Antic D Maksimovic B - The Modern French - New in Chess 2012

1
Table of Contents
Cover
Title
Key to symbols
Introduction
Aims of this book
My stylistic choices
Part_1_Chapter_1
Part_1_Chapter_2
Part_1_Chapter_3
Part_1_Chapter_4
Part_1_Chapter_5
Part_2_Chapter_1
Part_2_Chapter_2
Part_2_Chapter_3
Part_2_Chapter_4
Part_2_Chapter_5
Part_3_Chapter_1
Part_3_Chapter_2
Part_3_Chapter_3
Part_3_Chapter_4
Part_4_Chapter_1
Part_4_Chapter_2
Part_4_Chapter_3
Part_4_Chapter_4
2
3
Introduction to the Dynamic Caro-Kann
Introduction to the Dynamic Caro-Kann There are, broadly speaking, three different kinds of serious opening
books enter-ing the market these days. While making no value judgments about them, here they are:
1) The encyclopaedia. The author wants to dissect an entire opening, move by move, ending his or her
variations only when the game enters tablebase territory. Densely packed with confusing variations and
definitely not bedtime reading! It is impossible to fault their work ethic, but the long variations can be artificial
at times, and are not memorable or conducive to nurturing real interest from the reader. That has to come from
within.
Examples: anything by Boris Avrukh, Nikolaos Ntirlis, or Vassilios Kotronias.
2) The BuzzFeed essay. The author wants you to be captivated by their passion and dedication to their
opening. The books carry bombastic titles and inside they are laden with anecdotes, references to creative
processes that occurred in Soviet Ukraine in 1982, and the odd extremely brilliant improvement or line that
merits instant inclusion into your repertoire.
Examples: Steamrolling the Sicilian by Sergey Kasparov; Tiger’s Modern by (na-turally) Tiger Hillarp Persson;
The Flexible French by Viktor Moskalenko.
3) The personal archive. The author has almost certainly assembled the book ex-clusively from their existing
analyses, which may date back a fair bit, but would prefer that you did not know this. The sentence “I prepared
this in 2014 but then GM X played it and made me sad” is almost certain to come up. These books are like the
previous type in that they can be full of unrecognised gems (and they re-pay the effort of sifting through to see
which of their novelties are actually that special) but unlike the previous type, they tend to be written with an
unmistaka-ble undertone of professional rivalry.
Examples: anything by Negi Parimarjan or Milos Pavlovic.
Without intending any malice against any of the mentioned authors (I have learned something from all your
books!) we should note that there are limitations to each of the approaches. My fellow Thinkers Publishing
author Raja Panjwani, in his introduction to his The Hyper-Accelerated Dragon, distances himself from the
“Darwinian angst” characterising the 3rd category of book, and succeeds in perch-ing himself somewhere
between the 2nd and 3rd categories above.plistic fash-ion, while the latter might as well be “Black is better
because his back is to the bar so he won’t be distracted.”
4
Aims for this book
Aims of this book In the book you are now holding in your hand, I intend to straddle not only two, but all three
of the above categories, darting from rigorous theory to anecdote to jealously guarded novelty and then all the
way back. The book is intensely per-sonal, as befits an opening book about an opening I have played from both
sides no less than 60 times over 8 years. Hopefully, this depth of experience has helped me to bring together
these disparate ways of looking at the various lines into one coherent narrative. I have tried to write a one-of-a-
kind book, overflowing with not only novelties but also plans, useful context and passion. The range of
methods is quite astounding: I have used illustrative games (Chapter II.5) and ridiculous computer lines
(subchap-ter IV.3.2), meshed together qualitative and objective evaluations, sometimes for the same position,
and occasionally found some really great lines through trusting my gut feeling above the computer (subchapter
I.5.5). Regarding that last point, I would like to draw your attention to the following game, which provided me
with an early taste of how the human mind can some-times be better than the silicon one.
I Fernandez, Daniel (2367)
J Arkell, Keith (2434)
4NCL 2014
1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Nd7 5. c3 Ngf6 6. Nxf6+ Nxf6 7. Bd3 Bg4 8. Ne2
This position is discussed at slightly greater length in chapter III.3.
8... Bxe2?!
Suffice it to say that this is inaccurate, because of the next note.
9. Qxe2 e6 10. 0-0
10. f4!? exploits Black's lack of a bish-op-pair in ideal fashion: White will force through f5 on the next move.
10... Be7 11. b3
11. f4 g6 12. f5?! is still possible as a piece sacrifice (!) because Black's piec-es are stuck in awkward positions,
but it is nowhere near as strong as on the previous move. 12... gxf5 13. Bxf5 exf5 14. Rxf5 Nd5!?
11... 0-0 12. Bb2 Qc7
5
13. c4?!
Reflecting a serious gap in my under-standing: now Black's regrouping pro-ceeds as if on autopilot.
13... Rfd8
This may be the wrong rook, because of the next note.
14. Rfd1
'Forgiving' Black his previous move. 14. f4 g6 15. g4? was necessary, in or-der to bust out of the positional
head-lock I am in.
14... Rd7 15. a3 Rad8
This is what the computer would term an equal position, but in reality I think Black is already significantly
better.
16. Rac1 Qb8

A) 16... Ne8 was the move Black probably wanted to play, but because of the c-file there is 17. d5!?.
B) 16... g6 was also possible immedia-tely; not only does this create a square for the bishop or queen on g7 but
also it shuts down White's d5-based tactics because there will no longer be Bf5.
17. Qc2 g6 18. Be2 Ne8 19. Rd3 Bf6 20. Rcd1 Qd6!?
The beginning of a truly amazing con-cept, that I didn't understand until he played his next move. The queen is
of-ten found on c7 in the Caro-Kann, but here it takes a swing and winds up on... g7.
6
21. g3 Qf8 22. b4
22. Bf3 Qg7 23. Qd2 is the comput-er's favoured method of playing but af-ter a few more moves, beginning
with 23... h5! it recognises that Black is in fact better. Note that Black can choose his moment to go ...c5, or
avoid it completely.
22... Qg7 23. Qa4
23. Bf3 is the computer's opinion, but first of all, no human wants to give up d4 for free, and secondly, it just as
quickly finds a line ending in Black's fa-vour: 23... Bxd4 24. Bxd4 Rxd4 25. b5 cxb5 26. cxb5 Rxd3 27. Rxd3
Qa1+!
Setting up tactics based on ...Qd5 forks. 28. Kg2 Qf6 29. Rxd8 (29. Bxb7 Qf5!?) 29... Qxd8 30. Bxb7 Nd6 31.
Qc6 (31. Bc6 Nxb5! is the point) 31... Qb6 32. Qxb6 axb6 33. Bc6 Nc4? Black has a superior end-game: White
should probably sacrifice the a3-pawn right now for king activity but his king will have no way in and he will
be restricted to grovelling for a draw while Black manoeuvres round with his extra pawn.
23... a6!
7
Ensuring that White cannot make his pawn break without losing the d5-square.
24. b5 axb5 25. cxb5 cxb5 26. Qxb5 Nd6 27. Qb4 Nf5 28. Bf3
White has pulled out all the stops in an effort to get counterplay against Black's b7-pawn and to open the posi-
tion. However, it is, properly speaking, not enough for equality. .
28... Nxd4?!
This is hasty and throws away a per-fectly good bind that gave winning chances. Of course, the machine doesn't
'get' this and thinks the move is fine, but it is possible to defuse all the tricks it finds for Black in the subse-
quent notes.
A) 28... Ne7!?? was one possibility; Black occupies the d5-square, for which he has worked so hard, with his
knight, and then regroups his queen and possibly one rook, so as to attack the d4-pawn from a more active
angle.
B) 28... Qf8? was also good enough to maintain the edge; the idea is to play ...b6 next and hence stop that pawn
from being a target.
29. Kg2!
29... Nc6
A) 29... b5 30. Bxd4 Bxd4 31. Qxb5 e5 32. a4=
B) 29... Nc2!? would have led to an entertaining set of desperado moves: 30. Rxd7! Rxd7 31. Qf4 Ne1+ 32.
Kh3! Rxd1 33. Bxf6!
8
33... e5!! 34. Bxe5 (34. Bxg7 exf4 35. Bxd1 Kxg7 36. gxf4 Kf6?) 34... Nd3! 35. Qe3! Nxf2+!? (35... Nxe5 36.
Bxd1 Qf6=) 36. Qxf2 Qh6+ 37. Kg2 Rd2 38. Be2 Ra2?
30. Rxd7 Rxd7 31. Qf4 e5?!
31... Rxd1 was the last try: 32. Bxf6 g5 33. Bxg7 gxf4 34. Bxd1 Kxg7 35. Bf3!=.
32. Qa4 Rxd1 33. Qxd1 Qf8 34. Qb3
White might even have a little some-thing now, but I didn't feel like pushing my luck, and obviously while the
pawn deficit remains it can all still rebound in my face. A close shave and a lesson well learned, even if it might
not seem so superficially. 1/2-1/2
The grand old English maestro of the Caro-Kann clearly had a thing or two to teach me, and even to teach the
computer. We will examine the line more closely in chapter III.3. This game was a great learning experience for
me: never underes-timate the power of understanding, especially in simple positions, or of the role of
momentum in chess.
Momentum? Is this guy mad? Is this like some New Age, healing-crystal kind of chess analysis? Will he soon be
talking about the hidden orchestra on the d3-square or the feng shui benefits of having your knights face each
other at the start?
Not quite (though you will notice that a lot of good players have a knight-routine, and some have a bishop-
routine too.) Momentum is a very real phenomenon that has most visible effects when, for instance, only one
side has play, or a gambit is in the final stages of being neutralised.
9
Momentum is very much a qualitative concept, and as (to my mind as a seasoned computer analyst) a fairly
uncontroversial example of the same, it has a place in a chess book. However, that being said, it is important to
recognise and reinforce the line between qualitative observation (Only White can be better because of his
control of the c5-square) and subjective nonsense (I would prefer to play White because he has a knight against
a bishop).
The former attempts to balance out actual pros and cons, albeit in simplistic fash-ion, while the latter might as
well be Black is better because his back is to the bar so he wont be distracted.
10
My stylistic choices as a writer – and how you should use them as a reader
There is, however, a legitimate use of bias in writing, which is a second-hand one: deciding where to give
alternatives. For instance, I dislike accepting gambits, if I have another choice of comparable quality; or
entering IQP positions. Hence for both the completeness of my own repertoire and the retention of academic
integrity in this book, I have tried very hard to give alternatives to doing either of those things.
Likewise, I very much enjoy messy positions, particularly when the pieces left dangling en prise are not my
own, or they exist merely on a computer screen. Therefore, to avoid giving only lines which are ‘easy to
recommend but hard to play’ I have consciously corrected for my own tendency and given alternatives in many
such positions.
Due to the addition of such alternatives, as well as the historically relevant lines and annotated games, there are
a number of chapters which are surplus to requirements for forming a repertoire (II.5, III.4, IV.4) and two
others that can be omitted if you make certain choices (III.2, IV.3). However, I do think they are worth the time
of at least a majority of Caro-Kann players, else I would not have written them.
One place where I may have been lax in the provision of alternatives is in the use of pronouns: he/she/they and
the associated genitive forms his/her/their. Please do not take offence if I have lazily written just the male
forms: I wouldn’t take offence if a different author wrote ‘her’ all the time, and after all this is primarily a chess
book!
So as not to alienate the rare White player who has picked up this volume, I should point out that there are some
quite interesting White bits in here, even if I don’t consider them to get any objective advantage (IV.1, III.3,
parts of I.1), in addition to a number of strong replies that may prove useful to you: the act of explaining why
something hasn’t been chosen for Black is perfectly sufficient as a guide to playing the same rebuttal from the
White side.
Administratively, I have tried to organise the book into Parts, as outlined in the contents. References to other
places within the book are always within the same Part, so that they are completely self-contained. So, for
instance, if in II.3.2 you saw a reference to ‘line 5.3’, this means II.5.3. The numbering system is
straightforward: the Part number is followed by a chapter number and then a subchapter number. There are no
divisions within subchapters, and certainly no confusing codes like ‘B235b294’.
It remains to let you get on with studying my favourite opening. I don’t promise equality (largely a nebulous
concept anyway), or that the book is comprehensive (exhaustive opening books are a myth, though I have tried
and some try even harder), but I do promise you an interesting ride!
Daniel Fernandez
Budapest
February 2018
11
Part I:
Less common or critical attempts
Chapter 1
Early minor lines
1.e4 c6 2.--
Chapter guide
Chapter 1 – Early minor lines
1.e4 c6
1.1) 2.Ne2 and others
1.2) 2.Nc3 d5 minor lines (3.Qf3!?, 3.Nf3 Nf6 4.e5 Ne4 5.Be2!?)
1.3) 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.Be2 Bf5!
1.4) An extra option!
1.5) 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.d4
1.6) 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.Ne2 Qb6 6.d4 e6 7.Ng3
Introduction to Chapter 1
Every book needs to tackle its loose ends, either at the start or at the end. As you can see, I have gone for the
former approach, but I have split the material over three chapters. In the present one, I begin by tackling the
“deservedly minor yet still noteworthy” options, and I am aided in doing this by the fact that after 1...c6 it is
possible to ‘premove’ 2...d5.
It is possible to spend rather too much time on this kind of thing; therefore, I shall not preoccupy myself with
2.Na3, 2.h4 and so on. Instead, after some remarks on 2.b3 and 2.Ne2, which have been tried on occasion by
some rather good players, we turn our attention to a relatively major system, which is characterised by White
playing his knights to c3 and f3 in one order or another.
12
One of my most important findings in the Two Knights’ System was that White’s often-neglected option 5.Be2
is actually rather dangerous. I believe myself to have found equality against it, but the task was not easy, and
supporters of both colours should take note.
Line 1.6) then takes us a bit further along the path to the main line which will form the basis of the next chapter,
turning off with 7.Ng3 rather than the now-famous 7.Nfg1...
1.1) 2.Ne2 and others
1.e4 c6 2.Ne2!?
An interesting line, favoured by quite a few top players. It seems to me the rationale behind the move is to try
and get some sort of Advance or Tarrasch French position a tempo up (because Black has taken two tempi to
play ...c5). We are not going to trouble ourselves with 2.Na3, 2.Nh3 and other random things, but there are two
quite minor moves that are nevertheless somewhat intelligent and have been recommended in books before.
A) 2.b3!? d5 3.Bb2 dxe4
Position after: 3...dxe4
A1) 4.Nc3 is better, but more conciliatory and hence less popular. It should be met by a normal developing
scheme: 4...Nf6 5.Nge2 Bf5 6.Ng3 e6 7.Qe2 Nbd7 8.Ngxe4 Be7= for instance and now Black has the
option to initiate play on the queenside with ...Qc7 and either ...a5 or ...b5.
A2) 4.Qe2 Nf6 5.Nc3 Bg4! Forcing White to walk into a later ...Nd5. 6.Qe3 Bf5 7.h3 Nd5!? (7...Nbd7 as
recommended by Houska is also fine- that text follows the game Tu, H – Ni, H Ho Chi Minh City 2014,
which Black won in thematic fashion.) 8.Nxd5 cxd5 9.Ne2 Nd7³ It is not even clear to me how White
intends to try and regain the pawn.
B) 2.g3 d5 3.e5 c5 Now White’s f4-systems will be dealt with under 2.f4 d5 3.e5 in chapter 3, but independent
is 4.Bg2 Nc6 5.d3!? (5.Nf3 Bg4³) 5...Nxe5
13
Position after: 5...Nxe5
I am reluctant to recommend ...g6 as then f4! leaves our repertoire; or ...e6 as it closes in the c8-bishop.
However, either move is of course playable if Black does not wish to allow a repetition. 6.Qh5 Qc7 7.Bf4
Qa5+ 8.Kf1 (8.Nc3 Ng6=; 8.Bd2 Qc7 is an immediate draw by repetition) 8...Ng6 9.Bd2 Qc7 10.Qxd5 Nf6
11.Qb3 Rb8= Black has fully equalised.
C) 2.Nf3 d5 3.exd5 cxd5 4.Ne5
Position after: 4.Ne5
This is a minor third line. White admits that the position is equal, but tries to introduce some imbalance into it
by winning the bishop pair. The move has been chosen by Carlsen. Black can react in various ways, including:
14
4...e6 5.Bb5+ (5.d4 Nc6 6.Bb5 a6 7.Nxc6 Qc7 8.Ba4 Bd7=) 5...Nd7 6.d4 a6 7.Bd3 (7.Bxd7+ Bxd7 8.0-0
Rc8=) 7...Nxe5 8.dxe5 Ne7 9.0-0 Bd7=.
2...d5
2...e5!? is also good, for once, but it does not belong in a Caro-Kann book.
3.e5
Position after: 3.e5
3...c5
This is the main line, and I find it to be playable but not in the conventional way.
A) 3...f6 4.d4 g6 5.f4ƒ might be computer-sanctioned for Black, but isn’t clever
B) 3...d4!? Leading to interesting play, and most importantly not giving White the kind of game he was after.
The key thing to remember is to play ...c5, if possible, and attack the e5-pawn; typically with the queen but
maybe the knight or bishop. 4.b4 (4.c3 c5 5.b4 Qd5!=) 4...a5 5.Bb2 axb4 6.Nxd4 Qd5= Black was doing well
and eventually won in Abergel, T – Svane, R Germany 2014.
4.d4 Nc6!
4...cxd4?! 5.Nxd4 Nc6 6.c4 e6 7.Nc3 Bc5 (7...Nge7?!
15
Position after: 7...Nge7?!
8.Ndb5! is an opening trap known at least since Hillarp Persson, T – Willemze, T Hoogeveen 2007, which
continued 8...Nxe5 9.Qa4 N7c6 10.cxd5 exd5 11.Nxd5 Bd6 12.Be3‚) 8.Nxc6 bxc6 9.cxd5 cxd5 10.b4 Bf8
Reached by Keith Arkell in 3 of his Black games, and the position certainly appears robust at first glance, but
there are concrete problems with Black’s position.
5.c3 Bf5
A) 5...cxd4 6.cxd4 Bf5 7.Nbc3 e6 8.a3 Nge7 9.Ng3 Bg6 10.h4! h6 11.h5 Bh7 12.Bd3 Bxd3 13.Qxd3² Black
is very short of ideas, even if he also has relatively few immediate problems. If he could swap off one pair of
knights he would be fine.
B) 5...Bg4 was chosen by Anand and is a perfectly valid move too.
6.dxc5!?
6.Ng3 Bg6 7.h4 e6! is fine for Black, e.g. 8.h5 Bxb1 9.Rxb1 cxd4N (9...h6² Polster, W – Solozhenkin, E ICC
INT 2000) 10.Bb5 Nge7!?„
6...Nxe5!
16
Position after: 6...Nxe5!
To my mind, the only thing one has to remember about 2.Ne2.

6...e6 is played far more often but runs into concrete problems.
7.Nd4
7.Ng3 Bc8 8.b4 Nf6= It is almost sacrilege to end a variation on move 8, but here there is really not a lot to say
until more games are played. To my mind, Black is completely fine.
7...Bd7 8.Qh5 Qb8 9.Bf4 Nd3+ 10.Bxd3 Qxf4=
17
Position after: 10...Qxf4=
Black had equalised in Sedina, E – Botta, G Lugano 2011.
1.2) 2.Nc3 d5 minor lines (3.Qf3!?, 3.Nf3 Nf6 4.e5 Ne4 5.Be2!?)
1.e4 c6 2.Nc3
Undoubtedly the most mainstream of the ideas I have filed into chapter 1, and it is with this move that we shall
chiefly occupy ourselves in this chapter. As (almost) always, Black plays 2...d5.
2...d5
Position after: 2...d5
3.Nf3
The famous Two Knights system.

3.Qf3 An interesting minor system, noticed and then developed further by my friend FM Marcus Harvey. The
main point is that after ...d4 White plays Bc4 and Nce2, and we obtain some kind of strange Closed Sicilian.
3...dxe4 (3...d4 4.Bc4 e6 5.Nce2 c5 6.Qg3 is not all that comfortable for Black, particularly if you have never
seen it before or played the Sicilian.) 4.Nxe4 Nd7
18
Position after: 4...Nd7
I am aware that this is the most boring recommendation I could have made- not trying to treat Qf3 as anything
different but just continuing with the normal scheme of development. White will find nothing better than d4 and
c3 setups, which will lead us right back to the 5.Qe2 lines in Part III. One independent try: 5.b3!? Ngf6 6.Bb2
Nxe4 7.Qxe4 Nf6 8.Qe5!? Else ...Bf5. 8...Qd5 (8...g6 9.0-0-0 Bg7 10.Re1ƒ has the potential to be slightly
uncomfortable) 9.Nf3 Qxe5+ 10.Nxe5 Bf5=
3...Nf6
White players of this system tend to enjoy their bishop pair (the 3...Bg4 lines) and it is not clear to me how
good the small centre is if White refuses to go d4 himself.
3...Bg4 4.h3 Bxf3 5.Qxf3 e6 6.Qg3 Posing small and strange problems to Black’s development. I played in
more or less approved fashion: 6...Nf6 7.d3 b5 8.a3 g6 but after 9.Be2 White was more comfortable and
eventually won a topsy-turvy encounter in McShane, L – Fernandez, D London 2015.
4.e5 Ne4
19
Position after: 4...Ne4
5.Be2!?
One of the biggest surprises of this book was just how tough it was to crack this unassuming White 5th move.
A) 5.Nxe4 dxe4 6.Ng5 Qd5= leaves Black with an easy life as in Li, R – Smirnov, A Saint Louis 2017; White
is virtually compelled to release the central tension with d4 or d3.
B) 5.d3 Nxc3 6.bxc3 c5= leaves White with a position where d4 would be a waste of time, but it isn’t quite a
KIA either.
5...e6?!
Appears most solid, and it is useful to examine this first, for the lines that come later.
5...Nxc3 is logical, but my feeling is this does not actually equalise. 6.dxc3
20
Position after: 6.dxc3
A) 6...Bg4?! 7.h3 Bxf3 8.Bxf3 e6 9.0-0 Nd7 10.Qe2² was the starting point of my previous investigations;
however I concluded that Black seems to lack concrete ways to play and the momentum is clearly with White.
Both f4 plans and c4/cxd5/c4 plans are possible, and both are annoying.
B) 6...g6 7.h4! Black now essentially has to commit to an h-pawn move, and neither of them is particularly
desirable. 7...h5 (7...h6?! 8.Nd4!?‚; 7...Bg7 8.h5 Bg4² is ultimately nowhere near as terrible as it looks, but
that isn’t saying much.) 8.Nd4 Bg7 9.e6 0-0! 10.exf7+ Rxf7 11.Be3!² Either Nf3-g5 or simply g4 will follow
soon.
6.0-0 Be7
6...c5 7.Nb5! a6 8.d3!² gets an advantage for White, but only because Black is only attacking b5 once...
7.Nb1!
21
Position after: 7.Nb1!
Black is faced with a slightly awkward choice.
7...0-0
A) 7...f6 8.d3 Nc5 (8...Ng5 9.Nxg5 fxg5 10.c4²) 9.exf6 Bxf6 10.d4² Arribas Lopez, A – Solaesa Navalpotro,
L Madrid 2015.
B) 7...h6 8.d3 (8.h4?!³ might appear to be an interesting gambit line but is met by the geometrically appealing
...Qb6! threatening ... Ng3 and hence obligating White to play d4.) 8...Ng5 9.Nfd2 0-0 10.c4 f6 11.f4 Nh7
12.Nf3²
8.d3 Ng5 9.Bxg5
9.Ne1 f6!=
9...Bxg5 10.c4²
22
Position after: 10.c4²
The strategic battle is by no means over and Black has a big trump in the form of the bishop pair, but his lack of
space is very noticeable.
1.3) 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.Be2 Bf5!
1.e4 c6 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.Be2!? Bf5!
Position after: 5...Bf5!
Almost by a process of elimination. In the main line, White pushes this bishop back, and it is instructive to see
23
why it is better on d7 than c8.
6.Nh4!
A) 6.Nb1 e6 now fails to bother Black, who will get a position from the French (...Nc5-d7, ...c5, ...Nc6) but
with the light-squared bishop outside. For instance: 7.0-0 Be7 8.d3 Nc5 9.Be3 Ncd7 10.d4 (10.c4 dxc4
11.dxc4 c5=) 10...0-0 11.Nbd2 c5 12.c3 Nc6=
B) 6.d3 Nxc3 7.bxc3 e6 8.Rb1 b6 leads to considerable tempo gains but is ultimately harmless. White should
now continue with the f4-f5 push, or else his position makes no sense.
Position after: 8...b6
9.0-0 Be7 10.Nd4 Bg6 11.f4 c5! 12.f5! cxd4 13.fxg6 hxg6 14.cxd4 0-0 (14...Bg5!? 15.c4! Bxc1 16.Rxc1 Qh4
leads to a mess, but White is better in a standard game between good players...) 15.Be3 Bg5 16.Bf2=
24
Position after: 16.Bf2=
White’s queenside is rather weak and Black has more strategic ideas at his disposal, so even as a fan of the
slow effect of the bishop pair, I would prefer to be Black.
C) 6.0-0 e6 7.d4 Be7 8.Nxe4 Bxe4 9.Bd3 0-0 10.Be3 Bxd3 11.Qxd3 Nd7= is not particularly critical and
Black will get in ...c5 next.
6...Bd7
Position after: 6...Bd7
25
7.0-0
White makes a useful move and Black cannot do the same without ...e6.
A) 7.d3 Nxc3 8.bxc3 e6 9.Nf3 c5=
B) 7.Nxe4 dxe4 8.d3 e6 9.g3 exd3 10.Qxd3 c5!=
C) 7.f4 Qb6 8.Nxe4 dxe4∞ is strategically quite confusing, but Black is not worse. A good plan for him is
...Na6 and ...0-0-0.
7...e6 8.Nf3
8.g3 Be7 9.Ng2 was the enigmatic choice of one 2600+ player but after 9...Nxc3 10.bxc3 c5„ Black had set up
enough counterplay in Fedorchuk, S – Kovalenko, I Bastia 2014.
8...c5
Position after: 8...c5
8...Be7 9.Nb1 leads to the same problem as 5...e6; it is not at all clear to me that ...Bc8-d7 can be considered a
meaningful extra developing move.
9.Nb1
A) 9.Nb5 a6 Notice that the d7-bishop also attacks b5, so now d3 is just nonsense. 10.Na3 Ng5 11.Nxg5
Qxg5 12.d4 Qh4=
26
Position after: 12...Qh4=
Black has a good French-type position.

B) 9.Nxd5!? exd5 10.d3 Nxf2 (10...Nd6?! 11.exd6 Bxd6 12.d4²) 11.Rxf2 Be7 12.d4 Nc6 13.c3 0-0= Black’s
counterplay against the centre (...Qb6, ...Rac8, ...Bg4 etc) arrives in time.
9...Ng5
9...Be7? 10.h4! works for tactical reasons: 10...Bxh4 11.Nxh4 (11.d3?! Bxf2+ 12.Rxf2 Nxf2 13.Kxf2 f6∞)
11...Qxh4 12.d3 Ng5 13.g3 Qh6 14.f4 Qh3 15.Rf2!! Qxg3+ 16.Rg2 Nh3+ 17.Kf1 Qh4 18.Qe1± Black›s two
extra pawns do not help him extricate his knight.
10.Nxg5 Qxg5
27
Position after: 10...Qxg5
11.d4
11.f4 Qd8 is playable too, but White should not go without playing d4, as that square can become weak
otherwise. 12.d3 Be7 13.c4 dxc4 14.dxc4 Nc6=
11...Qd8
11...Qh4 does nearly force c3, but the time wasted may be regretted later. 12.c3 Nc6 13.Be3 cxd4 14.cxd4 Qd8
15.Nc3²
12.c4!
12.c3 a6! (12...Qb6 13.dxc5! is a French motif worth bearing in mind, and my feeling is it gives White an edge
here too) 13.Be3 cxd4 14.cxd4 Bb5= This equalises without much trouble.
12...dxc4
12...Nc6!? 13.cxd5 (13.dxc5!? d4 14.Nd2‰) 13...exd5 14.Be3
28
Position after: 14.Be3
This is fine for Black, but you must be prepared to play IQP positions. With the way I have chosen to write this
book (and indeed the way I play myself), I shall avoid giving such setups as the main choice, even if it would
save me quite a bit of work. Suffice it to say that White has substantial positional pressure, which can be
augmented by f4-f5 at the right time, unless Black is very careful, and I do not have the intuition for how to do
this. What follows is merely a best guess. 14...Qb6! Black needs to force through ...d4. (14...Be6 for instance
will probably end up sliding into an inferior position. 15.Nc3 cxd4 16.Bxd4 Rc8 17.Rc1 Be7 18.Bb5 0-0
19.Bxc6 Rxc6 20.f4ƒ) 15.dxc5 Bxc5 16.Nc3 d4! 17.Nd5 Qd8!
Position after: 17...Qd8!
29
18.Qc1!? Now Black equalises after some trials. (18.Qc2!? Be7 19.Bd2 0-0 20.Bd3 g6 21.Nxe7+ Qxe7 22.f4‚
could also be vaguely worrying)
A) 18...b6?! 19.Bg5‚ White has a lot of potentially unpleasant tactical possibilities (Bb5/a6, Bf6, a3/b4...)
B) 18...Be6 19.Nf4 (19.Qxc5 Qxd5 20.Qxd5 Bxd5 probably equalises for Black) 19...Qb6 20.Nxe6 fxe6
21.Bf4∞ Each of the two bishops functions better than Black’s single one on c5. The availability of ideas like
Bc4 and Qd1-g4 means that I am deeply skeptical of the computer’s “0-0 0.00.” Maybe if I got this in a game
I would castle long!
C) 18...Be7 19.Bf4 0-0 20.Qd2 Bf5 (20...Re8 21.e6!ƒ) 21.Nxe7+ Qxe7 22.Bb5 Qb4=
Position after: 12...dxc4
13.Be3
A) 13.Bxc4 Nc6! Not being afraid of ghosts. 14.d5 Nxe5 15.dxe6 (15.Bb3 c4 16.dxe6 Bxe6 17.Ba4+ Nc6³)
15...Bxe6 16.Qa4+ Bd7 17.Bb5 Bxb5 There is also ...Bd6. 18.Qxb5+ Qd7 19.Qe2 Qe6 20.Nc3 (20.Re1
Nd3!) 20...Nc6³ Black will complete his development with ...Bd6 and then determine which side is better for
him to castle on.
B) 13.Nc3 is a little slow, especially since White is not getting in d5. 13...Bc6! 14.Be3 cxd4 15.Bxd4 b5
30
16.a4 (16.Bf3 a6 17.a4 b4 18.Ne4 Be7 is completely harmless) 16...b4 17.Nb5 a6 18.Nd6+ Bxd6 19.exd6
Qxd6 20.Bxc4 Nd7 21.Re1 (21.Qg4 Ne5! 22.Bxe5 Qxe5 23.Rfe1 Qf6 24.Bxe6 0-0=) 21...0-0 22.Qg4 e5=
White’s bishop pair gives compensation for the pawn, but not more.
13...Bc6
Unexpectedly bringing about a Paulsen Sicilian setup.

13...cxd4!? I have consciously tried in this book to give options other than acceptance of gambits, even if the
acceptance is actually okay. Here, for instance, the position after 14.Qxd4 Nc6 15.Qxc4 Nxe5 16.Qc3 (16.Qe4
Bd6=) 16...Nc6 17.Rd1² looks slightly tricky to either exhaustively analyse or play in practice, even if it is
objectively fine.
14.Bxc4
14.dxc5 Nd7 15.Nd2 Bxc5 16.Bxc5 Nxc5 17.Nxc4 Qxd1 18.Rfxd1 Ke7= does not seriously trouble Black.
14...cxd4 15.Bxd4
Now we shall give two lines, one of which is risky and entails accepting a pawn sacrifice; the other one is
relatively simple and safe.
31
Position after: 15.Bxd4
15...b5!
‘Simple and safe.’

15...Be7!? – see chapter I.4.
16.Bb3
16.Be2 limits the bishop’s prospects and doesn’t work as well with f4-f5. 16...Be7 17.Qd3 (17.Nc3 b4 18.Nb5
0-0= is harmless; Black can play ...a6 next or even ...Bxb5 and ...Qd5.) 17...0-0 18.Rd1 Suggested as a test by
our silicon colleague. Now 18...b4 brings about a structure that I was very fond of in my 1.d4 b5 days. There
could follow: 19.f4 (19.a3 a5 doesn’t change much) 19...Nd7 20.Nd2 Bd5 21.Ne4 (21.Bf3 Nc5=; 21.Rac1 Rc8
22.Bf3 Nb6=) 21...Rc8 22.Qe3 Nb6=. Black’s strong centre serves as insurance and allows him all kinds of
plans, including ...Rc7/...Qa8, but also ...Bc4, or ...Bxe4 and ...Nd5.
32
Position after: 22...Nb6=
16...Nd7
17.Qc1!
A) 17.Nc3 Bc5= is absolutely nothing.

B) 17.Qg4 h5 18.Qf4 Bc5 19.Rd1 Qc7= is a little bit more dynamic but Black is still fine, and his short-
castling construction is not in serious doubt.
33
17...Rc8!
17...Qc8!? is worth considering, but may just be a bit slow after f4. 18.f4! (18.Qe3?! Qb7! 19.f3 Now f4
essentially won’t be played. 19...Be7 20.Nc3 b4!?=) 18...Qb7 19.f5 (19.Rf2 Be7=) 19...Bxg2 20.fxe6 (20.Rf2
Bd5 21.fxe6 Bxe6 22.Bxe6 fxe6 23.Qf4 0-0-0∞) 20...fxe6 21.Rf2 Bd5 22.Qf4 Bxb3 23.Qf7+ Kd8 24.axb3
Qe4 My analysis indicates that Black is alright, but it would be terribly remiss of me to advocate this as my
main suggestion.
18.Qe3
18.Qf4 Bc5=
18...Bc5!
Anyway! This is quite a significant point.
19.Bxc5 Nxc5 20.Qxc5
Else just ...Qb6 with equality.
20...Bxg2
Position after: 20...Bxg2
21.Qxc8
21.Qe3?! Bxf1 22.Kxf1 Qb6! Now White must agree to have his pawn structure crippled, because the
alternatives are worse. 23.Nc3 (23.Qd2 Qc6µ; 23.Qxb6 Rc1+µ) 23...Qxe3 24.fxe3 Rc5=
21...Qxc8 22.Kxg2 Qc5 23.Nc3 Qxe5∞

34
The mess continues, but with material level one of the main practical considerations becomes king safety, on
which grounds one should probably prefer Black, especially if he can trade a pair of rooks.
1.4) An extra option!
1.e4 c6 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.Be2!? Bf5! 6.Nh4! Bd7 7.0-0 e6 8.Nf3 c5 9.Nb1 Ng5 10.Nxg5
Qxg5 11.d4 Qd8 12.c4! dxc4 13.Be3 Bc6 14.Bxc4 cxd4 15.Bxd4 Be7
Position after: 15...Be7
This is the ambitious choice, even if it doesn’t look that way. In the two main lines (18.Ne4 and 18.Rfe1) White
must either sacrifice a pawn or accept a sacrifice of a pawn, and such imbalances favour the brave.
16.Nc3 0-0 17.Qg4 Nd7
35
Both sides have made the obvious developing moves and now White faces a real choice. Which rook, where?
Or something else entirely? I have tried to analyse the most obvious options.
18.Ne4!
Critical.
A) 18.Rfe1 Rc8 (18...Nb6 19.Bd3!² prevents the idea Black had, which was ...h5 with equality.) 19.Rad1
Position after: 19.Rad1
36
19...Ba4!? A tactic well worth remembering, not that any of us have a problem with that since the Qatar open
2016 where Carlsen played a similar shot in the tiebreaks... 20.Nxa4 Rxc4 21.b3 Rc6 22.Bxa7!? (22.Nc3 a6
23.Ne4 Qc7=) 22...Qa5 Black gets good compensation in all lines, or else his pawn back. The following lines
are illustrative and obviously don’t need to be remembered. 23.Bd4 (23.Qe2!? Rd8 24.Bd4 b5 25.Nb2 Nf8©
With the hidden tactical threat of ...Rxd4 and ...Rc2! Black has excellent compensation. Next he will double
rooks on the c-file and consider .. .Ng6 or ...Bb4/a3.) 23...b5 24.Qf3!
Position after: 24.Qf3!
A wrinkle, protecting b3 and thus preventing ...Qxa2. 24...Nb8 25.Nb2 (25.Nc3 Rd8 26.Ne2 Qxa2=) 25...Rd8
26.Rf1 (26.Qe3 Qa3 27.Nd3 Rc2= will take back the pawn under reasonable circumstances) 26...Rcc8 27.a4
(27.Qb7 Nd7 is the last way White can delay the threat of ...Qxa2 rather than deal with it, but now time is
truly up. 28.a4 Rb8 29.Qa7 Qxa7 30.Bxa7 Rb7 31.Bd4 bxa4 32.bxa4 Nb8= The pawn will come home in the
next few moves.) 27...Qb4
37
Position after: 27...Qb4
28.Qe3 (28.Qb7 bxa4 29.Qxb4 Bxb4 30.bxa4 Nc6 31.Bb6 Rxd1 32.Rxd1 Ra8!=) 28...Nc6! 29.Bb6 Rxd1
30.Rxd1 bxa4 31.Nxa4 Bg5! 32.Qxg5 Qxb3=
B) 18.Rad1
Allowing Black to gain tempo on e5, and thus a worse order. 18...Qa5 19.Rfe1 Bc5! The trade Black should
be making on positional grounds. (19...Rfd8? 20.Bxe6±; 19...Qb4?! 20.Bf1 Bc5 21.a3! Qxb2 22.Na4 Qxa3
23.Nxc5 Nxc5 24.Be3±; 19...Nb6!? 20.Bd3 Rfd8 21.a3 g6= is fine too) 20.a3 Bxd4 21.Qxd4 Qb6=
38
18...Qb8!
Position after: 18...Qb8!
Threatening to take on e5 and not being afraid of ghosts.
19.Nd6!
19.Nf6+ Kh8
Position after: 19...Kh8
A) 20.Nxd7 Bxd7 21.Bd3 Rd8 22.Qe4 g6=

39
B) 20.Bd3 gxf6 (20...Bxf6?! 21.Qh3 g6 22.exf6²) 21.Qh4 f5 22.Qxe7 Rg8 23.g3 Rg4 (23...Qf8!? 24.Qxf8
Nxf8 25.Be2 Ng6 26.f4 Be4= looks solid too; Black can continue with ...Rgd8-d7 and ...Ne7-d5.) 24.Bc3
Rg7= Black continues with ...Nb6-d5 (this is why the d4-bishop was driven back to c3!) and obtains good
counterplay on the kingside.
C) 20.Qh3!? gxf6!? leads to a long forced variation. There was nothing wrong with ...h6 either. 21.exf6 e5
22.fxe7 Rg8 23.f3 exd4 24.Bxf7 Qf4 25.Bxg8 Rxg8
Position after: 25...Rxg8
Practically speaking, this position seems good for Black. But the computer is able to ignore the mounting
threats on the g-file and long diagonal to reach a two rooks versus queen ending. 26.Rae1 (26.Kh1 Nf6!³ This
knight will come to e3 and cause no end of headaches for White’s position in general as well as his defense of
the e7-pawn.) 26...Ne5! 27.Qh5 (27.Kh1 Qf6! Threatens to scoop up the pawn, and after 28.Qh5 Ng6!³
White finds himself in tactical trouble on the long diagonal.) 27...Nxf3+
40
Position after: 27...Nxf3+
28.Qxf3 (28.Rxf3 Bxf3 29.e8=Q Bxh5 30.Qxh5 d3=; 28.Kh1 Qd2! 29.Rg1!! Nxg1 30.Qe5+ Rg7 31.e8=Q+
Bxe8 32.Qxe8+ Rg8 33.Qe5+=) 28...Bxf3 29.e8=R Rxe8 30.Rxe8+ Kg7 31.Rxf3 Qc1+ 32.Rf1 Qxb2= Black
draws with a little care.
19...Bxd6 20.exd6 e5
Position after: 20...e5
21.Bc3
41
21.Be3 Qxd6 22.Rad1 Qg6=
21...Qxd6 22.Bb4!
22.Rad1 Qg6=
22...Nc5 23.Rad1 Qe7 24.Rfe1!
A) 24.Bxc5 Qxc5 25.b4 Qe7 26.b5 Be8= The tricks have come to an end; White probably has compensation
right now but needs to cash in quickly before it evaporates.
B) 24.f4 a5! 25.Bc3 Nd7 26.fxe5 Nxe5 27.Qg3 Rae8= White will be unable to stop ...Kh8 and ... f6
consolidating. (27...Qc5+ 28.Rf2 Qxc4 29.Bxe5 g6 is playable but not clever.)
24...a5!
Position after: 24...a5!
Else Black may actually be in some trouble.

