposite-coloured bishops which
would only facilitate White’s attack. For example:
15...Ba6 16.Bd6 Rfd8 17.Bc5 Bd3 18.Qd2 Rd5
19.h3 Rad8 20.Bb4 Qb5 21.Rad1 Bc4 22.Rb1 Rb8
23.a4 Qb7 24.Qf4
We enter a “tunnel variation” now: 16...Rfd8
17.a3 (17.Be3 Bf8) 17...Nd3 18.a4 Qc4 19.b3 Qc3
20.Bd2 Bxd4 21.Nxa6 Qxa1 22.Qxa1 Bxa1 23.Rxa1
c5 24.b4². This should be defendable, but as long as
White keeps the rook, he retains practical chances.
After the immediate 12...Nxc3 13.bxc3 Ba6
14.Ba3 Re8 15.Ne5 Qd5 White could even leave his
queen on b3 – 16.Bc5 Bxe5 17.Rxe5 Qc4 18.h4 with
pressure.

Black has tried lately 12...Qc7!?.

His idea is to meet 13.Na4 or 13.Ne4 by
13...c5! 14.Nxc5 Bxd4 15.Nxd4 Qxc5. The game
Malakhov-Cheparinov, Zagreb 2018, went further
16.Nf3 (16.Rd1 is more testing, but Black holds
17
after 16...Rd8) 16...Qb6 17.Qxb6 axb6 18.Bd2 f6
19.Nd4 e5 20.Nc2 Bb7 21.Nb4 Rfc8 draw.

13.Bd2 Rb8 14.Qa3 Qe7 15.Qxe7 Nxe7

16.b3 Nf5 17.Rad1 Nxd4 led to another short draw –
18.Nxd4 Bxd4 19.Ne4 c5 20.Bf4 Rb7 21.Bd6 Rd8
22.Be5 Rf8 23.Bd6 Rd8 24.Be5 Rf8 25.Nf6+ Kg7
26.Ne8+ Kg8 27.Nf6+ Kg7 28.Ne8+ Kg8 29.Nf6+
½-½, Gledura-Safarli, Brest 2018. Perhaps the most
testing answer to 12...Qc7 is Matlakov’s:
13.h3!?
Bai Jinshi replied 13...Rb8 14.Qd1 c5?! (Or
14...Rd8 15.Ne4 h6 16.Bd2 Kh7 17.Rb1 Rb5 18.a4
Rb8 19.b4², Cheparinov-Alekseev, rapid, St
Petersburg 2018) 15.Nxd5 exd5, when the simple There is no future for the poor bishop on c8.
16.dxc5 Qxc5 17.Be3 would have been in White’s The only way to activate it is ...f6+...e5, but it would
favour. lead to yet another broken link in Black’s pawn
The computer suggests: chain.
13...Nxc3 14.bxc3 c5 15.Ba3 c4
16...Re8 17.Ne5 f6 18.Nc4 Qc7 19.Bd2 e5
20.dxe5 fxe5 21.Qa4

Some pawn will fall by itself, like a ripe

fruit.

21...Nb6 22.Nxb6 Qxb6 23.Qc4+ Kh8

24.Bc3

This allows 24...Qb5. Of course the endgame

I prefer White here because he has a clear
dark square strategy to follow: 16.Qb5!? Bd7
17.Qg5 or 16.Qb2 Rb8 17.Qe2 Rd8 18.Bc5 Ba6
19.h4.
is difficult, but I would prefer to keep the queens
here. 24.b3, intending a4, or 24.Be3 Qb5 25.Qc3
look more dominant.
24...Bf5?! 25.g4! Bc8 26.Re4

13.Ne4 Qb4 14.Qc2 a5 15.a3 Qb6 16.Nc5± Now it is curtains.

26...Qb5 27.Rae1 Qxc4 28.Rxc4 Rb8 29.h3

h5 30.gxh5 gxh5 31.Rh4 Kh7 32.Ne4 1-0

The threat Re3-g3 decides the game.

2. Buhmann – Lupulescu
Kings Medias (5.4), 03.11.2015

18
 1.Nf3 c5 2.c4 Nc6 3.Nc3 e5 4.e3 f5 14.b3 favours White.
13.0-0 Bb7 14.Qc2 0-0-0 15.Rd1 Qc7
16.Bd2 Bd6 17.Nf4. White’s king is safer.

5.d4 e4 6.d5

This is a semi-automatic reaction in such

positions, and it scores more than 61%. However,
the resulting endgame is difficult to win due to
White’s static doubled pawns. If you are looking for
sharper play, I would recommend the more
flexible:
6.Ng1!? Nf6
6...d6 7.Nh3 Nf6 8.Nf4 g6 9.dxc5 is awful
for Black;
This position is much more common without The insertion of 6...cxd4 7.exd4 is dubious –
the inclusion of e3 and c5. White’s main plan 7...d6 (7...Nf6? 8.d5 Ne5 9.d6!± a6 10.c5 g6 11.Bf4)
remains the same – to push d2-d4. The only problem 8.Nh3 Be7 9.Be3 Bd7 10.g3².
is where to retreat the f3-knight and whether to
retreat it at all. A decent possibility is to play first:
5.Be2.
The idea is to castle and meet ...e4 by Ne1.
Since 5...g6 6.d4! e4 7.Ng1 d6 8.dxc5 dxc5 9.Qxd8+
Nxd8 10.b3 Bg7 11.Bb2 Nf7 12.h4² is stably better
for White owing to the hole on d5, Black should try:
5...e4 6.Ng1 Nf6

White has a very nice choice. He could take

on c5 at once, or postpone it with 7.Nge2. For
instance: 7...d5 8.Nf4! dxc4 9.Bxc4 cxd4 10.exd4
Qxd4 11.Qb3 with fantastic piece activity.

6...exf3 7.dxc6 dxc6 8.Qxd8+ Kxd8 9.gxf3

Nf6 10.b3

7.d3
7.Nh3 d5 (7...Bd6 8.d4) 8.cxd5 Nxd5 9.0-0
Nxc3 10.bxc3 Be7 11.f3 exf3 12.Bxf3 Ne5 13.Be2
looks also in the spirit of the English, but Black
could sac a pawn – 13...c4! 14.Bxc4 Nxc4 15.Qa4+
Qd7 with some compensation.
7...d5 8.cxd5 Nxd5 9.Nxd5 Qxd5 10.Nh3
exd3 11.Nf4 Qd7 12.Nxd3 b6
Preparing to castle long. 12...Bd6 13.0-0 0-0
19
 18.Ne2².
16.Nxe4 Nxe4 17.Bxe4 g6 18.h4 Rf8 19.h5
Bf5 20.hxg6 Bxg6 21.Bxg6 hxg6 22.Rd3² and Black
cannot untie the bishop.
In most events White should prepare e4 with
Ne2, h4, Rhe1, Ng3, probably f4 and f3.

14.h4 g6 15.Ne2 Rhf8 16.Kc2 Rd7 17.Rd2

Rfd8 18.Bc3 Bc7 19.Ng3

Black is now faced with an important choice

– where to develop his king’s bishop. 10...g6 11.Bb2
Bg7 12.0-0-0+ Ke7 dooms him to a gloom defence
as White’s knight will be in full command on e5
after an exchange of the dark-squared bishops.
Therefore, Black should keep his bishop.

10...Bd6! 11.Bb2 Ke7 12.Bd3 Be6 13.0-0-0

Rad8

19...f4?
It is more natural to put the other rook on d8.
Let’s consider 13...Rhd8 14.Kc2
This move drops material. More solid was
19...b6 20.Rhe1 a6. Black prepares counterplay with
...b5 at a suitable moment. (20...Kf7 21.h5±). Then
21.e4 should be already good – 21...f4 (21...fxe4
22.Bxf6+) 22.Nf5+ Kf7 23.Nh6+ Ke7 24.Rg1 Nh5
25.Ng4 Kf8 26.Rgd1 with a passed pawn.

20.Bxf6+ Kxf6 21.Ne4+ Kg7 22.Nxc5 Re7

23.Nxb7 Rf8 24.exf4?!

24.e4 Bc8 25.Nc5 was simpler.

24...Bf5 25.Bxf5 Rxf5 26.Nd6 Rxf4 27.Ne4

Perhaps this is a critical position for the Rxf3 28.Rd3
6.d5-line. White’s main advantage is his more active
pieces, so he should aim to open the central files.
Moreover that the enemy king is on e7. However, an White has remained with only one extra
early e4 would be good only if Black committed a pawn, but it turned enough for the point:
tactical mistake. For instance, 14...Rd7 (14...g6! was
correct) allows 15.e4! fxe4
The exchange of the e6-bishop leaves a weak
pawn on f5 – 15...g6 16.exf5 Bxf5 17.Bxf5 gxf5
20

28...Rf5 29.f3 Kh6 30.b4 a5 31.a3 axb4
32.axb4 Be5 33.Ra3 Ref7 34.Ra5 Bd4 35.Nd6
Rxf3 36.Nxf7+ Rxf7 37.Rd1 Rf4 38.Ra6 c5 39.b5
Rxh4 40.b6 Be5 41.Kb3 Rh2 42.Ra2 Rh3+ 43.Ka4
g5 44.Rd5 Bf4 45.Rxc5 Rh1 46.Rf5 Rb1 47.Rxf4
gxf4 48.c5 Kg5 49.Ka5 Rc1 50.b7 1-0
3. Mamedyarov–Vachier-Lagrave
Biel (3.2), 24.07.2018
1.Nf3 c5 2.c4 Nc6 3.Nc3 e5 4.e3 Nf6 5.d4
5...e4
A major alternative is 5...cxd4 6.exd4 e4
7.Ne5 Bb4 8.Be2 Qa5
Actually, the rare 8...Qe7 is more solid –
9.Bf4 Bd6 (9...0-0 10.0-0 Bxc3 11.Nxc6 dxc6
12.bxc3 Qa3 13.Qb3) 10.Nxc6 (10.Bg5!? Nxe5
21
11.Nd5 Qd8 12.dxe5 Bxe5 13.0-0 looks
intimidating, but 13...d6 14.f4 exf3 15.Bxf3 Be6
16.Qa4+ Kf8 17.Rad1 Qc8 keeps Black on the
surface.) 10...dxc6 11.Qd2 0-0 12.Bxd6 Qxd6 13.h3!
White is planning Qe3, Ra1-d1-d2, with the idea of
d4-d5.
9.0-0
We get more than enough compensation for
the pawn after 9...Nxe5 10.dxe5 Qxe5 11.Bd2 Bc5
12.Nb5 0-0 13.b4 Be7 14.f4! Qe6 15.Be3 Ne8 16.f5
Qf6. Two correspondence games saw here 17.Rb1
b6 18.Bd4², while 17.Qb3!? b6 18.Rad1² looks more
natural to me.
6.Ne5
This leads to sharper play, but I recommend
the quiet:
6.d5! exf3 7.dxc6
7...fxg2
7...dxc6 8.Qxd8+ Kxd8 9.gxf3 Kc7
(Ponomariov believes that the king should stay on
e7 – 9...g6 10.b3 Bg7 11.Bb2 Ke7, but then the
c6-pawn will be weak. White retains a pull after
12.Ne4 b6 13.Bg2 Rd8 14.Be5 Be6
15.Ng5 Ne8 16.f4².) 10.b3 Be7 11.Bb2 Rd8
12.h4! g6 13.Ne4 Nxe4 14.fxe4 h5, when best is
15.Be2². White’s advantage in this pawn structure is
that he could produce a passed pawn on the kingside
by pushing e4-e5, e3-e4, f2-f4-f5. The game
Aronian-Vachier Lagrave, Karlsruhe 2017, does not
need comments: 15.Be5+ Bd6 16.Bf6 Re8 17.f3 a5
18.a4 Be7 19.Bc3 Be6 20.0-0-0 b6 21.Be2 Red8
22.Rdg1 b5 23.Kc2 b4 24.Be5+ Kb7 25.Bf4 Rd7
26.e5 Bf8 27.Rd1 Rxd1 28.Rxd1 Bg7 29.Bd3 Rd8
30.Bg5 Rd7 31.f4 Bg4 32.Be2 Bf5+ 33.Kc1 Kc7
34.Bf3 Bf8
35.e4 Bh3 36.f5 gxf5 37.Bxh5 fxe4
38.Rxd7+ Kxd7 39.Bxf7 Bg7 40.e6+ Bxe6
41.Bxe6+ Kxe6 42.Kd2 1-0.
MVL also tested 7...bxc6 8.Qxf3 d5 9.cxd5
Bg4, when strongest is 10.Qg3 Bd6 11.f4 cxd5
22
12.Nxd5, eating a pawn (Ponomariov).
8.cxd7+ Bxd7 9.Bxg2 g6 10.b3 Bg7 11.Qd6
Qb6
12.Qe5+!? Kf8 13.Bb2 Nh5 14.Qd5 Bc6
15.Qd6+ Kg8 16.Bxc6 Rd8 17.Qe7 bxc6 18.Rd1
Rxd1+ 19.Kxd1 Bf6 20.Qe4 Kg7 21.Kc2 Qc7 22.h3
Ponomariov suggests 22.Qg2!? Rd8 23.Rd1².
22...Rd8 23.Qf3 Kg8
In 2017 MVL played against Grischuk in
blitz 23...Be5 24.Rd1 Rxd1 25.Nxd1 Nf6 to draw
eventually.
Ponomariov offers 26.Bc1!? Qd6 27.Nb2
Nd7 28.Nd3 Ba1! with chances to defend.
24.Rd1 Rxd1 25.Nxd1 Be7. All this occurred
in Dominguez Perez-Vachier Lagrave, rapid Saint
Louis 2018. White has a small, but stable edge in a
safe position. He could manoeuvre his d1-knight to
d3 via b2 after 26.Bc1, and the c1-bishop could
place a second hit on c5 from a3. That makes an
exchange of the queens problematic for Black.
6...g6
 Besides the text, Black has successfully
tested 6...Be7 7.Be2 0-0 8.0-0 Re8. Now, instead of
9.b3, I propose to display activity on the queenside
with 9.a3 d6 (9...Bf8 10.Ng4) 10.Nxc6 bxc6 11.b4.
7.g4
7.Be2 Bg7 8.Nxc6 dxc6 9.dxc5 is parried by
9...Qa5!. That’s why White has tried 7.Rb1,
intending to eat a pawn with 7...Bg7 8.Nxc6 dxc6
9.dxc5 Qxd1+ 10.Kxd1. However, I do not like his
passive position after 10...0-0 11.h3 Re8 12.Bd2
Nd7 13.Na4 a5. Black will regain the pawn
following ...Bf8.
7...h6 8.Bg2 Bg7 9.h3
White awaits the opponent to castle. It would
be premature to take on e4 – 9.Nxe4 Nxe4 10.Bxe4
Qe7 with an initiative.
9...Qe7
9...0-0 10.Nxe4 Nxe4 11.Bxe4 was already
possible. Black’s king would be a target after
11...Qe7 12.Bd5 Nxe5 13.dxe5 Bxe5 14.e4.
10.b3 cxd4?!
The critical position for Mamedyarov’s idea
arises after 10...Nb4! 11.a3 Na6 12.0-0 d6
23
13.f4 0-0 14.Ra2. We have a dynamic
balance. Black is not obliged to take the knight. He
does not risk anything if he simply stays, e.g.
14...Nc7 15.Rd2 a5 16.Bb2 Ra6 17.Qe1 g5 18.d5
Nce8 19.Na4 Nh7 20.Qg3 Nhf6 21.Rdf2 Qc7. Still,
I would prefer his game, as White should constantly
calculate the consequences of ...dxe5. This capture
may not be in Black’s favour in most lines, but it
does not lose either. It just requires more responsible
moves. Here are a couple of illustrative lines:
15.Raf2 Nh7 16.Nxe4 dxe5 17.dxe5 Rd8
18.Qc2 Bd7 19.Bb2 Bc6 20.f5! gxf5 21.Ng3 Rd7
22.Bxc6 bxc6 23.Nxf5 Qe6
White still has a draw – 24.Nxg7 Kxg7
25.Rf6 Nxf6 26.Rxf6 Rad8 27.Bc3 Rd1+ 28.Kh2
R1d3 29.Rxe6 Nxe6 30.Qc1 Rd1 31.Qb2 R1d3=.
Or 15.a4 Rd8 16.Ba3 dxe5 17.fxe5 Nxg4
18.hxg4 Qh4 19.Nxe4 Bxe5 20.Qf3 Qh2+ 21.Kf2
Qh4+=.
The verdict is that the knight sac is correct,
but it does not promise White much. I think it was a
one-game-novelty from Mamedyarov. But it paid off
nicely, as we’ll see below.

11.exd4

White has won a healthy pawn, and he has

fair chances to convert it.

21...h5? 22.gxh5 gxh5 23.h4 Rac8 24.Bf3

Rxe1+ 25.Rxe1 Kf6 26.d6 b6 27.Re7 Rd8 28.Bxh5
11...Nxe5?! Be6 29.Rxa7 Rxd6 30.Kg2 Rd8 31.Bf3 Ke5
32.Ra4 1-0

This is too greedy. 11...0-0 12.Bf4 Nd8 was

unclear.

12.dxe5 Qxe5 13.Bb2 d5 14.Qe2

MVL said at the press conference that he

missed this move, but 14.Rb1 d4 15.Nb5 and even
14.Qd2 e3 15.fxe3 Qg3+ 16.Qf2 are also better for
White.

14...Qe7 15.cxd5 0-0 16.0-0 Re8 17.Rfe1

Bd7 18.Nxe4 Nxe4 19.Bxg7 Kxg7 20.Qxe4 Qxe4
21.Bxe4

24

Chapter 2. 1.c4 c5 2.Nf3 Nf6 3.Nc3 e6
Main Ideas
This chapter deals with probably the most
challenging set-up against the English Opening in
Black’s possession.
1.c4 c5 2.Nf3 Nf6 3.Nc3 e6
Black intends to meet 4.g3 by 4...b6, and
4.d4 cxd4 5.Nxd4 gives him a wide choice of decent
options. It suffices to say that five (!) of them yield
more than 50%. My proposition is:
4.e3. It considerably narrows the path before
Black.
4...d5 5.cxd5 exd5
Black chooses to play with an IQP. 5...Nxd5
25
reverses the roles – see Chapter 3.
6.d4
In my opinion 6.Bb5+ Bd7 simplifies the
enemy’s task, although White retains slight pressure.
In fact, this is not an English Opening
anymore. This position is well known from the
Semi-Tarrasch Defence. Black must now define his
plan about the future of his dark-squared bishop. It
is clear that it cannot remain for long on c5 after
dxc5.
One option is to preserve it on the a7-g1
diagonal, from where it would support the
equalising break ...d5-d4:
I believe that chances are even in this
scenario. Instead I suggest to fianchetto our king’s
bishop:
6...a6 7.g3!?
Now the isolated pawn loses most of its
attraction as the white king is well protected and the
g2-bishop is perfectly targeted on d5. The only
reasonable plan is ...c4 now or on the next turn.
Then White plays Nf3-e5xc6 and b3, with pressure
on the queenside without any risk.

The main line is 6...Nc6, preparing to shift

the c5-bishop to the b8-h2 diagonal. It is
double-edged as Black concedes the blockading
square d4, but gathers more forces against our king.
We answer 7.Bb5!

10.Nf3!. We attack d5 and exploit the

hanging state of the Bd6. Thus 10...a6 11.Bxc6 bxc6
would face 12.e4!, while 11...Bxc6 offers us an
excellent version of an IQP where the black bishop
is passive on c6.

We’ll often have to play against hanging

pawns on c6, d5 or c5, d5.
Positional goal 2
Note that we take on c6 only if Black spends
a tempo on ...a6. Otherwise we use the pin to force
the passive ...Bd7, and take on c6 by knight. Our
positional goals are:
Positional goal 1

It is impossible to win c6 against good

defence. In most cases our steps should be:
To force the enemy’s piece to passive places.
To execute the break in the centre e3-e4. It
will widen the scope of our pieces so they could be
We aim to saddle the opponent with an quickly transferred to the kingside.
isolated pawn, block it safely, and trade at least one Here are several examples:
minor piece. We can then transform this structure by
producing hanging pawns. Games 4, 7, 8 present a Analysis
good illustration of White’s plan.
Pay attention to the line 7...cxd4 8.Nxd4 Bd7
9.0-0 Bd6
26
 20.e4!? Bb5 21.e5 Qd8 22.Rh4ƒ. 18.Qe2 c4 19.g3 Qa5 20.Bc3 Qb5 21.Bxf6
gxf6 22.Nc3 Qa5 23.bxc4 dxc4². Black may have a
Giri-L’Ami, passer, but White’s pieces are clearly more stable.
Amsterdam 2015
Theoretical status

Most grandmasters believe that Black’s

set-up with an isolated pawn which arises in line B
is a solid equaliser against our plan with e3. Still,
with very precise move orders, White can reach
pleasant technical positions where he could squeeze
without any risk.

Of course White can win the c6-pawn here,

but Giri preferred to mount the tension with 25.e4!?
Qd6 26.e5.
A slightly improved version of the same idea
would be 25.Kh1 Bc8 26.e4.

Our primary goal should be to provoke one

of the c5-d5 pawns to move forth. That reduces their
dynamic and automatically gives us an easy game
since our pieces obtain more squares.

Kramnik-J.Polgar
Geneve 2013

27
 Chapter 2. 1.c4 c5 2.Nf3 Nf6 3.Nc3 e6
Step by Step
1.c4 c5 2.Nf3 Nf6 3.Nc3 e6
Black commonly plays 3...Nc6 first, to make
White show his hand. We answer 4.e3 e6 5.d4,
transposing. The move order with 3...e6 encourages
4.g3, but that would mean to fight with a timid
set-up against the Hedgehog 4...b6, which is not my
intention.
4.e3
Sometimes White tries 4.e4, hoping to get
the Maróczy Bind after:
4...d6?! 5.d4 cxd4 6.Nxd4 a6.
The known equaliser to 4.e4 is:
4...Nc6 5.Be2
5.d4 cxd4 6.Nxd4 Bb4! 7.Nxc6 bxc6 8.Bd3
e5! is in no way better for White.
28
5...d5!
5...e5 6.d3 d6 7.0-0 Be7 8.Ne1! 0-0 9.f4 exf4
10.Bxf4 offers White lasting pressure on the
kingside. The plan with a3+b4 is less promising.
6.exd5 exd5 7.d4 cxd4 8.Nxd4 dxc4 9.Nxc6
Qxd1+. White achieves amazing results after all 3
recaptures, but that is mostly due to Black’s wrong
defensive set-ups. For instance, 10.Nxd1 bxc6
11.Bxc4 Nd5 12.Ne3 Bb4+ 13.Bd2 Bxd2+ 14.Kxd2
Be6 15.Bxd5 0-0-0 16.Rhc1 Bxd5= is equal, as well
as 10.Bxd1 bxc6 11.Bf3 Bb7! – the Black king goes
to the queenside to defend both weaknesses, e.g.
12.0-0 0-0-0 13.Be3 Kb8 14.Na4 Nd5 15.Rfd1 Bd6
16.Bd4 f6=.
4...d5
We must also be prepared against the rare
option 4...a6.
The aim of this bizarre move is to avoid Bb5
after ...d5. The first players answer 5.d4 d5 6.cxd5
exd5 7.Be2 Nc6 (7...c4!?) 8.0-0 Bd6 9.dxc5 (9.b3 is
premature on account of 9...cxd4 10.Nxd4 Qc7)
9...Bxc5 10.b3 0-0 11.Bb2 Ba7 12.Rc1

We have reached by transposition one of the
main position of the Semi-Tarrasch. In practice
White often does not know what to do from here.
Look at Game 4 Rybka 3-Chessmaster 11,
CCRL 2008, for a detailed analysis. However, I
propose for a main weapon the surprising:
5.g3!?. It is extremely rare in modern chess
to see a completely new move at such an early stage.
Its advantage over 5.d4 is that White’s play here is
more straightforward, and he denies the opponent
any attacking possibilities.
5...d5 6.cxd5 exd5 7.Bg2 Nc6
7...d4? 8.exd4 cxd4 9.Qa4+.
8.0-0 Bd6 9.d4 0-0
9...c4 is thematically met by 10.b3 cxb3
11.Qxb3 Na5 12.Qb1 0-0 13.a4 Re8 14.Qa2 Bb4
15.Bd2 Be6 (or 15...Nc6 16.Qb1 Rb8 17.Rc1²)
16.Ne2 Bd6 17.Ng5. White has pressure down the
b-file, e.g. 17...Nc4 18.Nxe6 fxe6 19.Rfb1 Qd7
20.Bb4 Bxb4 21.Rxb4 Rac8 22.Nf4 Qe7 23.Nd3².
Here both 10.b3 cxd4 11.Nxd4 Bg4 12.Nxc6
bxc6 13.Qd3 and 10.dxc5 Bxc5 11.Ne2 Bg4 12.h3
29
Bf5 13.b3 give White a pleasant game. The bishop
stays very well on g2 against the isolated pawn.
We can meet 4...Nc6 5.d4 cxd4 6.exd4 d5 by
7.c5
7.cxd5 exd5 is less interesting, although
White scores pretty well.
a) 7...Be7 8.Bb5 0-0 9.0-0 Ne4 10.Re1
White can also take a tiny, but lasting edge
in knight vs bishop positions after 10.Bxc6 bxc6
11.Qc2 Nxc3 12.Qxc3 Qc7 13.g3 (13.Bg5) 13...a5
14.Bf4 Qb7 15.a4 f6 16.Rfe1².
10...Nxc3 11.bxc3 Bd7 12.Bf4 b6 13.cxb6
axb6 14.Re3
White’s pieces are dangerously pointed at
the black king. The game Ruck-A.Horvath, Austria
2005, went 14...Ra3 (14...h6 15.Bd3 Be8 16.Qd2)
15.Bxc6
A positional approach. It is sharper to keep
the bishop for an attack – 15.Qb1!?, (intending Bd3
or Bxc6) 15...Qa8 16.Ng5 g6 17.h4.
15...Bxc6 16.Ne5 Be8 17.Qg4
 17.Rg3! Bh4 18.Rh3 Bg5 19.Qc1².
17...Bf6 (17...f5! 18.Qh3) 18.Rg3
18.Rh3 Bxe5 19.Qh5 h6 20.Qxe5 f6
21.Qxe6+ Bf7 22.Qd6 wins a healthy pawn.
18...Qa8. Here 19.Bh6! obtains the upper
hand since 19...Rxa2 20.Rc1 Ra1? unexpectedly
loses to 21.Qd1 Rxc1 22.Qxc1 g6 (22...Kh8 23.Ng4)
23.Ng4 Be7 24.h4+–.

b) 7...Ne4 8.Qc2 Nxc3

8...f5 only temporary retains the the knight
on e4 as White will easily repel it by f3 – 9.Bb5 Be7
10.Ne5 Bd7 11.Bxc6 Bxc6 12.Bf4 (or 12.0-0 Bf6
13.Nxc6 bxc6 14.Ne2) 12...g5 13.Nxc6 bxc6 14.Be5
0-0 15.0-0-0!².
8...Qa5 only gives tempi on b4 – 9.a3 Be7
10.Rb1 Nxc3 11.b4².
9.Qxc3 Be7
The so called dark square strategy in action.
White’s pieces are more mobile.
17...Bb5 18.Rac1 Bc4 19.Ra3 Be7 20.Ne5
Bb5 21.Qg3².
4...b6!? is a clever way to reach Hedgehog
structures. For instance:
5.Be2 Be7 6.0-0 0-0 7.d4 cxd4 (7...d5 8.cxd5
exd5 9.b3 Bb7 10.Bb2 Nbd7 11.Ne5 is an anti-QGD
line) 8.Nxd4 Bb7 9.b3 a6
(9...d5 10.cxd5 Nxd5 11.Nxd5 Bxd5 12.Bb2
Nc6 13.Nxc6 Bxc6 is “almost” equal, but not a draw
– 14.Rc1 Bd5 15.Bf3 Bxf3 16.Qxf3 Rc8 17.Qg4 g6
18.Rxc8 Qxc8 19.Rc1 Qd8 20.Qd4 f6 21.Rc6 Kf7
22.Kf1 Qxd4 23.exd4²) 10.Bb2 d6 is a much
improved version of the main Hedgehog lines.

10.Bb5
White cannot keep his material advantage
after 10.b4 e5 11.Nxe5 Nxe5 12.dxe5 a5 13.Bb5+
Bd7 14.Bxd7+ Qxd7 15.bxa5 0-0.
10...0-0 11.0-0 Bd7 12.Bf4 a5 13.Rfc1!.
Aimed against ...b6.
13.Rfe1 b6! 14.Rac1 bxc5 15.Bxc6 Bxc6
16.dxc5 Rc8 17.Nd4 Bd7 18.c6 Bb4 19.Qe3 Bxe1
20.Qxe1 Re8 21.Qe5 Re7 22.cxd7 Rxc1+ 23.Bxc1
Qxd7 is hardly any better for White.
13...Re8 14.h3 Bf6 15.Bxc6 Bxc6 16.Qe3 a4
17.Rc3 Nonetheless, White can grab space with
11.f4!?, intending g4 or f5, and if 11...Nc6, then
12.Nc2.

A more practical approach is to transpose to

line B2 from the next Chapter 3 by:
30
 5.d4 cxd4 6.exd4 Bb7 7.Bd3 Be7 8.0-0 d5 A. 6.Bb5+ Bd7
9.Qe2

6...Nc6 7.d4 transposes to line B2.

This move order could be used to sidestep
the line 6.d4 a6 (which hardly should be avoided
anyway), were not for 6...Bd7.
Instead 7.Ne5 Qc7 8.Qa4 is probably
balanced:

9...0-0 (or 9...Nc6 10.cxd5 Nxd5 11.Rd1)

10.Rd1 Nc6 11.cxd5.

5.cxd5 exd5

5...Nxd5 – see the next chapter.

a) 8...Qxe5 9.Bxc6+ Bd7 (9...Kd8 10.Qa5+!)
10.Bxd7+ Nxd7 11.0-0 Bd6 12.f4 (12.g3 Qe6
13.Qb3 Nf6 14.Qb5+ Kf8ƒ) 12...Qe6 13.Qb3.
Although Black’s pawns are hanging, the poor
bishop on c1 gives Black enough counterplay –
13...Nf6 14.Qxb7 0-0 15.Qa6 Rfb8 16.b3 c4 17.Ba3
Bxa3 18.Qxa3 d4 19.exd4 Qd7 20.d5 cxb3 21.axb3
Nxd5 22.Qa4 Qe6=.

7.Bxd7+ Qxd7 8.0-0 Nc6

A. 6.Bb5+; B. 6.d4

A very important position arises after 6.b3

Nc6 7.Bb2. It is often reached through the
anti-Queen’s Gambit move order 1.Nf3 Nf6 2.c4 e6
3.e3 d5 4.b3 c5 5.Bb2 Nc6 6.cxd5 exd5 7.Nc3 or
similar. My opinion is that the early fianchetto does
not offer White any benefits after 7...Bd6 – see
Game 6 Mamedyarov-Aronian, blitz, Saint Louis
2018. Instead 7...d4 8.Na4 gives White some
practical chances – see Game 5 Kurylo-Crielesi,
ICCF 2018. 9.d4

31
 It is tempting to hamper ...c4 by 9.b3!?.
Then the lame 9...Be7 10.Bb2 0-0 11.d4
Rfd8 12.dxc5 Bxc5 13.Rc1 Qe7 14.Nb5 and 9...Bd6
10.Nb5 0-0 11.d4 Be7 12.dxc5 Bxc5 13.Bb2 Ne4
14.Rc1 Rfd8 15.Qe2 Rac8 16.Rfd1 Be7 17.Nbd4
give White the desired IQP position with a firm
control of d4. The problem lies in:
9...d4 10.Ne2 dxe3 11.dxe3 Be7 12.Qxd7+
Nxd7 13.Bb2 0-0 14.Rfd1 Rfd8 with gradual
equalisation.

9...c4

14.b3 cxb3 15.axb3 Bd6 16.Ba3

9...Be7 10.dxc5 Bxc5 11.b3 0-0 12.Bb2
should be pleasant, as well as 9...Ne4 10.Qa4.
White has only a marginal edge.
10.Ne5

B. 6.d4!?
It makes sense to damage Black’s queenside
pawn structure before pushing b3.
10.b3!? Bb4 11.Bd2 is double-edged, but it
is the best practical chance to impose sharp play.

10...Qe6!

10...Qc7 allows 11.e4ƒ, the point being

11...dxe4? 12.Bf4.

11.Qa4

11.e4 dxe4 12.Nxc6 bxc6 13.Bg5 Be7 is

already unimpressive.
We have a further split here:
11...a6 12.Nxc6 Qxc6 13.Qxc6+ bxc6 B1. 6...a6; B2. 6...Nc6

6...c4!? 7.Be2 Bb4 8.Bd2 is complex and

totally unexplored. White’s plan is b3, e.g. 8...0-0
9.b3 cxb3 10.Qxb3 Nc6 11.Ne5 Bd6 12.f4 with
dominance in the centre.

6...Bd6 7.dxc5 Bxc5 8.Bb5+ transposes, but

White has also tested successfully 8.Bd3 0-0 9.0-0
Nc6 10.h3, intending b3, Bb2.

B1. 6...a6
32

7.g3!?
This is less efficient here than after 4...a6.
The difference is that after ...c4 Black has the
possibility to develop his bishop to b4 instead of d6.
However, it still brings White more than 71%.
Perhaps Black players are unprepared for the
fianchetto. Aagaard/Ntirlis do not even mention it in
their book on the Tarrasch.
7.Be2 leads to Game 4 Rybka
3-Chessmaster 11 CCRL 2008, where Black’s
bishop remains active on the g1-a7 diagonal. That
provides him with good chances to equalise by ...d4.
7.b3 Nc6 8.Bb2 cxd4 9.Nxd4 Bd6 10.Be2
0-0 11.0-0 Re8 12.Nxc6 (12.Rc1 Nxd4 13.Qxd4
Be5 14.Qd2 d4 15.exd4=) 12...bxc6 13.Rc1 Bd7
leads to a typical position with hanging pawns. I
prefer to avoid it, as Black has chances for an attack.
7...Nc6
7...c4 8.Bg2 Nc6 9.Ne5 transposes to 8...c4.
Only 8.Ne5 b5 (8...Nc6) is of independent
significance. Hebden-B.Lalic, Hastings 2018, went
9.Bg2 Bb7 10.0-0 Nbd7 11.b3 b4? 12.Nxd7 Qxd7
13.Na4 Qb5 14.e4±. An obvious improvement is
11...Bd6 with tangled play. Perhaps simpler for
White would be 11.e4 dxe4 12.Qe2 Be7 13.a4 0-0
14.axb5 axb5 15.Rxa8 Qxa8 16.Nxd7 Nxd7
17.Nxb5.
8.Bg2 Be7
33
8...c4 is less popular, but it is more
challenging:
9.Ne5 Bb4
9...Be7 does not hamper b3. White was
clearly better after 10.0-0 0-0 11.b3 cxb3 12.Nxc6
bxc6 13.axb3, although the engine Gull 3 held the
draw against Texel 1 in 2014.
10.Bd2 0-0 11.0-0
11...Re8
11...Qd6 12.a3 Bxc3 13.Bxc3 b5 14.f3,
followed up by g4, hands White the initiative.
12.Nxc6 bxc6 13.b3. White has a structural
advantage and went on to win both tournament
games I found in my database:
13...a5 14.Qc2 Ba6 15.Rfd1 Rb8 16.bxc4
Bxc4 17.Na4 Ne4 18.Bxe4 Rxe4,
Iturrizaga-Perunovic, Lisbon 2017, when best is
19.Nb2 Be2 20.Rdb1 Bxd2 21.Qxd2 Bf3 22.Nd3
with a clear positional advantage. Quite curiously,
White’s knight is stronger than the daunting bishop
on f3.
13...Bg4 14.Qc2 Bxc3 15.Bxc3 cxb3
16.axb3², Tomashevsky-Borisek, rapid, Doha 2016.
9.dxc5 Bxc5 10.0-0 0-0
 I consider this version of an IQP position as
pleasant for White since ...d4 is not an imminent
threat and the g2-bishop safely defends the king.
11.b3 Be6
11...Bf5 is hardly any better. The simple
12.Na4
12.Bb2 Qd6 13.a3 is a decent alternative –
13...Rfe8 14.b4 Ba7 15.Ne2 Be4 16.h3 h6 17.Rc1
Qe7 18.Qb3 Rac8 19.Rfd1².
12...Ba7 13.Ba3 Re8 14.Bc5 Bxc5 15.Nxc5
Qe7 16.Rc1, Grachev-Dragnev, Minsk 2017,
consolidated to a small edge.
12.Bb2 Qe7 13.Ne2 Rac8
Emms made two draws from here, but I do
not think we should reconsider the classical chess
understanding about blocked isolated pawns. White
can count on lasting pressure. He could always
transform it in a bishop pair advantage with
Ne2-f4xe6, but perhaps he should take control of the
queenside dark squares first:
14.Nf4
Of course 14.Ned4 is not bad, but many
decades ago GM Bent Larsen observed that isolated
pawns should be taken, not blocked!
14...Rfd8
34
15.a3
15.Nxe6!? fxe6 16.a3 was already possible.
Hebden-Emms, Southend 2008, saw 15.Nd4
Nxd4 16.Bxd4 h6 (16...Bxd4 17.Qxd4 Qc5²) 17.Qd3
(17.Rc1!) 17...Bxd4 18.Qxd4 Qc5 19.Rfd1 Qxd4
20.Rxd4 Rd7 21.Rad1 Rc2, when 22.Nxd5 wins a
pawn, although Black has good chances to draw 3:4
pawns on the same wing after 22...Nxd5 23.Bxd5
Bxd5 24.Rxd5 Rxd5 25.Rxd5 Rxa2 26.Rd8+ Kh7
27.Rd7 b6 28.Rb7².
15...Ba7
15...a5 cedes the b5-square – 16.Nxe6 fxe6
17.Nd4 Bb6 18.Qe2 Rc7 19.Qb5.
16.Qd3².
B2. 6...Nc6 7.Bb5
This stab wins a tempo on development since
Black must put his bishop on the passive square d7.
B21. 7...a6; B22. 7...Bd6; B23. 7...cxd4
 B21. 7...a6 8.Bxc6+ bxc6 9.0-0 Bd6 10.dxc5
Bxc5

15...Ne4 16.Nb7 Qc7 17.Nxd6 Qxd6

18.Nd4. White has traded 2 minor pieces and has no
fear of a kingside attack anymore.
The computer finds a curious way of
11.b3 pursuing a repetition of moves: 13...a5 (instead of
13...Re8) 14.Na4 Ba6 15.Re1 Bb4 16.Bc3 Ba3
17.Rc2 Ne4 18.Bb2 Bb4. Fortunately, we have
11.e4 is commonly met by: 19.Qd4 f6 20.Rec1 c5 21.Qd1 c4 22.Nd4
11...0-0 12.Bg5 Be7, when after 13.exd5!
cxd5 14.Re1 h6 15.Bh4 Re8 16.Nd4 Bd7 17.Qd3
Rb8 18.Rab1² White’s d4-knight dominates the
Black light-squared bishop.
The acid test should be:
11...Nxe4 12.Nxe4 dxe4 13.Qxd8+ Kxd8
14.Ng5 Kc7 15.Bf4+ Kb6 16.Nxe4, as played in a
game between engines. I believe that Black should
draw after the blunt 16...Bf5 17.Nxc5 Kxc5, despite
White’s activity: 18.Bc7 Rhc8 19.Rfc1+ Kd5
20.Rd1+ Kc5 21.Rac1+ Kb5 22.Bd6 g6 23.a4+
Kb6=.

11...0-0 12.Bb2 Bg4

12...Bd6 is 3 times less popular, probably
because 13.Rc1 Re8 14.Na4 Bd7 15.Nc5 offers
White a good version of the hanging pawns.
White looks a little cramped, but in fact his
pieces have better prospects – 22...Re8 23.Nc3 Bxc3
24.Bxc3 Nc5 25.Rb1 Nd3 26.h3!² (the threat Nc6 is
stronger than its execution!). The knight on d3 is not
attacking anything, while Black has two weak
pawns, a5 and d5.

13.Rc1

13.Ne2 Bd6 14.Ne5 is also attractive as it

immediately forces exchanges. However, Black’s
rook is still on the a-file, and he could get
counterplay with 14...Bxe2 15.Qxe2 Qc7 16.Nf3
35
Ne4 17.Rac1 a5.

13...Bd6 14.Ne2!

The more minor pieces we trade, the weaker

are the hanging pawns. Have in mind that a 4-rook
endgame in this structure is practically lost for
Black! No wonder White has scored the
overwhelming 87% after
14...Rc8 15.Ne5. Look at Game 9
Giri-L’Ami, Amsterdam 2015. More natural is:

25...Red5 26.Kf1 Rxc2 27.Rxc2 Rd1+

28.Kg2 f5 29.Nc3 Rd7 30.Na4².

B22. 7...Bd6 8.dxc5 Bxc5 9.0-0 0-0 10.b3

Bg4 11.Bb2

14...c5 15.h3

15.Ng3 Bxg3 16.Bxf6 Qxf6 17.hxg3 Qd6 is

also somewhat better for White, e.g. 18.Rc4 Bd7
19.Re1 Rac8 20.e4 Bb5 21.e5 Qd8 22.Rh4.

15...Bxf3 16.gxf3 Re8 17.Bxf6 Qxf6

18.Qxd5 Rad8 19.f4 Bxf4 20.Qf3 Be5 21.Qxf6 11...Rc8
Bxf6 22.Rc2

11...a6 12.Bxc6 transposes to 7...a6.

Black’s split pawns cause him trouble in the
endgame. He does not seem to have a straight
equaliser – 22...Re5 23.Rfc1 Be7 24.Rc4 Rd2
25.R1c2
11...Bd6 is a positional achievement for us –
now the equalising ...d4 is no longer a threat. We
must now trade as many pieces as we can:
12.Be2 a6 13.Nd4 Bd7
Or 13...Be6 14.Nxc6 bxc6 15.Na4.
14.Bf3 Be6 15.Nxc6 bxc6 16.Rc1 Rc8
17.Na4², see Game 10 Kramnik-J.Polgar, Geneve
2013.

12.Rc1

36
 We should not allow Black to keep his
bishop on the a7-g1 diagonal as after 12.Be2 a6
13.Rc1 Ba7.

12...Bd6

The point is that 12...a6?! drops a pawn to

13.Nxd5! (Karjakin missed this hit in blitz against
Topalov).

13.Be2

The inclusion of 13.h3 Bh5 only helps Black
to avoid the exchange of his bishop for Be2:
14.Be2 Bb8 15.Nh4
15.Nb5!? Ne4 16.Nd2 Bxe2 17.Qxe2 Qa5
forces 18.Nxe4 dxe4. Black got rid of the isolator
and is only slightly worse.
15...Bg6! 16.Nxg6 hxg6. Without the
f3-knight White cannot prevent ...d4. His bishop pair
is not worth much in such positions.
13...Bb8
This is Black’s most aggressive plan.
13...Re8 14.Nd4 Bxe2 15.Ncxe2 strengthens
the blockade on d4. After 15...Ne4 White has a
pleasant choice between 16.Ng3 and 16.Nf5 Be5
17.Bxe5 with further exchanges of minor pieces.
13...a6 14.Nd4 Bxe2
The retreats look inconsistent:
14...Bd7 15.Bf3 forces the passive 15...Nce7
(15...Qa5 16.Nxc6 Bxc6 17.Ne2²), when White
could start pressurising the isolator.
14...Be6 15.Na4 Ne7 16.Rxc8 (or 16.Bd3
Bb8 17.Nxe6 fxe6 18.g3 e5 19.Rxc8 Qxc8 20.Ba3
Qe6 21.e4²) 16...Bxc8 17.Qd3 Bd7 18.Rc1 Ne4
19.Nc3 Nf6 20.Bf3².
14.Nd4
14.h3 Bh5 15.Nh4 might be objectively the
best continuation.
Then 15...Bxe2 16.Nxe2 Re8 17.Nf5 is nice
and safe for White. I only fear:
15...Bg6! 16.Nxg6 hxg6 17.Bf3 d4. Of
course White remains slightly better after 18.exd4
Nxd4 19.g3 Re8 20.Bg2, but a good player should
hold this.
14.Nb5 Ne4 15.Nd2 Bxe2 16.Qxe2 Qa5! is a
double hit so White has to take on e4 with just a
marginal edge – 17.Nxe4 dxe4 18.a4 g6.
14...Qd6 15.g3 Bh3 16.Rfe1²
White’s position looks slightly cramped, but
in fact his pieces stand well and the d5-pawn is a
target. See Game 8 Matlakov-Jakovenko, Porto
Carras 2018
B23. 7...cxd4 8.Nxd4
In 2018 Mamedyarov crushed Anand in a
blitz game with 8.exd4 Bd6 9.Bg5 Be6?! (everybody
plays this!) 10.0-0 0-0 11.Re1, when even the
relatively best line 11...h6 12.Bh4 Re8 is in White’s
favour – 13.Qd3 Rc8 14.Re2 Be7 15.Rae1².
Still, 9.Bg5 is a bluff. Black can sac a pawn
with 9...0-0! 10.Bxf6 Qxf6 11.Nxd5 Re8+ 12.Kf1
Qd8°. It is wiser to castle – 10.0-0 Bg4!, transposing

37
to a well know position with totally even chances.
8...Bd7 9.0-0
9.Nde2 a6 10.Ba4 Be6 11.Nf4 Bd6 12.Nxe6
fxe6 13.e4 0-0 is balanced.
9...Bd6
10.Nf3
This retreat is aimed against exchanges in the
centre followed by the freeing ...d4.
After 10.Be2 Nxd4 11.Qxd4 0-0 it transpires
that White’s threat on d5 is ineffective – 12.Bf3
12.Nxd5 Nxd5 13.Qxd5 Bc6°.
12...Bc6 13.g3 Re8 14.Rd1 Be5 15.Qd3 Qb6
16.Nxd5 Bxd5 17.Bxd5 Nxd5 18.Qxd5 Rad8 19.Qf3
Rxd1+ 20.Qxd1 Rd8 21.Qe2 Qd6 22.e4 Qd1+
23.Qf1 Qxf1+ 24.Kxf1 Rd1+ 25.Ke2 Rh1³.
Another possible treatment is 10.b3 0-0
11.Bb2 Re8 12.Nce2 a6 13.Bd3 Nxd4 14.Nxd4
Ne4=.
10...a6
This is Black’s most popular answer, but I
believe that it immediately gives White the upper
hand. Two bishop moves deserve more attention:
10...Be6 11.b3 0-0 12.Bb2 a6 13.Bd3
13.Bxc6 is less efficient than two moves ago
38
as the d6-bishop is protected and White lacks the
stab e3-e4 – 13...bxc6 14.Rc1 Re8 with
double-edged play.
13...Bg4 14.h3 Bh5
White can now transpose to 10...Bg4 with
15.Be2 Re8 16.Rc1!, or keep the blockade with the
committal:
15.g4!? Bg6 16.Bxg6 hxg6 17.Kg2, planning
a3, b4, Qb3.
10...Bg4!? 11.h3 Bh5 might be Black’s best
try to fight for a draw. In general, he needs his
bishop for an attack. Aagaard/Ntirlis explicitly
suggest that the bishop should stay on e6 where it
cannot be exchanged. The concept of ...Bg4 is based
on the idea to push ...d4 and reach full symmetry.
12.Be2 (or 12.b3 first) 12...0-0 13.b3
13.Nh4 Bxe2 14.Nxe2 14...Ne4 15.Nf3 Re8
16.b3 Qf6 17.Rb1 Ne5 18.Nfd4 Ng6 could become
dangerous if White allows ...Nh4.
13...a6 14.Bb2 Re8. This position occurred
only in the blitz game Carlsen-Grischuk, chess.com
3m+2spm 2017. The world champion chose:
 15.Nh4 Bg6 16.Nxg6 hxg6 17.Bf3 Be5 with
even chances. The drawback of Nh4 is that it
removes the control on d4. It would be interesting to
try some useful move:
15.Rac1, hoping to win a tempo after
15...Bc7 16.Nh4. However, Black has 15...Rc8
16.Nd4! (16.Nh4 Bg6=) 16...Nxd4 17.Bxh5 Nc6
18.Bf3 Be5 19.Ba1 d4 20.exd4 Bxd4. It looks that
Black has accomplished his goal, but 21.Na4 Bxa1
22.Bxc6! bxc6 23.Rxa1² assures White of a
long-term structural advantage. The second rank is
only temporarily weak – 23...Qxd1 24.Rfxd1 Re2
25.Kf1 Rc2 26.Rdc1 Rd2 27.Nb6 Re8 28.Nc4.
18.Rad1! g6 19.Nd4. Black is playing for
two results only. Play might continue 19...a6 20.Qb2
Rfd8 21.Ne2 Qc7 22.h3 Qe7 23.Bb1 and White
begins the siege of the d5-pawn.
11.Bxc6 Bxc6
The point of White’s previous move is that
11...bxc6 would face 12.e4!, winning a pawn.
12.b3 0-0 13.Bb2

Finally, 10...Bf5!? is the novelty White has achieved the positional dream of
Lysyj/Ovetchkin propose in the Chess Stars book the Semi-Tarrasch. He can squeeze here for the rest
The Hedgehog vs the English/Reti. It remains of the game without any risk. See Game 7
untested, but the resulting positions do not defer Ruban-Kharlov, Novosibirsk 1995.
much from what we have already seen after
10...Bg4. Lysyj/Ovetchkin’s line goes 11.Bd3 Be4
12.Nb5 Bb8 13.b3 0-0 14.Bb2 Bxf3 15.Qxf3
AlphaZero’s games inspired me to check
15.gxf3!? Re8 16.f4, but Black succeeds in trading
our most important piece with 16...a6 17.Nd4 Nxd4
18.Bxd4 Ba7=.
15...Be5 16.Bxe5 Nxe5 17.Qe2 Qb6

39
 Chapter 2. 1.c4 c5 2.Nf3 Nf6 3.Nc3 e6
Annotated Games
4. Rybka 3 – Chessmaster 11
CCRL 2008
1.c4 e6 2.d4 d5 3.e3 Nf6 4.Nf3 c5 5.Nc3 a6
6.cxd5 exd5 7.Be2 Nc6 8.0-0 Bd6 9.dxc5 Bxc5
10.b3 0-0 11.Bb2 Ba7 12.Rc1
12...Qd6
Aagaard/Ntirlis recommend this move as
their main repertoire, but the queen is hanging in
many lines at this place. Their alternative
proposition is better:
12...Re8!
Now the set-up with 13.h3 Bf5 14.Bd3 is
only equal on account of 14...Bxd3 15.Qxd3 d4, so
White should try the double-edged:
13.Rc2 Qe7 14.Rd2 Be6 15.h3
Mind the trap 15.Qb1 Rad8 16.Rfd1?!
(16.h3=) 16...Ng4! 17.Bd3 d4! with an attack.
40
15...Rad8 16.Bd3 h6. White has prevented
...d4, but his rook is clumsy on d2. He should be
constantly watching out for ...Ne4=. He may
continue 17.Re1 Qc7 18.Bb1 Bb8 with level
chances.
13.h3!?
This idea has only been tested in a couple of
old engine games and is virtually unknown. It is the
first link of the plan Bd3, Nc3-e2-d4. I believe that
Black should prepare ...d4, or he will have to suffer
with a blocked IQP.
13...Bf5
The thematic 13...Re8 14.Bd3 Bd7 15.Bb1
Rad8 is hardly good since 16.Ne2 Ne4 loses a pawn
to 17.Bxe4 Rxe4 18.Nc3 Ree8 19.Qxd5 Qg6
20.Qg5².
13...d4 14.exd4 Rd8 did not offer enough
compensation after 15.Na4 Nd5 16.Re1 Nf4 17.Qc2
Qh6 18.Bf1± Rybka 3-Hiarcs 13.1, CCRL 2010.
A better setting for the push ...d4 is 13...Rd8
14.Bd3 d4!? 15.exd4 Nxd4 16.Ne4 Nxe4 17.Bxe4
Qf4 18.Bxd4 Bxd4 19.Qc2 g6. Rybka 3-Stockfish
1.3.1 CCRL 2009, went 20.Rcd1, and White went
on to convert his marginal edge. Slightly more
challenging is 20.Rfd1 Bf6 21.Rxd8+ Bxd8 22.Bd5.
14.Bd3 Bxd3 15.Qxd3 Rad8 16.Rfd1 Rfe8
17.a3
 17...Qe7
17...Qd7 18.Ne2 Ne4 19.Ned4 Nxd4
20.Bxd4 Bb8 occurred in Junior 10-Rybka 3, CCRL
2008, when 21.Bb6! Rc8 22.Rxc8 Rxc8 23.Ba5
maintains pressure.
18.Ne2 Ne4 19.Ned4 Nxd4 20.Nxd4
The first stage of the game is over. White has
blocked the d5-pawn and safely sidestepped any
attempts of Black to attack. His advantage may not
be tangible yet, but it is very stable. Even the
machine cracks in the long run.
20...Bb6 21.Rc2 Qd7 22.Rdc1 f6
Black would be glad to trade both rooks, but
22...Rc8? loses the d5-pawn after 23.Rxc8 Rxc8
24.Rxc8 Qxc8 25.Ne2 Qc6 26.Nf4.
Here is an example where Black traded a
pair of rooks and controlled the c-file, but that did
not save him from suffering:
Keene-Fries Nielsen
Gausdal 1983
41
23.Re2 f6 24.Rd2 Bb4 25.Rd3 Qe6 26.h3
Qc6 27.Nd4 Qd7 28.Ne2 Qc7 29.Nc3 Bxc3 30.Rxc3
Qa5 31.Rxc8+ Nxc8 32.Qg4 Qc7 33.Bxf6, eating a
healthy pawn.
23.a4?!
The machine places its pawns “by the book”
– on light squares, to complement its bishop.
However, this is an example of play without a plan.
Let’s assess the position. White has
established a firm blockade on d4, and his king is
safe. That means, the opponent lacks any dangerous
plan to fear. Therefore, we should construct our own
plan of how to make progress. We have one clear
target – the d5-pawn. It is unrealistic to expect that it
will fall by itself.
The classics taught us that we should create a
second target.
The only reasonable place to seek such a
target is the queenside. Eventually it could be the
b-pawn. Thus we should think about b4+a4, having
in mind b4-b5-b6. In short, White should play
23.b4! instead of a4.

As a bonus, our queen finds a nice active
place on b3. We should not be afraid of ...Ne4-d6-c4
because the knight would be unstable there – e3-e4
could undermine its support.
Play might continue 23...Rc8 24.Rxc8 Rxc8
25.Rd1! (we need exactly one rook for the d-file!)
25...Rd8 26.Ne2 Bc7 27.Qb3².
23...Qd6 24.Qd1 g6 25.Qg4 Kf7 26.h4 h5
27.Qh3 Qe7 28.Qh2 Bc5 29.b4!
After some aimless manoeuvring, the engine
finally discovers the idea of b4 by purely tactical
means.
29...Bd6 30.Qh3 Bb8 31.b5 axb5 32.axb5
Kg8 33.Ne2 Kh7 34.Bd4
34...f5?!
42
The game was played in 2008. Today no
engine would weaken its kingside in such a manner.
They would stay tight with 34...Qf7 35.Qf3 Re7
36.b6 Qe6². The punishment came inevitably:
35.g3 Nd6 36.Nc3 Ne4 37.Qf1 Be5 38.Qd1
Bb8 39.b6! Qf7 40.Nb5 Re7 41.Rc8 Rxc8 42.Rxc8
Re8 43.Qc2 g5 44.hxg5 Kg6 45.Bf6 Qd7 46.Rxe8
Qxe8 47.Nd4 Qd7 48.Nf3 Nxf6 49.gxf6 Kxf6
50.Qc5 Qd8 51.Qd4+ Kf7 52.Kg2 Ke6 53.Qg7
Kd6 54.Qxb7 Kc5 55.Nd4 Qe8 56.Kh3 h4
57.Kxh4 Be5 58.Nxf5 Qd8+ 59.Kh3 Qxb6
60.Qe7+ Bd6 61.Nxd6 1-0
5. Kurylo – Crielesi
ICCF 2018
1.Nf3 d5 2.b3 Nf6 3.Bb2 Bf5 4.Nh4 Bc8
5.e3 c5 6.Nf3 e6 7.c4 Nc6 8.cxd5 exd5 9.Nc3 d4
(for 9...Bd6!? see next game) 10.Na4
10.exd4? led to a catastrophe after 10...cxd4
11.Qe2+ Be7 12.Nb5 Bg4 13.Qc4 Rc8 14.Nbxd4
0-0 15.a3 Re8 16.Be2 Nb4–+, N2-Jonny 4.0, CCRL
2010.
10...Bd7
10...Be7!? 11.Bb5 Bd7 is covered in the
comments to 11.Rc1.
10...d3 looks awkward for White, but in fact
this pawn is rather weak, and it does not hamper his
development so much – 11.Rc1 more active. The d3-pawn is a serious nuisance for
11.Bxf6!? gxf6 12.Rc1 is also pleasant. Black.
White will get more than enough compensation for
the exchange – 12...Nb4 13.Nxc5 Qd5 14.Na4 Bg4 This analysis shows that 9...d4 is not so
15.Nc3 Qa5 16.g3 0-0-0 17.Bg2 Nc2+ 18.Rxc2 smooth for Black .
dxc2 19.Qxc2².
11...b6 11...b6 12.exd4 Nxd4 13.Nxd4 cxd4
Or 11...Ne4? 12.g3! Nb4 13.Ne5± Nc2+
14.Rxc2 dxc2 15.Bb5++–.
12.Bxf6 gxf6 13.a3 Qd5 14.Rc3 Bb7
15.Rxd3.

11.Rc1

I think that White can pose more practical

problems to his opponent with 11.Bb5! Be7
11...a6 12.Bxc6 Bxc6 13.exd4 cxd4 14.0-0
Be7 15.Rc1.
12.0-0
After 12.Bxc6 Bxc6 13.exd4 0-0 14.0-0 Nd5
the strong centralised knight gives Black the
initiative. It turns out that winning the d4-pawn is 14.Ba6
not enough for an advantage because the a4-knight
is too far from its king.
12...0-0 I suppose that the bishop would be more
useful on e2. It was worth trying 14.Bxd4!? Bd6
14...Ba3 15.Bb2 Bxb2 16.Nxb2 0-0 17.Be2².
15.Qe2+ Be6 16.Qf3 0-0 17.Bxf6 Qxf6
18.Qxf6 gxf6. Still, it should be impossible to save
the extra d-pawn.

14...Be7 15.Bxd4 Bg4 16.Qc2 0-0 17.Bxf6

Bxf6 18.0-0=

White has won a pawn, but his knight is

13.Rc1!
13.exd4 Nxd4 14.Bxd7 Qxd7 15.Bxd4 cxd4
16.Rc1 d3 17.Re1 Ba3∞.
13.Re1 Nd5 14.exd4 Nxd4 15.Nxd4 cxd4
16.Bxd7 Qxd7 17.Bxd4 Nf4 is yet another example
where Black is pawnless, but his play is very easy.
13...Nb4 14.Bxd7 Nxd7 15.Ba3 Nxa2
16.Rc2 b5 17.Nxc5 Nxc5 18.Bxc5 d3 19.Rxa2 Bxc5
20.Qa1!. White is intending Qe5, Rb1, b4, Ne1.
Although his position looks restrained, his pieces are
clearly inferior to the black dark-squared bishop. I
even think that in an OTB game Black’s chances
should be higher. The rest is an example of
impeccable computer-guided defence.
18...g6 19.Nc3 Rb8 20.a4 Bc8 21.Bc4 Bb7
22.Rcd1 Rc8 23.Ne4 Bd4 24.Rde1 Re8 25.Kh1
Qh4 26.f4 Kg7 27.Qd3 Rcd8 28.Qg3 Qxg3
29.Nxg3 Rxe1 30.Rxe1 Bc5 31.d3 f5 32.Rd1 h6
33.h4 Kf6 34.Rf1 Re8

43

35.d4! Bxd4 36.Rd1 Bc5 37.Rd7 Re7
38.Rd8 Rg7 39.Re8 Re7 40.Rd8 Rg7 ½-½
6. Mamedyarov – Aronian
blitz Saint Louis 15.08.2018
1.Nf3 Nf6 2.c4 e6 3.e3 d5 4.b3 c5 5.Bb2
Nc6 6.cxd5 exd5 7.Nc3 Bd6 (7...Be7 8.d4)
8.Rc1
White’s move order has allowed the
double-edged 7...d4, but Black did not oblige. What
if we returned now to the Semi-Tarrasch with 8.d4?
Of course Black’s most challenging retort is 8...cxd4
9.Nxd4 0-0
I also investigated the thematic 9...Be5, but
we have here 10.Bb5!? Bd7 11.Nf3 or even
11.Nce2!? Qb6 12.Rc1. We see that the early b3
44
gave us additional possibilities in comparison with
the main line Semi-Tarrasch.
9...Nxd4 is also premature in view of
10.Qxd4 0-0 11.Nb5.
Unfortunately, I could not find any way of
extracting some benefits for White here:
10.Be2 (10.Nf3 Bg4 11.Be2 Bxf3 12.Bxf3
Be5=) 10...Nxd4 11.Qxd4 Re8 12.Rd1 Be5 13.Qd2
Bf5 14.0-0 Rc8 15.Bf3 occurred in 3 games, and
they all finished in a draw. The simplest is 15...Bxc3
16.Bxc3 Ne4 17.Bxe4 dxe4=.
Thus Mamedyarov’s idea to delay d4 is
principled, but in my opinion it is rather risky. It
deserves more attention if Black had played an early
...a6.
8...0-0 9.Na4 b6
A sterner test is 9...Qe7!? 10.Bb5
White is constricted after 10.Bxf6 Qxf6
11.Nxc5 b6 12.Nd3 Ba3.
10...Ne4
 storm will end to be in a teacup, as Black has
enough threats to draw.

12.Bb5 Ne7 13.0-0 Bg4 14.h3 Bh5 15.Be2²

White has obtained the dream IQP position

with a firm blockade on d4.

15...Rc8 16.Nd2 Bxe2 17.Qxe2 Nxd2

18.Qxd2 Bd6 19.Qd4

The c5-pawn is a target indeed, but our

king’s defence is rather neglected.
The direct 11.h3 Bd7 12.d3 Ng5 13.Nxg5
Qxg5 14.Nxc5 Bxc5 15.Rxc5 does not work in view
of 15...Qxg2 16.Ke2 Bxh3 17.Qg1 Bg4+ 18.Kd2
Qf3ƒ. Remains:
11.d3 Ng5 12.Bxc6 bxc6 13.Nxg5 Qxg5
14.Kf1 f5! 15.f4 Qe7 16.Kf2 Be6 17.Re1 c4 18.dxc4
dxc4 19.bxc4 Rab8, when I would feel ill at ease,
despite the extra pawn.

I’m afraid that the option of 9...Qe7 is

enough to discourage Mamedyarov’s followers. 19...Nf5

10.d4 Ne4 11.dxc5

Black decides to get rid of his weakness, but
he does not get any compensation for it. The
conclusion of the blitz fight was:

20.Rxc8 Qxc8 21.Qxd5 Rd8 22.Qe4 Bb8

23.Nc3 Re8 24.Qg4 Be5 25.Rc1 Qe6 26.Rd1 g6
27.Qe4 h5 28.Rd3 Qc8 29.Qa4 a6 30.Qd7 b5
31.Qxc8 Rxc8 32.Nd5 Bxb2 33.g4 Rc1+ 34.Kg2
hxg4 35.hxg4 Nh4+ 36.Kg3 g5 37.f4 Rg1+ 38.Kh3
Nf3 39.Ne7+ Kh7 40.Nf5 Bf6 41.Rd7 Kg6 42.Rd6
Rc1 43.Ne7+ Kg7 44.Nf5+ Kg6 45.Rxa6 Nd2
46.Ng3 Rc3 47.f5+ Kg7 48.Nh5+ Kf8 49.Nxf6
Rxe3+ 50.Kg2 Ne4 51.Ra8+ Kg7 52.Nh5+ Kh7
53.Re8 Re2+ 54.Kf3 Rf2+ 55.Ke3 Rxa2 56.Kxe4
Re2+ 1-0
11...Bxc5?!

In a blitz game Aronian understandably 7. Ruban – Kharlov

misses 11...bxc5! 12.Qxd5 Nxf2! 13.Qxc6 Nxh1 Novosibirsk 1995
14.Qxa8 Qe7 15.Rc3 Rd8 16.Be2 Bxh2!, and the
45

1.d4 d5 2.c4 e6 3.Nf3 Nf6 4.e3 c5 5.cxd5
exd5 6.Nc3 Nc6 7.Bb5 cxd4 8.Nxd4 Bd7 9.0-0 Bd6
10.Nf3 a6 11.Bxc6 Bxc6 12.b3 0-0 13.Bb2
13...Qe7
Black’s queen targets the kingside where
White is completely safe. Perhaps he should retain
the option of going to a5, as in the line 13...Re8
14.Ne2 Ne4 15.Rc1 Rc8 16.Qd3 h6 17.Rfd1 Qa5.
An improvement is 15.Ng3 Nxg3 (15...Rc8
16.Rc1 f6 17.Nd4 Bd7 18.Rxc8 Qxc8 19.Qd3)
16.hxg3 Qe7 17.Rc1 Rad8 (17...Qe4 18.Nd4 Bd7
19.Ne2)
Bacallao-Bruzon, Havana 2018, went 18.Qc2
f6 19.Nd4 Bd7 20.Nf5 Qe6 21.Nxd6 Qxd6=. Of
course White should not accept opposite-coloured
bishops. His main positional aim should be to
exchange the dark-squared bishops, or at least get
control of the diagonal a3-f8. That should be
46
possible after some manoeuvring with the queen, for
instance, 18.Qd4 f6 19.Rfd1 Kh8 20.Qc3 h6 21.Qa5
Bc7 22.Qd2 Bd6 23.Bd4. Combining threats on the
d5-pawn and on the c-file, White should be able to
put his bishop on c5, and then he could proceed with
the second stage of his plan – to push b3-b4-b5,
backed by a knight on c3:
In such position the tempi do not count
much. White should be patient and advance slowly.
He can always trade a pair of rooks, but not both.
14.Qd4 Bc5 15.Qh4
15...Ba3?!
An obvious positional mistake. Black will
obtain some temporary pressure on the queenside,
but in the long run the weakness of his dark squares
will begin to tell. To be fair, 15...Ne4 16.Qxe7 Bxe7
17.Ne2 is also pleasant for White – 17...Bb5
18.Nfd4 Bf6 19.Rfe1 Bxe2 20.Rxe2 Rac8 21.f3.
White allows a weakness on e3, but it is easy to
defend with the king – 21...Nd6 22.Kf2 Rfe8 23.Rd2
Bg5 24.Re1 h6 25.Ba3 Bh4+ 26.g3 Be7 27.Red1².

16.Bxa3 Qxa3 17.Qd4 Rac8 18.Rfd1 Rfd8

19.Ne5 Be8 20.f3

It is a bit early for this otherwise thematic

move. It is better to keep it for an endgame. White
could have improved his pieces with 20.Nd3 Bd7
21.Nf4 Be6 22.h3 h6 23.Rd3 or 23.g4 Qa5 24.Rd3
Qc5 25.Nce2

20...h6 21.Nd3 a5 22.Rd2?! 33...Rb2

White misses a simple trick, or he did not 33...Bxe4! 34.fxe4 Rc3 should be an easy
evaluate correctly the result of the exchanges. The draw.
logical continuation was 22.Nf4 a4
Or 22...Bc6 23.Nce2 a4 24.Ng3 axb3 34.Rd3
25.axb3 Qxb3 26.Nf5 Ra8 27.Nxg7±.
23.Rab1
34.Rd8+ Kh7 35.Nd2 leaves White
considerable chances.

34...Bxe4 35.fxe4 Kf8 36.h4 g6 37.Kf1?!

37.Kh2 Ke7 38.Kg3 Ke6 39.Kf3 h5 40.g3.

37...Ke7 38.g3 Ke6 39.Ke1 g5 40.hxg5

hxg5 41.a5

White admits that the whole idea of gong to

c1 was wrong – 41.Kd1 Ke5 42.Rd5+ Kxe4 43.Rb5
23...axb3 24.axb3. White is in command. He g4 44.Rxb7 f5 45.Rb4+ Kd3 46.Kc1?? Rf2 47.a5
could even mount an attack with g4, Rd2, although f4–+.
his main aim is an endgame with an extra pawn.
41...g4 42.Rd5 ½-½
22...Ne4! 23.Rc2 Nxc3 24.Rxc3 Rxc3
25.Qxc3 d4!? 26.exd4 Qd6 27.Re1 Bb5 28.Nf2
Qxd4 (28...a4!=) 29.Qxa5 Qd2 30.Qxd2 Rxd2
31.Ne4 Rc2 32.Rd1 Bc6 33.a4 8. Matlakov – Jakovenko
Porto Carras 15.10.2018

1.c4 c5 2.Nf3 Nf6 3.Nc3 e6 4.e3 d5 5.d4

Nc6 6.cxd5 exd5 7.Bb5 Bd6 8.dxc5 Bxc5 9.0-0 0-0

47
10.b3 Bg4 11.Bb2 Rc8 12.Rc1 Bd6 13.Be2 Bb8
14.Nd4

White has consolidated, and wants to occupy

the d4-square with Nb5. 19...a6? would just drop a
pawn to 20.Nxd5!.
14...Qd6
17.Nxc6 bxc6

This is a typical method of producing

weaknesses in White’s camp, but Black’s attack has
no chances for success in the current position.

15.g3 Bh3 16.Re1 Rfd8?!

This set-up is passive. More natural is

16...Ne5 17.Nf3! Rfe8 18.Nb5!
Aiming to trade 2 minor pieces instead of
one. 18.Nxe5 Qxe5 19.Bb5 Red8 20.Bf1 Bxf1
(20...Bg4 21.Qd3 Bf5 22.Qb5) 21.Rxf1 Qf5 22.Nb5
Rxc1 23.Qxc1 Ne4 is balanced – 24.f3 Ng5 25.Nd4
Qh3 26.Qd2 Be5=.
18...Qe7 19.Nxe5 Bxe5 20.Bxe5 Qxe5 I do not understand Jakovenko. He accepted
21.Qd4! – removing the queens. 21...Qxd4 hanging pawns without any active prospects in
22.Nxd4². return. Now simplest is 18.Qd4 c5 (or 18...Re8
19.Na4) 19.Qh4 with a clear advantage.
16...Nxd4 17.Qxd4 Qe6 is perhaps the best
way to go. 18.Na4 Bc7 19.Qd4 Ba5 20.Red1± Qe7
18.Qh4 21.a3 Rd6 22.b4 Bd8 23.Nc5 Bb6 24.Qh4 Bf5
White has some tricks after 18.Nb5 a6
19.Qh4!?, but 19...axb5 20.Bxf6 gxf6 21.g4 is
drawish.
18...Be5 19.Ba1

48
 It was better to keep the bishop – 29.Rxc5
Be4 30.Bxe4 Nxe4 31.Qxe7 Rxe7 32.Rc2 Rb7 33.f3
Nf6 34.Rdc1².

29...Rxf6 30.Rxc5 Be4 31.Bxe4 dxe4

32.Qh5 (32.Rd4²) 32...Qe6 33.Qe2 Qb3 34.Rd4
Qxa4 35.Ra5 ½-½

9. Giri – L’Ami
Amsterdam 11.07.2015

25.Bd3?! 1.c4 c5 2.Nc3 Nc6 3.Nf3 Nf6 4.e3 e6 5.d4

d5 6.cxd5 exd5 7.Bb5 Bd6 8.dxc5 Bxc5 9.0-0 0-0
10.b3 Bg4 11.Bb2 a6 12.Bxc6 bxc6 13.Rc1 Bd6
White has accomplished all his positional 14.Ne2
goals and is in full command. The next stage should
be to win material by some tactical device.
Especially the c6-pawn is hanging in many lines. This game is an illustration of the plans
Black can save it only at the price of weakening his against c6+d5 hanging pawns.
kingside. One attractive option was 25.Qg5 Bg6
25...Be4 26.Nxe4 Qxe4 27.Bxf6 Rxf6
28.Qxd5.
26.Be5 Rdd8 27.Bg4 Ra8 28.Nb3±.

25...Bg6 26.Bf1 Re8 27.Bg2

27.Bd4 h6

14...Rc8 15.Ne5 Bxe5 16.Bxe5 Qe7

In principle, White should be glad to take on

f6 in order to remain with a dominating knight
against a light-squared bishop. Therefore, 16...Re8
seems as a more logical move. Then 17.Bxf6 is still
possible – 17...Qxf6 18.Re1², but at least Black will
have an extra tempo.
28.a4± was easy to find. Other options are 17.Bd4 Nd7 18.f3 Bf5
19.Qd2 Qe7 20.b4 Ne5 21.Bxe5 Qxe5 22.Rc3 a5²
27...h6 28.a4 Bxc5 29.Bxf6?! and 17.Bb2 Nd7 18.Qd3 Bxe2 19.Qxe2 Re6 20.Rc2
a5 21.Rfc1².

49
 17.Bxf6 Qxf6 18.f3 Bd7 19.Qd4 Qe7
20.Nc3
The knight is going to c5 via a4. In my
opinion, 20.Nf4, intending Rfe1, e4, is more
straightforward.
20...Rfe8 21.Rfe1 h5?!
This pseudo-active move would have had
sense only if Black had serious prospects for an
attack. The case is just the opposite, so 21...h6 looks
more to the point. Another reasonable defensive
move is 21...a5, to hamper b4 and to prepare ...Ra6.
For instance: 22.Na4 Ra8 23.Qc5 Qd8 24.Rcd1 h6
25.e4, with pressure in the centre.
22.Na4 Rb8 23.Nc5 a5 24.Qc3 Rb5
25.e4!
50
This strike in the centre is White’s main
weapon against the c6+d5 pawn pair. It is also
effective without any minor pieces on the board
since it widens the scope of White’s rook from c5.
White could have transformed his positional
advantage into material with 25.Nxd7 Qxd7
26.Qxc6 Qxc6 27.Rxc6, immediately snatching a
pawn. The critical position arises after 27...a4
28.bxa4 Rb2 29.Kf1 Rxa2 30.Ra6. Here an
exchange of rooks would have been a decisive
mistake:
30...Rb8 31.Re2 Rxe2? 32.Kxe2 Rb2+
33.Kf1 Ra2 34.a5 is technically won. White pushes
his pawn to a7 and grabs the d5-pawn with his king.
However:
30...g6 31.Re2 Ra3 32.Kf2 h4! should result
in a drawn 4:3 endgame.
25...Qd6 26.e5² Qe7 27.a3
Again, 27.Nxd7 Qxd7 28.Qxc6 Qxc6
29.Rxc6 a4! is drawish.
27...Bf5 28.Qd4
 Correct was 59.Na6! Rc8 60.Kg2 Kg7 61.Rcf3 Rb7
62.Nc5 Re7 63.Nxe6+ fxe6 64.h4! Qc1 65.Rf6±.

59...Bd7

Black also misses his chance – 59...Ra8!

60.Rxc6 Ra3 61.Rcc2 Qh4 62.Rf3 Rb8, and it is
unclear how White could untie his pieces.

60.Ra3 Rbb8 61.Ra7 Be6 62.Ra6 Bd7

63.Ra7 Be6 64.Kg3 Ra8

White has established a total domination on

the dark squares, but his decisive break should come
from the kingside, to exploit the greater mobility of
his rooks. Obviously, it could be only g4, but Giri
takes his time.

28...Qg5 29.Kf2 Qg6 30.Re3 Qg5 31.Rce1

Qe7 32.Rc1 Qg5 33.Rcc3 Qe7 34.Kg1 Qg5 35.Kf1
Qe7 36.Kf2 Qg5 37.g3 Qe7 38.Kg2 Be6 39.h3 g6
40.Re1 Kg7 41.Rec1 Bc8 42.Re3 Be6 43.Re2 Kg8
44.Rd2 Bc8 45.Re2 Be6 46.Kh2 Kg7 47.Qc3 Kg8
48.Qd4 Kg7 49.Ree1 Kg8 50.Re3 Kg7 51.Re2 Kg8
52.Rc3 Kg7 53.Rg2 Kg8 54.g4 hxg4 55.fxg4 Kg7
56.Rf2 Qg5 57.b4 axb4 58.axb4 Kg8 65.h4?!

White begins to lose the thread, and ends up

in a worse position. 65.Ra5! Rxa5 66.bxa5 Ra8
67.Ra2± retained the edge.

65...Qh6 66.Nf4 g5 67.Nh5 Rxa7 68.Qxa7

gxh4+ 69.Kxh4 Kh8 70.g5? Qh7 71.Qd4 Qe4+
(71...Ra8 72.Qf4 Kg8µ) 72.Qxe4 dxe4 73.Nf6 Rb8
74.g6 fxg6 75.Kg5 Kg7 76.Rh2 (76.Nxe4 Rxb4
77.Nf6=) 76...Rb5?? 77.Ne8+ Kf8 78.Nc7 Bc4 1-0

10. Kramnik – J.Polgar

Geneve 27.06.2013

59.Nd3?!
1.c4 c5 2.Nf3 Nf6 3.Nc3 Nc6 4.e3 e6 5.d4
d5 6.cxd5 exd5 7.Bb5 Bd6 8.dxc5 Bxc5 9.0-0 0-0
White has finally made some progress – he 10.b3 Bd6 11.Bb2 Bg4 12.Be2 a6 13.Nd4 Bd7
has opened the f-file, and could proceed with Rf6. 14.Bf3 Be6 15.Nxc6 bxc6 16.Rc1 Rc8 17.Na4 c5
However, his last move is a tactical mistake – he
allows Black to activate his rook through the a-file.
51
 Unlike the previous game, Black prevents
the blockade on the c5-square. However, it is
impossible to support the pawn on c5 for long. In
most cases it should move to c4, surrendering the
d4-square.

Now that the knight is on c3, White has a

The best approach is to take control of the
dark squares with 18.Qd2! – aimed against ...Qa5,
...Bb4 – 18...c4 19.bxc4 dxc4 20.Rfd1 Be7 21.Bd4
Rb8
Here 22.Nb2 c3 23.Qxc3 Rc8 24.Qd3 Rxc1
25.Rxc1 Bxa2 26.Qxa6± wins a pawn. The position
remains complex, but White has chances to convert
it. However:
22.Nc3!? to keep the tension could be more
unpleasant, as 22...Qa5 stumbles into 23.Nd5.
stable edge owing to Black’s split pawns. Even the
strongest answer 24...Be5 does not solve all the
problems after 25.Nd5 Qc5 26.Bg4 Rfd8 27.e4! c3
28.Qf3 with pressure on the light squares.
24...Ba3?! 25.Rc2 Rfd8 26.Bd5 Bb4
27.Bxe6
Transforming one advantage (doubled
pawns) into another (weak king).
27...fxe6 28.Ne4 f5 29.Nf6+ Kf7

18.Qe2?! c4 19.g3 Qa5 20.Bc3 Qb5?! Now the game is over. The only move was
29...Kh8 30.Rd4! Be7 31.Nh5 c3 32.Rc4±.

It is difficult to criticize this move, although 30.Nd7 Kg6 31.Rxc4 Qb5 32.a4! Qxa4
it offers White an advantage. Black’s main chance is 33.Rcd4?!
the bad placement of the knight on a4. If White
succeeded in returning it into play, he would be
clearly better. Thus 20...Bb4! is obvious enough.
The difficult move is 21.Qb2 Ne4!! 22.Bxe4 dxe4
23.Bxg7 Rfe8, and it suddenly transpires that
White’s king is in more dangerous situation than the
black one. For instance: 24.Bh8 is not a threat owing
to 24...Bf8, while 24.Bf6 is dangerous after
24...Bh3. Remains:
24.Bc3 cxb3 25.axb3 Qh5 26.f3 Bf8 27.fxe4
Bh3 with considerable counterplay.

21.Bxf6 gxf6 22.Nc3 Qa5 23.bxc4 dxc4

24.Rfd1²

52
 Black is tight up and down. The obvious
threat is discovered check from e5, the hidden –
Rh4. Besides, 33.Rh4 was winning on the spot.

33...Qa5 34.Ne5+

34.Rh4 is mate in 14 according to the

computer!

34...Kf6 35.Qh5 Rxd4 36.exd4 Qc7

37.Qh6+ Ke7 38.Qh4+ Kd6 39.d5 Kxe5 40.Qxb4
Kf6 41.Qh4+ Kg6 42.dxe6 Qc2 43.Qd4 Qe2
44.Qd7 Rc7 45.Qe8+ Kf6 46.Qf8+ Kg6 47.Qd6
Rc2 48.e7+ Kf7 49.Qd5+ Kg6 50.Qg8+ Kf6
51.Qf8+ Kg5 52.Qg7+ Kh5 53.Qxh7+ Kg5
54.Qg8+ Kf6 55.Rd6+ Kxe7 56.Qd8+ Kf7 57.Rf6+
1-0

53
 Chapter 3. Symmetrical with ...e6, ...Nxd5
Main Ideas

This chapter deals with positions when Black

recaptures on d5 by knight. Compared to ...exd5, the
roles are reversed here. This time the isolated
d-pawn (or hanging pawns on c3+d4) will be ours,
and we should learn to use its bright sides. The main
position arises after:
1.c4 c5 2.Nf3 Nf6 3.Nc3 e6 4.e3 Nc6 5.d4
d5 6.cxd5 Nxd5 7.Bd3 Be7 8.0-0 0-0

Main Position I analyse it as line A. Contrary to K.Hansen

(in Yearbook 128), who seems to recommend 7...c4,
I’d be more afraid of 7...cxb4 8.Bb2 f6. White gets
sufficient compensation, and could easily regain the
pawn by Bd3, Qc2, but he cannot set up connection
between his heavy pieces. Play is chaotic and
double-edged. Let’s now return to the Main
Position.

Plans from the Main Position

The mere fact that White has tried 10(!)

Several paths lead to it, and they all have
their pluses and minuses.
Move orders
I mentioned above the line with early ...Nc6.
It deprives White of tricky lines without d4, but
Black misses the challenging setup 4...d5 5.cxd5
Nxd5 6.d4 cxd4 7.exd4 Nxc3!? 8.bxc3 Qc7 9.Bd2
Nd7.
On the other hand, 4...d5 5.cxd5 Nxd5
allows the trendy option 6.Nxd5 exd5 7.b4!?
different moves here suggests that he does not have
a clear path to the advantage. Our choice should
depend on what type of play we would like to
impose.
The most natural and sharp approach is to
point our pieces towards the Black king, and to try a
direct attack:
9.a3 Enabling Qc2+dxc5 or Bc2+Qd3.
Black’s most resilient set-up is 9...cxd4 10.exd4
Bf6!

54
 Black prepares ...Nce7, ...Bc8-d7-c6. The
standard development with ...Nxc3 and ...b6 offers
White considerable chances on the kingside thanks
to the pawn sacrifice h4 (after Re1). White has won
some model games from the diagram position,
implementing consistent dark-squared strategy with
11.Be4 Nce7 12.Ne5. Then he provokes ...g6, takes
on d5 and aims to remain with a knight on e5 against
a bishop on e6. See Game 11 Smyslov-Ribli,
London 1983. Apparently White’s game is easier in
this case, since he scores well in such positions.
However, Black should hold with good defence.
Our further plan assumes Bc4, Bg5, f2-f4-f5.
We might also “correct” the placement of our rooks
with Rf1 or Re1. We’d like to play with a pure
isolator rather than with a hanging pair (the latter is
easier to attack). Thus we should always consider
the option of Nxd5.
It seems that Black has no immediate need to
open our queen and bishop with 9...cxd4. He often
chooses:
9...b6 to await 10.Rd1, and only now
10...cxd4.

The modern trend for White is a more

complex, central approach. He does not waste time
on pawn moves like a3, but develops his pieces
following the scheme Qe2, Rfd1, Rac1:
9.Qe2!? and now Black is at a crossroads.
The point is that 9...cxd4 10.exd4 Bf6 is already
slow – the tempo saved on a3 counts! White grasps
the initiative with 11.Qe4 g6 12.Bh6 Re8 13.Rad1
and his pieces take the most active stands – Rfe1,
Ne5, e.g. 13...Nce7 14.h4!? Nf5 15.Bg5. So Black
commonly chooses 10...Nf6 to drag the “wrong”
rook to d1 – 11.Rd1 Nb4 12.Ne5!

Although 11.exd4 is still possible and leads

to complex positions, I would like to avoid this
version of the isolator when our king’s rook is on
d1. It is safer to enter a position with a small, but
lasting edge after:
11.Nxd5!? Qxd5 12.exd4 Bb7 13.Be4 Qd6
14.d5 exd5 15.Bxd5

55

Black’s pieces are somewhat hanging. That
should allow us to at least gain the bishop pair edge.
I devoted less than two pages to the main
lines, but you must be aware that this variation has
undergone several decades of intensive
development. Many different options have been
tested in hundreds of games. The play is not forced,
and both sides have a distractingly wide choice on
every move. In order to achieve good results in it,
you should before all understand the plans in the
different pawn structures which arise.
Let me discuss several basic strategic points.
d4 isolator or c3+d4 pawns
In most cases I prefer positions of this type:
Analysis
Our pieces are active, and Black lacks
targets. That allows us to prepare at leisure an attack
against the enemy king by pushing f2-f4-f5. Another
56
plausible idea is g4, Bg5-h4-g3, h2-h4.
K.Szabo-Petran
Hungary 2011
Black’s pieces have only defensive
functions, while White can unhurriedly increase the
pressure: 18.Bb3 Nfe7 19.Qf3 Rf8 20.Rfe1 Ra7
21.h3 Bd7 22.Bg5 f6 23.Nxd5+– fxg5 24.Nxe7+
Qxe7 25.Qe3 1-0.
In positions with a pawn on c3 we should
generate quickly concrete threats against the Black
king, even at the cost of a pawn. The modern
approach is to push h2-h4:
Curt Hansen-Ribli
EU-chT 1983
13.h4!?, intending Ng5, Qg4. 13...Bxh4
could be faced with 14.Qe2, threatening Qe4.
In the next example Black intends to bolster
his defence with Nf6:

White’s attack has no venom without a pawn
support. We should go 12.Ng5!, followed by 13.f4.
When to exchange on d5
The exchange on d5 is a recurring motif in
our version of the Semi-Tarrasch. Taking by knight
too early is generally drawish. For instance, in the
Main Position 9.Nxd5 Qxd5 10.e4 Qh5 11.dxc5 is
symmetrical. Although White has an extra tempo, he
cannot extract dividends from it.
Sometimes we should kill the d5-knight in
order to get access to the f4-square:
Anton Guijarro-Lewicki
Prague 2012
15.Nxd5 exd5 (15...Qxd5 16.Qf4 Bg7
17.Be4) 16.Qf4 with pressure.
It is more interesting to investigate Bxd5
when we believe that our knight would be stronger
57
than the Black light-squared bishop. Since the
Game 11 Smyslov-Ribli, it was believed that White
is better in this type of positions:
The ex-world champion played 17.Bxd5 to
impose consistent dark-squared strategy. The reason
for such an evaluation was the weakness of the
c6-square. As an example, without ...b6 Black
should be fine. Modern engines shake this
assessment and prove that Black can hold after
17...exd5!, but any decent player would take White
nonetheless.
I would add that the knight could be a strong
attacker in many positions where the other minor
pieces had been exchanged and Black had weakened
his kingside with ...g6.
Belikov-Kalegin
Alushta 2007
20...Qf6 21.Qg3 h5. Black’s problem here is
that he lacks any counterplay. The only weakness on
c3 is overprotected, so White can calmly improve
his position. The game went 22.Re3
(22.a4!22...Red8 23.Rae1 Kh7 24.a3 Rg8 25.Nd3
(heading for h5!) 25...Qe7 26.Nf4 Qxa3, and now
27.d5! exd5 28.Re7 Rcf8 29.Qg5 was winning
quickly.

The last example reminds me to call your

attention to the break d4-d5! Every so often in IQP
positions this break decisively opens the centre, or
simply offers White’s knights a nice stand on d4.
Analysis

Take a look. Black has 4 hits on d5 and there

are no threats along the e-file. Yet 15.d5! exd5
16.Bf5 Nb3 17.Rb1 Rc5 18.Bc2 Na5 19.Nd4
promises White the better prospects.

Theoretical status

I have the feeling that second players prefer

to avoid the Semi Tarrasch with ...Nxd5. I have
often heard that it is easier to play with an isolated
pawn after ...exd5. Practical results confirm such
claims. I looked in my database after 7.Bd3 and
found that White scores over 60% (and respectable
56% in correspondence and computer chess). More
importantly, if we take only the first forty games
between strongest players, the result is horrendous –
18-5 in White’s favour. Still, for best practical
results it would be preferable to vary between 9.a3
and 9.Qe2.

58
 Chapter 3. Symmetrical with ...e6, ...Nxd5
Step by Step

1.c4 c5 2.Nf3 Nf6 3.Nc3 e6 4.e3 d5

4...Nc6!? 5.d4 d5 6.cxd5 Nxd5 considerably

restricts the options of both sides. It might be
Black’s best way, if he aims for the positions of line
B. The only drawback of this move order is that
Black will miss the possibility 6...cxd4 7.exd4
Nxc3!? 8.bxc3 Qc7, as mentioned in line B.

5.cxd5 Nxd5 Yet, it is a dead draw after 11...0-0 12.Bf3

Nxd4 13.Qxd4 Be6 (13...Bc6=) 14.Rd1 b6 15.Bxd5
Bxd5 16.Qxd5 Qxd5 17.Rxd5 Rfd8=.

A. 6.Nxd5 exd5

In my opinion, the only plus of 6.Nxd5 over

A. 6.Nxd5!?; B. 6.d4
6.Bc4 does not fare well against 6...Nc6, and
even 6...Nf6!? (which has never been played!) is
interesting. Then 7.d4 converts to the QGA. In both
events the commitment of our queen’s knight to c3
makes the fight for an opening advantage more
difficult.
6.d4 is that it is still unexplored. Thus some of your
opponents may prefer 6...Qxd5, which gives White
nice attacking options: 7.b3 Be7 8.Bb2 0-0 9.Bc4
Qd8
9...Qh5 provokes 10.g4! Qh6 (10...Qh3
11.Bf1 Qh6 12.g5 Bxg5 13.Bxg7 Qxg7 14.h4‚)
11.h4 Nc6 12.g5 Qh5 13.Be2 Qg6 14.h5 Qf5
15.Qb1 Qxb1+ 16.Rxb1².
10.Qc2!
10.0-0 Nc6 11.d4 cxd4 12.Nxd4 Nxd4
13.Bxd4 Bd7 is boringly symmetrical. Bocharov
tried against Grachev, rapid Serpukhov 2018,
14.Qh5 Bc6 15.Rad1 Qc7 16.Bd3, and had to return
the queen to e2 after 16...g6=.
10...Nc6

6.Bb5+ Bd7 7.Be2 aims to reach the

following IQP position: 7...Nc6 8.0-0 Be7 9.d4 cxd4
10.Nxd5 exd5 11.Nxd4

59
 Black’s king. All this sounds nice, but I have mixed
feelings about this sacrifice. I’m mostly concerned
that the b4-pawn gives Black an active stand on the
queenside.

As a whole, the ensuing play is chaotic and

double-edged. White should not overestimate his
chances.

7...cxb4

C. Hansen claims in Yearbook 128 that

Black’s best choice is to decline the gambit with:
11.h4!. The plan with a kingside pawn storm 7...c4 8.Bb2 Bxb4 9.Bxg7 Rg8. I think that
has become standard in such positions. 10.Bb2 (10.Be5 Nc6 11.Bg3=) 10...Nc6 11.Qc2 Rg6
11...f5 12.Be2! takes the initiative.
This may look ugly, indeed, but 11...h6
12.g4! Nb4 13.Qb1 is dangerous. The attempt to
generate counterplay with 13...b5!? 14.Bxb5 Bb7 is
insufficient after 15.Ke2 Qb6 16.Bc4 Rad8 17.Rh3
Bd5 18.g5 h5 19.Bc3².
12.a3
After the weakening ...f5 we should switch
to play in the centre, so 12.0-0 Bxh4 13.d4 cxd4
14.exd4 is a plausible option – 14...Nb4 15.Qe2 Bf6
16.Bxe6+ Bxe6 17.Qxe6+ Rf7 18.Ne5.
The text aims to provoke new weaknesses
before castling:
12...a6 13.Ng5 Qd6 14.Qd1².

7.b4!? Black is bound to castle long, when d3

should open files against its king: 12...Qe7 13.h4!
(Anton Guijarro-Salem, Lake Sevan Martuni 2015,
saw 13.g3 Bh3÷) 13...h6 (13...h5 14.Nd4!?) 14.g3
Bd7 15.h5 Rg4 16.Kf1 0-0-0 17.Nd4 Rg5 18.d3.

8.Bb2 Nc6

8...f6!? should transpose. Black can hardly

save this move anyway. The idea of transferring the
knight to f6 proved bad in Kuzubov-Pe.Schreiner,
Hersonissos 2017 – 8...Nd7 9.Rc1 Nf6?! 10.Bb5+
(10.Bxf6) 10...Bd7 11.Bxf6 gxf6, when 12.Bxd7+!
Qxd7 13.Nd4 is strategically murky for Black. The
knight reigns over the board.
Modern chess is all about activity! White’s
idea is to put a bishop on b2, provoking ...f6. Then 9.Rc1
he could exploit the weak light squares around
60
 I also considered 9.Bd3 f6 10.Qc2 Be6!
The h7-pawn cannot be saved in view of
10...g6?! 11.h4.
11.Bxh7
White has regained the pawn, but Black
comfortably completes development with:

12.Bb5+

Similar is 12.Bd3 Qd7 13.f4 Bd6 14.Qh5+

Bf7 15.Qh4 h5 16.0-0 a5.

12...Kf7 13.Bd3 h5 14.f4 a5 15.0-0∞

11...Bd6, intending to hide the king on f8.
Also 11...Qd7 12.Nd4 Nxd4 13.Bxd4 Bd6 14.Bd3
Rc8 15.Qb1 a5 16.f4 Kf8 is totally unclear. White White has enough compensation for the
will have to launch a committal pawn storm on the pawn.
kingside, but he lacks coordination between his
heavy pieces.
B. 6.d4 Nc6
Lalith played 9.Nd4, when the paradoxical
9...h5! 10.Rc1 Rh6 should be fine for Black.
The delay of ...Nc6 gives Black the
9...f6 10.Nd4 interesting option:
6...cxd4 7.exd4 Nxc3!? 8.bxc3 Qc7

The only game with this set-up, Anton

Guijarro-Kasimdzhanov, blitz, Riyadh 2017, saw
10.Bb5 Bd7 11.Qb3 a6 12.Bd3 Be6 13.Nd4 Nxd4
14.Bxd4 Rc8 15.0-0 Be7 16.Qb1 g6 17.h4=.

10...Nxd4 11.Bxd4 Be6

Instead of ...Nc6, Black will bring his knight

on f6 to defend its king.
9.Bd2
61
 The attempt to activate the hanging pawns
with 9.Bb2 Nd7 10.a4 Bd6 11.Bd3 b6 12.0-0 Bb7
13.c4 is only enough for a dynamic balance. After
13...0-0 14.d5 Nc5 (14...h6) 15.Bc2 exd5 16.cxd5
Rad8 17.Qd4 f5 18.a5 Ne6 the chances are even.
However, if White does not push d5, Black will
prepare ...e5 – 14.a5 Rae8 15.h3 e5.
9...Nd7 10.Bd3. Black will castle, and will
aim for ...e5. We cannot break the defence without
including a pawn in the attack. Therefore, we must
play Ng5 and f4, while keeping the c-pawn on c3.
10...Bd6 11.0-0 (11.Ng5 Nf6) 11...b6 12.Ng5
Nf6 (12...Bb7 13.f4 Nf6) 13.f4!

Perhaps the best chance is to eliminate to a

Preventing ...e5 and ensuring a forepost on
e5 for the knight. In some lines f4-f5 could throw
more oil into the fire. Black must now decide
whether to repel the knight with ...h6, or live with
constant threats on e6 and h7. I think that in both
events White retains an initiative:
a) 13...0-0 14.Qe1!?
14.Qe2 Bb7 15.Rae1 h6 16.Nf3 Bxf4
17.Bxf4 Qxf4 18.Ne5 Qg5 19.g3 Rac8 20.c4 Rce8
21.Bc2 promises White full compensation, but
21...h5 22.Rf4 Qh6 holds on. We could triple our
heavy pieces on the f-file, just to maintain the
equilibrium.
14...Bb7 15.Qh4 h6 16.Rae1 Be7 (16...Nd5
17.Rf3) 17.Re3 Rac8 18.Rfe1 (18.Rh3 Ne4) 18...b5
19.Rg3 Kh8. Although White’s pieces look
daunting, I could not find anything decisive.
slightly better endgame. That said, the introductory
move does not hint for an endgame:
20.Rxe6!! (20.f5) 20...b4! 21.c4 Qd7
22.Rxe7 Qxe7 23.f5 Rce8 24.Nh3
24.Nf3 Bxf3 25.gxf3 Rg8 26.Bxh6 gxh6
27.Qxh6+ Nh7 28.Qxh7+ Kxh7 29.f6+ Kh8 30.fxe7
Rxe7 is about equal. We pursue an improved version
of this ending:
24...Rg8 25.Nf2 a5 26.Ng4 (26.Rh3 Nh7
27.Qf4°) 26...Nxg4 27.Rxg4 Qxh4 28.Rxh4 f6
29.d5. The two central pawns are very strong. The
key line is 29...Ba6 30.Kf2 Rc8 31.Re4 Rgd8
32.Ke3 Rc5 33.Kd4 Rdc8 34.Re8+!².
b) 13...h6 – the Black king has not castled
yet, so the knight must go back – 14.Nf3 Bb7
15.Ne5
Perhaps 15.Qa4+!? Bc6 16.Qc2 is stronger.
The idea is to play quickly Rae1 and push f5 before
Black does it. For instance: 16...0-0 17.Rae1 b5!
(17...Rac8 18.Ne5 Bb7 19.Qd1 Nd5 20.f5 exf5
21.Rxf5²) 18.Ne5 Rad8 (18...Nd5 19.f5) 19.f5 or
19.h3 first.
15...0-0 (15...Ne4 16.Qa4+ Kf8 17.Be1)
16.Qe2

62
 7...cxd4 opens prematurely the c1-h6
diagonal. After 8.exd4 the only justification of the
previous exchange is 8...Bb4.
8...Be7 9.0-0 0-0 10.Re1 Bf6 11.Be4 Nce7
gives White an improved version of the line 9.Re1 –
see below. White has an extra tempo, as he saved a3.
Both 12.Qd3 and 12.Ne5 maintain the pressure.
However, it turns out that White can ignore the hit
on c3 by:
9.0-0!, when 9...Nxc3 10.bxc3 Bxc3 11.Rb1
assures White of an edge:

All this occurred in Iturrizaga-Sethuraman,

Dubai 2018. White put his queen on e2 instead of e1
since ...h6 invites plans with g2-g4-g5. His set-up
paid off in the game after 16...Rad8?! 17.Rae1 Rfe8
(or 17...b5 18.Qe3 a6 19.Qh3), and now 18.Qe3!
Nd5 19.Qh3 f5 20.g4 Ne7 21.Bc4 Nc6 22.Nxc6
fxg4 23.Rxe6! assures White of an edge.
I believe that the best defence should be
based on ...f5, so 16...Nd5! is critical. Black plays
...f5 and blocks both White’s bishops. White cannot
prevent it since he has to take on e5 by the d-pawn
after 17.f5?! Bxe5. This is the difference with the 11...Bb4 12.Ne5 (or 12.Be4 a5 13.a4 Be7
line 15.Qa4+!?, where the knight on e5 is protected 14.Qc2) 12...a5 13.Nxc6 bxc6 14.Qg4.
by a rook.
7...Nxc3 8.bxc3 Be7 9.0-0 0-0 offers White a
7.Bd3 wide range of plans. Most frequent is:
10.Bb2!? b6 11.e4. The computer prefers to
provoke a weakness first:
7.Bc4 does not make much sense. Black 10.Qc2 h6 11.Rd1, and only then Qe2, Bb2,
should only take care of the d5-square: e4.
7...cxd4 8.exd4 Be7 9.0-0 0-0 10.Re1 Bf6!
10...Nxc3?! 11.bxc3 b6 12.Bd3 Bb7 13.h4! 8.0-0 0-0
has been devastating for Black in practice –
13...Bxh4 14.Nxh4 Qxh4 15.Re3 g6 16.Rg3±,
Anand-Morrison, Blackpool, 1988.
11.Ne4
11.Bb3 is best met by 11...Nde7, when
Kramnik’s sacrifice 12.Bf4 yielded him a victory
after 12...Bxd4 13.Nxd4 Nxd4 14.Bc4, but I’d
hesitate to recommend it.
11...b6 12.Nxf6+ Nxf6 13.Bg5 Bb7=. Black
can ignore the pin with ...Ne7 or ...Qd6.

7...Be7

63

9.Qe2
White’s best attacking set-up is Rfe1, Rad1,
Bg5, but it cannot be reached without Black’s help.
For instance, after:
9.Re1, Black does not take on d4, but keeps
our bishop closed – 9...b6! 10.Nxd5
10.a3 Bb7 11.Nxd5 Qxd5 12.e4 Qd7 has no
venom – 13.d5 exd5 14.exd5 Nd4 15.Nxd4 cxd4
16.Bf4 Rfe8=.
10...Qxd5
10...exd5 is also possible. Black only should
avoid the positional trap 11.b3 cxd4?! (11...Nb4!
12.Bb1 Be6 13.Bb2 Rc8=) 12.Bb5 Bd7 13.Bxc6
Bxc6 14.Nxd4².
11.e4 Qd8 12.dxc5 bxc5 13.Bf4 Bb7=.
Another major variation, which could be our
back-up line, is 9.a3!? (intending Bc2, Qd3) 9...cxd4
Avoiding 9...b6 10.Qc2 g6 (10...Nxc3!?
11.bxc3 h6 12.Rd1, followed by e4) 11.dxc5 bxc5
64
12.Rd1.
9...Nxc3 10.bxc3 b6 hands White the better
centre. See Game 13 Sargissian-Bogner, Batumi
2018.
10.exd4 Bf6
Practical experience has proved that Black’s
defence is very solid after ...Nce7, ...Bc8-d7-c6. For
instance:
a) 11.Be4!? Nce7 12.Qd3
Smyslov won a memorable game with
12.Ne5. Game 11 Smyslov-Ribli, London 1983, had
been a milestone for this line until Black found the
most solid set-up with 12...Bd7.
12...g6 13.h4!?
I propose this move since I would not like to
trade bishops after 13.Bh6 Bg7 14.Bxg7 Kxg7,
when a future ...f6 should hold.
13...Bd7 14.Bg5
14...Bc6 15.Bxd5 Nxd5 16.Ne4 Bxg5
17.hxg5.
 b) 11.Re1 Bd7!
The safest approach. 11...Nxc3 12.bxc3 b6 is
more unbalanced. It bolsters White’s centre, so he
could hurl the h-pawn forward – 13.h4!? Bb7
(13...Bxh4 14.Qe2ƒ) 14.Ng5 g6 15.Qg4 h5 16.Qg3
with tangible pressure.
12.Bc2 Nce7 13.Qd3 g6 14.Nxd5 Nxd5
15.Bh6 Bg7 16.Qd2 Rc8 17.Be4 Bc6 18.h4 Bxh6
(18...f6!?) 19.Qxh6 Nf6 20.Bxc6 Rxc6 21.Rac1=.

9.Be4!? cxd4 10.exd4 Nf6 11.Bxc6 bxc6

12.Na4 Ba6 13.Re1 Bb5 (13...Bc4 14.Ne5 Bd5
15.Bg5)

B1. 9...b6; B2. 9...cxd4

B1. 9...b6 10.Rd1 cxd4 11.Nxd5

11.exd4!? Nf6 transposes to line B, but we

should also consider 11...Nb4.

11...Qxd5

11...exd5 allows White to play against an

isolated pawn after 12.Bb5 Bd7 13.Bxc6 Bxc6
14.Bg5 Bxa4! (14...h6 15.Bxf6 Bxf6 14.Nxd4 Rc8 15.b3².
16.Nc5) 15.Qxa4 h6 16.Bxf6 Bxf6 is rather drawish:
12.exd4 Bb7 13.Be4 Qd6
a) 17.Rac1 Qb6 18.b3 Qb5 19.Qc4
19.Rc4 Rfd8 20.g3 Rd5 21.Kg2 a5 22.h4
Rad8 23.Re4 Rf5.
19...Rfd8 20.Ne5 Qxc4 21.Rxc4 c5! 22.Nc6
Rd5 23.dxc5 Rc8 24.Nxa7 Rdxc5 25.Rxc5 Rxc5
26.a4 Rc2=.

b) 17.Qxc6 Rc8 18.Qe4 Qb6 19.b3 Rfd8

20.Rad1 Rd5 21.Rd2 Rcd8 22.Red1 g6 23.g3 h5
24.Kg2 a5=.

Finally, 9.Nxd5 Qxd5 10.e4 Qh5 11.dxc5

Bxc5 12.Bf4 b6 13.a3 a5= is too symmetrical.

After 9.Qe2, we have a further split:

14.d5

This is the safest continuation. It assures

White of the small, but stable advantage of a bishop
65
vs knight. It may be insufficient against a top player
as Wesley So, so I understand Carlsen’s choice
against him in Saint Louis 2018:
14.g3 Bf6 15.Bg5, and he got pressure
following:
15...Qe7?! 16.Bxf6 Qxf6 17.Ne5 (stronger
was to double the rooks on the c-file). However:
15...Bxg5 16.Nxg5 h6 17.d5 exd5 18.Bxd5
Qg6= would have been fairly dry.
14...exd5
14...Nb4 lost a pawn to 15.Bg5 Rfe8
16.Bxe7 Rxe7 17.dxe6 Qxe6 18.Bxh7+ in
Pelletier-Prusikin, Switzerland 2005.
15.Bxd5 Qg6 16.Be4!?
Setting some problems to the opponent.
16.Ne5 Nxe5 17.Bxb7 Rab8 18.Ba6 is another line
where White is marginally better.
16...Qe6! 17.Re1 Rfe8 18.Qb5 Qd7
After 18...Qd6 19.Be3 Rad8 20.Rac1 Na5
21.Bxb7 Nxb7 22.Nd4 Qd7 23.Nc6 the white knight
occupies the c6-square.
19.Ne5
19.Bd2 Rad8 20.Bc3 Na5 21.Qxd7 Rxd7
22.Bc2 Nc6 23.Ba4 a6, and Black tames the
initiative.
19...Nxe5 20.Qxd7 Nxd7 21.Bxb7 Bb4
22.Rd1 Rad8 23.Bg5 f6 24.Be3
The two queenside pawns give White some
chances to squeeze.
B2. 9...cxd4 10.exd4 Nf6
10...Bf6 allows 11.Qe4 g6 12.Bh6 Re8
13.Rad1 Nce7 14.h4!? Nf5 15.Bg5, maintaining the
66
pressure.
11.Rd1 Nb4
White refrained from a3, so this leap looks
like a natural way to “punish” him. Still, the rare
11...b6 is a tough nut to crack. I investigated:
12.Bg5 Bb7 13.a3 Rc8 14.Bc2
14.Rac1 sets the positional trap 14...Na5?!
15.d5! exd5 16.Bf5 Nb3 17.Rb1 Rc5 18.Bc2 Na5
19.Nd4ƒ. Critical should be 14...h6 15.Bf4 Na5∞.
14...Re8
14...Nd5!? looks dangerous owing to 15.Qe4
g6 16.Qh4 Nxc3 17.bxc3 h5 18.Be4 Bxg5 19.Nxg5
Na5 20.Bxb7 Nxb7 21.g4, but 21...Nd6 22.gxh5
Ne4 neutralises the attack – 23.f4 Nxg5 24.fxg5
Rxc3 25.hxg6 fxg6 26.Qh6=.
White can try 15.Nxd5 Bxg5 16.Nc3 Bf6
17.d5 (17.Be4 g6 18.h4 Ba8 19.h5 Nxd4=) 17...exd5
18.Rxd5 Qc7 19.Qd3 g6 20.Rd7 Qb8 21.Nd5 Bg7
22.Rd1 with a pretty position, yet the only weakness
on f7 should be defendable.
15.h4
White cannot break the defence of the black
king without the help of the h-pawn. Game 12
Anton Guijarro-Mo.Nikolov, Linares 2017, is a fine
illustration of an effective dark square strategy after
15...Nd5?! 16.Nxd5 exd5 17.Qd3 g6. Black’s best
follow up is:
15...h6 16.Bxf6 Bxf6 17.Qe4 Kf8 18.Ba4
Kg8 19.Qf4∞.
 12.Ne5! 18...b6 19.Rh4. On the other hand, 18...Nh5
stumbles into 19.Nxf7! Kxf7 20.Rxh5 Bxg5
21.Rxh7+ Kf8 22.f4!! Bf6 23.d5 Qb6+ (23...Bxd5
In all the games White retreats the bishop to 24.Qxg6) 24.Kh1 Rc7 25.Rxc7 Qxc7 26.Ne4±.
c4 or b1, but that allows Black to develop his bishop
with ...b6, reaching a double-edged position. The This analysis suggests that Black should
idea behind 12.Ne5 is to counter: keep control of c6:
12...b6 with 13.Qf3 Nbd5 14.Nc6 Qd7 12...Bd7 13.Bc4! Rc8 14.Bb3 Bc6
15.Nxe7+ Qxe7 16.Bg5 Bb7 17.Nxd5 Bxd5 18.Qg3.
The pin is extremely nasty as something like
18...Rac8 19.Rac1 Bxa2? 20.Qh4 h6 would stumble
into 21.Bxh6! gxh6 22.Qxh6 and it is unclear how
Black could prevent a rook lift on the kingside, e.g.
22...Bd5 23.Qg5+ Kh8 24.Ra1 Rfd8 25.Qh6+ Kg8
26.Re1.

Another point is that:

12...Nxd3 brings our rook closer to the
enemy king – 13.Rxd3 Bd7 14.Rh3 Bc6
Or 14...Rc8 15.Bg5 Bc6 16.Qd3 g6 17.Qd2±.
15.Qd3 g6 16.Bh6 Re8 17.Bg5 Rc8 18.Rd1.
Black cannot wait, as Whites attack would play
itself after:
In this typical position White’s chances are
somewhat better.
He can take the c6-bishop – 15.Nxc6 Rxc6
16.Bf4 Nbd5 17.Be5 a6 18.Rd3, keeping the tension
since the c3-pawn is indirectly protected in view of
the threat d4-d5. More unpleasant for Black is to
retain the strong knight on e5:
15.Re1 Bd5 (15...Qxd4 16.Nxf7) 16.Nxd5
Nbxd5 17.Bg5

67
 Obviously we should attack on the kingside
here. White has two plausible plans :
to push f2-f4-f5;
to prepare g4, Bg5-h4-g3, h2-h4.
Here are some illustrative lines:

a) 17...Rc7 18.Rad1 h6 19.Bh4 a6 20.Rf1

The other plan is 20.a3 b5 21.g4, e.g. 21...a5
22.Bg3 a4 23.Ba2 b4 24.Ng6 fxg6 25.Qxe6+.
20...b5 21.f4 Qc8 (21...a5 22.a3 Qc8 23.Bxf6
Nxf6 24.Qxb5) 22.h3 Rd8 23.f5

23...exf5 24.Qf3 Qe6 25.Be1 Rcc8 26.Ba5

Re8 27.Qd3 g6 28.Rde1 Qd6 29.Bd2 (29.g4 b4
30.gxf5 g5) 29...Bf8 30.g4 with an attack.

b) 17...Re8 18.Rad1 Qb6 19.g4 Rcd8

20.Qf3ƒ.

68
 Chapter 3. Symmetrical with ...e6, ...Nxd5 Nxe5 23.Nh5 Nf3+ 24.gxf3 Nf5 25.Nxf6 Nxh6
Annotated Games 26.d5+–.

11...Nce7 12.Ne5

11. Smyslov – Ribli

Candidates sf2 London (7), 1983 I propose in the “Step by Step” section
12.Qd3 g6 13.h4!?, intending Bg5.