A) 24...Kh8 25.f4²
B) 24...Rad8 25.Rxd8 Rxd8 26.Qh5 b5 27.Bxc5 Qxc5 28.Bxf7+²
C) 24...b6 25.f4²
25.Bc3
25.Ba3 b5³
25...Nd7 26.b4
Most forcing and thus critical, but Black has equalised in any case.
42
26.Bd3 Rfd8 with ...f6 and ...Nf8 to come means that White’s compensation is fizzling out.
26...Nf6
Position after: 26...Nf6
Now White has a vast variety of queen moves, none of which promise anything.
27.Qf5
A) 27.Qg5 h6 (27...axb4 28.Rxe5 Ra5!? 29.Rxa5 bxc3² seems practically quite promising) 28.Qxe5 Qxe5
29.Rxe5 axb4 30.Bxb4 Rfd8=
B) 27.Qh4 Rfd8 28.Rb1 axb4 29.Bxe5 Nd5 30.Qxe7 Nxe7 31.Bxg7 Bd5=
27...Rad8! 28.Rxd8
28.Qxe5 Qxe5 29.Bxe5 axb4 30.Bd6 Rfe8 31.Bxb4 Rxe1+ 32.Rxe1 Bd5=
28...Rxd8 29.Rxe5 Qc7!
43
Position after: 29...Qc7!
30.Re1
Most sensible.
30.Rxa5 will lose in a practical game. 30...Be4 31.Qc5 (31.Qb5 Ng4! 32.g3 h6!! and Black’s attack will be too
much) 31...Rd1+ 32.Bf1 Qd8© Black threatens mate and even after White scrambles with Qe3 and Be1, Black
will have a raging attack.
30...Bd7 31.Qxa5 Qxa5 32.bxa5 Rc8 33.Bxf6 Rxc4 34.Bd8 Ra4=
44
Position after: 34...Ra4=
The extra pawn falls, leaving a drawn endgame.
1.5) 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.d4
1.e4 c6 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.d4
Position after: 5.d4
To my mind not a particularly logical move in the position, but it is played a lot, so merits separate attention.
5...Nxc3 6.bxc3 c5!?
Given the stamp of approval by Khenkin and Rasmussen among others.
7.c4
A) 7.Bd3 Not great, but it leaves Black in the odd position of really not wanting to make his next few best
moves. 7...c4! (7...Nc6 8.dxc5 e6 9.Be3 is probably a slight edge for White) 8.Be2
45
A1) 8...Nc6 has been played in conjunction with a queenside castling plan, but fails to convince me. 9.0-0 e6
10.Ng5N Improving over 10.a4, which was the actual order of the game but would have allowed Black
...Be7 which stops Ng5. 10...Bd7 11.a4 Be7 12.Nh3 Qa5 13.Qd2 0-0-0 14.Qe3 h6 15.Nf4ƒ
Position after: 15.Nf4ƒ `
White’s position supplemented by Ba3 ideas made perfect sense in an advance – French kind of way in
Karjakin, S – Arutinian, D Almaty 2016.
A2) 8...e6 9.Ng5 Following the Karjakin plan. (9.0-0 Be7 10.a4 0-0„ is what Black wants, because White’s
46
kingside attack is a long way from fruition and meanwhile the queenside is extremely vulnerable.) 9...Be7
10.Nh3 b5!
Position after: 10...b5!
Black crucially does not commit to castling either side and simply arranges ... b4 in the minimum number of
moves possible. He has very good counterplay, for instance: 11.0-0 Nc6 12.a3 Bd7 13.f4 a5 14.Be3 0-0
15.Qd2 b4„
B) 7.dxc5 e6 8.Bd3 Nd7= Munoz Pantoja, M – Lopez Martinez, J Catalunya 2014.
C) 7.Bb5+ Bd7 8.Rb1 e6 did not give many problems in Hess, M – Vogel, R Tegernsee 2017.
7...Nc6!
47
Position after: 7...Nc6!
8.Be2!?
Trying to keep the central tension alive.

A) 8.c3?! cxd4 9.cxd4 Bf5 10.Qb3?! Else also ...e6, but this is critical. 10...e6! 11.Qxb7? Bb4+ 12.Bd2
Bxd2+ 13.Kxd2 Qa5+ 14.Kd1 0-0!!© From a practical perspective, Black is already winning.
B) 8.dxc5 e6=
C) 8.cxd5 Qxd5 9.c4 Qe4+ 10.Be3 Nxd4 11.Nxd4 cxd4 12.Qxd4 Bf5= Tikkanen, H – Rasmussen, A Sweden
2017.
8...e6
8...cxd4?! This is wrong, but instructively so. 9.Nxd4
48
Position after: 9.Nxd4
9...e6 (9...Nxe5 10.cxd5 is certainly interesting, but ‘more so for White’ – I believe I have analysed this to a
concrete edge for him, but we omit this as it’s not relevant.) 10.0-0 Be7 11.cxd5 exd5 12.Nxc6?! (12.c3 0-0
13.Bf4² is the way for White to get a more fluent position) 12...bxc6 13.c4 0-0 14.cxd5 cxd5 15.Be3 Qa5!=
Black had equalised in Ziska, H – Rasmussen, A Vaxjo 2017.
9.cxd5
9.c3 is, as usual, absolutely nothing: 9...cxd4 10.cxd4 (10.Nxd4 Nxe5 is now fine since White has wasted time)
10...Bb4+ 11.Bd2 dxc4 12.Bxc4 0-0 13.Bxb4 Nxb4 14.0-0 b6=
9...exd5
9...Qxd5? leads to a fun attacking line for White: 10.c4 Qe4 11.d5! exd5 (11...Nd4 12.Nxd4 cxd4 13.0-0 Qxe5
14.Re1‚ with Ba3 to come) 12.cxd5 Nb4
49
Position after: 12...Nb4
13.e6! White is close to winning in a practical game, for instance 13...fxe6 14.dxe6 Be7 (14...Bxe6 15.Ng5+–)
15.0-0 0-0 16.Bg5 Qxe6 17.Re1!? (17.Rc1 Rd8! 18.Bc4 Rxd1 19.Rfxd1 Qxc4 20.Rxc4 Bf8² and Black can
hang on and claim he has some compensation) 17...Bxg5 (17...Bf6 18.Rc1 Qf5 19.Bxf6 Qxf6 20.Rxc5‚)
18.Nxg5 Qf5 19.Bc4+ Kh8 20.Nf7+±
10.0-0!
Position after: 10.0-0!
10...Be7!
50
A) 10...cxd4 11.Nxd4 Be7 12.c3! has already been assessed as good for White.
B) 10...c4 is the silicon recommendation, but I can’t help feeling Black suffers from a slight lack of targets.
11.Ng5 Be7 12.Nh3‚
11.c4!
A) 11.c3 enables Black to equalise by keeping it simple: 11...0-0 (11...c4?! 12.Bxc4! is fascinating and
probably good for White) 12.Bd3 cxd4 13.cxd4 Bg4=
B) 11.dxc5 Bxc5 12.Bg5 Qa5!=
11...0-0 12.Be3
12...cxd4!
Black’s last difficult opening decision was whether and how to release all the central tension. Doing it in this
way equalises.
13.Nxd4 Nxe5 14.cxd5 Qxd5 15.Nb5 Qxd1 16.Rfxd1 Bg4
Giving back the pawn to reach a simple and very level position.
16...Nc6 17.Bf3 has the potential to become quite unpleasant, with Rab1 and Nd6/c7 ideas at the forefront of
White’s mind.
17.Bxg4 Nxg4 18.Bxa7 Rfd8=
51
Position after: 18...Rfd8=
1.6) 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.Ne2 Qb6 6.d4 e6 7.Ng3
1.e4 c6 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.Ne2
The main line! We saw White’s alternatives in the previous subchapters.
5...Qb6
52
A) 5...e6 6.d3 Nc5 7.Ng3 doesn’t quite satisfy me.
B) 5...Bg4 6.Nfg1 Nc5² has been tried once by the great master of solid openings, Julio Granda, but this one
did not go well for him.
6.d4 e6
6...g6?! 7.c3 (7.Nfg1 f6=) 7...Bg7 8.Nf4! White should not worry himself with trying to trap the e4-knight as it
will feel awkward in Caissa’s own time. 8...f6 9.exf6 exf6 10.Bd3²
7.Ng3
7.Nfg1 is the subject of the whole next chapter, and is a very contentious line in modern opening theory!
7...c5 8.c3
Position after: 8.c3
8...Bd7
This move is endorsed by Dragan Solak and the idea is as plain as day.
8...Nc6 is the main move but I’m not so sure about it: 9.Bd3 Nxg3 10.hxg3 cxd4 11.cxd4 Bd7 and now for
instance 12.Kf1 as in Adams, M – Motylev, A Germany 2015. Black suffers from a slight lack of plans, which
is something that doesn’t apply if the light squared bishops are off.
53
Position after: 12.Kf1
9.Bd3
No other move suggests itself.
9...Nxg3
9...cxd4?!
Position after: 9...cxd4?!
This gives White an option which is unusual for these French-type positions. 10.Nxd4! (10.Nxe4 dxe4 11.Bxe4
54
dxc3 12.bxc3 Bc6=; 10.cxd4 may not be bad as such, but is certainly a bit compliant. Either ...Nxg3 and ...Bb5,
or ...Bb4 and ...f5, give Black a good game.) 10...Nc5 (10...Nxg3 11.hxg3 Nc6 12.Nf3² and Black suffers from
the same malady as in Adams – Motylev above, with the added problem that White can castle if he wishes.)
11.Bc2 (11.Be2 Nc6 makes less sense as the e2-bishop is more restricted now, and there are few options other
than taking on c6.) 11...Nc6 12.Nf3!²
10.hxg3 Bb5
Position after: 10...Bb5
11.Bc2!?
A) 11.0-0 Consider as encompassed within this all kinds of other White moves that allow the bishop trade: g4,
a3, a4, Be3 and so on. They are all harmless. 11...Bxd3 12.Qxd3 Nc6
55
Position after: 12...Nc6
Here, Black need not worry so much about the Advance French motif of dxc5 and b4, but White needs to do it
or else his position makes little sense. 13.dxc5 (13.Bd2 cxd4 14.cxd4 a5!= Black sets up ideas of ...Bb4 and
...Qa6 and generally has more to work with in this position.) 13...Bxc5 14.b4 Be7 15.Bf4 0-0= White’s space
advantage is balanced by the rigidity of Black’s centre and his better minor pieces.
B) 11.Rxh7
Position after: 11.Rxh7
It is only thanks to modern computers that such moves can even be considered. About 10 years ago, almost
any strong player would instantly recoil from taking such pawns, not least because of the voice of Fischer
56
reaching out from the grave: “Don’t do it”... 11...Rxh7 12.Bxh7 Nc6 13.Bd3 (13.Bf4 cxd4 14.Nxd4 Nxd4
15.Qxd4 Bc5 16.Qd2 0-0-0©) 13...Bxd3 14.Qxd3 cxd4 15.cxd4 Nb4©
C) 11.Bxb5+ Qxb5 12.a4 Qc4
Position after: 12...Qc4
and thanks to the many options created by this far-advanced queen, White doesn’t have a lot of options other
than 13.Qe2 Qxe2+ 14.Kxe2 Nc6 which is quite equal.
11...Nc6

57
11...Ba6?! 12.Ng5!?²
12.a4!?
12.Be3 is also legal, but this not really how you play when you’ve just foregone castling. 12...cxd4 13.Nxd4
Bc5! Now if White tries to keep any kind of tension he will invariably end up worse, e.g. 14.a3 Rc8 15.Qd2
h6³.
12...Ba6!
12...Bc4 13.a5 Qc7 is similar, but allows random a6-pushes at different points...
13.a5 Qc7 14.Ba4
14.Bf4 cxd4 15.Nxd4 Be7= leaves White with few options other than light-squared bishop trades.
14...cxd4! 15.Nxd4 Be7
16.b4
16.Bf4 could transpose to the previous note, whereupon Black has a pleasant choice. 16...0-0-0!? (16...Kf8
17.Bb5 Bxb5 18.Nxb5 Qb8= Black has a solid position and the better bishop.) 17.Bb5 (17.Nb5 Bxb5 18.Bxb5
transposes) 17...Nxd4 18.Bxa6 Nc6 19.Bb5 (19.Bd3 h5= With ...g5 to follow.) 19...a6! (19...d4!?∞) 20.Bxc6
Qxc6=
16...Bc4 17.a6
17.Bf4 a6=
58
17...0-0!!
Position after: 17...0-0!!
Whisking the king away from all the unpleasant tactics on the queenside.
18.Qh5
A) 18.axb7?! Nxd4 19.Qxd4 (19.bxa8=Q?? Qxe5+ 20.Be3
Now 20...Nf3+!!–+ takes lots of pieces with check before recapturing on a8.) 19...Qxb7 20.Bc2 (20.Be3 f5!³
59
closes off White’s attacking avenues, while White’s king continues to feel a cold breeze along the a6-f1
diagonal.) 20...g6 21.Qf4 Bd8!
Position after: 21...Bd8!
Black sets up ...f5 (and incidentally ...a5) while not allowing Rxh7 tricks followed by perpetual check. The
immediate ...f5 was fine but had that drawback. 22.g4 f5„
B) 18.Bc2?! g6 19.Rxh7?? Nxd4!–+
C) 18.Rxh7?? Qxe5+–+
18...Bd3!
60
18...Qxe5+ 19.Qxe5 Nxe5 20.Bc2 Rfb8 is a bit random, but possibly playable too.
19.Bc2
19.axb7 Nxd4–+
19...Nxd4
19...Bxc2 is fine if one is desperate to play on: 20.Nxc2 h6 21.Bxh6 White has to strike before his position
crumbles. 21...Qxe5+ (21...gxh6? 22.axb7 Rab8 23.f4‚) 22.Qxe5 Nxe5
23.axb7 (23.Bf4 Nd3+ 24.Kd2 Nxf4 25.gxf4 Rab8 26.Rh3 Bf6„) 23...Rab8 24.Be3 Rxb7= White can take a
pawn, but it will not be long for this world.
20.Qxh7+ Bxh7 21.Bxh7+=
61
Position after: 21.Bxh7+=
The game ends in perpetual check.
CONCLUSION TO CHAPTER 1
As we have seen, none of the lines in this chapter should particularly bother Black, but it is worth noting
precise moves in two of the lines – that is, 5.Be2 Bf5 (pick one of the two options) and the final line 7.Ng3.
In particular, in the final line, the idea of castling away, even at great material cost, is rather important.
Meanwhile, the 5.Be2 line is rather underexplored and I expect more research and endorsements from both
sides of the board.
62
Chapter 2
Two Knights main line
1.e4 c6 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.Ne2 Qb6 6.d4 e6 7.Nfg1 f6 8.f3 Ng5
Chapter guide
Chapter 2 – Two Knights main line
1.e4 c6 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.Ne2 Qb6 6.d4 e6 7.Nfg1 f6 8.f3 Ng5
2.1) Minor 9th and 10th moves for White (and c3 options)
2.2) 9.exf6 & 12.Bd3
2.3) 9.exf6 & 12.Nxe4
2.4) 9.exf6 gxf6 10.f4 Nf7!?
2.5) 9.exf6 gxf6 10.f4 Nf7!?: 17.Nxe6
2.6) 9.exf6 gxf6 10.f4 Nf7!?: 17.Qe2!?
Building on the work done in the previous chapter, we now reach the critical 7.Nfg1, which could not possibly
have any other idea than to proceed with 8.f3 on the following move, bringing us to the diagram on page 47.
As the reader may already know, I try to provide alternatives to three types of recommendations (gambit
acceptance, IQP positions and unmitigated mess) but have not avoided them entirely, especially the last one.
Without being unnecessarily dramatic, the final two lines of the chapter fall into this category. I believe Black’s
position to be significantly easier to play, unless White knows about 17.Qe2 (line 2.6) and then has a good
knowledge of the subtleties that follow. Even then, a quick flick through to the final diagram of the chapter
should convince you that Black has good chances.
63
However, there is an alternative given: on move 10, it is possible for Black to play his knight to e4 instead of f7,
a method which was endorsed by Anand in a ‘modern classic’ game against Maxime Vachier-Lagrave. With the
aid of some improvements to both that game and the subsequent analyses, it is possible for me to reasonably
confidently claim dynamic equality from that continuation, even though there is never any such thing as the
‘last word’ except in tablebase territory.
2.1) Minor 9th and 10th moves for White (and c3 options)
1.e4 c6 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.Ne2 Qb6 6.d4 e6
7.Nfg1
7.c3 c5 is a pair of moves which can be inserted by White at virtually any stage, both here and later. In this
case, of course, White keeps the option of transposing to line 1.5 with Ng3.
7...f6 8.f3 Ng5 9.exf6
This is almost universally played.

A) 9.c3 c5 may be inserted; White has no real independent choice but exf6 now.
B) 9.h4 Nf7 10.f4 Inspired in a relatively understandable, yet misguided way, by a certain line of the French...
64
Position after: 10.f4
10...c5 11.c3 (11.Nf3 cxd4 12.exf6 gxf6 13.Qxd4 Nc6! 14.Qxb6 axb6³ is objectively best, but nobody wants
to see their centre crumble in this way!) 11...cxd4! (11...Nc6³ as in Toufighi, H – Lupulescu, C Rasht 2017 is
not bad either, but I prefer to try and solve the problem of the c8-bishop sooner rather than later.) 12.cxd4 Bd7
13.Nf3 Be7 14.Nc3 0-0 15.Be2 (15.Bd3 Nc6 16.Na4 Qa5+ 17.Nc3 fxe5 18.fxe5 Nfxe5µ) 15...Nh6!‚ Black
has far too many ideas for his opponent to keep under control, for instance ...Be8-h5 and ...Nf5-g3.
C) 9.Ng3!? c5 10.Nh5 risks jumping the gun slightly.
Position after: 10.Nh5
65
10...Nd7 11.a4!? Careful analysis shows this to be the only convincing follow up. (11.exf6 Nxf6 12.Bxg5
Nxh5 13.f4 g6 14.Be2 is the forcing continuation but any out of ...Qxb2, ...h6 and ...Ng7 seems fine.)
11...cxd4 12.Bb5 Nf7! (12...Kd8! 13.exf6 gxf6 14.Nxf6 Nf7 is the same idea) 13.exf6 gxf6 (13...g6?!
14.Ng7+ Kd8 15.Qxd4 Bc5 16.Qh4∞ is very hard to play for Black, and we should make a special note that
16...Bxg1 17.a5! is much better for White.) 14.Nxf6+ Kd8 15.Bxd7 Bxd7 16.Ne2 e5 17.Nxd7 Kxd7=
Black’s weakened king position is compensated for by a very strong centre.
9...gxf6
Position after: 9...gxf6
10.c3
A) 10.Bxg5?! fxg5 11.h4 gxh4 12.Rxh4 c5µ was a disaster zone for White in Beerdsen, T – Vereggen, L
Assen 2017
B) 10.f4 is critical and will be seen for the rest of the present chapter.
10...c5 11.Nf4
White could obviously have played the last two moves in the other order.
11...cxd4 12.Qxd4
12.cxd4 Nc6 13.Be3 Bd6‚
12...Qxd4 13.cxd4 Nc6 14.Nge2 Nf7=
66
Position after: 14...Nf7=
Black is better developed and can use the extra time to generate play to compensate for the backward e6-pawn.
2.2) 9.exf6 & 12.Bd3
1.e4 c6 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.Ne2 Qb6 6.d4 e6 7.Nfg1 f6 8.f3 Ng5 9.exf6 gxf6 10.f4
This is White’s critical choice and I’m going to give two options for Black here.
67
10...Ne4
Surprisingly, this move is more solid than ...Nf7, which I analyse next. Interestingly, in both lines, Black’s king
ends up on c7. I shall not endorse one over the other, but analyse both to the best of my ability and let you make
your own mind up.
11.Ng3 Bd7
Endorsed by Anand, and who am I to disagree? There are other interesting moves:
A) 11...f5!? could be an interesting direction if one is into the Stonewall, and certainly looked convincing in
Sadzikowski, D – Bartel, M Warsaw 2017.
B) 11...Kd8!? 12.c3 Rg8∞ is maybe the most creative option.
C) 11...Nd7!? also heads for a Stonewall.
12.Bd3
A fully natural move and reasonably critical.

A) 12.Nf3 is solid but uninspiring and White will need to step carefully from here on in. 12...Nxg3 13.hxg3
c5 14.c3 Nc6
White has foregone his Qh5-trick and is behind in development. Therefore, he should make haste in getting
the queens off with Qb3, or else he will stand worse. 15.Qb3 Qxb3 (15...Qc7 16.Be3 is less comfortable, in
my opinion) 16.axb3 cxd4 17.Nxd4 Nxd4 18.cxd4 Bd6=
B) 12.c3 c5 will transpose if White takes on e4, and the independent tries do not inspire confidence.
C) 12.Qh5+!? Kd8 13.c3 c5 14.Nf3 Nxg3 15.hxg3 Nc6³ Next there will come ...Kc7, and it becomes clear
68
that White’s centre will crumble. This is of greater importance than the fact that White may win a pawn.
16.Qf7 Kc7 17.Qxf6 Be7 18.dxc5 Bxc5³
D) 12.a4!? could be an interesting direction, since Black’s best reply is ...a5 but after that there could be issues
with ...c5 later.
Position after: 12.a4!?
12...a5 13.c3 (13.Nxe4 dxe4 14.c3 h5 transposes; 13.Qh5+ Kd8 14.c3 c5„) 13...h5!? Suggested by Vidit and
seems best. 14.Be2 (14.Nxe4 dxe4 15.Be2 c5!= and counterplay of full value for a pawn has arrived.)
14...Na6 15.Bxh5+ Kd8
Position after: 15...Kd8

69
16.Nf3 (16.N1e2 f5³ is likely better for Black already) 16...e5!? (16...c5 17.Nxe4 dxe4 18.Nd2 may work for
Black, but my feeling is White does not have to think so much here but just target the e4 pawn until Black
pushes ...f5, and then come in with Nc4-e5.) 17.fxe5 Nxg3 18.hxg3 Bg4 19.Qc2 Threatening to come in to
g6, so Black must take the piece now. 19...Rxh5 20.Rxh5 Bxh5 21.Qf5 Bxf3 22.Qxf6+ Be7
23.Qh8+ (23.Qxf3 Nc7µ) 23...Kd7 24.Qxa8 Bxg2 If Black gets one tempo, he will play ...Nc7 and stand
much better, so White absolutely has to play e6 now. 25.e6+ Kxe6 26.Qc8+ Kf7= White has nothing better
than chasing the Black king around a light-squared diamond with checks, because the c1-bishop can never
develop and b4 is met with ...Nxb4.
E) 12.Nxe4 is seen next.
70
Position after: 12.Bd3
12...f5!?
12...Nxg3 cues up an interesting queen sacrifice, if Black wants it, but regrettably I am not sure what Vidit had
in mind following a subsequent piece sacrifice from White. 13.hxg3 Qxd4?! (13...Kd8 14.Nf3 c5„) 14.Bg6+
hxg6 15.Qxd4 Rxh1 16.Qxf6! Rxg1+ 17.Ke2± White emerges better in all lines.
13.Nf3 h5 14.Be2
This seems to be the only challenging answer.

14.Ne5 Qxd4³
14...Nf6!?
A fairly shocking move, pulling away the safety option that White’s g3-knight previously had, and making it
less likely that Black’s king will have to move.
15.Ne5!
The most critical move, and now Black also has to exercise some caution.
A) 15.Nh4 c5!=
B) 15.Ng5 h4 16.Nh5 Ne4 is fascinating, but with his knights sidelined in this way White will be unable to
pose major challenges to Black’s simple idea of ...c5.
The line could continue 17.g4 c5 (17...Na6!?= is ‘cleaner’) 18.gxf5 exf5 19.dxc5 Bxc5 20.Nxe4 dxe4 21.Qd5
Kd8∞ with continuing mess, but my further investigations show that Black is fine.
71
15...c5
16.Bxh5+
A) 16.Nxh5 Ne4! And once more with feeling! 17.Bf3! Initiating a bizarre line which really should not be
forced, but is. (After 17.0-0?! Nc6µ the centre is already being swamped.; 17.Bd3 might be alright but Black
gets an extra option of 17...Qd8!? which will repay investigation.) 17...cxd4 18.Bxe4 Bb4+! 19.Bd2 dxe4
20.Nf6+ Ke7 21.Nfxd7 Nxd7 22.Ng6+ Kf6 23.Nxh8 Rxh8 24.Bxb4 Qxb4+ 25.Qd2 Qxd2+ 26.Kxd2 e5!©
Position after: 26...e5!©

72
Black›s pieces are so well centralised that his compensation for the exchange endures into the endgame.
B) 16.0-0!? is logical in some sense, even though this is not a terribly logical position. 16...cxd4 There is also
just 16...Nc6 followed by quick queenside castling. 17.Bxh5+ (17.Kh1 h4! 18.Nh5 Ne4³) 17...Nxh5
18.Nxh5 Rh6 Note that ...Be7 would bring us back to 16.Bxh5+ territory. 19.Rf2! (19.Kh1 Qd8³ causes
embarrasment) 19...d3! 20.Nf6+! (20.Nxd3 Bb5µ) 20...Rxf6 21.Qh5+
Position after: 21.Qh5+
21...Ke7 Now White has no better than perpetual check with: 22.Qg5 dxc2 23.Ng6+= .
C) 16.Be3 Nc6 17.Bxh5+ Nxh5 18.Nxh5 0-0-0 19.Nf7 cxd4© will give Black huge compensation for his
exchange.
16...Nxh5 17.Nxh5 Be7
73
A critical position.
18.0-0
White should probably get the king to safety reasonably soon.

A) 18.Be3 Nc6 and White has no real moves other than castling anyway; for instance either this sample line
or 19.Ng7 Kd8 just play into Black’s hands. 19.Ng6?! cxd4! 20.Bf2 Rh6 21.Nxe7 Kxe7³
B) 18.Ng6 Rh7 19.Nxe7 Kxe7³
C) 18.Ng7+!? A very serious attempt to draw Black to the kingside. 18...Kf8 19.Qf3™
74
Position after: 19.Qf3™
C1) 19...Rh6!? There is a strong possibility of the knight on g7 being trapped, but it turns out White can
sacrifice this piece and force Black to walk a tightrope to a draw. 20.g4! Kxg7 21.gxf5 exf5 22.Rg1+ Kf8
23.Qg2 Qe6 24.Bd2 Bh4+! (24...Nc6 25.c4±) 25.Kf1 Rh7 26.Ng6+™ 26...Ke8 27.c4 Qe4 28.Nxh4 Qd3+
29.Ke1 Re7+ 30.Kd1 Ba4+=
C2) 19...Qd8! 20.Nh5 (20.0-0 Nc6µ; 20.g4 Bh4+ 21.Kf1 Nc6!!µ) 20...Be8 21.Ng3 Nc6= As soon as Black
can, he should run for the queenside.
D) Inverting the move order with 18.Qf3 Nc6 19.Ng7+ gets nowhere: 19...Kd8µ.
18...Nc6
18...cxd4?! runs into trouble after the unusual rook lift 19.Rf3! Nc6 20.Ng7+ Kd8 21.Rb3².
19.Be3
19.Ng7+ Kd8 20.Nf7+ Kc7 21.Nxh8 Rxh8© as usual gives Black amazing compensation.
19...cxd4!
Position after: 19...cxd4!
20.Bf2
20.Ng7+ Kd8 21.Nxc6+ bxc6! 22.Bxd4 Bc5 23.c3 Bxd4+ 24.cxd4 c5∞ is a mess, but not one I think Black
should be afraid of; he can play ...Rc8 and ...Kc7-b8 regardless of the situation in the centre.
20...Nxe5 21.fxe5 Rc8
75
This is the engine suggestion, and it seems reasonable, for instance:
22.c3
22.Bxd4 Bc5 23.Nf6+ Kd8 24.c3 Rh4! 25.Bxc5 Qxc5+ 26.Kh1 Kc7„
22...Qxb2 23.Bxd4 Qa3©
Black will almost unstoppably play ...b6 and ...Kd8-c7-b7.
2.3) 9.exf6 & 12.Nxe4
1.e4 c6 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.Ne2 Qb6 6.d4 e6 7.Nfg1 f6 8.f3 Ng5 9.exf6 gxf6 10.f4 Ne4 11.Ng3
Bd7 12.Nxe4
Position after: 12.Nxe4
As we have seen, there are a whole lot of alternatives, but nothing that prevents ...c5 from coming soon.
12...dxe4 13.c3?!
This lets White in for a bit of suffering.

13.Nh3!? c5 No reason not to play the push here either. 14.d5 exd5 15.Qh5+! Kd8 16.Qxd5 Nc6 17.Be3
Admittedly White has stopped long castling and is about to do that move himself, but with a precise 17...Ne7
18.Qxe4 Qxb2 19.Rd1™ 19...Kc7 Black prevents this. There is a very critical continuation now: 20.f5!? Else
...Bc6 or ...f5 gives Black a good game. Here he should be careful.
76
Position after: 20.f5!?
A) 20...Qb4+ 21.Qxb4 cxb4 22.Bf4+ Kc8 23.Be2² is to be avoided – Black’s development problems are of a
long-term nature and even the normally materialistic computers begin to see their mistake a few moves after
merrily chomping the f5-pawn.
B) 20...Bxf5 21.Bf4+! (21.Qf3 Rd8 22.Bf4+ Kc8 23.Rxd8+ Kxd8³; 21.Qf4+ Qe5³) 21...Kb6 Black needs to
be unafraid to enter this line. A detailed investigation shows that the position is balanced but it is White who
needs more accuracy. 22.Rd6+ Nc6 23.Rxc6+! (23.Qxf5 Bxd6 24.Bxd6 Qc3+–+ wins the d6-bishop.)
23...bxc6 24.Qxf5 Re8+ 25.Kf2 Qd4+
Position after: 25...Qd4+

77
26.Kf3 (26.Kg3 Bd6!³) 26...Qd1+ 27.Kg3 (27.Kf2 c4µ) 27...Rg8+ 28.Kh4 Rg6 29.Rg1 (29.a4 Bd6µ)
29...Bh6 30.Bd3 Bg5+ 31.Bxg5 fxg5+ 32.Qxg5 Rxg5 33.Rxd1 Rxg2„
Position after: 13.c3?!
13...c5!?
Here I suspect the man from Madras mixed up his move order, as the text seems much more natural than what
he played. White is instantly faced with great problems even to keep equality.
13...Na6 is a valid choice but while Black is probably objectively fine, he will have to spend more time than
White in what follows. 14.Qh5+ Kd8 15.Bc4
78
Position after: 15.Bc4
In the Anand source game there followed an exchange of minor inaccuracies: 15...Kc7?! (15...Bd6!? may well
improve; 15...f5!? 16.Qf7 Be7 is likely fine too)
A) 16.Qf7 asks a whole lot of pertinent questions, and I’m not sure what Black had in mind.
B) 16.f5!? was most natural, opening up Bf4+ against the king which has just moved into position, and also
okay. 16...e5 (16...exf5 17.a4±) 17.Qh4!²
C) 16.a4?! c5!
Position after: 16...c5!
17.Ne2 (17.f5 Nb4!!„) 17...Rd8 18.Be3 f5„ Black eventually triumphed in the very confusing game Vachier
Lagrave, M – Anand, V Saint Louis 2016, which nevertheless merits further study.
14.Qh5+
14.d5 is always a move to consider in these ...c5 lines. 14...exd5 15.Qxd5 Nc6! 16.Qxe4+
79
Position after: 16.Qxe4+
16...Ne7! Suddenly all Black’s pieces are working. Natural development leads White into even worse trouble.
17.Nf3 0-0-0 18.Be2 (18.Bc4 Nf5© keeps White›s king in the centre) 18...Re8 19.0-0 Nf5 20.Qc2 c4+ 21.Kh1
Qe6! A good maneouvre to know. Now Black’s pieces each get their own lines and ...Bc6, ...Bc5 etc can
follow.
14...Kd8
Position after: 14...Kd8
15.Qh4
80
15.d5 exd5 16.Qxd5 Kc7³ is another version of the central pawn-sacrifice.
15...Be7 16.dxc5 Qxc5 17.Qf2! Qxf2+!
17...Nc6 18.Be3 Qa5 19.a4!² with Bb5 and Rd1 in the pipeline is probably not the way Black wants to play;
suddenly it is White with the development advantage.
18.Kxf2 Nc6=
Position after: 18...Nc6=
The thematic way to play this position is with ...Kc7 and a quick ...e5. Black has a more comfortable game,
even though it remains objectively level.
2.4) 9.exf6 gxf6 10.f4 Nf7!?
1.e4 c6 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.Ne2 Qb6 6.d4 e6 7.Nfg1 f6 8.f3 Ng5 9.exf6 gxf6 10.f4 Nf7
81
This is the second alternative and one which can lead to even more messy, almost Poisoned Pawn Najdorf-like
positions. I shall not lie to you: this move is maybe objectively slightly worse than 10...Ne4, but the number of
nuances White needs to know in order to reach the final position of this chapter is beyond belief, and even then
it is not totally clear, at least to me, that he stands better. However, one thing at least is in our favour: White has
only one remotely logical 11th move!
11.Nf3 e5
Black will get fairly simple equality unless White is fully in command of the wrinkles that follow, beginning
with 13.Ng3! and 15.Nh5! and even so, Black is not worse if he has an equal amount of knowledge.
A) 11...Nd7 12.Ng3 f5 was tried by Saric, but I am not convinced by this late switch to a ...Ne4 setup.
13.Nh5 Rg8 14.c3 Be7 15.Qc2 Nf6 16.Nxf6+ Bxf6 17.Be3² Black later prevailed with an interesting queen
sacrifice in Bosiocic, M – Saric, I Valpovo 2017 but on an objective level White should be better with his
immediate ideas of Ne5 or a queenside pawn storm beginning with c4.
B) 11...Be7 12.c4!? dxc4 13.Nc3‚ is a pawn sacrifice espoused by Grischuk and Bartel, and it seems quite
dangerous upon further analysis.
12.fxe5
12.Ng3 e4 13.Nh4
82
Here 13...c5! was presumably preparation from Aryan Tari, and it seems to work well.
A) 14.Nhf5 Nc6!N (14...cxd4 15.Qxd4 Qxd4 16.Nxd4= was less good in Lu, S – Tari, A Doha 2016)
15.dxc5 Bxc5 16.Qxd5 Bf2+ 17.Ke2 0-0!!³ Black has a huge attack for just one pawn. The details can be
worked out by the interested reader.
B) 14.dxc5 Bxc5
Position after: 14...Bxc5
B1) 15.Qxd5 Bf2+ 16.Kd1 Bg4+ 17.Be2 0-0 18.Bxg4 Rd8 19.Qxd8+ Qxd8+ 20.Ke2 Bxg3 21.hxg3 Nc6³
83
Materially, White is fine, and he might even be better if he could get in Kf2-g1-h2(!) and Be3. But Caissa is
not that generous. Here he is in some trouble.
B2) 15.Be2™ 15...Bf2+ 16.Kf1 Bxg3 17.hxg3 d4„
Position after: 17...d4„
The position remains rather original, but Black should play normal moves and he will be fine, for instance
...Be6 and ...Nc6 next.
12...fxe5
Position after: 12...fxe5

84
13.Ng3!
A surprisingly strong continuation, leading to deliciously complicated positions.

A) 13.dxe5 Bg7! will regain the pawn in a convenient way, for instance 14.Nf4 (14.Qd4 Nd7! 15.Qxb6 Nxb6
16.Bf4 Nc4! 17.e6 Nd8!=) 14...0-0 15.Bd3 (15.e6 Bxe6³) 15...Re8=
B) 13.Nxe5 Nxe5 14.dxe5 Bg7 15.Nd4 0-0 16.Bd3 Qc7!
A nuance pointed out by GM Gujrathi. Black can regain the pawn, at a time of his choosing, with the b8-
knight. 17.Nf3 (17.Qh5 Qxe5+ 18.Qxe5 Bxe5=) 17...Nd7 18.0-0 Nxe5 19.Nxe5 Rxf1+ 20.Qxf1 Bxe5=
13...Be6
A) 13...Bg4 14.Be2 Bxf3 15.Bxf3 Qxd4 16.Qxd4 (16.Qe2!? is also promising) 16...exd4 17.Bd2² White had
good compensation in Christiansen, J – Risting, E Stavanger 2017.
B) 13...Be7² was my original recommendation but it was found not to be playable.
14.dxe5
14.Be2 Nd7= is very unlikely to trouble Black.
14...Bc5
85
Position after: 14...Bc5
15.a4
An interesting move-order attempt. White’s main idea is the immediate Nh5, to which we proceed next, but this
tries to transpose into a good version of that.
15.Bd3?! Nd7 16.Qe2 0-0-0© White is unable to castle and soon there will be not one but two Black rooks on
the central files. An exhaustive analysis might now throw up a Black advantage, but I shall leave that to others
and simply claim extremely good compensation.
15...0-0!
15...Bf2+ 16.Ke2 a5 17.Nh5² brings us to a good version for White of the next line.
16.Qd2!
Defending e5 and keeping the pieces reasonably well co-ordinated.

A) Normal moves such as 16.Bd3?! leave the e5-pawn hanging: 16...Nxe5 17.Nxe5 Bb4+–+.
B) 16.c3?! Bf2+ 17.Ke2 (17.Kd2 Nxe5³) 17...a5!µ The extra tempo makes all the difference: Black’s 0-0 is
much more useful than White’s c3.
C) 16.a5 Qc7 rounds up the e5-pawn.
D) 16.Qe2 a5!µ is too artificial for White and he has no obvious play.
16...a5!
86
17.Ng5!
At least giving Black some problems. 17.Bd3 Nd7 will end up dropping the e5-pawn unless White plays a
belated Ng5, but even there a nice trick exists: 18.Ng5 Bf2+! (18...Rae8 was considerably more violent but also
playable.) 19.Qxf2 Qxf2+ 20.Kxf2 Nxg5+ 21.Kg1 Nf7=
17...Nxg5 18.Qxg5+ Kh8 19.Bd3™
Now Black can choose between three broadly reasonable moves: ...Na6, ...Nd7 or ...Qb4. Here we pick up the
simplest.
19...Qb4+! 20.c3 Qg4
87
Position after: 20...Qg4
21.Qxg4
Everything else leads to simpler equalisation.

A) 21.Be2 Qxg5 22.Bxg5 Nd7= regains the e5 pawn.
B) 21.Be3 Qxg5 22.Bxg5 Nd7=
C) 21.Rf1 Nd7=
21...Bf2+ 22.Kd1
22.Kd2 Bxg4 23.h3 Be6= picks up the e5 pawn without further arguments.
22...Bxg4+ 23.Be2
23.Kc2 Nd7 24.Bh6 Rf7=
23...Bxe2+ 24.Kxe2 Bb6!
88
Position after: 24...Bb6!
White’s dodges to keep the e5-pawn are nevertheless at an end, for instance:
25.Rf1
A) 25.Kd1 Rf2=
B) 25.Be3 Bxe3 26.Kxe3 Nd7=
25...Rxf1 26.Nxf1 Nd7 27.Bf4 Re8 28.Re1 Bc7=
Position after: 28...Bc7=

89
2.5) 9.exf6 gxf6 10.f4 Nf7!?: 17.Nxe6
1.e4 c6 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.Ne2 Qb6 6.d4 e6 7.Nfg1 f6 8.f3 Ng5 9.exf6 gxf6 10.f4 Nf7 11.Nf3
e5 12.fxe5 fxe5 13.Ng3! Be6 14.dxe5 Bc5 15.Nh5!
Position after: 15.Nh5!
15...Nd7!
Forcing White’s reply, else Black will be able to castle while White cannot.
15...Bf2+ is overwhelmingly tempting, but detailed analysis shows that the two tempi are actually of higher
value than the fact that White’s king has had to relocate. 16.Ke2 Nd7 17.Ng7+ Ke7 18.a4!? (18.Nxe6 Kxe6
19.g3 is also worrying) 18...Bc5 19.Qe1! Ndxe5 20.a5 Qb4 21.Qxb4 Bxb4 22.Ra4 Bd6 23.Nxe6 Kxe6
24.Nxe5 Nxe5 25.g3² Black faces a long and unpleasant squeeze.
16.Ng7+ Ke7 17.Nxe6
17.Qe2 is the next line.
17...Kxe6
90
Position after: 17...Kxe6
We have reached a critical position where both kings are desperately unsafe. The Black one is wandering
around on the third rank behind a pawn cover of mixed colour, while the White one lacks a pawn cover but is
only on the first rank. However, I shall argue that Black’s position is easier to play: his next move is ...Ndxe5,
the following one likely to be ...Rae8, and his goal is to exchange dark-squared bishops.
18.c3!?
Most flexible by far, and causes the most problems.

A) 18.g3 Ndxe5 19.Bh3+ Kd6µ is not fast enough.
B) 18.Qe2 Ndxe5 19.c3 Rae8 20.Bf4
91
Position after: 20.Bf4
20...Kd7 (20...Kf6!? would be a bizarre choice for a human, but seems also to work out) 21.Nxe5+ Nxe5
22.0-0-0 h5!= Now g3 Ng4! and Kb1 Ng6! work out badly for White, so he probably has to play Bxe5 Be3+
when once more both kings are able to reach solid castled positions.
C) 18.a4 is useful when Black’s bishop is ‘loose’ on f2, but less so here. 18...Ndxe5
Position after: 18...Ndxe5
19.c3 (19.a5 Qb4+ 20.Qd2 Qg4!?ƒ) 19...Kd7 20.a5 Qd8 Black’s position will consolidate around a setup
featuring ...Kc7, for instance 21.g3 Qf6 22.Bf4 Rae8 23.Nxe5+ Nxe5 24.Kd2 Bd6 25.Kc2 Kc7=
92
18...Ndxe5
Position after: 18...Ndxe5
19.Nd4+
19.Kd2!? Quite possibly the only independent move, most of White’s other options having been dealt with in
the previous set of notes. I am reminded of my adventures on the chess24 server, playing incognito under the
name KingRunEveryGame, where I would habitually open with king walks to c7/g7 (c2/g2) in blitz games, and
try to get to the third rank in bullet ones. 19...Raf8 20.Kc2 Kd7 21.Bf4 Ng4
Position after: 21...Ng4

93
It transpires that Black is not all that threatened and will be able to exchange off the f4-bishop if he wants- with
...Ne3, ...Nd6 or ...Bd6. In fact, to maintain equality, White may have to opt for 22.g3 Ne3+ 23.Bxe3 Bxe3
24.Bg2 Kc7= when an odd kind of symmetry has appeared.
19...Kd7 20.Be2
Probably the best way of preventing Black’s castling-by-hand, and definitely preferable to the alternative in a
practical game. 20.g3 Rae8 21.Bh3+ Ng4+! is a cute wrinkle.
Position after: 21...Ng4+!
All lines now lead to draws. 22.Kf1 (22.Kd2 h5 23.Bxg4+ hxg4 24.Qxg4+ Kc7 25.Ne6+ Kb8 26.Qf4+ Bd6
27.Qxf7 Rhf8! Black will give perpetual check.) 22...h5 23.Bxg4+ hxg4 24.Qxg4+ Kc7! 25.Bf4+ Kd8 26.Qg6
(26.Ne6+?? Rxe6 27.Qxe6 Qxb2–+; 26.Rb1 Qa6+ 27.Kf2 Re4 28.Rbe1 Ne5!=) 26...Bxd4 27.cxd4 Qb5+
28.Kg1 Qxb2 29.Qf6+ (29.Rd1?! Re2–+) 29...Kc8 30.Qf5+=
20...Nd6 21.0-0 Raf8
94
Position after: 21...Raf8
22.Be3!
This forces Black to take the bull by the pawns and go ‘full Poisoned Pawn’.
A) 22.Bg4+ Patzer sees a check, but this is really not a good choice. 22...Kc7 23.Rxf8! Keeping the queen
defending g4. (23.Be3 Rxf1+ 24.Qxf1 Kb8µ) 23...Bxd4+! (23...Rxf8 24.Be3²) 24.Qxd4 Rxf8 25.Qxb6+ axb6
26.Be2 Ne4 27.Be3 b5= Technically, this endgame is equal, but I prefer Black due to his more active pieces
and ideas like ...Nc4.
B) 22.Bf4 Rxf4! 23.Rxf4 Qxb2 24.Kh1! Qxc3 25.Bg4+ Ke8=
95
Position after: 25...Ke8=
Curiously, neither side can do better here than Bh5-g4, ...Kd7-e8. It may appear that White’s king is now
much safer, but ...Ne4-f2 or ...Nd3 have the power to appear out of nowhere on not much notice.
22...Rxf1+ 23.Qxf1 Qxb2 24.Rb1 Qxc3 25.Qf4 Ng6
This is a little hair-raising, but objectively dynamically level.
26.Qg4+
26.Bg4+ Kc7 27.Ne6+ Kb8 28.Bxc5 Qxc5+ 29.Nxc5 Nxf4 30.Na6+= is comparatively civilised.
26...Ke8™ 27.Qe6+ Kd8 28.Rc1!?
A) 28.Qf6+ Ke8=
B) 28.Bg4 Kc7 29.Qd7+ Kb8 30.Nxc6+ Ka8–+ is a flash in the pan.
28...Qxc1+ 29.Bxc1 Bxd4+ 30.Kh1 Kc7„
96
Position after: 30...Kc7„
We have reached the kind of position from which Gawain Jones makes a living. Further analysis is possible, but
Black is fine: his ideas of ...Ne4 and ...Re8 cannot be held up much longer.
2.6) 9.exf6 gxf6 10.f4 Nf7!?: 17.Qe2!?
1.e4 c6 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.Ne2 Qb6 6.d4 e6 7.Nfg1 f6 8.f3 Ng5 9.exf6 gxf6 10.f4 Nf7 11.Nf3
e5 12.fxe5 fxe5 13.Ng3! Be6 14.dxe5 Bc5 15.Nh5! Nd7! 16.Ng7+ Ke7 17.Qe2!?
Position after: 17.Qe2!?
97
A strong computer suggestion which initiates a bizarre forced line followed by an intriguing set of endgame
subtleties.
17...Bg4 18.e6
18.h3 Rag8! 19.hxg4 Rxg7 20.e6? Rxg4µ
18...Kf8 19.h3
19.exf7 Kxg7 20.h3 Bf5!= doesn’t bother Black.
19...Bxf3 20.Qxf3!
20.gxf3 Kxg7 21.exd7 Rhd8! fizzles out, for instance: 22.Qe6 Qc7 23.Qg4+ Kh8 24.Bf4 Bd6 25.Be3 Ne5
26.Qh5 Rxd7 27.0-0-0 Ng6=.
20...Nde5™
Position after: 20...Nde5™
Now the White queen can go almost anywhere on the board.
21.Qf5!
A) 21.Qf6 Be7 22.Qf5 Bh4+ 23.Kd1 Qd4+ 24.Bd3 Kxg7 25.exf7 Raf8!= This is not quite the last hard move,
but in what follows Black’s play is just as easy(!) as White’s.
B) 21.Qf4 Kxg7 22.exf7 Ng6!=
C) 21.Qh5 Kxg7 22.exf7 Nxf7 23.Be2 Bf2+ 24.Kd1 Qd4+ 25.Bd3 Rae8=
D) 21.Qg3 Rg8! 22.exf7 Nxf7 23.Ne6+ Ke7 24.Ng5 h6!‚
98
21...Kxg7 22.exf7 Qb4+™
Position after: 22...Qb4+™
23.c3
23.Bd2 Qh4+ 24.Kd1 Qf6 White is now not as well co-ordinated as in some of the other endings in this line,
but neither does keeping the queens on with 25.Qh5 Rhf8„ particularly favour him.
23...Qh4+ 24.Kd1
24.Kd2 Nxf7=
24...Qf6 25.Qxf6+ Kxf6 26.Be2
26.Bf4 Nxf7 27.Kc2 Rag8=
99
Position after: 27...Rag8=
In general, this type of position favours White, but in this precise case I might prefer to be Black. Black has
better co-ordination and a coherent plan for holding White’s kingside in a bind.
26...Kxf7 27.Kc2 Kg7
Position after: 27...Kg7
Black should not be in too much trouble; White is simply not well enough developed. A good rule of thumb is
to develop the rooks to central files and then harrass White’s dark-squared bishop as much as possible. There
are too many variations going forward, but in my investigations I have not been able to find a very convincing
100
move-order for White. Nevertheless, if this possibility bothers you, then I suggest 10...Ne4.
This has been a mess-heavy and memory-heavy chapter. The main advantage from our perspective,
however, is that White’s memory work is both more extensive and less intuitive. First of all, we get to make
a choice between 10...Nf7 and 10...Ne4. While stylistically I prefer the former, the latter is also very
playable, especially when we consider the improvements 14...Nf6! in line 2.2) and 13...c5 in line 2.3). After
13...c5 a few of the lines play themselves, but one which may need to be memorised is the pawn sacrifice
with which we meet 14.d5.
In the 10...Nf7 line, everything is a lot more ‘up in the air’ and less explored. White’s best path with
13.Ng3 and 15.Nh5 has not been known for long, and in his ChessBase DVDs Vidit admits he changed his
choice of 10th move based on this continuation. According to my analysis, it is possible to go down this
route anyway, and in the resulting positions with castling ruled out for both sides, it is Black’s play that is
easier to follow (his rooks are connected, and his king has a refuge on c7). Nevertheless, due to the novelty
of the line, it would be no terrible thing to conduct a bit of your own analysis if you intend to go this route.
101
Chapter 3
King’s Indian Attack and Fantasy
1.e4 c6
2.f4, 2.d3
2.d4 d5 3.f3
Chapter guide
Chapter 3 – King’s Indian Attack and Fantasy
102
1.e4 c6
3.1) My line: 2.f4
3.2) Kings Indian Attack: 2.d3 d5 3.Nd2
3.3) Fantasy variation: 2.d4 d5 3.f3
Having examined White’s most ‘concrete’ sidelines in the first two chapters, we now turn our attention to three
sidelines that try to bring about an ‘understanding-based’ game, where White develops according to a set
formula, having decided on move 2 or 3 what the structure should look like.
The first line is one I have tried twice myself with White, unwilling to take on the main lines of the Caro from
the ‘wrong’ side. It aims for an endgame with White pawns on e4 and f4 and a cubbyhole for the White king on
c2. However, this comes at the price of development, and with accurate play Black is able to reach an equal yet
slightly more comfortable position.
The second line is the King’s Indian Attack, which should really be named the ‘Reverse Philidor’ as you shall
see. It is hard to write about, and even more so when it comes to variations, because despite the setup having
some popularity, there is no prospect of a White opening advantage. The question then becomes one of which
Black setup among a set of objectively equal choices offers him the greatest potential for posing practical
problems. I believe I have made a good selection.
Finally, we come to the Fantasy Variation, used by Irish IM Gavin Wall to beat me in August 2017 and hence
substantially delay my Grandmaster title. Unsurprisingly, I am biting back against it, and not giving White
players anything like what they want (hint: they want 3...dxe4) while simultaneously creating a different type of
sharp position.
3.1) My line: 2.f4
1.e4 c6 2.f4
103
As second moves go, this is surely one of the most natural after 2.d4. It could be played with one of two ideas
in mind- either an Advance Variation gone mad (3.e5) or an endgame with a slightly annoying space advantage
for White. Among those who have tried this is Nigel Short...
2...d5
2...e5 3.Nf3 may not be bad for Black, but is a King’s Gambit- and not the most critical way to play against it
either.
3.d3
Shortly before writing the present chapter, I tried this against the English IM Jovanka Houska – a noted Caro-
Kann specialist and author – with White.
3.e5 Bf5 4.Nf3
Position after: 4.Nf3
has been tried by some surprisingly good players from the White side, but now it really seems to me that after
4...e6 it may even soon be White struggling. A key point to remember is that Black does not have to
immediately address the development of his kingside pieces; in lines like 5.d4 h5 6.Be3 Qb6 7.Qc1 Nh6 8.c4
Nd7 9.Nbd2 c5 10.Nb3 Rc8 11.Qd2 Be7 12.Rc1 0-0 it can be several moves before he does so completely.
Black was more comfortable here in Morozevich, A – Svidler, P, Monte Carlo 2005.
3...dxe4
Probably simplest.
A) 3...e5!? is the maximalist choice, but there could easily be problems after 4.fxe5 dxe4 5.Nc3 since the
‘straightforward’ choices like ...Bb4 and ...exd3 admit to being worse. Maybe this can be salvaged with say
104
...Qa5 but it is unnecessary.
B) 3...g6 4.Nf3 Bg7 and similar options give White the flexibility of a range of anti-Sicilian approaches, in
which Black will probably need to subsequently play ...c5, leaving him a tempo down on the same.
C) 3...e6 4.Nf3 Be7 makes some sense, by analogy with the Fantasy Variation, but 5.Be2 again leaves Black
short of moves that don’t simply admit to having wasted time.
4.dxe4 Qxd1+ 5.Kxd1
Now Black should be a bit careful.
Position after: 5.Kxd1
5...Nf6
Detailed analysis shows that after 5...e5?! Black must play for compensation and cannot expect the pawn back
quickly; especially not after 6.fxe5 Bg4+?! 7.Be2 Bxe2+ 8.Kxe2² when White was already significantly better
in Fernandez, D – Houska, J, London 2017.
6.e5 Nd5
Another idea may be 6...Ng4?! 7.Ke1 h5 when White has been prevented from expanding with h3, g4 but
Black’s development still looks a bit questionable after 8.Bd3².
7.Nf3 Bf5
7...Bg4 8.c4 Nb6 9.b3 Na6 10.Kc2 gives White his desired regrouping and the knights may start looking
clumsy in a few moves.
8.Nd4
105
Position after: 8.Nd4
8...Bg6
8...Bg4+ 9.Be2 Bxe2+ 10.Kxe2 g5 is an engine idea but the circumspect 11.g3² leaves White for choice.
9.c4
9.e6 fxe6 10.Nxe6 Bf5= may well be what White should do by now.
9...Nb6 10.e6 Na6„
106
Position after: 10...Na6„
Black has solved his co-ordination issues, and will castle soon leaving White at risk of overextension.
3.2) Kings Indian Attack: 2.d3 d5 3.Nd2
1.e4 c6 2.d3 d5 3.Nd2
White tends to just be worse if he tries to play a KIA without the text, e.g. 3.Nc3 e5 4.g3 Nf6 5.Bg2 Bb4ƒ.
3...e5 4.Ngf3
Position after: 4.Ngf3
The starting position of the King’s Indian Attack, which in its purest form will follow up with g3, Bg2 and 0-0
like a reversed version of its namesake.
4...Bd6
Now White has quite a few moves.
5.g3
A) 5.d4!? This move has a lot more to commend it than it might initially appear. As I am unconvinced by the
conventional IQP positions that may result, and indeed would be happy to play them with White, I shall
recommend the more ‘Tarraschian’ approach of 5...exd4 6.exd5 Nf6!? 7.dxc6 Nxc6.
107
Position after: 7...Nxc6
Here Black still has an IQP, but it can’t be blockaded on the usual d4-square, and in the play that follows it is
quite often sacrificed, giving truly dynamic play. 8.Be2 There isn’t another logical way to continue
development, although arguably White could play first Nb3 and only then this. (8.Nb3 0-0 9.Nbxd4?! Nxd4
10.Qxd4 Bf5³ gives Black stupendous compensation; 8.Bd3 Qe7+ 9.Qe2 Qxe2+ 10.Bxe2 Bf5ƒ does not
promise White much fun either) 8...0-0 9.0-0 (9.Nb3 Bb4+ 10.Bd2 Qb6= gives typical Tarrasch equality)
9...Be6 10.Nb3 Qb6!?
Position after: 10...Qb6!?
An interesting pawn sacrifice.The alternative is to take on b3 and then play ...a6, ...Qc7, ...Rad8 et cetera,
108
which offers equality. 11.Nfxd4 (11.c3 Rfd8 12.cxd4 h6³; 11.Nbxd4 Rfd8 12.c3 Nxd4 13.Nxd4 Be5„)
11...Rfd8 12.c3 Nxd4 13.cxd4 h6 14.Be3 Bd5© Black has full compensation for his pawn.
B) 5.Be2 seems to me chiefly useful as an attempt to improve on 5.d4. 5...Nf6
B1) 6.0-0 is of course playable, and we reach a reversed Philidor. The extra tempo is not enough to give
White an edge, in fact it may not even give equality. Consider the following sample line: 6...0-0 7.c3 Nbd7
8.Re1 Re8 9.Qc2 Nf8 10.Nf1 a5
Position after: 10...a5
Black can continue with all the conventional plans: ...Ng6(-f4), ...h6/ ...Nh7-g5, or just queenside expansion.
Meanwhile White’s position is solid but lacks possibilities.
B2) 6.d4 Nxe4 I am even less convinced by Black’s straightforward IQP options here, e.g. 6... Nbd7. 7.dxe5
109
Position after: 7.dxe5
7...Nxd2
B2.1) 8.exd6 Nxf3+ 9.Bxf3 Qxd6 10.0-0 0-0© leaves White with seemingly just enough compensation for
the pawn but it will be very hard to claim more than that.
B2.2) 8.Qxd2 Bc7 9.0-0 0-0 10.c4 Bg4= led to a very ‘Petroff’ type of equality in Ljubojevic, L –
Dzindzichashvili, R Tilburg 1978.
B2.3) 8.Bxd2 Be7 9.0-0 0-0 10.b4 This is, for me, a slightly challenging plan to face. White’s idea is to
play some kind of minority attack on the queenside, exchanging one or two pairs of pawns and using the
resulting open files for his major pieces. Perhaps the best answer is to try and use the hole on c4
immediately. 10...Nd7 11.Re1 (11.b5 f6! 12.exf6 Bxf6 13.Rb1 c5„) 11...Nb6 12.a4
110
Position after: 12.a4
Doubtless there are other plans, but while there continues to be such a lack of praxis we can’t know which
is the best. This seems natural. 12...Nc4 13.a5 c5! Black can’t very well just sit and wait to be choked with
Bc3, Nd2, etc. 14.bxc5 Bxc5 15.Bg5 Qd7„ White’s structural problems and Black’s lack of breaks mean
that both sides have difficult positions to play.
5...Nf6 6.Bg2 0-0 7.0-0
Position after: 7.0-0
7...Re8
111
Houska recommends 7...Qc7. While there can obviously be nothing too badly wrong with a developing move
chosen by many 2600+ players in the past, I shall outline my (stylistic) issue with it. 8.b3!?
A) 8...Be6 will lead to White having all the light squares and easy piece play, while Black’s counter against
the d3-pawn and d4-square lacks bite: 9.exd5 cxd5 10.c4 Nc6 (10...d4 11.Ng5 Bg4 12.f3 Bh5 13.Nde4²)
11.cxd5 Bxd5 12.Bb2ƒ
B) 8...Re8 9.Bb2 Nbd7 looks solid, but following 10.d4 I didn’t find a totally satisfactory way for Black to
resolve the central tension, e.g. 10...exd4 11.exd5 c5 12.c3! dxc3 13.Bxc3 Nxd5 14.Nc4‚
C) 8...a5 9.d4 exd4 10.exd5 cxd5 11.Bb2 Nc6 12.Nxd4 Bg4 13.N2f3=
Position after: 13.N2f3=
This position is objectively equal, and yet I am uncomfortable recommending it, because three pairs of minor
pieces could easily come off by accident, and then Black would end up defending a nasty IQP position for the
rest of the game; Black lacks the dynamism to easily prevent this from happening.
Also, on a philosophical level, in these pages I have tried to give alternatives both to what I consider
unnecessarily bad versions of IQP positions, and to existing recommendations by other authors.
8.Re1
A) Following 8.b3 White at no time has room to play an advantageous d4-break: 8...a5 9.a3 h6!N 10.Bb2
Bg4=.
B) 8.d4 Nxe4! 9.Nxe4 dxe4 10.Nxe5 Nd7= once more grants a ‘Petroff type of equality’.
C) 8.c3 is undeservedly popular considering that it creates a significant weakness on d3. For instance:
8...Nbd7 9.Qc2 a5 10.a4 Nc5= and White should consider carefully what he intends to do about the opening
d-file and the possibility of a future ...Qd3.
8...Nbd7 9.b3 a5
112
Dreev’s ...Nf8-g6 also deserves attention at this point and has scored an amazing 78% for Black.
10.a3
Bringing us to the tabiya.

10.Bb2 a4 swiftly equalises for Black; best play now appears to be 11.Rb1 h6 12.bxa4 Rxa4 13.c4 b5„ when
Black appears to have no problems but neither does White so maybe he should choose this instead of the more
common text!
10...d4 11.Bb2
A) 11.Rf1 was inexplicable in Andriasian, Z – Motylev, A Tallinn 2016 ; simply 11...b5 would have been fine.
For instance 12.Nh4 a4 13.b4 c5 14.Nf5 Bf8³ and White’s f4-push arrives way too late; certainly about 4-5
tempi behind the equivalent in the KID.
B) 11.Nh4 b5 12.Nf5 is a marginally better execution but after 12...Bf8 a next move is not forthcoming; note
that tactically Black is doing well after 13.f4?! g6³.
11...a4
11...b5!? is also reasonable; as in Stroeher, W – Pezerovic, E Bayern 2007.
12.b4
113
Position after: 12.b4
12...c5
Forcing a decision.
13.c3!
A) 13.b5 Nb6³ quite likely spells the end of White’s b-pawn and compensation will almost certainly be
insufficient.
B) 13.Nc4 Bc7 14.b5 Nf8³
C) 13.bxc5 Bc7! Avoiding tricks based on Bxd4. 14.Nc4 Nxc5 15.c3 dxc3 16.Bxc3 Bg4!³
13...dxc3 14.Bxc3 b5
114
Now I believe White should liquidate the centre before it comes under pressure from ...Bb7, ...Rc8 et cetera.
15.bxc5 Bxc5 16.d4 exd4 17.Bxd4 Bxd4 18.Nxd4 Ne5! 19.N2f3!
19.Nxb5 Bg4–+
19...Bg4 20.h3 Bxf3 21.Nxf3 Qxd1 22.Raxd1 Nxf3+ 23.Bxf3 Re5=
Position after: 23...Re5=
115
The position has been clarified greatly and is almost exactly equal.
3.3) Fantasy variation: 2.d4 d5 3.f3
1.e4 c6 2.d4 d5 3.f3
The famous Fantasy Variation.
3...e6
A) 3...dxe4 4.fxe4 e5 5.Nf3 Be6 is another line that is possible to recommend, but it is less solid.
B) 3...Qb6 is in principle a good idea, but following c3 and Qb3 from White the position is not totally
comfortable. The following variation is not best play but serves to illustrate: 4.c3 Nf6 5.e5 Nfd7 6.Qb3²
Black will soon be forced into a less-than-desirable Advance French structure where he has spent tempi with
his knights but White hasn’t.
4.Nc3
A) 4.Be3 This gambit makes very ltitle sense, but disregarding it completely would be silly too. 4...dxe4
5.fxe4 (5.Nd2 exf3 6.Ngxf3 Nf6³ sees White struggling to get even basic compensation, due to Black’s plans
of ...Ng4 and ...c5) 5...Qh4+
116
Position after: 5...Qh4+
6.g3 (6.Kd2!? Nf6 7.Nf3 Nxe4+ 8.Kc1 Qd8 9.Bd3 Nf6© probably objectively fails to get quite enough
compensation, but it has a certain curiosity value.) 6...Qxe4 7.Qd2 Nd7 8.Bg2 Qf5! 9.c4 (9.Nf3 Qb5!³
prevents castling and gives Black some advantage) 9...Ngf6 10.Nf3
10...Ng4 Black gets developed without an excess of drama; however, it is very useful that he has wrinkles like
11.0-0 Nxe3 12.Qxe3 Be7 13.Ne5?! Qxe5!³ at his disposal.
B) 4.e5!? tries to take play into some kind of Advance French, but it’s not going to work as planned. 4...c5
117
5.c3 Qb6 6.f4 Bd7= White lacks a reasonable way to prevent ...Bb5, and the ensuing position will be either
equal or slightly favourable for Black.
4...Qb6!?
Position after: 4...Qb6!?
Now this move makes a lot more sense to me, striving to keep the c1-bishop firmly at home.
5.a3!?
Threatening to develop the c1-bishop anyway, since the pawn grab on b2 can now be met by Na4.
A) 5.Rb1 c5! 6.Be3 Nf6 7.Bb5+!? (7.e5 Nfd7 8.f4 Nc6 9.Nf3 transposes directly to a less-critical line of the
French, and here Black has equality: 9...cxd4 10.Nxd4 Bc5 11.Na4 Qa5+ 12.c3 Bxd4 13.Bxd4 Nxd4
14.Qxd4 b6=) 7...Bd7 8.Bxd7+ Nbxd7 9.exd5 Nxd5 10.Nxd5 exd5 11.Ne2 Be7
118
White can reach a large variety of IQP positions but there is no advantage in any of them.
B) 5.e5 c5 6.Be3 Ne7³ as usual spells death for White’s centre.
C) 5.Na4!?
Position after: 5.Na4!?
This is an interesting and maximalist move, which has not received the deserved attention.
C1) 5...Qa5+?! was seen in a high-level practical test, but it only encourages White’s idea. It is hard to find
improvements to Black’s play after this. 6.c3 Nd7 7.b4 Qc7 8.e5 b6 (Following the natural 8...b5 the
119
computers soon stop seeing logical things to do with Black’s position, e.g. 9.Nc5 a5 10.f4 Nh6 11.Nf3 axb4
12.cxb4 Nxc5 13.dxc5 Be7 14.Bd3²) 9.f4 c5 10.a3 Ne7 11.Nf3 Nc6
White was already comfortable in a game Popilski – Rodshtein, Acre 2016, and probably best now was
12.dxc5!N 12...bxc5 13.Be3².
C2) 5...Qc7! 6.Be3 (6.c3 is now too slow: 6...Nd7 7.b4 dxe4 8.fxe4 b5 9.Nc5 a5³) 6...dxe4 7.fxe4 e5= Black
has managed to get a good version of the usual Fantasy structure, because at some point the a4-knight will
need to return to c3.
5...Nd7
Awaiting the development of White’s dark-squared bishop helps keep us in pure Caro territory.
5...Nf6 is possible, but here after 6.e5 Nfd7 7.f4 c5 8.Nf3 Nc6 we reach another French position, where the
move a3 is of much more use than the move Rb1 in the earlier line.
6.Be3 Ngf6!
120
Position after: 6...Ngf6!
Now White does not have e5, and so we can play this move, intending ...c5 on the very next move. White has
tried a substantial variety of moves here, and none of them is the best.
7.Be2N
A) 7.Bd3 c5! Black is already almost better. The only way to stay equal is in fact to retreat with Be2.
Alternatives:
A1) 8.dxc5 Bxc5 9.Bxc5 Nxc5! (9...Qxc5 10.Qd2 0-0= Delgado Camara, L – Palacios Pena, G Badalona
2016) 10.Bb5+ Bd7 11.Bxd7+ Ncxd7³
A2) 8.Na4 Qc6! 9.Nxc5 Nxc5 10.dxc5 dxe4 (10...Bxc5 11.Qe2 0-0³) 11.fxe4 Bxc5 12.Qe2 Qb6! 13.Bxc5
121
Position after: 13.Bxc5
A2.1) The stars had aligned to make the surprising tactic 13...Qxb2!µ possible; White will lose either the
bishop or the rook on the next move, and none of the attempts to sacrifice the house and get the queen to d6
are working.
A2.2) 13...Qxc5? A great shame, though there was a silver lining for one of the brightest talents from my
‘old country’, who managed to hold a 2600 player to a draw. 14.e5 Nd5= Low, Z – Khairullin, I Bangkok
2016.
B) 7.Rb1 dxe4 8.fxe4 e5 9.Nf3

122
9...Qc7!?N (9...Bd6= was perfectly okay too in Das, A – Roiz, M Biel 2012) 10.Bc4 (10.Qd2 exd4 11.Nxd4
Bc5³) 10...Bd6 11.d5 0-0ƒ White is the one who must demonstrate equality here, because of the strength of
Black’s dark squares.
C) 7.Bf2 dxe4! (Previously only 7...c5 had been played, leading to playable if rather cramped French-type
positions.) 8.fxe4
Position after: 8.fxe4
Else Black is just better by normal methods, but here there is a resource worth noting. 8...Ng4! 9.Bh4 (9.Qxg4
Qxb2 10.Rb1 Qxc3+ 11.Kd1 e5µ) 9...Ne3 10.Qc1 (10.Qd3 Qxd4µ) 10...Nxf1 11.Nge2 Nxh2 12.Rxh2 e5µ
White has no real compensation for the pawn.
D) 7.Nge2 Qxb2
123
Position after: 7...Qxb2
Black should take this pawn, otherwise there is little point playing ...Qb6 at all. Note that White’s f1-bishop is
crucially not covering the b5 retreat square anymore. 8.e5 Ng8 9.g4 Qb6 10.Qd3 c5³ White has some
compensation, but it is not sufficient.
E) 7.b4 dxe4 (7...a5!? 8.b5 a4„ is worth exploring) 8.fxe4 e5
9.Nf3 (9.Bc4 Bd6 10.Nf3 0-0³ Black was already slightly better due to his dark squares and idea of ...Ng4 in
Velez Betancourt, N – Garcia Palermo, C Cienfuegos 1983) 9...a5N (9...Bd6 10.Qd2 Qc7 11.Rd1= Nabaty, T
124
– Rodshtein, M Jerusalem 2015) 10.b5 (10.Na4 Qc7 11.b5 Bd6³) 10...Qc7 White seems a little overextended
and will have his work cut out proving equality against even quite simple play, whereas Black has easy plans
of putting pieces on c5 and e5 – and playing ...Ng4 if appropriate.
7...dxe4
7...h5!? Black reintroduces the idea of ...dxe4 followed by ...Ng4. This idea works pretty well if White takes the
bait and plays 8.e5, because the piece sacrifice gives amazing compensation. However, there is 8.Nh3! dxe4
9.fxe4 Ng4 10.Bxg4! hxg4 11.Ng5² when I was unable to find any move for Black that did not run into a really
strong attacking counter.
8.fxe4 Be7
9.e5
9.Nf3 e5 10.Bf2 Qa5!= Black has a standard ‘good Philidor’ position.
9...Nd5 10.Nxd5 exd5 11.Nf3 f6 12.Qd2
12.0-0 fxe5 13.Bg5 e4= leaves White just enough compensation for the missing pawn.
12...0-0 13.0-0
A) 13.c4 looks critical, and indeed it is. In the lines that follow, Black will need to be fine sacrificing an
exchange for at least one (but usually two) pawns.
125
13...fxe5 14.cxd5 (14.dxe5 Bc5³) 14...e4 15.d6 Bxd6 16.Bc4+ Kh8 17.Ng5 Nf6 18.0-0 (18.Nf7+ Rxf7
19.Bxf7 Qc7³ and ...Bg4 next will leave White high and dry in the centre) 18...Qc7 19.Bf4 c5„
B) 13.e6 Nb8 14.c4 Bxe6³ is just a flash in the pan.
13...fxe5 14.dxe5 Qc7
15.c4!
126
White has to play with energy before Black either develops or takes on e5.
A) 15.Rae1 Nxe5 16.Nxe5 Qxe5 17.Bd4 Qd6³ demonstrates the latter.
B) 15.Bd4 Nc5 16.b4 Ne4 17.Qe3 Be6= demonstrates the former.
15...Nb6!?
It is also possible to take on e5 but this seems the most solid.
16.c5
Against any other answer it is obvious how Black should play. Hint: the c8-bishop goes to e6.
16...Nc4 17.Bxc4 dxc4 18.Qc3 Be6 19.Nd2 b5 20.cxb6 Rxf1+ 21.Rxf1 axb6 22.Nxc4 b5 23.Nd2 Bd5©
Position after: 23...Bd5©
Black has full compensation for White›s extra pawn, which is well blockaded in any case, and would continue
to be even if the major pieces came off.
Against ‘scheme-based’ White sidelines like these, there is little point doing memory work. Even in the
Fantasy Variation, knowing my main line up to move 14 or 15 should be good enough. More important is to
get a feeling for where the pieces go – doing a little experimentation with move-orders yourself would be a
good idea, so that you understand, for instance, the different characteristics of the positions after 9...a5 and
9...Nf8 in line 3.2).
127
Chapter 4
The Pseudo-Panov
1.e4 c6 2.c4 d5 3.cxd5 cxd5 4.exd5 Nf6
Chapter guide
Chapter 4 – The Pseudo-Panov
1.e4 c6 2.c4 d5 3.cxd5 exd5 4.exd5 Nf6
4.1) Intro and 5.Bb5+
4.2) 5.Qa4+
4.3) 5.Nc3 Nxd5 6.Nf3 Nc6 7.Bb5 g6 sidelines
4.4) 5.Nc3 Nxd5 6.Nf3 Nc6 7.Bb5 g6 8.Qa4 Nb6 9.Bxc6+
The Pseudo-Panov and its cousins the Panov-Botvinnik Attack and the Exchange Variation (next chapter)
occupy a convenient middle-ground between the sidelines and the main lines. All of them focus on getting
White’s pieces out quickly and keeping the position open, and this approach is always going to be a little bit
dangerous for the ill-prepared player. However, on another level, the lines can be considered quite drawish, as
once Black has demonstrated the small number of nuances required to equalise the game, there may not be
anything left in the position that either side can try and use to win with thereafter.
4.1) Intro and 5.Bb5+
1.e4 c6 2.c4!?
128
Position after: 2.c4!?
This is the so-called Pseudo-Panov, which is seen reasonably often. White defers (for one move) the choice of
whether to play lines involving trying to keep a lone isolated pawn at d5 alive, or transpose back into
mainstream Panov territory.
2...d5 3.cxd5
3.exd5 is of course playable for White, but after 3...Nf6 he is best advised to enter ‘normal Panov’ territory
with 4.d4, transposing to later lines in this chapter. This is because keeping the pawn with 4.dxc6 Nxc6 5.Nc3
e5© is miserable for White.
3...cxd5
3...Nf6 is still in some sense playable, but 4.dxc6 Nxc6 5.Nc3 e5 6.Nf3² should restrict Black’s compensation
to a minimum.
4.exd5 Nf6 5.Bb5+
The other two lines, seen next, are 5.Qa4 and 5.Nc3 Nxd5 6.Nf3.
5...Nbd7 6.Nc3 a6!
129
This move seems best; I am slightly sceptical of the endgame positions arising after 6...g6 and a later White d6-
push.
7.Qa4
A) 7.Be2 Nb6 8.Nf3 Nbxd5=

B) 7.Bxd7+ Qxd7 8.Nf3 Nxd5 9.0-0 e6 There was also the more maximalist 9...f6!? 10.Ne5 Nxc3 11.bxc3
Qd5= White has absolutely nothing as in Bodek, M – Ostrovskiy, A Crossville 2011.
C) 7.Ba4 b5 8.Bb3 Bb7 recoups the d5-pawn either by ...b4 or by ...Nb6xd5, leaving Black at least equal (and
in practical terms Black must be for choice).
7...Rb8
This is the main move and easily good enough.