1.d4 Nf6 2.Nf3 e6 3.c4 d5 4.Nc3 c5 5.cxd5 12...g6

Nxd5 6.e3 Nc6 7.Bd3 Be7 8.0-0 0-0 9.a3 cxd4
10.exd4 Bf6
Recently Svidler turned to 12...Nf5!? and
gradually equalised after:
13.Re1 Nde7 14.Be3 Nxe3. I suppose that
White should not allow this regrouping. He could
oppose it with:
13.Ng4!? with the idea of opening the centre
– 13...Nxd4 (13...Bd7 14.Qd3) 14.Qd3

11.Be4

In the previous White game in the match

Smyslov chose 11.Qc2 h6 12.Rd1 Qb6 13.Bc4 Rd8
14.Ne2 Bd7 15.Qe4 Nce7 16.Bd3 Ba4 17.Qh7+ Kf8
18.Re1 Bb5 19.Bxb5 Qxb5 20.Ng3 Ng6 21.Ne5 14...Kh8 15.Rd1 Nb3 16.Nxf6 Qxf6 17.Bxd5
Nde7, Nxc1 18.Raxc1 exd5 19.Nxd5².

Another sturdy set-up is 12...Bd7 13.Qd3 g6,

when White misses the resource h4+Bg5. This is the
difference between 12.Ne5 and 12.Qd3!. Although
White retains pressure after 14.Bh6 Bg7 15.Bxg7
Kxg7 16.Bxd5 exd5 17.Qf3 (17.Rfe1 f6) 17...Be6
18.Rad1, Black should be able to neutralise it.

13.Bh6 Bg7 14.Bxg7 Kxg7 15.Rc1 b6

16.Nxd5 Nxd5 17.Bxd5

and won with a direct attack after 22.Bxh6

69
 17...Qxd5? 24.f3?

Perhaps this recapture is based on wrong
calculation. It could have been justified only if
Black had tempi for ...f6. However, Smyslov does
not leave Ribli a chance for it. The blunt 17...exd5!
dooms Black to play for two results only, but the
bishop would control the g4- and h3-square. It is
unlikely that White could win the Q+N vs Q+B
endgame, although it looks pleasant for him. For
instance, the queenside play 18.Rc3 Be6 19.Qa4 is
ineffective due to 19...Qg5 20.Nf3 Qe7 21.Rfc1
Rfc8.
A critical moment. The only way to fix the
advantage was 24.Ra7! Ra8 25.Nxc6 Qxc6 26.h5!,
winning the battle for the dark squares. Smyslov’s
move allows 24...h6! 25.fxe4 hxg5 26.Nxc6 Rxc6
27.hxg5 Rc2=, when even his legendary technique
should not bring him a full point. The b-pawn is
bound to fall.
24...Qf5? 25.Ra7 Ba4 26.Re1 Rc2 27.b4
Bb3 28.bxa5 bxa5 29.Re4 h6 30.Qe3 Rb2 31.Rg4

18.Rc7 Bb7 19.Qg4 Rad8 20.Rd1

White could have been more concrete in his

actions. 20.h4! Qe4 21.h5 Qxg4 22.h6+! is hopeless
for Black – 22...Kxh6 23.Nxg4+ Kg5 24.Ne5 Ba6
25.Re1 f6 26.Nf7++–.

20...a5 21.h4

21.Qg5! is extremely nasty, because Black

cannot repel the queen from there – 21...h6? fails to
22.Rxf7+!. The white rook has reached the kingside and
the game is over.
21...Rc8 22.Rd7 Qe4 23.Qg5 Bc6
31...g5 32.hxg5 h5 33.Rg3 h4 34.Rg4 h3
35.g6 h2+ 36.Kxh2 Rh8+ 37.Kg3 Rxg2+ 38.Kxg2
Qc2+ 39.Qf2 Rh2+ 40.Kxh2 Qxf2+ 41.Kh3 Qf1+
42.Rg2 Qh1+ 1-0

70
 12. Guijarro – M. Nikolov
Linares 14.08.2017

1.Nf3 Nf6 2.c4 e6 3.Nc3 d5 4.d4 c5 5.e3

Nc6 6.cxd5 Nxd5 7.Bd3 cxd4 8.exd4 Be7 9.0-0 0-0
10.Qe2 Nf6 11.Rd1 b6 12.Bg5 Bb7 13.a3 Rc8
14.Bc2 Re8 15.h4

21...Bxg5

White has won the positional battle. He

prevented ...f6 and can switch his attention back to
the kingside.

22.hxg5 Nd8 23.Rxc7 Rxc7 24.Ne5 Bc8

25.Re1

15...Nd5?!

This game caught my attention because it is

closely related strategically with Smyslov-Ribli. It
presents model play on the kingside dark squares in
a symmetrical pawn structure. White’s opening
moves were not the best, and here 15...h6 16.Bxf6
Bxf6 17.Qe4 Kf8! would have been balanced.

16.Nxd5 exd5 17.Qd3 g6 18.b4!

Preventing ...Na5 and gaining space on the 25...f6?!

queenside. White’s ultimate positional goal is to
occupy the e5-square for his knight, like in the
previous game.
18...Rc7 19.Rac1 Qd6 20.Ba4 Rec8 21.Qd2
The Bulgarian grandmaster cracks under the
pressure. He understandably wanted to free his
kingside of the bind, but he only made new
weaknesses. It would be more stubborn to wait
passively. 25...Be6? loses to 26.Qf4, threatening
Nxg6, so he should go 25...Bf5 26.g4 Bc8, then
...Kg7.

26.gxf6 Qxf6 27.Bb3 Be6 28.b5 Kg7 29.a4

71
Qf8

Black is completely paralysed and cannot

protect his d5-pawn against threats like Qa2,
Ne5-d3-f4, Re5 or Re1-e3-f3.

30.Nd3 Qd6 31.Nf4 Bf7 32.g3 Qa3 33.Bxd5

Bxd5

11.Qc2

It is difficult to affirm that White should

provoke a weakness first, but this move is in no way
worse than 11.Qe2. From a human standpoint, it is
easier to play against a clear target.

11...g6 12.Rd1 Qc7 13.e4

34.Nh5+! Kg8 35.Nf6+ Kf7 36.Nxd5 Rd7

37.Qf4+ Kg7 38.Qe5+ 1-0 This set-up is more natural than c4, since it
assumes an attack against the enemy king. At the
same time, White threatens 14.d5!?, so Black’s most
precise move order is 13...Rd8 14.Be3 Bb7. In the
13. Sargissian – Bogner game both sides ignored the option 13...Bb7 14.d5!
Batumi 24.03.2018 exd5 15.exd5 Ne5 16.Nxe5 Qxe5 17.c4 Bf6 18.Ra2
Bc8 19.a4².

1.c4 e6 2.Nc3 d5 3.d4 Nf6 4.cxd5 Nxd5 13...Bb7 14.Be3 Rfd8 15.h4!
5.e3 c5 6.Bd3 Nc6 7.Nf3 Be7 8.0-0 0-0 9.a3 Nxc3
10.bxc3 b6
This approach has become semi-automatic in
such positions. I also noticed that the AI project
Alpha Zero tended to employ it against virtually any
set-up with g6.

15...Rac8 16.Qe2 Bf6 17.Rac1 Bg7

All the engines prefer to stop the h-pawn by

...h5, e.g. 17...Na5 18.Rb1 h5, but humans rarely
play like that.

18.h5 cxd4 19.cxd4 Qe7 20.hxg6 hxg6

72
 21...Bf6 22.Bxf6 Qxf6 23.e5

Positionally White looks fine – Black has

gaping holes on d6 and f6. However, tactics betray
him, as he loses the d4-pawn after 23...Qe7! 24.Qe3
Nxd4! 25.Nxd4 Rxc1 26.Rxc1 Rxd4 27.Qxd4 Qg5,
when White is lucky to dicover a perpetual check
after 28.Bf1 Qxc1 29.Qd8+.

23...Qf4?! 24.Qe3 Qxe3 25.fxe3 Ne7 26.Kf2

21.Bg5?

Sargissian has been quite consistent in his

aggressive intentions so far, but here he does not
dare to make the next logical step:
21.e5!!. You might wonder why I award this
thematic move with two exclamation marks. The
key is to see how White’s queen reaches the h-file:
21...Qxa3 22.Ra1 Qb4 23.Be4 Rc7 24.Qf1!!,
planning g4 or g3. It suddenly transpires that Black
may not live to convert his extra pawn. White has
many threats all over the board. An illustrative line Even in an endgame the strong centre offers
is White an advantage. Perhaps Black should have
24...Rdc8 25.g4 Ne7 26.Rdb1 Qc3 27.Rc1 now eliminate the white knight with 26...Bxf3,
(27.Bxb7 Rxb7 28.Rc1 Qb4 29.Rxc8+ Nxc8 30.Rc1 although both 27.Kf3 and 27.gxf3 Kg7 28.f4 would
Ne7 31.Ng5 is also tempting) 27...Qb4 28.Bd2 Qb2 retain the better prospects.
29.Bxb7 Rxb7 30.Rxc8+ Nxc8 31.Rb1 with total
domination: 26...Rxc1 27.Rxc1 Rc8 28.Rxc8+ Nxc8
29.e4 Kg7 30.g4

The golden rule in an endgame is to avoid

putting the pawns on the squares of the same colour
as the bishop. The simple 30.Ke3! a6 31.Nd2 Kf8
32.g3 Ne7 33.Nc4 Nc8 34.Nd6 was good enough,
but White judges that he would have fair winning
chances with a pawn on g5. This is a highly
committal decision since he might miss the invasion
path through g5 lately.

30...Kh6 31.g5+ Kg7 32.Ke3 Ne7 33.Nh2

Nc6 34.Ng4 Na5 35.Nf6 Bc6 36.Kd2 Kf8 37.Kc3
31...Qa3 loses to 32.Ng5 and 31...Qa2 to Ke7 38.Kb4 Ba8 39.Bb5 Nc6+ 40.Bxc6! Bxc6
32.Qb5. 41.d5

73
 41...a5+?

I’m not sure how White could break in after

41...Bb7 42.d6 Kd8. At least it is not trivial. It is
easier to defend the pawn on a7 than the one on b6.
In fact, Black will soon lose it after White puts a
knight on c4.

42.Kc4 Ba4 43.Ng4 Kd7?! 44.Ne3 Kc7

45.Kd4 Bb3 46.Nc4 Ba2 47.Nd6 Kd7 48.Nxf7 b5
49.d6 b4 50.axb4 axb4 51.Nh8 1-0

74
 Chapter 4. Symmetrical with ...b6
Main Ideas

1.c4 c5 2.Nf3 Nf6 3.Nc3 b6

8...Be7
If Black delays the development of his
kingside, he might never succeed in castling, as
Game 14 Kasparov-Salov, Barcelona 1989, teaches.
9.Bg2 a6 10.0-0 Qc7 11.Be3

One has to be a die-hard fan of the Hedgehog

to dare entering it via this move order. It offers
White the full range of plans against this set-up, so
your choice should be a matter of taste.
I offer a straightforward approach which I
have tested in my practice:
4.e4! (to force 4...d6) 5.d4 cxd4 6.Nxd4 Bb7
7.Qe2!? e6
7...Nc6 has some following, but then 8.Nxc6
Bxc6 9.Bg5!? Nd7

I also consider the plan with Bb2 in Game

15 Kiril Georgiev-Berg, Rijeka 2010, but Be3 is
somewhat easier to play.
11...0-0 12.Rac1 Nbd7 13.f4 Rfe8
(13...Rac8 14.g4!) 14.g4

10.0-0-0!? throws Black out of his comfort

zone. Instead of long manoeuvring he should think
about how to avoid a miniature. Besides, the calmer
10.Rd1 is also good enough.
8.g3

75
 That’s it. Any manoeuvring in the centre is 15.Bxe6! fxe6 16.Nxe6 Kf7 17.Ng5+ Kf8
useless. Our target should be the black king. We aim 18.Ba3 Bxd5 19.e5!!±.
to take all the space we can – g5, h4, f5. In my
opinion Black is in trouble here. Another typical break is e4-e5:

Tactical hits Analysis

We should never forget that the position after

7.Qe2 is no longer in the domain of the closed
openings. It is an open Sicilian (in relation to the
Kann), and we should treat it accordingly. Do not
miss thematic blows on e6:
Analysis

11.Nxe6! fxe6 12.e5².

Serper-Pisulinski
Miedzybrodzie 1991

16.Nxe6!! fxe6 17.Bxe6+ Kg7 18.Nd5 Qc6

19.Qd2 Nc5 20.Nxf6 Bxf6 21.Bd5 Qc7 22.e5±.

Analysis

76

11.e5! Bxg2 12.exf6 Bxf1 13.fxe7².
pieces around for ever.
In practice White most often chooses 4.g3
Bb7 5.Bg2 e6 6.0-0 Be7 and begins to fight for e4
with 7.Re1. I do not understand why not put it in
immediately on move 4?!.
Another popular move is 4.d4, but after
4...cxd4 5.Nxd4 Bb7 White has to play 6.f3 in order
to achieve e4. That completely changes his future
plans towards play on the queenside with b4.
My proposition 4.e4, then 7.Qe2, is the most
aggressive approach. It is based on the natural plan
of fianchettoing the king’s bishop and a subsequent
pawn storm with g4-g5. Practical experience
confirms that Black does not feel happy against it,
scoring less than 38%.

Move order with 2...b6

1.c4 c5 2.Nf3 b6 3.Nc3 Bb7 4.e4 e6! 5.d4

cxd4 6.Nxd4 Nf6 is less popular. It discourages us
of the plan with Qe2, so we should pay a special
attention to it.

I propose 7.e5 Ne4 8.Nxe4 Bxe4 9.Nb5 to

the lazy, and 7.f3! as more challenging. See Game
16 Shredder 12-Quazar 0.4, CCRL 2012.

Theoretical status

The Hedgehog has always had the reputation

of a somewhat dubious set-up, which has
considerable practical values. Computer assessments
are not especially helpful because the engines will
always award White an edge of about +0.50, even
though they would not offer any constructive plan. If
you listen to them, White should stay and shuffle his
77
 Chapter 4. Symmetrical with ...b6
Step by Step
1.c4 c5 2.Nf3 Nf6
Another possible move order is 2...b6!?
3.Nc3 Bb7 4.e4. It involves relatively unexplored
play:
a) 4...Nc6 5.d4 cxd4 6.Nxd4 e6 7.Be3
7.Bf4 Qf6 8.Be3 Bc5 9.Ndb5 Bxe3 10.fxe3
Qe5! is unclear.
7...Nf6 8.f3 transposes to line b.
b) 4...e6! 5.d4 cxd4 6.Nxd4 Nf6 7.e5!?
This is a decent way to avoid complications.
7.f3! is the most common and challenging
answer. Then 7...d6 8.Be3 Be7 9.Qd2! a6 10.g4!? is
a promising version of the English Attack. Safer is
to play in the Taimanov style – 7...Nc6! 8.Be3 with
a small edge. See Game 16 Shredder 12-Quazar 0.4,
CCRL 2012 for more details.
7...Ne4 8.Nxe4 Bxe4 9.Nb5 Nc6 10.h4 (or
10.Nd6+ Bxd6 11.Qxd6) 10...h6 11.Nd6+ Bxd6
12.Qxd6 f6 13.Bf4 Rf8 14.f3².
3.Nc3 b6
78
4.e4!
Aiming to force ...d6, which would block the
f8-bishop, and to gain a tempo for d4, Bd3.
4...d6
4...Bb7 5.e5 Ng8 6.d4 Bxf3?! (6...cxd4
7.Nxd4 Nc6 8.Bf4²) is altogether dubious as White
gets an overwhelming initiative after 7.Qxf3 Nc6
8.d5 Nxe5 9.Qe2 d6 10.f4 Nd7 11.g4 g6 12.h4±,
Djuric-Marinkovic, Svetozarevo 1990.
4...Nc6 5.e5
For 5.d4 – see Game 16.
5...Ne4 allows White to win a pawn with:
6.Ne2 f6 7.d3², but Black does have some
compensation. I would play for an attack with:
6.Bd3 Nxc3 7.dxc3 Qc7 8.Bf4.
5.d4 cxd4 6.Nxd4 Bb7 7.Qe2!?
White had a huge choice in this pure
Hedgehog position. A lot of ink has been spent on
plans with 7.f3, short castling and subsequent attack
on the queenside with b4.
I think that the most promising approach is
the pawn storm on the kingside:
7.f3!? e6 8.Be3 Be7 9.Qd2 a6
 10.g4!. See my analysis of this stab in the
annotations to Game 16 Shredder 12-Quazar 0.4,
CCRL 2012.
I consider 7.Qe2 as a main line mostly
because I have tested it in practice myself.
A. 7...Nc6; B. 7...e6
Delaying kingside development with
7...Nbd7 could be punished by central play, as
Game 14 Kasparov-Salov, Barcelona 1989, proves.
7...g6 8.g3 Bg7 9.Bg2 0-0 10.0-0 brings
White fantastic results probably because Black has
no counterplay after 10...Nc6
10...Nbd7 leaves too many pieces on the
board and Black cannot find room for the f6-knight
– 11.Rd1 Qb8 (11...Ne8 12.Bg5±) 12.f4
79
The threat e4-e5 is quite awkward. 12...h5
weaken the kingside – 13.f5 Qc7 14.Nd5 Bxd5
15.exd5±, while 12...Nc5 13.b4 e5 14.Ndb5 destroys
Black’s centre – 14...Ne6 15.Nxd6 exf4 (15...Nd4
16.Rxd4 exd4 17.Nxb7 dxc3 18.e5 Ne8 19.c5 bxc5
20.bxc5+–)
16.Nxb7±.
11.Nxc6 Bxc6 12.Bg5 Re8 13.Rfd1 Qc7
14.Rac1 Rac8
This position occurred in
Cheparinov-Ponkratov, Mali Losinj 2017. White had
lasting pressure after 15.Nd5 Qb7 16.b3 Nxd5
17.exd5 Bd7 18.Qd2 Qc7 19.Re1, but I would not
hurry with Nd5. Instead:
15.b3 Qb7 16.f3 retains more tension,
keeping Nd5 in reserve.
A. 7...Nc6 8.Nxc6 Bxc6 9.Bg5!?
Of course 9.g3 g6 10.Bg2 Bg7 11.0-0 0-0
12.Bg5, as in the above-mentioned game
Cheparinov-Ponkratov, is also possible. The bishop Kramnik-Adams, Tilburg 1997, went:
sortie leads the game course outside the typical
Maróczy Bind schemes.

9...Nd7

Actually, 9...g6 is possible, although White

retains a lasting structural advantage after 10.0-0-0
Qc7 11.Bxf6 exf6 12.h4 h5 13.Kb1 Bg7 14.f4².

10.Rd1

My style has evolved throughout the years

towards calm positional schemes, so I chose the safe
way. In my youth I would have castled long
(perhaps that’s why I was a junior world champion
back then!):
10.0-0-0!?
16.Nd5 Bxd5 17.Rxd5 Qc7 18.g3 Nf6
19.Rd2 b5 20.Bh3?! which turned out to be too
slow, and the game was eventually drawn. It was
time for a direct action:
16.h4! b5 17.g4 Nb6 18.c5 Nc4 19.Bc1±.

10...h6

10...g6 was possible since:

11.Rxd6 is now dubious – 11...h6 12.Bh4 g5.
Instead I like:
11.Qe3 h6 12.Bh4 Bg7
Or 12...Rc8 13.f4;
12...g5 13.Bg3 Bg7 14.Be2 Ne5 15.0-0 Ng6
16.b3 Nf4 17.Rfe1±.
13.Be2

10...h6
10...Qc8 11.Kb1 Nc5 12.f3 a5 13.h4
discouraged Black from trying to castle. In
Korchnoi-Gheorghiu, London 1980, he succeeded in
evacuating the king from the centre, but remained
very passive after 13...Qb7 14.h5 h6 15.Bh4 e6
16.Qd2 Qc7 17.Nd5 Qb8 18.Bd3 Ra7 19.Rhe1 Kd7
20.Bc2 Kc8 21.Nc3±.

I have also analysed 10...g6, when 11.Rxd6

is interesting – 11...h6 12.Bh4 g5 13.Rxc6 Ne5
14.Re6 fxe6 15.Qh5+ Nf7 16.Qg6 Qd4 (16...gxh4?
17.c5) 17.Nb5 Qc5 18.Bg3 Rc8 19.Qxe6 Rc6 13...Nc5
20.Qd5 e6 21.Qxc5 bxc5 22.Be2². 13...0-0 14.Rxd6 g5 loses here to 15.Rxc6
11.Be3 e6 12.Bd4 e5 13.Be3 Be7 14.Kb1 gxh4 16.0-0 Ne5 17.Rxh6+–.
0-0 15.f3 a6 14.b3
80
 It is unnecessary to sacrifice a pawn –
14.Nd5 Bxd5 15.cxd5 Bxb2 16.Bb5+ Kf8 17.0-0
Kg8 18.f4÷.
14...0-0 15.Nd5 Re8 16.0-0 Qd7 (16...Bxd5
17.cxd5±) 17.Bf3 Rac8 18.h3. White has a good
version of the Maróczy Bind as Black lacks
counterplay.

18...exf4 19.Bxf4 Ne5 20.b3 Qc6 21.Rd5

Rae8 22.Qd1 g5 23.Bxe5 dxe5 24.Rxf8+ Bxf8
25.Qf3 Be7 26.Rd6 Qa8 27.Rg6+ 1-0.

B. 7...e6
11.Bf4

Some players fear this Scheveningen set-up

Kramnik’s manoeuvre 11.Be3 e6 (11...g6?! because of 8.g4 Be7 9.g5 Nfd7 10.h4. However, this
12.Bd4) 12.Bd4 was also interesting, to prevent the plan is not so efficient with a queen on e2 since it
fianchetto of the black bishop. 12...Nf6 13.g3 Be7 does not defend g5 and hampers the bishop’s control
14.Bg2². over b5. 10...Nc6 gives enough counterplay:

11...e5

11...g5 is a serious weakening – 12.Bg3 Bg7

13.Qe3 Qc7 14.Be2 Bb7 15.0-0 Qc5! 16.Qc1 Be5
17.Nd5 Bxd5 18.exd5 Bxg3 19.hxg3 a5 20.Rd3 0-0
21.Re3 Rfe8 22.b3².

12.Be3 Be7 13.g3 0-0 14.Bg2 f5 15.f4! fxe4

16.0-0 Qc7 17.Bxe4 Bxe4 18.Nxe4²

The opening battle concluded with nice

pressure down the d-file. My game Kiril
Georgiev-Dimov, Pleven 2015, finished with:
81
Now 11.Be3 Nxd4 12.Bxd4 0-0 13.0-0-0 a6÷
is a double-edged position with mutual chances.
Perhaps White should take himself:
11.Nxc6 Bxc6 12.Be3 (12.f4 h6) 12...a6!
12...Nc5 13.Bg2 Qd7 14.f4 Qb7 is
ineffective – 15.Bd4 e5 16.Bxc5! is strategically
difficult for Black – 16...dxc5 17.0-0-0 0-0 18.Nd5
Kh8 19.Qd2 f6 20.h5 Bxd5 21.exd5 Bd6 22.h6 g6
23.Rde1 Qf7 (23...Qc7 24.Qc3+–) 24.gxf6 (24.Bh3
Qc7 25.f5²) 24...e4 (24...Qxf6 25.Re4 Rad8 26.fxe5
Bxe5 27.Rf1 Qd6 28.Qe3±) 25.Bxe4 Qxf6 26.f5
gxf5 27.Bc2±.
13.Bg2, and here 13...b5!? offers Black
sufficient play – 14.cxb5 axb5 15.Nxb5 (15.0-0 h6÷)
15...h6! 16.g6 0-0 17.0-0 Bxb5 18.Qxb5
Bxh4 19.gxf7+ Rxf7÷.
8.g3! Be7
If Black tries to force b3 at the expense of
...a6 or ...Be7 – 8...Nbd7 9.Bg2 Qc7, intending
10.0-0 Rc8, we could transpose to the main line with
10.Ndb5 Qb8 11.0-0 a6 12.Nd4, when Black’s
queen should soon return to c7 (or 12...Be7 13.f4
0-0 14.g4).
9.Bg2 a6
Other move orders should transpose, but they
also offer White additional possibilities. For
instance, 9...0-0 10.0-0 a6 11.e5±.
10.0-0 Qc7
82
11.Be3
Kotronias and Semkov recommend putting
the bishop on b2 in Attacking the Flexible Sicilian,
although not exactly in the same position. Look at
my Game 15 Kiril Georgiev-Berg, Rijeka 2010, for
detail.
11...0-0 12.Rac1 Nbd7 13.f4 Rfe8
We have two possible plans – g4 and f5.
Black takes measures against the latter, preparing to
defend the e6-pawn with his rook. Black’s second
choice is:
13...Rac8 14.g4!
Nobody has found this energetic stab in
practice. 14.f5 e5 15.Nd5 Qd8 16.Nc2 is
unimpressive, while 14.b3 Qb8 15.g4 Nc5 16.Bd2²
is pleasant for White, but b3 might be superfluous.
14...Nc5
 The trick is 14...Qxc4 15.Qd1!, and the
queen lacks retreats –15...Nc5 16.g5 Qd3 17.Qxd3±.
15.Bd2 h6
White’s attack is even stronger after 15...g6
16.g5 Nh5 17.f5 e5 18.Nc2 f6 19.fxg6 hxg6 20.Nb4
fxg5 21.Nbd5 Qd8 22.Rxf8+ Bxf8 23.Bh3 Nf4
24.Bxf4 gxf4 25.Qg2 Kh7 26.Kh1+–.
16.h4 g6 17.f5 h5 18.gxh5 Nxh5 19.fxe6
fxe6 20.Bh3 Ng7 21.Qg4 Bf6 22.Be3ƒ.

13...Nc5 14.Bf2 Rac8 15.f5 e5 occurred in

Nogueiras-Sl.Kovacevic, Linares 1997. Black does
not have any counterplay after 16.Nb3 Nxb3
17.axb3 h6 18.h4±.
14.g4
18.Ne6!! Denying ...Nf4. 18...fxe6 19.Bxh5
exf5 20.Nd5 Bxd5 21.cxd5 Nc5 (otherwise the
queen cannot block the f-pawn, e.g. 21...Qb7
22.exf5 gxh5 23.f6±) 22.exf5 gxh5 23.b4±.
17.Nxf5 Bf8 18.Nd5
18.Bf3 gxf5 19.Bxh5 Rxe4 20.Nxe4 Bxe4 is
not enough, but why give Black chances?!
18...Qd8 19.Rcd1 Rc8 20.b3. White owns 5
ranks of the board. The only possible counterplay
20...b5 is well countered by 21.c5! Bxd5 22.Rxd5
Nxc5 23.Qf3 f6 24.Bd4 Nxe4 25.gxf6 Nexf6
26.Rxd6 Bxd6 27.Qd5+ Kf8 28.Nxd6 Qe7 29.Nxe8
Rxe8 30.Qc6 Kg7 31.Qxa6 Rd8 32.Qb6±.

15.g5 N6d7 16.h4 Qd8

14...Nf8 The only game I know,

14...g6 15.g5 Nh5 is more active, although
Black would still be struggling following 16.f5 exf5
16...Bf8 17.Bf3! e5
Iskusnyh-Mchedlishvili, Tula 1999, finished
16...Bd8 17.h5 Rc8 18.b3 Qb8 19.Qb2 f6 20.h6 g6
21.Rcd1 Qc7 22.Ndb5 axb5 23.Nxb5 Qb8 24.Nxd6
Re7 25.gxf6 Rf7 26.e5 Bxg2 27.Qxg2 Rc7 28.b4
Ra7 29.a4 1-0.

83
 17.f5 Ne5 18.b3±

Black is suffocating. In fact, Stockfish

claims that he is lost already since he lacks
counterplay. For example, 18...f6 would face 19.g6.

84
 Chapter 4. Symmetrical with ...b6
Annotated Games
14. Kasparov – Salov
World Cup Barcelona 03.1989
1.Nf3 Nf6 2.c4 b6 3.Nc3 c5 4.e4 d6 5.d4
cxd4 6.Nxd4 Bb7 7.Qe2 Nbd7 8.g3 Rc8
If Black wanted to play the Hedgehog, he
should have better started with 7...e6 and ...Be7. His
“original” move order quickly leads him into
trouble.
9.Bg2 a6 10.0-0 Qc7
Salov already faces difficulties in completing
development. 10...e6 allows 11.Nxe6! fxe6 12.e5
Bxg2 13.exf6 Qxf6 14.Kxg2 Be7 15.Re1 Ne5
16.Bf4 0-0 17.Bxe5 dxe5 18.Ne4
11.b3 e6
11...b5 12.Nd5! Qc5? 13.Be3 bxc4 simply
does not work – 14.b4 Qa7 15.Nb5 Qb8 16.Ba7+–.
The text also allows Kasparov to switch to
tactics:
12.Nd5! Qb8
12...exd5 13.exd5+ is hopeless, especially
85
against Kasparov. He would not even need to find
some extraordinary moves. Simple mobilisation of
forces should do the job:
13...Ne5 14.f4 Qc5 15.Bb2 Nfd7 16.fxe5
dxe5 17.Kh1+–;
13...Kd8
14.a4!
Of course 14.Re1 Ne5 15.f4 is good enough,
but the text is just crushing. White is planning a5,
Bd2.
14...Ne5 15.f4 Ng6 16.a5 Be7 17.axb6 Qd7
18.Be3 Ke8 19.f5 Ne5 20.Ne6!+–;
13...Be7 14.Nf5! Ne5 15.Nxg7+ Kd8
16.Bb2+–.
13.Rd1
This is too sophisticated. Kasparov aims to
discourage ...exd5, but he certainly had more
threatening moves in his disposal. For instance:
13.Bh3 h5 14.Re1 Ne5 15.Bg5 exd5 16.exd5
Be7 17.Nf5 Rc7 18.f4±, or:
13.f4 h5 14.Bb2 Qa8 15.b4 Be7 16.Nxe7
Kxe7 17.e5±.
13...g6
13...exd5 14.exd5+ Kd8 15.Nc6+ Bxc6
16.dxc6 Nc5 17.b4 Ne6 18.Bb2 is hopeless for
Black, so Salov decides to fianchetto the bishop.

14.Bg5?!
The world champion betrays his patent style.
14.Bh3! was much more energetic, to prevent
castling. I guess he missed the hit 14...Bg7 15.Bxe6
fxe6 16.Nxe6 Kf7 17.Ng5+ Kf8 18.Ba3 Bxd5
18...Nxd5 19.cxd5 Kg8 20.Qg4 Nf8 21.Qf3
Qc7 22.Bxd6 Qd7 23.e5±.
19.e5!! Nxe5 20.Qxe5 Rc5 21.Bxc5 bxc5
22.Qe2 Qe8
Or 22...Bf7 23.Re1 Qc7 24.Rad1, winning
the d6-pawn.
23.Qxe8+ Nxe8 24.cxd5 Bxa1 25.Rxa1±.
14...Bg7 15.Bxf6 Nxf6
15...Bxf6 16.Nxf6+ Nxf6 17.e5! Bxg2
18.exf6 Bh3 19.Rd2 is strategically lost for Black.
16.Nxb6 Rd8?
86
The critical moment of the game. Only
16...Rc7 allowed Black to keep his position together,
although White remains better after 17.Na4 (17.e5?
dxe5 18.Qxe5 0-0). After the text White opens up
the centre.
17.e5! Bxg2
Apparently Black missed 17...dxe5 18.Nc6!±
when he retreated his rook to d8.
18.exf6 Bxf6 19.Nxe6! fxe6 20.Qxe6+ Be7
21.c5!!
The only winning move. Black cannot
answer 21...Bc6 in view of 22.Re1 Qc7 23.Nc8!!+–.
21...Bb7 22.Re1 Qc7 23.c6 Bxc6 24.Rac1
Rd7 25.Nxd7 Qxd7 26.Qc4 Bb7 27.Qc7 Rf8
28.Qb8+ Kf7 29.Rc7 1-0
The next game considers the set-up with
Bb2:
15. Kiril Georgiev – Berg
EU-ch Rijeka 12.03.2010
1.Nf3 Nf6 2.c4 c5 3.Nc3 d6 4.d4 cxd4
5.Nxd4 b6 6.e4 Bb7 7.Qe2 e6 8.g3 Nbd7

Note that the move order 8...Be7 9.Bg2 a6
(9...0-0 10.0-0 a6 11.e5±) 10.0-0 Qc7 does not
impede the queenside fianchetto. 11.b3 is possible
since:
11...b5 12.Ba3 bxc4 13.Qxc4 Qxc4 14.bxc4
is difficult for Black – 14...Nbd7 15.Rab1 Rb8
16.Rfc1. White’s pieces generate a lot of tactical
threats. For instance:
16...0-0 stumbles into 17.e5, while 16...g6
(preparing...e5), faces the spectacular 17.Nd5!! exd5
18.cxd5 0-0 19.Nc6 Bxc6 20.dxc6 Nc5 21.e5! dxe5
22.Rxb8 Rxb8 23.Bxc5±. Thus Black should go:
11...0-0 12.Bb2 Nbd7 13.f4 Rfe8
13...b5!? is a bold pawn sac – 14.cxb5 axb5
15.Ndxb5 Qb6+ 16.Qf2 Qa5 17.a4 d5 18.exd5 Bc5
19.Nd4 exd5 20.Ncb5 Ba6 21.Rfc1².
14.Kh1 Bf8
This is an improved version of Berg’s play.
Black has not weakened f6, so the plan with Rae1
and e5 is pointless. However, White’s attack is
87
dangerous: 15.g4! Nc5
15...g6 16.e5!?;
15...h6 16.g5 hxg5 17.fxg5 Nh7 18.h4 g6
19.Qe3 Bg7 20.Rad1 Ne5 21.Qg3ƒ.
16.b4 e5 17.fxe5 dxe5 18.Nf5 Ne6
19.g5! (dragging the knight to a perilous
square) 19...Nxg5 20.Rac1 a5 21.Nd5 Nxd5 22.cxd5
Qd8 23.h4 axb4 (23...g6 24.hxg5 gxf5 25.Rxf5+–)
24.hxg5 Ba6 25.Qg4 Bxf1 26.Bxf1 g6 (26...Rxa2
27.Bb5 Rxb2 28.Bxe8 Qxe8 29.Nh6+ gxh6
30.gxh6+ Kh8 31.Qg5) 27.Nh6+ Bxh6 28.gxh6
Rxa2 29.Rc2 Ra5 30.Rf2 Rc5 31.Qg3ƒ.
9.Bg2 Qc7 10.0-0 Rc8 11.b3 Be7!
11...a6 allows 12.Nd5! – Game 14 Kasparov
– Salov.
12.f4 a6 13.Bb2
13...g6?!
 This move was probably meant to prepare
...e5, or Black just wanted to wait with his castle 18.Kg1 dxe5 19.fxe5 Nh5
until I define my plan. However, it weakens f6. That
encouraged me to abandon g4 in favour of the break
e5! The normal follow up was:
13...0-0 14.g4 g6 15.g5 Nh5 16.Bh3!?. The
threat on e6 is especially strong with the bishop
lurking from b2:
16...Ng7
16...Nc5 17.b4 e5 (17...Nd7 18.Nxe6 fxe6
19.Bxe6+ Rf7 20.Nd5 Bxd5 21.Bxd5+–)

20.Ne4!

20.g4 Ng7 21.Nf3² was less ambitious.

20...Nxe5 21.g4 Ng7 22.Nxe6?

18.Nf5 Ne6 19.Nh6+ Kg7 20.fxe5 Bxg5
21.Nf5+ Kg8 22.Nxd6 Nef4 23.Qf3±.
17.Rad1 Rcd8 18.Qg2 e5 19.Nde2 f6
20.gxf6 Bxf6 21.Kh1 Kh8 22.Nd5 Bxd5 23.exd5
exf4 24.Bxf6 Nxf6 25.Rxf4².
Regaining the pawn, but throwing my
advantage away. 22.g5!± Nh5 (22...Nf5 23.Nxf5
exf5 24.Nf6+ Bxf6 25.gxf6+–) 23.Nc2 would have
given me a huge edge.

14.Rae1 0-0 15.Kh1?!

The immediate 15.e5!? was possible –

15...dxe5 16.fxe5 Ne8 17.Bxb7 Qxb7 18.Rd1 Ng7
19.g4². But the computer also finds a really crushing
option:
15.Bh3!! Rce8 16.Nxe6 fxe6 17.Bxe6+ Kg7
18.Nd5 Qc6 19.Qd2 Nc5 20.Nxf6 Bxf6 21.Bd5 Qc7
22.e5±.

15...Rfe8 16.e5 Bxg2+ 17.Kxg2 Qb7+

17...Nh5! was playable for Black – 18.Nf3
Nc5 19.Rd1². The point is that 17...Qb7+ allowed
me to trade queens 18.Qf3± with a tangible edge
with any risk. However, I judged that my attack
should be very strong.
88
23...Nd7? drops a whole queen to 24.Qf3 and
Nf6+, 23...Nc6 24.Ne3 (going for g4) is also
hopeless, so the only move is:
23...Qb8 24.Qf2 Rf8 25.Bxe5 Qxe5 26.Nf6+
Bxf6 27.Rxe5 Bxe5 28.Qxb6. OTB Black has
chances here, but objectively the c-pawn should
decide the game. My mistake led to a draw rook
endgame. I tried to complicate it, but to no avail.

22...Nxe6 23.Bxe5 Ng5 24.Nxg5 Bxg5

25.Qg2 Qxg2+ 26.Kxg2 Bd2 27.Rd1 Rxe5
28.Rxd2 Rc7 29.a4 b5 30.cxb5 axb5 31.a5 Ra7
32.b4 Re4 33.Rb1 Rxg4+ 34.Kf3 Rg5 35.Rd6
Rf5+ 36.Kg3 Rc7 37.Rb3 Rc1 38.h3 Rg1+ 39.Kh2
Ra1 40.Kg3 Rg1+ 41.Kh2 Ra1 42.Kg3 h5 43.Rd4
Rg1+ 44.Kh2 Ra1 45.Kg3 ½-½

16. Shredder 12 – Quazar 0.4

CCRL 40/40 16.03.2012
This is an interesting version of the English
attack. The differences with the original set-up
1.c4 Nf6 2.Nc3 c5 3.Nf3 b6 4.e4 should favour White. For instance, the counterplay
with ...d5 is totally ineffective – 10...d5? 11.cxd5
exd5 12.g5 Nh5 13.exd5.
10...h6 11.h4 Nc6 12.0-0-0 Nd7 13.Kb1 is
similar to 10...Nc6, but the weakness of the kingside
may tell.
10...Nc6 11.0-0-0 Nd7 (11...0-0 12.h4)
12.Kb1 0-0 13.h4 Nxd4 14.Bxd4. This position has
never occurred in practice. To me, the only sensible
try for Black is to prepare ...b5. Let’s consider
14...Rb8 15.g5 Bc6 16.Qc2!?

4...Nc6

This move brings an original flavour to the

game. Black usually builds up directly the
Hedgehog with 4...d6 5.d4 cxd4 6.Nxd4 Bb7 7.f3!?
e6 8.Be3 Be7. I like here 9.Qd2 a6 10.g4! (10.0-0-0
allows 10...h5 11.Kb1 Nbd7 12.Be2 Qc7 13.h3 h4,
player112-yakimov, gameknot.com 2013).
Setting up an X-ray on the c6-bishop –
16...b5 17.cxb5 axb5 18.Nd5! Bxd5 19.exd5 e5
20.Be3±.
16...Qc8 17.Bh3!
Just for the record, even 17.Nd5 is possible –
17...exd5 18.exd5 Ba8 19.Bd3, but Black could
return the piece – 19...Ne5 20.Bxh7+ Kh8 21.f4
Qxc4 22.fxe5 Bxd5 23.Qxc4 Bxc4 24.exd6 Bxd6
25.Be4 Rfe8 26.Bc6².
89
 17...b5 18.g6!
This trivial for the English Attack blow is
winning here – 18...hxg6 19.h5 g5 (19...gxh5
20.Bxg7+–) 20.h6 gxh6 21.Bf5+–.
5.d4 cxd4 6.Nxd4 Bb7 7.f3 e6 8.Be3 Qb8
Other options are:
8...Bc5 9.Qd2 0-0 10.0-0-0 Qe7 11.Nxc6
The flank attack 11.g4 is not dangerous
when the black bishop is active – 11...Nxd4 12.Bxd4
d6 13.Be2 Rad8.
11...Bxc6 12.Bxc5 Qxc5 13.Qd4 Qa5
14.Kb1². White has a comfortable space advantage.
8...Bb4!? – preparing ...d5. A logical retort is
9.Qc2 0-0 10.0-0-0, when 10...d5 11.Nxc6 Bxc6
12.Kb1 Qe7 13.exd5 exd5 14.Bd4 leaves White with
the more active pieces.
9.Qd2 Be7
90
10.Be2
Suba and Korchnoi won their games with
10.0-0-0 0-0 11.g4. However, their opponents
missed 11...Ne5!, when 12.g5 Nxf3!? would be
messy. On the other hand, the prophylactic 12.Be2
Rc8 13.Bf4 allows the counter-attack 13...Ba6
14.Ndb5 Bxb5 15.cxb5 Bb4 16.Kb1 Bxc3 17.bxc3
Ne8÷. Short castling looks safer.
10...0-0 11.Rd1 Rd8
11...a6 12.0-0 Qc7 occurred in a game
between engines.
Instead of 13.Nxc6 dxc6 14.c5 b5 15.f4 Rfd8
16.Qc2 e5 17.f5 a5= White should take aim at b6:
13.Qe1 Bc5 14.Qf2 d6 15.Na4 Nxd4
16.Bxd4 e5 17.Be3².
12.0-0 Nxd4
The transition to a Hedgehog structure with
12...d6 is not easy to play since almost all Black’s
pieces are not on their proper places:
12...d6 13.Qe1 Ne5
An attempt to redeploy the knight to the
“right” square. 13...a6 14.Qf2 Nd7 15.Na4 Na7
(what else?!) 16.Rc1² is awkward for Black.
14.f4 Ned7 15.Bf3
 After the text the centralised black queen
keeps the position together.

18...bxc5 19.f5 Qe5 20.fxe6 fxe6 21.Bf3

Bxf3 22.Rxf3 dxc4 23.Rg3+ Kf7 24.Rf1+ Ke7
25.Qa5 Rd3 26.Rg4

15...Bf8
15...Qc8?! would be ignored – 16.g4! Qxc4
17.g5 Ne8 18.Rc1 Bf8 19.Be2 Qc7 20.Qh4±.
16.f5 Re8 17.fxe6 fxe6 18.b3 Ne5 19.Bg5
Nfd7 20.Be2 a6 21.Qg3. Black has only one good
piece, but his kingside is all weak. White threatens
Bh5 or Rf1-f4-h4.
26...Rad8 27.Qxa7+ R3d7 28.Qa5 h6 29.a4
13.Bxd4 Bd6 14.f4 Nxe4 15.Nxe4 Bxe4 Rd1 30.Qc7+ Qxc7 31.Rg7+ Ke8 32.Rxc7 Rxf1+
16.Bxg7 Bc5+ 17.Bd4 d5 33.Kxf1 Ra8 34.Ke2 Rxa4 35.Kd2 Ra1 36.Rxc5
Rg1 37.g3 Rg2+ 38.Kc1 Rxh2 39.Rxc4 Rg2
40.Rc3 Kd7 ½-½

18.Bxc5

This move looks strong in conjunction with

the break f5, but Black is able to neutralise the
tactical threats. White should have chosen a
dark-squared strategy – 18.Qe3! Qc7 19.b4 Bxd4
20.Rxd4, and he is somewhat better because of his
safer king. For example, 20...Rac8 21.Rfd1 Re8
22.a3 f5 23.Bf3 Qg7 24.cxd5 exd5 25.Bxe4 dxe4
26.h3².
91
 Chapter 5. Maroczy Bind
Main Ideas

In this chapter I deal with lines when Black

plays 2...g6. We should take all the space we can,
and reach a Maróczy Bind set-up, as on the
following diagram:
1.c4 c5 2.Nf3 g6 3.e4! Bg7 4.d4 cxd4
5.Nxd4 Nc6 6.Be3 Nf6 7.Nc3

• Plan with ...Be6, ...Qa5, ...Rc8, ...a6,

preparing ...b5.

Many games reach this position via the

Hyper Accelerated Dragon move order 1.e4 c5
2.Nf3 g6, and it is well tested at all levels. Fischer
won as Black a memorable game against Bent
Larsen in the semi-final Candidates match in 1971
to cause a boost of popularity of this set-up.
Subsequent play mostly depends on Black’s
choice, so we should be aware of their 2 most
important plans:
• Dark square strategy with ...Nxd4, ...a5,
...Nd7, ...Qb6.
Sometimes Black also plays ...e6, ...Qe7,
hoping for ...d5, but it is more of a sideline against
my recommended repertoire.
I’ll elaborate more on these plans below. It is
important to remember that whatever line you
consider as White, it should be checked against both
scenarios.
White does not have a clear-cut plan to
pursue. His idea is to restrict the opponent’s pieces
and to gradually overrun him in the centre.

The critical moment is when exactly Black

exchanges on d4.

1. Black takes on d4 before we played Be3

Then we lead out our bishop to g5: 3...Nc6

4.d4 cxd4 5.Nxd4 Nf6 6.Nc3 Nxd4 7.Qxd4 d6
8.Bg5 Bg7 9.Qd2 Be6 10.Rc1 0-0 11.Bd3 Qa5
92
12.0-0 Rfc8 13.b3 a6 14.Rfe1
We are perfectly set for an offensive in the
centre with Nd5 and exd5, or with h3, f2-f4-f5.
Black should better wait for Be2 – 6...d6
7.Be2 Nxd4 8.Qxd4 Bg7 9.Bg5!. Again we should
think of further f2-f4-f5.
2. Black forces Be3
It would be good to activate our bishop on
g5, but Black has a move order which drags it to e3
– 3...Bg7 4.d4 cxd4 5.Nxd4 Nc6 6.Be3 Nf6 7.Nc3.
The game takes another course, when we should
forget about aggressive stabs like f4 and concentrate
on fighting Black’s dark square strategy. Again, it is
critical when Black takes on d4.
a) If he does it before we lost a tempo on
Qd2, we develop the queen on d3 –
7...Nxd4?!8.Bxd4 d6 9.Be2 0-0 10.0-0 Bd7
11.Qd3!, or 7...0-0 8.Be2 d6 9.0-0 Nxd4 10.Bxd4
Bd7 11.Qd3!
93
Black fails to reach the set-up ...Qb6+Nc5
because 11...a5 12.b3! (to meet ...a4 by b4) 12...Bc6
does not force f3. Thus White has time to mount a
tactical device down the d-file – 13.Rad1! Nd7
14.Bxg7 Kxg7 15.Qd4+ Kg8 16.Bg4! Qb6?!
(16...Nc5 17.e5!) 17.Bxd7 Qxd4 18.Rxd4 Bxd7
19.e5±.
b) Black awaits Qd2 with 7...0-0 8.Be2 d6
9.0-0 Bd7. Although I pay ample attention to
10.Qd2 Nxd4 11.Bxd4 Bc6 12.f3 a5 13.b3 Nd7
14.Bf2! in the “Step by Step” section, I recommend
to counter-exploit Black’s “exploitative” strategy by
saving the knight:
10.Nc2!. The point is that Black’s best retort
to Nc2 is the attack on c4 with Be6, but he lost
already a tempo on ...Bd7.
Nd5 in the middlegame
In the Maróczy Bind White can always play
Nd5. All the engines encourage this move and give
generous evaluations, especially Stockfish. They
overweigh White’s space advantage, but they miss
the fact that White is unable to break through.
Sometimes our advantage could look
enormous, and still it could be insufficient for
cashing a full point.

Analysis

In practice Black hates to stay totally passive

and often seeks “counterplay” with 18...e6?!, but
that only helps White.

The more minor pieces White retains, the

I did not find a way to break on the kingside
here. 30.g4 may become even double-edged after
30...fxg4 31.hxg4 b5!.
The maximum White could get is to push b4:
Analysis
higher his winning chances are. For instance, if he
kept the dark-squared bishops, he could try to
combine play on both flanks. Also his chances on
the kingside would be somewhat better than in
positions with only a B:N, especially if Black
provided a lever:
Fusco-S.Alonso
Buenos Aires 2006

23.g3², h4, Bh3, g4.

28.Re6 Nf6 29.Rg3, and it is still insufficient
after 29...Ng4. However, in the general case without
weaknesses Black should be able to hold:
Carlsen-Lie
Drammen 2004 M.Yilmaz-Vocaturo
Batumi 2018

94
 Almost all tournament and correspondence
games finished in a draw after 17...Ne4 18.Be3.
I believe that White has more practical
chances to win if he takes with the c-pawn, because
that widens the scope of his bishop. Of course, that
would make sense if Black had weakened his light
squares and White could penetrate through the
c-file:
georgedimitrov-boeien
net-chess.com 2014
Afer 24...Bxd5 25.cxd5 White would have
been clearly better because of the weakness of c6
and b6.
Khalifman-M.Petrov
Porto Carras 2018
95
Here the pawn is on b7, so 17...Rfc8 retains
good drawing chances. Khalifman traded all the
heavy pieces and won the ending with his bishop
pair, but only with the help of his opponent.
My final verdict is that we should play Nd5
mostly if we have another knight. Thus I
recommend 10.Nc2! after 9...Bd7. Without knights,
I would prefer to manoeuvre around and await a
mistake, as in Game 18 Kurnosov-Maze, Biel 2012.
Break in the centre
Our best chances to win are based on a
timely break in the centre with c5 or e5. Stay sharp
and do not miss it when an opportunity appears!
Grandelius-Bejtovic,
Sweden 2015
19.e5! dxe5 20.fxe5².
 Here is a more complex example:

Analysis

16.e5 Re8 17.exd6 (17.Bg4±) 17...exd6

18.Qxd6² (18.Bf3±).

22.e5! dxe5 Rahman-Gillani

22...Ne8 23.Nd5 Qd8 24.Bxb6+– New Delhi 2009
23.Bxc6
23.fxe5 Nxe5 24.Nb4 – a double hit on a6
and d5, 24...Nxf3+ 25.Qxf3 Qb7 26.Ncd5±
23...Bxc6 24.Nb4 exf4 25.Bxf4 Qb7
26.Bxb8+–.

Hracek-Roganovic
Hungary 2015

23.c5!+– Qa7 24.e5 dxe5 25.fxe5 Nh7 26.e6.

Analysis

17.e5! f5 18.Be2 Ne6 19.Qh4±.

Ragger-Bezold
Munich 2016

96
 18.c5! dxc5 19.Qxc5 Qb8 20.e5±.
And here is a final example where White just
strangles the opponent in the centre:
Mareco-A.Hoffman
Montevideo 2017
21.f4 Bc6 22.a3 b6 23.Re3 Qf6 24.b4 axb4
25.axb4 Na4 26.Bxa4 Bxa4 27.Rd3±. The d6-pawn
is doomed.
Theoretical status
The latest book (published in 2018) on the
Accelerated Dragon is written by Panjwani.
He covers two plans for Black. I do not pay
much attention to the Breyer Variation 7...Ng4
8.Qxg4 Nxd4 9.Qd1 e5 10.Bd3 d6 11.0-0 0-0
because White has an obvious advantage and
can press for a win with normal positional moves as
b3, Ne2 etc.
97
Panjwani’s main recommendation is 7...0-0
8.Be2 d6 9.0-0 Bd7. I studied this variation several
years ago for my students, and came to the
conclusion that White should preserve his knight
with 10.Nc2!. Then he should consolidate and make
a few prophylactic moves as 10...Rc8 11.Rc1! Ne5
12.b3 Qa5 13.Bd2!
13...Qb6 14.Kh1 Bc6 15.f3². White has
preserved all his pieces and Black will soon become
to wriggle in the Maróczy binds.
As a whole, my own practice with both
colours has convinced me that playing this line as
Black is somewhat masochistic. Tournament results
confirm my assessment.
 Chapter 5. Maroczy Bind
Step by Step
1.c4 c5 2.Nf3 g6
I mentioned in the previous chapter that we
meet 2...Nc6 by 3.e3. Thus the move orders with
2...g6 or 2...Nf6 3.Nc3 g6 4.d4 are the only paths to
the Maróczy Bind.
3.e4!
Perhaps this is the cleanest way. 3.d4 might
face 3...cxd4 4.Nxd4 Nf6 5.Nc3 d5 6.cxd5 Nxd5
7.Ndb5 Nxc3. The endgame may be slightly in
White’s favour, but why to trade queens so early if
we have better options.
A. 3...Nc6; B. 3...Bg7
A. 3...Nc6 4.d4 cxd4 5.Nxd4 Nf6 6.Nc3 d6
6...Nxd4 7.Qxd4 d6 8.Be2 transposes.
However, White may opt to develop the bishop on
the more active square d3:
8.Bg5 Bg7 9.Qd2 Be6
Or 9...0-0 10.Bd3 a5 11.0-0 a4 12.b4².
10.Rc1 0-0 11.Bd3 Qa5 12.0-0 Rfc8 13.b3
a6 14.Rfe1
98
White has achieved a more active set-up than
in the lines with Be2, Be3. The Bg5 and the Re1 are
eyeing the e7-pawn.
14...Rc5
14...b5 is not a good idea owing to 15.Nd5
Qxd2 16.Bxd2 Bxd5 17.exd5 bxc4 18.Bxc4 Kf8
19.b4±.
15.Be3 Rc6 16.h3 Nd7. We have been
following the game Ivanchuk-Tiviakov, Benidorm
2008. White could gain the edge with 17.f4! b5
18.f5 bxc4 (17...Nc5 18.Bb1 b5 19.f5 Bd7 20.Bd4±)
19.Bb1 cxb3 20.fxe6 fxe6 21.Ne2².
7.Be2
7.Be3?! stumbles into 7...Ng4 8.Nxc6 Nxe3
9.Nxd8 Nxd1 10.Rxd1 Kxd8 11.c5 Be6.
7.Nc2 is less efficient than in line B1 since
Black has not played ...Bd7 and could lead out the
bishop to e6. White still obtains excellent results
after 7...Bg7 8.Be2 0-0 9.0-0 though.
7.f3 is the other normal move, but Black then
gets an improved version of line B2, since White
had to play f3 too early:
7...Nxd4
Also interesting is 7...Bg7 8.Be3 0-0 9.Be2
Nd7!?, since 10.Nc2 Bxc3+ 11.bxc3 Nc5 is rather
unclear.
8.Qxd4 Bg7 9.Be3 0-0 10.Qd2 Be6
10...Bd7?! 11.Be2 Bc6 leads to line B2 with
a tempo less.
11.Rc1 Qa5
 White looks better, but concrete analysis
reveals enough defensive chances – see Game 19
Ivanchuk-Kiril Georgiev, Belgrade 1997.
7...Nxd4!
Taking on d4 before White had time for Be3
or Nc2. The delay of this exchange leads to lines I
cover in B1 and B2.
8.Qxd4 Bg7 9.Bg5
This is the only reasonable try to exploit the
early exchange on d4.
After 9.Be3 0-0 we should decide where to
retreat the queen:
a) 10.Qd2 Be6
10...Nd7 11.0-0 Nc5 12.f3 a5 is another
thematic plan.
11.0-0 Qa5 12.Rab1 Rfc8 13.b3
99
13...Ng4!
After 13...b5 14.b4 Qc7 15.Nxb5 (15.c5 a6)
White’s queenside pawns are more dangerous –
15...Qd7 16.Rfc1 Nxe4 17.Qd1 a5 18.a3.
14.Nd5 Qxd2 15.Bxd2 Kf8=.
b) 10.Qd3, leaving d2 free for the bishop,
would have been the perfect place if Black had not
10...Nd7!
The plan with 10...Be6 11.0-0 Qa5 is
insufficient. (11...Nd7 12.b3 Nc5 13.Qc2 a5 is
passive – 14.Rfd1 Bd7 15.f3 Rc8 16.Rac1 b6
17.Qd2 Bc6) 12.Rac1 Rfc8 13.b3 a6 (13...Nd7
14.Bd2! reveals the main idea behind 10.Qd3. White
repels the Black queen and launches an attack after
14...Nc5 15.Qe3 Qd8 16.f4 or 16.h4!?) 14.Bd4 Rab8
15.Qe3 Nd7 16.Kh1 b5 17.f4 Bxd4 18.Qxd4 b4
(18...f6 19.cxb5 axb5 20.b4 Qa3 21.Rc2²) 19.f5 Qe5
20.Qxe5 Nxe5 21.fxe6 bxc3 22.exf7+ Kg7
23.Rxc3².
11.Bd4
Or 11.f3 Qa5 12.Rc1 b6 13.0-0 Bb7 14.Rfd1
Rac8 15.Qd2÷.
11...Nc5 12.Qe3 Bxd4 13.Qxd4 e5! 14.Qd2
Ne6=, Gipslis-Giorgadze, URSS 1972.
9...0-0
 15...b5 stumbles into the typical for the Bg5
set-up 16.f5² Bd7 17.fxg6 hxg6 18.e5.