7...g6!? could possibly improve even further, but it isn’t necessary.
8.Bxd7+
130
Position after: 8.Bxd7+
8...Qxd7!
8...Bxd7?! 9.Qb3 Here I am a little sceptical of Black’s position. 9...b5 10.d3 e6 (10...b4 11.Ne4 Rb5 is a
silicon variation that I don’t really believe either) 11.Bg5 exd5 12.Nf3= To me, it looks like White has the
long-term trumps and any Black play is based on tricks only.
9.Qxd7+
White has to tread carefully to avoid being substantially worse.

A) 9.Qf4 Ra8 10.Qf3 b5³
B) 9.Qb3 Qg4³
9...Bxd7 10.Nge2 Rd8 11.d4
11.f3 Bc8= takes back the pawn even more trivially.
11...Bf5 12.Nf4 g5!
131
Position after: 12...g5!
13.Nfe2 h6 14.h4 Rg8 15.hxg5 hxg5=
We have been following the game Arakeljan, A – Pogosian, M Dubna 2015. White is out of tricks and Black
will recapture on d5 next, with a game that is at least equal for him.
4.2) 5.Qa4+
1.e4 c6 2.c4 d5 3.cxd5 cxd5 4.exd5 Nf6 5.Qa4+
Position after: 5.Qa4+

132
This is another, slightly more interesting try at keeping the pawn, which has some subtle points to it and thus
gives Black a reasonable number of practical problems. One of this line’s main adherents is the Ukrainian GM
Olexandr Bortnyk.
5...Nbd7!
5...Bd7 6.Qb3 Na6 has been recommended, but after 7.Nc3 it is difficult to believe claims of full equality.
Some brief sample lines: 7...Nc5 (7...Qb6 8.Bc4 Qxb3 9.Bxb3 Nc5 10.Bc2 e5 11.dxe6 Nxe6 12.Nf3 Bc5 13.0-
0 0-0-0 14.d3² Ivanchuk, V – Burgess, M Groningen 1986) 8.Qd1
Position after: 8.Qd1
8...g6 (8...e5 9.dxe6 Nxe6 10.Nf3 The jury may still be out, but I am not going to recommend this gambit line.
It seems to me that White untangles and develops fast enough, e.g. 10...Bc6 11.Bc4²) 9.Nf3 Bg7 10.Bc4 0-0
11.0-0 Rc8 12.Re1 a6 13.d4 Na4 14.Nxa4 Rxc4 15.Nc5² White was better, even before we begin to mention
Black’s tactical issues, in Vyzmanavin, A – Smagin, S Moscow 1984.
6.Nc3 g6
133
Position after: 6...g6
7.Bc4
7.Nf3 Bg7 is utterly pointless as an independent sequence, because if White’s centre is not secured quickly with
d3 then he will run into all kinds of problems. 8.Qb3 0-0 9.Bc4
A) 9...Nc5 is also possible, but seems a little convoluted. 10.Qa3! Nce4 11.0-0 Nd6 (11...Bf5!?© would be a
‹truer› gambit approach, and maybe the best way to play.) 12.d3 a6 13.Bb3 b5 Black will regain the pawn in
one way or another, but it is possible for White to ask a few searching questions before that happens, e.g.
134
14.Bf4 Bb7 15.Be5!?ƒ with a small initiative for White due to the tactical weakness of the e7-pawn.
B) 9...b5!
B1) 10.Nxb5 Rb8 11.Qd1 (11.Qa3 Nb6 12.d3 Nfxd5³ is getting very unpleasant already, and the a7-pawn is
very much an irrelevance.) 11...Qa5 12.Nc3 Nb6 13.Be2 Nbxd5³ Black has regained the first pawn and
stands well. He can generate play over the whole board with ideas like ...Ba6 and ...Rfc8, and can consider
playing for ...e5 at some moment too.
B2) 10.Bxb5 Rb8 11.Qc4 (11.Qa3 Nxd5!³) 11...Nb6 12.Qe2 Nbxd5 13.0-0 Nh5
Position after: 13...Nh5
Objectively, Black has nothing more than ‘very full’ compensation. However, it is very easy for White to go
wrong, for instance: 14.d4 Nhf4 15.Bxf4 Nxf4 16.Qe4 (16.Qe3! Bh6 17.Kh1= is unnatural but best)
16...Bh6 17.Rfe1 a6 18.Bf1 Bb7 19.Qxe7 Bxf3 20.Qxd8 Rfxd8 21.gxf3 Rxb2³ Here, the d4-pawn falls very
soon indeed to ...Bg7 ideas, and White will be suffering for a long time with his doubled f-pawns.
7...Bg7 8.d3 0-0 9.Nf3 a6
Setting up the well known cheap trick ...b5 followed by ...Nb6.
10.Qa3
Thus reaching the tabiya. Black has two moves here.
135
Position after: 10.Qa3
10...b6
As played by Grischuk.
10...Nb6!? This may also be playable, but I have not found conclusive equality, and since the position is simple
and the opening non-critical by White, that is what I seek. But it would be remiss of me not to mention a way
Black can roll the dice, since not every player is like me. 11.Qb3 e6!?
Position after: 11...e6!?
A really interesting double pawn sac, courtesy of the Greek grandmaster Mastrovasilis. This is almost winning,
136
regardless of your own rating, if your opponent is below 2300, and very quickly shifts to being rather
inadvisable if they are above 2450. While that might seem cynical to you, it is in fact not an unreasonable way
to assess a move which is intrinsically a gamble. (11...Bg4 12.Ne5 Qc7! 13.0-0!!² was a great find in Bortnyk,
O – Burmakin, V Manacor 2016) 12.dxe6 Nxc4 13.exf7+ Rxf7 14.dxc4 Bg4
Position after: 14...Bg4
Now White needs to preserve his calm and play one of two awkward-looking continuations, or else he quickly
gets into a very unpleasant situation indeed. 15.0-0! (The engines give the implausible maneouvre 15.Be3!?
Bxf3 16.gxf3 Qd7 17.Nb1 but Black gets amazing slow-maturing play after 17...Nh5! 18.Nd2 Nf4 19.Bxf4
Rxf4 20.0-0-0 Qf5©) 15...Bxf3 16.gxf3 Qc8 17.Kg2! A great move, not getting nervous about the g-file but just
calmly protecting that which needs to be protected i.e. the h3-square and f3-point. (17.Ne2 Nd7© is a great
move to know.; 17.Bf4 Qf5 18.Bg3 Qxf3 19.Rfe1 Nh5‚ and since White can’t actually get the queens off,
Black’s play continues.) 17...Qf5 18.Rd1 Re8 19.Nd5!² My sense is that with ideas like Qd3, Ne3 or Bf4-g3 in
the air (depending on
137
Position after: 19.Nd5!²
Black’s response) White is now some way towards neutralising Black’s play, even though Black can keep
trying to make magic happen with, say, ...Nd7.
11.0-0 Bb7
12.Re1
A) 12.Be3 immediately may be even less inspiring due to an extra pawn sacrifice Black has. 12...b5 13.Bb3
138
a5!? Of course ...Nb6 was okay too. 14.Nxb5 Nxd5 15.Bxd5 Else either ...Nb4 with a double attack, or ...
Nxe3 and ...Qb6 with another. 15...Bxd5© Black has more than adeaquate compensation, based chiefly on the
targets on d3 and b2.
B) 12.Bf4?! This bishop move is probably worse than no move at all, despite protecting the d6-point for a
later pawn advance! 12...Rc8! The idea of this move is ...b5 and ...Nc5, and concretely White is in a bit of
trouble. 13.Rfe1 (13.d6 e6µ White has no more answers. When Black has ...b5 and ...Nc5 in, a further ...Nh5
will win the d6-pawn, to say nothing of the embarrassment that could be caused to the White queen by ...b4
ideas.) 13...b5 14.Bb3 Nc5 15.d4 Nd3 16.Qxe7
Position after: 16.Qxe7
16...Bxd5! Not allowing White ...Nxe1 Qxb7, which could be messy. 17.Nxd5 Nxd5 18.Qxd8 Rfxd8µ
12...Re8
139
Position after: 12...Re8
13.Be3
It is pointless to analyse variations where White lets the pawn be simply taken back with ...b5 and ...Nb6,
because unlike the positions from early notes of the last line, Black is nearly fully developed and White’s queen
has been on something of an odyssey.
13.Ng5 Rc8! This should be a familiar idea by now. 14.d6 e6³ Rest assured that the piece sacrifices do not
work and that ...b5 and ...Qb6 can follow.
13...b5 14.Bb3 Nb6
14...a5 is less interesting here for an odd reason: the f7-pawn is no longer defended twice! 15.Nxb5 Nxd5
16.Ng5! and while White is still far from neutralising the compensation, he has been able to solve the issue of
the f3-knight with tempo.
15.Bxb6
15.d6
140
15...Bxf3! (15...exd6 is given by Houska, but after 16.Ng5 d5 17.Qb4 Black’s pieces don’t work well together
and there is a risk of a dark-squared bind.) 16.dxe7 Rxe7 17.gxf3 Nbd7 It might appear that Black wants ...Ne5,
but in fact the real idea is ...Bf8. White’s best way out of both is 18.Qd6 but after 18...Ne5 19.Qxd8+ Rxd8
20.Kg2 Nxd3 21.Red1 Red7 22.Rab1 Nh5ƒ he is nevertheless in trouble, since the b2-pawn hangs for tactical
reasons.
15...Qxb6 16.Qb4
Position after: 16.Qb4
141
16.Rxe7 b4!–+
16...Qd8
So far we have followed Bortnyk, O – Grischuk, A ICC INT 2016.

16...Qd6 was played earlier between the same two players, but 17.Qxd6 exd6 18.a3 Nd7 19.d4 seems destined
to lead to rook trades, after which Black will be suffering with his pawn deficit since he lacks ways to attack d5.
17.Rac1N
Trying to be clever about the return of the pawn.

A) 17.Re5 is a little bit unsound, but interesting. 17...Nd7
A1) 18.Rg5?! Rc8! Best, focusing on ...Nc5 and eliminating the b3-bishop rather than accumulating more
material. (18...Bf6 19.Qh4 Kg7³ and although Black stands marginally better, White’s pieces were
aggressively enough placed for him to whip up a storm and win in Bortnyk, O – Kononenko, D Tallinn
2016.) 19.h4 Nc5 20.h5 Nxb3 21.Qxb3 e6³ White finds himself in quite a precarious position.
A2) 18.Re3 Nb6 19.d4 Rc8 (19...Bf8!³) 20.a4 Nxd5 (20...Nc4©) 21.Nxd5 Bxd5 22.axb5 axb5 23.Qxb5
Black’s compensation slowly evaporated in Varga, Z – Lizak, P Budapest 2015.
B) 17.a4?! has been a major choice in practical games. 17...Nxd5
142
Position after: 17...Nxd5
B1) 18.Nxd5 does net a pawn in the end, but White’s position does not inspire confidence. 18...Bxd5
19.axb5 axb5 20.Qxb5 Bxf3 21.gxf3 Rxa1 22.Rxa1 e6³ All of White’s pawns are isolated and he lacks any
sort of way to simplify the position, while Black can continue with ...Rf8 and ...Qf6 and take material back at
leisure.
B2) 18.Bxd5 Bxd5 19.axb5 axb5 20.Qxb5?! Bxf3 21.gxf3 Rb8 22.Qc6 Rxb2µ Bortnyk, O – Grischuk, A
ICC INT 2016.
C) 17.Qc5?! Rc8 18.Qa7 Rc7³
17...Rc8
17...Nxd5 18.Qe4! e6 19.Nxd5 exd5 20.Qxe8+ Qxe8 21.Rxe8+ Rxe8 22.d4 looks psychologically rather
unpleasant and certainly in a game between equally matched humans Black would consider himself to have
worse chances here.
143
Position after: 17...Rc8
18.Qh4
The last attempt to keep the pawn tactically.

A) 18.a4?! Nxd5 19.Nxd5 Rxc1 20.Rxc1 (20.Nxe7+?! Qxe7 21.Qxe7 Rxe7 22.Rxc1 bxa4 23.Bxa4 Bxb2³)
20...Bxd5 21.axb5 Bxf3 22.gxf3 e6!‚ Delabaca, R – Postny, E Paleochora 2008.
B) 18.Ne5 Qd6! 19.Qxd6 exd6 20.Nc6 Rxe1+ 21.Rxe1 Bxc6 22.dxc6 Kf8!= Black has completely equalised
and a purist might even prefer his position due to the c-file being more relevant than the e-file.
18...Bxd5
A) 18...Nxd5 19.Ng5 e6 20.Qxh7+ Kf8© is a little hair-raising, but playable. It is, though, still unclear what
Black is really doing about the pawn deficit.
B) 18...h6!?= could be another option, really trying to slow down and regain the pawn only when it is
maximally convenient.
19.Nxd5 Rxc1 20.Nxe7+ Rxe7 21.Rxc1 Qxd3=
144
Position after: 21...Qxd3=
Black has regained the pawn and White does not have meaningful tactical tries to prove an edge from here.
4.3) 5.Nc3 Nxd5 6.Nf3 Nc6 7.Bb5 g6 sidelines
1.e4 c6 2.c4 d5 3.cxd5 cxd5 4.exd5 Nf6 5.Nc3
Position after: 5.Nc3
The only genuinely ambitious way for White to return the pawn is with this move.
145
5...Nxd5 6.Nf3
A) 6.Qb3 Nb6!? A maximalist move, though there is also ...e6 or ...Nf6. 7.Nf3 Be6=
B) 6.Bc4 Nb6 7.Bb3 Nc6 8.Nf3 Bf5 9.d4 e6= is also not very threatening; White will manage the d5-break
but it will only serve to vacuum off all the pieces.
6...Nc6 7.Bb5
Played to try and make Black play ...e6 with his c8-bishop still at home. However, I am going to recommend
avoiding that anyway.
7.d4 Bg4 transposes to the Panov Endgame, as considered in line B12.
7...g6!?
Position after: 7...g6!?
‘Doch!’ as the Germans would say. This move entails a pawn sacrifice.
8.Qa4
A) 8.0-0 Bg7 is not a helpful insertion from White’s point of view: after 9.Qa4 0-0 10.Bxc6 Nb6 his best
choice is to return to the main line with Qh4. 11.Qh4 (11.Qb5?! bxc6 12.Qxc6 Bf5³) 11...bxc6 12.d4 f6 See
move 12 of the main line.
B) 8.Qb3 a6! This is a nice ‘clever’ move. (The alternative 8...Be6 9.Nd4 led to a less desirable position in
one recent heavyweight encounter.) 9.Ba4
146
Position after: 9.Ba4
9...Nb6!N (9...Nf4 10.d4 Nxg2+ 11.Kf1 b5∞ was suitably bizarre in the game Lagarde, M – Piorun, K
Cappelle-la – Grande 2014, and could be interesting to analyse for fun, but it is a little unnecessary.) 10.Ne5!?
(Normal moves which lose the bishop pair either on c6 or a4, such as 10.Ng5 do not pose Black any real
problems, for instance: 10...e6 11.Bxc6+ bxc6 12.d4 Bg7 13.0-0 0-0 14.Nf3 Qc7=) 10...Be6 11.Nxc6 Bxb3
12.Nxd8+ Bxa4 13.Nxa4 Nxa4 14.Nxb7 Bg7=
Position after: 14...Bg7=
Black can regain the pawn at a time of his choosing; however it may in fact be even stronger to simply place
the rooks on c8 and b8 without taking anything.
147
8...Nb6 9.Qh4
We shall see the acceptance of the pawn with 9.Bxc6 next, and no other move makes any sense at all.
9...Bg7 10.d4 0-0
Position after: 10...0-0
11.Bxc6
11.Bh6!? leads to interesting but minor complications and a (non-compulsory) queen sacrifice. 11...Nxd4
12.Nxd4 Bxh6 13.Rd1 Bg7! (There was also 13...e5 when White obtains compensation for a pawn but not
more.) 14.Ne6 (14.Nc6 Bxc3+ 15.bxc3 Qc7³) 14...Bxe6 15.Rxd8 Raxd8 16.0-0 a6³ White is in some trouble.
11...bxc6 12.0-0
12.Bh6 is also principled. 12...Bf6 13.Qf4 (13.Bg5 Ba6!µ led to a smooth win for the talented player on the
Black side of Szabo, K – Csonka, B Hungary 2013) 13...Re8 14.0-0 Nd5 15.Nxd5 cxd5= Black has equalised
and his development in what follows looks relatively easy: ...Qb6, ...Bf5, ...Rac8.
12...f6!?
12...Re8 13.Re1 Nd5 is also playable as in Pelletier, Y – Sebag, M Cap d’Agde 2013.
13.Bh6
Otherwise Black’s play is self-explanatory: ...Nd5, ...Bf5, ...Rb8 et cetera.
13...g5! 14.Qh5 Qe8 15.Qxe8 Rxe8 16.Bxg7 Kxg7=
148
Position after: 16...Kxg7=
White’s slightly better development and structure is balanced by the advantage of the bishop over knight and
Black’s control of the d5-square.
4.4) 5.Nc3 Nxd5 6.Nf3 Nc6 7.Bb5 g6 8.Qa4 Nb6 9.Bxc6+
1.e4 c6 2.c4 d5 3.cxd5 cxd5 4.exd5 Nf6 5.Nc3 Nxd5 6.Nf3 Nc6 7.Bb5 g6 8.Qa4 Nb6 9.Bxc6+ bxc6
10.Qxc6+
Position after: 10.Qxc6+
149
10...Bd7
Note that 10...Qd7 is not entirely without merit, but my feeling is that Black may be walking a finer line there.
11.Qe4
This has been almost unanimously chosen.

11.Qc5 Bg7 12.0-0 0-0 13.Re1 Re8 14.d4 Rc8 15.Qg5 Bc6³ Melero Morales, J – Cenal Gutierrez, R Asturias
1995.
11...Bf5!?
This is maximalism, trying to play ...e5 and also stop or delay White’s castling.
12.Qe5
12.Qe3 Qd3 is a great bind for Black, as long as he makes sure not to let White improve his structure too much.
13.h3 (13.b3 Bg7 14.Ba3 0-0³) 13...Bg7 14.g4 Qxe3+ 15.fxe3 Bd3 16.b3 Nd5©
Position after: 16...Nd5©
White›s problems with passivity are not going away any time soon.
12...f6
150
Position after: 12...f6
13.Qb5+
13.Qe3 Nd5 14.Nxd5 Qxd5 15.d4 Qb5ƒ
13...Kf7 14.d4
14.Qb3+ Be6 15.Qc2 Bc4³ is exactly what Black wants.
14...Be6! 15.Be3
15.b3 prevents ...Bc4, but this wasn’t Black’s only idea. 15...Rc8 16.Bd2 Bg4! 17.Qd3 Bxf3 18.gxf3 e5 19.d5
Bc5=
White cannot hold on to the pawn any longer.
15...Rc8 16.Nd2
151
White indicates a wish to castle despite the possible cost.
16...Nd5 17.0-0
17.Nde4 Nxc3 18.Nxc3 Bc4 19.Qa4 Bg7‚ is likely to end in tears for White since Black will complete his
development on the very next move.
17...Nxc3 18.bxc3 Rxc3 19.Ne4 Rc7=
Black has regained his material and has a solid position.
152
Position after: 19...Rc7=
The first and third lines of this chapter were relatively easy to ‘de-fang’ and it only takes a few moves.
However, right in the middle of the two is the subtle line with 5.Qa4+, a speciality of the Ukrainian GM
Olexandr Bortnyk. In exchange for a relatively large corpus of theory to memorise, Black gets reasonable
winning chances, courtesy of the imbalance in the position (extra doubled pawn for bishop pair.) In my
opinion, it is important to memorise the moves up to and including Black’s 16th (16...Qd8!) and have some
idea what is going to happen after that. The internet games between Bortnyk and Grischuk are important,
as is the model game (for Black) Delebaca-Postny.
The final line of this chapter is rather different and offers the first example in the book of what I have come
to think of as the most basic imbalance in the Caro-Kann: White has a proper extra pawn (not doubled) but
Black has the bishop pair and possibly better coordination in an open position. Unfortunately, in this early
example, we don’t get to keep the imbalance for long, because Black regains the pawn, though this time it is
at the cost of king safety and development rather than the bishop pair. The final position is at once
dynamically balanced and drawish – it depends on the level of the players concerned!
153
Chapter 5
Exchange and Panov-Botvinnik variations
1.e4 c6 2.d4 d5 3.exd5 cxd5
Chapter guide
Chapter 5 – Exchange and Panov-Botvinnik variations
1.e4 c6 2.d4 d5 3.exd5 cxd5
5.1) Exchange Variation: 4.Bd3 and other alternatives to 4.c4
5.2) 4.c4 Nf6 5.Nc3 Nc6: 6.Bg5 minor lines
5.3) 4.c4 Nf6 5.Nc3 Nc6: 6.Bg5 dxc4 7.d5
5.4) 4.c4 Nf6 5.Nc3 Nc6: 6.Nf3 & “Panov endgame”
5.5) 4.c4 Nf6 5.Nc3 Be6!?: intro and 6.c5 (knight sac declined)
5.6) 4.c4 Nf6 5.Nc3 Be6 6.c5 (knight sac accepted)
5.7) 4.c4 Nf6 5.Nc3 Be6 6.Qb3
5.8) 4.c4 Nf6 5.Nc3 Be6 6.Nge2!?
This rather long chapter (possibly the most substantial in the whole book) deals with White’s options after
3.exd5 cxd5. Clearly, it is possible to transpose into these lines from the previous chapter (though doing it the
other way round is difficult) and indeed there are substantial ideological similarities.
The Exchange Variation (4.Bd3) has never been particularly popular, primarily because even if White is able to
secure the d4-square as an outpost following a Black ...e5 break, he will have to pay the price, which is
generally the bishop-pair. By following the order I recommend here (4...g6!) Black keeps the choice between
154
entering the just-mentioned type of going, and simply following up with ...Bf5.
In the following three lines, we examine the absolute main lines of the first part of this book, characterised by
the further moves 4.c4 Nf6 5.Nc3 Nc6. First is the gambit continuation 6.Bg5 dxc4 7.Bxc4: while in general I
give alternatives to gambit acceptance, here the gambit is reasonably easily neutralised, and we swiftly move on
to the main continuation 7.d5, a memory-intensive variation replete with pawn sacrifices and other imbalances.
I think Black should choose between 10...e5 and 10...e6, and then learn a further 6-8 moves against White’s
major possibilities, while not paying a great deal of attention to the engine evaluations (they aren’t useful in all
sharp positions!) Finally, we deal with the Panov Endgame and the deviations along the way to it: this is
possibly the ‘tabiya’ with the highest draw rate that isn’t actually deliberately played with the aim of short
draws in mind.
Then, because that wasn’t enough work for either me or you, I give you another choice: the bizarre and original
5...Be6, which has been adopted a fair amount but not enough for it to be termed a main line in its own right.
White has no fewer than three critical lines, and in the first I bring to you a knight sacrifice that almost defies
computer analysis, though I have given it a good try. I believe the idea to be novel. With the aim of fostering
creativity in the Caro-Kann I have given a fairly comprehensive overview of 5...Be6 and would very much
appreciate it if some readers would take over the carrying of the baton. Chess is for enjoying, after all!
5.1) Exchange Variation: 4.Bd3 and other alternatives to 4.c4
1.e4 c6 2.d4 d5 3.exd5 cxd5
Position after: 3...cxd5
4.Bd3
The most major of White’s choices other than c4. Here, I shall deviate somewhat from Black’s most played
option and elaborate instead on one that goes directly to the piece configuration that he wants.
4.Nf3 Bf5!?
155
A) 5.c4 e6 6.cxd5 Qxd5 The pawn recapture is also playable, but could require some precision as White has
an extra tempo in a fully open and otherwise symmetrical position. The critical continuation is: 7.Nc3 Bb4
8.Bd2 Bxc3 9.bxc3 Nf6 10.Be2 h6= Black has acheived equality (...Nc6 and ...0-0 cannot be stopped by
normal means) and the momentum in the position is with him.
B) The maximalist continuation, seeking to exchange the ‘bad’ bishop for White’s f1-bishop rather than the
f3-knight. There are, as far as I can see, two significant answers. 5.Ne5 Nd7
B1) 6.Nc3 Ngf6 7.Be3 (7.g4 Be6 8.g5 is creative, but after 8...Ne4= White has no advantage, either in
development or otherwise; Black will be able to fianchetto his f8-bishop.) 7...g6 8.Be2 Bg7 9.0-0 h5!= Black
has equalised.
B2) 6.Bb5!? Ngf6 (6...a6?! 7.Bxd7+ Bxd7 8.0-0ƒ demanded accurate answers of Black in Al Hindi, A –
Piscitelli, J Mexico City 1980; White wants Qf3 and c4 against most moves.) 7.0-0 a6! (7...e6 8.Bg5±
Hidalgo Santana, F – Damaso Tacoronte, C Santa Cruz de Tenerife 2011) 8.Bd3 (8.Bxd7+ Nxd7 is nothing,
unless White goes for the speculative 9.Nxf7!? Kxf7 10.Qf3 e6 11.g4, however here too the solid 11...Qf6!=
is easily good enough.) 8...Bxd3 9.Nxd3 e6= Black has equalised and a purist could argue that only Black
could stand better, due to the better bishop.
4...g6!?
4...Nc6 is most played, and indeed I have played it twice.
156
The issue I have with the move is a stylistic one: the extra option Black gains (...e5) is shown in the following
play to rarely be of much use, while the option lost (of developing this knight to d7) could well be valuable.
5.c3 Nf6 (5...e5?! 6.dxe5 Nxe5 can be considered, but the weakness of the IQP is likely to outweigh any
temporary activity, for instance after 7.Qe2 Qe7 8.Be3 Nxd3+ 9.Qxd3 Be6 10.Ne2ƒ)
A) 6.Bf4 Bg4 (6... g6!) 7.Qb3 Qd7 8.Nd2 e6 9.Ngf3 Bxf3 10.Nxf3 a6 11.0-0 Be7 12.Rae1 0-0 13.Qc2 Rfc8
14.Ne5² Black’s position was solid, but uninspiring. After further outplaying me in the next 20 moves my
opponent collapsed in a rook endgame in Jarmany, J – Fernandez, D Aberystwyth 2014.
B) 6.Bg5!? is objectively not an issue, but stylistically it poses a problem if Black is not ready for the bishop-
pair-for-f-pawns strategic imbalance.
C) 6.h3 e5
157
Now this is playable, although ...g6 should still be preferred on principle. 7.dxe5 Nxe5 8.Nf3!? The move that
asks the most searching questions about Black’s structure. (8.Qe2 Ne4!„ is Black’s point- a pawn sacrifice
yielding very convincing play.; 8.Bc2 Bc5 9.Ne2 0-0 10.Nd2 Re8„ was comfortable for me in Satyapragyan,
S – Fernandez, D Parramatta 2009) 8...Nxd3+ 9.Qxd3 Be7 10.0-0 0-0 11.Be3 White has good results and
Black lacks actual plans, see for instance Alekseev, E – Bocharov, D Khanty-Mansiysk 2016.
5.Nf3
This is never a particularly amazing move for White to play, but it does demand a deviation from the standard
...Bf5 response, and so seems to be the least worst choice.
A) 5.c3 Bf5! 6.Bxf5 gxf5 7.Nf3 e6= is completely harmless.
B) 5.Bf4
158
can be met with the unorthodox 5...Bh6!? recognising that in fact both of White’s bishops are marginally
better than their counterparts, and so both bishop trades favour Black. Now for instance: 6.Bxh6 (6.Qd2 Bxf4
7.Qxf4 Qb6=) 6...Nxh6 7.Qd2 Nf5 8.c3 0-0 9.Nf3 Nc6 10.0-0 f6 11.Re1 Nd6= Next can be either a standard
minority attack plan or ...Bf5/g4.
C) 5.Ne2 is met as usual by 5...Bf5!
6.Bxf5 gxf5 7.Bg5 Nd7 and development will proceed as normal with ...Ngf6, ...e6, and either ...Rg8 or
159
...Bg7/...0-0. 8.Qd3 Qb6 9.0-0 e6 10.Nbc3 Ngf6=
D) 5.Qf3 Nc6 6.c3 sees White risking being worse while trying to stop ...Bf5: 6...h5= (6...Bg7!?).
5...Nc6
So Black has lost the flexibility regarding where to develop this knight, but White has also committed his g1-
knight.
6.0-0!
The only move which doesn’t allow any of Black’s desired bishop trades to be carried out immediately.
A) 6.Bf4 Bg4 7.c3 Bh6=
B) 6.c3 Bf5=
6...Nf6
6...Bf5 7.Bxf5 gxf5 8.c4±
7.Re1 Bg7 8.Bf4 0-0
160
Now both sides have exhausted their useful moves, and White will have to commit on the queenside.
9.c3
Otherwise the queenside pawn formation gets attacked easily.

A) 9.h3 Qb6 wins a pawn while not actually falling behind much in development; following 10.Nc3 Nxd4
11.Nxd4 Qxd4 12.Qf3 Be6= White needs to think very seriously about how to equalise, although it is still
possible.
B) 9.Nbd2 Nb4 wins the bishop pair thanks to a fork motif on c2, then play could continue 10.Nf1 Nxd3
11.Qxd3 b6= and Black has equalised.
9...Bf5! 10.Bxf5 gxf5
161
There is little point analysing this further, or even evaluating it, but I will comment that from a human
perspective Black has a coherent (and relatively easier) position and the momentum is with him.
5.2) 4.c4 Nf6 5.Nc3 Nc6: 6.Bg5 minor lines
1.e4 c6 2.d4 d5 3.exd5 cxd5 4.c4
Reaching the main position of the proper (i.e. not ‘pseudo’) Panov-Botvinnik Attack. Note that White could
162
equally well have played 2.c4 and 4.d4.
4...Nf6 5.Nc3
This is the traditional order of development; moving the other knight first gains little but cues up an extra ...Bg4
pin option.
5...Nc6
There are now two moves worth examining: Nf3 and Bg5. A few minor alternatives can be quickly dispatched.
6.Bg5
A) 6.c5 e5! 7.dxe5 Ng4 8.Qxd5 Qxd5!? Black gets compensation for his pawn after the pedestrian 8...Be6
too.
9.Nxd5 Bxc5
10.Nc7+ Ke7 11.Nxa8 (11.Nh3? Bb4+ 12.Bd2 Bxd2+ 13.Kxd2 Rd8+ 14.Ke1 Rb8³) 11...Bxf2+ 12.Ke2
Nd4+ 13.Kd1 Bf5© White is in huge trouble and Black›s attack plays itself.
B) 6.cxd5 Nxd5 7.Bc4 (7.Nf3 Bg4 is the next subchapter) 7...Be6! 8.Bb3 g6= Black has a perfectly sound
position. His subsequent play will probably involve ...Bg7, ...0-0, ..h6, ...Qd7, ...Rac8 and ...Rfd8.
6...dxc4
163
7.Bxc4
The main move is 7.d5, seen next.

7.Nf3!? h6 8.Bxf6 exf6 9.Bxc4 Bd6 This line is something of a modern fashion, having been employed by both
Giri and Aronian. 10.0-0 (10.d5 Ne5 11.Bb5+ Bd7=) 10...0-0 11.h3 Otherwise Black’s play is too easy.
11...Bf5 12.Bd3
Simplest, and stopping any tricks based on ...Nb4, but this doesn’t give White anything either. 12...Qd7
164
13.Bxf5 Qxf5 14.Qb3 Rab8 (14...Na5 15.Qa4 Rfc8= was also fine in Ivanchuk, V – Aronian, L Nice 2009)
15.Rfe1 Rfd8 In the play that follows, Black is always rock-solid even if one feels White should be getting
somewhere. 16.Rad1 a6 17.a3 Bf8 18.d5 Ne5 19.Nxe5 fxe5 20.Ne4 Qg6 21.Qf3 Rd7 22.h4 h5 23.Ng3 Bd6
24.Nf5 Re8 25.Re4 f6 26.Kh1 Qh7 27.Rc4 g6 28.Nxd6 Rxd6 29.Rdc1 Qf7 30.Rc7 ½-½ Svidler, P – Giri, A
Germany 2010.
7...Qxd4
7...h6 8.Be3 is fine for Black, but he has very little choice other than to play ...e6 either now or very soon,
which is against the spirit of my repertoire.
8.Qxd4
This has been nearly universally played.

A) 8.Nb5 Qb6 doesn’t leave White a lot of options other than a Be3-d2 repetition. 9.Nf3 (9.Qb3 e6³ and
Black will get developed with ...Bb4 and ...a6 soon; 9.Be3 Qa5+ 10.Qd2 Nb4!³) 9...Be6!
Position after: 9...Be6!
Now White can obtain just enough compensation in a few ways, but cannot seriously trouble Black. 10.Bf4
(10.Qe2 a6 11.Bxe6 axb5 12.Bb3 e6=) 10...Rd8 11.Qc2 (11.Qe2 Bxc4 12.Qxc4 Rd7 13.0-0 a6 14.Nc3 e6³)
11...Nd5! 12.Bxd5 Qxb5 13.Bxe6 fxe6 14.Bd2 g6© Unfortunately, to stop Black from castling White has to
give up that right himself.
B) 8.Qb3 e6 9.Nf3 (9.Nge2 is similar but allows Black the added option of 9...Qg4) 9...Qb6 10.Qxb6 axb6
11.Nb5 Bb4+ 12.Ke2 0-0= Black’s development cannot be stopped, though White is probably able to recover
the b6-pawn in some way.
C) 8.Qe2 Qg4!
165
Position after: 8...Qg4!
9.Nf3! (9.Nb5 Qxg5 10.Nc7+ Kd8 11.Nxa8 Qa5+ 12.Kf1 Bd7µ is worth noting) 9...Be6 10.Bxe6 Qxe6
11.Qxe6 fxe6 12.Nb5 Nd5= White has compensation, but his bind will not last forever; Black can free
himself with the fianchetto.
8...Nxd4 9.0-0-0 e5
10.f4!
166
A) 10.Nge2 Be6! 11.Nxd4 Bxc4 12.Rhe1 is not a critical test; for instance, 12...0-0-0 13.Rxe5 Ng4= will do.
B) 10.Nf3 merely requires Black to keep his nerve a little. 10...Nxf3 11.Bxf6 (11.gxf3 Be6 12.Bxe6 fxe6
13.Rhe1 Bb4³) 11...gxf6 12.Bb5+ Ke7 13.Nd5+ Kd6! 14.gxf3 Bh6+ 15.Kb1 Bf5+ 16.Bd3 (16.Ka1 Bc2µ)
16...Be6 17.Be4 Rad8!³
Position after: 17...Rad8!³
White will be unable to regain more than a pawn, and his best case is to enter an endgame with a rook and
bishop each of similar quality, but doubled f-pawns for him.
10...Bg4 11.Nf3 Bxf3 12.gxf3
167
Position after: 12.gxf3
12...0-0-0
12...Rc8 was the main move, but I have been unable to find good reasons to take Black’s side after 13.fxe5
Rxc4 14.exf6 Ne6 15.Rhe1 g6 16.Kb1².
13.fxe5
13.Nb5 is interesting, but a picturesque wrinkle makes Black’s position equal: 13...h6 14.Bh4 Nxb5 15.Rxd8+
Kxd8 16.fxe5 Nd6! 17.exd6 Bxd6 18.Bxf7 Kc7=.
13...Nxf3 14.Rxd8+ Kxd8
I have not been able to find an edge for White here in any of the lines, which are given in no real order.
15.Bxf6+
A) 15.exf6 Nxg5 16.h4 Ne6 17.Rf1 gxf6 18.Rxf6 Kc8! 19.Rxf7 Nd8 20.Rf1 Bg7=
B) 15.Bf4 Nh5 16.Rf1 Nxf4 17.Rxf3 Be7!=
15...gxf6 16.exf6 Ne5
17.Bd5
17.Bb3 Rg8 18.Bc2 h5 19.Re1 Bh6+ 20.Kb1 Ng4 21.Ne4 Bg5=
17...Kc7 18.Rg1 Kb8!=

168
This last move avoids, would you believe, forks with Ne8 in case White ever sacrifices an exchange on g7.
Black can now proceed with normal development: ...Bh6, ...Rd8, ...a6, ...b5.
5.3) 4.c4 Nf6 5.Nc3 Nc6: 6.Bg5 dxc4 7.d5
1.e4 c6 2.d4 d5 3.exd5 cxd5 4.c4 Nf6 5.Nc3 Nc6 6.Bg5 dxc4 7.d5
This is deservedly the main move.
7...Ne5 8.Qd4
A) 8.Nf3 This is a less significant try that demands only one really precise move from Black. 8...Nxf3+
9.Qxf3 h6
A1) 10.Bh4?! Bg4µ
A2) 10.Bxf6 exf6 11.Bxc4 Bd6 is alright for White, but he needs to play respectfully and keep the bishop
pair under control. He is not better. A bad continuation would be 12.Bb5+?! Kf8³ and Black has only been
encouraged to continue with his desired construction of ...g6, ...Kg7, and gifted an extra tempo for ...a6
besides.
A3) Not the only choice, but best. Black needs to take a liberal attitude to the weakening of his kingside
structure in this line! 10.Be3 e5
169
11.Qg3!? (11.Bxc4 Bd6 will either transpose or allow Black a straightforward short castling.) 11...Bd6! A
crucial tactic enables Black to develop anyway. Without it he is significantly worse. 12.Bxc4 (12.Qxg7 Rg8
13.Qxh6 Rg6= leads to a perpetual attack on the queen.) 12...e4! (12...Nh5?! 13.Bb5+!² Anything else
allows Black a hassle-free short castling, but this gets an edge.) 13.Bb5+ Ke7
Position after: 13...Ke7
14.Bf4 (14.Qxg7?? Rh7–+) 14...Bxf4 15.Qxf4 Qd6 16.Qe3 Rd8 17.Rd1 Kf8= Next there will follow ...a6
and ...Bf5 and Black is fine.
170
B) 8.Bxc4!? This recovers the pawn, but that is really all that can be said for the move. 8...Nxc4 9.Qa4+ Bd7
10.Qxc4 h6!
Position after: 10...h6!
This move again! 11.Bf4 (11.Be3 e5= leads either to a similar lifting of the central bind or to Black’s desired
construction with ...Bd6 and ... 0-0.) 11...e6 12.dxe6 (12.d6?! is principled, but the pawn will not survive:
12...Bc6 13.0-0-0?! Rc8!µ) 12...Bxe6 13.Qb5+ Qd7 14.Qxd7+ Nxd7= Black has ceased to be under pressure.
C) 8.Bf4 Ng6 9.Bg3 e5= is similar to the main line, but White having fewer options.
8...h6
But now there is a real parting of the ways.
171
Position after: 8...h6
9.Bf4
A) 9.Qxe5 hxg5 10.Bxc4 Qb6! Black sets up the ‘cheap’ ...Qxf2.
Position after: 10...Qb6!
A1) 11.Bb5+ Bd7 12.Bxd7+ Nxd7 13.Qe2 Rc8³

A2) 11.0-0-0 is not really where White wants to be castling: 11...Bd7 12.d6 Rc8! 13.Bb3 Rc5 14.Qe3 Ng4µ.
A3) 11.Nb5 Qxf2+ 12.Kxf2 Ng4+ 13.Ke2 Nxe5 14.Nc7+ Kd8 15.Nxa8 Nxc4 16.Nf3 g6³ A little further
172
analysis reveals that the a8-knight is toast, meaning Black restores material equality, but must be better due
to White’s bad king and his own bishop pair.
A4) 11.Nf3 The only move to cause Black any trouble. 11...g4 12.Ng5 Bd7!
Not rocket science, but Nb5 was by now a serious threat, and this is the best move to parry it. 13.0-0 (13.d6
e6 14.Nb5 Rc8 Interestingly, there was also long castling, but that might be one adventure too many.
15.Nc7+ Rxc7 16.dxc7 Bd6 17.c8=Q+ Bxc8 18.Qb5+ Bd7 19.Qxb6 axb6 20.h4 Bc6³) 13...Rh5 Threatening
...Nh7. (How often does one get to write that?!) 14.Qf4 (14.Bd3 e6 15.dxe6 Bd6 16.exd7+ Kf8µ) 14...Rc8
15.Bd3 Qc7 16.g3 Qxf4 17.gxf4 g6„
B) 9.Bh4 Ng6 10.Nf3 (Notice that 10.Bg3 transposes to the main line.) 10...Nxh4 11.Qxh4 g5
173
There was little wrong with ...g6 as well, but it makes sense to avoid Bb5+ if possible. 12.Qxc4 Bg7 13.Rd1
White should take measures to delay ...e6 if possible. A sample continuation-possibly not the best- could be:
13...0-0 14.Be2 Qd6 15.0-0 a6³ Black will complete development on the next move and stand slightly better
due to the bishop pair. In this line, ...g5 is not particularly a weakening move.
9...Ng6
10.Bg3
174
Now Black can choose to sacrifice a pawn in either of two ways, and both seem fine.
10.Bxc4 Nxf4 11.Qxf4 g5 I also analysed 11...a6, but a recurring theme was that when Black did get around to
playing ...g6(5) there would follow Qd4 and Ne4. The following game should illustrate Black’s play nicely.
12.Qd2 Bg7 13.Nge2 0-0 14.0-0 e6 15.d6 Bd7 16.f4 Qb6+ 17.Kh1 Qc5 (17...Ng4!?³) 18.fxg5 hxg5 19.Bb3
Bc6 20.Ng3 Rad8 21.Rad1 Rd7 22.Qc1 Rfd8‚ Yilmaz, M – Shimanov, A Konya 2011.
10...e6
This has been chosen in some fairly high-profile encounters.

10...e5!? 11.Bxe5 (White is morally obligated to accept the challenge; 11.Qxc4 a6!? followed by the swift
development of both bishops and a possible ...Nh5 is not a good outcome from the opening for him.) 11...Nxe5
12.Qxe5+ Be7
A) 13.Bxc4?! 0-0 14.Nf3 (14.0-0-0 compounds the error: 14...Ng4 15.Qe2 Bf5 16.Bd3 Re8 17.Qc2 Bxd3
18.Rxd3 b5–+) 14...Bd6 15.Qd4 Re8+ 16.Be2 b5 17.a3 Bb7 18.0-0 a6 19.Rad1 Rc8³ Black looks to have
more than sufficient compensation for the pawn, which in any case he threatens to regain in a variety of ways.
B) 13.0-0-0 Ng4 14.Qxg7 Bg5+ 15.Kb1 Bf5+ 16.Bd3 Bxd3+ 17.Rxd3 Bf6 18.Qxg4 cxd3 19.Nf3© My
problem with 10...e5 is that I believe White’s compensation here; Black’s king position looks decidedly
worrying.
175
11.d6!?
11.Bxc4 is safer, if a little unambitious: 11...exd5 12.Rd1 Be7 13.Bxd5 0-0 14.Nge2 Nxd5 15.Nxd5 Be6=.
11...Ne7
11...h5!? should be very seriously considered. Certainly the intuitive continuation for White leads to a
disadvantage: 12.Nf3 h4 13.Nxh4 Nxh4 14.Bxh4 Qxd6 15.Qxc4 Bd7³.
12.Rd1
A) 12.Qxc4 Nf5 13.Nb5 Bxd6 14.Nxd6+ Nxd6³ leaves White with insufficient compensation.
B) 12.0-0-0 Nf5 13.Qxc4 Bd7 14.Nb5 There is little else that can be advised against ...Rc8, ...Qb6 and the
slow retrieval of the d6-pawn. 14...Bxb5 15.Qxb5+ Qd7 16.Nf3 a6
176
17.Qb4? White had to go for the ending, but it was unpleasant in any case. Now 17...Rd8µ would have
improved slightly over Fougerit, V – Demuth, A Chalons en Champagne 2015 which was already a
convincing performance.
12...Nf5
12...Ned5 was less convincing in Anand, V – Seirawan, Y Amsterdam 1992.
13.Qxc4 Bd7
177
14.Qb3
A) 14.Nf3 again allows Black an optimal configuration, which is hard to shake even by violent tactical
methods: 14...Qb6 15.Nb5 Rc8 16.Nc7+ Kd8 17.Ne5 Nxd6 18.Rxd6 Bxd6 19.Nxf7+ Ke7 20.Nxd6 Rxc7
21.Nf5+ exf5 22.Bxc7 Qxb2µ.
B) 14.Nb5 Bxb5 15.Qxb5+ Qd7 16.Nf3 a6 is similar to the Fougerit – Demuth game above, and again
White’s best here is 17.Qxd7+ Nxd7 18.Bf4 g5 19.g4= when he should maintain the balance.
14...Nxg3 15.hxg3 Bc6 16.Bb5 Bxd6 17.Bxc6+ bxc6 18.Qb7 0-0 19.Qxc6 Nd5©
Black has completed his development. White now has a choice between accepting a pawn and coming under
fire on the e-file, and just finishing his own development and hoping the bishop is not too much superior to his
knight in what follows.
5.4) 4.c4 Nf6 5.Nc3 Nc6: 6.Nf3 & “Panov endgame”
1.e4 c6 2.d4 d5 3.exd5 cxd5 4.c4 Nf6 5.Nc3 Nc6 6.Nf3 Bg4
As very often in this book, I am recommending a path which does not imply knowledge of the standard IQP
positions. If the reader is comfortable with that structure, he can feel free to play 6...e6. (However, it will be
impossible to completely avoid such structures – White can often obtain versions which are less favourable for
him even within our repertoire.) Also, I would like to commend the work done on 6...g6 by IM Houska; that
move is fully playable also.
7.cxd5
178
A) 7.Bg5!? This is a minor move but nevertheless noteworthy. 7...Qa5! More adventurous than the slightly
tired 7...e6, which can lead to marginally worse symmetrical positions. Meanwhile, I don’t trust the line
7...Ne4 8.cxd5 Nxg5 9.dxc6 either, and think it favours White. 8.Bxf6 Otherwise there was little point in Bg5,
as Black will have ...Ne4 with tempo soon. 8...exf6 9.Qe2+ (9.Be2 dxc4 10.Nd2 Be6 11.Nxc4 Bxc4 12.Bxc4
0-0-0= Li, R – Shimanov, A Saint Louis 2017) 9...Kd8 10.Qd2 (10.cxd5?! Bb4 11.0-0-0 Ne7µ is close to
being winning for Black!; 10.0-0-0?! Nb4 11.a3 dxc4!µ) 10...Bxf3 11.gxf3 Bb4 12.0-0-0 Ne7
The position is difficult for both sides, but it is easier to see what Black should play next (...Rc8; answer c5
with ...b6 or Rg1 with ...g6.) Tests are needed here but my preliminary analysis indicates that Black is fine.
B) 7.Be2 e6 8.0-0 Be7= is deservedly unpopular for White, as Black has far more useful moves he can make
prior to the inevitable release of central tension.
C) 7.Be3 e6 8.c5 Be7= is similar; after castling Black will be able to play challenging pawn moves on the
queenside.
7...Nxd5 8.Qb3
This is most critical and almost always played.

8.Bc4 e6 9.0-0 Be7= is not the kind of IQP position Black should worry about- his light-squared bishop is
outside the pawn chain, White has no Bg5 pin or d5 push, and the d4-pawn is already under some kind of
threat.
8...Bxf3 9.gxf3
179
Position after: 9.gxf3
9...e6
This path is already quite well-trodden; the originality which is hopefully evident in the rest of the book has
taken a brief break.
Another move is 9...Nb6 but after brief investigations I decided I could not recommend that.
10.Qxb7 Nxd4 11.Bb5+
11.Nxd5 Qxd5 12.Qxd5 exd5 13.Kd1 is not a totally pointless idea from White but he can at any rate never be
better after logical development, e.g. 13...Bd6³ with ...Ne6 and perchance ...Bf4 to follow.
11...Nxb5 12.Qc6+
12.Qxb5+ Qd7 13.Qxd7+ Kxd7 14.Nxd5 exd5 is the same endgame as the text, except Black has the choice of
moving his f8-bishop next. There is no way that can constitute an improvement for White.
12...Ke7 13.Qxb5 Qd7 14.Nxd5+ Qxd5 15.Qxd5 exd5
180
Position after: 15...exd5
We have reached the famous Panov Endgame, which is also reachable from the Alapin Sicilian or potentially
the Symmetrical English. I would advise the reader to go and research the plans and motifs for both sides at
some length, because it is impossible to do justice to this position here. I shall confine myself to some brief
variations and the comment that the position is essentially equal.
16.0-0 Rd8 17.Rd1
This must be one of the most natural ways to play: rooks to c1 and d1.
17.b3 Ke6 18.Bb2 f6=
181
Position after: 18...f6=
is another main path: Black can continue with ...Bd6-e5 and ...Kf5 over the next few moves. As a rule of
thumb, it is not a big deal if White gets a rook to the seventh and begins taking pawns, as long as Black keeps
calm and plays with his d-pawn.
17...d4
17...Ke6 18.Be3 Rd7 19.Rac1 is rather problematic.
18.Bf4 Ke6 19.Rac1
Position after: 19.Rac1
19...g5!?
The direct solution, and one which I believe works.

After the preparatory 19...Rd5, 20.Rc8 g5 21.Bd2 Bg7 22.Rc6+² sees Black get tied in various knots.
20.Be3 Bg7 21.Bxg5
21.Rc6+ Kd5 22.Rc7 Be5 23.Rxa7 Ke6 24.Bd2 Ra8 Now Black was probably better and even went on to win
in: Predojevic, B – Anastasian, A Kemer 2007.
21...Rc8 22.Kf1 Kf5 23.Bd2 Be5 24.h3 d3
182
Position after: 24...d3
25.Be3
The computer move is 25.Ba5 but it may not have recognised that after 25...Rc2 26.Rxc2 dxc2 27.Rc1 Rb8
28.Rxc2 Rxb2 29.Rxb2 Bxb2= the extra pawn is entirely devoid of any significance at all.
25...Rxc1 26.Rxc1 Bxb2
There are likely a few ways play may proceed now but one illustrative one is given.
27.Rc7 Rd8 28.Ke1 Kf6!
A strong move, ensuring that Black has compensation for the pawn which is about to fall.
29.Bxa7 d2+ 30.Kd1 Bd4=
183
Position after: 30...Bd4=
White has nothing and could even need to be careful in a few moves time, whereas Black’s ideas play
themselves.
5.5) 4.c4 Nf6 5.Nc3 Be6!?: intro and 6.c5 (knight sac declined)
1.e4 c6 2.d4 d5 3.exd5 cxd5 4.c4 Nf6 5.Nc3 Be6!?
Position after: 5...Be6!?
This move is a speciality of the Swedish grandmaster Agrest, and leads to novel play.
184
6.c5
One of the critical moves, and the others are also seen.
6.Nf3?! dxc4 7.Ng5 Bd5!³ can be quickly dispensed with.
6...g6
We now have a position where the static features are probably already in Black’s favour.
6...Nc6 7.Bb5 g6 8.Nge2 Bg7 9.0-0 0-0 10.f3 gives White a slightly un pleasant bind over the centre and does
not fit in with the pioneer spirit of 5...Be6!
7.Qb3!?
White is trying to play fast; this is clearly critical.

A) A good example of what could happen if White does ‘nothing’ is 7.Nf3 Bg7 8.Bb5+ Nc6 9.Ne5 Bd7=.
B) 7.Bb5+ Bd7= immediately bothers Black even less as he will be able to recapture on d7 with the queen.
C) 7.Bd3 Bg7 8.Nge2 0-0 9.0-0 Nc6 10.f3 b6= Lopez de Turiso, J – Cruz Lledo, P San Sebastian 2015.
7...Nc6!?
Position after: 7...Nc6!?
As played by a young Bassem Amin – later the first Arab member of the 2700 club.
A) 7...Qc8 may well be fine but has a slightly passive feel to it; there are now two pieces tied down to d5 and
one to b7. 8.Bb5+ Nc6 9.Nge2 Bg7 10.0-0²
B) 7...Qc7 8.Nb5 followed by Bf4 is not a good outcome for Black
185
8.Qxb7
A) 8.Be3 Bg7! Black can merrily continue hanging the b7-pawn.

A1) 9.Qxb7?! Bd7 gives Black irrepressible activity, for instance 10.Qa6 (10.Qb3?! Rb8 11.Qc2 e5!³)
10...Rb8 11.Bb5 Nb4 12.Bxd7+ Qxd7 13.Qe2 Qf5©
A2) 9.Nf3 0-0! 10.Qxb7? After anything else, ... b6 undermines not only the White centre but also the whole
raison d’etre of Qb3. 10...Bd7 11.Bf4
Black now has a whole bunch of violent possibilities, but my favourite is 11...g5!? Trying to recover the
pawn with ...g4, which will work unless White takes on g5. 12.Nxg5 (12.Bxg5 Rb8 13.Qa6 Nb4 14.Qe2
Bf5µ) 12...e5 13.dxe5 Rb8µ Black has a huge initiative.
B) 8.Nf3 Qc7 may superficially appear no better than the queen developments in the last notes but the
difference is like night and day. The f3-knight can now be harrassed by a timely ...Bg4, whereas on e2 it could
not. It also does not support Bf4. Finally, Black’s queen helps prepare ...b6. The critical continuation is the
psuedo-active 9.Nb5 Qb8 10.g3 but after 10...Nh5!= Black’s next moves will almost definitely be ...Bg7, ...
0-0, ...b6, and White’s forces will find themselves awkwardly placed. The obvious 11.Ng5?! fails to put a dent
in this, e.g. 11...Bg7 12.Nxe6 fxe6 13.Be3 a6 14.Nc3 Bxd4ƒ.
8...Nxd4 9.Nb5
9.Bb5+ Nxb5 10.Nxb5 Rc8 gives White nothing; he can only repeat moves after 11.Bf4 Qa5+ 12.Bd2 (12.Kf1
Bg7 13.Nc7+ Rxc7 14.Bxc7 Qxc5µ) 12...Qd8=.
9...Nc2+ 10.Kd1 Rc8!!
186
Position after: 10...Rc8!!
As far as I am aware, a totally new idea, and so I shall leave the variations as a fairly sketchy outline. I have
satisfied myself – and hopefully I shall satisfy you also – that White faces a tremendously difficult task, even if
somehow he emerges with an edge.
11.Bf4!!
An even more surprising reply.

11.Kxc2 Bg7 There isn’t even any forced play. Black could conceivably castle kingside and shrug at his
opponent for a few moves before commencing offensive operations!
A) 12.Be3 0-0 13.Nd4 (13.Nf3 Ne4 14.Rd1 Nxc5³) 13...Ng4 14.b4 Bd7³ This is a good illustration of
Black’s less-is-more attitude to attacking
B) 12.b4 a5 13.a3 0-0 There is also ...Ng4. 14.Bb2 d4!?µ White’s queen is in tactical trouble, and after a
follow-up like Qf3 Nd5 it transpires that the rest of his position is too.
11...Bg7 12.Nc7+
12.Bc7 Qd7 13.c6 Bg4+ 14.Kxc2 Qf5+ gives Black full value for the piece: 15.Bd3 Qxf2+ 16.Ne2 0-0
17.Raf1 Qc5+ 18.Kb1 Ne4© and even now White has ‹castled› his pieces are still horrifically un-coordinated.
12...Kf8 13.Nxe6+
13.Kxc2 Bf5+ followed by ...Nh5 will be crushing.
13...fxe6
187
Position after: 13...fxe6
14.Rc1?!
14.Kxc2 is examined next and probably leads to a very slight plus after good play.
14...Nd4
I initially tried to make 14...Nh5 work, but there are a number of issues there.
15.Nf3
A) 15.Ba6? This completely natural move gets punished in an amazing way. 15...Qa5!!
188
Position after: 15...Qa5!!
A remarkable move, leading to some absolutely spectacular variations.

A1) 16.Qxc8+ Kf7 17.Qb7 Ne4 18.Nh3 (18.Be3 Bh6µ) 18...Rb8!!
Position after: 18...Rb8!!
A second rook sacrifice! 19.Ng5+ (19.Qxb8 Qxa6 20.Ng5+ Nxg5 21.Rc3 Ne4–+) 19...Nxg5 20.Bxb8 Ne4
21.Rc2 Qxa2‚ The two knights in the centre lend tremendous weight to the mating attack.
A2) 16.Ne2 Kf7! 17.Bd2! (17.Nxd4 Qa4+ 18.Nb3 Qxf4³) 17...Qxa2 18.Nxd4 Qa4+ 19.Rc2 Ne4!∞ Of
course, there was also ...Qxd4 with an equal game, but this is more fun and may even be better for Black. I
189
shall omit the details.
B) 15.Nh3 e5 16.Bxe5 Ne4
17.Ba6 The engine suggestion, and now Black has a choice. (17.Bxd4 Bxd4 18.c6 Kg7„ gives Black
seriously good play for a pawn) 17...Rxc5 (17...Qa5 18.Qxc8+ Kf7= is the by-now-thematic rook sacrifice,
which I have seen so often now that it has begun to look almost unimaginative. There are some cute drawing
lines here.) 18.Rxc5 Nxc5 19.Bxg7+ Kxg7 20.Qb4
190
20...Qc7! This is a nuance worth knowing. Now play enters a roughly balanced rook and knight endgame.
21.Qxd4+ e5 22.Qc3 Nxa6 23.Qxc7+ Nxc7= Black has good co-ordination and is well centralised. This
compensates for the favourable 2-v-1 thing White has going on on the queenside.
15...Ne4 16.Ba6 Rxc5
Position after: 16...Rxc5
Unfortunately, the rook sacrifices don’t really work here.
17.Rxc5 Nxc5 18.Qxa7 Nxf3 19.Qxc5
19.gxf3 Qa5 20.Bg5 Kg8!–+
19...Qa8 20.Qb6 Nd4=
Next will come ...Kf7 and Black’s co-ordination will compensate fully for the bishop pair and connected
queenside passers.
5.6) 4.c4 Nf6 5.Nc3 Be6 6.c5 (knight sac accepted)
1.e4 c6 2.d4 d5 3.exd5 cxd5 4.c4 Nf6 5.Nc3 Be6!? 6.c5 g6 7.Qb3!? Nc6!? 8.Qxb7 Nxd4 9.Nb5 Nc2+
10.Kd1 Rc8!! 11.Bf4!! Bg7 12.Nc7+ Kf8 13.Nxe6+ fxe6 14.Kxc2 Ne4
Here Black has extremely good compensation, but as we shall see it can be mitigated.
191
15.Nh3!
A) 15.Re1 is most natural, but commits to a sacrifice of the queen for a bag full of other stuff: 15...Nxc5
A1) 16.Qxa7 Ne4+ 17.Kb1 Ra8 18.Qe3 Qa5 19.a3 g5!
White is unable to keep his extra piece, since 20.Bg3? walks into 20...Nd2+! 21.Kc2 (21.Ka2 Qb4–+)
21...Rc8+ 22.Kd1 Qa4+ 23.Ke2 Nb3!–+
192
This incredible quiet move derives its strength from the fact that White’s kingside is still asleep and his king
walk is at an end.
A2) 16.Qb5 e5 17.Bc1 Rb8 18.Qxc5 (18.Qe2 e4³ is great for Black) 18...Rc8 19.Qxc8 Qxc8+ 20.Kb1 Qf5+
21.Ka1 Qxf2 22.Re3 Bf6³
B) 15.c6 stays equal, but Black’s play is still quite easy. 15...e5 16.Be3 (16.Bd2 Nxd2 17.Kxd2 Rc7 18.Qa6
Qd6 19.Bb5 Qb4+ 20.Kc2 e4 21.Rb1 d4‚; 16.Bg3 Rc7 17.Qb3 Rxc6+ 18.Kd1 Qa5©) 16...Rc7 17.Qa6
(17.Qb3 d4 18.Rd1 Rxc6+ 19.Kb1 Qc8‚ looks pointless to even try for White) 17...Qd6 18.Bb5 d4 19.Rd1!
Nc5!
20.Qa3 a6 21.Bc4 (21.b4 Ne4 22.Ba4 Rxc6+ 23.Bxc6 Qxc6+ 24.Kb2 Nc3©) 21...Qxc6 22.Kb1 Ne6!
Regaining the piece; now hostilities suddenly cease. 23.Rc1 Qe4+ 24.Qd3 Qxd3+ 25.Bxd3 dxe3 26.fxe3
Rxc1+ 27.Kxc1 Nc5 28.Kc2 a5=
C) 15.Kb1 looks natural too, but there follows 15...g5 16.Ba6
193
Position after: 16.Ba6
(16.Be3 Rb8; 16.Bc1 Nxc5 17.Qb5 Qc7µ) 16...Qa5!! Anyone surprised? 17.Qxc8+ Kf7 18.Qb7 gxf4 19.Nf3
(19.Nh3 Be5!? 20.Qb5 Nc3+ 21.bxc3 Rb8³; 19.a3 Bxb2=) 19...Nxc5 20.Qb5 Qxa6 21.Qxa6 (21.Qxc5?!
Rc8; 21.Ng5+?! Kf6 22.Qxc5 Rc8 23.Qd4+ Kg6 24.Qd1 Rb8 25.b3 Kxg5!³) 21...Nxa6=
D) 15.Nf3 e5 16.Be3 Nxc5 17.Bxc5 Rxc5+ 18.Kb1 Bf6ƒ is likely to turn into a model exposition of Black’s
long-term counterplay.
15...Nxc5

194
16.Qxa7
16.Qb5 e5 17.Bc1 Rb8 is ‘perpetual queen’.
16...e5 17.Bd2!
A) 17.Bc1 Ra8 18.Qxc5 Rc8 19.Qxc8 (19.Qc3 e4 20.Qxc8 Qxc8+ 21.Kb1 d4‚) 19...Qxc8+ 20.Kd1 Qg4+
21.f3 Qa4+ 22.Ke1 e4©
B) 17.Be3 d4 is never going to be a helpful move-pair for White to insert.
17...Ne4+ 18.Kd1 Nxd2 19.Kxd2 e4
20.Rb1
White wants to avoid losing the b2-pawn and run away with Be2, Ke1-f1.
A) 20.Ng5 Qd6 21.Rc1 Rb8‚ The attack rages on.
B) 20.Be2 Bxb2 21.Rab1 Ra8 22.Qc5 (22.Ng5 Bc3+!) 22...Rc8 23.Qb5 Bc3+ 24.Kd1 Qd6©
20...Qd6 21.Qa3
21.Qa4 Bh6+ 22.Ke1 Kg7 gives Black decent compensation too.
21...Bc3+ 22.Kd1 Bb4 23.Qb3 Kg7²
195
Position after: 23...Kg7²
With the aid of some machinery Black has detained the enemy monarch in the centre. Black has almost
sufficient compensation.
A machine line goes: 23...Kg7 24.Rc1 Qb6 25.Qe3 Qa5 26.Be2 e5 27.Qb3 Be1! 28.Rc2 Rxc2 29.Kxc2 Qd2+
30.Kb1 Qxe2 31.Qxd5 Kh6∞ and then while White might be better with good play, it has not been possible to
find out exactly how in 15 minutes of cloud engine usage...
5.7) 4.c4 Nf6 5.Nc3 Be6 6.Qb3
1.e4 c6 2.d4 d5 3.exd5 cxd5 4.c4 Nf6 5.Nc3 Be6!? 6.Qb3!?
196
Position after: 6.Qb3!?
Just as with 6.c5 g6 7.Qb3, Black can continue in enterprising vein.
6...Nc6 7.Nf3!
A) 7.cxd5? Nxd5 8.Qxb7 Ndb4!µ

B) 7.Qxb7? Nxd4 8.cxd5 Nc2+–+
7...Na5™
A) 7...Qd7 8.c5 or;

B) 7...a6 8.c5 risk giving White exactly what he wants from 6.c5, minus the development issues.
C) 7...Qb6 8.c5 Qxb3 9.axb3²
D) 7...dxc4 8.Qxb7 Qc8 9.Qxc8+ Rxc8 10.Be3² could well be playable but from my investigations I am not
convinced.
8.Qb5+ Bd7
9.cxd5!!
If White does not find this idea he is of course worse for quite straightforward reasons.
9...e6
Necessary.
197
A) 9...a6 10.Qd3 b5² was the immediate continuation of the extremely messy encounter Nadanian, A –
Riazantsev, A Khanty-Mansiysk 2013, but although Black ultimately outplayed his opponent, White’s play is
easy to improve.
B) 9...Bxb5 10.Bxb5+ Nd7 11.Ne5² leaves Black with no real way to regain the d5-pawn.
10.Qe2! Bb4!
10...Be7 11.dxe6 Bxe6 12.Qd1 0-0 13.Bd3 will probably lead to White unravelling, though there is still a
measure of compensation after say 13...Nd5².
11.dxe6
11.Bg5 0-0 12.dxe6 Bxe6 is totally unconvincing; White needs to make haste developing his kingside, not mess
around with pins.
11...Bxe6
Position after: 11...Bxe6
12.Qb5+
A) 12.Qc2 0-0 13.Be2 Rc8 14.0-0 h6© gives Black fantastic positional compensation.
B) 12.Qd1 0-0 13.Bd3 Rc8 will compel White to give back the pawn since 14.Bd2? Nc4 spells nothing but
trouble.
12...Nc6 13.Be2!
A) 13.a3 leads to positions where one could be excused for thinking Black had played some gambit line of the
Nimzo. 13...Bxc3+ (13...Ba5!? may also be worth investigation) 14.bxc3 0-0
198
15.Be2 (15.Bd3 Qa5 16.Qxa5 Nxa5=; 15.Bg5 Re8 16.Be2 Qa5 17.Qxa5 Nxa5 18.Bxf6 gxf6= Black will
regain the pawn and continue playing actively.) 15...Ne4 16.Qb2 Na5! Black hits c3 much too fast, e.g. 17.0-
0 (17.Nd2 Nxd2 18.Bxd2 Bc4©) 17...Rc8 18.Bd3 (18.Bd2 Nb3³) 18...Bd5!=
B) 13.Qxb7 Bd5 leaves White scrambling somewhat; he now has to find a second queen sacrifice.
14.Bc4! (14.Bb5 0-0 15.Bxc6 Rb8‚) 14...Rb8! (14...0-0 15.Bxd5 will let White consolidate his pawn)
15.Bxd5!! Rxb7 16.Bxc6+ Rd7 17.Ne5 0-0 18.Nxd7 Nxd7 19.0-0 Nf6© The chances are roughly balanced in
199
spite of the odd material distribution.
C) 13.Bd3 is a worse deployment of the bishop: 13...a6 14.Qa4 0-0
15.Be3 (15.0-0 Nxd4!= is a great tactical argument in favour of Be2 rather than Bd3) 15...Bd5! 16.0-0 Bxf3
17.gxf3 Qd7 18.d5!? (18.Ne4 Nd5 with ...f5 and ...Rad8 to come also leaves the compensation in no doubt)
18...Bxc3 19.dxc6 Qxd3 20.cxb7 Rab8 21.bxc3 Rxb7©
13...a6 14.Qd3!
14.Qa4 is not the best retreat: 14...0-0 15.0-0 Re8 16.a3 (16.Bg5 b5 17.Qd1 h6 18.Bxf6 Qxf6©) 16...Bxc3
17.bxc3 Bd5 18.Qd1 Rc8 also gives great positional compensation, though there is one critical line going
forward, which is 19.c4 Bxf3 20.Bxf3 Nxd4 21.Bxb7 Rxc4 22.Bxa6 Ne2+ 23.Kh1 Rd4 24.Qc2 (24.Qb3
Ng4‚) 24...Nxc1 25.Raxc1 (25.Rfxc1 Ng4‚) 25...Qa5 winning back the pawn.
14...0-0 15.0-0 h6 16.Be3 Re8
200
White now has to decide how to develop his rooks.
17.Rac1!
A) 17.Rfd1 Bxc3! seems to give Black enough. 18.bxc3 Na5 Now there are only a few things White can do
about Bc4. 19.Ne5 (19.Nd2 Rc8=) 19...Rc8 20.Bf3 (20.Bf4 Nc4 21.Qc2 b5=) 20...Bd5 21.Bf4 Bxf3 22.Qxf3
Qd5 23.Ng4™ 23...Ne4 24.Rac1 Nc4©
B) 17.Rfe1 Na5 18.Ne5 Rc8 19.Rac1 Nc4 is also enough.
C) 17.Rad1 Bxc3 18.bxc3 b5= leaves not only ‘c3’ but also ‘a2’ vulnerable; there might follow 19.Qc2 Rc8
20.Qb2 Na5 when White is totally passive and ...Bc4 or ...Nc4 are unstoppable. Black also has ...Nd5 and
...Qc7 ideas in the bank.
17...Bxc3 18.bxc3
18.Qxc3 Nd5 19.Qa3 Qf6©
18...b5 19.Qb1
Other moves seem to give Black a purchase on the c4-square.

A) 19.Qd2 Rc8 20.Rfe1 Na5 21.Ne5 Nc4 22.Bxc4 Bxc4 Now, in order to prove something, White has to
sacrifice: 23.Bxh6 gxh6 24.Qxh6 Re6 25.Qg5+ Kf8∞ but there is nothing particularly obvious for the piece,
and Black is threatening to unravel with ...Ke7 and ...Qg8.
B) 19.a4 Bc4 20.Qd1 Nd5=
C) 19.Qc2 Rc8 20.a4 (20.Qb2 Na5©) 20...Nb4 21.Qb2 Na2 22.Rc2
201
Position after: 22.Rc2
22...Nd5! 23.Bd2 Bf5 Black’s position looks decidedly tenuous if not for this move. 23.Bd2 Bf5! White must
now think about how to keep equality; one possibility is 24.Ne5 Bxc2 25.Qxa2 Bxa4 26.Bf3! Rxe5 27.dxe5
Nxc3 28.Bxc3 Rxc3 29.e6 Bb3 30.Qxa6 Bxe6 31.Qxb5=.
19...Bd5 20.Rfe1
20.Nd2 Na5 may ironically end up giving Black his desired clamp on c4: 21.Rfe1 Qc7 22.a4 Nc4©.
20...Qa5!
20...Na5 21.Ne5±
21.c4
Risky, but the only way to stop White’s extra pawn from winging its merry way home. 21.Nd2 Qxa2=
21...bxc4 22.Bxc4
202
22...Rab8
22...Bxf3?! 23.gxf3 Qh5 is appealing, but White has just enough accurate moves to come out the other side.
24.Bxf7+! Qxf7 25.Rxc6 Nd5.Now 26.Qg6! Qxf3 27.Qg3! Qxg3+ 28.hxg3 Rac8 29.Rd6!± is a very annoying
string of accurate moves for White.
23.Bb3
23.Qd3 Nb4 24.Qf5 Qa3= regains the pawn and Black seems fine.
23...Qa3 24.Nd2
After 24.Rc3 Ne4 25.Qc1 (25.Rd3 Qd6) 25...Qxc1 26.Rcxc1 Nb4 Black keeps a bind.
24...Na5 25.Qd3
25.Rc3 Bxb3 26.axb3 Nd5=
25...Nxb3 26.axb3 Qb4 27.h3²
203
Position after: 27.h3²
White still has a pawn, albeit a well-blockaded one, so Black’s compensation is always in danger of slipping
away, even if personally I wouldn’t mind taking his position after 27...Qb7.
5.8) 4.c4 Nf6 5.Nc3 Be6 6.Nge2!?
1.e4 c6 2.d4 d5 3.exd5 cxd5 4.c4 Nf6 5.Nc3 Be6!? 6.Nge2!?
Position after: 6.Nge2!?
This equally creative response has become reasonably well known. It tends to lead to open positions where
204
Black has the bishop pair but it is not particularly well-placed; typically the dark-squared bishop is vulnerable
to being exchanged. Nevertheless, by contrast to the other two lines I do not believe White gets anything of note
here.
6...dxc4 7.Nf4 Bg4 8.f3 Bd7 9.Bxc4 e6
9...Nc6 10.d5 Na5 11.Bd3² is nice and easy to play for White.
10.d5!
10.0-0 Bd6
11.d5 (11.Ne4 Qc7 12.Nxd6+ Qxd6 13.Be3 Bc6 14.Rc1 0-0 15.Qd2 Nbd7= Svensson, B – Agrest, E Sweden
2010) 11...Qc7 12.dxe6 fxe6 13.Qd2!? (13.Nxe6 Bxh2+ 14.Kh1 Qxc4 15.Nxg7+ Kf7 16.Kxh2 Kxg7³ doesn’t
give White quite enough) 13...g5 14.Nfe2 Bxh2+ 15.Kh1 Rg8= Black will manage to either complete
development or get a strong attack with ...Rg6-h6.
10...Nxd5!?
10...e5 is the tried and tested alternative, but I’m not so sure about it. 11.Nd3 Bd6
205
12.Qb3! As played by a young Katerina Nemcova- now she is a WGM. (12.Bg5 Bf5„; 12.Be3 Na6! Arranging
counterplay before White can castle. 13.0-0 Rc8 14.Bxa6 bxa6 15.Nf2 0-0„) 12...Bf5!? 13.g4!? The engine
move, but taking on b7 immediately was similar. 13...Bg6! 14.Qxb7 Nbd7 15.Qc6² The jury is still out to
some extent, but Black has no obviously inspiring continuation.
Position after: 10...Nxd5!?
11.Bxd5
A) 11.Nfxd5? Qh4+ is of course the point.

206
B) After 11.Ncxd5 exd5 12.Bxd5 Bb4+ 13.Bd2 Qe7+ 14.Qe2 Qxe2+ 15.Kxe2 Bxd2 16.Kxd2 Bc6 17.Rhe1+
Kf8= it isn’t clear to me that White has anything.
11...exd5 12.Qe2+
12.Nfxd5 Bb4!?
Position after: 12...Bb4!?
There is also 12...Bc5 13.Ne4 Bc6, but this seems to equalise cleanly, even if the ‘proof’ takes a while:
13.Qe2+ (13.Nxb4 Qh4+ 14.Kf1 Qxb4 15.Nd5 Qc4+ 16.Qe2+ Qxe2+ 17.Kxe2 Na6 18.Kf2 Be6 19.Nf4 0-
0=; 13.Qd4 Bxc3+ 14.Qxc3 0-0 15.0-0 Bc6 16.Ne3 Bd7 17.b3 Be6 18.Bb2 f6 19.Rfd1 Qb6 20.Kh1 Nc6=;
13.0-0 Bxc3 14.bxc3 0-0 15.c4 Nc6=) 13...Be6 14.Nxb4 (14.Nf4 Qb6 15.Nxe6 Qxe6 16.Qxe6+ fxe6=)
14...Qh4+ 15.Qf2 Qxb4 16.Be3 b6 17.0-0 0-0=.
12...Be7 13.Nfxd5 Nc6
207
14.Bg5
Forcing ...f6 for an obscure reason.