16.Qe3

10.Qe3

10.Qd2 commonly transposes.

10...Be6 11.0-0 Qb6 12.Qd2 Rfc8 13.b3 a6 16...Re8

14.Rac1 Qa5

White discouraged 16...b5 in view of 17.e5.

The game Bu,Xiangzhi-Zvjaginsev, Sochi
2016, saw 14...Qb4 15.f4 b5 16.f5 Bd7 17.fxg6
hxg6.
The consistent 16...Rac8, which revives the
threat ...b5, is met by 17.Rb1 Qd8
17...R5c6 18.f5 Bd7 19.Kh1 Qe5 20.Qf3
gxf5 21.exf5 Rc5 22.Bxf6 exf6 23.Qxb7 Bxf5
24.Rbd1².
18.a4 a5 19.Bh4! (to prevent the defensive
sacrifice ...Rxg5) 19...Nd7 20.h3±.

17.h3

17.Bh4 is also slightly better for White. I (as

Black) beat Almasi with 17...Bh6, but 18.h3! b5
19.g4 would have been awkward. A better retort to
17.Bh4 is 17...Rcc8, when 18.Kh1 allows 18...b5,
and 18.Be1 – 18...Qc5².

Here the Bg5 reminded of himself, enabling 17...b5

the decisive shot 18.e5! Qc5+ (18...dxe5 19.Bxf6)
19.Kh1 Qxe5 20.Bf3 b4
Or 20...bxc4 21.Bxa8 Rxa8 22.Rce1 Qc5 In comparison to the previous comment,
23.Rxe7±. White has an extra luft – 17...Rcc8 18.Kh1 b5 19.f5
21.Nd5 Nxd5 22.Bxd5+–. Bd7 20.Bxf6 exf6 21.Qa7±.

15.f4 Rc5 18.e5 Nd7 19.Ne4 Rcc8 20.cxb5 Rxc1

21.Rxc1 dxe5 22.bxa6 h6 23.Bh4 exf4 24.Qd2

100
 The position is very sharp, but White’s pawn
look dangerous.

B. 3...Bg7 4.d4 cxd4

Pay attention to some tricky lines, which

have a common feature – Black is trying to snatch a
pawn at the cost of development. They should be
punished with energetic play involving sacrifices:
4...Qa5+ 5.Nc3 Nc6
5...d6 6.d5 Bxc3+ 7.bxc3 Nf6
9.c5!? This line requires a little
memorization indeed, but the effort is worth it. Still,
the positional 9.Bd3 is a decent alternative.
9...dxc5 10.Bf4 Kf8 11.Rc1 a6 12.Bc7 Qf6
13.Na3 b5 14.Bd3!
Ivanchuk chose against Topalov in 2000
14.Rxc5, when 14...d3! 15.Bxd3 Qxb2 16.Qb3!
Qxb3 17.axb3 is only marginally better.
14...c4 15.Bxc4 bxc4 16.Nxc4

8.Qc2! Nxe4 9.Bd3 The pawn does not

compensate our space advantage – 9...Nf6 (9...f5
10.0-0) 10.0-0 0-0 (10...Bg4 11.Nh4 0-0 12.Bg5±)
11.Bh6 Re8 12.h3±.
6.d5 Nd4 7.Bd2 d6
7...Nxf3+ does not pose any problems –
8.Qxf3 d6 9.Be2 Nf6, when any decent move offers
a clear advantage – 10.Qd3±, 10.h3±, 10.0-0±.
8.Nxd4 cxd4 9.Nb5 Qd8 10.Bb4! (10.Bc1
transposes to 4...d6, which is also in White’s favour) Black is suffocating.
10...Nf6 11.e5!±. A correspondence game went 16...Qg5
17.0-0 Nf6 18.Nb6 Bh3 19.Bg3 Bg4 20.h4, and
4...d6 5.Nc3 (Of course 5.d5 is also good.) Black did not survive. I also checked:
5...Nc6 6.d5 Nd4 7.Nxd4 cxd4 8.Nb5 Qb6 16...Bh6, when both 17.Rc2 Bf4 18.Bb6 Qg5
19.g3 Bh3 20.Qxd4 and 17.0-0 Bxc1 18.Qxc1 are
unpleasant for Black.

4...Qb6 5.dxc5! Qxc5 6.Be2 d6 7.0-0 Nc6

8.h3 Be6 9.Be3 Qb4

101
 11...Be6 transposes to the main line after
12.Nd4 Nxd4 13.Bxd4, but White has also 12.0-0
Rc8 13.Rac1.
12.Na4!?
This is the only way to exploit the early
sortie of the enemy queen. The trivial approach is
12.0-0 a4 13.Nd4 Nxd4 14.Bxd4 Qa5.
12...Nd7
12...Be6 is less provocative –
13.0-0 Nd7 14.Nd4 Nxd4 15.Bxd4 Bxd4+ 16.Qxd4
Nc5 17.Nc3.
After the text, besides 13.0-0, White could
also try 13.0-0-0!±.

10.Nbd2 Qxb2 11.Rb1 Qxa2 12.Ng5 Nf6 7.Nc3

13.c5! 0-0 14.Ra1 Qb2 15.Nxe6 fxe6 16.Nc4+–.

5.Nxd4 Nc6 6.Be3 Nf6 7.f3 could face 7...Qb6!? 8.Nf5 Qxb2 with
unorthodox play. In practice White does not cope
well.
A rare sideline is 6...Qb6 7.Nb3 Qd8.
Black’s idea is take on c3 should we lead out our 7...0-0
knight. We have to play 8.Qd2, losing control of g4.
Then 8...Nf6 forces the committal 9.f3 which might 7...Nxd4 8.Bxd4 d6 9.Be2 0-0 10.0-0 Bd7
not have been in White’s plans. Perhaps that could 11.Qd3! transposes to the main line.
explain his unsatisfactory results in tournament
chess. Surprisingly, even in correspondence and 7...Ng4 8.Qxg4 Nxd4 9.Qd1 e5 10.Bd3 d6
computer chess White cannot prove an edge. Let’s 11.0-0 0-0 is a second-rate variation with obvious
delve deeper: positional flaws.
9...0-0
9...a5 is premature, because the b3-knight
goes for a4 to underline the weakness of b6 –
10.Nc5 a4 (10...d6 11.Na4) 11.Nc3 a3 12.b3 d6
13.N5a4±.
10.Nc3 d6 11.Be2

Panjwani devoted to it no less than 8

chapters, but in my opinion they are just a
make-weight to his book.
The latest trend for White is to grab space on
the queenside with 12.a4, intending to push a5, Qa4.
A key moment is that Black cannot prevent it, as
11...a5 12...a5 would be met by 13.Nb5 Bd7 14.Bxd4±. So

102
Black commonly answers 12...Be6 13.a5 f5 (Or
13...Qd7 14.Qa4 Qxa4 15.Rxa4 Rfc8 16.Nd5 Bxd5
17.exd5, Saric-Navara, Batumi 2018. White is
certainly in control, and he won the game, but I
think that Black has very good chances to hold the
draw.) 14.exf5 Nxf5 15.Qe2
15...Kh8! (15...Nxe3 16.Qxe3 Rc8 17.Qxa7
Rf7 18.Be4² leaves White in most lines with a
dominating knight over a dark-squared bishop.)
16.a6. All this occurred in Navara-Plichta, Poland
2018. Now instead of the positional capitulation
16...b6?!, Black should have played 16...Nxe3
17.Qxe3 Rb8! with fair chances to equalise, e.g.
18.Rfb1 e4! 19.Nxe4 d5.
The central approach is more unpleasant for
Black:
12.Qd2 Be6 (12...f5 13.exf5) 13.Rad1 a6
14.b3 Rc8 15.Ne2
15...b5
15...Nxe2+ 16.Bxe2 Rc6 17.a4 is rather dull
for Black.
16.cxb5 Nxb5
103
White is clearly better in this position. He
has 2:1 pawn majority on the queenside and pressure
down the d-file. He could force play with 17.Bxb5
axb5 18.Nc3 Qa5 19.Nd5 Qxd2 20.Rxd2 Rfe8 21.a4
bxa4 22.bxa4 f5 23.a5², but it is even better to start
rolling the pawns immediately:
17.b4! Nc7 18.a3 f5 19.Bb6². Black has less
space and two targets on a6 and d6.
8.Be2 d6
After 8...b6 White could choose the active
Sicilian set-up with:
9.f4 Nxd4 10.Bxd4 d6 11.0-0 Bb7 12.Bf3
and the quieter:
9.0-0 Bb7
10.Nxc6
10.f3 e6 11.Qd2 d5 12.Nxc6 Bxc6 13.cxd5
exd5 14.e5 Ne8 15.f4 f6 16.exf6 Nxf6 17.Rad1 is
too modest albeit safe.
10...Bxc6 11.f3 e6 12.Qd6 Ne8 13.Qd2 f5
14.exf5 gxf5. In general, this pawn structure is
slightly better for White owing to the open d- and
e-file. I like the idea of putting pressure on the
queenside with:
15.a4 Be5
Or 15...Qe7 16.Nb5 e5 17.Bd3 Qe6 18.Rad1
Rf6 19.b3.
16.b4
It would be difficult to win after 16.Kh1 Qh4
17.f4 Bxc3 18.Qxc3 Nf6.
16...Qf6 17.Ra3² or Rc1², planning a4-a5.
9.0-0 Bd7
 Black’s best plan is based on ...a7-a5-a4,
...Qa5. However:
9...a5 is too early owing to 10.Ndb5 or 10.f3
Nd7 11.Ndb5. Also premature is:

9...Nxd4 10.Bxd4 Bd7 because after 11.Qd3!

B1. 10.Nc2!; B2. 10.Qd2

B1. 10.Nc2! Rc8

I do not know why the plan with 10...a5 is

only 4th most frequent here. It is hardly worse than
Black fails to reach the set-up ...Qb6+Nc5. 10...Rc8. Perhaps Black is afraid of the possibility
The reason is that 11...a5 12.b3! (to meet ...a4 by b4) b2-b4 in some lines (after Qd2). After 10...a5 White
12...Bc6 does not force f3 (e4 is protected by the has two distinct approaches:
queen!), and White gets a vital tempo for 13.Rad1
Nd7 14.Bxg7 Kxg7 15.Qd4+ Kg8 16.Bg4!. It now a) 11.f4 a4 12.Na3 Qa5 13.Kh1! Rfd8
transpires that 14.Qe1
16...Qb6?! stumbles into a pin down the
d-file – 17.Bxd7 Qxd4 18.Rxd4 Bxd7 19.e5±.
The same is 16...Nc5?! 17.e5. That means,
Black has to return the knight to f6, but White then
launches his own attack on the kingside:
16...Nf6 17.Bh3 Qb8 18.f4 (or 18.Qd3)
18...Qa7 19.e5 dxe5 20.fxe5 Ne4,
Grandelius-Bejtovic, Sweden 2015. Here, instead of
21.Nd5, best is 21.Nxe4 Bxe4 22.Qxa7 Rxa7
23.Rd7±.

Black may opt for a passive stand, but that

does not yield him decent results: 11...Bc6 (instead
of 11...a5) 12.b4 b6 13.Rad1 Rc8. White’s main
plan should be f4, but after a solid preparation. For
instance:
14.f4 could be countered with 14...e5. So we
play first:
14.f3 and wait to exchange the dark-squared
bishops.
Another plausible plan is to retreat the
bishop first: 13.Be3 Rc8 14.Rac1 Nd7 15.f4.
104
White is planning to take as much space as
possible. The only way to counter it is to push ...f5,
but:
14...Ne8 15.Rd1 f5 16.Rd5 cuts off the
queen from the kingside. The attack on c4 is also
ineffective:
15...Qb4 16.Nab5 Na5 17.Nd5 Qxe1
18.Rfxe1 Nc6 19.Bb6 Rdb8 20.Nbc7 Nxc7 21.Nxc7
Bxb2 22.c5 with an extra exchange.

b) It is less committal to trade queens, as I
played once:
11.f3 a4 12.Qd2
12.Na3!? Qa5 13.Nd5 may be even stronger.
12...Qa5 13.Rfd1 Rfc8 14.Na3 Be6 15.Nd5.
Without queens Black lacks any counterplay. See
the instructive Game 17 Kiril Georgiev-Strikovic,
Marrakesh 2010.
10...a6 is commonly met by:
11.f3 Rc8 12.Qd2 Ne5 13.Na3 (I beat Nenad
Ristic with 13.b3?! b5 14.cxb5 Qc7? 15.Bd4 Nxf3+
16.Bxf3 e5 17.b6, but 14...Qa5! 15.Bd4 Bxb5 is
roughly equal.) 13...Be6, when 14.Nd5 Bxd5
15.exd5 e6 16.dxe6 fxe6 17.b4 d5 18.c5 is only
marginally better.
I suppose we should play more energetically:
11.Rc1 (or 11.f4 Rc8 12.Rc1) 11...Rc8 12.f4!
Although 11.f4 yields fantastic results, I do
not like to give conterplay with 11...Ne8! 12.Bd2
f5!.
A little prophylaxis does not hurt, so 11.Kh1,
as Ivanchuk and Hracek played, should not be bad.
Then 11...Be6 12.f4! Rac8 13.Bd2!² forced Black’s
queen back.
I like the idea of driving the queen back, but
prefer to achieve it by:
11.Rb1!? Rfc8 (a correspondence game saw
11...Be6 12.f4 Rfc8 13.a3 Qd8 14.f5 Bd7 15.g4ƒ)
12.f4 Qd8

13.h3. White’s intention is to grab as much

space as possible. If Black tries 13...a6 14.a4 Na5,
we defend everything with 15.Na3 Nc6 16.Qd2. We
have consolidated and we’ll soon be able to go
forward – 16...Rab8 17.b4 e5 (17...a5 18.bxa5)
18.fxe5 dxe5 19.a5².

Now 12...Na5 13.b3 b5 is plain bad owing to

14.e5 Ne8 15.Nd5±.
12...b5 is an interesting gamble for blitz, but
Black does not have full compensation for the pawn
after 13.cxb5 axb5 14.Bxb5 (14.a3²) 14...Ng4 15.a4
Nxe3 16.Nxe3 Nd4 17.Bxd7.
12...Qa5 13.a3 is good enough, but even
stronger is 13.b4 Nxb4 14.Nxb4 Qxb4 15.e5±.
Thus it remains the waiting:
12...Re8 13.Kh1 Kh8 which occurred in the
correspondence game Volodarsky-Gonzaga Grego,
ICCF 2017 – 14.Nd5 Nxd5 15.exd5 Nb8 16.Bd4 e6
17.dxe6 Bxe6 18.Ne3 Nc6 19.Bxg7+ Kxg7 20.f5
Bd7 21.Rc3². Black even managed to survive, but it 11.Rc1!
is easy to take sides here.

10...Qa5 is not a good idea either. Panjwani recommends as slightly better for
105
White 11.f3 Be6 12.f4 (12.Qd2 Nd7 13.Rac1 Re8)
12...Qa5 13.Rc1 Qa6 (13...h5 14.h3 Nd7 15.Rb1)
14.b3 (14.Na3 Nxe4 15.Nxe4 Bxb2 16.Nb5 Bxc1
17.Bxc1 Qxa2 18.Nbc3 Qa5)
14...Ng4 15.Bxg4 Bxg4 16.Qd2 (16.Qxg4
Bxc3) 16...Bd7 17.f5 Qa5 18.Nd4 Rfe8 19.Rcd1 a6
20.a4 Ne5 21.h3. I find all these moves unnatural,
and especially the idea of playing f2-f3-f4 in two
moves.
11...Ne5
11...a6 12.f4 transposes to 10...a6.
12.b3 Qa5
12...a6 offers a pleasant choice between:
13.Nb4 Neg4 14.Bf4 Ne5 15.Qd2 Nc6
16.Nbd5 Nxd5 17.Nxd5² and:
13.Nd4 Nc6 (13...Neg4 14.Bd2 Qb6 15.Nc2;
13...Qa5 14.Nd5 Nxd5 15.exd5 e6 16.dxe6²) 14.Qd2
Qa5 15.Rfd1².
12...b5 is dubious in view of 13.Nxb5 Bxb5
(13...Nxe4 14.Nxa7 Ra8 15.a4²) 14.cxb5 Nxe4
15.Nd4 Qd7 16.f3². The c6-square will be a fantastic
outpost for our knight.
13.Bd2 Qb6
106
14.Kh1 Bc6 15.f3²
White has neutralised all Black’s stabs and
stands obviously better. The game
Delchev-Bit-Narva, Hersonissos 2018, went further:
15...e6 16.Be3 Qc7 17.Nd4 a6 18.Qd2 Rfd8
19.Nxc6
19.Bg5 Bd7 20.Rfd1 Qa5 21.f4 Nc6 22.Nc2
h6 23.Bh4 Be8 24.h3 Qc7 25.Bf3 Qa5 26.b4 Qc7
27.a4± is a fair alternative.
19...bxc6
19...Nxc6 20.Nd5 exd5 21.cxd5 Qe7
22.dxc6±.
20.f4 Ned7 21.Bf3± e5, when strongest is
22.f5! with a strategically won position.
Another option is 15...a6 16.Nd5 Qd8
16...Bxd5 17.exd5 e6 (or 17...Ned7 18.g4 e5
19.dxe6 fxe6 20.Be3 Qc7 21.g5 Nh5 22.f4) 18.f4
Ned7 19.Be3 Qc7 20.dxe6 fxe6 21.Bf3 Kh8 22.g4
e5 23.f5±.
17.Be3 Ned7 18.Qd2 b6 19.Rfd1 Bb7
20.Ncb4
 13.b3

If Black succeeds in playing ...a4, ...Qa5, he

would have a decent game.

13...Nd7

20...a5 21.Nxf6+ Bxf6 22.Nd5 Bg7 23.Bg5

Re8 24.Bf1 Bc6 25.g3 Nc5 26.Bh3².

B2. 10.Qd2

A major alternative is the plan without f3. It

assumes that White will defend the e4-pawn with
Bd3, Re1, and will change later the pawn structure
with Nd5. It is easy to understand, safe, and assures
White of long-term pressure. However, I believe that
if Black discovers the best defensive stand, he
should be able to hold on.
10...Nxd4
10...Ng4 11.Bxg4 Bxg4 is positionally
dubious, as White gets rid of his only bad piece. We
put a knight on d5, b3, Rac1, and look for a break.
That could be c4-c5. As a rule, we should be happy
to trade our remaining bishop and a knight to reach a
N:B position.
14.Bf2!
A small, but important detail. The main
move has been 14.Be3 Nc5 15.Rab1, when
Panjwani recommend15...e6!? 16.Rfd1 Be5,
threatening ...Qh4 in some lines. Another argument
in favour of Bf2 is that the bishop would not be
hanging in the event of a future break ...d6-d5. A
notable protagonist of this retreat is Kramnik.
14...Nc5
A rare alternative is 14...f5 15.exf5 gxf5

11.Bxd4 Bc6 12.f3 a5

12...Nd7 13.Bf2 transposes.

12...Qa5 provokes b2-b4 – 13.a3 Rfc8 14.b4

or 13.Rab1 Rfc8 14.b4 Qd8 15.Rfd1, which is in
White’s plans anyway. In the latter line White saves
a3, since 15...a5 would be met not by 16.a3, but by
16.b5 Be8 17.f4.

107

We have met this structure when we
discussed 8...b6 instead of 8...d6. Here the Black
pawn on a5 prevents the recommended plan with a4,
b4, a5, but it will be a cause of concern for Black in
future. We should put our pieces in the centre and
keep manoeuvring:
16.Rad1 Kh8 (16...e6 17.Na4) 17.Nd5 e5
18.Kh1².
15.Rab1
The rook should support a3 and b4, although
this plan depends on Black’s set-up and may be
cancelled if he opened the centre.
15...e6
The only way to obtain counterplay on the
kingside. Black often chooses a waiting game by
leaving the queen on the queenside:
15...Qb6 16.a3 Qd8 17.Bd1 b6 18.Bc2 Qb8
19.Rfc1 Rc8 20.b4² (Panjwani).
Similar is 15...Qc7 16.Rfc1 Rfc8 17.Bd1
Kramnik chose 17.Nd5 Bxd5 18.Qxd5, but
18...e5!, followed by ...Nc5-e6-d4 would be difficult
to break.
17...Qd8 18.a3. See Game 18
Kurnosov-Maze, Biel 2012.
15...Be5 hints for aggressive intentions with
...f5, so we may alter our plans and keep our rook on
the kingside – 16.Rbd1, although 16.Rfd1 f5 17.exf5
Rxf5 18.Be3 is not bad at all.
108
16.Rfd1
Mareco won two games with 16.Bd1!?. This
manoeuvre is primarily meant to prepare b4, but it
might also support f4-f5. In a similar position
Panjwani recommends 16...Qc7 (16...Qe7 17.Bc2
Rfd8 18.Rfe1) 17.Bc2 Rfd8 18.Rfd1 Be5, when:
19.a3 a4! does ensure counterplay against the
c4-pawn. However, the absence of the queen from
the kingside may encourage White to change plans
with:
19.Re1!?, preparing f4-f5. For instance:
19...b6 20.f4 Bg7 21.f5 Qe7 22.Bd4, or 20.Rbd1
Rac8 21.f4².
16...Be5
17.Bf1!
Strangely, in his highly detailed book on the
Hyper Accelerated Dragon Panjwani ignores this
prophylactic move, which occurred in 9 games.
Instead he only mentions the rare 17.Nb5 Bxb5 with
a tenable position.
In order to understand White’s approach, we
should elaborate a little on the position on the last
diagram.
White’s main advantage is not that he has a
good plan, e.g. a3,b4 or g3, f4. His biggest asset is
space. As long as he preserves the grip on the centre,
his chances will always be slightly better. Thus his
primary task should be to anticipate any
counterplay. The moves Bf2 and now Bf1 are a step
in that direction. Black’s only activity could be
based on ...f5, but it would weaken his king. Thus
White is not afraid of spending tempi on
prophylaxis, and even on several other
noncommittal moves like Nc3-e2-d4. It is more
difficult to find such “free” moves for Black.
On the contrary, it was early to play 17.a3
owing to 17...Qf6 18.Bd4 (Marin erred with
18.Nb5?, when 18...Bxb5 19.cxb5 a4! would have
grabbed the initiative.) 18...Bxd4 19.Qxd4 Qxd4+
20.Rxd4 Rfd8. Then Black plays ...e5, ...Ne6, and
the b3-pawn is obviously weak.

17...Qf6

This move deprives Black of counterplay

with ...f5, but it is ineffective anyway:
17...Qe7 18.Ne2 Rfd8 19.Nd4 Be8 (19...d5
20.exd5 exd5 21.Qe3±) 20.Re1
20.a3 is also possible – 20...a4 21.b4 Nb3
22.Nxb3 axb3 23.f4 Bf6 24.b5 b2 25.Bb6 Rd7
26.Qe3 Bg7 27.Rd2².
20...Bg7 21.a3.

18.Bd4

White is slightly better. He has a clear plan

to produce a passer on the queenside. The game
Yemelin-Suran, Czech Republic 2016, went:
18...Rad8 19.a3 Qe7 20.b4 axb4 21.axb4 Bxd4+
22.Qxd4 e5 23.Qe3 Na4 24.Nd5±. Another
possibility is:

18...Qe7 19.a3 Rfd8 20.b4 Na4

A key point is that 20...Bxd4+ 21.Qxd4 e5
22.Qd2 Ne6 fails to 23.b5 Be8 24.Nd5±.
21.Bxe5 dxe5 22.Qe3 Nxc3 23.Qxc3 Ba4
24.Rd2 Rxd2 25.Qxd2 b6 (25...Rd8 26.Qb2)
26.Qe3². White has an extra “half pawn”.

109
 Chapter 5. Maroczy Bind is already strong:
Annotated Games

17. Kiril Georgiev – Strikovic

Marrakesh 27.06.2010

1.d4 d6 2.Nf3 Nf6 3.c4 g6 4.Nc3 Bg7 5.e4

0-0 6.Be2 c5 7.0-0 cxd4 8.Nxd4 Nc6 9.Be3 Bd7
10.Nc2 a5 11.Qd2 a4 12.f3 Qa5 13.Rfd1 Rfc8

19...f5 20.exf5 gxf5 21.g5 Bxd5 22.cxd5

Nb4 23.Bb5±.
17.cxd5 Nb4 18.Bb5 Ne8 19.g3 Kf8 20.Bf1².

16.Qxa5 Nxa5 17.cxd5 Nd7 18.Rab1 Nc5

19.Bb5

14.Na3

Morozevich likes this treatment. The

alternative approach is to push b4 – 14.Bf1 Be6
15.b4 Nxb4 16.Nxb4 Qxb4 17.Rab1 Qa5 18.Bb6
Qh5 19.Rb5 Bh6 20.Qe1. White’s rook on b5 makes
the difference here – 20...g5 21.Bd4 Bg7 22.Nd5
Bxd5 23.exd5 g4 24.Be2 gxf3 25.Bxf3 Qg5
26.Rxb7 Rxc4². Time to take stock. The Black queenside
pawns are both weak, and he lacks any counterplay.
14...Be6 15.Nd5 Bxd5 As a rule, this pawn structure with cxd5 gives White
considerable winning chances if he has a knight.
Otherwise exd5, with the idea of b4, c4-c5, would
It is more precise to exchange the queens on be more promising.
d2 – 15...Qxd2 16.Rxd2 Bxd5
16...Nd7 17.g3 (Black’s only counterplay is 19...Kf8 20.Kf1 b6 21.Rd2 h5 22.g3 Bf6
based on ...f5, but 17.g4 f5 18.exf5 gxf5 19.g5 is not 23.h4 Kg7 24.Ke2 Ra7 25.Rc1 Rb8 26.Rcc2 Kf8
good since Black has 19...Bxd5 20.cxd5 Nb4 27.Bd4
21.Bb5 Nf8. We should wait for Black to shift his
knight to the kingside first.) 17...Ra5 (17...f5 18.exf5
gxf5 19.Rb1 Bxd5 20.Rxd5 f4!? 21.gxf4 Nc5 I have fixed the opponent’s pawns on light
22.Nc2 Rf8 23.Rh5²) 18.Kg2 Nc5, and now 19.g4! squares, so it should be good to trade the other
110
bishop.
27...Kg7 28.Ke3 Bxd4+ 29.Rxd4 Nab7
30.Rb4 Kf6
31.f4
This move is positionally correct, but 31.Nb1
Kg7 32.Nc3 was more direct. Now the only way to
not lose a pawn is 32...a3, when 33.b3, followed by
Rd4, b4 should win easily.
31...Kg7 32.Be2
Curiously, Megabase claims that I blundered
a piece with 32.Bd7?? and Strikovich did not notice
to take it! My scoresheet tells another story:
32...Na5 33.Bb5 Nab7 34.Nc4 Na5 35.Nd2
a3 36.bxa3
Of course 36.b3 is also possible, but the
game had a relatively short control at 45 minutes
without any seconds per move, and I decided to play
it safe, avoiding any future traps with ...Nxb3.
36...e6 37.dxe6 fxe6 38.Rd4 Rd8 39.a4 Kf7
40.Nf3 Ke7 41.Rcd2 Raa8
111
42.Rb2
42.f5! gxf5 43.exf5 exf5 44.Rd1! (44.Rf4
Ne4) would have won– 44...Ne4 (44...Nab7
45.Rd5+–) 45.Kf4.
42...Rac8 43.Rdb4 Rc7 44.Rd2 Rcc8
45.Ng5 d5 46.Rbd4 dxe4 47.Rxd8 Rxd8 48.Rxd8
Kxd8 49.Nxe4 Ke7 50.Kd4 Ncb7 51.Kc3 Nd6
52.Nxd6 Kxd6 53.Be8 1-0
My light-squared bishop has finally eaten the
pawns which I fixed thirty moves ago!
18. Kurnosov – Maze
Biel 27.07.2012
1.e4 c5 2.Nf3 Nc6 3.d4 cxd4 4.Nxd4 g6 5.c4
Bg7 6.Be3 Nf6 7.Nc3 0-0 8.Be2 d6 9.0-0 Bd7
10.Qd2 Nxd4 11.Bxd4 Bc6 12.f3 a5 13.b3 Nd7
14.Bf2 Nc5 15.Rab1 Qc7 16.Rfc1 Rfc8 17.Bd1
Qd8 18.a3 Nd7 19.Be2 Nc5 20.Bd1 Nd7

21.Ne2
I chose this game for it offers a good idea of
possible plans in a waiting game. White’s previous
moves prepared b3-b4, and the only sensible try now
was to push the b-pawn. However, White suddenly
decides to manoeuvre aimlessly, without committing
himself with a plan. This is possible of course, but it
would have been better to leave the pawn on a3, to
avoid the weakness on b3. Let’s investigate 21.b4
axb4 22.axb4 Nb6
22...Ne5 invites f4 – 23.Be2 Ra3 24.f4
24...Nxc4 25.Bxc4 Rxc3 26.Rxc3 Bxc3
27.Qxc3 Bb5 28.Bxf7+ Kxf7 29.Qh3 Kg8 30.f5ƒ.
23.Bb3 Be8 24.Ne2 Ba4 25.Bxa4 Nxa4
26.Rc2 Nb6 27.c5. White retains a slight pull, but
nothing tangible.
21...b6 22.Nd4 Bb7 23.Be2
112
White’s bishop does not have any business
on d1 anymore. White often employs it on the h3-c8
diagonal after g3, so e2 is a step in the right
direction.
23...Qf8
A multipurpose move. In the game Black
played later ...h5, ...Kh7, and ...Bh6. He also could
seek counterplay with ...f5 in some lines. Another
known idea is ...h5, ...Kh7, and ...Qh8, but it does
look weird.
24.Be3
I prefer 24.Nb5 Bc6 25.Nc3, to support e4 in
the event of f4.
24...h5 25.Nb5 Nc5
26.b4?!
Opening the a-file in Black’s favour. It was
better to keep manoeuvring with 26.Nc3 Kh7
27.Rc2.
For instance, the thematic 27...Bh6 28.Bf1
Bxe3+ 29.Qxe3 Qh6 would be pleasant for White –
30.f4 e5 31.b4 axb4 32.axb4 Nd7 33.Qd3±. Let’s
check:
27...Bc6 28.Bf1 Bh6 29.g3! (intending Bh3)
It is very important to evaluate correctly the
following endgame: 29.Nd5 Bxd5 30.Bxh6 Qxh6
31.Qxh6+ Kxh6 32.cxd5
 It looks that White should be better in view
of the weakness of c6 and the possibility of b4.
However, it is a draw, even if White wins a pawn –
32...Nd7 33.Rc6 Ne5 34.Rxb6 Kg5 35.Rb7 h4 36.h3
Rc2 37.Rxe7 Ra2 38.a4 Rb8 39.Ra7 Rxa4=. Black’s
pieces are much more active.
White’s space advantage could be useful
only with queens. For instance, the attempt:
29...f5?! 30.exf5 gxf5 31.Nd5 Bxd5 32.Bxh6
Qxh6 33.Qxd5± would be risky for Black.
26...axb4 27.axb4 Nd7 28.Bf1 Kh7 29.Qf2
Bh6 30.Ra1
30.Re1 Bxe3 31.Qxe3 Ba6 32.Na3 Qh6
33.Qf2 Bb7 34.Nb5 Ba6 is balanced.
30...Bxe3 31.Qxe3 Qg7 32.Nc3 Qh6 33.Kf2
Provoking ...g5!? with sharp play. Instead
33.f4 e5 34.g3 exf4 35.gxf4 Nf6„ is double-edged.
Maze opts for a safe ending:
33...Rxa1 34.Rxa1 Ra8 35.Rxa8 Bxa8
36.h4 Bc6 37.Qxh6+ Kxh6= 38.g3 Kg7 39.Ke3 e6
40.f4 Kf8 41.Be2 Ke7 42.Bd1 Kd8 43.Ba4 Bxa4
44.Nxa4
113
44...f5?
Perhaps Black missed White’s answer.
44...Ke7 (or 44...Kc7) was a draw – 45.Kd4 Kf6
46.e5+ dxe5+ 47.fxe5+ Kf5 48.b5 f6 49.exf6 Kxf6
50.c5 bxc5+ 51.Nxc5 Nb6=.
45.e5! dxe5 46.fxe5 Nxe5 47.Nxb6 Kc7?
This last mistake loses. Black could have
still saved a half point with 47...Ke7 48.c5 Nc6
49.b5 Na7 50.Nc8+ Nxc8 51.b6 Nd6!.
48.c5 Kc6 49.Nc8 Kb5 50.Ne7 Nc4+ 51.Kf4
e5+ 52.Kg5 Kxb4 53.c6 Kc5 54.c7 Nb6 55.c8Q+
Nxc8 56.Nxc8 e4 57.Kf4 Kd4 58.Ne7 e3 59.Kf3 f4
60.Nxg6 fxg3 61.Nf4 Ke5 62.Kxe3 1-0
19. Ivanchuk – Kiril Georgiev
Belgrade 11.11.1997
1.c4 c5 2.Nf3 g6 3.d4 cxd4 4.Nxd4 Nc6 5.e4
Nf6 6.Nc3 d6 7.f3 Nxd4 8.Qxd4 Bg7 9.Be3 0-0
10.Qd2
 12...Rfc8 13.Be2 a6 14.Na4

White can also trade queens with Nd5, but

the endgame should be drawn after 14.h4 b5 15.Nd5
Qxd2+ 16.Kxd2 Nxd5 17.cxd5 Bd7 18.g4 Bb2
19.Rc2 Rxc2+ 20.Kxc2 Be5 21.Bd2 Rc8+ 22.Kd3=.

14...Qxd2+ 15.Kxd2 Nd7

10...Be6

The other major plan is 10...a5 11.b3 Be6

Caruana-Carlsen, Sinquefield 2014, saw the
more aggressive 11...a4 12.b4 Be6 13.Rc1 Nd7
14.Be2 a3 15.Qd1 Nb6 16.Qb3 Rc8 17.Nb5 Bb2
18.Rc2 Qd7 19.Nxa3 Bxa3 20.Bxb6 Ra8 21.0-0 Ra4
22.b5 Rfa8 with some counterplay for the pawn.
12.Rb1 Nd7 13.Be2 Nc5 14.0-0 Qb6
16.g4

Today I see about 400 games in this position,

and Black’s counterplay with ...f5 is already well
tested. However, in 1997 things looked awful for
him – he had lost nearly all the previous games. I
did not improve his score either, but the game
helped the second players to come across the best
defence.
My current evaluation is that Black has fair
chances to neutralise the enemy initiative.

16...f5! 17.exf5
Black is passive, but solid. White lacks a
good concrete plan, so he should start manoeuvring,
for instance: 15.Nb5, Rfd1, Bf1. 17.gxf5 loses control of f4 after 17...gxf5
18.Rhg1 (18.Nc3 fxe4) 18...Kh8 19.Bd3 f4! 20.Bf2
11.Rc1 Qa5 12.b3 Bf7=.

17...gxf5 18.h3 Rf8 19.f4

12.Nd5 is unimpressive owing to 12...Qxa2!
(12...Qxd2+ 13.Kxd2 Bxd5 14.cxd5²) 13.Nxe7+
Kh8 14.Bd4 Rae8 15.Nd5 Bxd5 16.cxd5 Rc8
17.Be2 Rxc1+ 18.Qxc1 Nd7=.

114
 24.hxg4 d5

19...Nf6?!

The only effective counterplay is based on
the break ...d6-d5.
However, the immediate 19...d5 20.cxd5
Bxd5 does not equalise due to 21.Rhd1 fxg4
22.hxg4 e6 23.Nc3 Bc6 24.g5!, when Black’s knight
is severely restricted: 24...Rad8 (24...h6 25.Bc4
Rae8 26.Rg1 hxg5 27.Rxg5 Kf7 28.f5; 24...e5 25.f5
Rxf5 26.Bg4 Rff8 27.Ke1) 25.Kc2 Rc8 26.Bc4 Kh8
27.Kb1².
The right way of achieving the break is:
19...Rad8! 20.Nc3 d5! 21.cxd5 Nf6 22.Bb6
Rd6 23.Bc5 Rd7 24.Ke3 Rc7 25.dxe6 Rxc5 26.Bf3
fxg4 27.hxg4 Nxg4+= with opposite-coloured
bishops.
20.Rhg1 Rad8
I
cannot allow White to consolidate, but the text does
not solve my problems either.
25.f5 Rd8 26.Ke1 d4?!
Ftacnik suggested 26...Bf7, but it is also dull
– 27.g5 Ne8 28.Nxd5 Bxd5 29.Rd1 Bc3+ 30.Kf2
Rcd6 31.cxd5. The d-pawn costs an exchange, e.g.
31...Ng7 32.Bf4 R6d7 33.f6 exf6 34.gxf6 Bxf6
35.Bg4 Re7 36.d6 Re4 37.Kf3 Rd4 38.d7+–.
The rest is easy.
27.Rd1 Rcd6 28.c5! Nd5 (Or 28...Bxb3
29.cxd6 Bxd1 30.dxe7 Re8 31.g5) 29.cxd6 Nxe3
30.dxe7 Re8 31.fxe6 Rxe7 32.Ne4 1-0

It is already late for 20...d5 in view of

21.Nc5.

21.Bb6 Rc8

21...Rde8 22.Bf3 also leaves White with the

upper hand – 22...Bc8 23.Be3 e5 24.fxe5 dxe5
25.gxf5 e4 26.Be2 Bxf5 27.Nc5².

22.Nc3 Rc6 (22...d5 23.gxf5 Bxf5 24.Nxd5²)

23.Be3 fxg4

23...d5 24.gxf5 Bxf5 25.cxd5 Rd8 26.Bd4

Rcc8 27.Bf3 is grim, so I exchange first.

115
 Chapter 6. Anti-Grünfeld
Main Ideas

1.c4 Nf6 2.Nc3 d5 3.cxd5 Nxd5 4.Nf3

See Game 24 Bu Xiangzhi-Zeng, China

2018.
4...c5 is a cunning introduction to the
Grünfeld, and presents a serious theoretical
challenge.

I believe that the most testing way of If we attempt to transpose to Chapter 3 with
fighting the Grünfeld is to enter the main lines. Thus 5.e3, Black can deviate by 5...Nxc3! 6.bxc3 g6.
4.d4 should be the best move. However, if Black has Although I discovered some interesting ideas in the
exceptional memory and had learned well his lesson, hot line 7.h4!? Bg7 8.h5 Nc6 9.Be2, I think that it is
he could often equalise without making any move only a matter of time for Black to find how to
on is own. neutralise White’s slightly “anti-positional” flank
This chapter deals with two alternative assault.
systems where White attempts to outplay the
opponent in relatively unexplored lines.
4...c5
4...g6 is less flexible and allows us to attack
at once with 5.h4. In fact this is a purely positional
stab, which awaits 5...Bg7 to offer a trade of queens
with 6.e4 Nxc3 7.dxc3!. These endgames are, as a
rule, better for White if Black cannot oppose the
construction ...f6+...e5. In our case, the bishop
would be misplaced on g7.
It is wiser to meet 5.h4 by 5...h6 6.e4 Nxc3
7.dxc3 Qxd1+ 8.Kxd1, but again, practical
experience has seen Black struggling to find an
employment for his light-squared bishop.
After all, White has not established yet any
superiority in the centre, to have the right to count
on a successful flank attack. His main problem is the
poor queen’s bishop.

In my opinion, 5.e4 Nb4 6.Bc4 Nd3+ 7.Ke2

Nf4+ 8.Kf1 deserves a greater attention.

116
 Black is still probing several approaches
against White’s provocative set-up. While 8...Ne6
9.b4!! cxb4 10.Ne2 is more or less acknowledged as
dangerous for Black, the counter-sac 9...g6 10.bxc5
Bg7!? still hides secrets. I believe that we should
play for activity and return the pawn with 11.Bb2!.
Svidler recommends for Black 8...Nd3!?
9.Qe2 Nxc1 10.Rxc1 e6.
My improvement over the above-mentioned
game is 15.Kg1! with a strong initiative.
I expect more high-level tests of 5.e4, as the
other trend e3 followed by h4 does not inspire much
confidence.

We claim a space advantage with 11.e5 and

pass the ball into Black’s court.
The latest high-level game
Cheparinov-Gopal, Dubai 2016, saw the plan
11...Nc6, ...Bd7, ...Qb6 with the clear intention of
castling long. Black has two weaknesses in this
scenario – the f7-pawn (Cheparinov attacked it with
Qe2-e3-f4), and c5. We can hit the latter with h4,
Qe3, Be2, Ne4. See Game 21. Analysing it, I
understood that the rook lift Rh1-h3-g3 is not
always obligatory as we could connect the rooks by
Kg1-h2.

117
 Chapter 6. Anti-Grünfeld
Step by Step

1.c4 Nf6 2.Nc3 d5 3.cxd5 Nxd5

5...h6
5...Bg7 6.e4 Nxc3 7.dxc3 Qxd1+ 8.Kxd1
Nd7 is scoring very well for White after both 9.Nd4
c5 10.Nb3 and 9.h5! Nc5 10.h6 Bf8 11.e5. I guess
that the latter line would be the choice of the AI
Alpha Zero, which values highly pawns on h6.
4.Nf3
An interesting, but risky approach is 5...Bg7
6.e4 Nb6. We go 7.d4 Bg4 8.Be3
4.Qb3 Nxc3 5.dxc3, 4.e3 g6 5.Bc4 (5.Qb3
Nb6 6.Nf3 Bg7 7.d4 0-0 8.Be2 Be6 9.Qc2 Bf5
10.Qb3=) 5...Nb6 6.Bb3 Bg7, and 4.g3 c5 5.Bg2
Nc7 are fine for Black.

4...c5

4...Nxc3 5.bxc3 g6 gives White additional

options:
6.e4 Bg7 7.Ba3 0-0 8.d4 Nd7 9.Be2 c5
10.0-0 b6 11.e5 is double-edged. Perhaps best is to
include:
6.h4 h6 7.d4 with a promising position.
4...g6 offers White a wide choice:
5.Qb3 Nb6 6.d4 Bg7 7.Bf4;
5.Qa4+ Bd7 6.Qh4 Bc6.
I will consider the most topical plan, based
on e4, followed up by dxc3.
5.h4
The immediate 5.e4 Nxc3 6.dxc3 Qxd1+
7.Kxd1 is well met by 7...f6 8.Be3 e5. Therefore, we
should provoke weaknesses first. A good news
would also be ...Bg7 since the bishop is better off on
f8.
Now 8...0-0 is dangerous – 9.Qd2 Bxf3
10.gxf3 e5 11.d5 h5 12.0-0-0², but 8...Nc6 9.d5 Ne5
10.Be2 Nxf3+ 11.gxf3 Bh5 12.f4 Bxe2 13.Qxe2
Qd7 14.a4 also gives White an initiative – 14...c6 (or
14...Bxc3+ 15.bxc3 Nxa4 16.c4 a6 17.h5±, Ghaem
Maghami-Puranik, Moscow 2017) 15.a5 Nc8 16.h5
Nd6 17.Rd1².
6.e4 Nxc3 7.dxc3 Qxd1+ 8.Kxd1. Black’s
main problem is that he cannot find a good place for
his light-squared bishop. See Game 24 Bu
Xiangzhi-Zeng, China 2018.

118

White has two ways to avoid theoretical
discussion in the Grünfeld:
A. 5.e3; B. 5.e4!?
They both lead to complex play with plenty
of unexplored lines, although Black is catching up
quickly. In my opinion, 5.e4 hides more venom, and
should be the weapon of choice. The attempt to
avoid the Grünfeld by:
5.d4 Nxc3 6.bxc3 g6 7.h4 Bg7 8.h5 is
unconvincing – 8...cxd4 9.cxd4 Nc6 10.e3 e5, or
7.Bf4 Qa5 8.Qd2 Bg7 9.e3 0-0 10.Be2 Nc6 11.Rb1
b6=.
A. 5.e3 Nxc3
5...g6 allows tricks like 6.Qa4+ Bd7 7.Qb3,
when 7...Nxc3 8.Qxc3 is a double attack.
5...e6 6.d4 should transpose to Chapter 3.
5...Nc6 6.Bb5
6.Bc4 Nxc3 7.bxc3 g6 is a good version of
the Grünfeld for Black.
After 6...Nxc3 7.bxc3 Bd7 8.0-0 e6 9.d4 Be7
White builds a strong centre with Rb1, e4, Bf4, and
aims for d4-d5.
6.bxc3
The structure after 6.dxc3 Qxd1+ 7.Kxd1 is
slightly better for White only if Black’s bishop had
gone to g7, which is a wrong place for it. Here
7...Nc6 8.e4 b6 9.Kc2 Bb7 10.Bf4 h6 is fairly equal.
119
6...g6
A1. 7.h4; A2. 7.Bb5+
Another version of line A1 is 7.Qa4+ Nd7
8.Ba3 Qc7 9.h4, when 9...h6 puts White’s concept to
the test.
A1. 7.h4
If you ask me what is the reason behind this
move, I would not be able to answer. Frankly, I do
not see any benefits of including:
7.h4 h6!?.
Cummings mentions 8.h5 g5 9.d4, then Bd3,
but Black will put comfortably ...Bg4∞.
I still analyse 7.h4 because it is fashionable
and you might want my opinion on it. It is intriguing
for me that such a natural move as 7...h6 is not the
main line and is rarely seen. I could not even find a
single fresh game in correspondence and engine
chess. If you do meet it, you would have to
improvise. Here is a model of clear and purposeful
play from White:
8.Ba3 Qc7 9.Qb3 Bg7
9...Be6 10.Bc4 Bxc4 11.Qxc4 Nd7 12.0-0
Rc8 13.Rab1 Bg7 14.d4 Qa5 15.Qb3 b6 16.c4 e6∞.
10.d4

10...cxd4 (10...Be6!) 11.cxd4 Nc6 12.Rc1
Be6 13.Bc4 Qa5+ 14.Ke2 Bxc4+ 15.Rxc4 0-0
16.Ra4 Qc7 17.Rc1 Qd7 18.d5 Ne5 19.Re4! Nxf3
20.Rxe7², Grandelius-Tari, Fagernes 2016.
7...Bg7 8.h5 Nc6
8...0-0 is highly provocative, but not that
bad. Still, it helps White to define his plan. He opens
the h-file and tries to bring his queen there – 9.hxg6
hxg6 10.Ng5 e5
Si.Williams-Hill, Telford 2018, saw
10...Qd5, when 11.f4! Nd7 12.e4 Qc6 13.Qf3 would
have been dreadful.
11.Ne4 Qe7 12.Qf3 Be6 13.Qg3 with
prospects for an attack.
Let’s return to the main line.
White has tested 6 different moves.
The original intention of Pachman, who
introduced 7.h4, was to transfer the queen to h4 at
once – 9.Qa4 Bd7 10.hxg6 hxg6 11.Rxh8+ Bxh8
120
12.Qh4 Bf6 13.Qh7 Be6, Pachman-Lau, Munich
1979. However, all this is perfectly comfortable for
Black.
Then White players decided to wait for short
castling before attempting to attack on the h-file.
They tried Qb3+Ng5 to provoke weaknesses and to
activate the king’s knight. However, the straight
9.Qb3 could be neutralised by the “dumb” 9...Qb6
10.Ba3 Qxb3.
An interesting discovery was 9.Ba3 first, and
if:
9...Qa5, 10.Rh4!? protects indirectly the
bishop. If we prolong this variation: 10...Bd7 11.Ra4
(11.Qb3 Rc8) 11...Qc7 12.Bxc5 Ne5, we see that
Black wins the exchange with unclear consequences.
On top of all, Duda-Tomczak, Warsaw 2018, saw
another method of defending c5:
9...b6 10.Rc1 (10.Qa4 Qc7; 10.d4 0-0)
10...Qd5
Duda opted for 11.Qb3 Be6 12.c4 Qd7
13.Bb2 0-0 14.Bxg7 Kxg7 15.d4 cxd4 16.exd4 Bg4
17.d5 Bf3, when 18.Qxf3 Ne5 19.Qe3 would have
assured him of the initiative. Again, the trade of
queens should be enough for Black to hold –
11...Qxb3 12.axb3 Bb7.
Finally, Cummings recommends 9.Bc4 0-0
10.hxg6 hxg6 11.Ba3 Na5 12.Be2. I’m not sure that
White equalises in this line after 12...Qc7 13.Ng5
Rd8 14.Qa4 f6. In any case the manoeuvre
Bf1-c4-e2 does not impress me.
The analysis of the above variations inclined
me to give my vote for:
 9.Be2

This is not a simple developing move. I have

in mind the really weird manoeuvre Kf1, Qe1-h4!?
after something like Ng5, f3 first. An illustrative line
is:
9...b6 10.Kf1 Qd6 11.d4 0-0 12.Ng5!?,
which is much more straightforward than:
12.Ba3 Bf5 as in the game
Aronian-Nepomniachtchi, Saint Louis 2017.
12...Bb7 13.hxg6 hxg6

Black has deflected our attack, but our pieces

are slightly better coordinated, and we still keep our
central pawns mobile. That makes our practical
chances higher, although the computer would level
them.

b) 14...Rfc8! 15.f3 cxd4, when Black has to

find the sacrifice on d4 after 16.Ne4 Qe6 17.exd4
Nxd4!= or 16.exd4 Nxd4! 17.cxd4 Qxd4 18.Rb1
Rc2 with sharp play.

9...Bf5 10.Qb3 b6

14.Qe1!?
Another idea is to double the heavy pieces
on the h-line with Rh3+Qh1, threatening to mate by
Rh8. For instance: 14.Rh3!? Rad8 (14...Rfd8
15.Qb3 Rf8 16.Kg1 Rad8 17.Qa4 Bc8 18.dxc5
Qxc5 allows 19.Rh8+! Bxh8 20.Qh4 Qxg5 21.Qxg5,
although Black has considerable compensation for
the queen) 15.f4 Bf6! 16.Nh7 Rfe8 17.Kf2 Kg7
18.Qh1 Rh8 19.Ba3÷.
a) 14...Na5 (on 14...cxd4 we take 15.exd4)
15.f3 f6 16.Nh3 Kf7 17.Kf2 g5 18.Bd3 Rh8 19.Qe2
Rad8 20.Ba3 Bc8 21.Rad1 Qc7 22.g4
Perhaps Black could try:
10...Qd6 11.Ba3 b6 12.d4 0-0 13.Rd1 Na5
14.Qb5 Qc7 15.dxc5 bxc5 16.0-0.
10...0-0 is also interesting since 11.hxg6
hxg6 12.Qc4 e5 13.Qxc5 is risky – Black has
compensation for the pawn after 13...Bd3 14.a4
Bxe2 15.Kxe2 e4 16.Nd4 Rc8 17.Qb5 Re8.
11.Ng5 e6

121
 b6) 10.d4 Qc7 11.Bxd7 Nxd7 12.e4 Rfd8 is level
since 13.Bg5 could be parried by 13...h6.

8...Bg7 9.0-0 0-0 10.d4

We have been following the game

Aronian-Svidler, Saint Louis 2017. It saw 12.f4 h6
13.Qa4 Rc8 14.Ne4 g5 15.Bb5 Qd5 16.Nf2 Qd6
17.Ne4 Qd5 18.Nf2 Qd6 19.Ne4 ½-½.
I think that White can fight for an edge with:
White has achieved more than 62% in this
12.Qa4 Rc8 13.Ne4 0-0 position, but that does not necessarily means that he
is better.

Or 13...Qd5 14.f3 (14.h6 Be5 15.f3 Bxe4 10...Qa5!

16.fxe4!? is sharper – 16...Bg3+ 17.Kf1 Qe5 18.Rh3
0-0 19.Kg1) 14...Bxe4 15.Qxe4.
10...Nc6 11.Rb1 is awkward, as Black must
14.g4 Bxe4 15.Qxe4² place on a passive place either his queen (11...Qc7
12.e4) or the knight (11...Na5 12.e4).

11.Qb3 Bc6! 12.Bd2 Nd7 13.c4 Qa6

14.Bc3=
A2. 7.Bb5+ Bd7!

Gelfand reached twice this position in blitz

The check from b5 mostly counts on games and drew against Nepomniachtchi and
7...Nbd7, when White should expect a slight Svidler.
advantage after 8.0-0 Bg7 9.d4 0-0 10.a4!, planning
a4-a5-a6 – see Game 22 Banusz-Danin, Zalakaros
2017. B. 5.e4!?

8.Be2
This move shuns a lot of move order tricks
and transpositions, so typical for 5.e3. At the same
8.a4 Bg7 9.0-0 (9.d4 Qa5=) is insufficient time it immediately takes the initiative.
owing to 9...Bxb5 10.axb5 Qd3 11.Qb3=.
5...Nb4
8.Qb3 Bg7 9.0-0 0-0 10.d4 Nc6=.

8.Rb1 Bg7 9.0-0 0-0 (9...Bxb5!? 10.Rxb5 5...Nxc3 6.dxc3 Qxd1+ looks drawish at a

122
glance, but White scores surprisingly well, both in
practical and computer chess. White commonly
enjoys a small, but stable edge. It is mostly due to
the c5-pawn, which limits the scope of the Black
bishop. White should concentrate on the kingside
and take space with h4. See Game 20
Timman-Kamsky, rapid Paris 1991.

White’s pieces are slightly more active and

that could be transformed into more stable pluses.
For instance:
12...g6 (12...a6 13.Ba4) 13.0-0 Bg7 14.Rd1
0-0 15.Bxc6 bxc6 16.Qc5.