14.Be3 0-0 15.Rd1 Nb4!= Neither this nor the equally strong ...Bh4+ is possible with the f-pawn at f6.
14...f6 15.Be3
15.Bf4 0-0 16.Rd1 Rf7 17.0-0 Qf8 is similar.
15...0-0 16.Rd1 Rf7 17.0-0 Bf8„
208
Position after: 17...Bf8„
Black is fine, though he may need some care to unravel completely.
It is time to congratulate anyone who has made it through this whole chapter (the last of the ‘less critical
lines’ part of the book, though as you will appreciate before long, that is very much relative and there have
been some challenging lines here too.)
That means not only the standard segments (though hopefully you have made your choice between 10...e6
and 10...e5 in 5.3) by now) but also the 5...Be6 digression. My sense is that ultimately, White should allow
and then accept the knight sacrifice, despite the position not being easy, and that White preserves a little
something at the end. It is important to be honest. However, the amount of memory work needed on both
sides (especially White’s) means that an imbalance on that front is sufficient to outweigh the slightly
undesirable outcome of absolutely perfect play – and the final position is not exactly clear either. If I were
to learn the variation from scratch, I might memorise the critical line up to 19...e4. Happy hunting. Next up
– the Advance Variation.
209
Part II:
Advance Variation
Chapter 1
Various White systems
1.e4 c6 2.d4 d5 3.e5 Bf5 4.--
Chapter guide
Chapter 1 – Various White systems
1.e4 c6 2.d4 d5 3.e5 Bf5
1.1) 4.Nc3 e6 5.g4 Bg6 6.Nge2 h5!?
1.2) 4.Nc3 e6 5.g4 Bg6 6.Nge2 c5
1.3) 4.h4 h5 5.Bg5
1.4) Main line: 4.h4 h5 5.Bd3 Bxd3 6.Qxd3 Qa5+
1.5) Main line with 7.Bd2 Qa6 8.e6 Qxd3 9.cxd3 f6 10.Nc3
1.6) 4.h4 h5 5.c4
1.7) 4.Nd2 and 4.Be3
Let’s start off gently. There are a number of new minor lines to think about after 3.e5 Bf5 – as well as old ones
that have faded from the public consciousness. A good example of the latter is 4.Nc3 e6 5.g4 Bg6 6.Nge2,
against which I give two recommendations. The imbalance is White’s dynamic piece play against his crumbling
centre and weak light squares; overall, modern engine analysis has tended to prove that Black has good play
here and my findings are not substantially different. Meanwhile, the whole 4.h4 complex is a relatively modern
one, and it would be wise to pay attention to lines 1.4) and 1.5) as my feeling is White probably has not yet
210
found the right way to execute the pawn sacrifice. Related to these lines is the annotated game 5.3).
Finally, we wrap up the chapter with a look at White’s ambitious, yet concretely not-quite-there attempts to
keep the ...c5 break under control by placing a knight on b3.
1.1) 4.Nc3 e6 5.g4 Bg6 6.Nge2 h5!?
1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nc3 e6 5.g4 Bg6
6.Nge2
White needs to hurry with this, otherwise ...h5 will erode his kingside light squares and leave him worse. The
text gives him the idea of Nf4, which discourages the aforementioned push from Black.
A) 6.h4 h5 7.g5 c5³
B) 6.Be3?! h5 7.g5 Ne7ƒ
6...h5!?
This can possibly be played anyway, though it is rather esoteric. I shall give White’s path to an advantage, but it
is nevertheless clear to me that some Black players could wish to enter this line. The next line deals with a more
mainstream option for Black.
7.Nf4 Ne7 8.Nxg6!
A) 8.gxh5 Bf5 9.Bd3 Bxd3 10.Qxd3 c5 11.dxc5 Nd7 and Black regains one of the pawns and gets full
compensation for the other, e.g. 12.Be3 Nxe5 13.Qb5+ Qd7 14.Qxd7+ Kxd7 and in fact only he can stand
better.
211
B) 8.Nxh5 c5³
8...Nxg6 9.gxh5 Nh4 10.Qg4!
Position after: 10.Qg4!
This is the only critical move. Against anything else, ...c5 at least equalises.
10...c5 11.dxc5 Nc6 12.Bd2!
A) 12.Bb5 Bxc5 13.Qxg7 Rxh5³

B) 12.Bf4 is most natural, and leads to a very rich position, where Black is nevertheless not worse: 12...g6!
13.hxg6 (13.0-0-0 Rxh5 is fine for Black, e.g. 14.Bc4 Qc8!=) 13...Nxg6 14.Bg3
212
Position after: 14.Bg3
B1) 14...d4?!N is tempting but Black has broken his structural integrity and now White has a promising
piece sacrifice: 15.0-0-0! (15.Ne4 Qa5+ 16.Kd1 Ngxe5„) 15...Qa5 16.Bc4!? (16.Nb5 Qxa2 17.Nd6+ Bxd6
18.exd6 Qa1+ 19.Kd2 Qa5+ 20.c3 0-0-0„) 16...dxc3 17.Bxe6‚.
B2) 14...Ngxe5 15.Bxe5 Nxe5 16.Bb5+ Nc6 17.0-0-0 Qh4! (17...a6 worked out for Black in Nolte, R –
Pantsulaia, L Hyderabad 2013, however there is an unpleasant idea for White here: 18.Nxd5N 18...axb5
19.Rhe1±) 18.Qe2 Be7
213
We have reached a position that is hard to play for both sides. It is approximately dynamically equal. The
obvious try is 19.Bxc6+ bxc6 20.Qa6 but Black is on time with 20...Qf4+! 21.Kb1 Qc7„.
12...g6 13.hxg6 Nxg6 14.0-0-0 Bxc5 15.f4 Qh4
Position after: 15...Qh4
16.Qf3²
Objectively, Black does not have enough compensation, but his plans are easy enough (...0-0-0, ...Nd4, maybe
...Qh3 if the f1-bishop moves away, pressure the f4-pawn) and White needs to be quite precise.
1.2) 4.Nc3 e6 5.g4 Bg6 6.Nge2 c5
1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nc3 e6 5.g4 Bg6 6.Nge2 c5
214
7.h4
7.Be3
A) 7...h5!? 8.Nf4
A1) After 8...Bh7 Black obtains an undoubtedly inferior version of the main line (7.h4) assuming White
responds with the correct Nxh5. Meanwhile, 9.dxc5?! Nc6 was already good for Black in Murray, D –
O’Donnell, C Bunratty 2015.
A2) 8...Nc6!?N To my mind a really remarkable idea. Black recognises that doubled g-pawns are not a
massive weakness, and that in fact that structure can be useful to control the f5-square a bit better. 9.Nxg6
fxg6
215
Position after: 9...fxg6
A2.1) 10.gxh5?! cxd4 11.Bxd4 Nge7! Black could already be better; the imminent ...Nf5 will hunt the d4-
bishop down in mid-board. White needs to take action immediately, but even after the best continuation,
12.Bb5 Nf5 13.Qg4 there is a phenomenally strong move for Black: 13...Qh4! 14.Qxh4 Nxh4 15.0-0-0
Rxh5³.
A2.2) 10.Qd3 This seems critical, but as long as Black is prepared to move his king, as in the lines
10...cxd4 11.Bxd4 Nge7 12.Nb5 Kf7! 13.0-0-0 (Or 13.c4 Nxd4 14.Qxd4 Qa5+³) 13...hxg4 14.c4 dxc4
15.Qe4 Nd5³ White is really struggling to prove compensation.
A2.3) 10.Ne2 is most principled but Black is straightforwardly fine after 10...hxg4 11.c3 Qd7= and his
next plans like ...cxd4 and ...Nge7-f5 are not exactly rocket science either.
A2.4) 10.Bb5 Nge7 11.dxc5 a6 12.Ba4 (12.Be2 hxg4 13.Bxg4 Nf5³) 12...Qa5 13.a3™ 13...0-0-0 14.b4
Qc7 15.Ne2!
216
Position after: 15.Ne2!
Now accurate play is required from both players to reach an equal ending: 15...hxg4 16.Bxc6 Nxc6 17.Nd4
g3 18.fxg3 Qxe5 19.Qd3 Nxd4 20.Qxd4 Qxd4 21.Bxd4 e5= Black’s central play and ideas down the h-file
compensate for his undoubted structural problems.
B) 7...Nc6 as recommended by Vidit is also fine, but I think he is a bit overenthusiastic in his espousal of it,
and also I feel a moral duty to expand the range of options available to my chosen side. I shall confine myself
to demonstrating a minor problem I have with his main line, for those who own that work: 8.dxc5 h5 9.Nd4
hxg4 (9...a6= could be preferable) 10.Bb5 Nge7 11.Qxg4 Qc7 12.0-0-0 Bh5 13.Qh3 0-0-0 14.Bxc6!?N
14...Nxc6 15.Ncb5‚ White whips up a nice attack against the castled Black king.
7...h5 8.Nf4
217
8...Bh7
8...cxd4 9.Nxg6 fxg6 10.Qxd4 Nc6 is concretely not as convincing as the 7.Be3 h5 8.Nf4 Nc6 line, because of
the interesting gambit line 11.Qd3! Nxe5 12.Qg3‚.
9.Nxh5
9.g5 is White’s second option, and has been chosen by people like Shirov. 9...cxd4 10.Nb5 (10.g6 is too
ambitious: 10...Bxg6 11.Nxg6 fxg6 12.Qxd4 Ne7³) 10...Be4 11.f3 Bf5 12.Nxd4 Ne7 13.Nxf5 Nxf5 14.Bd3
Qc7= Black is fine here as long as he follows up with standard light-square themed moves like ...g6 and ...Nc6.
9...Nc6 10.dxc5
10.Bb5?! is not a strong move but has a trick or two: 10...cxd4 11.Qxd4 (11.Bxc6+ bxc6 12.Qxd4 Bxc2³ and
White does not have enough play or development to offset the bishop-pair.) 11...Nge7
218
Position after: 11...Nge7
12.Bh6!? Trying to win a brilliancy prize. (12.Bg5 a6! 13.Bxc6+ Nxc6µ Black was already significantly better
in Ehlvest, J – Bareev, E Moscow 2001 and others.) 12...a6! 13.Bxg7 Bxg7 14.Nxg7+ Kf8µ White lacks full
compensation for the piece, for instance see Simacek, P – Deviatkin, A Pardubice 2003.
10...Bxc5
11.Bb5
219
A) 11.Bg5 Qb6 12.Qd2 is somewhat critical in that it obliges Black to begin taking pawns. 12...Qxb2 13.Rb1
Qxc2 14.Qxc2 Bxc2 15.Rc1 d4! After this last move, which is a good one to know, there is no way Black can
be worse.
B) 11.Bg2 Bg6!? The most forcing continuation. 12.Bg5 (12.Nf4 Rxh4µ; 12.Nxg7+ Kf8 13.Nxe6+ fxe6
14.h5 is initially thrown out by the engine but White’s compensation is decidedly dubious after 14...Bh7µ)
12...Qb6 13.0-0 Bxh5 14.gxh5 Be7!
This position is an unmitigated mess and both kings are weak. However, there are relatively few lines in
which White is able to develop an attack against the king, and so the human evaluation should, like the
computer one, favour Black slightly.
C) 11.Nxg7+ Kf8 12.Nh5 Once more, the various ideas connected with Nxe6+ do not convince, and
encouraging Black’s queen out with gain of tempo is almost certainly counterproductive.
220
12...d4 13.Nb5 (13.Nb1 Qd5 14.Rh3 Nxe5³) 13...a6 14.Na3 Nxe5³ Not withstanding the popularity of this
line for White in the past, I think it is safe to end here and state that Black is certainly better, with his superior
king safety, central control and overall co-ordination.
11...Qc7 12.Bxc6+ Qxc6 13.Qf3
13.0-0 d4 14.Ne2 0-0-0µ and with ...d3 imminent (amongst other things) White is in huge trouble.
13...0-0-0 14.Nxg7
221
Position after: 14.Nxg7
14...Ne7!
A good find from Vidit, hugely strengthening the threat of ...d4. While I am still not sure it promises a 100%
foolproof advantage, his analysis of the then- novelty and subsequent game with it are fairly convincing.
14...d4 15.Qxc6+ bxc6 16.Na4© eventually led to a White win in Topalov, V – Gelfand, B Dortmund 2002
although right here there seems nothing wrong with Black›s position.
15.Bg5 d4 16.Qxc6+ Nxc6 17.Na4!
17.Bxd8 Rxd8 18.Nd1 Nxe5³ and Black’s bishop pair completely dwarfs White’s material advantage in
importance.
17...Bf8 18.Bxd8 Kxd8 19.0-0-0 Bxg7 20.f4 Be4„
Position after: 20...Be4„
White is not without certain practical chances connected with the kingside pawn massif and the d6-square, but
objectively Black should be slightly better here. This last move I consider to be a slight improvement over a
recent game Liu Guanchu – Vidit, S Chengdu 2017, which continued 20...Bh6.
1.3) 4.h4 h5 5.Bg5
1.e4 c6 2.d4 d5 3.e5 Bf5 4.h4
222
Position after: 4.h4
4...h5
4...c5!? is worthy of some attention, but a detailed look will reveal too many problems for it to be analysed
here.
5.Bg5
This looks logical, using the slightly weakened g5-square and making ...e6 difficult for a move or two, but
Black has at least two acceptable answers.
5...Qb6
This makes White play an unsound gambit or else give concessions. I am not recommending 5...f6, because
after 6.Bf4! Black should probably play ...Qb6 anyway, and that version is worse than this one, even though
Black might still be equal there.
6.Bd3
Not the best move, but the only critical one.

6.b3 e6 7.c3 c5= and only Black can think about being better.
6...Qxd4!
6...Bxd3 cannot be recommended, since after 7.Qxd3 Black cannot take the material. 7...Qxb2? 8.e6! Qxa1
9.Qb3!± See, inter alia, Perunovic, M – Lalith, B Dubai 2009. Therefore, Black would have to try and claim
equality with 7...Qa6 or 7...e6, which is a worse approach than taking the material in the correct way on the
previous move.
223
7.Nf3 Qg4
8.Bxf5
A) 8.Be2 is less critical. There follows 8...Qb4+ 9.Nbd2 e6 10.0-0 Nd7 and Black seems to have consolidated
his pawn pretty well. There is no need to take another on b2- this just makes life more complicated. White’s
compensation is a bit lacking, e.g. 11.c4 Be7!³.
B) 8.0-0 Bxd3³ deprives White of most of his dynamism and all of his queen-trapping tricks, so Black can get
on with ...e6 and ...Nd7 and be better.
8...Qxf5 9.0-0!?
Cuing up an interesting piece sacrifice.

9.c4 leaves the king in the centre too long: after 9...Qe4+! 10.Be3 e6
224
Black has gained two tempi, one from the king and one from the c4-pawn. 11.cxd5 (11.Qb3? Qxc4 12.Qxb7
Qb4+µ is worth pointing out)
A) Vidit also gives 11...Bb4+ 12.Nbd2 Qxd5 but I find this less solid after the gambit continuation 13.0-0
Nd7 14.Qc2!? when Black might be objectively for choice but he gets prevented from castling and White has
two more developed pieces with which he is constantly able to gain time.
B) 11...Qxd5 12.Qc2 (12.Qe2 Qb5!µ; 12.Nbd2 Nd7µ) 12...Nd7 13.Nc3 Qa5 14.0-0-0 Ne7³
9...f6
225
10.Nd4
10.exf6 gxf6 11.Be3 Qg4–+ leads to a position where Black has the pawn and the compensation.
10...Qxe5 11.Re1 Qd6
Now Black wants to play ...e5.
12.Ne6 Nd7!
12...fxg5 13.hxg5 Nd7 14.Qf3³ seems less clear to me.
13.Bf4™ 13...Ne5 14.Nxf8
14.Ng5
Position after: 14.Ng5
This is trying a bit too hard. Black has an accurate move now which will leave White feeling sorry for himself.
14...Nh6! 15.Bxe5 (15.Qxh5+ g6 16.Qe2 Nf5–+; 15.Nd2 Nhg4–+) 15...fxe5 16.c4 0-0-0 17.cxd5 cxd5µ White
has some play, but it would be misleading to describe it as compensation.
14...Kxf8
226
Position after: 14...Kxf8
15.c4!?
There is only one way White can make play here and that is by blasting the centre open.
15.Qe2 Nh6 16.Bxe5 fxe5 17.Qxe5 Ng4 18.Qxd6 exd6µ
15...dxc4 16.Qe2
16.Nd2 Rd8µ
16...Qd3!
Not allowing the counterplay to germinate.
17.Bxe5 Qxe2 18.Rxe2 fxe5 19.Nd2!
19.Rxe5 Rd8µ
19...Rh6 20.Nxc4
20.Rxe5 Rd6 21.Nxc4 Rd5³ is also a solid extra pawn.
20...Re6 21.Nxe5 Ke8³
Black has a solid extra pawn and can follow up by developing his rook to d8 and his knight to either f6 or f5.
1.4) Main line: 4.h4 h5 5.Bd3 Bxd3 6.Qxd3 Qa5+
1.e4 c6 2.d4 d5 3.e5 Bf5 4.h4 h5 5.Bd3 Bxd3 6.Qxd3 Qa5+

227
Position after: 6...Qa5+
On principle, this is the right way to go, placing the queen at a6 to interfere with White’s development. Against
all one try, this idea equalises virtually instantly, and so there are lots of stub lines.
7.Bd2!
White forces e6 through.

A) 7.Nd2 e6 8.Ngf3 (8.Ne2 Qa6=) 8...Qa6 9.c4 Now Black can choose between a few ways of playing, like
the creative ...Bb4!? but there is also just the simple 9...Ne7= going for ...Nf5 and ...Be7.
B) 7.Nc3 e6 It is desirable to prevent White’s e6-based ideas before playing ...Qa6, as otherwise White would
obtain a better version of 7.Bd2. 8.Nge2 (8.Nf3 Qa6 9.Qxa6 Nxa6=) 8...Qa6 9.Qh3!? A slightly different
place for the queen, but Black just follows the formula and is equal, e.g. 9...Ne7 10.0-0 Nf5 11.Bg5 Nd7=
228
Position after: 11...Nd7=
with two possible plans: either ...Be7/...0-0 or ...g6/...c5 can follow next but in either case Black is okay.
7...Qa6 8.e6!
Position after: 8.e6!
If not for this move, it would not be clear why White has blocked the check in a worse way with his previous
move.
8...Qxd3 9.cxd3
229
9.exf7+!? This is a significant option, but is decidedly non-forcing after 9...Kxf7 10.cxd3 so I shall break with
the style of the book and deal with it through the annotated game in chapter 5 (line 5.3, Kollars – Can).
9...f6!
The maximalist continuation.

Vidit gives 9...fxe6 as main, which could also be playable, but I think White’s play is a bit easy. 10.Nf3 Nd7
11.0-0 Nh6 12.Re1
Position after: 12.Re1
12...e5! This is needed to equalise. (12...Nf5 13.Rxe6 g6 14.Na3 Bg7 15.Rae1² and I think White is already
better because his moves are obvious (Nc2, Bg5, R6e2 and then prepare either Ne3 or Ne5) while a Black
break is less easy to see.) 13.dxe5 Nf5 (13...Nc5 14.Bxh6 Rxh6 15.d4²) 14.Nc3 Nc5 15.d4 Ne6 16.Ne2 g6=
Black is passive, but does have a good structure in compensation. I have given this option due to my philosophy
of always offering an alternative to the acceptance of a gambit, unless the gambit is truly terrible; however, I
prefer 9...f6 as there is always a danger of constriction in positions like this one, where you can never be sure
whether ...c5 is really going to be a good idea.
10.Ne2
A) 10.Nh3 is more or less equivalent, and it stands to reason that the main moves worth considering are
knight developments.
B) 10.Nf3 is not the right trajectory for the knight: 10...Na6 11.Bf4 Nh6 12.Nc3 Nf5³.
230
Position after: 12...Nf5³
with ...g6 and ... 0-0-0 to come, and then several mixtures of the moves ...Bh6, ...Ng7, ...Rd6 and ...Nc7 are
possible, with the net result being that the e6-pawn is gone.
C) 10.Nc3 – see the next subchapter.
10...Nh6
The position is very schematic by nature and Black seems to do very well if he just plays, almost on autopilot,
231
...Nh6-f5 and ...Na6-c7, without thinking too much about it. That is the essence of the following lines, which
are extremely good for Black, even if he might have some clever slight improvements in different places.
11.Nf4
A) 11.Bxh6? is rarely going to be a good idea: now the g6-square is not weak and Black can target the e6-
pawn in various ways, supported by a timely ...g5. 11...Rxh6 12.Nbc3 Na6 13.Nf4 Nc7 14.Kd2 0-0-0!³ (The
pawn trade 14...g5 could also leave Black slightly for choice.)
B) 11.Ng3 lacks a threat and Black can just play ...Na6-c7. 11...Na6 12.Bf4 (12.0-0 g5„) 12...0-0-0 13.Nc3
Nc7³
C) 11.Nbc3 Nf5 12.Nd1
...is a coherent idea, but a bit slow: 12...Na6 13.Ne3 Nxe3 14.Bxe3 g6 15.Nf4 Rg8³ In the long run, the e6-
pawn will not survive against the idea of ...Nc7, ...0-0-0, ...Rd6, and then ...Bh6 and/or ...g5 as necessary.
11...Nf5 12.Nc3
White probably has to play some such gambit, and another better version will be seen next. It would be
reasonable to ask why the present line is being seen in so much detail, since at the end it is White walking a
narrow tightrope to equality; however, from another perspective, all gambits are somewhat dangerous and in
this particular one Black has very obvious immediate problems with development- if they are not solved in a
precise fashion Black will be dead within 10 moves and his remains exhibited in the Museum of Miniatures
with the caption ‘Light Square Catastrophe’.
A) 12.Be3 Na6! (Vidit gives 12...g6!³ followed by some very elegant variations, but there is more than one
way to skin a cat.) 13.Nc3 Nc7³ Next there will follow ...Rg8 and ...g6, and there is not much White can do
about it.
232
B) 12.Ng6 Rh7 13.Be3 Na6µ is similar but worse; White is not even defending the e6-pawn and the
momentary misplacement of Black’s h7-rook is not particularly relevant.
12...Nxd4
13.Rc1
13.0-0-0 Na6 14.Be3 Nf5³ does not help particularly
13...Na6 14.Nce2 Nxe2 15.Kxe2 Nc7 16.Kf1!?
A great computer move. Now White gets full compensation for his pawn.
A) 16.Rh3 g5 17.hxg5 fxg5 18.Ng6 (18.Nxh5 Rg8³) 18...Rg8 19.Nxf8 Kxf8 20.Re1 Nxe6 21.Kf1 Rg6
22.Rxh5 Kg7³ White has just a bit less than full compensation.
B) 16.Ba5!? This move makes a commendable effort to tackle the problem at its root, viz. the knight on c7. In
case of the computer-sanctioned ...0-0-0 I could yet see Black going down in a painful ‘good knight versus
bad bishop’ type ending, however a strong reply is 16...Nb5!
233
Position after: 16...Nb5!
Threatening ...g5, and this forces evasive action. 17.Rce1 (17.Rhe1 g5µ) 17...b6! Else White can keep some
kind of bind, but this move seems to free Black, though he must still give the pawn back. 18.Bc3 (18.Bd2 g6!
sees Black escape for tactical reasons due to his control of the d4-square and some tactics on the bishop’s new
diagonal: 19.Nxg6?! Rg8 20.Nf4 Bh6µ) 18...Rh6 19.Kf3 (19.Kd1 0-0-0 20.Bd2 g6³ likely favours Black,
who will be playing ...Nd6 or ...Nd4 next – there is no really good discovery White can play with his knight.)
19...g6
20.Rh3 (20.a4 Nc7³ only helps Black, who now threatens ... g5 for real.) 20...Rh7! 21.Nxg6 Bh6³ Both sides
234
have undergone some contortions in the last few moves, but I think White’s have been more severe. Certainly
Black’s game looks close to being free, with his pieces out of their cages on the kingside and a possible motif
over the next few moves of trapping the White bishop.
16...Rg8!
Position after: 16...Rg8!
17.Ba5
A) 17.g4!? is rather grandiose in aim but Black can nevertheless just play his idea of 17...g5³
B) 17.Re1 g6µ and this is just a pawn.
17...b6 18.Rxc6 0-0-0 19.Bb4 Kb7 20.Rxc7+!
20.Rc1 a5 21.Ba3 g6 22.g3 Nb5 23.Kg2 Nxa3 24.bxa3 Bh6 25.Kf3 Rd6ƒ with a messy ending where only
Black can be better
20...Kxc7 21.Ng6
235
This bind is painful enough that Black is well advised to give back his gains immediately.
21...Rd6! 22.Bxd6+ Kxd6 23.Ke2 Kxe6 24.Nf4+ Ke5 25.Nxh5 g6 26.Ng3 Rh8=
Black can still claim some kind of moral edge on the grounds of pawn-structure, but with reasonable play on
both sides this should be a draw.
1.5) Main line with 7.Bd2 Qa6 8.e6 Qxd3 9.cxd3 f6 10.Nc3
1.e4 c6 2.d4 d5 3.e5 Bf5 4.h4 h5 5.Bd3 Bxd3 6.Qxd3 Qa5+ 7.Bd2! Qa6 8.e6! Qxd3 9.cxd3 f6! 10.Nc3
236
One way or another, White should try and play against the kingside light squares (especially g6) and it turns out
this is the best way to accomplish that.
10...Nh6 11.Nce2!
11.Bxh6 is again not a bright idea, because the Black rook on h6 is a surprisingly good piece. 11...Rxh6
12.Nce2 Na6 13.Nf4 Nc7³ Black can follow up with ...g6, ...0-0-0, ...Bg7 basically regardless of what White
plays, and always has a bail-out option of ...g5 in his back pocket.
11...Nf5 12.Nf4!
As seen in the last line, if White does not at least try to constrict the Black kingside then he is in trouble.
12...Nxd4 13.0-0-0 g5!
After detailed analysis of the attempts by Black to keep all his material, I am unconvinced by them, and so am
recommending this bail-out. Almost all of the following lines simplify rapidly.
14.hxg5
14.Ng6 Rg8 15.Nxf8 Rxf8 16.hxg5 fxg5 17.Rxh5 Nxe6=
14...fxg5 15.Nxh5
15.Nge2 Bg7 16.Nxd4 Bxd4 17.Nxh5 Na6 18.Bxg5 Nc5=
15...Rg8 16.Re1
237
16.Nf3 Nxe6 17.Rde1 g4! 18.Rxe6 gxf3 19.gxf3 Nd7=
16...Na6 17.Bc3
17.Nf3 Nxf3 18.gxf3 0-0-0= with ...Bg7 to follow in most cases.
17...Nb5 18.Nf3 0-0-0
Position after: 18...0-0-0
19.Ne5
This seems like the only really challenging setup. Now none of the following lines are forced, but note Black’s
26th, as it is of use in a number of positions where White attempts to re-sacrifice the pawn.
19.g4 Nxc3 20.bxc3 Rd6= Black will gang up on the e6-pawn, probably ceding the g5-pawn in return.
19...Nxc3 20.bxc3 Nc7 21.Nf7 Re8 22.g4
22.Re5 Rg6 23.Rhe1 g4= and Black will break free with ...Bh6 and ...Rf8 next.
22...Rg6 23.Re5 Rxe6 24.Rxe6 Nxe6 25.Kc2 Kd7 26.Re1
238
26...Ng7! 27.Nxg5
27.Ng3 e5=
27...Nxh5 28.gxh5 Bg7=
White’s nibble has evaporated.
1.6) 4.h4 h5 5.c4
1.e4 c6 2.d4 d5 3.e5 Bf5 4.h4 h5 5.c4 e6
239
6.Nc3
6.Bg5!? f6!? Not a popular reply, for some strange reason. 7.exf6 gxf6
A) 8.Bf4 Na6 9.a3 (9.Nc3 Nb4 10.Rc1 e5³ leads to messy complications that favour Black) 9...Ne7 10.Nc3
Qd7ƒ Black will castle long and then consider options like ...e5.
B) 8.Be3 Qb6 9.Qd2 dxc4 Threatening ...Bxb1! 10.Nc3 (10.d5 Qb4³) 10...Qa6! 11.Nge2 (11.b3 b5 12.bxc4
bxc4³) 11...Nd7 12.Ng3 Ne7!
Position after: 12...Ne7!

240
Seeing through to what really matters. Nxf5 isn’t even a real threat.
13.Be2 (13.Nxf5 Nxf5 14.Qe2 Nb6³) 13...Bd3 14.Rd1 (14.Bxh5+ Ng6 15.Nce2 0-0-0³) 14...0-0-0 15.Bxd3
cxd3 16.Qxd3 Qxd3 17.Rxd3 Nb6= Only Black can claim to be better in this endgame.
6...Ne7 7.Nge2
A) 7.Bg5 Qb6 8.Qd2 Nd7 9.Nf3 (9.Nge2 dxc4 10.Ng3 Bg6 – see 9.Bg5; 9.c5 Qa5 10.Nge2 b6 and after a
subsequent ...f6 White’s centre will be revealed to indeed have been built upon sand.; 9.0-0-0 f6 10.exf6 gxf6
11.Be3 Qa5µ is bad news for White) 9...f6
10.Be3 (10.Bf4 Ng6 11.Bg3 fxe5 12.cxd5 cxd5 13.dxe5 Bb4= Black is certainly not worse.) 10...Bg4! This
and Black’s next point out the precise nature of the problem with Be3. 11.Nh2 Nf5 12.Nxg4 hxg4= Black
will take on e3 and castle long.
B) 7.Nf3 Bg4!? 8.Bg5 Qb6 9.Qd2 Bxf3 10.gxf3 Nf5„
7...Nd7 8.Ng3 Bg6
241
9.cxd5
9.Bg5 doesn’t promise any advantage. 9...Qb6

A) 10.Na4 Qa5+ 11.Bd2 (11.b4 Qc7³ sees White left over-extended) 11...Qc7 12.c5 Nf5 13.Nxf5 Bxf5=
B) 10.Qd2 dxc4
B1) 11.Nge4 could be objectively better, but the position simplifies rapidly. 11...Bxe4 12.Nxe4 Nf5
242
13.0-0-0 Bb4 14.Qf4 (14.Qc2!? Nxd4 15.Qxc4 Nf5 16.a3 Be7 17.Bxe7 Kxe7„) 14...Nf8! 15.Bxc4 Ng6
16.Qf3 Be7 17.Bxe7 Ngxe7=
B2) 11.Bxc4 0-0-0 12.Na4 (12.0-0-0 is a forced variation at the end of which Black should be happy with
his compensation for an exchange: 12...Nxe5 13.Na4 Qc7 14.Bf4 Nxc4 15.Bxc7 Nxd2 16.Bxd8 Nc4
17.Bxe7 Bxe7 18.Nc3 Rd8„) 12...Qc7 13.Rc1 Nxe5!
Position after: 13...Nxe5!
It is a good idea to simplify the position as quickly as possible, because the computer doesn’t understand
White’s play well enough. 14.Bf4 Nxc4 15.Bxc7 (15.Rxc4 e5 16.Bxe5 Qxe5+ 17.dxe5 Rxd2 18.Kxd2 b5µ)
15...Nxd2 16.Bxd8 Nf3+! 17.gxf3 Kxd8 18.Nc3 Nd5© Once more, Black should be satisfied with what his
exchange has bought.
9...Nxd5 10.Nxd5
10.Be2 is probably less critical. 10...Qa5 11.Bd2 0-0-0 12.Nxd5 (12.0-0 as given by Vidit will lead to problems
with the h4-pawn: 12...Be7³) 12...Qxd5 13.Bc3 f6„
10...cxd5 11.Bd2 Be7 12.Be2
12.Ne2 Rc8! 13.Rc1 Rxc1 14.Qxc1 Bf5=
12...Rc8
243
Position after: 12...Rc8
Black ‘threatens’ a repetition.
13.Bxh5
13.Nxh5 Bc2 14.Qc1 Bd3=
13...Bxh5 14.Nxh5 g6 15.Nf4 Rxh4 16.Rxh4 Bxh4=
1.7) 4.Nd2 and 4.Be3
1.e4 c6 2.d4 d5 3.e5 Bf5
244
Position after: 3...Bf5
Now we shall examine a slightly different plan from White, one which tries to stop Black from playing ...c5 in a
convenient way.
4.Nd2!?
4.Be3 would be the first move of the other implementation. It may be marginally less accurate as it encourages
...Nf5 from Black in the future. 4...e6 5.Nd2 Nd7
A) 6.Ngf3 h6 7.Be2 Ne7 8.0-0 Bh7 9.Nb3 (9.c4 dxc4 10.Nxc4 Nd5= is again nothing for White; this is
normal for when he goes c4 without waiting for ...Nc8 or ...c5. See for instance Mammadova, G – Gasanov, E
Nakhchivan 2014) 9...Nf5 10.Bd2 Be7
245
Now it seems all of White’s plans involve playing g4 soon, as otherwise Black may even consider ...h5 and
...g5-g4 himself. 11.g4 (11.Bd3 0-0 12.g4 Nh4 13.Bxh7+ Kxh7 14.Nxh4 Bxh4 15.f4 Be7 16.f5 c5„)
11...Nh4 12.Nxh4 Bxh4 13.f4 0-0 14.f5 a5 15.a4 c5„
B) 6.Nb3 is of course the same as the 6.Be3 note (in the 4.Nd2 variation).
C) 6.f4 Nh6
C1) After 7.Be2 Ng4³ Black graps the bishop pair and will be well equipped to open the centre with ...c5
246
and/or ...f6.
C2) 7.h3 Bg6 8.g4 f6 9.exf6 (9.Ngf3 Qb6! forces concessions) 9...gxf6 10.c4 Qb6 11.Qb3 Nf7„ Both sides
have their trumps here, and the position is likely equal.
C3) 7.Ngf3 Be7 8.Be2 0-0 9.h3 Unfortunately, the structure means that if White wishes to castle, he had
best play this first, to avoid tricks with ...Ng4. However, the move implies the gambit of a pawn: 9...Bh4+
10.Nxh4 (10.Kf1 Bg3³) 10...Qxh4+ 11.Bf2 Qxf4 12.g3 Qg5 13.g4 Be4 14.Nxe4 dxe4„
Position after: 14...dxe4„
White has reasonable play for the pawn, but only in computer chess. Over the board, if White doesn’t swap
queens really soon, his king will find itself under an attack which is really difficult to defuse (... c5, ...f6 et
cetera).
4...e6 5.Nb3
247
Position after: 5.Nb3
The whole point is, of course, to play this, and discourage Black from playing ...c5. Here, a previous work has
recommended ...c5 anyway, followed by ...Bxc5 – which may well be playable, but first of all I have concrete
problems with one of the lines , and second of all, on a stylistic level, if I am to give up the bishop pair in this
structure I would probably rather give up the light-squared one. All of which brings me to my recommendation.
5...Nd7 6.Nf3
6.Be3 is not really a favoured move-order; I suspect this is because it allows Black to gain tempo off the e5-
pawn and White doesn’t really have a good way to protect it. 6...c5 7.dxc5 a6 8.f4 (8.Nf3 Bg4=) 8...Nh6 9.Nf3
Rc8 10.Be2 (10.c6 Rxc6 11.Nfd4 Rc8=) 10...Nxc5 11.c3 Bg4=
6...Ne7!? 7.Be2
248
7...Nc8
The solid option, playing to acheive ...c5 under optimal circumstances, and not accepting anything less.
7...c5?! was my previous choice, trying to accelerate the process a bit. Unfortunately, there is no sure-fire way
to safely regain the pawn which is thus sacrificed, although the line is probably still somehow playable.
A) 8.dxc5 Nc6 There is now no way for White to keep the pawn; for instance after Be3 Black can play ...Be4,
...Be7, castle and only then take back on e5. One of my opponents chose 9.Bf4 Nxc5 10.Nbd4 Bg6 11.0-0
Be7 12.c4 0-0= but this was comfortably equal for Black. In this position, he now blundered a pawn to an
instructive tactic with his next move (a3) which gave me my first Grandmaster norm! Izso, D – Fernandez, D
Budapest 2014.
B) 8.Nxc5! Nxc5 9.dxc5 Nc6 10.Be3 Bg4 11.0-0
249
11...Be7 (11...Bxf3?! 12.Bxf3 Nxe5 13.Bd4±) 12.Bb5! This retains a slight edge. (12.Bd4 Bxf3 13.Bxf3 and
now in Skytte, R – Hermansson, E Holbaek 2001, 13...0-0 14.c3 Qb8= would have sufficed to take back the
pawn – either ...Nxe5 or ...a5 and ...Qa7 will follow.) 12...Qb8!? 13.Bd4 0-0 14.Bxc6 bxc6 15.h3² White has
a good grip on his extra pawn, though the half-open b-file and bishop pair offer Black a measure of
compensation.
8.0-0
250
White will have to do this eventually and Black’s plan is the same anyway.
For instance, White can start by playing 8.a4 Be7 9.a5 a6 10.Be3 Na7 but by now either here or on the very
next move he will need to castle.
8...Be7 9.a4 0-0 10.a5 a6 11.Be3 Na7 12.c4
White can try and play without c4, but ultimately he will face problems preventing Black from going ...c5, e.g.
12.Bd3 Bxd3 13.Qxd3 Nb5 14.c3 Rc8=
12...dxc4
13.Nbd2
13.Bxc4 Nb5= is also not a problem; it matters little how many tempi it takes. Black to put a piece on d5,
because of how dominant the piece will be after that. The idea is ...Nc7-d5 or ...Be4-d5.
13...Nb5 14.Nxc4 Be4 15.Nfd2 Bd5=
Black can follow up with ...f6(5) or ...c5 next, and has counterplay of full value.
This was not, by any stretch of the imagination, a memory-intensive chapter, but there are, to my mind, two
key points.
The first was the improvement 14...Ne7, first seen in the game Liu-Vidit in line 1.2). I think this move well
and truly puts paid to White’s aspirations of getting an edge from the 4.Nc3 and 5.g4 system.
251
The second is the precise manner in which Black forces equality after White’s 12th in line 1.6). After some
other continuations, White is able to stir up trouble; the idea of the perpetual attack on the White queen
with ...Bc2-d3 is worth remembering.
Other than that, the chapter is more about understanding than memory, a pattern that will be reinforced
when we tie up the remaining loose end from it, in line 5.3).
252
Chapter 2
Short system Recommendation 1
1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nf3 e6 5.Be2 Ne7
Chapter guide
Chapter 2 – Short system, recommendation 1
1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nf3 e6 5.Be2 Ne7 6.0-0 c5 7.c4 Nbc6 8.dxc5 d4
2.1) Intro and 9.Qa4!?
2.2) Main line: 9.Qb3 Qd7 10.Rd1 Ng6 11.Nc3 f6 12.exf6 gxf6 13.--
2.3) Main line with 13.Nd5
2.4) Main line with 13.Nd5 (18.Qg3 Raf8 19.--)
2.5) Main line with 13.Nd5 (19.Nf6 & 19.Bh6!?)
The Short System, characterised by quick kingside development and the meeting of ...c5 by c4, is one of the
most difficult lines for any Caro-Kann player to face, and as we will see in the final chapter of this Part, one
which has been espoused by players right up to the very top.
In this chapter, I shall advocate the sharp 5...Ne7, which leads in our main line to one of the most irrational
positions in modern opening theory (another contender is found in Part I, chapter 2). The fearless Maxime
Vachier-Lagrave has been found on the White side of this position, and we examine a game he played with
Caro specialist Ding Liren in some depth in line 2.5). However, apart from this, there are numerous deviations
(2.1 through 2.3) and no fewer than four other significant options on move 19.
If you find this all a bit chaotic (perhaps even without looking further at the lines) I will not think less of you for
proceeding straight to the next chapter, in which I give an alternative.
253
2.1) Intro and 9.Qa4!?
1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nf3
The Short system is characterised by White holding back on c4, except possibly in certain lines where Black
plays ...c5 quickly, in which case White would want to open the position as soon as possible. As such, White is
not playing Nc3 either, but just developing pieces first.
4...e6 5.Be2
5.Be3 is the only other sensible move-order for reaching the Short System, but it is worse, because now
5...Qb6! forces concessions. Against moves like Qc1, b3, etc. Black can simply play ...c5 with equality. It
remains to briefly check 6.Nbd2!? Qxb2 7.Bd3 (7.Rb1 Qxa2µ) 7...Bxd3 8.cxd3 Qa3, however, while this
analysis could undoubtedly be continued and it would show a small amount of compensation for White, the
position is not critical and Black is in zero danger of being worse.
254
Position after: 8...Qa3
Most likely if he just continues normally with ...Ne7-f5, the position is objectively ‘³’.
5...Ne7
Not the most common, but this leads (in our main line) to some fascinating and irrational positions, where any
sane players without exhaustive preparation would likely just agree a draw to save themselves the randomness.
Later we examine an alternative repertoire based on 5...c5, which is much more solid.
6.0-0
6.c4 is rarely a good choice without Black having himself played ...c5 or at least expended tempi on a lengthy
walk with the king’s knight: 6...dxc4 7.Bxc4 Nd5=.
6...c5
255
Now White has a genuine choice, although in practice he has rarely chosen any move except the text.
7.c4
A) 7.Na3!? An interesting and flexible choice that was given ‘airtime’ in a recent White repertoire book. My
preferred move for Black is now: 7...Be4!? Forcing the question over White’s d4-pawn.
A1) 8.Nb5 Nf5!=
A2) 8.c4 Nf5 9.Bg5 Qd7 leaves White’s centre in meltdown while Black’s is well supported by his bishops;
White does not have enough dynamic options to avoid the position petering out into equality and might even
have to be more careful.
A3) 8.dxc5 This is now critical and is clearly the best way to play against the main weakness in Black’s
camp- the d6-square. 8...Nec6
256
Position after: 8...Nec6
A3.1) 9.Nd2 is a bit convoluted and doesn’t really threaten Black as long as he takes his time to regain the
pawn, e.g. 9...Bg6 10.Nb3 Be7 11.f4 a5! 12.Be3 a4 13.Nd4 Bxc5=
A3.2) 9.Bg5 Be7 10.Bxe7 Qxe7 11.Nb5 looks dangerous, but following 11...0-0= and ...Na6 next it
becomes clear that Black will regain his pawn with a knight recapture at some point in the next three
moves.
A3.3) 9.Be3 Nd7 10.Nb5! Bxf3 11.Bxf3 Ncxe5 12.Be2 a6 I don’t trust either way of taking the c5-pawn;
this move seems safe to me. 13.Nd6+ Bxd6 14.cxd6 Rc8„
Position after: 14...Rc8„

257
The position is objectively level. White has done well to force us to give up the bishop pair, but at the same
time, moves like ...Nc4 or ...Nf6 are coming soon, and the d6-pawn may not be long for this world.
Meanwhile, there is no real attack against the black king.
B) 7.dxc5 Nec6 8.Be3 Nd7
9.c4 (9.Bb5 is an interesting tactic, but I think Black can ignore it: 9...Bxc5 10.Bxc5 Nxc5 11.Nd4 0-0!=)
9...dxc4 10.Nbd2 Bxc5 11.Nxc4 0-0„
Position after: 11...0-0„
258
There will be a trade of dark-squared bishops and at some moment Black will play ...Bg4 (e4) to compensate
for his lack of dark-square control by play against e5.
7...Nbc6 8.dxc5 d4 9.Qa4!?
Position after: 9.Qa4!?
White almost invariably picks 9.Qb3, on which much more later.
9...Ng6 10.Rd1 d3!
After the more obvious 10...Bxc5, there followed the unpleasant 11.b4 in Tabatabaei, M – Svane, R Moscow
2017.
11.Be3
It is to White’s advantage to have the c5-pawn protected before embarking on forcing variations.
11.Bxd3 Bxd3 12.Ne1 Ngxe5³
11...Be7 12.Bxd3
White cannot delay the tactics any longer: 12.Nc3 0-0µ and in the process of castling away, Black has not only
increased his king safety but also unpinned the c6-knight.
12...Bxd3 13.Qb3 Ngxe5 14.Nxe5 Nxe5 15.f4 Nxc4 16.Rxd3 Qa5
259
Position after: 16...Qa5
17.Nd2
17.Bf2 Bxc5!µ
17...Nxe3 18.Qxb7 0-0 19.Nb3!
White is worse if he takes the bishop immediately, but in this version, the position remains equal.
19...Qa4 20.Qxe7
20.Rxe3 Bxc5 21.Nxc5 Qd4=
20...Qxf4„
260
Position after: 20...Qxf4„
Black’s active pieces compensate for his worse majority.
2.2) Main line: 9.Qb3 Qd7 10.Rd1 Ng6 11.Nc3 f6 12.exf6 gxf6 13.--
1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nf3 e6 5.Be2 Ne7 6.0-0 c5 7.c4 Nbc6 8.dxc5 d4 9.Qb3
This is most critical.
261
9...Qd7!
Possibly the only move to equalise for Black. We shall briefly dispense with the instructive alternatives to
explain why I am not recommending them- some notable grandmasters have in their time endorsed most of the
following.
A) 9...Qc8 was recently played as an attempted improvement (over his previous ...Qd7) by Caro expert Aryan
Tari, but I think his opponent’s later play is easy enough to improve. 10.Rd1 Ng6 11.Nxd4 Nxd4 12.Rxd4
Bxc5 13.Rd1 0-0 14.Nc3 Qc6
15.Nb5! (15.Bf3 Qc7 16.Na4 Nxe5= was the game continuation in Vidit, S – Tari, A Moscow 2017) 15...Be4
Else Nd4 and White gets an edge more easily. Initially, I was impressed by the resources in Black’s position,
but then a set of wrinkles for White began to emerge, of which I show the main one. 16.Bf1 a6 (16...Nxe5
17.Qg3 f6 18.Be3² with Nd4 to follow might be playable for Black, but it’s not pleasant.) 17.Nd4 Qc7
18.Qe3! Qxe5 19.b4! Bxb4 20.f4!²
262
Position after: 20.f4!²
White stands better in the two-pieces-for-rook game that follows.