7.Ke2 Nf4+
6.Bc4

7...Nxc1 8.Rxc1 a6! is playable, if Black is

After 6.Bb5+ N8c6 (6...Bd7 7.a3 N4c6 8.d4) well prepared for the ensuing very sharp play.
7.d4 (7.0-0 a6=) 7...cxd4 8.a3 simplest is 8...Bd7, 8...Nc6?! is worse owing to 9.Bb5 Bd7
although: (9...Bg4 10.Na4) 10.d4 with a huge lead in
8...dxc3 9.Qxd8+ Kxd8 10.axb4 cxb2 development.
11.Bxb2 e6 is also solid. 9.d4
9.Nxd4 Nxd4 10.Bxd7+ Qxd7 11.axb4 e6 You can also test the opponent by 9.e5,
12.Be3 Rd8 13.0-0 Bxb4 14.Nd5 Ne2+ 15.Qxe2 intending e6 or Bd3-e4. For instance: 9...g6 10.e6
exd5 16.Rfd1 0-0 17.Rxd5 Qe7=. fxe6 11.Qb3 Nc6 12.Bxe6; 9...e6 10.d4 b5 11.Bd3
cxd4 12.Be4 Ra7 13.Qxd4 Rd7 14.Qe3 Be7 15.Rc2
6...Nd3+ 0-0 16.Rhc1; 9...Nc6 10.e6 fxe6 11.Re1 b5 12.Bd3
g6 13.Be4 Bb7 14.Kf1 Bh6 15.a4 with balanced
play.
6...Be6 is best met by 7.Bb5+!? 9...b5 10.Bd5 Ra7
7.Bxe6 Nd3+ 8.Kf1 fxe6 9.Ng5 Nc6
10.Nxe6 Qd7 11.Nxc5 Nxc5 12.Qh5+ g6 13.Qxc5
e6 14.Qc4 Bg7 15.h4 Nd4 is dangerous for White,
although he is not worse objectively.
7...Bd7
7...N8c6 8.d3 a6 9.Bxc6+ Nxc6 10.Be3
practically wins a pawn as 10...b6?! 11.d4 cxd4
12.Nxd4 Nxd4 13.Bxd4 is obviously in White’s
favour. Of course 10...Bg4 11.Bxc5 Bxf3 12.gxf3 is
not clear at all, but let Black prove enough
compensation.
8.a3 N4c6 9.d4 Nxd4 10.Nxd4 cxd4
11.Qxd4 Nc6 12.Qe3

123

a) 11.Ne5!? e6 12.Bc6+ Ke7 13.Qd2 f6
14.Nd5+!! exd5 15.Rxc5 dxe4 16.Rhc1 (16.Rd5
Bg4+=) 16...Be6. This crazy position occurred in
Jakovenko-Sutovsky, Poikovsky 2016.
White opted for 17.Qe3, when 17...Qb6
18.Qxe4 Kd8 19.Rd5+ Ke7 20.Rdc5 could have led
to a draw. We could test Black by:
17.d5!?, and if he finds a series of only
moves all the way up to move 35, he should hold the
draw:
17...Qd6 18.Qe3 Bf5 19.Rxb5 Rd7 20.Nc4
Qxh2 21.d6+ Kf7 22.Rxb8
Or 22.Rxf5 Nxc6 23.Qxe4 Bxd6 24.Qxc6
Re7+ 25.Kf3 Rd8 26.Rd1 Re6 27.Nxd6+ Rdxd6
28.Rxd6 Qxd6 29.Qxd6 Rxd6=.
22...Bxd6 23.Nxd6+ Rxd6 24.Rb7+ Kg6
25.Bxe4 Qh5+ 26.Qf3 Qxf3+ 27.gxf3 Re8 28.Rg1+
Kh5 29.Rgxg7 Bxe4 30.fxe4 Rxe4+ 31.Kf3 Ra4!
32.Rxh7+ Kg5 33.a3 Rf4+ 34.Ke2 Re4+ 35.Kf1
Rd2=. Black’s rook on d2 ensures enough
compensation for the pawn.
124
b) 11.dxc5 e6 12.c6
Komodo 7 won a game following 12.Qd4
exd5 13.exd5 Qf6 14.Rhd1 Qxd4 15.Nxd4 g6 16.c6.
Black is holding here, but he is playing for two
results only.
12...b4
Or 12...exd5 13.Nxd5 Bd6? (13...Rc7
14.Qd4) 14.Qd4.
13.Qd4 Rc7
13...Qe7 14.Na4 exd5 15.Rhe1 dxe4 16.Kf1
f5 17.Nb6 Be6 18.Ng5 Qxg5 19.Qe5 Qd8 20.c7
Rxc7 21.Qxe6+ Re7 22.Qxf5 keeps the attack.
14.Na4 exd5 15.exd5
15...Bd6!
Best defence. K. Hansen recommends
15...Qf6?!, but if we dig deeper, we find Black
struggling in this endgame – 16.Rhe1 Qxd4
17.Nxd4 Bd6 18.Kf1+ Kd8 19.Nb6 h5 20.h3 a5
21.Kg1 Ba6 22.Nf5 Bf4 23.Rcd1 g6 24.Re4.
15...Be7?! 16.Qxg7 Bf6 17.Qh6 Re7+
18.Kf1 was also better for White in
Giri-Harikrishna, Wijk aan Zee 2017.
16.Qxg7 Re7+ 17.Kf1 Rf8 18.Qd4. White
has full compensation, but 18...f5 19.Nb6 Re4
20.Re1 Kf7 21.Rxe4 fxe4 22.Nd2 Bf4 23.Nxc8
Nxc6 24.Qc5 Qxc8 25.dxc6 Qd8 26.Qh5+ leads to
perpetual check.
8.Kf1
 If Black wanted to play 12...Nc6, he did not
need 11...a6. We answer 13.Bd3 Bd7
Or 13...Be7 14.Rh3 Qc7 15.Qe3².
13...b5 offers a pawn – 14.Bxh7 Rxh7
15.Qe4, but we could also follow our main plan –
14.Be4! Bb7 15.Rh3 Be7 16.Qe3.
14.Rh3 Qc7
Hou,Yifan-Tari, Fagernes 2016, saw
14...Rc8 15.Re1 (aimed to protect e5 in the event of
Ng5, Qh5. The immediate 15.Ng5 was stronger –
15...h6 16.Qh5) 15...b5?! (15...Be7 16.Ng5ƒ)
16.Ng5! Nd4 17.Qg4 h6 18.Qh5±.
B1. 8...Nd3!; B2. 8...Ne6

B1. 8...Nd3! 9.Qe2 Nxc1 10.Rxc1 e6 11.e5

It does not matter whether we play first

11.h4 or 11.e5.
I have also analysed variations where White
refrains from e4-e5 in favour of h4-h5, but I did not
find any benefits from that after both 11.h4 Nc6
12.h5 h6 and 12...Bd7.

15.Qe3! (preparing Be2, Ne4) 15...Nb4

16.Be2 0-0-0 17.a3 Nd5 18.Nxd5 exd5 19.Rg3 Kb8
20.b4 and Black’s centre crumbles down, e.g.
20...d4 21.Qd3 g6 22.Qb1 Bc6 23.bxc5 Ka8 24.Rg4
Rd5 25.Nxd4 f5 26.exf6 Qh2 27.Nf3±.

13.Bd3 Bb7 14.Rh3!

11...a6

Black has also tried two other set-ups – with

...Bd7+Qb6, and ...Bd7+Qc7, then long castling.
White follows the scheme h4, Kg1, Qe3 when Black
castles on the queenside, Be2 or Bf1. Then he
attacks the f7- and the c5-pawn. I consider all this in
detail in Game 21, Cheparinov-Gopal, Dubai 2016.

12.h4 b5
125
 14.a4 Be7 15.axb5 axb5 is unclear as Black
gets counterplay on the queenside. White should not
distract himself from the main battlefield.
An important question is whether to insert h5
h6. In my opinion, ...h6 is a useful move, and I do
not see a necessity to provoke it, at least at the early
stage of the opening.
14...Be7 15.Be4 Nc6 16.Qe3 Qd7 17.Rg3 g6
18.Kg1
We have fulfilled our plan. Our king is safe,
while Black cannot castle. Our next step should be
to open the centre, e.g. 18...h6 19.a3 Rb8 20.Qf4 b4
21.Ne2 Na5 22.d4².
B2. 8...Ne6
9.b4!!
126
This ingenuous sacrifice was introduced in
1980 by the famous analyst Polugaevsky, but his
original idea to meet 9...cxb4 by 10.Nd5 was
unconvincing. The tide turned when Lerner
improved it with 9...cxb4 10.Ne2!
10...Nc7
The alternative set-up is 10...Nc5 11.Qc2 e6
12.d4 Ncd7. Here White players often continued
development with 13.Bb2, but I prefer to remain
true to the spirit of the line with Kf1: 13.h4!? Nb6
14.Bd3 Bd7 15.Nf4ƒ.
11.d4 e6 12.h4! Nd7 13.Bd3!
Kasparov played 13.Bg5 Be7 14.Rc1.
Stronger was 13...Nf6, when g5 is hardly the best
place for the bishop.
13...Nf6 14.Bb2 Bd7 15.Rh3 Be7 16.Nf4
White’s pieces are dreadful. The bishops
trade does not help much – 16...Bb5 17.Kg1 Bxd3
(Our attack is also strong after 17...a5 18.Bc2 a4
19.Rc1 b3 20.axb3 a3 21.Ba1 a2 22.b4 Bxb4 23.Rg3
Ba3 24.d5 exd5 25.e5 Ne4 26.Rxg7 Ra6 27.e6.)
18.Qxd3 Rc8 19.Rd1 0-0 20.Rg3±. The computer
proposes the bold:
16...0-0 17.Qe2 e5 18.dxe5 Ng4 19.Rg3
Bxh4 20.Nxh4 Qxh4

White’s dominance is so big that he can
safely hide the king in the centre – 21.Ke1 Rad8
22.e6! Bxe6 23.Rc1, winning material.
An interesting alternative to 9.b4 is 9.d3,
having in mind 9...Nc6 10.h4!. I like the concept,
but Black has found a solid retort in 9...g6! 10.h4
h5! – see Game 23 Khalifman-Paravyan,
Kaliningrad 2015.
The best answer to the bold 9.b4 is
considered to be the even bolder:
9...g6!? 10.bxc5 Bg7
11.Bb2!
Both sides are trying to outsacrifice each
other! Instead of taking on e6 White is planning d4,
followed up by Nd5. In such strategically
unbalanced positions initiative is more important
127
than pawns. Let’s consider:
11.Bxe6 Bxe6 12.d4 Nc6 13.Be3 Qa5
Black’s game is playing itself.
14.Ne2
14.Rc1 0-0-0! 15.h3 (15.Ne2 f5! 16.Ng5
Bxa2 finished in Black’s favour, Garriga-Gopal,
Caleta 2017) 15...Rhf8 (peparing ...f5) is balanced –
16.d5 Bxc3 17.Qc2 or 16.Kg1 f5.
14...Bc4 15.d5 Nb4
15...Bxa1 16.Qxa1 f6 (16...0-0-0 17.Nd2)
17.dxc6 bxc6∞.
16.Bd4 Bxe2+ 17.Qxe2 Bxd4 18.Nxd4 Qxc5
19.Rd1 Rc8 20.g3 Qc4=.
11...Nc6 12.h4! 0-0
Black should bear in mind that the threat
Bxe6 is still in the air. For instance, 12...h5 allows
13.Bxe6
13.d4 Ncxd4 14.Nd5 Nxf3 15.Bxg7 Nxg7
16.Qxf3 0-0 17.Rb1² is also possible.
13...Bxe6 14.d4 Qa5 15.Ne2 Bc4 16.Qc2².
12...Nxc5 is no better owing to 13.d4 Nxd4
14.Nd5.
13.h5 Nxc5 14.d4 Nxd4 15.Nd5 Nxf3
16.Bxg7 Kxg7 17.Qxf3 g5 18.Qe3 Ne6 19.h6+ Kh8
20.Rd1 Qd6²
I suppose that White might improve this
sequence of moves, but the position on the diagram
suits me enough. White’s pieces are much more
active, but the cherry on the cake is the h6-pawn.
Thanks to it even endgames are in White’s favour.

A natural follow up is 21.Bb3, while the

computer suggests 21.Kg1.

128
 Chapter 6. Anti-Grünfeld
Annotated Games

20. Timman – Kamsky

rapid, Paris 1991

1.Nf3 Nf6 2.c4 c5 3.Nc3 d5 4.cxd5 Nxd5

5.e4 Nxc3 6.dxc3 Qxd1+ 7.Kxd1

It is commonly good to trade light-squared

bishops because our remaining pieces are more
active. White is better here: 26...b6 27.h4 h5 28.g5
Nd4+ 29.Bxd4 cxd4 30.Ne4 dxc3 31.gxf6 Bf8
32.Rg6 Ke8 33.Rg8 Rc6 34.Rh8 a6 35.Rh7 b5
36.axb5 axb5 37.Rb7 Bd6 38.Rxb5 Bc7 39.Rb7 Bd8
40.Rb8 1-0.

Li-Ostenstad
Fornebu 2017

7...f6

An ambitious plan. Black does not want to

cede space on the kingside.
Another popular approach is 7...Nc6 8.Be3
e6 9.Kc2 Bd7 10.Be2 Be7. I believe that White
should focus on the kingside in this line.
Accordingly, we should put our queen’s rook on d1
and aim for h4-h5, e.g. 11.Rad1 0-0-0 12.b3 f6
13.h4. Naturally, play is not forced, so I see no
reason to provide concrete variations. Instead I’ll
show you some practical examples to clarify what White has fixed the g7-pawn and could
we are aiming for. prepare a further pawn advance on the flank. Black
sought counterplay: 21...f5 22.exf5 exf5 23.Bf4 Nf6
Martirosyan-Babujian 24.Rxd8+ Rxd8 25.Ne5 Bd5 26.Ng6 Kf7 27.Nxe7
Yerevan 2016 Kxe7 28.Re1 Kf8 29.g4 fxg4 30.fxg4 Re8 31.Bd2
Kf7 32.c4 Be4+ 33.Kd1 a5 34.Bc3 Bc6 35.b3 Re3
36.Kd2 Rg3 37.Bd3 Bd7 38.Be5 Rxg4 1-0

Krasenkow-Tomczak
Lublin 2009

129
 19...f5 20.gxf5 gxf5 21.Be6+ Kc7 22.Bxf5².
Black did not allow the opponent to produce a
passer on the h-file, but he made a weakness on e5.
Greenfeld-Aghayev
Tortoreto 2015
Black decided to keep his bishop, but this
way all his minor pieces remained passive: 19...Bc8
20.Nd5 Ne7 21.f3 Nxd5 22.Bxd5 g6 23.Rhd1 h5
24.Bf2 g5 25.a6 b6 26.b4 Kc7 27.bxc5 Bxc5
28.Bxc5 bxc5 29.c4 Rd6 30.Ra5 Bxa6 31.Rda1 Bb7
32.Bxb7 1-0
It is important to keep a rook! Without it we
would not be able to attack the weak enemy pawns.
8.Be3 e5 9.Nd2 Be6 10.Bc4! Bxc4 11.Nxc4
Nc6
130
12.a4
This move is in no way weak, but White had
a better idea, which would work against long
castling. Instead of entrenching the c4-knight, he
could create a weakness on c5 with a3, b2-b4xc5.
For instance: 12.Ke2 0-0-0
The point is that 12...b5 only facilitates
White’s play after 13.Nd2 Kf7 14.a4 b4 15.Nc4.
13.a3 Rd7 14.b4 Nd8 15.Rad1 Rxd1
16.Rxd1 b6 17.bxc5 Bxc5 18.Bxc5 bxc5. The first
stage is finished, and the next step should be to
undermine e5 – 19.g3 Kc7 20.f4 exf4 21.gxf4 Ne6
22.f5 Ng5 23.Ke3 with serious space advantage.
12...Kf7 13.Ke2 Be7 14.Rhd1
Timman stakes all on a central break.
You could also push h4-h5 and trade a pair
of rooks. Then you could try Nc4-e3-d5.
14...Rhd8 15.Rd5 Ke6 16.Rad1 b6
 17.f4! exf4 18.Bxf4 Rac8 19.Na3?
Such a familiar picture! You follow a good
plan, get a tangible positional advantage, and you
know that you should switch to a calculating phase.
But there is nobody to tell you when is the exact
moment for that! Timman found a concrete
manoeuvre – Nc4-a3-b5, but it is too hasty. He
should have kept on squeezing the opponent with
19.h4 Kf7 20.g4.
19...Ne5??
This is a double fault, I would say – both
positional and tactical. It is more than obvious that
White’s knight would be dominating the bishop.
Besides, Black could have launched a counter-attack
with 19...g5! 20.Bg3 Na5 21.Nb5 Rxd5! 22.exd5+
Kd7 23.Nxa7 Ra8, and suddenly Black’s pieces
develop great activity – 24.Nb5 Nb3 25.Bc7 Rxa4
26.Bxb6 Ra2 27.Rb1 c4 with decent compensation
for the pawn.
20.Bxe5 fxe5 21.Nc4 Bf6
131
Black is positionally dead. White only needs
a second target, and the easiest way to make it was
22.a5!. Timman let a great part of his advantage slip
away, but eventually won:
22.Nd6 Rb8 23.Kf3 h5 24.Nb5 Rxd5
25.Rxd5?! Rb7 26.c4 Be7 27.Ke2 a6 28.Nc3 Rb8
(28...Rd7!²) 29.Rd3 a5 30.Nd5 Bd8 31.Rh3 g6
32.Rg3 Kf7 33.Rf3+ Kg7 34.g3 Rb7 35.Rd3 Kf7
36.Rf3+ Kg7 37.h3 Rd7 38.h4 Rd6 39.Kf2 Rd7
40.Kg2 Rf7 41.Rb3 Rb7 42.Kh3 Rd7 43.g4 hxg4+
44.Kxg4 Rd6 45.Rh3 Re6 46.h5 Rd6 47.b3 Bf6
48.Rh1 Re6 49.Rf1 Be7 50.Rf3 Kh6 51.Rh3 Kg7
52.hxg6 Kxg6 53.Rh8 Bd6 54.Rd8 Kf7 55.Rd7+
Kg6 56.Rb7 Bf8 57.Nxb6 Rf6 58.Nd5 Rf1
59.Rb6+ Kf7 60.Rb5 Rb1 61.Kf5 Rf1+ 62.Kxe5
Bg7+ 63.Kd6 Bd4 64.e5 Re1 65.Rb7+ Kg6 66.e6
Be5+ 67.Kd7 Bf6 68.Nxf6 Kxf6 69.Rb6 Kg7
70.Rd6 Kf8 71.Rd5 Re3 72.Rxc5 1-0
21. Cheparinov – Gopal
Dubai 18.4.2016
1.Nf3 Nf6 2.c4 c5 3.Nc3 d5 4.cxd5 Nxd5
5.e4 Nb4 6.Bc4 Nd3+ 7.Ke2 Nf4+ 8.Kf1 Nd3
9.Qe2 Nxc1 10.Rxc1 e6 11.e5 Nc6 12.h4 Bd7
 13.Qe3!?
Sometimes White plays 13.Rh3, but this
move is essential when we need to attack the
kingside. Meanwhile, we have more urgent matters
to attend to. Black’s last move displays his intention
to castle. If we do not hinder his plan, he will
proceed with ...Qb6(c7), 0-0-0, possibly ...h6,
...Kb8, with fair chances. A move like Rh3 does not
pose any concrete problem. I was more attracted by
13.Kg1 at first, but analysis convinced me that it
was too slow. Cheparinov chooses to oppose the
enemy plan with the manoeuvre Qd2-e3-f4, hitting
f7. I like his move for other reasons too – from e3
the queen squints at c5 and prepares Be2, Ne4.
Our second target is c5!
13...Qb6
This seems more active than 13...Qc7!?, but
the latter hits e5, hampering the idea Nc3-e4.
The computer promises an edge after
14.Bb3, but it does not realise that the forced line
14...0-0-0 15.Ne4 Nxe5! 16.Nxe5 Qxe5 17.Nxc5
Qxe3 18.Nxd7+ Kxd7 19.Ba4+ Ke7 20.dxe3
132
20...Kf6 21.Rc7 Bd6 22.Rxb7 Rb8 23.Rd7
Rhd8 24.Rxd8 Rxd8 25.Ke2 Rb8 is a dead draw
despite the extra pawn. Yet it could be White’s best
option as the alternatives are:
14.h5 0-0-0 15.Rh4 transposes to
Nepomniachtchi-Bok, Moscow 2016: 15...Be7
(15...Kb8!∞) 16.Rg4 g6 17.Nb5?! Qb6 18.Be2,
when 18...gxh5 would have been pleasant for Black;
14.Kg1 0-0-0 15.Bf1 Kb8 16.Rh3 Be8
17.Ne4 Nxe5 (Black always meet Ne4 by this
capture) 18.Nxc5 Ng4 19.Qg5 f6 20.Qxg4 Bxc5
21.d4 h5 22.Qxe6 Qd6 23.Bc4=;
14.Rh3 0-0-0;
14.Qf4 0-0-0 15.Qxf7 Nxe5 16.Nxe5 Qxe5
17.Re1 Qf5=;
14.Be2 0-0-0 15.Ne4 (15.Na4 b6 16.d4 Nxd4
17.Nxb6+ axb6 18.Nxd4 Kb8=) 15...Nxe5 16.Nxe5
Qxe5 17.d4 Qd5 18.Nxc5 Bc6=.
As you see, 13...Qc7 leeds to rather drawish
positions if White attempts to play energetically in
the centre.
14.Qf4 h6
 15...Be7

Perhaps Black should have tried 15...Na5

16.Be2 Bc6, although 17.Rh3 Be7 18.Ne4 Qb4
19.d4 takes the initiative.

16.Rh3 Rf8 17.Rg3 g6 18.Qxh6 0-0-0

19.Qf4 Kb8 20.Kg1

15.b3

White plays it solid. His most principled try

was 15.Kg1!. Its idea is to connect the rooks via the
first rank. Thus we could capitalise on having
delayed Rh3. For instance:

a) 15...Na5 16.Bd3 Bc6 17.Be4 Bxe4

18.Nxe4 Qxb2 19.Kh2! Be7 20.d4! c4 21.Qg4 0-0

20...Nd4?!

Black plugs the d-file and remains without

counterplay. It was better to prepare ...Nd4 with
20...Be8. It is unclear what White could do here. For
instance:
21.Ne4 Nd4 22.Nxd4 Rxd4 23.Bf1 Rh8 and
he should sacrifice the queen in order to maintain
the balance – 24.Rg4 Rxh4 25.Rxh4 g5 26.Rh8 gxf4
27.Rxe8+ Bd8 28.Nxc5 f3 29.gxf3 Rd5 30.b4 a5
31.Ne4=.

22.Rb1! 21.Nxd4 cxd4 22.Ne4 Bc6 23.Nf6

Driving the queen away from the kingside.
The immediate 22.d5 exd5 23.Ng3 Nc6 24.Nf5 g6
25.Nxh6+ Kh7 26.Qf4 Nd8 27.Rb1 Qc2 is only a 23.Ng5 was more active – 23...Rd7 24.Bd3
draw. a6 25.Be4±.
22...Qa3 23.d5 exd5 24.Ng3 Kh7 25.Rb5
Rae8 26.Nh5 Rg8 (26...g6 27.Nf4 Nc6 28.h5 g5 23...Qa5
29.Nxd5±) 27.Ng5+!±. The rook from h1 suddenly
becomes the deciding factor.

b) 15...Ne7 16.Bb3 Bc6 17.Ne4 Ng6 18.Qe3

h5 19.Rh3ƒ.
133

24.a4?!
White has two weaknesses – on a2 and d2.
He should have defended them with 24.Rc2 Rh8
25.Bd3, retaining the better chances.
24...Rh8 25.Re1
Chaparinov does not feel the danger. He
should have exchanged pieces with 25.Bd3 Bxf6
26.exf6+ e5 27.Rg5 exf4 28.Rxa5 Rxh4 29.f3=.
25...a6 26.Bd3 Ka8 27.Rh3 Bxf6 28.exf6=
Rh5 29.Re2 e5 30.Qg3 Qc5
31.f3?
This move breaks the connection between
White’s heavy pieces and the b3-pawn. 31.Re1 still
134
kept the balance. Gopal could have punished the
mistake with 31...Bd7! 32.Rh1 Qd6 33.Qe1 Qxf6µ.
The rest of the game featured mutual mistakes:
31...Qd6 32.Qe1 Qxf6 33.Be4? Bd7 34.g4
Rhh8 35.Qg3 Qd6 (35...Qb6–+) 36.Bd3 Rde8
37.h5 f5? (37...gxh5 38.Rxh5 f6 39.Reh2 Rhf8µ)
38.hxg6 Rxh3 39.Qxh3 Qxg6 40.Rg2 e4 41.gxf5
Qxf5 42.Qxf5 Bxf5 43.fxe4 ½-½
22. Banusz – Danin
Zalakaros 24.05.2017
1.c4 c5 2.Nf3 Nf6 3.Nc3 d5 4.cxd5 Nxd5
5.d4 Nxc3 6.bxc3 g6 7.e3 Bg7 8.Bb5+ Nd7 9.a4!
0-0 10.0-0
10...b6
Black decides to ignore the positional threat
a4-a5-a6.
10...a6!? brings him better results. In my
opinion White should retreat to e2:
11.Be2!. White’s design is to gain space with
e4-e5. Then the bishop should protect the f3-knight
and indirectly the d4-pawn.
The advantage of this move over 11.Bd3 is
seen in the line 11...e5 12.Ba3 Re8 – ...e4 is not a
fork, and White has 13.Rab1 or 13.Qb3! The latter is
possible because 13...cxd4? loses to 14.Bc4.
11...b6 12.e4!. White has only tried Ba3 and
Qb3. I believe that his activity on the queenside
could only be effective if he had a better centre.
Thus e4 is indispensable.
12...Bb7 (12...Nf6 13.Bd3) 13.e5 cxd4
14.cxd4 Bd5 15.Be3
The bishop is the proud owner of the
d5-square, but Black lacks good prospects for his
other two pieces. He would hardly survive without
undermining the e5-outpost. However, that would
create new weaknesses in his position. I analysed
two possible scenarios:
a) 15...f6 16.Rc1
The normal place for the rook is b1 (16.Rb1
a5 17.exf6), but the computer wants to drag the
enemy king to h8 first.
16...Kh8 17.Rb1 a5 18.Qd2 fxe5 19.Nxe5
Be4 20.Nxd7 Qxd7 21.Rb5²;
b) 15...Qb8 16.Re1 Qb7 17.Rb1 f6 18.Bf1
Rad8 (18...Bxf3 19.gxf3 fxe5 20.Bg2!±) 19.exf6
exf6 20.Qe2 Kh8 21.Bd2 Ra8 22.Qe3 Rfc8 23.Bd3².
White’s heavy pieces could penetrate through e7.
Another typical Grünfeld set-up is:
10...Qc7 11.e4 Rd8
135
12.Bc4!, underlying the weakness of f7.
The point is 12...cxd4? 13.Bxf7+ or 12...e6
13.Bg5. 12...Nf6 13.Qb3 e6 14.e5 Nd5 15.Bg5 Ne7
16.Rfe1 h6 17.Bf6 is also better for White. Perhaps
he should play:
12...Nb6 13.Ba2 e6 14.Bg5 Re8 15.a5 Nd7
16.Qd3².
11.a5 Bb7
With a hindsight, 11...bxa5 deserves
attention, in order to stop the pawn’s march to a6.
11...cxd4 12.cxd4 bxa5 is inferior on account
of 13.Bd2.
12.Qa4 Nf6
12...Nb6 13.Qxa5 cxd4 14.cxd4 Bb7
occurred in Gustafsson-Baskin, Germany 2015.
15.Be2 (15.Ba3!? Bxf3 16.gxf3 Qd5
17.Bxe7 deserves attention – 17...Qxf3 18.Bh4 h5
19.Bg3 Qg4 20.Qd2 h4 21.Be2 Qg5 22.f4 Qe7
23.Bf2. I guess that White’s centre and bishop pair
promise him an edge.) 15...Qc7 (15...Qd7 16.Ba3 should not give up any space.
Rfc8) 16.Bd2 Rfc8 17.Rfc1 Qd8 18.h3 e6 19.Rxc8
Rxc8 20.Bb4 Bf8 21.Bxf8 Qxf8 22.Ne5 Qe7 23.Bf1 13...Nf6 14.Ne5 Rc8
Nd7 24.Qb5 Nxe5 25.Rxa7 Nf3+ 26.gxf3 Rc7
27.f4².
13.Qxa5 Bb7 14.Ne5 a6 15.Bd3 Rc8 16.Nc4 14...Nd5 15.Bb2 e6 16.f3± Nxe3 17.fxe4
Be4 17.Be2 Bc6 18.Ne5 Bb7 19.Nd3 Ne4 20.Bf3². was pretty hopeless for Black in Comtesse-Pheby,
ICCF 2012.
12.a6!
15.f3 Ba8 16.e4 cxd4 17.cxd4 Ne8

The only logical conclusion of the idea

a2-a4-a5. I played 12.axb6?! axb6 13.Rb1 Nf6
14.Bd3 Qc7 15.Re1 c4! 16.Bf1 Ne4 17.Qc2 f5 and
Black was in command, Kiril Georgiev-Krasenkow,
Leon 2010.

18.Be3

I prefer to regroup by 18.Rd1!² Nd6 19.Bf1

Qc7 20.Bb2. The idea is to push d5 – 20...b5 21.Qb4
12...Be4 Rfd8 22.d5 Nf5 23.Nxg6 hxg6 24.exf5.

In Datler-Feco, ICCF 2014, Black provoked
c4 with 12...Bd5 13.Qe2 Qc7 14.c4 Be4
14...Bxf3 15.Qxf3 cxd4? fails to 16.Qb7±.
15.Bb2 e6 16.Rad1 Rad8 17.Rd2 cxd4
18.exd4 Bc6 19.Rfd1 Rfe8
19...Bxb5 20.cxb5 Nf6 gains the d5-square,
but surrenders c6 – 21.Ne5 Nd5 22.Nc6 Rd7 23.g3².
20.h4 Bh6 21.Rc2 Bg7 22.h5 Rc8 23.Rcc1².
In all the lines White’s main trump is the a6-pawn. It
could decide any endgame.
13.Qa4
18...Nd6 19.Be2 e6
Black is struggling against White’s centre.
19...Rc3 20.Bf2 Qc8 21.Rad1 f6 22.Nd3 Bh6
23.Nb4².
20.Rfc1 Qe7 21.Qa3 Rfd8 22.Nc4 Bf8
23.Nxd6 Qxd6 24.Qxd6 Bxd6 25.Bg5 Rf8 26.Bb5
As I noted before, the endgames are difficult
for Black.

13.Ng5 Nf6 14.Nxe4 Nxe4 15.Bb2 Rc8

16.Bd3 Nd6 was balanced in Korobov-Volokitin,
Rivne 2016, although White went on to win. White

136
 17.Qxa4+ Nc6 18.Nxc6 bxc6 19.Ke2 Qb2+ 20.Bd2
Qb5 21.Qb3 e6 22.Rab1 Qxb3 23.Rxb3²) 15...e6
16.b4 Nb8 17.Rb1 Qc7 18.Bxd7+ Nxd7 19.Qa4 b6
20.g3 Bd6 21.bxc5 bxc5 22.Kg2 0-0 23.Rhc1².
12.Nbxd4

26...Kg7 27.e5 Bb4 28.Bd7 Rc3 29.Kf2

Rb8 30.Rab1 Ba5 31.Rxc3 Bxc3 32.Ke3 Ba5
33.Bf6+ Kh6 34.g4 Bd5 35.f4 Bc4 36.Rg1 g5
37.fxg5+ Kg6 38.Bc6 Bd5 39.Bb5 1-0

12...cxd4
23. Khalifman – Paravyan Or 12...Nxd4 13.Nxd4 Qxd4 14.Bd2 Qxb2
Kaliningrad 01.07.2015 15.Rb1 Qd4 16.Qa4+ Bd7 17.Qa5±.
13.Rh3 with prospects for an attack.

1.Nf3 Nf6 2.c4 c5 3.Nc3 d5 4.cxd5 Nxd5 10.h4 h5!

5.e4 Nb4 6.Bc4 Nd3+ 7.Ke2 Nf4+ 8.Kf1 Ne6 9.d3

As a rule, Black should not allow h4-h5. For

instance: 10...Nc6 11.h5 Bg7 12.Be3 Ned4 13.Rc1
0-0 (13...Bg4 14.Bxf7+) 14.hxg6 hxg6 15.Qd2 Nxf3
16.gxf3 Ne5 17.Qe2 b6 18.Rd1 e6 19.Bb5².

11.Ng5

White could win the c5-pawn, but he would

remain with passive pieces and without a plan after
11.Be3 Bg7 12.Na4 Nc6 13.Bxe6 (13.Nxc5 Nxc5
14.Bxc5 0-0) 13...Bxe6 14.Nxc5 Bg4 15.Nxb7 Qb8
16.Na5 Qxb2 17.Rc1 Nxa5 18.Qa4+ Bd7 19.Qxa5
0-0 20.Kg1 Bg4 21.Kh2 Qf6. Black has enough play
9...g6 against the weak white pawns.

11...Nc6!
Perhaps 9...Nc6 10.h4 h5 would transpose if
Black played ...g6 later. The plan with 10...Ned4
yields worse results: 11.Nb5 e6
Or 11...Nxb5 12.Bxb5 Qb6 13.Ba4 Bd7
14.Be3 Rd8 (14...e6 15.b4 Nxb4 16.Bxd7+ Kxd7
17.Ne5+ƒ) 15.a3!? (15.b4 Nxb4 16.Ne5 Bxa4
137
 15.Rg3! 0-0-0 16.Rxg6 Ne5 17.Rg3 Nxd3
18.Qb3 was clearly better for White.

15...b6 16.Qa4?! (16.Rg3) 16...0-0-0

17.Rd1 Kb7 18.Rg3 Rg8?! (18...Bg7)

Black takes a passive stand and passes the

initiative back to White. A good continuation would
be 19.Ne2 and Black cannot prevent d4 because:

12.Nxe6

At first glance 12.Be3!? Bg7 13.Bxe6 Bxe6

14.Nxe6 fxe6 15.Bxc5 wins a healthy pawn.
However, White cannot establish coordination
between his rooks. That assures Black of enough
counterplay. In my analysis I reached some
paradoxical positions, e.g. 15...0-0 16.Rh3 Qa5
17.Be3 Rad8 18.Qb3 Rxd3 19.Qxe6+ Kh7 20.Rg3
Rf6 21.Qc4 Ne5 22.Qa4 Qxa4 23.Nxa4 Ng4 24.Ke2
19...e5 20.Bxc5! bxc5? 21.Qb3+ is a double
hit. Khalifman opts for a strategic solution – he seals
down the black bishop.

19.e5!? Nb4

Or 19...Nd4 20.Qxd7+ Rxd7 21.f4.

20.Qxd7+ Rxd7 21.Ke2 Nc6 22.f4 Bh6

23.Rf3 Rgd8 24.g3

24...Rxe3+ 25.fxe3 b5=.

12...Bxe6 13.Bxe6 fxe6 14.Rh3 Qd7?!

Black decides to hide his king on the

queenside, but that leaves the g6-pawn weak.
14...Bg7! 15.Rg3 Kf7 holds on – 16.f4 Qd7 17.e5
Rad8 18.Ne4 b6 19.Ng5+ Kg8=.

15.Be3
138
 24. Bu Xiangzhi – Zeng
China 27.11.2018

1.Nf3 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5

5.h4 h6 6.e4 Nxc3

The cage is complete, and White’s advantage

is clear.

24...Nb4 25.Bg1 Nd5 26.d4 c4 27.Ne4 Rc7

28.Bf2 Nb4 29.a3 Nd5 30.Be1 Rdc8 31.Bd2 Rd8
32.Rc1 Rdc8 33.Rc2 Rc6 34.Rf1 R6c7 35.Rfc1
Rc6 36.Kf3 a5 37.Ke2 R6c7 38.Be3 Rc6 39.Nd2 c3
40.bxc3 Nxc3+ 41.Kf3 Nd5 42.Rc4 a4 43.Ke2 7.dxc3
Rxc4 44.Nxc4 b5 45.Nb2 Rc3 46.Rxc3 Nxc3+
47.Kd3
7.bxc3 requires some experience in the
Grünfeld – 7...Bg7 8.d4 c5 9.Be3
9.Rb1 0-0 10.Be2 b6 (Ivanchuk use this
move to beat Mareco in 2018) 11.Be3 (11.0-0 e6)
11...Bb7 12.Bd3 Nd7 13.e5 e6 14.Rh3 looks
promising for White, but 14...h5 15.Rg3 Rc8
16.Ng5 Nxe5! gives Black counterplay.
9...Qa5

47...Nb1

After 47...Nd5 White transfers his knight to

c5 to win the e6-pawn.

48.Bc1 Kc6 49.Nd1 b4 50.axb4 a3 51.Nc3

Bf8 52.Kc2 1-0 White has tested here only 10.Qd2 Nc6
11.Rb1 or 11.Rc1. I would borrow a modern idea
from the “original” variation without the inclusion
of h4 h6:
139
 10.Nd2!? Bd7 11.Rc1 cxd4 12.Nc4 Qa4, and 0-0-0
even without queens White has a slight initiative. Note that the black pawns on the kingside
are fixed on light squares, so 18...Bxb3+ would not
7...Qxd1+ 8.Kxd1 Nd7 equalise.
19.Nd4 Bf7 20.f4².

The alternative is 8...Bg4, employing the 10.Kc2 Bg7 11.Nd2 Be5 12.Be3 Nf6 13.f3
problem bishop. Nh5 14.Bc4 Bf4

However, if Black took on f3 after 9.Kc2, he

would hand the opponent the bishop pair and a
better control of the centre:
9...Bxf3 10.gxf3 h5 11.f4 Nd7 12.f5!²,
Naiditsch-Ni Hua, Wijk aan Zee 2010. That’s why
the second players prefer:
9...Nd7, when you should remember to
develop firstly your king’s bishop:
10.Bc4 (or 10.Bb5!? a6 11.Bc4).
The idea is to discourage 10...e5?! in view of
11.Nh2, and 10...Bxf3 11.gxf3 e5 12.f4 also favours
White, e.g. 12...Bd6 13.fxe5 Nxe5 14.Bb5+ c6
15.Be2².
10...Bg7 11.Bf4 Nc5 12.Nd2 Be6 13.f3 0-0-0
14.Be3 Bxc4 15.Nxc4 Nd7 16.a4 a6 17.h5 g5 18.g3
Rhg8 19.Rad1. White puts his hopes on his better
bishop.
9.Bf4 c6
15.Bf2
White has more space, so he wants to keep
more pieces. Still 15.Bxf4 Nxf4 16.g3 Ne6 17.b4
Bd7 18.a4 was a decent alternative.
15...0-0 16.a4 Kh7 17.Rad1 f5?!
A horrible weakening. White did not have an
invasion square down the d-file, but now his rook
will target e7, which is much more vulnerable.
17...b6 18.Nb3 c5, planning ...Bg3, was only slightly
better for White
18.Rhe1 Bg3 19.Bxg3 Nxg3 20.Kc1± f4
21.Nb3

9...Nc5 10.Nd2 Ne6 11.Be3 h5 12.a4 Bh6

13.Bxh6 Rxh6 solves one of Black’s problems, but
meanwhile White seizes space on the other wing:
14.b4 Ng7
14...a5 only helps White’s play after 15.Nb3
axb4 16.cxb4.
15.Bb5+ c6 16.Be2 Be6 17.Kc2 f6 18.Nb3

140
 Black kept the e-file closed, but White’s
knight has obtained access to e6. That decides the
game eventually. Besides, the c8-bishop has lost his
only available place to go. White could use that to
double his rooks on the d-file and win the battle for
the 8th rank.

21...g5 22.hxg5 hxg5 23.Rd2+– g4 24.Nc5

Kg6 25.Red1 gxf3 26.gxf3 Rh8 27.a5 Kf6 28.Nd7+
Bxd7 29.Rxd7 Rh5 30.Rxb7 Rxa5 31.Rdd7 Rc5
32.b3 a5 33.Rxe7 a4 34.Re6+ Kg5 35.Rg7+ Kh5
36.Rd6 Rxc4 37.bxc4 a3 38.Kb1 a2+ 39.Ka1 Ra3
40.Rd2 Kh6 41.Rg4 Rxc3 42.Rh2+ Nh5 43.Rgh4
1-0

141
 Chapter 7. Anti-Nimzo
Main Ideas

White commonly plays 1.c4 to avoid the

Nimzo/QID set-up. However, he also has to take
into consideration his repertoire against the Queen’s
Gambit. If he is a fan of the Carlsbad’s structure
with cxd5, after 1.c4 Nf6 2.Nc3 e6 his only choice is
3.e4, which is the subject of Chapter 8. Objectively,
3.Nf3! offers more chances, but you should be ready
to meet 3...d5 by 4.d4. (I discuss my attitude
towards 4.e3 in Chapter 11.) However, in that case
you should also be prepared for the Ragozin and the
the Vienna Variation, which are beyond the scope of Then we put our rook on d1 and aim for
this work. f4-f5 or e4-e5 (after ...g6). Note that the thematic
plan with g4-g5 is ineffective with a bishop on c2
since we cannot attack the h5-square after ...g6,
...Nh5.
Our bishop pair can often be brought to life
by a pawn or even piece sacrifice. For instance:

Serper-Epishin
Soch 1984

I analyse two moves from here:

A. 3...b6

I’m not a fan of the Sicilian Hedgehog, but I

must admit that this introduction towards it has more
merits than 1...c5 followed by b6. The point is that
4.e4!? Bb7 forces the ugly 5.Bd3. Although we
cannot use the set-up with Qe2 from Chapter 4, we 19.f5!! Nxe5 (Or 19...exf5 20.e6!; 19...Bxe5
still possess an aggressive plan on the kingside, 20.fxe6 fxe6 21.Rf7 Nf4 22.Rxf4 Bxf4 23.Nxh7;
based on f4+f5 or e5. We aim to reach the following 19...exf5 20.e6) 20.Nxf7, winining the exchange.
position:
5...c5 6.0-0 d6 7.Bc2 Be7 8.d4 cxd4 9.Nxd4 Romanishin-T.Petrosian
0-0 10.b3 a6 11.Bb2 Nbd7 12.f4 Qc7 13.Kh1 Rfe8 Yerevan 1975
14.Qe2

142
 only point at one nuance. You should not allow 8.e3
Bg4! 9.Be2 e4! 10.Ng1 Bxe2 11.Nxe2 Nbd7 12.Ng3
Ne5 13.0-0 Re8 14.Bb2 Nd3=. The right move order
with 8.Bb2 is a good workaround as 8...Bg4 could
be attacked by 9.h3 and g4.

In the middlegame White should obligatory

open the long diagonal to activate his b2-bishop:

16.Nd5! exd5 17.Nf5!!

White actually played 17.exd5?! Nxd4
18.Qxd4 Rde8 19.f5! Bd8 20.Qh4 Re5!, and Black
should have held.
17...dxe4 (17...gxf5 18.Rg3+ Kh8 19.exd5
Rg8 20.Re3+–) 18.Rg3 Qc7 19.Qh5 (intending
£h6+–) 19...Rfe8 20.Qh4‚.

36.c5! dxc5 37.Bc3!‚.

B. 3...Bb4 4.Qc2 0-0 5.a3 Bxc3 6.Qxc3 If White opens the g-file, he might not even
leads to a completely different game. I propose to need d4:
fianchetto only our queen’s bishop and put the other
one on e2, exactly in this order!

14.f4! Qh4 (or 14...exf4 15.exf4 Qh4

16.Bxb7²) 15.fxe5 Rxe5 16.f4.
6...d6 7.b4 e5 8.Bb2
One last example of the importance of the
The modern theory of this line has been g-file in this line:
“written” by the games between the best chess
players in the world – the AI project AlphaZero and
Stockfish, in London 2017/2018. They showed how
to use the power of the bishop pair in seemingly
peaceful and safe positions. Be sure to look at the
commented games – they speak for themselves. I’d
143
 AlphaZero’s move 24.gxf3!! abruptly
changed the slow pace of the game. White’s rooks
took command from g1 and f4!

The other important set-up is 6...b6 7.b4

Bb7 8.Bb2 d6 9.e3 Nbd7 10.Be2 c5 11.0-0 Qe7

White has two possible plans here:

d2-d4-dxc5 + b5; a4-a5.
However, 12.d4 Ne4 is double edged as
Black has a series of easy active moves – ...f5,
...Rf6.

I think that we should not let the opponent to

attack us.

I recommend to postpone d4, making some

other useful moves first. At the same time we should
be always ready to counter ...e5 by d4. The only way
to do it is 12.Rfe1!, protecting Be2. We can follow
up with d3, Qb3, a4, even h3, and only then d4. That
should give us a lasting pull in a safe position.

144
 Chapter 7. Anti-Nimzo
Step by Step

1.c4 Nf6 2.Nc3 e6 3.Nf3

7...Be7

7...e5 would have been too easy, if White did

not have 8.b4 Be7 (8...cxb4 9.Nd5) 9.bxc5 bxc5
10.Rb1 Qd7, Bauer-Medvegy, Budapest 2017.
A. 3...b6; B. 3...Bb4 White should use his initiative to launch an attack
against the enemy king: 11.Nh4! g6 12.d3 0-0
3...c5 4.e3 is covered in Chapter 2, while I 13.f4².
discuss 3...d5 in Chapter 11.
8.d4 cxd4 9.Nxd4 0-0 10.b3
A. 3...b6 4.e4

I do not like the plan with Be3 here – 10.f4

Of course it is always possible to turn to the Nc6 11.Nxc6 Bxc6 12.Qe2 a6 13.a4 Rc8÷.
Petrosian System with 4.a3 Bb7 5.d4, or to the main
line QID. I do not consider here 4.g3 since Black 10...a6
has been quite confident against this set-up lately.

4...Bb7 5.Bd3 10...Nbd7 11.Bb2 transposes.

10...Nc6 could be met by 11.Nxc6 Bxc6
12.Qd3, creating the threat Nd5. Black should
White does not have any serious alternatives: neutralise it with 12...g6, and 13.Bg5 d5 does not
5.Qe2 c5 6.e5 Ng8 7.d4 cxd4 8.Nxd4 Nc6 equalise owing to 14.exd5 exd5 15.Rad1.
9.Nxc6 Bxc6÷;
5.e5 Ne4 6.Bd3 Ng5; 11.Bb2 Nbd7
5.d3 d6 6.g3 Be7 7.Bg2 c5 8.0-0 0-0 9.Ng5
Nfd7 10.h4 Nc6 11.f4÷.
11...Qc7 12.Qe2 Nbd7 transposes.
5...c5 6.0-0 d6 7.Bc2 11...Nc6 12.Kh1 Qd7 13.Nxc6 is similar to
the previous note – 13...Bxc6 14.Qd3 b5 15.cxb5
axb5 16.Nd5! Bxd5 17.exd5 g6 18.dxe6².

12.f4 Qc7 13.Kh1 Rfe8

(13...Rac8 14.Qe2)

145

14.Qe2
I prefer this queen development because it is
connected with a clear plan – White prepares e4-e5.
The queen lift 14.Qf3 (heading for h3 or g3)
is less dangerous than it looks – 14...Bf8 15.f5 (or
15.Qg3 g6) 15...e5 (15...Ne5 16.Qg3) 16.Nde2 b5
17.g4 h6 18.h4 d5 19.cxd5 b4 20.g5÷.
The engines recommend 14.Qd2. In my
opinion, this idea is not too natural as it dooms the
bishop at c2 to a sad future.
The “secret” of their assessment is that they
do not look for a constructive plan. They only see
that White conserves his grip in the centre, and that
is enough for them. As a result, Stockfish (at depth
47!) produces the following “masterpiece” after
14...Bf8: 15.Rad1 g6 16.Nde2 Rac8 17.Bb1 Nc5
18.h3 Qe7 19.Ba1 Red8 20.Qe3 Bh6 21.Bc2 Rc7
22.Kh2 (+0.67) etc. I explored more sensible ideas
for White:
14...Bf8
14...Rac8 is not the best place for the rook. It
would stay better on d8, in opposition to the Qd2.
15.Rae1 Bf8 16.Nf3 g6 (16...b5 17.e5) 17.Ng5 Re7
18.Ba3 Nc5 19.e5 Ne8 20.Be4 h6 21.Nf3 with
tangible pressure.
15.Rad1
15.Rae1 Rad8! 16.Nf3 g6 17.f5 exf5 18.exf5
Rxe1 19.Rxe1 Ne5 provides counterplay, for
example, 20.Ne4 Nh5 21.Bd4 Bg7 22.Qf2 Nxf3
23.Bxb6 Qc6 24.gxf3 Re8 25.Rd1 Nf6 26.Kg2
Nxe4=.
15.Nf3 Rad8 16.Rad1 g6 17.f5 is not much
better – 17...exf5 18.exf5 Bg7 19.fxg6 hxg6 20.Ng5
146
Ne5 21.Qf4 b5 with typical Sicilian counterplay.
15...Rad8 16.Qf2!? (correcting the queen’s
placement) 16...g6, when White seizes the initiative
with 17.f5 exf5 18.exf5 Bg7 19.Nd5², transposing to
the main line with 16.Qf2.
14...Bf8
If Black saves this move, we get additional
options:
14...g6 15.Rad1!?
15.Rae1 Rad8 discourages e4-e5 in view of
the hanging Nd4. The computer assures me that
16.f5 exf5 17.Nxf5 Bf8 18.Ne3 or 18.Nd4 favours
White, but I have played such positions with Black
and find them rather resilient. So we make a more
useful move.
15...Rad8 (15...Bf8 16.e5) 16.f5!?
16.Qf2 Bf8 17.f5 is also possible – see the
last diagram.
16...exf5 17.exf5 Bf8 18.Qf2 Bg7 19.Nd5².
15.Rad1
We need our rooks on d1 and f1. 15.Rae1 is
also possible, but I could not find a good follow up
after 15...g6.
The wrong move order 15...Rad8 16.Nf3 g6
allows the brilliant blow 17.e5 Nh5 18.Ng5 Bg7
19.f5!! Nxe5 (19...Bxe5 20.fxe6 fxe6 21.Rf7;
19...exf5 20.e6) 20.Nxf7 exf5 21.Nxd8 Qxd8
22.Nd5².
16.e5 (16.Nf3 Nh5) 16...Nh5 17.Be4 Nxf4!
and White can only maintain the balance after
18.Qf3 Bxe4 19.Qxf4 Bf5 20.Nxf5 exf5 21.Nd5
Qb8 22.exd6 Qxd6 23.Qd4 f6 24.Qh4=.
15...Rad8
Here 15...g6 16.e5 Nh5 17.Be4 Nxf4?
already loses to 18.Rxf4 dxe5 19.Ndb5 and Rxd7.
15...Rac8 stops 16.f5 in view of 16...b5, but
the break e4-e5 is effective:
16.Nf3 Bc6 (16...b5 17.e5) 17.e5 Bxf3
18.Rxf3 dxe5 19.fxe5 Qxe5 20.Ne4 Qxb2 21.Nxf6+
Qxf6 22.Rxd7 Qg5 23.Rfxf7 Rcd8 24.Rxd8 Rxd8
25.Qxe6 Kh8 26.Rf1².
 White has several ways of expanding his
initiative from here:
16.Nf3!? g6 17.e5 dxe5 18.fxe5 Ng4 19.Ne4
Bxe4 20.Qxe4 h5 21.Qe2 Nh6 22.Ng5 Nf5 23.Bxf5
exf5 24.e6 fxe6 25.Nxe6 Rc8 26.Nxc7² – Black’s
7th rank is weak.
16.f5!? e5 17.Nf3 b5 18.Bd3 bxc4 19.Bxc4
Nc5 20.Ng5 Rd7 21.Nd5²;
16.Qf2 g6 17.f5 exf5 18.exf5 Bg7 19.Nd5
Bxd5 20.cxd5 Ne5 21.Rc1 Qd7 22.Nc6 Rc8 23.Qd2
Rc7 24.Bb1².
B. 3...Bb4 4.Qc2
The only plus of 4.g4 is that it leads to
chaotic play. As a rule, this advance makes sense
when Black has made passive moves on the
queenside, like ...c6 and b6. In our case, Black can
semi-automatically answer 4...h6 5.Rg1 (5.Qc2 c5)
5...d5, asking White to show his hand.
147
White commonly chooses here 6.Qa4+ Nc6
7.Ne5 ( 7.Nd4 Rb8), when 7...Qd6 8.Nxc6 bxc6 9.a3
Bxc3 10.dxc3 Bd7 would leave him somewhat
uncoordinated. I also investigated:
6.a3 Be7 7.d4 a6!? with complex play –
8.cxd5 exd5 9.Qb3= (9.g5 hxg5 10.Bxg5 Be6
11.Qc2 Nbd7 12.0-0-0 Rc8÷) or 8.g5 hxg5 9.Bxg5
dxc4 10.Ne5 Nc6 11.Nxc6 bxc6 12.e3 Rb8 13.Qc2
Nd5.
4...0-0 5.a3 Bxc3 6.Qxc3
This is AlphaZero’s pet line for White. It has
scored several memorable victories against
Stockfish.
B1. 6...d6; B2. 6...b6
6...a5 basically wastes a tempo at an
important moment. Play may transpose after 7.b4 d6
8.Bb2! e5 9.e3 Bg4.
 The only independent line is 8.e3 Ne4 9.Qc2
Ng5 10.b5 Nxf3+ 11.gxf3 – look at Game 27
AlphaZero-Stockfish 8, London 2018. It is one of
the most beautiful games I have ever seen!
However, I’m not sure White has more than a draw
against best defence.
B1. 6...d6 7.b4 e5
8.Bb2!
An important detail – White delays e3.
AlphaZero tested once 8.e3 Bg4! 9.Be2 e4 10.Ng1
Bxe2 11.Nxe2, but the weakness of its light squares
allowed Stockfish to equalise with 11...Nbd7
12.Ng3 Ne5 13.0-0 Re8 14.Bb2 Nd3 15.f3 Nxb2
16.Qxb2 Qe7.
I checked 12.0-0 Ne5 13.Nf4, to prevent
...Nd3. Then Black has 13...Rc8! 14.d3 (14.d4
Ng6=) 14...c5! 15.bxc5 exd3=.
The above examples show that the
evaluation of the bishop trade depends on the tempi.
We should avoid it in this concrete position, but
we’ll soon see that it is not so efficient later.
8...Re8
8...Bg4 is already pointless as we have 9.h3
Bh5 10.g4.
8...a5!? does not look like the best way of
spending a tempo in the opening. White has the
bishop pair, so he should welcome open files. And if
148
Black refrains from taking on b4, White could take
himself on a5 at some moment. Still, 8...a5 is often
played, as it is the only neutral move which retains
the option of ...Bg4:
9.e3
9.g3 is a viable alternative, but the highest
chess instance today – AlphaZero, played invariably
e3.
9...Bg4
9...axb4 10.axb4 Rxa1+ 11.Bxa1 Bg4
12.Be2 e4 13.Nd4 Bxe2 14.Nxe2 Nc6 15.0-0 Ne5
16.Ng3 favourably differs from the
above-mentioned game of AlphaZero – Black is a
tempo down, and White’s bishop is not on b2. Thus
...Nd3 would be senseless. Chuchelov-Taimanov,
Fuerth 1999, went 16...Qd7, when White could
attack the e4-pawn with either 17.h3 (preventing
...g4) 17...h5 (17...Qe6 18.c5) 18.Qc2 or 17.c5 Re8
18.cxd6 cxd6 19.Qc2.
10.Be2
White could prevent ...e4 by 10.d3, but that
would hamper his future plans. For instance, after
10...Bxf3 11.gxf3, his normal goal would be to push
f4 twice, but then the e3 pawn would be weak.
10...axb4
10...Bh5 is not a human move, although it
does not change much – Black will play it anyway
after h3. See Game 25 AlphaZero-Stockfish 8,
London 2018. It illustrates play without ...axb4.
11.axb4 Rxa1 12.Bxa1 Nbd7 13.d3 Qe7
14.0-0 h6 15.h3 Bh5 16.Bb2 Ra8 17.Ra1 Qe8
18.Rxa8 Qxa8
 This position occurred in 10...Nbd7
Gustafsson-Almasi, Reggio Emilia 2009. Instead of
trading queens prematurely with 19.Qa3. White
should keep manoeuvring, cashing on his space A key moment is that 10...e4?! stumbles into
advantage on both wings. Plausible continuations 11.h3! exf3
are 19.Qb3 b6 20.Ne1, or 19.Nd2 Qa6 20.Nf1 Or 11...Bh5 12.g4 exf3 13.Bxf3 Bg6 14.g5
intending f2-f4-f5. Nbd7 15.gxf6 Ne5 16.Be2².
12.hxg4 fxe2 13.g5 and Black lacks the
9.e3 Bg4 e8-square for the knight.
Thus White succeeds in completing
development, and his chances are slightly higher in
9...Nbd7, now or on the previous turn, is view of his bishop pair.
inconsistent as it blocks the bishop. After 10.d3 Nf8
11.Be2 Zhigalko chose against Meribanov in 2016: 11.d3

a) 11...Ng6 12.0-0 Bd7 13.a4 a5 14.b5 c5

15.Nd2 Bf5 16.Bf3 Qd7, when 17.Bd5 is obviously 11.0-0 is also possible – 11...e4 12.Nd4 Bxe2
pleasant for White. 13.Nxe2 Ne5 14.Nf4 a5 15.d3.

b) 11...a5 12.0-0 Bg4 was Stockfish’s choice 11...h6 12.h3 Bh5 13.0-0 Qe7²
against AlphaZero. See Game 26.

10.Be2

The real fight is yet to begin. Unfortunately

149
my database does not offer any practical example, (19.Bf3) 19...Qc7 20.Qd1 Ra8 21.Rxa8+ Bxa8
but the annotated games of AlphaZero should shed 22.Qa1², hoping to break on the kingside later on.
some light on White’s plans. Note AlphaZero’s
manoeuvre Rfd1 and Qc3-e1. Generally speaking, 8.Bb2 d6
White’s main positional goal should be to shake the
e5-pawn, which stops our bishop from revealing its
full potential. On the queenside, he could take on a5
if Black pushed it, to distract his opponent’s pieces
with the task of regaining the pawn.