B) 9...Qc7?! This leads to interesting play, where White needs an as yet unnoticed improvement. 10.Na3 a6
11.Qa4 0-0-0?! (11...Rd8 12.b4 Ng6 13.Bd1 Rd7 14.Nc2² is the lesser evil, but still unpleasant) 12.b4 Ng6
13.Nb5!!N (13.Bg5 is given by Dreev, but following 13...f6 14.exf6 gxf6 15.Bxf6 Nf4 16.Bd1 Rg8! 17.g3
Be7 18.Bxe7 Nxe7!∞ Black obtains full compensation in the messy position.) 13...axb5 14.cxb5 Now careful
analysis shows that the best Black can do is to sacrifice his queen. 14...d3 15.b6 dxe2 16.bxc7 exf1=Q+
263
17.Kxf1 Kxc7 18.b5 Bd3+ 19.Ke1 Bxc5 20.bxc6± White should win with some care.
C) 9...Ng6?! 10.Qxb7 Qc8 11.Qxc8+ Rxc8
might appear to gain back both the pawns, but White can extract a concession- the bishop pair. The following
game is also cited by Dreev and contains no noteworthy inaccuracies from either side: 12.a3 a5 13.b4! axb4
14.axb4 Nxb4 15.Nxd4 Bxc5 16.Nxf5 exf5 17.Ra5² Kosteniuk, A – Bulmaga, I Astana 2013.
10.Rd1 Ng6 11.Nc3 f6!?
11...Bxc5 12.Qb5² has been analysed extensively - the verdict is that White wins the bishop pair in a fairly
open endgame position and so is slightly better.
12.exf6 gxf6
264
This position has been played from the Black side by two members of the 2700 club, so it should not be
underestimated!
13.Qa4!?
This is an interesting attempt by White to get ‘something like’ the previous note. The most critical move (the
violent 13.Nd5) is seen next. Other minor lines are below:
A) 13.Qa3 keeps the pawn, but the doubled c-pawns are both weak. 13...Rd8 14.b4 (14.Be3 e5©) 14...d3
15.Be3 Clearly the most natural move. (15.Bf1 a5! 16.bxa5 Nge5 17.Nxe5 Nxe5© leads to Black having fully
adequate compensation; next there can follow (even if not ...Nxc4) either ... Qc6 or ...Qd4, targeting every
aspect of White’s structure.) 15...h5
265
Making a generally useful move and challenging White to do the same. Black is fine and here are some
sample lines: 16.b5 (16.Ne1 Nge5 17.b5 Nd4=) 16...Nce5 17.Qxa7 Qc8! 18.Nxe5 Nxe5 19.Bf3 (19.Bf1
Nxc4 20.c6 bxc6 21.b6 Rh7 22.Qa4 Nxe3 23.fxe3 Rhd7!„) 19...Nxf3+ 20.gxf3 Rg8+ 21.Kh1 Be7„ The
sources of Black’s counterplay are hopefully obvious and his position seems easier to play.
B) 13.Na4?! also keeps the pawn, but this is a terrible idea. The knight is immobilised, as are White’s
queenside pawns. 13...e5³ Black can continue with ...Rg8, ... 0-0-0, etc.
13...Bxc5
266
14.Qb5
14.b4 Bxb4 15.Nxd4 Nxd4 16.Rxd4 Qxa4 17.Nxa4 Ne5=
14...b6 15.Na4
15.b4 is a better version of the last note, but Black is still fine: 15...a6 16.Qa4 Bxb4 17.Nxd4 Nxd4 18.Rxd4
Qxa4 19.Nxa4 Rb8=.
15...0-0
15...e5!? 16.Nxc5 bxc5 17.Qxc5 h5© is another possibility, but it cannot be to White›s advantage that Black
gets to castle with the text.
16.Nxc5 bxc5 17.Qxc5 e5
Black has good compensation here, and while it is a hard position to analyse exhaustively, the sources of his
play should be obvious. Possible plans include ...Nd8-e6-f4; ...a5/...Nb4, or even just playing for ...d3 and ...e4.
18.Ne1
18.Bh6 forces Black’s hand a little bit more than might be a good idea from a practical perspective. 18...Rfc8
19.Qd5+ Else Black gets great play based on ...d3 and ...Nd4 motifs. 19...Qxd5 20.cxd5 Nb4 21.d6 Rc2„
18...a5 19.Bd2
This seems most natural; I think most people would be intimidated by the idea of ...Nb4 and the ensuing queen-
267
trap ideas. The computer bears out this impression.
19...e4!?©
Position after: 19...e4!?©
Here a number of moves were possible, but I chose the one that illustrates an idea I didn›t even mention, i.e.
...Nge5. Black is completely fine in this tactically and strategically rich position.
2.3) Main line with 13.Nd5
1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nf3 e6 5.Be2 Ne7 6.0-0 c5 7.c4 Nbc6 8.dxc5 d4 9.Qb3 Qd7! 10.Rd1 Ng6 11.Nc3
f6!? 12.exf6 gxf6 13.Nd5!
268
Position after: 13.Nd5!
13...Kf7 14.g4!
A) It makes little sense to move the knight back so soon, and indeed, 14.Ne3 Nf4 15.Bf1 Rg8= strongly
encourages White to walk into the perpetual check with ...Nh3-f2. There was also a more combative move:
15...Be4 would probably repay investigation.
B) 14.Qb5 Qc8 15.b4 (15.Nxd4 Nxd4 16.Rxd4 a6µ) 15...Bc2 16.Rd2 (16.Rxd4 a6 17.Qb6 exd5µ) 16...Be4
17.Nxd4 Rg8! Black develops adequate play, e.g. 18.Nc3 Bxg2 19.Kxg2 e5 20.Nf3 Qh3+!! 21.Kxh3 Nf4+
269
22.Kh4 Ng2+=.
C) 14.Nxd4? Nxd4 15.Rxd4 Bxc5µ
D) 14.Rxd4!? Against accurate defense this exchange sacrifice does not acheive anything, e. g. 14...Nxd4
15.Nxd4 Be4 16.Qh3!
Position after: 16.Qh3!
16...Bxd5 17.cxd5 Qxd5 18.Be3 Bxc5 and White will have to dig deep for his compensation after 19.Nxe6.
14...Bxg4 15.Ng5+ fxg5 16.Bxg4 Nce5
270
Position after: 16...Nce5
17.Bxg5!
17.Qh3? h5!–+ immediately won a piece in Praneeth, S – Vidit, S Thiruvarur 2015.
17...Bxc5 18.Nf6?!
18.Qg3!? is a logical novelty, which has been ‘visited’ many times and so it will be played before long. It will
occupy us for the rest of the chapter.
18...Qc6
Now this was in fact slightly more pleasant for Black in Vachier Lagrave, M – Ding, L Wijk aan Zee 2015; it is
possible White was even still in preparation but then just mixed up his lines a little bit with this. I give one of
the computer draws without much comment, and the game itself will be transposed into later.
19.Bh5™ 19...h6!N 20.Bh4 Kg7 21.Qg3 Rhg8 22.Re1™ 22...Kh8 23.Nxg8 Rxg8 24.Bf6+ Kh7
271
25.Bxe5!
25.Bxg6+? Rxg6 26.Bxe5 Rxg3+ 27.Bxg3 d3–+ causes White to miss a critical tempo, and this makes all the
difference!
25...Nxe5 26.Rxe5 Rxg3+ 27.hxg3 d3
Threatening the bizarre skewer ...Qd6. However, after...
28.Rae1=
White’s counterplay is enough to secure a draw.
2.4) Main line with 13.Nd5 (18.Qg3 Raf8 19.--)
1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nf3 e6 5.Be2 Ne7 6.0-0 c5 7.c4 Nbc6 8.dxc5 d4 9.Qb3 Qd7 10.Rd1 Ng6 11.Nc3
f6 12.exf6 gxf6 13.Nd5 Kf7 14.g4 Bxg4 15.Ng5+ fxg5 16.Bxg4 Nce5 17.Bxg5 Bxc5 18.Qg3!?N 18...Raf8!
272
Position after: 18...Raf8!
Now we arrive at a late critical juncture, though not the latest one in the book, by some way. White’s most
critical moves are 19.Nf6 and 19.Bh6, on which more later, but minor moves are not necessarily trivial to meet
either.
19.Bh5
A) 19.Re1 is logical, but just a little too slow to be truly critical. 19...Nxg4 20.Qxg4 h6!
A1) Following 21.Qf3+ Kg8 22.Nf6+ Kg7 White can choose between the obvious perpetual, and a
simplification such as 23.Nxd7 Rxf3 24.Bd2 Be7! 25.Rxe6 Rf5„ when between the threatened knight trap
(...Rd8) and the pressure on the f2-pawn, Black has fully sufficient compensation.
A2) 21.Bf6 h5 22.Qh3 Rh7
273
Position after: 22...Rh7
White should now seek desperate solutions to the problem of his tangled pieces. 23.Be5 Nxe5 24.Rxe5 Kg8!
One of the stranger ways in which Black can threaten the d5-knight in this line- and there are always new
ones. 25.Rg5+ (25.Qxe6+ Qxe6 26.Rxe6 Rg7+ 27.Kf1 d3 28.Rf6 Rgf7³ is ‘safer’ but still marginally more
pleasant for Black) 25...Rg7 26.Rxg7+ Kxg7 27.Kh1 Rf5
Position after: 27...Rf5
28.Qg3+™ (Strangely enough, after 28.Rg1+ Kh6–+ White is unable to get past the defensive setup of rook
on f5 and queen on f7, at least before Black either wins the knight or arranges ...Rxf2 and ...Qf3.) 28...Kh6
29.Nf4 Qc6+ 30.f3 e5 31.Rg1 Qf6 32.Nd3!= Black can now consent to an equal endgame after ...Rxf3, or
274
try his luck again with ...Bd6.
B) 19.b4 Nxg4 20.Qxg4 h6 makes White scramble somewhat to avoid an unpleasant position. His best is
21.Bf6! Rhg8 22.Kh1 when after 22...Qc6= the chances are approximately level.
19...Kg7
20.f4!?
20.Bf4 is very ‘engine’: White’s position appears to be falling apart but he just about has a draw. 20...Nxc4
21.Bxg6 hxg6 22.Be5+ Nxe5 23.Qxe5+ Kf7 24.Qf4+ Kg7 25.Qe5+=
20...Nxc4! 21.Nf6!
After 21.Bf6+ Kf7 22.Bxh8 Qxd5 23.Bf3 d3+ 24.Kh1 Qf5©
275
Position after: 24...Qf5©
Black still has a raging attack while White›s is taking a bit of a break. I would say this is easier to play for
Black, although the machine confidently asserts that 25.Bc3 Rg8 is ‘0.00’.
21...d3+ 22.Kh1 Rxf6 23.Rxd3!
23.Bxf6+ Kxf6 24.Qg5+ Kg7 25.Qxc5 Rc8 seems to give Black some kind of tenuous grip on the reins of the
position
23...Bd6 24.Bxf6+ Kxf6
276
Position after: 24...Kxf6
Reaching -not unusually for this book- a position where both kings are rather weak, though here there are stacks
of alternatives given, starting right on move 3!
25.Rg1
White seems to have to play for Qg5+, one way or another.

A) Following the original 25.Rd4 Qb5 26.Qc3?! it is even Black who whips up an attack: 26...Bb4 27.Qxc4
Qxh5 28.Rf1™ 28...Qh3‚ and White appears to be in some kind of trouble.
B) 25.Qg5+ immediately only helps Black get safe: 25...Kg7 26.Rad1 (26.Rg1 Qc6+ 27.Bf3 Qc5!³) 26...Qc7
Now White will have to use one of his drawing lines, the simplest being 27.Bxg6 hxg6 28.Rxd6 Nxd6
29.Qe5+ Kh7 30.Qxd6 Rc8! 31.Qxe6 Qxf4=.
25...Rg8 26.Qg5+
Else Black plays ...Qb5 and may even be more comfortable. A sample line now is:
26...Kf7 27.Rgd1 Qe7 28.Qg2 Kf8=
Position after: 28...Kf8=
The position continues to be a horrendous mess, but Black’s knights are every bit as tricky as White’s heavy
pieces.
2.5) Main line with 13.Nd5 (19.Nf6 & 19.Bh6!?)
1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nf3 e6 5.Be2 Ne7 6.0-0 c5 7.c4 Nbc6 8.dxc5 d4 9.Qb3 Qd7! 10.Rd1 Ng6 11.Nc3
f6!? 12.exf6 gxf6 13.Nd5 Kf7 14.g4! Bxg4 15.Ng5+ fxg5 16.Bxg4 Nce5 17.Bxg5! Bxc5 18.Qg3!?N
277
18...Raf8! 19.Nf6!?
Position after: 19.Nf6!?
Given as almost winning by cloud engines, even on fairly high depth, but after some persuasion they recognise
the resilience of Black’s position. We reach in our main line an endgame that is almost certainly a draw with
best play (and White needs to play very well too) so even though the engines give ‘²’ I shall not be discouraged
from recommending the line.
19.Bh6 Rd8!
Position after: 19...Rd8!
278
Black threatens, unbelievably, to take the d5-knight after an exchange on g4. So White’s replies are slightly
constrained by that.
A) 20.Bh3 Rhg8! 21.Bg5 h6 22.Bxd8 Qxd8 23.Bg2! exd5 24.Bxd5+ Qxd5 25.cxd5 Nf4 26.Kf1 Rxg3
27.hxg3 Nxd5„ The rooks do not co-ordinate well yet, and the d-pawn is far advanced, so this seems to give
Black an acceptable level of play. There was an interesting alternative in 23...Kf8!? if this turns out to not be
satisfactory.
B) 20.Bh5 leads to an amusing repetition where Black wants to trade queens and White is avoiding it:
20...exd5 21.Qxe5 Qg4+ 22.Kf1! Qh3+ 23.Kg1 Qg4+=.
C) 20.Ne3?! Qc6 21.Ng2 Nxg4 22.Qxg4 Rhg8‚ is too passive for White.
D) 20.Nc3?! Bf8! is also quite passive: to avoid really major headaches White has to play 21.Bh5! Bxh6
22.Qxe5 d3!„ when his position is delicate but there is still a fair amount of end-to-end play.
E) 20.b4! Bf8 21.Bg5 Nxg4 22.Qxg4 exd5 23.Qf3+ Ke8 24.Bxd8 Qxd8 25.Rxd4 Bxb4
Position after: 25...Bxb4
We have seen an essentially forced variation for the last few moves. Now Black wants ...Qh4, ...Be7 and
...Rf8 to balance his attacking and defensive responsibilities. 26.Rb1! This is most challenging. (26.Rxd5 Qh4
27.Rad1 Be7 28.Qf5 Qxc4= holds the balance; 26.c5 Bxc5 27.Rxd5 Qc8= similarly holds; Black’s defensive
moves are not difficult to find.) 26...Qg5+ 27.Kf1 Rf8 28.Qxd5 (28.Qe2+ Be7 29.Rxb7 Qf6 30.Qb2 dxc4= is
another line in which White’s major piece attack has been successfully mitigated; White should probably now
be more careful.) 28...Qxd5 29.Rxd5 a5 30.a3 Bxa3 31.Rxa5 Rf3!
279
Position after: 31...Rf3!
It is important to keep rooks on. After 32.Rxb7 Be7= Black should draw with a modicum of care. Objectively,
this probably represents a better winning try for White than 19.Bh5 or 19.Nf6, but it is still not bad for Black.
19...Qc6 20.Bh5
White has to keep this bishop a) alive and b) on the d1-h5 diagonal, and allowing ...d3 with tempo is senseless.
Thus, the text is forced.
20...Kg7™
Now we have transposed back into the source game, which proceeded down an inferior line (though this, like
much else in the position, is not obvious, immediately or otherwise, to the non-machine-aided eye.)
21.b4!
This is what the engines want, and capturing the pawn leads to nothing but trouble after f4. Both sides must
continue to play with tempo!
21.Re1 Rxf6 22.Rxe5 Rf7! 23.Bxg6 hxg6 24.Bf4?! (24.Rae1 Bd6 25.Bf4 Rf6=) 24...Be7 25.Re2
280
So far following Vachier Lagrave, M – Ding, L Wijk aan Zee 2015, where 25...Rh5!µ would have in all
likelihood left White feeling rather sorry for himself; the most obvious threat is ...Rxf4 and ...Rg5+.
21...Be7!
22.b5!
White gets the Black queen away from pressuring f3. The fact that it can take a pawn is really not relevant, even
281
though that’s Black’s best.
A) 22.f4?! Bxf6 23.Bxf6+ Rxf6 24.fxe5 Rf5 25.Bxg6 hxg6 26.Rxd4 does leave White a pawn up, but you
have to be quite delusional to think that’s even close to the most important feature of this position. In the real
world, 26...Rhh5!‚ comes and gives Black a very good game based on obvious moves directed against the
open White monarchs.
B) 22.Ne4? Qxe4 23.Bxe7 Rf5µ leaves White in an even more immediate tangle.
22...Qxc4 23.Rac1!
A) 23.Bxg6 Bxf6 24.Be4 looks scary, but instead of worrying about White’s discovered attacks Black is able
to set up his own: 24...Rhg8! 25.Kh1 Ng6 26.Rac1 Qe2= and Black’s queen has reached its ideal defensive
post.
B) After 23.Ne4?! Bxg5 24.Nxg5
Position after: 24.Nxg5
24...Kh6!! it transpires that not only must White worry about the threat to his h5-bishop, but also there is
nothing better he can do with it than 25.Bxg6 Nxg6µ when Black has fended off White’s attack and stands
ready to counter with ...Qd5, ...Nf4, ...Rhg8, etc.
23...Qxb5
282
Position after: 23...Qxb5
24.Ne4
Only this precise order causes Black significant problems.

A) 24.Rb1!? leads to some great drawing lines: 24...Qa5 25.Rxb7 Rxf6 26.Rxe7+ Nxe7 27.Bd2+ (27.Bxf6+
Kxf6 28.Qh4+ Kf5= and White has no better than the inventive perpetual Bg4-h5.) 27...Kf8 28.Bxa5 Rg8
29.Bc7 Rxg3+ 30.hxg3 N5c6© White has compensation for his pawn, but not more.
B) 24.a4?! is not a ‘real’ gain of tempo since on a5 the Black queen is moderately useful, protecting the e5-
and c7- squares. 24...Qa5 25.Ne4 Bxg5 26.Nxg5 Rhg8!!³
C) 24.Rc7?! Rxf6 25.Rxe7+ Rf7³
24...Qd5 25.Bxe7 Qxe4
283
Position after: 25...Qxe4
26.Re1!
26.Bxf8+ Rxf8© gives Black amazing compensation based on ...Nf3 or ...Nf4 ideas, so White must try and
disturb the e5-knight prior to taking any material.
26...Qf5 27.Bd6! Nd3 28.Rc7+ Kh6 29.Bxg6 Qxg6 30.Bxf8+ Rxf8 31.Qxg6+ hxg6 32.Rxe6
32.Rd1 Nf4 33.Rxb7 Rc8© still offers Black amazing compensation.
32...Rxf2 33.Rd7 Rd2!∞
We have almost reached tablebase territory and still White is not able to consolidate his advantage. The best he
might be able to do is Re4, obliging Black to take on a2, and then we would get a fascinating endgame.
You have just seen a whole stack of completely irrational positions in quick succession. Obviously, it is
pointless asking anyone to memorise all of the top continuations, unless they possess a photographic
memory.
What I will say though is that for most purposes, it will probably suffice to just memorise up to the move
18...Raf8, and let the chaos take care of itself after that.
284
Chapter 3
Short system Recommendation 2
1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nf3 e6 5.Be2 c5
Chapter guide
Chapter 3 – Short system, recommendation 2
1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nf3 e6 5.Be2 c5 6.Be3 cxd4 7.Nxd4 Ne7
3.1) 6th, 7th, and 8th move minor lines for White
3.2) 8.c4 Nbc6 9.Qa4 a6 10.cxd5!?
3.3) 8.c4 Nbc6 9.Qa4 a6 10.Nc3
3.4) Main line: 8.0-0 Nbc6 9.Bb5 a6 10.Bxc6+ bxc6 11.b4
3.5) Main line with 11.c4 Qd7 12.cxd5
3.6) Main line with 11.c4 Qd7 12.Nc3
My second recommendation against the Short system is one which I share with the Indian grandmaster Vidit. It
is the main move 5...c5, which almost inevitably proceeds with the follow-up 6.Be3 cxd4 7.Nxd4 Ne7
(alternatives are seen in the first line). Play is substantially quieter than in the previous chapter, though the price
we pay is an increased risk of asphyxiation, especially in lines 3.3) and 3.6).
Unless you are one of those maniacs who intends making the previous chapter their exclusive choice (I salute
you!) this is probably the single most important chapter in the book, with every line being somewhat critical.
The most important new contribution here, in my opinion, is the rehabilitation of 11...Qc8 in line 3.3) – I have
explained why the alternatives do not merit being recommended. The move requires a string of tactical nuances
to justify it, and while they hold up well to analysis, they deserve to be memorised, at least up to the diagram in
285
the note to White’s 18th.
3.1) 6th, 7th, and 8th move minor lines for White
1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nf3 e6 5.Be2 c5
This is the second line I shall be recommending against the Short system. The two lines require the same
amount of work, but in this one, the value of each move is a bit lower, whereas in the other, the value is higher
but White will also definitely have done significant homework.
6.Be3
This is the critical test and scores a full 15% better than the alternatives!
A) 6.0-0 Nc6 7.c3 (or indeed the same White moves in the order) lets Black get the ‘improved French’ setup
which is so often touted as a benefit of the Caro – Kann. For the price of one tempo, his bishop has managed
to fly over the chain of pawns on light squares. 7...cxd4 8.Nxd4 (8.cxd4 Nge7 Planning an assault on the d4-
pawn, which may be aided by ...Bg4xf3 if necessary. 9.Nc3 Nc8 10.Be3 Nb6 11.Rc1 Be7= or 9.Be3 Bg4
10.Nbd2 Nf5=) 8...Nge7
286
Position after: 8...Nge7
A1) 9.Qa4 Qd7= threatens ...Nxd4 and so Black is fine

A2) 9.Bg5!? tries to tie Black’s pieces up, but Black can play ‘past’ the g5-bishop. 9...Nxd4 10.cxd4 Qb6
11.Qd2 Nc6 12.Rd1 (12.Be3 Bb4 13.Nc3 0-0= is similar) 12...Bb4 13.Nc3 0-0 14.a3 Bxc3 15.Qxc3 Rac8=
A3) 9.Nxf5 Nxf5 10.Bd3 g6
A3.1) 11.Bxf5 gxf5= is not threatening; Black can play on the kingside next with ...Qh4 or ...Rg8, or
through the centre with ...d4.
287
A3.2) 11.Qe2 Bg7 12.f4 0-0 13.Kh1!? (13.Nd2 doesn’t stop ...f6 either. 13...f6 14.exf6 Qxf6=) 13...f6
14.g4 Nfe7!? (It is possible to not give the pawn away, e.g. 14...Nh4 15.exf6 Qxf6 16.Be3 e5„) 15.exf6
Rxf6 16.g5 Rf8 17.Qxe6+ Kh8© Black has compensation here. What it makes sense to remember is that
...d4 and ...Nf5 are coming next.
A3.3) 11.f4 h5 12.Kh1 d4= is not inspiring for White, partly due to the h-file motifs.
A3.4) 11.Re1 Bg7
12.f4 (12.Bf4 walks into future x-rays on the f-file, since ...f6 is unstoppable: 12...0-0 13.Qd2 f6ƒ) 12...0-0
Now Black wants ...d4 or ...f6. The main move given by Vidit, and the most logical, is 13.Nd2 and here I
am convinced by his analysis of 13...d4 although ...f6 was fine too. 14.Nb3 (14.Ne4 dxc3 15.bxc3 Nb4!=;
14.c4 Ne3 15.Qe2 f6ƒ) 14...dxc3 15.bxc3 Qb6+ 16.Kh1 Qf2!? 17.Rf1 Ne3=
B) 6.c4 looks vaguely threatening too, so let us take a quick look. 6...dxc4 7.Bxc4 Nc6! 8.Be3 cxd4 9.Nxd4
Nxd4 10.Qxd4 Qxd4 11.Bxd4 a6= It transpires that Black’s development is not significantly behind White’s,
and makes up for in quality what it lacks in quantity.
6...cxd4 7.Nxd4 Ne7
288
Reaching a critical juncture, White’s choices at which will occupy us for the rest of this chapter. Right here,
however, we shall just look at some minor moves, leaving 8.c4, 8.Nd2 and 8. 0-0 for later.
8.Bg5
A) Against the natural yet innocuous (and therefore little-played) 8.Nc3 I shall recommend, for once, avoiding
giving up the bishop-pair. 8...Bg6
289
A1) 9.0-0 is too slow; White’s light-square play needs this time to get off the ground. 9...Nec6 10.f4 Nxd4
11.Bxd4 Nc6 12.Bf2 (12.g4 Bxc2³) 12...Rc8= Black may follow up with ... h5 and ...Be7, with a
comfortable position.
A2) 9.h4 h5 10.Ncb5 was tried in Saric – Laznicka, Bundesliga 2015, but it should get nowhere: 10...Nec6
11.c4 Bb4+ 12.Kf1 dxc4=
A3) 9.f4 Nbc6 10.Ncb5 Nf5 11.Nxf5 Bxf5 12.0-0
White has probably one of the best versions of the ‘Knight swap in advance Caro’ complex, but it is not
quite enough for an objective advantage: 12...a6 13.Nd4 Nxd4 14.Bxd4 Rc8 15.Bd3 Bxd3 16.Qxd3 g6„
B) 8.Nd2 An unassuming option, but nevertheless one quite enthusiastically recommended by Dreev. The
idea is simply to reinforce White’s d4-knight with Nd2-f3, castle and only then think about c4. 8...Nbc6
9.N2f3 Be4!
290
Black needs to not only attack f3 but also reinforce the light squares in his centre and queenside. Originally, a
more maximalist approach with ...Bg4 was intended, but White is able to engineer a favourable c4-break, and
can probably get an edge providing he is willing to make some incisive sacrifices. 10.0-0 a6 11.c4 White has
played in such a way that it is hard to think of other things he can reasonably do. (11.Rc1 Rc8 will likely
transpose in the end too. If White does absolutely nothing, e. g. 12.h3 Qd7 13.c3 then it is reasonable for
Black to take the knights off. 13...Nxd4 14.Nxd4 Nc6=) 11...dxc4 12.Bxc4 Rc8 13.Rc1 Nxd4
B1) 14.Qxd4 Bxf3³ is given by the engine as a realistic option for White, but here it has overrated the bishop
291
pair. In the endgame that follows White lacks any sort of break at all and only Black can have winning
chances.
B2) 14.Nxd4 Bd5 15.Qe2 Rxc4 16.Rxc4 Bxc4 17.Qxc4 Nd5=
8...Qd7!
The structure after the two minor piece swaps is not so bad that we need to get worked up about it.
9.Bxe7
9.c4 leads to another forcing line: 9...Nbc6 10.Nxf5 Nxf5 11.cxd5 Qxd5 12.Qxd5 exd5 13.Nc3 h6 (13...d4!? is
possible but more loosening, in my opinion) 14.Bd2 Nfd4 15.Nxd5 0-0-0 16.Ne3 Nxe5=
292
Position after: 16...Nxe5=
Black can snap off White’s bishop pair at will, and complete development the following move. Then the
symmetry of the position means it cannot be evaluated as anything except equal.
9...Bxe7 10.Nxf5 exf5 11.Nd2
Now I believe White should give a pawn and Black should take the pawn in one way or another, but there are
some subtleties.
293
11...Qe6
11...f6!? 12.e6 might be a slightly better version for White, echoing Carlsen – So, Tata Steel 2018 and others.
12.f4
12.Nf3 0-0 13.0-0 Nd7= will snap off the e5-pawn and White cannot do more than some day recovering it on
d5. Note that 14.Qd4?! Bc5 15.Qf4 f6ƒ would not be a great idea for White.
12...f6 13.Nf3 fxe5 14.Nxe5 Nc6=
Position after: 14...Nc6=
Black will probably proceed by castling long; a fighting game can ensue.
3.2) 8.c4 Nbc6 9.Qa4 a6 10.cxd5!?
1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nf3 e6 5.Be2 c5 6.Be3 cxd4 7.Nxd4 Ne7 8.c4 Nbc6
294
Position after: 8...Nbc6
9.Qa4
9.Nc3 Nxd4 10.Bxd4 dxc4= is at least fine for Black, who can continue with the usual ...Nc6, ...a6, ...Be7 stuff
in case White takes on c4, or ...Bd3 otherwise.
9...a6 10.cxd5!?
This forces an unusual kind of play from Black. The main move is 10.Nc3, on which more next.
A) 10.Na3 is thematic, but the position rapidly becomes simplified and it is hard to try and play for anything
with White. 10...Qa5+ 11.Qxa5 Nxa5 12.0-0 dxc4 13.Nxc4 (13.Nxf5 Nxf5 14.Bb6 Bxa3! 15.bxa3 Nc6 16.f4
g5!= Leko, P – Le, Q Dortmund 2010) 13...Nxc4 14.Bxc4 Be4 15.Rac1 Bd5
295
16.Be2 Rc8 (16...g5!? is an interesting idea from Vidit that the reader can explore if they like.) 17.f4
(17.Rxc8+ Nxc8 18.Rc1 Kd7 19.a3 Ne7 20.f4 Nc6= was similar in Adams, M – Ragger, M Germany 2011)
17...Rxc1 18.Rxc1 Kd7= Black followed up with ...g6, ....Nc6 etc and just drew in Svidler, P – Epishin, V St
Petersburg 1997.
B) 10.0-0 dxc4 11.Rd1 Bd3 12.Bxd3 cxd3 13.Rxd3 Qa5=
C) 10.Nd2 Qd7! 11.0-0 Nxd4 12.Qxd7+ Kxd7 13.Bxd4 Nc6 14.Bb6 Nxe5 15.cxd5©
White can make just enough trouble to claim compensation for his pawn, but Black’s development is good
enough to keep his opponent’s play firmly within that limit.
296
Position after: 10.cxd5!?
10...Nxd5!?
10...b5!? 11.Nxc6 Nxc6 12.Qf4 Nb4 13.0-0 Nxd5 14.Qg3 Nxe3 15.fxe3 Bg6 is another option, but I doubt
Black’s solidity after a4.
11.Nc3!
A) 11.Nxc6 Qd7 12.0-0 Qxc6 13.Qxc6+ bxc6=

B) 11.0-0 Nb6 12.Qd1 Nxd4 13.Qxd4 Qxd4 14.Bxd4 Nd5=
11...Nxe3 12.fxe3
12.Nxc6 Qd7 13.fxe3 bxc6=
12...Qb6 13.0-0 Bg6 14.Bf3 Rc8 15.Qc4
Position after: 15.Qc4
Threatening Na4. Against other moves Black can simply play ...Be7 and castle.
15...Na7 16.Qb3 Qxb3 17.axb3 Rb8!=
Black should continue with ...Bb4(e7) and ...0-0.
3.3) 8.c4 Nbc6 9.Qa4 a6 10.Nc3
1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nf3 e6 5.Be2 c5 6.Be3 cxd4 7.Nxd4 Ne7 8.c4 Nbc6 9.Qa4 a6 10.Nc3 dxc4
297
11.0-0-0
White has to make tactical problems for his opponent within a fairly short time frame.
11.Rd1 leads to some forcing variations to which there is nothing much new to add: 11...Bd3 12.Bxd3 cxd3
13.Rxd3 b5! 14.Ndxb5 (14.Qb3 Nxe5 15.Rd1 Qc8 16.0-0 N7c6³) 14...Qxd3 15.Nc7+ Kd7 16.Nxa8
Position after: 16.Nxa8
16...Nd5 (16...Ng6!? could be one try to play for a win) 17.Nxd5 exd5 18.Qg4+ (18.Nb6+ Kc7 19.Nxd5+ is a
try but 19...Qxd5 20.0-0 Be7 21.Qxa6 Rb8= defends at least adequately) 18...Kd8 19.Bb6+ Ke7= Both kings
298
are so appallingly weak that the game will end in perpetual within a few moves; neither side should try to avoid
these possibilities altogether, and should be open to forcing the draw themselves if need be.
11...Qc8!?
Position after: 11...Qc8!?
This move is only given as a secondary option by both Dreev and Vidit, but I believe that on the level of
principle it is the best move, i.e. better than ...Qa5 which has been main thus far. This is just because the
position will become messy in any case and as Black I would prefer that my compensation for the bishop pair
take the form of pressure on the enemy king and pieces, rather than some kind of short-term strategic
consideration which can probably be defused by patient play by White against all but the most accurate
execution. That is, if I am to be forced to do memory work, I would like my opponent to have to do it too!
12.Nxf5 Nxf5 13.Bb6 Bb4!
13...Be7 is possibly under-rated, trying to just castle and then harrass White’s b6-bishop: 14.g4 Nh4 15.f4 0-0
16.Qxc4 Bd8² However White can keep an edge by simple enough means, e.g. taking on d8 and then Kb1, or
Bc5-d6.
14.f4
14.Rd8+ should of course be noticed, but Black’s pieces co-ordinate extremely well. 14...Qxd8 15.Bxd8 Rxd8
16.a3 (16.Bxc4 Rd4³) 16...Bxc3 (16...Ba5!? is the engine move) 17.bxc3 0-0 18.Qxc4 (18.f4 Ne3³) 18...Rd5
19.f4 Rc8 The game continues to be tremendously complicated, but Black has huge practical chances as well as
objective equality.
14...0-0 15.Bf3 Nfe7
Both sides have made their indispensable moves, with the exception of Black’s ...f6, which will follow next
299
against almost anything.
Position after: 15...Nfe7
16.g3
Given by Dreev.
16.Rhf1 is another move which consolidates White’s f4-pawn against the coming ...f6 thrust. Here I
recommend: 16...Rb8! 17.g3 In place of this you may insert any ‘nothing’ move e.g. Rd2. (17.Ne4?? Nd5–+;
17.Bxc6 bxc6! 18.Qxb4 Qb7=) 17...Ba5! 18.Bxa5 b5 19.Qa3 b4= Black will regain his piece, with continuing
pressure against the White queenside to compensate for having the worse minor piece.
16...f6
Black has to free his game ‘immediately if not sooner’ so it is fortunate that this break works.
17.exf6 Rxf6 18.Rhe1
18.Ne4 is the obvious move, trying to play against Black’s tactically vulnerable pieces. Unfortunately, this can
cut both ways. 18...Rf5 19.a3 Nd5! 20.Bf2 Ne5!
300
A) 21.fxe5 b5 22.Qc2 Rxf3 23.axb4 a5!ƒ If we take note that d3 is a good square for the f3-rook, we can see
that Black has a really good attack developing.
B) 21.Be2 Now with accurate play Black can force an equal ending: 21...Nd3+ 22.Bxd3 cxd3+ 23.Kb1 Qc4!
24.Nd2 Qc2+ 25.Qxc2 dxc2+ 26.Kxc2 Rc8+ 27.Kb3 Bxd2 28.Rxd2 Nf6=
C) 21.axb4 Nxf3 22.Nd6 Qf8 23.Qd7 Black has obvious and great compensation in case White takes on f5
immediately. 23...Rf6 24.Nxb7 Qf7!=
18...Kh8
301
Renewing the threat of ...e5. The position is not settled, so I recommend readers do a little further research, but
my analysis indicates Black is comfortable here. Some sample lines:
19.Be4
A) 19.Re2 Bxc3 20.bxc3 e5 21.Bxc6 (21.Qxc4 exf4 22.gxf4 Ng6„) 21...Qxc6 22.Qxc6 Nxc6 23.Rd7 exf4
24.gxf4 Ne5!=
B) 19.Qc2 e5 20.a3 Bxc3 21.Qxc3
Position after: 21.Qxc3
B1) 21...exf4!? 22.Bxc6 (22.Rxe7 Nxe7 23.Rd8+ Qxd8 24.Bxd8 Rxd8µ) 22...Nxc6 23.Rd8+ Qxd8 24.Bxd8
Rxd8 25.gxf4 b5² White is slightly better according to the machines, but the position is a hard one to
understand or to play for either side.
B2) 21...Ng6! 22.h4 exf4 23.h5 Nge5„
19...Qg8!?
Trying for ...Bxc3 and then ...Nd5 at no material cost.