B2. 6...b6 7.b4 Bb7

7...a5 offers a choice:

8.b5 keeps more pieces – 8...Bb7 (A
correspondence game saw 8...d5 9.cxd5 Qxd5 10.e3
c6 11.Bc4 Qf5, which opens play in favour of
White’s bishops.) 9.d3 d6 10.g3 and subsequent play
on the kingside. 9.e3
In line B1 we saw that AlphaZero does not
mind opening the a-file. Thus the main answer
should be: 9.g3 Nbd7 10.Bg2 is of equal worth in any
8.Bb2 axb4 9.axb4 Rxa1+ 10.Bxa1 c5 11.e3 aspect – its frequency and winrate are very close to
d6 12.Be2 e5 13.b5 (or 13.0-0 Re8 14.Bb2 Nc6 9.e3. White’s castling position is safer with a
15.b5 Ne7 16.d4) 13...Re8 14.d3 Nbd7 15.0-0 Bb7 fianchettoed bishop, but it would be more difficult to
16.Bb2 Nf8 17.Ra1 Qb8 find a constructive plan in the middlegame.

9...Nbd7 10.Be2 c5

10...e5!? is not too clear.

After 11.0-0 Ne4 12.Qc2 Ng5 13.Nxg5 Qxg5
14.f3 a5 White may have two bishops, but he lacks a
clear plan. More tension keeps:
11.d3, but then 11...c5 gives Black a better
version of the main line – in it we always meet ...e5
by d4, while now we’ll have to work for it.
Nevertheless the tempo is not vital, and White
retains good chance after:

Cheparinov-Almasi, Varadero 2016, went

18.Nh4 Ng6 (18...d5!? 19.Qe1 d4 20.e4) 19.Nxg6
hxg6 20.Bf1 d5 21.d4 Ne4 22.Qc2 dxc4 23.Bxc4
cxd4 24.exd4 Rc8, when White missed 25.Bxf7+!!
Kxf7 26.Qb3+ Ke7 27.f3 with a strong attack, e.g.
27...Ng5 28.Ba3+ Kd7 29.dxe5 Qxe5 30.Rd1+ Kc7
31.Qc2+ Kb8 32.Bd6+ Qxd6 33.Qxc8+±.

Another option is 18.Nd2 Ne6 19.Qb3

150

12.0-0 Re8 13.Rfe1 Rc8 14.Qb3 Qc7
15.Rac1 Qb8 16.d4², Bu,Xiangzhi-Dao,Thien Hai,
rapid Doha 2006.
As you understand, the most principled
answer should be:
11.d4!?, for instance: 11...exd4 12.Qxd4 Qe7
13.0-0. The Bb2 is now active, we should not fear an
attack, so I’d gladly take White here.
10...Qe7 11.0-0 e5 poses the same dilemma
as 10...e5.
12.d3 was the choice of Smyslov and most
GMs, but it is likely to cost White a tempo
eventually. Again, the most consistent is:
12.d4 Rfe8 13.Rfe1.
11.0-0 Qe7
After 11...Rc8 we could play the
prophylactic 12.Qb3 Qe7 13.Rfe1 as in the main
line.
11...e5 should be automatically countered by
12.d4 exd4 13.exd4 Re8 14.Rfe1 Nf8 15.d5±,
Agrest-Johansson, Sweden 2008.
12.Rfe1!?
It would be easy to understand this strange
move (Nimzowitsch would have been glad to see it
– he paid special attention to such manoeuvres) if
we delve deeper into the normal 12.d4 Ne4
12...Rac8 loses the focus of the battle, and
White is slightly better after 13.dxc5 bxc5 14.b5.
151
For instance: 14...d5?! 15.a4 Nb6 16.Nd2
(16.cxd5!?) 16...Rfd8 17.a5 Nxc4 18.Nxc4 dxc4
19.Rfd1 Rd5, M.Gurevich-Aagaard, Torrelavega
2007, 20.f3+–. Or 14...e5 15.Nd2 Ne8 16.a4 f5
17.Rae1 Nef6 18.f3.
13.Qb3 (13.Qd3 is not much different)
13...f5
Black’s game is too easy from a strategical
point of view – he will try to mount an attack against
our king.
Unfortunately, 14.bxc5 bxc5 15.Qxb7 is only
equal after 15...Rab8 16.Qxa7 Ra8 17.Qb7 Rfb8
18.Qc6 Rc8 19.Qb5 Rcb8 20.a4 Rxb5 21.axb5 Qf8
22.Rxa8 Qxa8 23.Ra1 Qb7 24.h4 g6 25.Ra6 h6=.
That explains why everybody chooses:
14.Rad1, but White does not have much
following 14...Rad8 (another thematic idea is
14...cxb4 15.axb4 b5!? 16.Nd2 Nb6) 15.Nd2 Nxd2
16.Rxd2 Rf6 17.d5 Rh6 18.g3 Nf6, e.g. 19.bxc5
bxc5 20.Bxf6 Rxf6 21.Rb1 Rf7 22.Rdb2 Ba8 23.Bf3
exd5 24.cxd5=.
On the other hand, d4 remains White’s
natural plan, and it would be more efficient if Black
misplaced some of his pieces. Even a move like ...h6
would be helpful for us since it would deny the rook
lift ...Rf8-f6-h6 which we saw above. These
considerations made me elaborate on 12.Rab1. It
looks like a useful waiting move, until I noticed
12...e5. Then the “obligatory” retort 13.d4? simply
loses the bishop on e2. Hence – the idea to protect
the bishop with the rook. Besides, it enables Bf1.
12...Rfc8
 12...e5 13.d4 Ne4 14.Qb3 (14.Qc2!)
14...Rab8 15.Rad1 was pleasant for White in
Ubilava-Huss, Zaragoza 1996.

We treat 12...Rac8 like 12...Rfc8 – 13.d3

Rfe8 14.Nd2 (14.b5!? a6 15.a4) 14...e5,
Vaisser-Rozentalis, Sverdlovsk 1984, when I like
most 15.d4².

13.d3

White’s position is more flexible. He can

pick up the best moment for d4, while Black should
wait. Ruck-Miroshnichenko, Austria 2009, went
13...Nf8 (13...a6 14.a4 e5 15.Nd2) 14.Nd2 Ng6,
when White should have taken prophylactic
measures:
15.Qb3
Ruck actually played 15.d4 Rc7 (15...cxb4
16.axb4 b5!?) 16.Rad1 (16.Bf1) 16...Rac8 17.dxc5
dxc5 18.b5 Rd7 19.Nf1 Rxd1 20.Rxd1 Rd8
21.Rxd8+ Qxd8 22.f3 Ne8 23.a4 a5, when 24.bxa6
Bxa6 25.a5 still retained some pull – 25...bxa5
26.Qa3 Qc7 27.Bc3 Nd6 28.Bxa5 Qc6 29.Nd2 Ne5
30.Qa2².
15...e5 16.d4 Ne4 17.Nxe4 Bxe4 18.g3 Nf8
19.f3 Bg6 20.Bf1 a5 21.Bh3².

152
 Chapter 7. Anti-Nimzo
Annotated Games

25. AlphaZero – Stockfish 8

match, London 18.01.2018

1.Nf3 Nf6 2.c4 e6 3.Nc3 Bb4 4.Qc2 0-0

5.a3 Bxc3 6.Qxc3 d6 7.b4 e5 8.e3 Bg4 9.Be2 a5
10.Bb2 Bh5 11.d3 Qe7 12.0-0 Nbd7 13.h3 h6

Now the most consistent follow up, 20.Qg3,

allows some counterplay – 20...axb4 21.axb4 Rxa1
22.Rxa1 d5 23.Qxe5 Qxb4„ or 20...b5 21.cxb5 Nd5
22.Qe1 Rab8, so the introductory 20.b5!?, similar to
the next game, seems a wise approach.
Stockfish’s choice to give up a pawn
transfers the fight in the centre.

16.bxa5 Na4 17.Rab1 Nd7 18.Ba1 Nac5

At the cost of a doubled pawn Black secured

14.Rfd1 the c5-square for its knight and fixed targets on the
a-file. Its future intentions comprise ...Ra7, ...Rea8,
even ...f5.
In another game AlphaZero preferred to wait
for a few moves with 14.Qc2 Rfe8 15.Rfc1 b6
16.Rf1 Bg6 17.Rfd1, when the “impatient” 17...e4
opened its bishop – 18.dxe4 Nxe4 19.Bd3 Ng5
20.Nd4 Bxd3 21.Qxd3 Qe4 22.Nb5 Qxd3 23.Rxd3
Rec8 24.h4 Ne6 25.f4².

14...Rfe8 15.Qe1 Nb6

It would be interesting to see AlphaZero’s

plan against some waiting move like 15...b6. If the
previous moves had been played by a human GM,
he would probably follow up with something like
16.g4 Bg6 17.Nh4 Bh7 18.Ng2 Nf8 19.f3 Ne6 Interestingly, even at a great depth Houdini
promotes as first lines passive waiting moves like
19.Rb2, 19.Rb5 or 19.Qd2, while Stockfish
demonstrates a better dynamic feeling, and around
depth 37 discovers the merits of the following break
in the centre:

153
 19.d4! exd4 20.Nxd4 Bg6 21.Rb2 f6 22.Ra2 Rd6 41.Qg3 Rad8 42.Bf5 Kf8 43.Rxd6 cxd6
Ne5 23.Bc3 Kh8 44.Ba1 Qe7 45.Bd3 Nc4

It would have been safer to rid White of his

bishop pair by 23...Ne4.

24.Nf3 Be4 25.Nxe5 fxe5 26.f3 Bc6

A controversial choice, which eventually

costs the game. 26...Bf5 keeps control of the b1-h7
diagonal.

27.Qg3 Qf7 28.Bb4 Ba4 29.Rc1 Bc6 30.Rf1

Re6 31.Kh2 Kh7 32.Qh4 Rf6 33.Rd2 Nb3?!

46.a4!+–
This move regains the pawn, but allows
White to activate the bishop. 33...Kg8 was more
resilient. Showing the reason behind the retreat
44.Ba1. Although White is conducting an attack on
34.Bd3+ Kh8 35.Rdd1 Nxa5 the kingside, the AI always plays all over the board
and does not miss any opportunity to increase the
pressure with stabs on the opposite wing. For
example, after 46...d5 47.axb5 Bxb5, its rook enters
play from b1!.

46...Na3 47.Qf4+ Qf7 48.Qh4 Rd7 49.Qg4

b4

The game will finish when White’s rook

joins in the attack. For instance, 49...bxa4 50.Bg6
Qe7 51.Rc1 prepares Bb2 and Rc4.

50.a5 Qe7 51.a6 Rd8 52.Qf4+ Ke8 53.Bb2

Nb5 54.Qg4 Qg5 55.Qc4 Nc7 56.Qc2 Kd7 57.Bf5+
36.c5! dxc5 37.Bc3! Ke8 58.Bg4 Nd5 59.f4 Qe7

After a seemingly quiet manoeuvring in

somewhat passive position White suddenly develops
a crushing attack. The leading role is for the bishop
pair. I think that this tragic outcome was a result of
Black’s consistent underestimation of the potential
power of the enemy’s pieces. Now Stockfish realises
that it is busted, but it is already late.

37...b5 38.Bxe5 Re6 39.Bb2 Kg8 40.Rfe1

154
 We saw in the previous game the manoeuvre
Rfd1+Qe1, preparing a queenside activity. Here the
AI shows an alternative. Its primary goal is to push
d4 in order to activate the b2-bishop.

14...h6 15.Bc3 b6

Black stubbornly refuses to exchange on b4,

obviously assessing that the open a-file could only
be in White’s favour. But then, why did it play ...a5
at all?! I believe that Black’s defence would be
simpler after 15...axb4 16.axb4 Qd7. Against a
60.e4! Nxf4 61.e5 bishop pair, it is advisable to maintain a symmetrical
pawn structure without weaknesses.

At last the dozing rook wakes up to seal the 16.b5

victory.

61...h5 62.Qf2 d5 63.Qxf4 hxg4 64.Qxg4
Ra8 65.Bc1 (yet another quiet move!) 65...Kf7
66.Bg5 Qe8 67.e6+ 1-0
26. AlphaZero – Stockfish 8
match, London 18.01.2018
1.Nf3 e6 2.c4 Nf6 3.Nc3 Bb4 4.Qc2 0-0
5.a3 Bxc3 6.Qxc3 d6 7.b4 e5 8.Bb2 Nbd7 9.e3 Re8
10.d3 Nf8 11.Be2 a5 12.0-0 Bg4 13.h3 Bh5
A surprising decision for Stockfish. If it were
a regular engine with White, we could muse upon
the imperfections of computers’ evaluating methods
and how they sometimes tend to overestimate space
and weaknesses.
However, AlphaZero operates on a
completely different basis! It does not care about
weak squares. It just plays many games against itself
and uses statistical data, measuring the probable
outcome after any move. Apparently it found that
even with a closed queenside it retained good
chances to win on the other wing. In that sense its
decisions are utmost pragmatic and not biased by
any human experience.

16...N6d7 17.Rad1 Nc5

14.Qc2

155
 should not forget that the engine was programmed
by humans. Since Tarrasch’s time we fear flaws in
our pawn structure, especially close to our king.
Stockfish also expected 24.Nxf3 with just a slight
edge. Yet opening the g-file dramatically changes
the prospects in White’s favour. It is enough to look
at the next diagram to understand in what a mess
Black engaged owing to a wrong positional
evaluation. Even the best defender in the world
proved unable to neutralise White’s pressure.
Correct was 23...Qe7 24.Bxe4 Qxe4 with a tenable
position.

24...Nfe6 25.Kh2 Nxd4 26.Rxd4 Kh7

18.Ba1 27.Qc2+ g6 28.Rf4 Qe7 29.Rg1 Rg8 30.h4 h5
31.Rg5 Kh6

Implacably preparing d4.

After the immediate 18.d4 Bg6 would be
awkward. White anticipates it, and now the queen
has a retreat to b2, avoiding ...Na4 or Nb3.

18...Bg6 19.Qb2 Na4 20.Qa2 Nc5 21.d4

exd4 22.Nxd4 Be4

According to the book Game Changer,

AlphaZero’s expected score here was 70.2%. The
rest does not need any comments. It is pure delight
to watch the AI increasing its domination with the
help of its king.

32.e4 Ne6 33.Rf6 Nxg5 34.hxg5+ Kh7 35.f4

Rae8 36.Qd3 Rg7 37.f3 Kg8 38.Qd4 Kf8 39.Bc3
23.Bf3! Bxf3? 24.gxf3!! Rg8 40.a4 Rd8 41.Kh3

This is the critical moment of the game. One The expected score has risen to 82.6.
move earlier White refrained from the mundane
f3+e4 because it would leave it with a “bad” 41...Rd7 42.f5 gxf5 43.Rxf5 Qe6
light-squared bishop. Although the AI loves space, I
have observed that it values much higher piece
activity. If AlphaZero does not see a way to activate
a piece, it seeks to exchange it. Meanwhile it set up
a positional trap and Stockfish failed into it! We
156

44.Kh4 Re7 45.Qd5 Rg6 46.Kxh5 Re8
47.Bf6 Qd7 48.Kg4 Rc8 49.Qc6 Qe8 50.Qxe8+
Kxe8 51.Rd5 Rxf6 52.gxf6 Kd7 53.Kf5 c6
54.bxc6+ Kxc6 55.f4 Rh8 56.e5 1-0
27. AlphaZero – Stockfish 8
match, London 18.01.2018
1.c4 Nf6 2.Nc3 e6 3.Nf3 Bb4 4.Qc2 0-0
5.a3 Bxc3 6.Qxc3 a5 7.b4 d6 8.e3 (8.Bb2!) 8...Ne4
9.Qc2 Ng5
10.b5!? Nxf3+ 11.gxf3 Qf6 12.d4 Qxf3
13.Rg1 Nd7
13...Qh5 14.Bb2 Nd7 15.Be2 (15.Qe4 Qf5
16.Qg2) 15...Qxh2 16.Rg3 transposes.
157
14.Be2 Qf6
Another transposition is 14...Qh3 15.Bb2
Qxh2 16.Rg3.
15.Bb2 Qh4 16.Rg4 Qxh2 17.Rg3 f5
18.0-0-0
To me, here is the critical position of the
game! The manoeuvre ...Nd7-f6-e4 looks strong.
The key line is:
18...Nf6! 19.Bf3 (19.Bd3 Ne4) 19...Ne4!
20.Bxe4 fxe4 21.Rdg1 Rf7 22.Qxe4 g6 (22...Qxf2?
23.Qh4) 23.f4 (23.d5 e5 24.Rxg6+ hxg6 25.Qxg6+
Kf8 26.Qg8+ Ke7 27.Qg5+ is perpetual) 23...Qe2
24.d5 e5, and again 25.Rxg6+ is a draw.
Thus the only reasonable way to keep the
tension is 19.Rdg1 g6 20.f3 (20.Qd1 Ne4 21.Rh1
Qxf2 22.Rf3 Qg2=) 20...Qh6 21.Qd2 b6
22.c5!? Nd5 23.Rg5°, but I would not feel
comfortable with two pawns less and without
imminent threats.
18...Rf7 19.Bf3 Qh4 20.Rh1 Qf6 21.Kb1 g6
22.Rgg1 a4 23.Ka1 Rg7
Somewhat imperceptibly, White has imposed
total control. All Black’s pieces remained very
passive. AlphaZero feels that it is time to open up
the centre:
24.e4 f4 25.c5±
White needs just one open file for his rooks.
25...Qe7 26.Rc1 Nf6
27.e5! dxe5 28.Rhe1!! e4 (28...Nd5 29.c6)
29.Bxe4 Qf8 30.d5 exd5 31.Bd3!
The final touch! Like in the first commented
158
game, the AI has found a way to clear the obstacles
before its tremendous bishops battery. For “mere” 3
pawns!
31...Bg4 32.f3 Bd7 33.Qc3 Nh5 34.Re5 c6
35.Rce1 Nf6 36.Qd4 cxb5 37.Bb1+–
37...Bc6 38.Re6 Rf7 39.Rg1 Qg7 40.Qxf4
Re8 41.Rd6 Nd7 42.Qc1 Rf6 43.f4 Qe7 44.Rxf6
Nxf6 45.f5
45...Qe3
The endgame is lost, again because of the
tremendous bishops!
46.fxg6 Qxc1 47.gxh7+ Kf7 48.Rxc1 Nxh7
49.Bxh7 Re3 50.Rd1 Ke8 51.Ka2 Bd7 52.Bd4 Rh3
53.Bc2 Be6 54.Re1 Kd7 55.Kb2 Rf3 56.Re5 Rg3
57.Re3 Rg2 58.Kc3 Rg4 59.Rf3 Ke8 60.Rf2 Rg3+
61.Kb4 Rg4 62.Rd2 Bd7 63.Ka5 Rf4 64.Be5 Rf3
65.Rd3 Rf2 66.Bd1 Bc6 67.Kb6 1-0

159
 Chapter 8. 1.c4 Nf6 2.Nc3 e6 3.e4
Main Ideas

1.c4 Nf6 2.Nc3 e6 3.e4

13...Nc6 (In the event of 13...Qe7 I propose

This
move is principled and aggressive. It demands from
both sides considerable home preparation. As most
lines which lead to a direct pawn clash in the centre,
it is well developed, and Black has established solid
equalisers against many variations. Our book also
contributes to their cause!
I’ll show you on which directions White’s
latest investigations are focused. But let’s see first
Black’s answer:
to keep the queens – 14.Qb3! d6 15.0-0-0) 14.Qc5
Qe7 15.Qxe7+! Nfxe7 16.0-0-0 d5 17.Be2!. White
has lasting pressure here, see Game 29
Zhigalko-Vaibhav, Lille 2013.
B. 3...d5 has eventually come to the fore.
4.cxd5 and 4.e5 yield the same decent 60%, so I
analyse both of them. In my opinion, the latter gives
more practical chances:
4.e5 d4 5.exf6 dxc3 6.bxc3 Qxf6

A. 3...c5 is the older main line. It has faded

out of fashion because of the following pawn
sacrifice:
4.e5 Ng8 5.Nf3! Nc6 6.d4 cxd4 7.Nxd4
Nxe5 8.Ndb5 a6 9.Nd6+ Bxd6 10.Qxd6 f6 11.Be3
Ne7 12.Bb6 Nf5 13.Qb4!

It looks that White should be better after

7.d4, intending Nf3, Bd3, but 7...e5! throws a
spanner into the works. The tactical justification of
Black’s idea is the pawn sac 8.Qe2 Be7! with
excellent compensation, while 8.Nf3 exd4 9.cxd4
Bb4+ trades bishops and White loses a vital
attacking resource. Dubov has tried 9.Bd3, but his
idea looks like a one-game novelty.

To sum up, if White exchanged any of his

160
bishops, his chances to develop an initiative n the
kingside would evaporate. This conclusion drew my
attention to 7.Nf3!?. The idea is to delay d4 and
avoid exchanges even at the cost of a tempo – 7...e5
8.Bd3!? Bd6 9.0-0 0-0 10.Be4. Black’s most
challenging defence is:
7...b6! 8.d4 Bb7 9.Bd3 Bxf3 10.Qxf3 Qxf3
11.gxf3

15.c5!ƒ.
Smerda-Mudra
ICCF 2012

We can reach this pawn structure in different

ways – e.g. 9.Bg5 Bxf3 or 9...h6 10.0-0 Bxf3.
Keeping the queens with 10.gxf3!? is much
sharper and double-edged. At first I liked a lot, but
the computer always finds coolers. You should try
10...Nd7 11.Be3!? Bd6 12.Rg1 h6 13.f4!?, when
Black must find the exchange sac 13...Bxf4!

Structures with split pawns

15.c5! bxc5 16.Bf4 cxd4 17.Rfe1 Kf7
18.Bc4 Re8 19.Rb7 Ba3 20.cxd4 with strong
Have another look at the last diagram. The pressure.
engines evaluate it in White’s favour because of the
better centre and the bishop pair. White scores rather Grandelius-Oparin
well OTB, in email and engines games. He plays Gjakova 2016
practically with the draw in the pocket as Black has
no active plan of his own. Yet my feeling tells me
that Black should be able to make a draw. The
good news is that apparently second players do not
know how to achieve it, and engines cannot give
them adequate advice! So let’s see how to exploit
their mistakes!
If Black managed to push ...c5 in a safe
position, he should draw. Thus our task is to get
there ahead of him!

Goganov-Shevchenko
Moscow 2016

161
 23.c5 Bf4, when 24.Bc6 or 24.h3 retain the 27.c5! Rdd8 28.Bc6 Rbc8 29.c4±.
initiative.
Finally, an example where the c5-break is
Chiron 2-Sting SF 4.7 not so efficient:
CCRL 2014
Goganov-Kovalenko,
Izhevsk 2012

20.c5! Be7 21.c4ƒ.

Williams-Norman 17.c5 bxc5 18.dxc5 Bxc5 19.Bf4 Bd6, when

ICCF 2013 20.Rxd6! cxd6 21.Bxd6 f6 22.Rb1 Rc8 23.Bb5 is
equal.

Remember that our advantage is dynamic. It

could be developed only by a break in the centre,
which would activate our bishops and the rook on
the b-file. Without it, we would stay nicely, the
computer will be assuring us we are fine, but the
game should be eventually drawn. We can do
nothing on the kingside and our target should be the
a-pawn.

Theoretical status

162
 This system is known as the Flohr-Mikenas
Attack. It has been in hectic development in the last
2-3 years, so we cannot talk about established main
lines. Only 3...c5 4.e5 Ng8 5.Nf3 Nc6 6.d4 cxd4
7.Nxd4 Nxe5 8.Ndb5 a6 9.Nd6+ Bxd6 10.Qxd6 f6
11.Be3 Ne7 12.Bb6 Nf5 13.Qb4 Nc6 14.Qc5 Qe7 is
a tunneled route. In The Full English Opening C.
Hansen considers only 15.0-0-0, which is rather
balanced.
I advocate 15.Qxe7+!.

3...d5 is much more eventful. White has

tested many ideas against it, but second players have
corked up several important novelties to neutralise
them. I preach to delay d4, when the most
challenging line is 4.e5 d4 5.exf6 dxc3 6.bxc3 Qxf6
7.Nf3 b6 d4 Bb7 9.Bd3 Bxf3. Hansen does not
consider this pawn structure at all, so my coverage
presents a new theme for discussion. Recent games
confirm that Black players does not understand well
how to set up his pieces, so that give us chances to
score some points before they learn their lesson.

163
 Chapter 8. 1.c4 Nf6 2.Nc3 e6 3.e4
Step by Step

1.c4 Nf6 2.Nc3 e6 3.e4

5.Nf3

Actually, 5.d4 cxd4 6.Qxd4 brings better

This system became very popular in the
70ies, and has returned triumphantly to the top in the
last few years.
practical results, so it could serve us as a backup
line. However, Black has no difficulties after 6...Nc6
7.Qe4 d6 8.Nf3 dxe5 9.Nxe5 Bd7 10.Nxd7 Qxd7
11.Bf4 Nf6 12.Qe3 Bb4 13.Rd1 Qe7 14.Be2 0-0
15.0-0 Rfd8=.

A. 3...c5; B. 3...d5 5...Nc6 6.d4 cxd4 7.Nxd4 Nxe5

3...Nc6 4.d4 Bb4 (à la Nimzowitsch) 5.e5

Ne4 6.Qc2 Nxc3 7.bxc3 Be7 is solid, but passive. It is better to struggle with an extra pawn
than for nothing after 7...Nxd4 8.Qxd4 Ne7 9.Bf4
3...e5!? is not ridiculous at all. I tested it Nc6 10.Qe3 Bb4 11.Be2.
unsuccessfully against Halkias in 2013, but the
cause of my difficulties was that after 4.Nf3 Nc6 8.Ndb5
5.Be2 I chose the overly timid 5...d6. Instead 5...Bb4
6.Nd5 Bc5 would be only slightly better for White.

A. 3...c5 4.e5 Ng8

8...a6

164
 After 8...f6 it is best to transpose to the main
line with 9.Be3 a6 10.Nd6+.
13...Qe7!? is less popular, but it is a worthy
9.Nd6+ Bxd6 10.Qxd6 f6 alternative. In fact, White has not proved a tangible
advantage after:

Black intends to contend for the d6-square
with ...Nf7. It is difficult to play the whole game
with dark-squared holes:
10...Nc6 11.Be3 Nge7 12.Bd3 0-0 13.0-0!
White usually castles long in this line, but here his
king would be safer on the kingside.
13...Ng6 14.c5 with a steel grip.
a) 14.Qxe7+ Nxe7! (intending ...d7-d5 or
...d7-d6).
He has adequate compensation for the pawn
after 15.f4 Nf7 16.0-0-0 d6, but at some point he
will have to take on d6 which will even the chances.
The best option is:
15.c5 d5! 16.cxd6 Nd5 17.Nxd5 exd5

11.Be3 Ne7

There is no much sense in 11...Nf7 as

besides the thematic retreat to b4 – 12.Qb4 Ne7
13.Bb6 Nc6 14.Qc5 Qe7 15.Qe3 d6 16.f4 f5
17.0-0-0 with excellent compensation, White has the
additional option of 12.Qg3 Ne7 (12...g6 13.Bd3
Ne5 14.Bd4°) 13.Qxg7 Nf5 14.Qg4 Nxe3 15.fxe3
Qb6 16.0-0-0 Qxe3+ 17.Kb1 Kf8 18.g3 with active
play in the centre.

12.Bb6 Nf5
18.Rc1!
18.Rd1 Be6 19.f4 Nf7 20.f5 Bxf5 21.Rxd5
Ne5 would turn the tabels in Black favour, because
his king will comfortably block the d6-pawn.
18...Nc6 19.g3 Be6 20.Bg2 Kd7 21.0-0.
White retains some pressure, but Black should hold
this.

b) 14.c5 d5! 15.cxd6 Qxd6 16.Qxd6 Nxd6

17.0-0-0 Ndf7, Zugic-Bluvshtein, Toronto 2004,
18.Be2 Bd7 19.Rhe1 poses more problems, although
the absence of queens reduces White’s attacking
potential. I believe that White should seek new
ways. He would have more chances with queens:

13.Qb4! c) 14.Qb3! d6 15.0-0-0 0-0

White should not give a tempo for ...d6 –

13.Qc5 d6 14.Qa5 Qd7 15.f4 Nc6 16.Qa3 Nce7
17.0-0-0 Qc6∞.

13...Nc6

165

The b6-bishop hinders Black’s counterplay
while we could safely press in the centre.
Undoubtedly, it is easier to conduct White here:
16.f4 Ng6 17.g3 Bd7 18.Bd3 Bc6 19.Rhe1
Qf7 20.Kb1 Rfe8 21.Bc2ƒ. Black has no threats so
we could carefully prepare our offensive with a3,
Ka1, and eventually g4.
14.Qc5 Qe7
The reason behind the queen’s pendulum
manoeuvre was to drag the black pawn on c5, where
it is bound to perish: 14...d6 15.Bxd8 dxc5 16.Bb6
Black can finish development with 16...Bd7,
but 17.0-0-0 Ne5 18.Ne4 or 17.Ne4 Ncd4 18.0-0-0
leave White with active pieces and the bishop pair.
Alternatively:
16...Ne5 (peventing 17.Ne4 in view of
17...Nd7) 17.Bxc5 Bd7 18.f4 Ng6 19.g3 Rd8
neutralises White’s immediate threats, but 20.b4! or
166
20.Rd1 underline the fact that White has a clear plan
to advance on the queenside. Finally:
16...e5 17.0-0-0 Ncd4 18.Bxc5 Bd7 19.Bd3
occurred in Giri-Dubov, Novi Sad 2016. Although
the Black knight on d4 looks intimidating, it is not
threatening anything from that square and it might
hang in some lines. For instance:
19...Bc6 20.Rhe1 Bxg2 21.f4 Rc8 22.b4².
Dubov chose instead:
19...Rc8, when the obvious 20.Bb6 would
have retained a pull – 20...Bc6 (20...Rc6 21.c5)
21.Rhe1 0-0 22.f3.
15.Qxe7+!
15.0-0-0 Qxc5 16.Bxc5 d6! 17.Bxd6 Nxd6
18.Rxd6 Bd7!, followed by ...0-0-0, is pretty
balanced, as proven by several correspondence
games.
White should aim to regain the pawn in more
favourable circumstances.
15...Nfxe7
This recapture prepares the freeing ...d7-d5.
15...Kxe7 offers White a tempo for g4 – 16.0-0-0 d6
(16...h5 17.c5) 17.g4 Nh6 18.f4. Now 18...Nxg4
introduces new imbalances which favour White’s
bishop pair and active rooks – 19.Rg1 f5 20.h3 Nf6
21.Rxg7+ Kf8 22.Rg3².
16.0-0-0
(or 16.Be2 d5 17.0-0-0)
16...d5 17.Be2
 A critical position for the 3...c5 line.
Computers may be able to defend it, but OTB
Black’s task is not trivial at all. His problem is that
in the rest of the game he’ll have to make many
important decisions based on intuition and not on
calculation since the forced play is over. See Game
29 Zhigalko-Vaibhav, Lille 2013.
B. 3...d5 4.e5
4.cxd5 exd5 5.e5 Ne4 is another topical line.
Capturing the pawn with 6.Nxe4?! has
proved to be dubious, so we should consider 6.Nf3
and 6.d4:
a) 6.d4 Bb4 7.Qa4+ Nc6 8.Bb5 0-0!
8...Bd7 9.Ne2 Nxc3 10.bxc3 Be7 11.Nf4!
(11.0-0 gives Black time to block the kingside with
11...a6 12.Bxc6 Bxc6 13.Qc2 0-0 14.f4 f5) 11...0-0
(Now 11...a6 is dubious owing to 12.Nxd5! 0-0
13.Be2 Nxe5 14.Qc2 Ng6 15.Nxe7+² with a bishop
pair.) 12.0-0! a6 13.Bxc6! Bxc6 14.Qc2 Qd7 15.a4
167
White’s space advantage promises him an
attack. The thematic break 15...f6 stumbles into
16.e6, and 15...Rfe8 16.Re1 Bg5 17.Nd3 keeps the
strong knight. In this position White should not push
f4, which would cage in the c1-bishop. Instead he
should attempt a rook lift on the kingside.
9.Ne2
9.Bxc6 a5! 10.Ne2 bxc6 11.0-0 Ba6 12.Rd1
should be balanced after 12...Qd7 or even 12...f6.
9...Nxd4!!
The point of Black’s 8th move! 10.Nxd4 c5
11.Nde2 d4 12.a3 dxc3 13.0-0 Nd2 and tactics work
fine for Black: 14.axb4
Or 14.bxc3 Nxf1 15.axb4 a6 16.Qc2 Bf5
17.Qxf5 Qd1 18.Bb2 Rfd8 19.Nd4 Ne3+ 20.Rxd1
Nxf5 21.Bc4! cxd4 22.cxd4 Rdc8 23.Bb3 a5
24.bxa5 Rxa5 25.h4!? Nxh4 26.Rd3=.
14...Nxf1 15.Kxf1 (15.bxc3 a6 16.Qc2 Bf5
as above) 15...a6 16.Nxc3 axb5 17.Qxa8 cxb4
18.Qa5=.
b) 6.Nf3 Bf5 7.d3 Nxc3 8.bxc3 c5 9.d4

This line was very hot in 2018, with
Nakamura and Aronian testing it with both colours,
while Carlsen, Grischuk, Khalifman, and others took
White only.
The most ambitious and tangled retort is
undoubtedly 9...c4, but I could not find anything for
White after:
9...Qa5!? 10.Bd2 Nc6
It transpires that the noncommittal 11.Be2
Be7 12.0-0 0-0 13.a4 Rac8 suits Black so White
should try:
11.c4! Qd8 12.Qb3 Be6 13.Qxb7 Rc8
Black is holding here:
14.Ng5
Or 14.Rc1 Be7 15.Be2 0-0 16.0-0 Qb6.
14...Nxd4 15.Nxe6 fxe6 16.Rb1 Be7 17.Bd3
Nc6!
17...0-0 18.0-0 Rb8 19.Qa6 Qd7 20.f4 is
pleasant for White, but Anand’s improvisation
17...Kf7?! is worse: 18.0-0 Nc6 19.Be2! (Perhaps
Black missed this move. Opening the e-filewith
19...Nxe5 would be disastrous after 20.Bf4+–)
168
19...Nd4 20.Bd3 Nc6 21.Rfd1! Rc7 22.Qb3 Nd4
23.Qb2 (23.Qc3!?²) 23...Rd7 24.cxd5 Rxd5,
Nepomniachtchi-Anand, Batumi 2018, when
strongest is 25.Be4 Rxe5 26.Qb7±.
18.f4 (18.cxd5 Rc7!) 18...Kf7!?. I’d be
curious to see what White players could cook up
here. It looks pretty balanced to me.
4...d4
4...Nfd7 cedes a lot of space – 5.cxd5!
5.d4 dxc4 (5...c5 6.cxd5 exd5 7.Nf3 Nc6
8.Bb5 a6 9.Bg5²) 6.Bxc4 c5∞.
5...Nxe5 6.d4 Ng6 7.dxe6 Bxe6 8.d5².
5.exf6 dxc3 6.bxc3
6.fxg7 cxd2+ often leads to equal endgames,
even after 7.Bxd2 Bxg7 8.Qb3. For instance: 8...Nd7
9.Rd1 b6 10.Bc3 (or 10.Be2 Bb7 11.Bf3 c6 12.Bc3
Bxc3+ 13.Qxc3 Qf6) 10...Bxc3+ 11.Qxc3 Qf6=,
Harutyunian-Oparin, Batumi 2018.
6...Qxf6
7.Nf3
The older main line 7.d4 has been
neutralised by 7...e5
7...b6 could face 8.Be2 (Aronian
experimented with 8.h4 Bb7∞) 8...Bb7 9.Bf3 Bxf3
10.Nxf3 Bd6 11.Qa4+ c6 12.0-0 with good practical
results for White.
 8.Nf3
8.Qe2 Be7! is considered safe for Black.
8...exd4 – see line B11.

7.Bd3 is a recent attempt to evade the set-up

with ...b6 (in view of Be4). However, 7...Bd7! 8.Nf3
Bc6 9.Be2 (9.Bc2 Bxf3) 9...Bd6 10.d4 b6 11.0-0 0-0
12.Bg5 Qg6 13.Qd2 h6 14.Bf4 Nd7= was level in
the blitz game Caruana-Anand, Saint Louis 2018.

White’s pressure on the queenside could

become tangible following 13...b6 14.a4.

7.g3 e5 8.Bg2 Nc6 9.Nf3 Bc5 10.0-0 0-0

11.d3 looks passive for White.

B1. 7...e5

B1. 7...e5; B2. 7...c5; B3. 7...b6!

We can choose here between two promising
7...Nc6 8.Bd3 Ne5 variations: B11. 8.d4 and B12. 8.Bd3!.
8...g5?! is a rubbish move, which should be
met calmly – 9.0-0 Bd6 10.g3 g4 11.Nh4 Qg7 B11. 8.d4
12.Be4 0-0 13.d4±, Houdini 3-Bouquet 1.6, 2013.
9.Be4 Nxf3+ 10.Bxf3
10.Qxf3 Qxf3 11.Bxf3 is almost the same, The main drawback of this move is that it
only White’s king is on e1. Kermer-Packroff, ICCF allows an exchange of the bishops:
2016, went 11...Rb8 12.a4 a5 13.d4 b6 14.Be3 f6 8...exd4! 9.Bg5 Qe6+ 10.Be2 Be7 11.cxd4
15.Bc6+ Kf7 16.Rb1 Ba6 17.c5 Bc4 18.cxb6 Rxb6 11.Bxe7 d3! gradually equalises since Black
19.Rxb6 cxb6 20.Bb5 Bxb5 21.axb5. White won should be able to tame White’s fleeting initiative.
after 21...Bd6 22.Kd2 Rc8 23.Rc1 Ba3 24.Rc2 Ke7 11...Bxg5 12.Nxg5 Qg6 13.f4 0-0 14.0-0
25.Kd3 Bd6 26.c4 Rd8 27.Re2 Rc8 28.f4 g6 29.Rc2
Rd8 30.Kc3 Rc8 31.Kb2 Bb4 32.Kb3 Kd6 33.c5+
1-0.
10...Qe5+ 11.Qe2 Qxe2+ 12.Kxe2 Rb8
13.d3

169
 In 2017 Harikrishna chose here 14...Nc6, and balanced position.
a year later he decided to include first 14...h6 15.Nf3 11...h6 12.h5
Nc6, with a satisfactory game in both events.
However, in the latter line White might try 15.Bh5
Qa6 16.Nxf7 Rxf7 17.Bxf7+ Kxf7 18.f5 Qf6 19.Rf4
Kg8 20.Qf3 Nc6 21.Re1 Bd7 22.Qd5+ Qf7 23.f6
gxf6 24.Rxf6 Qxd5 25.cxd5 Nb4 26.d6 Nd5
27.Rxh6 Kg7 28.Rh5 Nf6 29.Rc5². He does not risk
anything with his rook and 2 pawns vs a bishop and
knight.
Lately Black prefers to keep the tension, but
it is a risky approach:

8...Nc6?! 9.Qe2!?

This move should be a big surprise for your White had been scoring well here until the
opponents. Although it is not entirely new, White’s game Anand-Karjakin, London 2017 which
investigations were focused on other lines: introduced the novelty:
Dubov beat nicely Nakamura with 9.Be2 12...Qa6! and Black was fine after 13.Be3
exd4 10.Bg5 Qg6 11.0-0 Nd7 14.Bd3 Ba3=.
11.cxd4 Bb4+ 12.Bd2 Bxd2+ 13.Qxd2 0-0
14.d5 Ne7= (Cummings). 9...Bg4 10.d5 Bxf3 11.gxf3
11...dxc3! Looks risky, but it will probably
be analysed up to a draw very soon.
12.Bd3 Qd6 13.Re1+ Be7 14.c5!? Qxc5
15.Bxe7 Nxe7 16.Re5 Qd6 17.Bb5+ Kf8 18.Qe2 f6

11...Ne7

19.Rd1 fxe5 20.Nxe5! Bf5 21.Rxd6 cxd6
22.Nd7+ Kf7 23.Qc4+ Be6 24.Qf4+ Nf5 25.g4.
Now instead of 25...Rad8? Black could have played
25...Rhc8 26.gxf5 Bxd7 27.Bxd7 c2 28.Qc1! with a
draw, according to the computer.
9.Bg5 Qg6 10.d5 Nb8 11.h4
The sharp attempt 11.Bd3 f5 12.h4 e4 13.h5
Qd6 14.0-0 h6! 15.Bh4 Be7 concludes with a
170
11...Nb8 occurred only once, but might be
the better option. White still retains an initiative with
12.Rg1! – see Game 28 Petursson-Lagunow, Bern
1996.
12.Rb1 b6
12...0-0-0 looks dangerous, but it is not easy
to crack Black’s castling position:
13.Rg1 Kb8
Black cannot entrench himself with 13...Qd6
14.Be3 c5 owing to 15.Qb2 Rd7 16.Bh3 f5 17.Qb5
and he is tied up and down.
White’s task now should be to provoke ...b6
which could be attacked later by a4-a5.
14.a4. All the engines favour White,
although I doubt that we could break through. Note
that White does not rely only on an attack. His
“centralised” king assures him of an edge in most
endgames. For instance: 14...Ka8 15.a5 Rb8 16.Qe4
Qa6 17.f4ƒ f5 18.Qxe5 Ng6 19.Qxc7 Bd6 20.Qd7
Rhe8+ 21.Kd1 Bxf4 22.Bxf4 Nxf4
23.Qb5 Qxb5 24.Rxb5±.
13.h4 Ng6 14.Qe4 Bc5 15.Bg5 Qd6 16.h5
Ne7 17.f4 f6 18.fxe5 Qxe5 19.Qxe5 fxe5 20.Bh3
0-0
White has a superior centre and the more
active pieces. That allowed him to win the sharp
171
endgame after
21.h6 g6 22.Be6+ Kh8 23.Rb2, Almarza
Mato-Mason, ICCF 2017.
I would prefer to keep the pawn on h5:
21.Be6+ Kh8 22.Be3. We want to play a
rook endgame, where our active king should be the
prevailing factor thanks to the break c4-c5. For
instance: 22...Bxe3 23.fxe3 Nc8 24.Bxc8 Raxc8
25.Ke2 Kg8 26.Kd3 Rf2 27.Ke4 Rg2 (27...Rxa2
28.Ra1) 28.Rbg1² or 22...Bd6 23.Ke2 Nf5 24.Bxf5
Rxf5 25.a4 Kg8 26.Kd3 Re8 27.c5 bxc5 28.Rb7 Ra8
29.Kc4².
All that said, 9.Qe2 is very committal and
requires energetic play from White. Otherwise the
split pawns could remain a heavy burden.
B12. 8.Bd3!? Bd6
8...Nc6 9.0-0 Bg4 10.Be4 Bd6 is a bad
set-up in view of 11.Rb1. Black could save the pawn
with 11...Bxf3 12.Bxf3 0-0 since 13.Rxb7?! would
stumble into 13...e4, but 13.d3 Nd8 14.Re1² should
be good enough.
9.0-0 0-0
Black could discourage Be4 by 9...Na6,
intending 10.Bc2 Bg4 11.d4 Bxf3 12.Qxf3 Qxf3
13.gxf3
The threats c4-c5 and d4xe5 force Black to
take on d4, but that improves White’s pawn
structure and opens the e-file for a check:
 13...exd4 14.Ba4+!? albeit defendable for Black.
14.Re1+ Kd8 15.cxd4 c5 is not too clear as 10.0-0 Bd6 11.Be3 0-0
Black can exchange a pair of rooks from e8. The
bishop’s check prevents it in the event of 14...c6
15.Re1+ Kd8 16.cxd4 c5 17.d5±.
14...Kd8 15.cxd4 c5 16.d5 Kc7 17.f4!
The point of White’s strategy. He denies the
e5-square to the knight, discouraging the manoeuvre
...Na6-b4-d3-e5. Kuljasevic-Grigorov, Konya 2014,
went further:
17...Nb8 18.Rb1 b6, when the rook lift along
the third rank would have produced weaknesses:
19.Rb3! Nd7 20.Ra3 a5 21.Rg3 (Kosten).

10.Be4 Nc6 11.Qc2 h6 12.Rb1 Na5

White takes the initiative with natural

moves: 12.Qe2 Qe7 13.Rfe1 b6 14.Rad1 Bb7
15.d5ƒ.

9.Bg5 Qf5 10.cxd4 Bb4+ 11.Bd2

White’s pieces are clearly more active. He

can develop his initiative with 13.d4 exd4 14.c5
Bxc5 15.cxd4 Bd6 16.d5, for instance: 16...b6
17.Bb2 Qd8 18.Bh7+ Kh8 19.Bd3 Bb7 20.Qc3 Qf6
21.Qxf6 gxf6 22.Bxf6+.

Bareev-Rozentalis, Germany 2001, saw the 11...Qa5

materialistic 13.Qa4 Bf5 14.Bxf5 Qxf5 15.Rb5 b6
16.c5 Be7 17.cxb6 axb6 18.Rxe5 with an extra
pawn. 11...Nc6 12.Bxb4 Nxb4 13.Rb1 Qa5 14.Qd2
Nc6 15.Bd3 transposes.
B2. 7...c5 8.d4 cxd4 12.Bd3 Nc6 13.Rb1 Bxd2+ 14.Qxd2
Qxd2+ 15.Kxd2².
The restrictive stand 8...h6 9.Bd3 Nc6 also
has some following. White had the slightly better endgame in
9...cxd4 10.cxd4 Bb4+ 11.Bd2 Bxd2+ Wang Hao-Saduakassova, Hanoi 2018.
exchanges a piece, but 12.Qxd2 Nc6 13.Be4 0-0
14.0-0 Rd8 15.Bxc6 bxc6 16.Qe3 is rather dull

172
 B3. 7...b6! 8.d4
We could prevent new damages in our pawn
structure by 8.g3 Bb7 9.Bg2, but 9...Bd6 10.d4 h6
11.0-0 Nd7, followed by ...c5, is level.
8...Bb7 9.Bd3
9.Bg5 Bxf3 10.Qxf3 Qxf3 11.gxf3 Be7!? is
no better. You’ll understand that after reading the
main line. I found an interesting example of how
helpless are the engines in such purely strategic
positions. Black was clearly ill at ease in its shoes in
the following game:
11...Nd7 12.0-0-0 Bd6 13.Kc2 h6 14.Be3
Ke7 15.h4 c6 16.f4 Nf6 17.Bg2 Kd7 18.h5 Rhg8
19.Bf3 Rac8
Black adopted a wrong set-up on the
kingside (h6 instead of g6), and put its king on the
d-file to allow the stab:
20.c5! Be7 21.c4ƒ, Chiron 2-Sting SF 4.7,
CCRL 2014.
I also investigated:
9.Be2 Bd6 10.Bg5
White had not enough compensation
following 10.c5?! bxc5 11.Qa4+ Bc6 12.Bb5 Bxb5
13.Qxb5+ Nd7 14.0-0 cxd4 15.Bg5 Qg6,
Furman-Taimanov, Moscow 1955.
10...Qf5 11.0-0 h6 12.Bh4 0-0 13.a4 Nd7
14.a5 e5=.
9...Bxf3
173
9...h6!? might be a slightly improved version
of the main line because White’s king is no better on
g1 while ...h6 could enable counterplay with g5.
10.0-0
We have to play this anyway, now or after
10.a4!? a5. (10...Bd6 11.a5 Nd7 12.0-0 Bxf3
13.Qxf3 at least gives White an initiative on the
queenside. The a-pawn could be unpleasant on a6 in
some endings.) Note that after 11.Ne5 (instead of
11.0-0) 11...Bd6! 12.Qe2 Black can even take on g2,
although White has enough compensation.
10...Bxf3
After 10...Bd6 White can save the knight
with 11.Nd2 0-0 (11...Qh4 12.g3 Qd8 13.a4! 0-0
14.a5 Nd7 15.Ne4ƒ) 12.Qh5, retaining a flexible
pawn formation.
11.Qxf3 Qxf3 12.gxf3 Nd7
12...Nc6 allows White to push c5 – 13.Be4
Kd7 14.Rd1 Rd8 15.c5! and Black quickly lost in
Goganov-Shevchenko, Moscow 2016, following
15...bxc5? 16.dxc5+ Kc8 17.Be3 Rxd1+ 18.Rxd1.
The a7-pawn is weak. Of course Black should have
entrenched himself with 15...Ne7, but 16.c4ƒ
activates White’s central pawns.
13.a4 a5
I explain this pawn structure in Game 30
Williams-Norman, ICCF 2013. The annotations will
also be helpful for grasping the essence of the main
line with the immediate 9...Bxf3.
10.Qxf3
10.gxf3 Nd7 is rather risky for White , but it
is the best way to impose sharp, strategically
unbalanced play.

In practice Black does not fare well,

although engine analysis claims that he should be
fine. The most critical line looks:

a) 11.Be3!? Bd6 12.Rg1 h6 13.f4

Another complex line is 13.Ke2 Rd8
(13...Bxh2 14.Rg4 g5 15.Qa4 Rd8 16.Qxa7∞) 14.h4
Kf8 15.Qa4∞.

14...Kf8!
14...g5 15.h3 c5 16.Rd1 Bf4 17.Qc2 Ke7
18.Ke2 was pleasant for White in
Iordachescu-Gajewski, Legnica 2013.
15.0-0-0 Bxh2 16.Rg4 e5 17.Bc2 Re8
18.Ba4 c6 19.Re4 g5 20.dxe5 Nxe5 21.f4!=.
However, note that it is White who should make the
draw.

10...Qxf3 11.gxf3
13...Bxf4!
In Aleksandrov-Riazantsev, Loo 2014, Black
did not brace himself for this sacrifice and was
worse after 13...Rd8 14.Qg4 g6. White puts his rook
on d1, the king on f1, and enjoys a nice attacking
position. In some lines f4-f5 could be the needed
break.
14.Qf3 Bxe3 15.Qxa8+ Ke7 16.Qg2 Bg5
17.Ke2. This position may be equal for the engines,
but OTB White is more prone to mistakes.

b) 11.Be4 Rd8 12.Rg1

White’s pressure quickly comes to an end
after:
12.Qa4 e5 13.Be3 Be7 14.Qxa7 exd4
15.cxd4 Bd6 16.0-0-0 0-0;
12.h4 h6 13.Be3 Bd6 14.Qa4 (Or 14.Qd2 c5
15.Ke2 Qe7 16.Rag1 Nf6 17.h5 Qc7 18.Bd3 Kf8
19.a4 Rc8÷.) 14...0-0 15.Bg5 hxg5 16.hxg5 Qxg5
17.Bh7+=.
12.Be3 Bd6 13.Qd2 Qe7 14.Rg1÷.
12...h6 13.Be3 Bd6÷ 14.Qd2
In practice White scores over 60% in both
OTB and engine games. The computer is
enthusiastic about White’s chances and sometimes
its evaluation peaks at +1. However, it fails to
realise that White has no plan to break through.
He could cross the fourth rank only if Black chose a
wrong set-up. Let’s look at several examples:
11...Nd7 12.a4 a5 13.Rb1.

11...Bd6 12.a4?! Nc6 13.Be4 Kd7 and White

cannot advance further as a4-a5 is impossible.
174
 Evidently, he should hinder this set-up. That
is possible only with 12.Be4! (12.Be3 Nd7 13.a4 a5
14.Be4 Rd8 15.Rb1 0-0 16.f4 f5 was unassailable in
Dolin-Yartsev, ICCF 2013) 12...c6 13.f4! (ensuring
a retreat to f3)
Black is aiming for a fortress of the
following type: 13...Kd7 14.Be3?! Kc7 15.a4 a5!
16.h4 g6! 17.h5 Nd7 18.Ke2 Nf6 19.Bf3, no matter
what

19.d5! exd5 Here, instead of 20.Bxf5 (which

also brought Nakamura a full point), White had
20.Bd4! Rh6 21.Bg7 Re6+ 22.Be5±.

12.Kd1

Black’s idea is to cage in the c1-bishop as in

Remember this position! It is a dead draw.
However, White had 14.f5! Re8 15.Be3,
opening play in his favour. Black could meet 13.f4
by:
13...g6 14.a4 a5, but 15.Rb1 treads on the
sore point at b6.
These considerations led me to the
conclusion that Black should kill two birds with one
shot:
the game Stockfish-Komodo, TCEC 2014: 12.0-0
Nd7 13.a4 Bd6 14.a5 Kf7 15.Bc2 g5, and White
struggled to make a draw.
12.a4 Nc6 and 12.d5 g6! are also fine for
Black so White should try to put pressure on e6:
12...Kf7 13.Kc2

11...f5!

Preventing both Be4 and f4-f5.

The recent game Nakamura-Duda, rapid
Saint Louis 2019, saw another attempt to neutralise
White’s queenside plan:
11...Nc6 12.Be3
Perhaps 12.Rb1! was more precise. The idea
is to meet 12...Na5 by 12.Be3 as in the game. The
difference is 12...0-0-0 13.Ke2 – this move is more
useful than Be3 at this point. Of course 12...Bd6 lets
in 13.c5 bxc5 14.dxc5 Bxc5 15.Bb5. This position is equal. Black could even push
12...Na5 13.Rb1 Be7 14.Ke2 f5 15.c5 g5 ...f4 at some point, isolating the bishop. Plausible
16.f4 g4 17.h3 gxh3 18.Rxh3 h5 options are:
13...Nc6 14.h4 Bd6 15.h5 Ne7 16.Bg5 h6
17.Be3 c5=;
13...Nd7 14.a4 Bd6 15.Be3 a5=.
175
 Chapter 8. 1.c4 Nf6 2.Nc3 e6 3.e4 could break through with c4-c5, Kd3-c4 which is
Annotated Games often decisive. Accordingly, White should seek to
trade his light-squared bishop for the knight:
17.Bd3
28. Petursson – Lagunow 17.Be3!? g6 18.Rg5 is also possible, but
Bern 1996 instead of the passive 18...Bd6 19.h4ƒ, Black has
18...0-0-0 19.Rxe5 Bg7 20.Bh3+ Kb8 21.f4! Rhe8!
22.Rxe8 Bxc3+ 23.Ke2 Rxe8, reducing the material.
1.c4 Nf6 2.Nc3 e6 3.e4 d5 4.e5 d4 5.exf6 To be sure, White’s bishop pair is still stronger after
dxc3 6.bxc3 Qxf6 7.Nf3 e5 8.Qe2 Nc6 9.d4 Bg4 24.Rg1.
10.d5 Bxf3 11.gxf3 17...Bc5 18.Ke2 0-0 19.Rg2 Rf7 20.Be4 Bd6
21.f3 Nc5 22.Rag1. White will try to put pressure on
the h7-pawn with the help of a rook lift via the third
rank. Whenever Black takes on e4, the
above-mentioned break c4-c5 will be looming.

12...Nd7 13.Bh3 Bd6

This position looks scary as White has

14.Bxd7+ Kxd7 15.c5, but Black is holding with
only moves:
15...exf4 16.Kd1
16.Qg4+ Kd8 17.cxd6 Qxc3+ 18.Kf1 cxd6
19.Qg5+ f6 20.Qxf4 Qxa1 21.Qxd6+ Kc8 22.Qe6+
is a perpetual.
11...Nb8 12.f4 16...Be5 17.Rb1

Of course this is the most principled line, but

objectively White can retain a slight pull only with:
12.Rg1! Na6 13.Bg5 Qd6 14.f4 f6 15.fxe5
Qxe5 16.Qxe5+ fxe5

17...Kc8 (17...Rhd8 18.c4) 18.Qg4+ Kb8

19.Re1 Rd8 20.c4 with complex play.

14.0-0 0-0-0 15.c5 Bxc5 16.fxe5 Qh4

17.Kg2 Kb8

White’s pawns are all split, but his pieces are

more active. The most important factor is his king’s
closeness to the queenside. In a rook endgame he
176
 18.f4 24...Qh3? (24...Rdg8) 25.Rad1?

Despite White’s enormous space advantage
and bishop pair, the game remains balanced after
18...Nb6! 19.c4 f6 20.a4 g5 21.Be6 gxf4 22.a5 f3+
23.Rxf3 Rhg8+ 24.Rg3 Rxg3+ 25.hxg3 Qd4
26.axb6 Qxa1 27.bxc7+ Kxc7 28.Bf4 Qg1+
29.Kh3=.
Black’s attempt to undermine the centre at
once could have passed the initiative to White:
18...g5 19.fxg5
White returns the favour. 25.Rf3 was equal.
25...Rdg8 26.d6 Rg2?
(26...cxd6 27.exd6 Ka8µ)
27.dxc7+ Kc8 28.Qe4 R2g6 29.Rd8+ Rxd8
30.cxd8=Q+ Kxd8 31.Rb1 Qd7 32.Qf3 Ke8
33.Bg3 Bb6 34.Rd1 ½-½

29. A.Zhigalko – Vaibhav

It would have been stronger (and safer!) to
exchange the queens – 19.Qg4 Rhg8 20.Qxh4
gxh4+ 21.Kf3. White’s protected passer on e6
would generate nasty threats – 21...Nb6 22.e6 f6
23.Bf5 h5 24.Rd1 Rxd5 25.Rxd5 Nxd5 26.Be4 c6
27.Bd2².
19...h6 20.g6 Rhg8 21.Bxd7?
Lille 2013
1.c4 e6 2.Nc3 Nf6 3.e4 c5 4.e5 Ng8 5.Nf3
Nc6 6.d4 cxd4 7.Nxd4 Nxe5 8.Ndb5 a6 9.Nd6+
Bxd6 10.Qxd6 f6 11.Be3 Ne7 12.Bb6 Nf5 13.Qb4
Nc6 14.Qc5 Qe7 15.Qxe7+ Nfxe7 16.0-0-0 d5
17.Be2

21.Bf4 Rxg6+ 22.Bg3 Rdg8 23.Rf3 Qe7

24.Re1 h5 25.e6 is level. White’s incomprehensible
move (was he afraid of ...Nb6?) turns the tables.

21...Rxg6+ 22.Kh1 Rxd7

(22...Rdg8! 23.Be3 Rg2µ)

23.c4?! Rd8 24.Bf4

177

17...Kf7
Black hopes to buy himself off with a pawn,
but we should not oblige.
17...dxc4 18.Bxc4 Bd7 19.Ne4 Nc8 20.Be3
occurred in Zajsek-Peczkowski, ICCF 2015. The
game followed a completely logical course until its
last move:
20...Na5 21.Bd3 Bc6 22.b4 Bxe4 23.Bxe4
Nc6 24.Bc5 N8e7 25.Rd6 Rd8 26.Rhd1 Rxd6
27.Rxd6 e5 28.a4 f5 29.Bc2
29...h5
29...e4 30.b5 axb5 31.axb5 Nc8 32.bxc6
might be insufficient for a full point, but 32.Rd1
Ne5 33.Rd5 Nd7 34.Rxf5± is good enough. After
the text White suddenly played 30.Re6 and the
opponents signed a draw!? After the obvious:
30.b5! axb5 31.axb5 Nd8 32.b6 Rh6
33.Ba4+ Nec6 34.Rd5 Black would be tied up and
down – 34...Kf7 35.Rd7+±.
178
More consistent is 17...Bd7!?, trying to trick
White with the move order, e.g.:
18.Rhe1 dxc4 19.Bxc4 Ne5 20.Bb3 Rc8.
Thus we have no choice, but “concede” to regain the
pawn:
18.cxd5 exd5 19.Nxd5 Nxd5 20.Rxd5 Be6
21.Rd6 Ke7! 22.Bc5 Kf7. The worst is behind
Black, but he’ll still have to work hard for the draw
because of the bishop pair.
18.Rhe1
Another natural move is 18.f4!? to deny
...Ne5.
Mikhalevski-Vaibhav, Moscow 2012, went
18...Bd7 19.g4 Rhe8 20.Bf3 Rac8=. It is better to
activate the rook and see Black’s set-up. For
example: 19.Rhe1 Rae8 20.Bh5+ (to forbid ...Ng6)
20...g6 21.Bf3. White keeps the tension, and he can
regain the pawn whenever he wants.
18...d4!? takes the bull by the horns. We
have to simplify:
19.Bxd4 Nxd4 20.Rxd4 e5! 21.fxe5 Nc6
22.Re4 Nxe5 23.Rf1 Ng6
Here White played 24.h4 h5 25.c5 and Black
held the draw in Cannon-Cardoso, ICCF 2016. It
turned out that the extended white pawns on the
kingside were easy to exchange. Perhaps White
should try 24.c5² at once.
18...Bd7 19.Ne4
Houdini proposes here 19.b3! (to have Kb2
in some lines) 19...Rhe8 20.cxd5 Nxd5 21.Nxd5²,
and we should also check the idea from the previous to take committal steps in clearly superior positions.
paragraph: Instead they secretly hope that the opponent will
19...d4 20.f4 e5 21.fxe5 fxe5 22.Ne4 h5 lose by himself. All too often this tactic costs them
23.h3. White retains pressure. their advantage, as was the case in the current game.

19...dxe4 20.Rxd7 Rab8 21.Bc7 Rbc8 24...Kf6 25.Kb2?

22.Red1

This is already too much. 25.f3² was

White has excellent compensation, but Black indispensable. Now Black obtains counterplay on
should draw with good defence: the kingside and all three results become possible.
22...Nb8! 23.Bxb8 Rxb8. White has now
several ways to develop his initiative: 25...Ng6 26.Kc3 Nge5 27.R7d2 Rbc8 28.b4
24.c5 a5 25.c6 Rhc8! (25...bxc6 26.R1d6±) g5 29.a4 Ne7 30.Kb3 f4?
26.Bb5 bxc6 27.Bc4 Rb4 28.b3 Rcb8 29.R1d6 Ke8
30.Bxe6 R4b7 31.Kd2 f5 32.Ke3 g6=;
24.R1d6 a5 25.Rb6 f5 26.Rbxb7 Rxb7 Black underestimates the weakness of the
27.Rxb7 Kf6 with a tangled ending; e4-pawn. 30...h5, 30...g4 or 30...N7g6 were messy.
24.f3!? e3 25.R1d3 Rhc8.
Vaibhav’s next move offers White a precious 31.Rd4 Kf5 32.b5
tempo and dooms him to a gloom defence.

22...f5?! 23.Bb6! White had not spotted yet the original mating
net, which eventually decided the game. 32.Re1!
N7c6
23.f3 exf3 24.Bxf3 is also possible, although
the position after 24...Kf6 25.c5 Nb8 26.Bxb7 Nxd7
27.Rxd7² is double-edged.

23...Rb8 24.b3!?

This move, besides its chess value, has a

strong psychological impact. Zhigalko underlines
the fact that Black cannot disentangle his pieces.
White could have also undermined the enemy centre
with 24.f3 Ne5 25.Rc7, but chess players often hate

179
 33.Rxe4!! Kxe4 34.Bg4+ Kd3 35.Rd1+ Ke4
36.f3+ Nxf3 37.gxf3+ Ke5 38.Re1+ Kf6 39.Rxe6+
Kf7 40.Re4 was close to winning!
32...axb5 33.axb5 h5
(33...Ng8! kept Black in the game)
34.Re1! Ng8 35.Rxe4
This is even more destructive now as
35...Kxe4 36.Bg4+ Kd3 37.Rd1+ Ke4 38.Rd4
mates!
35...Nf6 36.Rxe5+ Kxe5 37.Bg4+ Kd6
38.Rxe6+ Kd7 39.Bf5 1-0
30. Williams – Norman
ICCF 2013
This game reveals some strategic traps which
await Black in the position on the next diagram. It
shows that even computer help cannot save Black if
he chose a wrong set-up.
1.c4 Nf6 2.Nc3 e6 3.e4 d5 4.e5 d4 5.exf6
dxc3 6.bxc3 Qxf6 7.Nf3 b6 8.d4 Bb7 9.Bd3 h6
10.0-0 Bxf3 11.Qxf3 Qxf3 12.gxf3 Nd7 13.a4 a5
180
14.f4?!
This is mostly a defensive move. It does
takes control of the centre, but loses the initiative. At
the same time White should rely on dynamic factors.
His only way to break the enemy front is c4-c5. It
would allow White’s rook to penetrate to b7.
Therefore, he should play concrete chess:
14.Rb1!
Now 14...c5 would face 15.d5 while 14...g5
15.f4 would help us to open files.
14...f5?! allowed White to execute his main
threat – 15.c5!
15...bxc5 16.Bf4 cxd4 17.Rfe1 Kf7 18.Bc4
Re8 19.Rb7 Ba3 20.cxd4 with strong pressure in
Smerda-Mudra, ICCF 2012.
14...Bd6 15.Be4 Rd8 (15...0-0-0 16.c5²)
16.Rb5 g5 17.Bc6 Ke7 18.Be3 Bf4 19.Kh1. White’s
rooks are more active, e.g. 19...Kd6 20.Bxd7 Kxd7
21.Bxf4 gxf4 22.Rg1 Rhg8 23.Rg4 Rg6 24.c5.
 Note that 14.Rb1 also prevents 0-0-0, which Rb8
would be good after 14.Be4 0-0-0 15.f4 g6=.

14...Bd6 White was patiently improving his pieces

while his opponent clearly had no idea what he was
doing. The punishment is close.
14...f5! would cut across White’s plan and
would leave the c1-bishop caged in.

15.Re1 0-0

Black follows the engines’ advice, but

15...g6 16.Be4 0-0-0 17.Rb1 c5= would be safer.
Short castling deprives Black of counterplay on the
kingside. Besides, the Black king would be more
useful on c7.

16.Be4 Rab8 17.Kg2

27.c5!± Rdd8 28.Bc6 Rbc8 29.c4 Nd7

30.Rb1 Bf8 31.Bb7 Rb8 32.Be4 Rbc8 33.cxb6
Nxb6

The line 33...cxb6 34.c5! bxc5 35.Bb7 Rc7

36.Bd2+– reveals the full power of the bishop pair.

34.c5 Nc4 35.Rc1 Nxe3+ 36.fxe3 Rb8

37.Rcb1 Rxb5 38.Rxb5 g5 39.f5 exf5 40.Bxf5 h5

White’s positional dominance is so big that

he does not even need the a-pawn to win. For
17...Nf6 example: 40...Ra8 41.Bd7+–, intending Bc6 and the
central pawns decide the game.

Although this is the first line of Houdini, it is 41.Rxa5 Kf6 42.Be4 Re8 43.Ra6+ Ke7
a step in a wrong direction. A good long-term 44.Bc6 Rc8 45.a5 Kd8 46.Bd5 Rb8 47.Bxf7 1-0
solution would be to prevent the option of c4-c5
with 17...c5.

18.Bf3 g6 19.Rb1 Kg7

Black stubbornly refrains from the “ugly”

...c7-c5 which betrays bad understanding.

20.Be3 Be7 21.Rb5 Rfd8 22.Reb1 Rbc8

23.Bb7 Rb8 24.Bc6 Rbc8 25.R1b3 Rd6 26.Bb7

181
 Chapter 9. Anti-Slav and Chebanenko
Main Ideas

1.c4 c6 2.Nf3 d5 3.e3 Nf6 4.Nc3

7.Bb5+!? Nc6 8.Bxc6+ bxc6 9.0-0

What are the main pluses of my proposition?

It is no secret that many players begin with

1.c4 mainly to avoid the most studied lines of the
Slav and the Meran. I would not claim that White
could rip substantial benefits from delaying d4, but
at least he should be able to drag the opponent into
relatively uncharted waters. Accordingly, the focus
of my research was to find decent, positionally
sound ideas, where Black would not find ready
receipts in the books and databases.
Let’s start with the classical Slav retort:

A. 4...Bg4
White
Cummings recommends here 5.Qb3 Qb6
has created a permanent flaw in the enemy pawn
6.Ne5 Be6 7.d4, hoping to gain two bishops with
structure. Black’s c-pawn will remain a juicy target
Na4-c5. I tried it in practice (see Game 31 Kiril
for many moves ahead, binding his pieces with its
Georgiev-Terbe, Arad 2018), and it is a fair plan,
defence.
but it is nearly impossible to offer you any clear plan
That assures White of an edge if he managed
to follow. The kingside pawns stand symmetrically
to trade queens. In such event the play would be for
on their initial squares, and both players could
two results only.
arrange them in many different configurations. A
Our heavy pieces would be more active – the
long complex fight in a queenless middlegame is
rooks will double on the c-file and the queen is close
ahead. Besides, I came upon on a more
to the black king.
straightforward idea:
I have found only 19 games in my database.
5.cxd5!?
Black is obviously ill at ease in this position
as he achieves just 16%.
Now 5...cxd5 6.Qb3 Bc8 is totally passive
You’ll see the concrete moves in the “Step
and brings Black the horrifying 17%, 5...Nxd5 gives
by Step” section. Here I’ll add some positional
up the centre, so he commonly parts with his bishop:
guidance.
5...Bxf3 6.Qxf3 cxd5
Analysis
182
 Our ultimate strategic goal is to block the
c-pawn on c6. White could then claim a clear edge.
Black will probably anticipate the above
scenario by pushing ...c5. Then the attack on c5 with
Na4, Ba3 is ineffective if Black’s queen is on the
queenside. However, we can try to exploit its
absence from the kingside by mounting a kingside
attack or striking at the centre with e3-e4. Note that
e4 is effective only when Black cannot answer ...d4.
Bishoff-Bagirov
Netherlands 1998
19.e4± is strong as 19...d4 drops the pawn.
The opposite example is:
Levin-Van Foreest
Netherlands 2015
183
16.Rc2=. Our queenside pieces have no
prospects and we lack space.
The computer might be seducing you with an
early d4, claiming some advantage, but then our
dark-squared bishop would become completely
useless. We should not be able to win a game with
such a bad piece!
The second part of the chapter deals with the
Chebanenko-like:
B. 4...a6!?
This set-up has always been a tough nut to
crack. I hesitate which move order to recommend
against it. It is possible to start with 5.d4, having in
mind to meet 5...Bf5 by 6.Be2!. However, 5...b5
6.b3 Bg4 7.Be2 e6 8.h3 allows 8...Bh5. Finally I
decided to stake on the surprise effect, so my
proposition is:
5.Be2!?
 The only drawback of this move order is that
it allows a transition to the Queen’s Gambit
Accepted, but White has been scoring extremely
well lately after:
5...dxc4 6.Bxc4 b5 7.Be2! e6 8.0-0 Bb7 9.d4
c5 10.dxc5
White’s plan is to trade queens and attack the
queenside with Nd2-b3, a4. His dream position is:
Fressinet-Karjakin
blitz, Beijing 2012
17...g4 18.Be3 gxf3 19.gxf3 Rg8+ 20.Kf1 e5
21.Rac1 Bd8 22.Nac5±.
Black can prevent the binding a4-a5 by
taking on a4, but that does not save him from
trouble:
Kramnik-Tomashevsky
Moscow 2012
184
White has two targets – on a6 and e6. The
game went 28....Nb4
(28...Rc6 29.Nxe6 Kxe6 30.Raxd5)
29.Rd6 Rc8 30.f4 Rac7 31.Nxe6 Rc2+
32.Kf3 Rxb2, when best was 33.f5±. Kramnik won
anyway.
The advantages of delaying d4 are best seen
in the line 5...b5 6.b3 Bg4 7.h3 Bh5 8.cxd5 cxd5
9.g4 Bg6 10.g5 Ne4 11.Bb2 Nxc3 12.Bxc3 Nc6
Imagine that no one has reached this position
in practice! White has rich possibilities after 13.a4!
b4 14.Bb2.
Theoretical status
I have not seen a satisfactory receipt for
White against the Chebanenko. For instance,
Cummings recommends 4...a6 5.Qc2 e6 6.b3 c5
7.Bb2, omitting the most logical retort 7...Nc6
8.cxd5 exd5∞, when Black even has an initiative.
Delchev analyses 7.a3 Nc6 8.Bd3, but a
correspondence game suggests that Black is fine
after 8...Be7.
K. Hansen considers 5.b3, but it deprives
White of several active possibilities, based on Qb3
or Qa4. Black could answer 5...Bg4 6.Be2 e6 7.h3,
when both 7...Bh5 and 7...Bxf3 equalise.

Thus my analysis of 5.Be2 contributes to the

theory of the Chebanenko and gives White players a
new weapon against this sturdy system.

185
 Chapter 9. Anti-Slav and Chebanenko
Step by Step

1.c4 c6 2.Nf3 d5 3.e3

Now White prevented ...b5 and obtained an

advantage with 10.a4! e6 11.e4 Be7 12.0-0 0-0
13.g3 Nb6 14.Bf4 Qd8 15.Ne5!,
Sodomski-Hietanen, ICCF 2015.

4.Nc3
3...Nf6

A. 4...Bg4 (in the Slav style); B. 4...a6

3...e6 4.b3 will most likely transpose to the (Chebanenko style)
Anti-Meran after 4...Nf6 5.Nc3 Bd6. An additional
options for White is 5.Bb2 Bd6 6.d4 0-0 7.Bd3 b6 4...e6 5.b3 is the subject of the next chapter.
8.0-0 Bb7 9.Nbd2, intending to meet 9...Nbd7 by
10.Ne5 and 9...c5 by 10.cxd5 exd5 11.dxc5. 4...Bf5 5.cxd5 cxd5 6.Qb3 Bc8 is covered
via the move order 4...Bg4.
3...Bf5 is simply met by 4.Qb3
Or 4.cxd5 cxd5 5.Qb3 Qc7 6.Nc3 e6 7.Nb5
Qb6 8.Qa4 Nc6 9.Nbd4 Nge7 10.b3². A. 4...Bg4 5.cxd5
4...Qc7
4...Qb6 5.Qxb6 axb6 6.cxd5 cxd5 7.Nc3 e6
8.Na4 Nd7 9.b3 Ngf6 10.Bb2 Be7 11.Nh4². The only sensible alternative is 5.Qb3 Qb6
5.Nc3 e6 6.Nh4². 6.Ne5 Be6 7.d4 – see Game 31 Kiril
Georgiev-Terbe, Arad 2018.
3...Bg4 4.Qb3 is good again. 4...Qc7 5.Ne5
Be6 6.d4 Nd7 7.Nc3 dxc4 8.Bxc4 Bxc4 9.Nxc4 5...Bxf3
Ngf6

5...cxd5 6.Qb3 Bc8 is passive, but difficult to

crack.
For instance, 7.Ne5 Nc6 8.d4 e6 9.f4 Be7
10.Bd3 0-0 11.0-0 Nd7 12.Bd2 Ndxe5 13.fxe5 Bd7
14.Ne2 f6 is roughly equal. We could try instead:
7.d4 e6 8.Bd3 Nc6 9.0-0 Be7
After 9...Bd6 the only plausible plan for
White is to open the centre with e4: 10.e4 dxe4
11.Nxe4 Be7 (this is the flipside of ...Bd6) 12.Be3

186
0-0 13.Rac1. This set-up in IQP positions, with Be3 12.Ne5 Bxe2 13.Qxe2 0-0 14.Rac1 Nb6
and Rac1 instead of Bg5+Rad1, was recommended 15.Nd3 Rfe8 16.Rfd1±.
by Akiba Rubinstein.
6.Qxf3 cxd5 7.Bb5+

Creating a target. 7.g4 e6 8.g5 only spends

two tempi without apparent benefits – 8...Nfd7 9.h4
Nc6 10.d4 h6 11.g6 f5÷.
7.d4 e6 8.Bd3 Nc6 9.Bd2 Be7 10.0-0 0-0
11.Qe2 Rc8 12.Rac1 Bd6 13.f4 a6 14.g4 looks
active only at the surface. In fact Black does not
have bad pieces, nor weaknesses.

7...Nc6 8.Bxc6+ bxc6 9.0-0

10.Bd2 Nobody has played 9...e5. We should meet it

Grischuk and Ibragimov opted for 10.e4, but by 10.d3 Be7 11.e4 Qd6 (11...0-0 12.Bg5) 12.Bg5
Black did not threaten ...e5, so we could keep the Rd8 13.b3 with the better pawn structure.
centre undefined.
10...0-0 11.Rac1 Bd7 (11...Na5 12.Qd1) 9...e6 10.b3
12.Ne5 Nxe5 13.dxe5 Ng4 14.f4. White is better as
sooner or later Black’s knight should land on f5,
when Bxf5 will offer us a dominating knight.
Erdos-Hajnal, Hungary 2017, went 14...Bc5 15.Nd1
Bb6 16.h3 Nh6 17.Bb4 Re8 18.Bc5² f6 19.Bxb6
axb6 20.exf6 Qxf6 21.Qxb6 Bc6 22.a3±.

5...Nxd5 surrenders the centre. Any

reasonable plan based on e4 should offer White the
better chances. Even the most restrained approach
gave White a nice edge in Lisitsin-Capablanca,
Moscow 1935:
6.Be2 e6 7.d4 Nd7 8.0-0 Qc7 9.Bd2 Bd6
10.Ne4 N7f6 11.Nxd6+ Qxd6

10...Be7

10...Bd6 weakens the f6-square and binds the

black queen with the defence of the knight:
11.Bb2 0-0 12.Rac1 Rc8
Van Foreest decided to spend a tempo on
12...e5, when his opponent Levin should have
refrained from 13.d3 and e4 in favour of 13.Rc2 Qa5
14.Rfc1 Rac8 15.Qe2 Rc7 16.Na4 Rfc8 17.Nc5².
13.Na4 Qe7, Peralta-Palacios Perez, Madrid
2002. Here 14.Bxf6! Qxf6 15.Qxf6 gxf6 16.Nc5
accomplishes White’s positional goal.

187
 11.Bb2 0-0 12.Rac1 Qa5 13.Rc2 c5
If Black leaves the pawn on c6, we could try
to block it by a piece – 13...Nd7 14.Rfc1 Rac8
15.Na4 Qa6 16.Qd1 Rfe8 17.Qf1 Qb7 18.h3 Bf8
(18...c5?! 19.Ba3)
19.Bd4!?
Tomashevsky-Ansell, Kallithea 2008, saw
19.d3 g6 20.Qe2 e5 21.e4, which is a wrong plan in
view of 21...d4!=.
19...Ba3 20.Re1 e5 (20...c5 21.Bb2 –
exchanges are in White’s favour.) 21.Nc5!.
14.Rfc1
Ivanchuk-Vitiugov, Jurmala 2016, saw 14.d3
Qa6 15.Qe2 Nd7 16.Na4 Rfc8 17.Rfc1 Bf8 18.h4 h6
19.e4 d4!÷. As ugly as Black’s queenside pawns
seem, they effectively fend off our pieces. The
moral is:
We should not play e4 if Black can answer
...d4.
14...Nd7
14...Qa6 does not change anything since we
can always shift the queen with 15.Qe2 (or even
Qf3-d1-f1) 15...Qa5 16.Na4 c4 17.Bc3 Qa6 18.Nb2
Qxa2 19.bxc4 Rfc8 20.Bd4 dxc4 21.Nxc4².
188
15.Ne2!?
My favourite treatment of this pawn
structure is to transfer the knight to the kingside
instead of a4. It may not be a game changer, but
poses concrete problems.
15...Qa6 16.Nf4 Rfc8
16...Rad8 17.Nh5 g6 allows us to trade a
piece with 18.Nf6+².
17.Qg4 Bf6 18.Bxf6 Nxf6 19.Qe2².
B. 4...a6 5.Be2!?
I’m not too keen on entering the Mainline
Chebanenko, since Black has learned rather well
how to make draws after, for instance, 5.d4 Bf5
6.Nh4 Bc8! (6...Bg4 7.Qc2 e6 8.h3 Bh5 9.g4 Nfd7
10.Ng2) or 6.Qb3 Ra7 7.Nh4 Bc8 8.Bd2 b5 9.c5
e5=. Of course, White could play:
5.d4 Bf5 6.Be2!, which could transpose to
my mainline. The only independent variation is
6...e6 7.Nh4 Be4 (since we lack d3), but it should be
slightly better for White.
5.Qc2 throws the opponent out of the
Chebanenko, as:
5...Bg4 6.Ne5 Bh5 gives White a nice choice
between 7.Qb3 Nbd7 8.d4 and 7.g4!? Bxg4 8.Nxg4
Nxg4 9.Qb3². However:
5...e6, intending ...c5, underlines the clumsy
stand of White’s queen on the c-file:

The point of Black’s plan is the pawn sac

14.Bf3 0-0! 15.Nxd5 Nxd5 16.Bxd5 Bxd5 17.Rxd5
Rac8=. Best results yields:
14.Na4, but 14...Bd6 15.Nb6 Bxh2+
a) 6.b3 c5 7.a3 (7.Bb2 Nc6 8.cxd5 exd5∞) 16.Kxh2 Qd6+ 17.Kg1 Qxb6 18.Qa4+ Bd7 19.Qf4
7...Nc6 8.Bd3. This was recommended by Delchev, 0-0 20.b3 Bc6 21.Bb2 Ne4 22.Rac1 f6 is nothing
who considers only Black’s fianchetto on g7. The special for White, who still has to prove full
game Sandström-Stengelin, ICCF 2014, showed that compensation for the pawn.
8...Be7 9.Bb2 0-0 10.Ne2 b6 11.cxd5 exd5 12.Ng3
Re8 13.Rc1 h6 is balanced. The position remains 5.b3 is not too challenging. Black can answer
very complex though. 5...Bg4 6.Be2 e6 7.h3, when both 7...Bh5 and
7...Bxf3 equalise.
b) 6.d4 c5! 7.cxd5
Or 7.a3 dxc4 8.Bxc4 b5 9.Be2 cxd4=; My proposition is a relatively rare option, so
7.b3 Nc6 8.dxc5 Bxc5 9.Bb2 0-0 10.Rd1 we could count on the effect of surprise. Main
Qe7! 11.Be2 Rd8 12.0-0 Bd7 13.cxd5 Nb4 14.Qc1 retorts are:
Nbxd5=, Scherer-Spreemann,
www.remoteschach.de 2013.
7...exd5 8.Be2 Nc6 9.0-0 Be6
9...cxd4 10.Nxd4 Nxd4 11.exd4 Be7 12.Qb3
b5 should also equalise.
10.Rd1
10.dxc5 Bxc5 11.Rd1 Qe7 12.a3 Rc8 13.b4
Ba7=.
10...cxd4 11.Nxd4 Nxd4 12.Rxd4 Bc5
13.Rd1 Qe7

B1. 5...Bg4; B2. 5...b5; B3. 5...dxc4.

The thematic 5...Bf5 is hardly the best

answer in view of 6.0-0 e6 (6...h6 is often played
against 5.d4, but here it is dubious in view of 7.Qb3
Ra7 8.d3!) 7.Nh4 Bg6. White should not hurry to
189
capture the bishop, although 8.Nxg6 hxg6 9.d4 is in
his favour. Instead 8.d4 leaves all his options open,
e.g. 8...Nbd7 9.g3 Be7 10.Nxg6 hxg6 11.b3 0-0-
12.Bb2 b5 13.Qc2, intending to meet 13...Rc8 by
14.c5.

5...g6 6.0-0 Bg7 7.d4 0-0 is the Kamsky’s

pet line in the Schlechter Variation. White does not
need much theory to obtain good results after 8.a4 or
8.Bd2.

5...e6 6.0-0 b5
6...dxc4 7.Bxc4 transposes to line B3, while
White gets also the additional option of 7.a4 c5
8.Bxc4 Nc6 9.Qe2 Be7 10.Rd1 0-0 11.d4.
7.b3 Bd6 8.d4 Nbd7 9.Qc2
Now 9...Nc6 loses a pawn to 10.Ne5 Bxe2
11.Nxc6, and 9...Nbd7 10.Qd4! Nb6 11.Na4 is not
pretty either.
In practice White chooses 6.Qb3 Ra7 7.h3
Bxf3 8.Bxf3, but I believe that our queen does not
belong to b3 unless it gains concrete benefits. The
game Critter 1.6a-Houdini 5.01 CCRL 2016, went
8...e6 9.d4 b5 10.c5 a5 11.0-0 Be7 12.Rd1 0-0
13.Qc2 Qc7 14.a3 Nbd7 15.e4 dxe4 16.Nxe4 a4
17.Nd6 Ne8 18.Nxe8 Rxe8 19.Qe4 Rc8 20.Bf4 Qd8
21.Qe2 Nf8 22.Rd3 Rd7 23.Rad1 Ng6 24.Bg3=.

6...Bxf3 7.Bxf3

9...b4
9...0-0 10.e4 dxe4 11.Nxe4 Nxe4 12.Qxe4
Bb7 13.c5 Be7 14.Bd3 Nf6 15.Qh4 Re8 16.Bg5 h6
17.Bxf6 Bxf6 18.Qe4 Qd5 19.Qh7+ Kf8 20.Rad1
Ke7 21.Bc2 Rh8 22.Qd3 Rad8 23.Rfe1 Qd7 24.Ne5
Qc7 25.Qg3 Qa5 26.Nxf7! was crushing in
Riazantsev-Korobov, Dubai 2014.
10.Na4 Bb7 11.a3 bxa3 12.Bxa3 Bxa3
13.Rxa3 Qe7 14.Rfa1 0-0, Delchev-Golubovic,
Kastel Stari 1997. White has a strategically winning
position, and 15.c5 would have been the simplest
solution.
7...e6
B1. 5...Bg4 6.h3!?
Amazingly, this position has never occurred
in practice. The tactical justification of my novelty is
The most challenging timing for this move the pawn sacrifice 7...dxc4 8.b3! cxb3 9.Qxb3 Ra7
since the retreat is dubious – 6...Bh5 7.Qb3 Ra7 10.0-0 e6 11.d4 Be7 12.a4 0-0
allows 8.cxd5 cxd5 9.Qa4+!

190
 Pardubice 2015.

13.a5
The immediate 13.e4 allows 13...c5 and the 10.cxd5 cxd5 11.e4! dxe4 12.Nxe4 Nd5
b8-knight receives the c6-square. 13.Rd1 is another 13.Qb3 Qb6 14.Qa4²
interesting option.
13...Nbd7
The knight has to go to a passive place since White has the more active pieces and the
13...c5 loses to 14.Qb6. bishop pair.
14.e4 with more than enough compensation.

8.0-0 B2. 5...b5 6.b3 Bg4 7.h3

In such positions the only constructive plan
for White, in my opinion, is to push e4.
The alternative is 8.b3 Bd6
8...d4?! broke the symmetry in White’s
favour – 9.exd4 Qxd4 10.0-0 Be7 11.Bb2 0-0
12.Qc2 Qf4 13.Ne2 Qc7 14.d4 Nbd7 15.Rad1ƒ,
Sambuev-S.Kasparov, Montreal 2009.
9.Bb2 0-0 10.d4 Nbd7 11.0-0 Re8. White is
considerably better according to the engines, but in
fact he lacks a plan.
8...Nbd7 9.d4
Again, it is good to define the future of the
bishop right away.
7.Bb2 e6 8.Ne5 Bxe2 9.Nxe2 Bd6 10.Rc1
0-0 11.0-0 looks a decent alternative, only after
11...Nfd7 we should retain more pieces – 12.Nf3!.
7...Bh5
7...Bxf3 8.Bxf3 e6 is solid, of course.
But not 8...e5 9.Bb2 Ra7 10.cxd5 cxd5 11.g4
h6 12.Rc1 Be7 13.0-0 0-0 14.Bg2 with expansion on
the kingside.
9.d4 Be7 10.0-0 0-0 11.Qc2 Nbd7

9.cxd5 cxd5 10.d4 is also possible, but I

want to sidestep 10...Bb4, which would forbid
e3-e4.

9...Be7

9...dxc4 10.Qe2 Bd6 11.Qxc4 0-0 12.Rd1

comes accross White’s plan, Gagare-Kassani,

191
 Now in Gulko-Short, Candidates playoff,
New York 1994, White opened a file on the
queenside with 12.a4!? Rc8 13.axb5 axb5 14.c5! e5
15.Bd2 Ra8 16.Na2 Qc7 17.Nb4. White should gain
control of the a-file. His bishop pair and pressure on
c6 promise him a lasting initiative.
8.cxd5 cxd5 9.g4 Bg6 10.g5 Ne4 11.Bb2
Nxc3 12.Bxc3 Nc6
13.a4! b4 14.Bb2
This sharp unexplored position is suitable to
play for a win. For instance:
14...h6 15.Nh4 Be4 16.f3 hxg5 17.fxe4 Rxh4
18.exd5 Qxd5 19.Bf3 Qe6 20.Rc1 Rc8 21.Rg1.
White’s raging bishops control the board;
14...Qd6 15.Rg1 e5 16.Rc1 Bf5
192
17.Rxc6!? Qxc6 18.Bxe5 Bxh3 19.Nd4 Qc8
20.Bh5 Bc5 21.Qa1 Rg8 22.Ne2 Kf8 23.Bf3 Qd7
24.Nf4 Rd8 (24...Be6 25.Nh5) 25.Nxh3 Qxh3
26.Bg4 Qh4 27.Qc1 Be7 28.Bc7 Qxg5 29.Bxd8
Bxd8 30.Qc5+ Qe7 31.Qxd5. The opposite-coloured
bishops only hinder Black’s defence;
14...Be4 15.Rc1 Qd6 16.Rg1 e5 17.h4 h6
18.h5 hxg5 19.Nxg5 Bf5
20.f4!?
20.e4 dxe4 21.Bc4 Ra7 22.Qe2 is also
interesting.
20...exf4 21.Bg4 Ne7 (21...Bxg4 22.Qxg4
fxe3 23.Rg3!‚) 22.Bxf5 Nxf5 23.Qf3!? or 23.Qg4
Qd7 24.Qxf4ƒ.
B3. 5...dxc4
This move looks the most consistent retort to
5.Be2 since White will have to lose a tempo. Then
Black will lose a tempo, too, to enter the Queen’s
Gambit Accepted.
6.Bxc4
6.a4 Be6!? is dynamically balanced – 7.Nd4
Bd5 8.Nxd5 cxd5 9.b3 e5 10.Nf3 e4 11.Nd4 Nc6
12.Bb2 Bc5 13.bxc4 Bxd4 14.exd4 dxc4 15.Bxc4
Nxd4 16.0-0 0-0 17.d3. White has enough
compensation for the pawn, but that’s all.
6...b5
7.Be2!
Stayed the knight on b1, 7.Bb3 would have
been promising because after a4, ...b4 would have
been without a tempo. In the current situation, 7...e6
8.0-0 Bb7 9.e4 c5 10.d4 cxd4 11.Nxd4
11...Nc6! is close to equal, as
193
Delchev/Semkov point out.
11...Nbd7 is equally played, but it leaves
White the initiative – 12.Be3 Rc8 (12...Nc5 13.Qe2
Be7 14.Rac1 0-0 15.Rfd1 Qb8 16.f3) 13.Nxe6!?
fxe6 14.Bxe6 Qe7 15.Bh3 Rd8 16.Qe2 b4 17.Nd5
Nxd5 18.exd5 Nf6 19.Rfe1 Rxd5 20.Qf1 Kf7
21.Bf4+–, DisasterArea 1.54-Dirty 3Nov2015,
CCRL 2015.
12.Nxc6 Qxd1+ 13.Rxd1 Bxc6 14.f3 Bc5+
15.Kf1 Ke7 16.e5 Nd7 17.Bg5+.
Here in Korobov-Lesiege, Baku 2016, Black
erred and quickly lost after 17...f6? 18.exf6+ gxf6
19.Re1 e5 20.Bh4 b4 21.Nd5+ Bxd5 22.Bxd5 Rac8
23.Rad1 Rc7 24.f4 Bd6 25.Rd3 h5 26.Bb3 Rh6
27.Rd5 Rc6 28.fxe5 Nxe5 29.Rdxe5+ 1-0. However,
the calm king retreat neutralises most of White’s
pressure:
17...Ke8! 18.Rac1 h6 19.Bd2 a5.
7...e6 8.0-0 Bb7 9.d4 c5 10.dxc5
10...Qxd1
The most frequent move. 10...Bxc5
11.Qxd8+ Kxd8 is similar. See Game 32
Cheparinov-Banusz, Sibenik 2016.
11.Rxd1 Bxc5 12.Nd2! Nbd7
12...0-0 13.Nb3 Bb4 (13...Bb6 14.a4 b4
15.a5 Bd8 16.Nc5) as in Kramnik-Tomashevsky,
Moscow 2012, transposes to the main line after
14.Bd2 Nbd7 15.a4.
 13.Nb3 Bb4

13...Bb6 allows White to show the main idea

of White’s set-up:
14.a4 b4 15.a5! Bd8 16.Na4 Be7 17.f3 g5
18.e4 g4 19.Be3 gxf3 20.gxf3 Rg8+ 21.Kf1 e5
22.Rac1 Bd8, Fressinet-Karjakin, blitz, Beijing
2012, when simplest was 23.Nb6 Rb8 24.Nc4+–.

14.Bd2 0-0 15.a4 bxa4 16.Nxa4 a5 17.Bxb4

axb4 18.Nac5 Bd5

This occurred in the game Topalov-Wang

Hao, Stavanger 2013. Instead of 19.f3 Rxa1 20.Rxa1
Rc8= White had to keep his rook on the d-file –
19.Rxa8 Rxa8 20.f3 Bxb3 21.Nxb3 g5 22.Kf2 Ra2
23.Rd2, and his king is close to the b-pawn. For
instance:

23...Ne5 24.Nc5 Ra1 25.e4 Nh5 26.g3 Nf6

27.Rd1 Ra2 28.Rb1 Nfd7 29.Nxd7 Nxd7 30.Ke3
Kg7 31.Kd4².

194
 Chapter 9. Anti-Slav and Chebanenko
Annotated Games

31. Kiril Georgiev – Terbe

Arad 07.08.2018

1.c4 c6 2.Nf3 d5 3.e3 Nf6 4.Nc3 Bg4 5.Qb3

Qb6 6.Ne5

18.dxe5 Bxe5 19.Bc3²,

Ponomariov-Ivanchuk, Dagomys 2010;
13...Rc8 14.Nb6 Rd8 15.f3 0-0 16.Kf2 h5
17.h4 Rfe8 18.Rhd1 Bc8 19.b4 Bd7 20.Na4 Rc8
21.Be2 g6 22.Nb6 Rb8 23.e4 dxe4 24.fxe4 Nxe4+
25.Ke3 Nf6 26.Rf1 Nd5+ 27.cxd5 exd5+ 28.Kd2²,
Equinox 3.20-NirvanaChess 2.3, CCRL 2016.
11.Nc5 a6

6...Be6

6...Qxb3 7.axb3 e6 8.Nxg4 is similar –

White takes the bishop pair and tries to do
something on the kingside.

7.d4 Nbd7 8.Nxd7 Nxd7?!

The best answer is to hide the king on the

queenside – 8...Bxd7 9.Na4 Qxb3 10.axb3 0-0-0.
10...a6 11.Bd3 e6 12.Bd2 Bd6 13.Ba5 leaves
White the slightly better chances:
13...Ke7 14.c5 Bb8 15.Kd2 e5 16.Nb6 Ra7
17.f3 h5
The engines promise White some edge after
12.b4 e6 13.b3 Be7 14.Be2 h5 15.h4 or 12.Bd2 e6
(12...Ne4 alters the pawn structure, which is an
achievement for White – 13.Nxe4 dxe4 14.h4².)
13.f3 Bd6, but it would be difficult to break through
the centre. That’s why I recommend as a main line
5.cxd5, where the engines give a smaller edge, but
White has a more clear plan.

9.cxd5 Qxb3 10.axb3 Bxd5 11.b4!

This move forces Black to open the centre

since 11...e6 12.e4 Bb3 loses the bishop after
195
13.Ra3. It is true that 13...Bc2 14.Kd2 Bxe4 hold it, without committing any apparent mistake.
15.Nxe4 Bxb4+ 16.Kc2 Bxa3 17.bxa3² gives Black
enough material, but my pieces would be in 17...Nb6 18.Bc3 Ke7 19.Rfa1 a6 20.Be2
command. Rhd8 21.Bf3 Ke6 22.b5+–

The alternative is 11...f5 12.Nxd5 cxd5

13.Bd3 e6. We should seek to push b5 here –
14.Bd2 Bd6 15.Ke2 Kf7 16.Rhc1 a6 17.f3 Rhd8
18.Ra5 Nf6 19.b5.

11...e5 12.Nxd5 cxd5 13.Bd2

22...Nc4

(22...axb5 23.Re1+!)
23.R5a2 a5 24.b3 Nd6 25.Bxa5 Re8
26.Bg4+ Ke7 27.Bb4 Rxa2 28.Rxa2 h5 29.Re2+
1-0

13...exd4
32. Cheparinov – Banusz
Sibenik 12.10.2016
General considerations suggest that Black
should maintain the symmetry, but the game course
proved that it was a difficult task. Perhaps he should 1.d4 d5 2.c4 c6 3.Nc3 Nf6 4.e3 e6 5.Nf3 a6
have allowed an isolator – 13...Bd6 14.dxe5 Bxe5 6.Bd3 dxc4 7.Bxc4 b5 8.Be2 c5 9.0-0 Bb7 10.dxc5
15.Ra2 0-0 16.g3 Nf6 17.Ke2 a6 18.Bg2 Rac8 Bxc5 11.Qxd8+ Kxd8 12.Nd2
19.f4².

14.exd4 Be7 15.Bb5

Suddenly it transpires that Black cannot

connect his rooks. For instance:
15...a6 16.0-0 Rd8 17.Bd3 0-0 18.b5²;
15...Bf6 16.0-0 Bxd4 17.Bf4 a6 18.Rfd1 Ba7
19.Rxd5 axb5 20.Re1+ Kd8 21.Red1±.

15...Kd8 16.0-0² Bf6 17.Ra5!

My next step is to get to the b7-pawn from

b5. As simple as this position looks, Black failed to 12...Ke7
196
 manoeuvring with his two bishops.

12...Nbd7 13.Nb3 Ke7! transposes to the 14.f3 Nbd7 15.Na5

next comment.
The alternatives are inferior:
13...Bb4 is probably best met by 14.Rd1 (It
is slightly more precise than 14.Bd2 Nc5) 14...Ke7
15.Bd2 Rhd8 16.a3 Nc5 17.Ne4 Nxb3 18.Bxb4+
Ke8 19.Rxd8+ Rxd8 20.Nxf6+ gxf6 21.Rd1 with the
bishop pair advantage.
15.e4 Nc5 16.Na5 Rac8?! 17.Be3 Ba8,
Mesaros-Studer, Le Castella 2018, 18.a4! is
excellent for White, but 16...Bc7! 17.b4 Bb6 18.Kh1
Rac8 should hold on.
15...Ra7

13...Bb6 14.a4 b4 15.a5 Ba7 16.Na2 Bd5

17.Nd4 e5 18.Nf3!±.
13.Nb3 Bd6
In my opinion 13...Nbd7, giving the bishop,
is the lesser evil. Then 14.Na5 Rab8
The only game, Swierz-Bartel, Shorzow
2013, saw 14...Ra7?! 15.Rd1 Ba8 16.Bd2 Bb6
17.b4².
15.Bd2 Ne4 16.Nxe4 Bxe4 17.Rfc1 Rhc8
18.a4 looks very promising for White, but the forced
line 18...bxa4 19.Bc3 Bd6 20.Nc4 Nc5 21.Nxd6
Kxd6 22.f3 Bd5 23.e4 Nb3 24.exd5
finishes with the unexpected:
After 15...Nc5 White has two ways to
destroy the enemy queenside:
16.a4 b4 17.Na2 Bc7 18.Nxb7 Nb3 19.Nxb4
Nxa1 20.Bc4 Bb6 21.a5 Rhc8 22.axb6
22.b3?! Nxb3 23.Bxb3 Rxc1 24.axb6 Rxf1+
25.Kxf1 Nd7 26.Bc4 Nxb6 27.Bxa6 is drawish.
22...Rxc4 23.Na5 Rxb4 24.Nc6+ Ke8
25.Nxb4 Nd7 26.Bd2 Nb3 27.Bc3 g6 28.b7 Ra7
29.Rd1 Nbc5 30.Rd6 Rxb7 31.Nxa6;
16.b4 Bc7 17.a4

17...Nd5 18.Nxd5+ Bxd5 19.axb5 axb5

20.Bd2 Nb3 21.Nxb3 Bxb3 22.Bxb5².

16.Bd2
24...a3!!= (24...exd5 25.Rd1+–).
Therefore, White should consent to a small
plus after: 16.a4! b4 (16...Bb4 17.axb5) 17.Na2 was
more direct:
14.Nxc5 Nxc5 15.b4 Nce4 16.Nxe4 Nxe4
17.Bb2. The game Rychagov-Rozhkov, Voronezh
2016, continued 17...Rhc8 18.Rfc1 f6 19.f3 Nd6
20.Bd4 Rxc1+ 21.Rxc1, and White went on to win,

197
 31...Ke6 held on a little longer – 32.h4 Rf8
33.Kf2 h6 34.h5 f5 35.Kg3 fxe4 36.fxe4+–. The
e5-pawn is doomed.

32.Bxe5 1-0

17...Rb8 18.e4 Bc7 19.Be3 Raa8 20.Nxb7

Rxb7 21.Nc1± with domination.

16...Bc7

Or 16...Bb4 17.Nb3 Nc5 18.Nd4 Rc8 19.e4.

17.b4± Rc8 18.Rfc1 Bxa5 19.bxa5 Ne5

20.e4 Nc4

21.Bxc4 bxc4

The opposite-coloured bishops do not help

Black because the b7-bishop is horribly restricted –
21...Rxc4 22.a4 b4 23.Ne2 Rxc1+ 24.Rxc1±.

22.Na4 Nd7 23.Be3 Raa8 24.Rab1 Bc6

25.Nb6 Nxb6 26.axb6 Bb7 27.Bf4 Rc5 28.Bc7 f6
29.Rb4 e5 30.Rbxc4 Rxc4 31.Rxc4 Kf7??

198
 Chapter 10. Anti-Meran
Main Ideas

1.c4 c6 2.Nf3 d5 3.e3 Nf6 4.Nc3 e6 5.b3

Before we delve into the different plans from

it, I should point out some nuances in the:

Move order

The schemes with e3 and b3 have come to
the fore when White discovered the idea of gaining
space with g4-g5. Delchev’s book The Modern Reti
from 2012 was the first to popularise this plan, and it
quickly found many followers. Unfortunately, I have
a cold shower for all the fans of magical solutions:
Black can sidestep the plan with g4 with a
clever move order:
Whenever Black plays ...Nbd7, we should
answer Qc2, preparing to meet ...e5 by cxd5 and
Nb5. Otherwise we develop the bishop on b2 first.
The early ...Nbd7 offers us the powerful,
albeit double-edged, plan with Rg1 and g4. For
instance:
5...Nbd7 6.Qc2 Bd6 7.Bb2 0-0

5...Bd6! 6.Bb2 0-0!

I stubbornly tried here 7.Qc2 in one game,

but Black took the initiative with 7...e5! 8.cxd5 cxd5
9.Nb5 Nc6.
Therefore, we cannot avoid:
7.Be2! Nbd7 8.Qc2, reaching the following
Anti-Meran tabiya:

8.Rg1! e5 9.cxd5 Nxd5 10.Ne4 Bc7 11.g4.

White’s attack could quickly become overwhelming,
and we do not even need some extraordinary
ingenuity to finish the opponent off.

Iturrizaga-Biag
chess.com 2017

199
 20.Nh2±. White’s attack is playing itself, and we could
even throw in a piece to open the diagonals for our
akjoltoy-misterija bishop pair. I give enough examples in the
gameknot.com 2013 theoretical chapter.

8...a6 is another non-developing move, so

9.Rg1 should be possible, but I do not recommend
it. Instead I prefer to switch to standard play:
9.d4 e5
9...b5?! 10.0-0 Bb7 11.c5
10.cxd5 cxd5 11.dxe5 Nxe5 12.0-0 Be6
13.Rac1

White’s king is safe in the centre – 26.f4!

exf4 27.Bxf6 Rxf6 28.Nxf6 Rxf6 29.Rf1 Rf8
30.Qc4±.

Let’s now return to the bitter reality and

consider the tabiya after move 8. Although we had
to play the modest Be2, Black is still not immune
against the plan Rg1. It still works wonders against:
8...b6?! – 9.Rg1! Bb7 10.g4. Most players prefer to play against the
isolator in such positions. White’s game is
undoubtedly easier.

Undoubtedly the most challenging

continuation is:
8...Re8 9.0-0 e5 10.cxd5 Nxd5

200
 We got a typical Sicilian structure with
colours reversed. Black’s piece are passive, so we
should not fear an attack. On the other hand, it is not
clear what we could do ourselves. The thematic
break d4 would only activate the black bishop. If
Black’s knight were on c6, we could prepare b4-b5,
but with the pawn on c6, b4-b5 does not make much
sense. The only constructive plan I see for White is
to put the rooks on e1 and d1 and push e4 to gain
space in the centre. Then we could think about an
activity on the queenside or even d4. Play may
continue 11.Ne4 Bc7 12.Rad1 (12.Ng3) 12...Nf8
13.d3 Qe7 14.a3 a5 15.Rfe1 Bf5 16.Ng3 Bg6 17.Bf1
Nd7 18.e4.

Theoretical status

Commentators tend to overestimate White’s

chances in the position of the last diagram.
Cummings “sidesteps” altogether the line 8...Re8
9.0-0 e5!?. Hansen arrives at the position from the
Sicilian Reversed, contenting himself to cite the
game Grischuk-Kramnik, Wijk aan Zee 2011, where
Black simply made a gross positional mistake.
Delchev gives much more detail, but it all
turns around the same Grischuk-Kramnik game. As
a whole, I feel that the theoretical coverage of the
Anti-Meran has been not too useful from a practical
standpoint. Thus White still has enough ground for
innovations.

201
 Chapter 10. Anti-Meran
Step by Step
1.c4 c6 2.Nf3 d5 3.e3 Nf6 4.Nc3 e6
a) 8.h3 does not seem the most useful move
in this position. Let’s check 8...e5
Then 9.cxd5 Nxd5 10.Ne4 Bc7 11.Ng3 Re8
would be similar to 8.Be2 Re8, only White played
h3 instead of Be2, which is not something to boast
about. However, the sharper alternative is worse:
9.d4?! e4 10.Nd2 Re8 11.g4 h6 12.0-0-0 Nf8
13.Kb1 Ng6 14.Be2 Bd7 15.Rhg1 Bb4 16.g5 hxg5
17.Rxg5 b5

5.b3 Bd6!

5...a6 6.Bb2 b5 defines Black’s plan too
early. We could exploit that by developing our
bishop on d3 instead of the common e2 – 7.Qc2 Bd6
8.d4 0-0 9.c5 Bc7 10.Bd3.
6.Bb2 0-0!
6...Nbd7 offers White additional options,
based on the Bayonet attack with g4:
7.Qc2 0-0
Now besides 8.Be2, which transposes, White
also has:
It is obvious that Black has taken the
initiative – 18.Rh1 bxc4 19.bxc4 Nh7 20.Rg3 Rb8
21.Nb3 Be6, Kupsys-Efremov, ICCF 2014.
b) 8.Rg1! e5 9.cxd5
9.g4 d4!? 10.exd4 exd4 11.Nxd4 Re8+ looks
very pleasant for Black – 12.Be2 Bxh2 (12...Nc5)
13.Rh1 Nc5 14.Rxh2 Qxd4 15.0-0-0 Rxe2ƒ.
9...Nxd5 10.Ne4 Bc7 11.g4

White has a clear-cut plan for attack, while

his king is rather safe in the centre. See Game 33
Hacker-Kahl, ICCF 2013.

202
 7.Be2!
7.Qc2 is a sloppy move order. Black takes
the initiative with 7...e5! 8.cxd5 cxd5 9.Nb5 Nc6
10.Nxd6 Qxd6 11.d4 Nb4 12.Qd1 e4 13.Ne5 Ne8
14.a3 f6.
7...Nbd7
7...e5!? is possible – 8.cxd5 cxd5 9.Nb5 e4
10.Ne5 Ne8 11.Nxd6 Nxd6 12.f3 f6 13.Ng4 Nc6
It would be too much to part with the second
bishop – 13...Bxg4 14.fxg4 Nc6 15.0-0 d4?! 16.b4!
d3 17.Qb3+ Kh8 18.Bd1 Qe7 19.Rc1 Rac8 20.g5
fxg5 21.Bg4±, Gonda-Savchenko, Budapest 2013.
14.0-0 Be6 15.Qe1 with complex play. This
line is still untested.
8.Qc2
A. 8...Re8!; B. 8...Qe7; C. 8...b6?!; D. 8...a6
A. 8...Re8 9.0-0 e5
9...dxc4 10.bxc4 e5 11.d4 exd4 is an
independent line (11...Qe7 and 11...e4 transpose to
9...dxc4/line C2.) 12.exd4 Nf8 13.h3 Ng6 14.Rfe1
203
This is Dreev’s favourite set-up for Black.
Practice shows that if White takes space with c4-c5,
followed by Bc4, he enjoys a lasting pull. Therefore:
14...Be6 looks indispensable. Still White’s
better centre offers him decent manoeuvring options
– 15.Bf1 Qd7 16.Rad1 Bf5 17.Qb3 Rxe1 18.Rxe1
Rd8 19.Nd1 Nf4 20.Ne3 Be6 21.Ng5ƒ.
We meet 9...a6 by 10.d4, having won a small
battle. Indeed, the move ...a6 is hardly an
achievement for Black. The only way to justify it is
to push 10...b5.
10...dxc4 11.bxc4 e5 is a worse version of
another line. We answer, for example, 12.h3,
intending to meet 12...e4 13.Nd2 Qe7 by the
thematic 14.c5 Bc7 15.Ba3 or even 14.f3!? exf3
15.Rxf3.
11.e4 b4 12.e5².
In my opinion, the plan with 9...b6!? when
White has already castled, is Black’s most solid
option. We have nothing better than 10.d4 Bb7
11.Rad1 Qe7! 12.Rfe1, reaching the following
topical position:

White has two possible plans:
To push e4
To take on d5 and fianchetto the
light-squared bishop (or put it on d3).
It has been found that the otherwise logical
12...Rac8 does not address well the first plan, as the
rook occupies a closed file after 13.e4 dxe4 14.Nxe4
Nxe4 15.Qxe4. Carlsen-Topalov, Morelia/Linares
2007, went 15...Bb4 16.Rf1 Ba3 17.Bxa3 Qxa3,
when 18.Ne5 Rcd8 19.Qe3 Nxe5 20.dxe5 c5 21.Rd2
h6 22.Rfd1 gains control of the d-file.
Lately the main line has been:
12...Rad8 13.g3!?
13.Bf1 is probably premature due to 13...e5
(but not 13...dxc4 14.bxc4 e5 15.g3ƒ, as in my game
Korobov-Kiril Georgiev, Bilbao 2014).
See more details in Game 34 Le Quang
Liem-Wang Hao, rapid Huai’an 2017.
10.cxd5 Nxd5 11.Ne4 Bc7
204
12.Ng3
This move is in no way obligatory, but
sooner or later White will probably play it anyway.
His main active plan is Rfe1, Bf1, e4, and eventually
d4. For instance:
12.Rad1 Nf8 (12...N7f6 13.d3 Qe7 14.Ng3)
13.d3
13.d4 is not a good idea – 13...f5 14.Nc3 e4
15.Ne5 Be6.
13.a3 is a normal Sicilian move. A game
between engines went 13...Bf5 14.b4 Nd7 15.b5
cxb5 16.Bxb5 Rc8 17.Bd3 Bg6 18.Qb3 N7b6
19.Bb1 f5?! 20.Ng3 e4 21.Nd4 Qd7 22.d3 f4
23.Ngf5 Bf7 24.Qb5±, Gull 2.2-iCE 2.0 CCRL
2014.
13...Qe7 14.a3 a5 15.Rfe1 Bf5 16.Ng3 Bg6,
Polugaevsky-Piket, Aruba 1994, 17.Bf1 Nd7 18.e4
N5f6 19.Nh4 Qc5 20.Qd2 Qf8 21.b4 axb4 22.axb4
Ra2 23.Ra1 Rea8 24.Rxa2 Rxa2 25.Ra1 Rxa1
26.Bxa1 Qd8 27.Ngf5 with a slight pull.
12...Nf8
This set-up leaves a free lane for the f-pawn.
12...N7f6 13.d3 Qe7 14.a3 c5 15.Rac1 b6 16.Rfe1
Bb7
We should prepare e4:
17.Bf1! Rac8 18.e4 Nf4 19.Nf5 Qd8 20.g3
Ng6 21.b4².
13.a3 Qe7!
 It is important to prevent d4.
Grischuk-Kramnik, Wijk aan Zee 2011, saw
13...a5?! 14.d4! Ng6 15.Bd3 exd4 16.Nxd4 Bg4
17.Bf5±.
14.b4 a6 15.Rac1 Ng6
White’s position is more flexible, and he
could continue in the Taimanov style with 16.Bc4
Nb6 17.Bb3 Bg4 18.Qe4.
B. 8...Qe7 9.0-0 Re8
9...e5 10.cxd5 Nxd5 11.Ne4 (11.d3!?)
11...Bc7 12.Ng3 Re8 13.a3 a5 is the typical
Reversed Sicilian.
9...b6?! 10.cxd5 exd5
10...cxd5 11.Nb5 Ba6 12.Nfd4 Rfc8 13.Nc6
Qf8 14.Nxd6 Qxd6 15.Bxa6 Rxc6 16.Qd1².
11.Nd4 Nb8 12.Rac1² g6 13.f4.
9...dxc4 10.bxc4
The practically untested idea of Delchev
10.Bxc4!? is in the spirit of the English, but it
cannot pose serious problems to Black – 10...b5
11.Be2 Bb7 12.Ng5 Rac8 13.Nce4 Nxe4 14.Nxe4
Bc7 15.f4 f5=.
10...e5 11.d4 Re8
This position is known from the Anti-Meran
with 6.Qc2. It has always had the reputation of
somewhat better for White, but the fact that he has
tested 16 (!) different moves suggests that things are
not too clear. Black will push ...e4 and will attempt a
205
kingside attack. In the majority of games White
chose to undermine Black’s centre with f3, which
gave Black considerable piece play against the
weakened kingside. I propose to show character and
ignore the enemy threats:
12.a4!? e4 13.Nd2 Nf8 14.a5 Ng6
Or 14...a6 15.Rfb1 h5 16.Qd1 Bg4 17.h3.
15.Rfb1
The point of our defence. We prepare Qd1. I
do not see substantial dangers for us. For instance:
a) 15...h5 16.Qd1 h4 17.c5 Bc7 18.h3 a6
19.Nc4 Be6 20.Bc1². The b7-pawn is under
pressure.
b) 15...Nh4 16.g3 (16.Qd1!?) 16...Qd7 17.a6
b6 18.c5².
c) 15...a6 16.Na4
16.Nf1 Nh4 17.Qd1 is also possible –
17...Bc7 18.Ng3 Qd8 19.d5 Nf5 20.dxc6 bxc6
21.Qxd8 Bxd8 22.Nxf5 Bxf5 23.c5².
16...Bg4 17.Qd1 Bxe2 18.Qxe2 Rab8
19.Bc3².
In all the events our mobile central pawns
maintain enough tension.
9...a6 10.d4 e5 11.dxe5 Nxe5 12.cxd5 cxd5
leads to the same pawn structure as in the main line.
White could find a more useful move than 13.Rac1,
e.g. 13.Nxe5! Bxe5 14.Bf3².
10.Rac1!?

A cunning waiting move. By leaving d4 free,
we discourage 10...b6 in view of 11.cxd5 exd5
12.Nd4.
10...a6
10...Nf8 11.d4 Ng6 12.g3 puts the
manoeuvre of Black’s knight to the test.
11.d4 e5 12.cxd5 cxd5 13.dxe5 Nxe5
14.Na4!?
As a rule, exchanges favour us. 14.Rfd1 Be6
15.Nd4 Rad8 keeps more tension.
15...g6 16.Nxe6!, intending 16...fxe6 17.f4,
gave Ivanchuk the upper hand against Khenkin,
rapid, Riyadh 2017.
16.Nf5 Bxf5 17.Qxf5 Ba3 18.Bxa3 Qxa3
19.Rc2 Rd6 20.h3, Gajewski-Sebag, Montpellier
2015. Black’s queen is too active. The simple
20...Rc6! would have been roughly equal.
14...Bg4 15.Nxe5 Bxe2!
In the source game Kraus-Nitish, Prague
2017, Black quickly remained pawnless after
15...Bxe5?! 16.Bxe5 Qxe5 17.Bxg4 Nxg4 18.Qc7
Rac8 19.Qxe5 Nxe5 20.Nb6 Rxc1 21.Rxc1 Nc6
22.Nxd5.
16.Qxe2 Bxe5 17.Rfd1 Rac8 18.Bxe5 Qxe5
19.Qb2 Qxb2 20.Nxb2
206
White retains slight pressure in the endgame.
C. 8...b6?! 9.Rg1!
The die is cast. White sets for an all-out
attack, with excellent chances to win a miniature.
9...Bb7
Dreev chose twice in blitz 9...Ne5 to reduce
White’s striking force. Zvjaginsev correctly opted
for 10.Nxe5 Bxe5 11.f4 Bc7, when the straight
12.g4! e5 13.g5 Ne8 14.0-0-0 exf4 15.cxd5 cxd5
16.Bd3 is already winning.
10.g4
10...Qe7
Planning to trade the dangerous bishop from
a3.
Alternatively:
10...c5 11.g5 Ne4 12.Nxe4 dxe4 13.Qc3 f6
14.gxf6 Nxf6 15.Ne5±.
10...e5 11.g5 Ne8 12.h4 Nc7
Perhaps Black should seek counterplay with
12...b5 13.h5.
12...e4 13.Nd4 Be5 14.d3 occurred in
Papin-Ter Sahakyan, Chennai 2013. (14.Nf5!? b5
15.d4±) 14...exd3 15.Bxd3 g6, and here 16.f4 Bg7
17.cxd5 cxd5 18.f5± was crushing.
13.h5!. The position is ripe for destructive
tactics.
13.Bd3 gave Black a tempo after 13...g6
14.h5 Nc5 15.Be2, although White retained the
attack after 15...e4? (15...N7e6)
16.Nxe4!! Nxe4 17.d3 Bb4+ 18.Kf1 Bc8
19.dxe4 Bh3+ 20.Rg2 Ne6 21.a3+–, Giri-Michiels,
Montpellier 2015.
13...Re8
13...f5 14.gxf6 Qxf6 15.0-0-0 Rae8 16.Kb1
Re7 17.Rdf1 Ne6 18.d4 e4 19.Ne1±.
13...Ne6 14.0-0-0 b5 15.Bd3 e4 is similar to
Giri’s game – 16.Nxe4! dxe4 17.Bxe4 f5 18.gxf6
Nxf6 19.Bf5 Bc8 20.h6+–.
14.0-0-0 b5 15.g6 fxg6 16.Ng5 Nf6
You can test here your attacking skills:
17.f4! exf4 18.Nxh7 Nxh7 19.Qxg6±.
11.g5 Ne8 12.h4 Rc8
207
12...e5 blocks the main diagonal, but the
debacle comes on the light squares – 13.h5 g6
14.Nh4 Nc5 15.0-0-0 d4
Or 15...Rd8 16.b4 Nd7 17.f4 Bxb4 18.f5+–.
16.b4! dxc3 17.Bxc3 Ne6 18.hxg6 hxg6
19.Nxg6 fxg6 20.Qxg6+ N8g7 21.Bd3+–.
13.h5 Ba3 14.Bxa3 Qxa3
I have been following the game
Nepomniachtchi-A.Muzychuk, Jerusalem 2018.
The direct attack 15.Bd3 g6 16.Nh2 Qd6
17.Ng4 leads to a huge positional advantage:
17...Ne5 18.Nh6+ Kg7 19.Be2 f5 (19...c5 20.Nb5
Qd8 21.Qc3 f6 22.f4±) 20.d4 Nf7 21.0-0-0±.
Black’s pieces are passive, and his king is in a
mating net.
D. 8...a6 9.d4
9.Rg1 is possible, but I do not think it is
worth the risk. Black could choose between the
sharp 9...b5 10.g4 bxc4 11.bxc4 Rb8 12.Rb1 e5
13.cxd5 cxd5 14.g5 Ne4 15.Nxd5 Ndc5 16.Ba3
Rxb1+ 17.Qxb1 Be6°, Yurov-Jung, ICCF 2017, and
9...e5 10.g4 e4 11.g5 exf3 12.gxf6 g6 13.Bxf3 dxc4
14.Ne4
 as an improvement. Perhaps the most precise is:
14.Bd3 Re8 15.f4 Bc7 16.e4±.

10.cxd5 cxd5 11.dxe5 Nxe5 12.0-0 Be6

12...Bg4 helps White to exchange more

pieces – 13.Nxe5 Bxe5 14.Bxg4 Nxg4 15.h3 Nf6
16.Rac1 Rc8 17.Rfd1, Jovanic-Sedlak, Varazdin
2004, “when Black is playing for two results only”
(Delchev).

13.Rac1

14...cxb3 15.axb3 Be5 16.h4 Bxb2 17.Qxb2

Re8=. In both events Black’s pieces are better Delchev recommends 13.Nxe5 Bxe5 14.f4 –
coordinated. see Game 35 Hacker-Dolgov, cr 2013.

9...e5 13...Rc8 14.Qb1

There is no reason to play 9...Qe7 first. After

10.0-0 e5 11.dxe5 Nxe5 12.cxd5 cxd5 we get the
same pawn structure as in the main line, with the
black queen committed to e7. We can continue
13.Nxe5 Bxe5 14.Bf3 Rd8 15.Rfd1 Be6 16.h3 with
a pull.

9...b5 10.0-0 Bb7 11.c5 is a sad position for

Black, who scores the poor 25% – 11...e5 (or White
will play e4 himself) 12.Nxe5 Nxe5 13.dxe5 Bxe5

14...Nfg4

14...Qe7 15.h3 Rfd8 16.Rfd1 Bb8 17.Nxe5

Bxe5 18.Bf3 Bb8 19.Ne2, Nyzhnyk-Erenburg,
Norfolk 2017, drops a pawn to 19.Nxd5.

15.Nxe5 Bxe5 16.h3 Bh2+

16...Nf6 17.Bf3 Qa5 18.Ne2 offers White a

It is clear that White’s plan should be based nice version of an IQP position with exchanged
on f4. The only question is when is the best moment dark-squared bishops.
for it. Delchev-Brunello, EU tch. 2011 saw the too
straightforward: 17.Kh1 Bb8 18.Bxg4 Bxg4 19.Qd3
14.f4 Bc7 15.e4². Delchev suggested 15.a4
208
 White enjoys a pull in a safe position. The
only game in my database, Mamedyarov-Andreikin,
Tashkent 2014, went:
19...Qd6 20.f4 Be6 21.Rcd1 Qc5 22.Nxd5
Bxd5 23.Qxd5 Qxd5 24.Rxd5 with a fairly sound
extra pawn.
19...Be6 looks more stubborn. White’s game
is only slightly better after 20.Ne2 Qd6 21.f4 f6
22.Rfd1. His knight should be dominant, if he
succeeds in trading the dark-squared bishops.

209
 Chapter 10. Anti-Meran
Annotated Games

33. Hacker – Kahl

ICCF, 10.02.2013

1.Nf3 d5 2.c4 c6 3.e3 Nf6 4.Nc3 e6 5.b3

Nbd7 6.Qc2 Bd6 7.Bb2 0-0 8.Rg1 e5 9.cxd5 Nxd5
10.Ne4 Bc7 11.g4

16...Kh8. This occurred in Ivanovic-Szerlak,

ICCF 2015. It makes sense here to enable long
castling with 17.Rg2!, e.g. 17...f5 18.gxf6 Nxf6
19.0-0-0 with an attack down the g-file.

12.g5 Nf8

12...Kh8 13.h4 Qe7 14.h5 Nb4 15.Qc3 Nb6

16.a3 Bf5 17.d3 N4d5 18.Qc2 a5 gave White a
strong initiative in Sigurdsson-Johansen,
Lechenicher SchachServer 2012. A practical player
11...Re8 would have kept the tension with 19.0-0-0 or
19.Nh4. The game course was not bad either – 19.h6
g6 20.Ned2 Bc8 21.e4 Nf4 22.d4±.
I do not understand why most practical
games feature this passive move. 11...Qe7 is more 13.h4 Bf5 14.a3
consistent, to employ the f8-rook on the f-file. It
does not discourage White from his plan, but:
12.g5 f5 13.gxf6 N7xf6 14.Bc4 Kh8! In Docx-Gagunashvili, Warsaw 2013, White
allowed Black to hold in Hacker-Evtushenko, ICCF decided to save a3 and won quickly after 14.h5!?
2013, after 15.Neg5 b5 16.Be2 e4 17.Nh4 a5 Ne6 15.Rc1 Nd4
18.Nxe4 Be5 19.Ng5 Bxb2 20.Qxb2 Ng4 21.Nhf3
h6 22.h3 Nxf2 23.Kxf2 hxg5=. It is more interesting
to try:
12.h4! a5 13.a3 N7f6 14.Nxf6+ Nxf6 15.g5
Ng4 16.Be2!
In Von Weizsaecker-Vinchev, ICCF 2012,
most of the material disappeared from the board
after 16.Bd3 f5 17.gxf6 Qxf6 18.Bxh7+ Kh8 19.Be4
Nh2 20.0-0-0 Nxf3 21.Rg3 Bf5 22.Rxf3 Bxe4
23.Rxf6 Bxc2 24.Rxf8+ Rxf8 25.Kxc2 Rxf2=.

210
 16.Bxd4 exd4 17.Nf6+ Nxf6 18.Qxf5 Ne4?
19.g6 fxg6 20.hxg6 dxe3 21.fxe3 h6 22.Qf7+ Kh8
23.Rh1 Bg3+ 24.Ke2 1-0. The only defence was
18...Nxh5! 19.Bc4 g6 20.Qxf7+ Kh8 21.Be2 with
some initiative. However, Black has survived the
first wave. I think that a little prophylaxis does not
hurt.

14...Bg6 15.d3

White’s plan is to prepare h5 by Ng3 or Rh1,

so d3 is useful anyway. B.Maksimovic-Ali Marandi,
Skopje 2016, saw instead 15.Be2 Kh8 16.0-0-0 a5
17.d3 a4 18.b4 Nd7, and Black was already hopeless
in view of 19.Rh1 Bh5 20.Nxe5.
We can take stock. The opening is over, and
Black has escaped the worse scenarios. Still, he has
not equalised completely. He should be constantly
concerned about his king. More importantly, I do
not see any plausible counterplay for him.

20.Rg4 Qc7 21.Rag1 Nge7 22.h5 Rad8

23.Qc2 Nf5 24.Nd2 Rd7 25.Nf3 Qd6 26.Re4 Rde7
27.Kd2!

Underlying the fact that Black is tied up and

down. He is helpless against the battery Bb2+Qc3,
followed by a decisive f4. But first the king goes
unhurriedly to a safe heaven.

15...Ba5+ 27...Nb6 28.Kc1 c5 29.Nh2 Nd7 30.Ng4

Kf8 31.Kb1 Re6 32.Qc3 f6

It looks wise to trade the “bad” bishop for a

knight. 15...f5 only made weaknesses in
Wojtaszek-Krysztofiak, Leba 2004, 16.gxf6 Nxf6
17.0-0-0 Nxe4 18.dxe4 Qf6 19.Ng5±.

16.Nfd2 Bxd2+ 17.Qxd2 Bh5 18.Be2 Bxe2

19.Kxe2 Ng6

33.h6! fxg5 34.hxg7+ Kg8 35.Qc4 Qe7

36.Bxe5 h5 37.Rh1 b5 38.Qd5 Nxe5 39.Rxe5 Nxg7
40.f4 Qd8 41.f5 Qxd5 42.Rxd5 Nxf5 43.Rxh5
211
Nxe3 44.Rdxg5+ Kf8 45.Nxe3 Rxe3 46.Rxc5 a6
47.Kc2 Kg7 48.Kc3 Rf3 49.Rhd5 Re6 50.Kb4 Rf1
51.Rg5+ Kf6 52.Rh5 Kg7 53.Rc7+ Kg6 54.Rch7
Ref6 55.R5h6+ Kg5 56.Rxf6 Rxf6 57.Ka5 Rd6
58.b4 Rxd3 59.Kxa6 Rxa3+ 60.Kxb5 Kf6 61.Kb6
1-0

34. Le Quang Liem – Wang Hao

rapid Huai’an 2017

1.d4 Nf6 2.c4 e6 3.Nf3 d5 4.Nc3 c6 5.e3

Nbd7 6.Qc2 Bd6 7.Be2 0-0 8.b3 b6 9.0-0 Bb7
10.Bb2 Qe7 11.Rad1 Rad8 12.Rfe1 Rfe8 21.dxc5 bxc5 (21...Qxc5 22.b4 Qc8 23.Rfe1)
22.Rfe1 Qc7 23.Bf3 Bxf3 24.Nxf3 Rxd1 25.Rxd1
Rd8=.
More complex is:
14.Bd3 c5!
14...e5 15.e4 dxc4 16.Bxc4 b5 17.Bf1 exd4
18.Nxd4 Be5 19.Nf5 Qc5 20.Qc1 a6 21.Re2 is a
typical Meran structure, where White has some
space advantage, akjoltoy-aboveoeuf,
gameknot.com 2013.
15.cxd5 exd5

13.g3!?

The plan with e4 leads to completely drawn

positions. For instance: 13.h3 h6 14.e4 dxe4
15.Nxe4 Nxe4 16.Qxe4 Nf6 17.Qc2 Bb4 18.Rf1
Ba3 19.Ne5 Bxb2 20.Qxb2 c5

It is unclear how to proceed from here. After

16.Nh4 Bb8 17.Nf5 Qf8, Black will repel the knight
with ...g6.

13...a6

Black has made only one draw out of 6 game

after 13...c5, but this move is perfectly playable.
14.cxd5 exd5 15.Bf1 Rc8!
15...a6 might turn superfluous – 16.Bg2 Rc8
212
17.Qb1 h6 18.Nh4 c4 19.Nf5 Qf8, Ding 19.Nb3.
Liren-Topalov, Shamkir 2018. Now 20.e4 dxe4
21.bxc4 Rxc4 22.Nxe4 Nxe4 23.Bxe4 Bxe4 16...exd5 17.Nxd5 Nxd5 18.cxd5 Qe4
24.Rxe4 Rxe4 25.Qxe4² would have been in White’s
favour – his queen and knight are rather menacing.

15...Nf8 16.Bg2 Ne6 is a wrong set-up –

Black’s pawns hang after 17.dxc5 bxc5 18.Nh4, Le
Quang Liem-Navara, Saint Louis USA 2017.

19.Qc3

19.Qe2 b5 20.Bg2 kept the balance, while

after the text, 19...Nf6! 20.Bg2 Qa4 would have
16.Bh3 cxd4 17.Nxd4 Bb4 18.f3 a6 19.Rc1 probably cost White the d5-pawn. Still, the position
g6 20.Qd1 h5 21.Rc2. White has the centre under remains unclear – 21.Qd3 Bxd5 22.Bxf6 Be4
control, but it would be difficult to make progress. 23.Qe2 gxf6 24.Nh4.

14.Bf1 dxc4 15.bxc4 c5 19...Ne5 20.Nh4 Bf8

16.d5?! 21.Bg2

This break is premature as it leads to 21.f4 Ng4 22.Bg2 Qb4 would lead to a
multiple exchanges. White should have exploited his balanced queenless middlegame.
space advantage by keeping the tension with
16.Nd2! cxd4 17.exd4 h6, and only now 18.d5 Nc5 21...Qc4?!
213
 d4.

The critical moment of the game. 21...Qa4 14...Bc7

was indispensable, with the idea 22.f4? Nc4 23.Bc1
Nd6, blocking the e-pawn. Thus White would have
to play 22.Qb3 Qb5 23.Qc2 c4∞.
22.f4 Qxc3 23.Bxc3 Nc4 24.e4
Delchev considers 14...Bb8 15.Rad1 Re8
16.Kh1². However, the game, Critter 0.90-Deep
Sjeng 2010, CCRL 2011 was drawn after 15...Qc7!

Now White’s pawns are stronger as they

constrain the enemy pieces.

24...b5 25.e5 b4 26.Ba1 Na3 27.Bb2 Nb5

28.d6 Bxg2 29.Kxg2 (29.Nxg2!±) 29...c4 30.Nf3 c3
31.Bc1 f6?

(31...a5! 32.f5 h6)

32.d7+– Re7 33.e6 f5 34.Nd4 c2 35.Nxc2
Nc3 36.Rd2 Ne4 37.Rxe4 fxe4 38.f5 g6 39.Rd5
gxf5 40.Nd4 1-0

16.Qd3 Re8 17.Bf3 Ba7 18.f5 Bd7 19.Nxd5

Nxd5 20.Qxd5 Bxe3+ 21.Kh1 Bb5 22.Rfe1.
35. Hacker – Dolgov 16.Qc1 is no improvement due to 16...b5
ICCF, 10.02.2013 17.Bf3 Ba7 18.Nxd5 Nxd5 19.Bxd5 Qxc1 20.Bxc1
Rac8 and Black regains the pawn thanks to the
resource ...Rxc1. Perhaps he should try:
1.Nf3 d5 2.c4 c6 3.e3 Nf6 4.Nc3 e6 5.b3 16.Bf3 Ba7 17.Qf2 Rad8 18.Ne2, threatening
Bd6 6.Bb2 0-0 7.Be2 Nbd7 8.Qc2 a6 9.d4 e5 Bxf6.
10.cxd5 cxd5 11.dxe5 Nxe5 12.0-0 Be6 13.Nxe5
Bxe5 14.f4 14...Bd6!? 15.Rad1 Qe7 16.Bf3 Rac8
17.Qd3 Ba3

The idea of this move is rather simple – Now 18.Bxa3 led to a quick draw in
White hopes to win quickly the d5-pawn with the Haugen-Hagen, ICCF 2014 – 18...Qxa3 19.h3 Qa5
help of f4-f5 or to trade the dark-squared bishops via 20.Nxd5 Nxd5 21.Bxd5 Rc3 22.Qe4 Bxd5 23.Qxd5
214
Qxa2 24.Rd3 Rxd3 25.Qxd3 g6 26.Rd1 ½-½.
White could have preserved the tension with
18.Ba1!? Bc5 19.Kh1 Rfd8 20.h3.

15.Rad1 Rc8

Junior 10-Fruit 2.2.1 CCRL 2006, saw

15...Re8 16.Bf3 Bb6, when instead of forcing the
play with 17.f5 Bd7 18.Nxd5 Nxd5 19.Rxd5
Rxe3!=, White should have made a prophylactic
move – 17.Kh1. The point is that 17...Bxe3? would
be a mistake owing to 18.f5 Bd7 19.Nxd5+– Nxd5
20.Bxd5.
White has some practical chances:
16.Bf3 Re8 22...Rc2 23.Rf2 Rdc8 24.Bxd5 Bb6 25.Bd4;
22...b5 23.Rf2 Kf7 24.Rd3 g6 25.h3; 22...Bc3
23.Rf2 b5 24.Rd3 with the better bishop.
After the text the game sails forth to an easy
draw:

17...Ba5 18.Qd3 g6 19.Rc1 Bf5 20.Qd1

Rxe3 21.Nxd5 Nxd5 22.Qd4 Qf6 23.Rxc8+ Bxc8
24.Qxf6 Nxf6 25.Bxf6 Bc3 26.Rc1 Bxf6 27.Rxc8+
½-½

17.Kh1?!

I think that only 17.Qf2 Ba5 18.Ne4 leaves

White some chances. (18.Nxd5 Bxd5 19.Bxf6 Qxf6
20.Bxd5 b5 21.Kh1 Bb6=) For instance:
18...Nxe4 19.Bxe4 f6 20.Qh4 (20.Bf3!?
might be the best option) 20...f5 21.Qxd8 Rexd8
22.Bf3

215
 Chapter 11. Anti-Queens’s Gambit
Main Ideas

At first I did not intend to discuss 1.c4 e6 at

all. My deep conviction is that White’s best option is
to opt for the main lines of the QGD. I joked that if
after 2.Nf3 d5 he happened to touch the wrong pawn
and had to play 3.e3, his best follow up remained
d4! No matter how well explored is the QGD today,
it remains the king of the closed openings, with
plenty of different strategic plans. The only detail is
that if you play only the Carlsbad structures with
cxd5, you should counter the move order 1...Nf6
2.Nc3 e6 by 3.e4. Now Black could delay ...c5, but that should
not bother us as we have a clear plan to pursue:
Although I did not change my mind about 6...b6 7.cxd5 exd5 8.d4 Bb7 9.Bd3 Nbd7 10.0-0 a6
3.d4!, I finally decided to add a chapter on the 11.Ne5! intending Qf3 or f4.
Anti-QGD since we have already done half of the
job anyway – the position after 3.e3 Nf6 4.Nc3 c5
5.d4! is covered in chapters 2 and 3!

The main line is 6...c5 7.cxd5 with an

Thus I had to deal “only” with the plan with
4...Be7 5.b3 0-0 6.Bb2.
important split. I should warn you that you should
not expect any theoretical advantage after both
7...Nxd5!? and 7...exd5. The difference is mostly in
the character of play. Taking by knight gives us a
better control on the centre, so we could try some
sharp attacks with 7...Nxd5 8.Qc2, then Bd3, h4.
However, the clever 8...b6! 9.a3 Bb7 should be a
signal to slow down and switch to central play. Our
chance is that this nuance is still little known.

Line B2 deals with 7...exd5 8.d4

216
 Here the plan with 8...b6 is again not very
challenging. Black does not pressurise d4, and that
gives us the opportunity to develop our bishop on
the active square d3, aiming for the same scheme
with Ne5: 9.Bd3 Bb7 10.0-0 Nbd7 11.Ne5.
The big problem is 8...Nc6. Now 9.Bd3?!
would be pointless owing to 9...cxd4, so we have to
content ourselves with 9.Be2 Ne4! and our last
chance for an opening advantage is probably
10.dxc5 Bf6

11.Rc1, when Black must find the far from

obvious novelty 11...Nb4! or remain pawnless after
11...Nxc5?! 12.Nxd5!.

As you see, we have a couple of tricks in our

bag, so the Anti-QGD is still viable.

217
 Chapter 11. Anti-Queens’s Gambit B. 4...Be7!? practically forces White to
Step by Step fianchetto the bishop – 5.b3 0-0 6.Bb2 c5

6...b6 without ...c5 is another problem

1.c4 e6 variation. If we want to be consistent with our main
line against 6...c5, we should choose 7.cxd5 exd5
8.d4 Bb7 9.Bd3 Nbd7 10.0-0
After 1...Nf6 White has the additional option
of 2.Nc3 e6 3.e4, although my opinion is that 3.Nf3
d5 4.d4! is best. If you hate main lines, you play
4.e3. The rest of the chapter is for you.

2.Nc3 d5 3.e3 (3.d4!) 3...Nf6 4.Nf3

Black has often played here:

a) 10...a6. We answer with our thematic
11.Ne5 Bd6 12.f4 (12.Ne2 ia also possible), and
12...Ne4 13.Rc1 is easier to play with White, as we
have a clear idea what to do later – the rook lift
Rf1-f3-h3.

You should not be surprised that Black
scores around 50% in this position. He has several
decent set-ups, and White often overpresses because
he feels he has the “advantage of the first move”!
Still, White has found a few active ideas lately.
b) 10...Ne4 11.Rc1 Nxc3 12.Rxc3 Bd6
13.Qc1!. We aim to trade dark-squared bishops by
Ba3. The point is that 13...Qe7 stumbles into
14.Rxc7! grabbing two central pawns for the
exchange.

A. 4...c5; B. 4...Be7!? 7.cxd5

A. 4...c5 5.d4!
Mamedyarov tried against Nakamura, rapid
Saint Louis 2018 the sneaky 7.Rc1.
Now the play transposes to Chapter 2. For
instance, I consider in line B1 5...a6 6.cxd5 exd5
7.g3, and the less principled 7.Be2 leads to Chapter
2/Game 4 Rybka 3-Chessmaster 11 CCRL 2008.
5...Nc6 6.cxd5 Nxd5 is the subject of
Chapter 3.
5...dxc4 transposes to the QGA. I deal with it
in Chapter 9 Game 32 Cheparinov-Banusz, Sibenik
2016.

218

It hopes for 7...Nc6 8.cxd5 exd5 9.Na4 b6
10.d4, temporarily winning a pawn – 10...Be6!
(10...cxd4 11.Nxd4) 11.dxc5 bxc5 12.Nxc5 Ne4
13.Nxe4 dxe4 14.Qxd8 Rfxd8 15.Nd4 Bb4+ 16.Kd1
Rac8. Black regains the material, but in a slightly
worse ending after 17.h3 Nxd4 18.Rxc8 Rxc8
19.exd4 a5 20.Bb5 g6 21.Rf1 f5 22.Bc1 Bc3
(22...Kf8! is close to equal) 23.Be3 Bxd4 24.Bxd4
Rb8 (24...Rd8 25.Ke2 Rxd4 26.Rc1²) 25.Be2 Rd8
26.Kc2 Rxd4 27.Rd1². Nakamura’s choice:
7...dxc4 8.bxc4 b6 9.g3 Bb7 10.Bg2 Nc6
allowed Black to get an improved version of a well
known Reti system, where Rc1 is obviously
superfluous.
Now B1. 7...Nxd5!? is more tangled, while
B2.7...exd5 should equalise with precise play.
B1. 7...Nxd5!? 8.Qc2
If we try to trick the opponent with the move
219
order and start with 8.a3, 8...b6! throws a spanner in
the works (8...Nc6 9.Qc2!). Still, nothing wrong has
happened, and we can transpose to the main line
with 9.Qc2.
White has also tried to improve with 8.Qb1,
when the same 8...b6 solves all the problems.
Beliavsky’s choice 8...Nc6 9.Bd3 f5
(9...h6÷) 10.0-0 Qd6 is also not bad.
9.Bd3
9.Nxd5 exd5 10.Be2 (10.d4? cxd4µ)
10...Ba6.
9...h6 10.Nxd5 exd5 11.0-0 Nc6 12.Bb5 Bd7
13.d4 Rc8 14.Rd1 Bg4 15.dxc5 bxc5=.
8.Nxd5 exd5 9.d4
A fresh try was 9.Be2 Nc6 10.d4
10...b6 (10...Qa5+ 11.Qd2 Qxd2+ 12.Kxd2
b6 is simler.) 11.0-0 a5 12.Rc1 Qd6 13.Qd2 Rd8
14.Rfd1 Qg6 15.dxc5 bxc5 16.Ne5 Nxe5 17.Bxe5
Bf5= Harikrishna-Caruana, Saint Louis 2019.
9...Nc6 10.dxc5 Bxc5 11.a3 Bg4 12.Be2
d4=.
8.Be2 Nc6 9.0-0 Bf6 10.Qc1 b6 11.Nxd5
Bxb2 12.Qxb2 Qxd5 13.a3 a5 could be dangerous
only for White, Ivkov-Korchnoi, Cannes 1986.
8...b6!?
White hoped for 8...Nc6 (the most frequent
answer!) 9.a3 b6 10.h4!? – introduced by Karjakin
against Anand.
Actually, 10.Nxd5 Qxd5 11.Bd3 h6 12.Bh7+
Kh8 13.Be4 may be stronger. For example: 13...Qd6
14.d4 (14.Rg1!?) 14...Bb7 15.dxc5 Qxc5 16.Qxc5².
10...f5 11.Nxd5 exd5 12.d4 Be6 13.Rd1 Rc8
14.Ba6 Rc7 15.dxc5 bxc5 16.0-0 h6 17.Bc4 Kh8
18.Rd2 Bd6 19.Rfd1², Roiz-Zanan, Israel 2018.
8...Nb4 aims to provoke a3 in order to
weaken b3. However, it also gives us a tempo:
9.Qb1 b6 10.a3 Nd5
11.Bd3!?
11.Qc2 transposes to the main line.
11.Nxd5 exd5 12.d4 is totally even –
12...Nc6 13.Bd3 (13.Bb5 Bd7 14.0-0, Hou
Yifan-Bluebaum, Karlsruhe/Baden Baden 2018,
14...Na5=.) 13...h6 14.0-0 Bg4 ½-½,
Tazelaar-Krimbacher, ICCF 2017.
11...f5
11...h6 12.g4 Bb7 13.Ne4 deserves a test.
12.Nxd5
12.h4 looks already pointless – 12...Bb7
13.h5 Bf6 14.Qc2 h6 15.Rh3 Nd7³.
12...exd5 13.0-0 Be6 14.Qc2 Nd7 15.Ba6
Nf6 16.Rac1 Bd6 17.g3. I prefer White here.
220
9.a3
9.Nxd5 exd5 10.d4 Nc6 11.Be2 Be6=,
Karjakin-Kramnik, rapid Moscow 2018.
9...Ba6
9...Bb7 poses more complex problems.
The idea is to meet 10.h4 by 10...Nf6 with
double-edged play. I analysed:
10.Bd3 f5
10...Nf6 is more natural. See the instructive
computer Game 36 DiscoCheck 4.3-LoopList 6.00
CCRL 2013.
11.0-0 Nc7 12.Rfd1 Bf6 13.Rac1 Qe7
14.Bf1 Nc6 15.Ne2 with a tangled position. This
line has not been tested in practice yet.
10.Bxa6 Nxa6 11.h4!?
White has an initiative.
11...Nac7 looks solid, but after 12.Ne2 Bf6
13.Ng5 (otherwise it is difficult to justify 11.h4)
13...g6 14.Nf3 Bxb2 15.Qxb2 f6 Black’s position
looked a bit shaky in Morozevich-Caruana,
Thessaloniki 2013. White keeps the pressure with
16.h5 g5 17.d4, and Morozevich’s 16.b4 was also
slightly better for White.
The latest game in this line,
Andreikin-Tomashevsky, Satka 2018, saw 11...f5.
Instead of 12.Nxd5 Qxd5 13.Qc4 White should have
kept more pieces in order to use the weakness of e6
and e5 – 12.Rc1 Rc8 13.Ne2 Nac7 14.d4 Nb5 15.0-0
 yields White less than 50%.

9...Bb7 10.0-0 Nbd7

The point is that 15...Bxh4 (15...cxd4

16.Qd3) 16.dxc5 Rxc5 17.Qd3 Nbc7 (17...Ndc3
18.Ned4 e5 19.Nxh4²) 18.Nxh4 Qxh4 19.Be5 offers
White excellent compensation for the pawn, and
15...Qd7 16.dxc5 Bxc5 17.Qd3 Rfd8 18.Rfd1 is also
pleasant for him owing to the d4-square for the
knight.
B2. 7...exd5 8.d4
8.Rc1 is hardly very wise owing to 8...d4
9.Na4, when Black can equalise with 9...dxe3
10.dxe3 Qxd1+ 11.Rxd1 Nc6, or sac a pawn –
9...Nc6!? 10.Nxc5 Bg4©.
11.Ne5
We take the chance to claim some space
until it is still open. We aim for an attack with f4,
Qf3.
The main move 11.Rc1 allows 11...Bd6 with
a fine game.
11...Rc8
Or 11...Bd6 12.Nb5 Bb8 13.f4 a6 14.Nc3.
11...a6 might be a waste of time. Then
12.Qf3 is possible, planning 12...Qc7 13.Qh3 Rfd8
14.f4. However, it is more solid to play against the
d5-pawn: 12.Ne2!? g6 13.Nf4 Re8 14.Rc1 Bd6
15.Nxd7 Nxd7 16.Be2 Nf6 17.Bf3 Ne4 18.g3².

12.Ne2

The above-mentioned manoeuvre 12.Qf3 g6

13.Qh3 is still a fair alternative.

12...cxd4 13.Nxd7 Qxd7

B21. 8...b6; B22.8...Nc6

B21. 8...b6 9.Bd3

The “conservative” development 9.Be2

221

Now both 14.Nxd4, Ostovic-Pribelszky,
Zalakaros 2018, and 14.Bxd4! give White the
somewhat bet ter game. He can improve his pieces
with a2-a4-a5, or Nf4.
B22. 8...Nc6 9.Be2 Ne4!
Recommended by Mikhalevski.
The most popular continuation is 9...Bg4
10.dxc5 Bxc5 11.0-0 a6 12.Rc1 Ba7. This is almost
identical with a line I analysed in Chapter 2, only
instead of ...Qd6 Black put ...Bg4 here. However,
this “small” difference tips slightly the balance in
White’s favour after 13.h3 Bf5 14.Bd3 Bxd3
15.Qxd3 d4 16.exd4 Nxd4 17.Rcd1. The point is
that Black cannot keep on eliminating the material
with 17...Nxf3+ 18.Qxf3 since 18...Qc7 19.Ne2
would be horrendous.
9...Be6 10.0-0 Rc8 11.dxc5 Bxc5 12.Rc1
Qe7 is playable. The computer suggests to activate a
piece – 13.Bd3!?, not being afraid of 13...Bg4 in
view of 14.h3 Bh5? 15.g4.
Humans prefer the thematic 13.Nb5, without
big success after 13...Ne4 though.
9...a6 10.0-0 Re8 11.Rc1 cxd4 12.Nxd4 is
slightly better for White. Compare this position to
Chapter 2 – the black bishop is on e7 instead of d6.
That reduces Black’s attacking potential.
10.dxc5
222
10.0-0 Bf6
10...cxd4!? 11.Nxe4 dxe4 12.Nxd4 Qb6= is
a decent alternative.
11.Rc1
11.Na4 Bg4 12.Rc1 (12.h3 Bxf3 13.Bxf3
cxd4 14.exd4 Qd6) 12...cxd4 13.Nxd4 Bxd4
14.Bxd4 Bxe2 15.Qxe2 Nxd4 16.exd4 Re8=.
11...cxd4 12.Nxe4 dxe4 13.Nxd4 Nxd4
14.Bxd4 Bxd4 15.Qxd4 Qxd4 16.exd4 Be6 17.Rc7
Rfc8 with a quick draw in Nakamura-Le Quang
Liem, chess.com INT 2019.
10...Bf6
11.Rc1
Cummings also analyses 11.Nd4 Nxc5
12.Nxc6 bxc6 13.0-0 Bf5 14.Rc1 Re8 as in
Farago-Acs, Balatonlelle 2002, when 15.Ba3 (Ribli)
is not much of an improvement. Play is balanced
after both 15...d4 16.Nb5 cxb5 17.Rxc5 Bg6
18.Bxb5 dxe3! and 15...Ne4 16.Nxe4 Bxe4 17.Bd3
Qa5 18.Bd6 Qxa2.
11...Nb4!
This novelty evens the chances. Cummings
and Mikhalevski mention only 11...Nxc5?!, missing
the obvious 12.Nxd5 Bxb2 13.Rxc5 Ba3 14.Rc2
Nb4 15.Rd2 Nxd5 16.Rxd5 Bb4+ 17.Kf1. Not only
has White an extra pawn, his pieces are also more
active – 17...Qc7 18.Rd4 Qe7 19.Bd3 g6 20.h4 h5
21.g3².
 12.Nd4 Nxc3 13.Rxc3 Nxa2 14.Rc2 Nxc3
15.Rxc3 Qa5=.

223
 Chapter 11. Anti-Queens’s Gambit
Annotated Games

36. DiscoCheck 4 – LoopList 6

40/40 CCRL 22.07.2013

1.Nf3 Nf6 2.c4 c5 3.Nc3 d5 4.cxd5 Nxd5

5.e3 e6 6.Qc2 Be7 7.a3 0-0 8.b3 b6 9.Bb2 Bb7
10.Bd3

Black (temporarily!) bolstered its castling

position, but at the cost of putting its knights on the
last rank. So White strikes from the other wing:

17.b4 cxb4 18.axb4 Qe7 19.d4!

Complex play all over the board! 19...Qxb4

20.Ne4 g6 21.Ra1 threatens, apart from other things,
to win the exchange. Black correctly rejects this
offer, but fails to find the best defence – 19...b5,
restricting White’s bishops.

10...Nf6 19...Nd6? 20.e4 Bh4 21.b5! axb5 22.Nxb5

Rc8 23.Qd2 Ne8

After 10...f5 we should also castle. However,

I would like to mention the game between engines 23...Bxf2 loses to 24.Rg2 Bh4 25.Ba3.
GreKo 10.2-Tucano 2.00, CCRL 2013, in which the
machine tried the speculative pawn march 11.h4?! 24.Bd3
Bf6 12.h5 Nxc3?! 13.Bxc3 Bxc3 14.Qxc3 Rf6
15.Ke2 a6 16.Rh4 Bxf3+?! 17.gxf3 and the open g-
and h-files proved fatal for Black.

11.0-0 Bxf3

The games of the AI AlphaZero showed how

dangerous could be to open the g-file against a white
bishops battery. Both 11...Nbd7 12.Rac1 Qc7
13.Ne4 h6 and 11...Nc6 12.Ne4 Nxe4 13.Bxe4 f5
were balanced.

12.gxf3 a6 13.Kh1 Ra7 14.Rad1 Rd7

15.Be2 Ne8 16.Rg1 Bf6
Black is beyond salvation. The imminent
224
catastrophe on the g-file could be postponed only by
24...g6, when the standard positional sacrifice 25.d5!
exd5 26.e5 should implacably decide the game –
26...Ng7 27.Nd6, when 27...Rxd6 28.exd6 Qxd6
29.Be4 Bf6 is not a fortress – 30.Bxd5 Kf8 31.Qe3!
Bxb2 32.Bb7 Qe6 33.Bxc8 Qxc8 34.Qxb6 is a triple
hit (on b2, d8 and f6).

24...Kh8 25.Rg4 h5 26.Rg2 Qf6 27.Qe3 g6

28.Rdg1 Qg7

29.d5! Qxb2 30.e5 Kg8 31.dxe6 Rxd3

32.Qxd3 Qxe5 33.exf7+ Kxf7 34.Qxg6+ Ke7
35.Qh7+ Ke6 1-0

225
 Chapter 12. Odds and Ends
Step by Step
Most authors offer a full English repertoire. I
adhere to a more practical approach. Some major
openings simply should not be avoided – the King’s
Indian, the Modern Defence, the Dutch, even the
Queen’s Gambit. In this chapter I only discuss how
to transpose smoothly to the main lines.
King’s Indian Defence
1.c4 Nf6 2.Nc3 g6 3.e4
This anti-Grünfeld move order is the only
reasonable way to aspire to the advantage. The
answer 3...d6 4.d4 will give us the KID without any
concessions, but Black has a way to steer the game
into an original track.
3...e5!? 4.Nf3 Bg7
After 4...Nc6 5.d4 exd4 6.Nxd4 Bg7
simplest is 7.Nxc6
7.Be3 Ng4 8.Qxg4 Nxd4 9.Qd1 c5 10.Bd3
might be difficult to win.
7...dxc6 8.Qxd8+ Kxd8 9.Be3 Ng4 10.Rd1+
Ke8 11.Bd4². Our aim here should be to exchange
everything and make a passer. For instance –
11...Bxd4 12.Rxd4 Be6 13.b3 Ke7 14.f3 Nf6 15.Be2
Rad8 16.Rxd8 Rxd8 17.Kf2 Rd2 18.Ke3 Rc2
19.Kd3 Rb2 20.Rb1².
226
5.d4!?
I believe that it is better to have a small
positional advantage than an extra pawn with
passive pieces. However, in practice the pawn eaters
are twice as much, and they score 71%! So have a
look at:
5.Nxe5 0-0!
Or 5...Nxe4 6.Nxe4 Bxe5 7.d4 Bg7 8.Bg5 f6
9.Be3².
6.Nf3
6.d4 Re8 7.Bd3 c5 8.Nf3 cxd4 9.Nxd4 d5
10.cxd5 Nxd5 11.0-0 Nxc3 12.bxc3 Nc6 13.Nxc6
bxc6 14.Be3 (14.Qc2 Be6 15.f4 Qa5 16.e5 f6)
14...Be6 15.Qc2 Qa5 16.Rfc1 also keeps the pawn,
but not an edge – 16...Rab8 17.Bf1 c5 18.h3 Qa3
19.Qd2 a5=.
6...Re8 7.d3 d5 8.cxd5 Nxd5 9.Nxd5 Qxd5
10.Be2 Qd6. White had a clear extra pawn, but no
idea how to convert it in Game 37 Svane-Duzhakov,
Riga 2018.
5...exd4 6.Nxd4 0-0 7.Be3 Re8 8.Bd3! c6
8...d6 gives us a tempo to consolidate –
9.Qd2 c6 10.h3.
9.0-0
 9...Na6
9...d5 10.cxd5 cxd5 11.Bb5 Rf8 12.exd5
Nxd5 13.Nxd5 Qxd5 was tested in
Tomashevsky-Sutovsky, Ohrid 2009. Now 14.Rc1!?
assures White of a huge lead in development since
14...Qxa2 is dangerous – 15.Bc4 Qxb2 (15...Qa5
16.Bxf7+!) 16.Qd3, and Black should give up his
queen for a rook, bishop and a pawn – 16...Nc6
17.Rb1 Qxb1 18.Rxb1 Nxd4 19.Bxd4 Bf5 20.Qe2
Bxd4 21.Rxb7. Although this might be insufficient
for a full point, White can squeeze for many moves.
10.h3!
10.Re1 Nc5 looks shaky, but we cannot rip
dividends – 11.Nf5
Or 11.Nxc6 dxc6 12.Bxc5 b6 13.Be3 Ng4
14.Bf4 Bd4 15.Bg3 Qf6.
11...gxf5 12.Bxc5 d6 13.Bd4 f4.
10...d6
(10...Nc5 11.Nxc6)
11.Re1 Nc5 12.Qc2
227
White has to part with the d3-bishop, but he
preserves pressure down the d-file. The critical line
is 12...Nxd3 13.Qxd3 d5
13...Nd7 14.Rad1 Nc5 15.Qc2 Qa5 16.f3 a6
17.a3 Qc7 18.Bf2 Qe7 19.b4 Ne6 20.Qd2² Be5
21.Nxe6 Bxe6 22.f4 Bg7 23.Nd5.
14.exd5 cxd5 15.Rad1 dxc4 16.Qxc4 Bd7
17.Ndb5
17...Bxb5 18.Nxb5 Qa5 19.Nd6 Re7
20.Qb3². Black is unable to escape from the binds.
For instance:
20...Rb8 21.Re2 Qh5 22.Red2 b6 23.Qc2
Ne8 24.Rd5 Re5 25.Nc8 Rxc8 26.Qxc8 Rxd5
27.Rxd5 Qxd5 28.Qxe8+ Bf8 29.Bh6 Qd1+ 30.Kh2
Qd6+ 31.g3 a5 32.b3 Qe7 33.Qb8 Qd6 34.Qa8 Qc5
35.Be3 Qc7 36.h4±;
20...Ne4 21.Nxe4 Rxe4 22.Qxb7 Rae8
23.Re2 Qxa2? 24.Qxe4+–.
Modern Defence
 1.c4 g6 2.Nc3!?

2.e4 e5 3.Nf3 Bg7 4.d4 exd4 5.Nxd4 Nf6

6.Nc3 0-0 7.Be3 Re8 8.Bd3 transposes to the
previous main line.
Another option is 3.d4 Nf6 4.Nf3 exd4 5.e5
Ne4 6.Qxd4 Bb4+ 7.Kd1 f5 8.exf6 Nxf6

Black has a huge choice here, but none of it

is subject of our book.

Dutch

The best thing you can do when facing the

Dutch is to opt for the main lines and enjoy a small
Robert Ris published an article on this line in edge!
NIC Yearbook 127. His conclusion is that play is Although White has reasonable schemes
unclear, and I agree with him. For instance: with e3 and b3 if Black attempts the Stonewall,
9.Qh4 (9.Bg5 0-0 10.Qh4) 9...Nc6 against the Leningrad he cannot aspire to the
9...d5!? 10.Bg5 0-0 11.Kc1 d4 12.a3 Be7 advantage without d4 – sooner or later. It’s difficult
13.c5 Qd5 14.Nc3 Qxc5 15.Nxd4 Rf7 16.Nb3 Qb6 to trick the opponent with the move order since his
17.Bc4 Be6 18.Re1 Bxc4 19.Qxc4 Qc6 20.Qxc6 first moves are the same against anything.
was drawn in Ellis-Barsallo, ICCF 2017.
10.Bg5 0-0 11.Kc2 Be7!
11...d6 12.Bd3 Bc5 13.a3 Qe7 14.Nbd2 Qf7
15.b4 Bb6 16.Rhe1 Bf5 17.Bxf5 gxf5 18.Re2²,
Tazelaar-Muñiz Pardiño, ICCF 2017.
12.Nc3 d5 13.Rd1 Bf5+ 14.Kc1 Ne4 15.cxd5
Nxg5 16.dxc6 Nxf3 17.Qc4+ Kg7 18.gxf3 Bd6
19.cxb7 Qg5+ 20.Rd2 Rab8÷.

2...Bg7 3.d4

228
 Chapter 12. Odds and Ends
Annotated Games

37. Svane – Duzhakov

Riga 10.08.2018

1.Nf3 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 e5 5.Nxe5

0-0 6.Nf3 Re8 7.d3 d5 8.cxd5 Nxd5 9.Nxd5 Qxd5
10.Be2 Qd6 11.0-0 c5 12.Qc2 Nc6 13.Be3 b6

Now 19.Bd2!² should be in White’s favour –

he can eventually put the bishop on c3. Aronian
chose instead:
19.f5 Qe7 20.Qd2 Bxe3 21.Qxe3 Qe5
21...Nd4 was preferable – 22.fxg6 hxg6
23.Bd1° since the text allows 22.b4², shifting the
focus of the battle. Instead White played:
22.Rab1 and went on to win with an attack
on the kingside.

14...Be6 15.Qb1

14.Rad1!?
15.a3 was also possible. Perhaps White was
afraid of 15...Na5, missing that 16.e5! Qd7 17.Qc3
It is clear that Black has good compensation would neutralise the threat ...Bb3, gaining space in
for the pawn, but on the other hand, White does not the process.
risk much with these two pawns in the centre. His
chance would come if he could open a file on the 15...h6 16.Rfe1
wing – with f4-f5 or b4. Aronian-Grischuk,
chess.com INT 2016, went 14.a3 Bg4 15.Ng5 Be6
15...Bd7!? looks more natural. Without its White clearly put his hope on the manoeuvre
bishop Black still should be able to keep the Be3-d2-c3
balance, but he will lack threats.
16.Nxe6 Qxe6 17.Kh1 Rad8 18.f4 Bd4 16...Rad8

229
 22.Nxd4 Bxd4 23.e5?

Although this move does not change

dramatically the position, it is serious strategic
mistake. White had instead a clear way to improve
his pieces – 23.a3 Qe7 24.Rc1 Qg5 25.Bxd4 Rxd4
26.g3 h5 27.Re3 Qf6 28.h4 Qe7 29.Qd2, and his
better centre plus an extra pawn should give him an
attack. Perhaps White had something similar in
mind, but rushed to put the e-pawn on “the right”
colour first, but forgot that the enemy bishop will
rule over the light squares?!
17.h3
23...Qe7 24.b3?! Qg5 (24...a5!) 25.Bxd4
Rxd4 26.Re3 h5 27.Rde1
White could have save this prophylaxis for a
while – 17.Rd2!?, since 17...Bg4 allows 18.d4.
17...Nd4 18.Bxd4 Bxd4 19.Nxd4 Qxd4 20.Rc2 also
suits White.

17...Kh7 18.Bf1 Re7 19.Bd2 Red7 20.Bc3

27...h4

I like 27...Bf5 28.Rg3 Qe7.

28.Qb2 Rf4 29.Qc1 Rb4 30.Qc3 a5 31.Re4

White has completed the redeployment of his Rdd4 (31...Bf5!) 32.Rxd4 Rxd4 33.Re4
pieces, and he will be waiting for an opportunity to
move forth. He is still threatening nothing, but
Black’s task is not so easy – he has too many
candidate moves to consider.

20...Qf8 21.Qc2 Nd4

This is already a small victory. 21...a5 or

21...Qe7 were good waiting moves, and 21...Nb4!
22.Bxb4 cxb4= deserved attention.
230
 33...Bxh3?

A horrible mistake. Black has played so

many moves with just a positional compensation,
and suddenly he forces the events at the most
unfortunate moment.

34.Rxd4 cxd4 35.Qxd4 Bxg2? 36.Bxg2 h3

37.Qd5 Kg7 38.f4 Qxf4 39.Bxh3 Qg3+ 40.Bg2
Qe1+ 41.Bf1 Qg3+ 42.Kh1 Qf2 43.Qg2 Qe1
44.Qe2 Qh4+ 45.Kg1 Qg5+ 46.Bg2 Qc1+ 47.Kf2
Qf4+ 48.Bf3 Qh2+ 49.Kf1 Qf4 50.Kg2 Qd4 51.a4
Qc3 52.Bd5 1-0

231
 Index of Variations

Chapter 1. 1.c4 c5. Symmetrical with ...g6

7

1.c4 c5 2.Nf3 Nf6 3.Nc3 e6 4.e3 d5 5.cxd5

Nxd5 (6.Bc4 64)
6.Nxd5 64
6.d4 Nc6 (6...cxd4 7.exd4 Nxc3 67) 7.Bd3
Be7 8.0-0 0-0 9.Qe2 (9.Re1 69; 9.a3 70; 9.Be4!? 70)
2.Nf3 (2.Nc3 10) 2...Nf6 (2...Nc6 3.e3 9...b6 71
11) 3.Nc3 Nc6 (3...g6–Ch.5) 4.e3 g6 (4...d5 12) 5.d4 9...cxd4 72
cxd4 6.exd4 d5 7.cxd5 Nxd5 8.Qb3 14
Chapter 4. Symmetrical with ...b6 81
Chapter 2. 1.c4 c5 2.Nf3 Nf6 3.Nc3 e6 27

1.c4 c5 2.Nf3 Nf6 (2...b6!? 84) 3.Nc3 b6

4.e3 d5 (4...a6 31; 4...Nc6 31; 4...b6 33) 4.e4 d6 5.d4 cxd4 6.Nxd4 Bb7 7.Qe2!? (7.f3!? 85)
5.cxd5 exd5 7...Nc6 86 (7...Nbd7 85; 7...g6 85)
6.Bb5+ 33 (6.b3 33) 7...e6 88
6.d4 a6 35 (6...c4 35)
6...Nc6 7.Bb5 a6 37 Chapter 5. Maróczy Bind 99
7...Bd6 39
7...cxd4 40

Chapter 3. Symmetrical with ...e6,

...Nxd5 59

232
 1.c4 c5 2.Nf3 g6 3.e4 Nc6 105
3...Bg7 4.d4 cxd4 5.Nxd4 Nc6 6.Be3 Nf6
(6...Qb6 109) 7.Nc3 0-0 (7...Ng4 110) 8.Be2 d6
9.0-0 Bd7 (9...Nxd4 111) 10.Nc2 112
10.Qd2 115
Chapter 6. Anti-Grünfeld 125
1.c4 Nf6 2.Nc3 d5 3.cxd5 Nxd5 4.Nf3 c5
(4...Nxc3 127; 4...g6 127)
5.e3 Nxc3 6.bxc3 g6 7.h4 129
7.Bb5+ 131
5.e4!? 132
Chapter 7. Anti-Nimzo 153
233
1.c4 Nf6 2.Nc3 e6 3.Nf3 b6 156 (3...d5 –
Ch.11)
3...Bb4 4.Qc2 0-0 5.a3 Bxc3 6.Qxc3 d6 159
6...b6 162
Chapter 8. 1.c4 Nf6 2.Nc3 e6 3.e4 173
3...c5 177
3...d5 4.e5 (4.cxd5 exd5 5.e5 180) 4...d4
5.exf6 dxc3 6.bxc3 Qxf6 7.Nf3 182
7...e5 (7...Nc6 182) 8.d4 183
8.Bd3 185
7...c5 186
7...b6 187
Chapter 9. Anti-Slav and Chebanenko 199
 1.c4 c6 2.Nf3 d5 3.e3 Nf6 4.Nc3 Bg4 5.cxd5 1.c4 e6 2.Nc3 d5 3.e3 (3.d4!) 3...Nf6 4.Nf3
204 (5.Qb3 204) c5 237
4...a6 5.Be2 Bg4 208 (5...Bf5 207; 5...e6 4...Be7 5.b3 0-0 6.Bb2 c5 7.cxd5 Nxd5 238
207) 7...exd5 8.d4 b6 241
5...b5 209 8...Nc6 241
5...dxc4 210
Chapter 12. Odds and ends 245
Chapter 10. Anti-Meran 217

1.c4 Nf6 2.Nc3 g6 3.e4 e5 245

1.c4 c6 2.Nf3 d5 3.e3 Nf6 4.Nc3 e6 5.b3 1.c4 g6 2.Nc3!? 247
Bd6 6.Bb2 0-0 (6...Nbd7 220) 7.Be2 Nbd7 8.Qc2
Re8 221
8...Qe7 223
8...b6 225
8...a6 226

Chapter 11. Anti-Queens’s Gambit 235

234