19...Nd5 20.Bxd5 exd5 21.Nxd5 Bxe1 22.Nxf6 Qe6!= is also fine
20.Rd7 Nd5
302
21.Rxb7
A) 21.Nxd5 leads to incredible lines: 21...exd5 22.Bxd5 Qc8 23.Rc7 (23.Bxc6 Rxc6 24.Rd8+ Qxd8 25.Bxd8
Bxe1 26.Qd1 c3!? 27.Qd7 cxb2+ 28.Kxb2 Bc3+ 29.Kb3 h6=) 23...Qf5
Position after: 23...Qf5
24.Be4 (24.Rd1 Bd6 25.Rxb7 Nb4 26.Bd4 Nd3+ 27.Rxd3 cxd3 28.Bxf6 Rc8+ 29.Bc3 Qxd5= We know a
draw by perpetual check is coming soon, but it is not clear who is doing the scrambling for the draw!)
24...Re8!! 25.Bxf5 Rxe1+ 26.Kc2 Re2+=
303
B) 21.Bxd5 exd5 22.Rd1 Bxc3 23.bxc3 d4„
21...Rc8 22.Qxa6 Nxc3 23.bxc3 Bxc3„
Black has enough play and co-ordination to compensate for his nominally individually worse minor pieces and
queen.
Position after: 23... Bxc3„
3.4) Main line: 6.Be3 cxd4 7.Nxd4 Ne7 8.0-0 Nbc6 9.Bb5 a6 10.Bxc6+ bxc6 11. b4
1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nf3 e6 5.Be2 c5 6.Be3 cxd4 7.Nxd4 Ne7 8.0-0 Nbc6
304
Position after: 8...Nbc6
Now White has to do something about the e5-pawn, so his options are slightly limited.
9.Bb5
A) 9.c4 Nxd4 10.Bxd4 dxc4 11.Qa4+ (11.Bxc4 a6= will see the queens come off under quite auspicious
circumstances; White has to be careful not to end up worse.)
B) 9.f4 Now Black has quite a few options, but the best seems to me to be: 9...Nxd4 10.Bxd4 Nc6 11.c3
(11.Be3 Qa5 12.Nd2 Bc5³ is even more comfortable) 11...g5!?
Position after: 11...g5!?
There is also equality after normal moves like ...Be7 or ...Rc8, but this move tries for a bit more, i.e. the total
disintegration of the White centre.
B1) 12.Qb3 gxf4 13.Qxb7 Nxd4 14.cxd4 Bh6ƒ leads to an interesting game which takes more ideologically
from the Sveshnikov than the Caro-Kann.
B2) 12.Bb5 gxf4 13.Rxf4 a6 14.Bxc6+ bxc6„ is a dynamically balanced mess, but I would prefer to take
Black due to the greater ease of finding his ideas.
B3) 12.Nd2?! is way too placid. 12...gxf4 13.Rxf4 Nxd4 14.cxd4 Bh6 15.Rf1 0-0 16.Nf3 Qb6³
B4) 12.g4!? Bxb1 13.Rxb1 gxf4 14.Rxf4 Qg5
305
15.Qf1 (15.Qb3 is interesting, but after a while the queen will be the only attacking piece, so there is no real
chance to play for more than a draw. 15...Nxd4 16.Qxb7 Nxe2+ 17.Kh1 Qxf4 18.Qxa8+ Kd7=) 15...h5
16.Rxf7 Be7!„ Black develops good play on the kingside and against the e5-pawn, and will probably castle
long.
9...a6
9...Bg6 is the only other commonly-played move, but Vidit convincingly demonstrates a route to a White edge.
10.Bxc6+ bxc6 11.b4
This is a quieter approach than 11.c4, which we see next. I advocate just retreating the bishop and then playing
...a5, as follows.
11...Bg6
306
Now White can basically choose between two thematically consistent moves, if we bear in mind Black’s threat
of ...Qb8.
12.Nd2
Instead, 12.f4?! would give Black a slight edge after 12...a5 13.c3 Nf5 14.Nxf5 Bxf5. Another critical option
is:
12.c3 a5 13.a3 Qc7
A) 14.f4 lets us play a nice line where White’s moves are basically forced: 14...Nf5 15.Nxf5 Bxf5 16.Nd2 c5!
17.Nb3 c4 18.Nxa5 Ra6!ƒ White’s knight will be unable to get out of the trap without great concessions being
made. Meanwhile Black can get on with the rest of the game, notably the ...f6 break when he feels like it. If he
has a rook on the b-file, then maybe ...Rxa5 is also an idea.
B) 14.Qa4!? Qd8!
307
Challenging White to make use of his last move. Obviously, ...axb4 is now a fairly serious threat.
B1) 15.Nxc6 Qd7 16.b5 Bd3 17.Rd1 Bxb5 18.Qxb5 Nxc6=
B2) 15.Qb3 ends up just getting in the way: 15...Qc7 (15...Nf5!? 16.Nxc6 a4 17.Qd1 Nxe3 18.fxe3 Qb6
19.Nd4 Be7©) 16.f4 Nf5 17.Nxf5 Bxf5 18.Nd2 a4³ This is clearly a better version for Black of the 14.f4
line.
B3) 15.b5 Bd3 16.Rd1 cxb5 17.Nxb5 Bxb5 18.Qxb5+ Qd7= Next will come ...Rc8, either with or without
the queen trade on d7, and Black is fine.
12...Nf5
After 12...Qb8 I found surprising concrete problems in the line 13.a3 Qxe5 14.c4‚
13.Nxf5 Bxf5 14.c3 a5 15.a3 Qc7
308
16.Bd4
16.f4 c5 17.Nb3 c4 18.Nxa5 Ra6 was a previously mentioned point.
16...c5 17.bxc5 Bxc5 18.Bxc5 Qxc5 19.Qa4+ Kf8=
White has taken Black’s castling rights, which compensates for his worse minor piece, but he is unable to use
the b-file for attack, so Black will be able to get his pieces out. White’s best use of the time will be to somehow
liquidate his weakness on c3.
3.5) Main line with 11.c4 Qd7 12.cxd5
1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nf3 e6 5.Be2 c5 6.Be3 cxd4 7.Nxd4 Ne7 8.0-0 Nbc6 9.Bb5 a6 10.Bxc6+ bxc6
11.c4
309
We have arrived at possibly the most unforgiving line of the whole 3.e5 Bf5 complex. Black has to do a
significant amount of memory work here- there is no point lying to you.
11...Qd7 12.cxd5
Against White’s more minor choices, ...c5 tends to be good enough.

A) 12.Qa4 c5! 13.Qxd7+ Kxd7 14.Nb3 dxc4 15.Nxc5+ Kc6 16.Rc1 Nd5!„
B) 12.Qe2 c5! 13.Nxf5 Nxf5 14.cxd5 Qxd5 15.Nc3 Qc6=
C) 12.c5 Bg6 13.Nc3 Nf5 14.Na4 Ra7!=
310
Position after: 14...Ra7!=
In a closed position, Black’s development disadvantage is not of such great importance, and his solid structure
and bishop pair bode extremely well in the medium term.
D) 12.Na3 tries to get a better version of the text, but I think after 12...Rb8 13.cxd5 Nxd5 14.Nxf5 exf5
15.Nc4 Be7= both sides have minor pieces of reasonable quality and Black’s structural weaknesses are hard
to exploit.
12...Nxd5 13.Nxf5 exf5
Position after: 13...exf5
Now the position is quite interesting. In general, Black will acheive compensation for his dreadful structure in
the form of a better minor piece.
14.Bd4
A) 14.Qd4 Be7 15.Nc3 Nxe3 16.Qxe3 0-0=

B) 14.Nd2 Nxe3 15.fxe3 Bc5 16.Nc4 g6 17.Qxd7+ Kxd7 18.b4 (18.g4 Ke6 19.Rac1 Rad8 20.Kf2 Be7„)
18...Ba7 19.Rfd1+ Ke7 20.Kf2 c5=
14...Rb8 15.Nd2!?
A) 15.g3 is the engine recommendation, but after the logical re-routing of the knight to e6 it has nothing:
15...Nc7=.
B) 15.Qf3 Be7 16.e6!? The only challenging move, but it’s not enough. 16...Qxe6 17.Bxg7 Rg8 18.Bc3 Qe4=
C) 15.Nc3 Nc7
C1) 16.Rc1 Ne6 17.Ne2 c5 (17...Be7 18.b3 was a bit less clear-cut as an equalising method in Vachier
311
Lagrave, M – Navara, D Baku 2016) 18.Bc3 Be7=
C2) 16.Ne2 Ne6 17.Bc3 c5! 18.Ng3 (18.Qxd7+ Kxd7 19.Rfd1+ Kc6 20.Nc1 c4 21.b3 Bb4!=) 18...g6
19.Qf3!?
Position after: 19.Qf3!?
Keeping the queens on is really a disruption of the equilibrium; the position is pretty equal otherwise.
19...Bg7!N (19...Qb7 20.Qxb7 Rxb7 21.Rfd1 Be7 22.Nf1= might still have been equal, but White had the
c4-square and hence a considerable practical advantage, in Shirov, A – Vidit, S Douglas 2016) 20.Rfd1 Nd4
21.Qe3 Qe7!= White has to fight to stay equal.
15...Be7
312
16.e6!?
Other moves don’t promise much, e.g. 16.Nc4 0-0 17.Qf3 Qe6= It is, however, important to remember that
Black will be playing for ...f6 or even ...g5, and not for ...c5.
16...Qxe6 17.Bxg7
17.Re1 was seen in a high-profile test. 17...Qh6! (17...Qg6 18.Bc5 Rb7² was ultimately successful for Black in
Vachier Lagrave, M – Vidit, S Doha 2016, but he must have regretted the lack of control the queen had over the
d2-knight.) 18.Bc5 Rb7
313
Position after: 18...Rb7
19.Qa4 (For instance, Black can now castle immediately in the line 19.Bxe7 Rxe7 20.Rxe7+ Nxe7 21.Qe2 0-
0=) 19...0-0 20.Bxe7 Rxe7 21.Nf3 Rxe1+ 22.Rxe1 Qf6= White can regain his pawn- probably in a variety of
ways in which Ne5 is the simplest - but there will not be any advantage.
17...Rg8 18.Bc3
18...Rg4
Black’s pawns are all isolated now, but he has active play in all parts of the board, including the g-file.
19.Re1
A) 19.Qf3 Nxc3 20.bxc3 Qd5=

B) 19.g3 Nxc3 20.bxc3 Rd8 21.Re1 Re4!=
19...Nxc3 20.bxc3 Qd5 21.Nf3
21.f3 Rg6= and with ...Re6 coming next Black will be fine.
21...Qxd1 22.Raxd1 c5 23.Re5
23.Ne5 Re4=
One last awkward maneouvre is required.
23...Rf4!
314
Position after: 23...Rf4!
24.Rde1
24.Kf1 or any other move can be met by 24...Rb2 25.a3 f6= and Black gets his activity going by normal
methods.
24...Re4! 25.R5xe4 fxe4 26.Rxe4 Rb2 27.a4
27.g4 Rxa2 28.Ne5 Bf6!= doesn’t prove anything either as long as Black is unafraid of ghosts!
27...Rc2 28.g3
28.Re3 Rc1+ 29.Ne1 Kf8= regains the pawn under even better circumstances.
28...Rxc3 29.Ne5 Bd6=
There is no particular reason to fear the discovered checks. Black is fine in this ending and if White oversteps
the mark, the c-pawn will be quick to correct him.
3.6) Main line with 11.c4 Qd7 12.Nc3
1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nf3 e6 5.Be2 c5 6.Be3 cxd4 7.Nxd4 Ne7 8.0-0 Nbc6 9.Bb5 a6 10.Bxc6+ bxc6
11.c4 Qd7 12.Nc3!
315
Position after: 12.Nc3!
This move leads almost by force to an endgame which it will be hard for Black to win, although it is
complicated enough that White can easily overpress. I have tried to give as much detail as possible.
12...dxc4 13.Na4
13.Nxf5 leads to rapid simplifications, though a wrinkle or two helps make Black’s equalising task easier.
13...Qxd1
A) 14.Nxd1 Nxf5 15.Rc1 0-0-0 16.Rxc4 Kb7 17.Nc3 (17.g4 Nxe3 18.fxe3 Be7 19.Nc3 Rhf8! 20.Rcf4 f6
21.Ne4 h5!=) 17...Be7 18.Ne4 Rd5=
B) 14.Rfxd1 Nxf5 15.Bb6 Now the next few moves are forced. 15...Ne7! 16.Rd4 Nd5 17.Ba5 Bb4 18.Bxb4
Nxb4 19.Rxc4 Nd3 20.f4 Nxb2 21.Rxc6 0-0=
C) 14.Nxg7+ Bxg7
316
Position after: 14...Bxg7
C1) 15.Nxd1 wins a pawn, but the pawn will be in terrible health. 15...Bxe5 16.Rc1 Nd5 17.Rxc4 (17.Bd2
0-0 18.Rxc4 c5=) 17...Nxe3 18.fxe3 0-0 19.Rxc6 Rfc8!©
C2) 15.Rfxd1 Bxe5 16.Bd4 Bxd4 17.Rxd4 Rb8 18.Ne4 Nf5 19.Rxc4 Rxb2 20.Rxc6 0-0 21.Rxa6 Rc8© is the
same as Raxd1 except for the position of the a1-rook-both are fine for Black.
C3) 15.Raxd1 Bxe5 16.Bd4 Bxd4 17.Rxd4 Rb8 18.Ne4 (18.Rb1 Rb4 19.a3 Rb3 20.Rxc4 0-0=) 18...Nf5!
19.Rxc4 Rxb2 20.Rxc6 0-0 21.Rxa6 Rc8© Black has full compensation for the pawn.
13...Nd5 14.Nxf5 exf5
317
Position after: 14...exf5
The position now resembles the ones after 12.cxd5. Just like there, an important motif for Black to bear in mind
is the idea of giving the c-pawn back by advancing it.
15.Rc1
A) 15.Qe2 Be7 16.Rac1 0-0 17.Rxc4 Nxe3 18.Qxe3 Rad8 19.g3 Rfe8= Black is prepared to target the e5-
pawn, and meet f4 with ...f6.
B) 15.Bd2 Qe6! In this specific case, it makes sense to reserve the option of ...Bb4 rather than playing ...Be7
immediately. 16.Rc1 Rb8 17.Qc2 (17.Qe2 Bb4 18.Rfd1 c3! 19.Nxc3 0-0=) 17...Bb4 18.Bxb4 Nxb4 19.Qxc4
Qxc4 20.Rxc4 Nxa2 21.Rxc6 Nb4=
C) 15.Qc1 Be7 16.Qxc4 (16.Bc5 c3! breaks White’s coordination and allows Black to castle without
structural problems, e.g. 17.Bxe7 Qxe7 18.Nxc3 0-0=) 16...Nxe3 17.fxe3 Rd8!
Position after: 17...Rd8!
18.Rac1 (18.Rad1 Qxd1 19.Rxd1 Rxd1+ 20.Kf2 0-0=) 18...0-0! 19.Qxc6 Qxc6 20.Rxc6 g6 21.Rxa6 Rd2©
Black obtains full compensation for his sacrificed material.
D) 15.Bd4 Qe6!? The plan is to put the d5-knight on d3, and let the rest of the position worry about itself.
Vidit’s 15...c5 is equally sharp and equally playable, but there is an even greater diversity there of ways for
White to force a draw, and it is harder to play for Black if one has not memorised precisely. 16.Nc5 (16.Rc1
Nf4 17.Nb6 Rd8=) 16...Bxc5 17.Bxc5 Nf4 18.Qf3 (18.Bd6 f6 19.exf6 0-0-0= White is now obliged to play
Qa4 and give up the d6-bishop for perpetual check.) 18...Nd3 19.Bd6 f6!„
318
Position after: 19...f6!„
Both sides have dominant minor pieces in their opponent’s turf, but I prefer Black because it is easier to think
of things for his rooks to do even with the constraint of White’s d6-bishop. Not to mention, Black can always
play ...Rhe8, but White can’t move a rook to e1.
E) 15.Qc2 is the simplest to dispense with: 15...Nxe3 16.fxe3 Qd3=.
15...c3! 16.Nxc3 Nxe3
319
17.fxe3
17.Qxd7+ Kxd7 18.fxe3 g6 This is a bit of a ‘tabiya’. If Black got two tempi for, say, ...Ke7 and ...Bg7, then
he would already be better. Hence, White has to act now, and both sides have to be quite familiar with the
endgame subtleties that follow. 19.e4! (19.Rfd1+ Ke7 20.Ne2 Bg7 21.Nd4 Bxe5 22.Nxc6+ Kf6=) 19...Bc5+
20.Kh1 Rhe8 21.Rfd1+ Ke7
Position after: 21...Ke7
A) 22.exf5 ironically makes it less likely that Black will be a pawn down. 22...Be3 23.Rc2 gxf5 24.Rf1
(24.Re1 Bd4 25.Ne2 c5 26.Ng3 f4= Evtushenko, S – Schmidt, M ICCF 2015) 24...f4 25.Ne2 Kf8! 26.Nxf4
Rxe5 27.g3 Rd8=
B) 22.Ne2! Be3 23.Rc3 f4 (Vidit gives a long line commencing 23...Bh6 and proceeding through a number of
intricate subtleties in a position with rook and knight and four versus rook, bishop and three before reaching a
safe draw.) 24.g3 g5 25.gxf4 (25.Kg2 Red8 26.Rxd8 Rxd8 27.gxf4 Bxf4=; 25.Rxc6 Rac8© is a terrible idea
for White; Black gets active immediately.) 25...Bxf4 26.Nxf4 gxf4
320
27.Rd6 The only way to keep going. 27...Kf8 28.Kg2 (28.Rcxc6?! gives Black the chance to play some quite
aggressive chess deep into the ending, though it might still be fine: 28...Rxe5 29.Rxa6 Rae8 30.Rd4 Rg5!
31.Ra3 Rc8!ƒ) 28...Rxe5 29.Kf3 Rae8 30.Rdxc6 Rxe4 31.Rxa6 Re2=
17...g6
18.e4
321
A) 18.Qf3 Bh6 19.Na4! (19.Rcd1 Qc7 20.e4 Qxe5 21.exf5 Qe3+=) 19...0-0! 20.Nb6 Qd2 (20...Qa7 21.Rxc6
Rae8 22.Nc4 f4= was more convoluted but also ultimately equal in Caruana, F – Navara, D Baku 2016)
21.Nxa8 Qxe3+=
B) 18.Na4 Qc7 19.e6!? fxe6 20.Qd4 Bg7 21.Qc4 0-0 22.Qxe6+ Qf7© will give Black adequate
compensation.
18...Qxd1 19.Rfxd1
It makes sense to keep up the pressure on c6.

19.Rcxd1 f4 20.Na4 (20.Rxf4 Bg7=) 20...Rb8 21.b3 (21.Rxf4 Rb4 22.b3 Bg7 transposes) 21...Rb4 22.Rxf4 Bg7
23.e6 Bd4+= Having prevented Rd6, Black can go ahead and take on e6, and then cement his position with
...c5.
19...Bg7 20.exf5 Bxe5
Position after: 20...Bxe5
21.b3
A) 21.Ne4 Bxb2 22.Nd6+ (22.Rxc6 0-0=) 22...Ke7 23.Rxc6 Rhd8! 24.Rc7+ Kf8! 25.Rxf7+ Kg8 White is
tangled up and Black has the ...Be5 threat, so Black is fine. For instance 26.fxg6 (26.Rc7 gxf5=) 26...Be5
27.gxh7+ Kh8 28.g4 Bxd6 29.Kg2 Rab8„
B) 21.Na4 0-0! 22.fxg6 fxg6 (22...hxg6 23.Rxc6 Rfc8 24.Rdc1 Rxc6 25.Rxc6 Rd8 26.Rxa6² illustrates the
problem, though even this should be drawn with good play and the machine doesn’t see an issue (0.13))
23.Rxc6 Rf4 24.Nb6 Raf8„ Black has enough play for the pawn.
21...gxf5 22.Ne2 Ke7 23.Rc5
322
23.Rxc6 Rhc8! 24.Rdc1 Rd8!„ is an amusing ‘mirror’ of the rook maneouvre in the last line!
23...f6 24.Kf2 a5
Black is thinking of ...Ra6 and ...a4 ideas, so White should stop being pretentious and just take on c6.
25.Rxc6
25.Nd4 Bxd4+ 26.Rxd4 a4 27.bxa4 (27.Rxf5 axb3 28.axb3 Ra2+=) 27...Rhd8 28.Rxd8 Rxd8 29.Rxc6 Rd2+=
25...Rhc8 26.Rdc1 Rd8!=
Black’s better minor piece compensates fully for his broken kingside structure.
This chapter has been an interesting juxtaposition of sneaky attempts to win based on minor points in
obscure endgames, and lines which can be neutralised quickly by accurate yet essentially unique
continuations. Clearly, the former are far more dangerous, and the final subchapter contains a huge
number of them. One of the important things is to develop a sense for when Black has compensation for his
pawn in the ’12.Nc3 endgame’, and when not; the line can be fine and the call hard to make.
323
Chapter 4
The Arkell-Khenkin Variation
1.e4 c6 2.d4 d5 3.e5 c5 4.dxc5 e6
Chapter guide
Chapter 4 – The Arkell-Khenkin Variation
1.e4 c6 2.d4 d5 3.e5 c5 4.dxc5 e6 5.a3 Nc6 6.Nf3 Bxc5 7.b4 Bb6 8.Bb2 Nge7 9.Bd3 Ng6 10.0-0
4.1) 5.Be3 alternative
4.2) Main line with 10...Nf4?!
4.3) Main line with 10...0-0 11.Re1 f6 12.exf6 gxf6 13.c4 Nf4 14.cxd5
4.4) Main line with 10...0-0 11.Re1 f6 12.exf6 gxf6 13.c4 Nf4 14.Bf1
It would be a crime against humanity to write a Caro-Kann book without mentioning 3...c5, which (depending
who you believe) is either a forced slight edge for White, or a beautiful way to cut through the brutal memory
work of the previous few chapters and equalise without fuss. The truth, of course, is somewhere in between.
By far the most logical and critical move for White is 4.dxc5. There is no particular reason to dwell on 4.Nf3 or
4.c3, as these moves essentially lead to a ‘bad French’ for White. However, it does bear mentioning that 4.c4 is
playable for White, and that Black should have an answer in mind. As 3...Bf5 is already covered in depth (over
a few chapters) our repertoire is already complete, and thus I see no real reason to rehash the already adequate
answers of earlier authors. Instead I will add something new, which is a way to achieve what looks very much
like equality after 4.dxc5, beginning with 4...e6.
Against 5.Be3 I find it impossible to recommend quite a few of the main moves, so I have plotted out what I
consider to be the absolute best move-order, drawing on the contributions of a great number of practical experts
324
on this opening. Provided Black follows that, he achieves dynamic equality without a great deal of fuss –
certainly less than in the Short System.
After 5.a3 I find a lot of the available analysis rather confusing, and while generally excellent, the books by
both Houska and Dreev miss some important nuances. I shall first explain why Black needs to castle on move
10, and then proceed to chart narrow paths to endgame equality against both the important White tries on move
14.
Note that, as I try to be a conscientious and thoughtful author, I have a different definition of ‘endgame
equality’ than the computer: if White has +0.17 but after 10 completely natural moves by both sides it has
inexplicably crept up to +0.47, or if White is the only one with any possible winning attempt, then the endgame
is to be avoided, notwithstanding the precise number on the screen. Momentum is important, as is ease of
playing – as you will see from my rather philosophical remarks in the introduction...
4.1) 5.Be3 alternative
1.e4 c6 2.d4 d5 3.e5 c5 4.dxc5
Position after: 4.dxc5
To me, it does not make much sense to even consider moves other than 4.dxc5. The likes of 4.c3, 4.Nf3, etc.
will all end up giving Black a good French where his bishop is able to leave the nest before ...e6 is played.
4...e6 5.Be3
In this variation, transpositions are everywhere and both sides have to be wary of being ‘move-ordered’. As
such, we shall examine an unusual volume of Black options before ‘deciding’ on the main one.
A) 5.Qg4 h5 6.Qg3 h4 7.Qg4 Nh6 8.Bxh6 Rxh6 9.Nf3 is a slightly wacky yet well-trodden line. However,
Black should not worry. He has a vast number of moves here, of which I shall just mention two.
325
9...f5! (9...Bxc5!? 10.Qxg7 Bf8 11.Qg4 Qb6©) 10.Qd4 Bd7 11.Nbd2 Qc7 12.Nb3 b6 13.cxb6 axb6©
B) 5.Bd3 Nc6! 6.Nf3 Bxc5 is not problematic for Black unless White follows up with the a3, b4 type stuff
from lines 4.2 – 4.4).
5...Nd7
I am not recommending the most-played move 5...Nh6?! because after 6.Nf3 I find it impossible to find a good
answer for Black, even though ‘there must be one’.

326
A) 6...Nc6 7.c3 Nf5 8.Qd2! This is unstereotyped but rather good. (Against 8.Bd4 there is 8...Bd7= and I
believe Black’s position again.) 8...Nxe3 9.Qxe3² Now or in the next few moves, Black will have to play ...b6
and begin digging for ultimately specious compensation, or else be squished.
B) 6...Nd7 gambles quite heavily that having avoided 6.Bxh6, White will shrink from 7. Bxh6 also. 7.Bxh6!
There follows a nearly-forced line: 7...gxh6 8.c4 dxc4 9.Bxc4 Bg7 10.Nc3! Nxe5 11.Qxd8+ Kxd8 12.0-0-0+
Ke7 13.Nxe5 Bxe5 14.Kc2 Bd7 15.Rhe1² and at the end Black faces a rather depressing endgame.
C) 6...Nf5 This is recommended in one Black repertoire book, but personally I find the move to be of doubtful
value, as Black does not have a good answer to the obvious Bg5. 7.Bg5! Black would ideally like to arrange
to take back on c5 with the bishop and not the queen, and indeed ...Qa5 8.c3! is not favourable for him, but
neither is the attempted improvement 7...Qc7 which is probably best met by straightforward development.
(7...Be7 8.Bxe7 Nxe7 9.c4² does not inspire confidence either.) 8.Nc3 Bxc5 9.Bd3 Nc6 10.0-0!
There is really no need for White to go into the complications that arise if he tries to win material
immediately. 10...h6 The move pair ...h6/Bc1 could also have been inserted at a number of other points.
(10...Nfd4 11.Nxd4 Nxd4 12.Be3² is also rather unpleasant) 11.Bc1 Nfe7 (11...0-0 12.Bxf5 exf5 13.Nxd5²)
12.Nb5 Qb6 13.c3² White’s central blockade on the dark squares is beginning to look rather impressive.
6.Bb5
327
Position after: 6.Bb5
6...Ne7!
This is an interesting line, and was given a stamp of approval by Caro expert GM Igor Khenkin recently.
6...Qc7 is less good, for instance after 7.Nf3 Bxc5 8.Bxc5 Qxc5 9.Qe2‰, which was seen in Caruana, F –
Ushenina, A Wijk aan Zee 2008, or even the more maximalist 7.b4!?.
7.c3
328
Here Nf3 and Bg5 make much less sense than in our main line, because Black has ...Qa5+ winning a piece.
A) 7.Nf3 Nf5 8.Qd2?! Nxe3 9.Qxe3 Qa5+ 10.Nc3 a6 11.Bxd7+ Bxd7³
B) 7.Bd4 is a valid attempt at keeping the pawn alive. 7...a6
B1) 8.Bxd7+ Bxd7 will transpose to the main line, as ...Nf5 is coming next and c3 is indispensable at some
stage or other. One example of trying to avoid c3 shall suffice: 9.Nf3 Nf5 10.0-0 Nxd4 11.Qxd4 Rc8 12.b4
b6! 13.cxb6 Rc4 14.Qd3 Rxb4 15.Nbd2 Rxb6 16.c4 Be7!„ Black has completed his development and can
already think about being better.
B2) 8.Ba4 Nc6 9.Bxc6 bxc6 10.b4 a5
11.a3 (11.c3 Ba6 12.Nf3 g5µ looks even worse; White is not able to prevent ...g4 in the long term.) 11...Qg5
12.g3 Nxe5! 13.Nd2! (13.f4 Qg6 14.Nd2 Ng4ƒ) 13...Ng6 14.f4 Qf5 15.Ngf3 f6!= If Black had played
...Ba6, he would still have been struggling for equality against a White setup involving h4, Kf2 and Nf1-e3.
But this forces through ...e5 and equalises cleanly.
7...a6!?
A) 7...Nf5 8.Bd4 a6 seems less accurate due to the unusual 9.Bd3! Nxc5 10.Bxf5 exf5 11.Ne2² when I would
tentatively award White an advantage based on the central blockade.
B) 7...Nc6 may also equalise, but is a little unambitious. A sample line could be: 8.Nf3 Bxc5 9.Bxc5 Nxc5
10.0-0 0-0 11.Nbd2 Bd7= when there is still an appreciable chance that White’s central blockade will work
out, and Black is trying to play around it with (say) ...Ne7-f5, ...Qb6 etc. This is objectively acceptable but I
prefer delaying the regaining of the pawn.
8.Bxd7+
329
8.Ba4 Nc6 9.Nf3 Nxc5 is Black’s point.
Now there could follow 10.0-0 Nxa4 11.Qxa4 and now the most sophisticated move is 11...Qc7! which doesn’t
quite attack the e5-pawn, but ensures by pressuring it that White will be unable to set up with Qg4-g3 or Qg4
and Bf4. 12.Nbd2 Bd7 13.Qf4 f6! 14.Bd4 Be7 15.Rfe1 0-0 16.Qg3 a5„
This type of position will be well known to French players. Black will at some moment decide that the tension
does not favour him any more, and proceed with either ...f5 or (though unlikely) ...fxe5. However, until that
moment, White’s d2-knight is unemployed and he
Position after: 16...a5„

330
must think about Black’s interesting ideas such as the doubling of rooks behind the f6-pawn. (It was also
possible to execute this immediately with 16...Rf7.)
8...Bxd7
Position after: 8...Bxd7
9.Nf3
A) 9.Bd4 Nf5 will transpose, unless White tries to get clever with Ne2: 10.Ne2 Bb5 11.0-0 Rc8 12.b4 b6
13.cxb6 Nxd4 14.Nxd4 Bxf1 15.b7 Rb8 16.Nc6 Qg5 17.Kxf1 Rxb7
B) 9.Ne2 Nf5 10.Qd2 Qc7 11.Bd4 f6! 12.exf6 e5
331
13.f7+ (13.Be3 gxf6³ unifies Black’s pawns nicely) 13...Kxf7 14.Be3 Nxe3 15.fxe3 Qxc5 16.0-0+ Kg8= The
position is somewhat unbalanced, but Black should not worry.
9...Nf5 10.Bd4 Nxd4
11.Qxd4!?
This move is at least somewhat ambitious.

11.cxd4 b6 The main thing to remember here is that Black has compensation as long as he doesn’t allow an
exchange of queenside pawns. Exchanging queens is fine! 12.cxb6 Qxb6 13.Qb3 (13.Qd2 Rb8!µ) 13...Rb8
14.0-0 Be7 15.Nbd2 0-0 16.Qxb6 Rxb6©
332
Position after: 16...Rxb6©
The most likely scenario is that White will at some point give up a- and b-pawns for Black›s a-pawn, leaving a
drawn ending, but Black has some degree of choice in whether and when he accepts this transition. For
instance, he could choose to play ...Ba3 instead of ...a5-a4 against White’s b3, or ...Ra8 and ...a5-a4 in place of
...Ba4 against White’s Nb3.
11...Rc8 12.b4 b6 13.cxb6 Rc4 14.Qd1
Position after: 14.Qd1
14...Qxb6
333
14...Re4+ is playable as well, but signs up for the following awkward-looking line, which I would avoid if I had
the choice: 15.Kf1 Bb5+ 16.Kg1 Qxb6 17.Nbd2 Re2 18.Nd4 Rxe5 19.N2f3∞.
15.0-0 a5 16.a3 Be7©
Black›s compensation is obvious, and so are his moves: ...Rfc8, ...f6, possibly ...Be8-g6.
4.2) Main line with 10...Nf4?!
1.e4 c6 2.d4 d5 3.e5 c5 4.dxc5 e6 5.a3!?
An interesting concept and advocated by Alexey Dreev in his ambitiously named “Attacking the Caro-Kann”.
As we will see, the move is not without venom, but there are ways for Black to handle the line.
5...Nc6 6.Nf3 Bxc5 7.b4 Bb6
7...Be7!? deserves a lot more attention than it currently receives, and leads to a game with significantly
different character than 7...Bb6. 8.b5! (8.Bd3 f6 9.Bb2 Nh6 10.0-0 Nf7„ illustrates perfectly the point of
7...Be7: there will be no blockade on e5!) 8...Na5 9.Bd3 Bd7 10.0-0 Qc7!N
334
Position after: 10...Qc7!N
A) 11.Qe2 is given by Dreev but now 11...Nc4! gets sufficient counterplay, for instance 12.a4 (12.Bxc4 Qxc4
13.Qxc4 dxc4 14.Nc3 f5!∞) 12...a6 13.bxa6 Rxa6„
B) 11.Nbd2! A necessary improvement. 11...Nh6 (11...Nc4? 12.Bxc4 dxc4 13.Ne4±) 12.a4 a6 (12...f5?!
might be strategically desirable but there is a concrete problem after 13.exf6 Bxf6 14.Ba3!) 13.Ba3 (13.Qe2
0-0 14.Bb2 Rfc8„) 13...Bxa3 14.Rxa3 0-0 Reaching a confusing position. I suspect White is better, but it is
hard to prove. The wealth of ideas present (...Qc5 or ...Nb7-c5, ...Nf5, ...f6, etc) means it would not be hard as
Black to outplay a weaker opponent from here.
Returning to 7...Bb6:
335
8.Bb2
8.Bd3 is the other main move: 8...Nge7 9.0-0 Ng6 and now 10.Bb2 would transpose back into the text, so we
analyse the independent possibility of 10.Re1 while noting that 10.Qe2 would run into either ...Nd4 with a
forced draw (!) or ...a5 with good chances of a Black edge. 10.Re1 f6!
Position after: 10...f6!
A) 11.Bxg6+ hxg6 12.Bb2 is given as main by Houska but this is not worrying for a variety of reasons,
including the following: 12...fxe5 13.Nxe5 0-0 14.Nd3 (14.Nf3 e5! 15.Bxe5 Bg4³ is a devastating pin, not
least because after Nbd2 there would be ...Nxe5 and ...Bd4.) 14...e5 15.Nc3 e4 16.Nc5 Rf5³ Black has a
much nicer position and can attack with ideas like ...Qh4, ...Rh5, ...Bc7.
B) 11.exf6 Qxf6 12.Bxg6+ hxg6 13.c3! (13.Ra2 fails in its objective of delivering a dark-squared blockade
since after 13...0-0 14.Bb2 there is 14...e5!„) 13...0-0 14.Be3 g5! (14...Bc7 15.Nbd2 e5 16.Qb3 Be6 17.Ng5²
looks uncomfortably like White has consolidated) 15.Bxg5 (15.h3 Ne5=; 15.Bxb6 axb6 16.h3 Nxb4!= is
cute) 15...Bxf2+ 16.Kxf2 Qxg5 17.Kg1 Qf6=
336
Position after: 17...Qf6=
Black has undoubled his pawns and can proceed with ideas like ...Bd7-e8-g6(h5.) He need not worry too
much about the blockade on e5 as even if that happens, material will be greatly reduced and White’s c3-pawn
will be almost as weak as e6.
8...Nge7
A) 8...g5!?² Not as bad as it looks, but not recommendable either.

B) 8...f6 9.Bd3² leads to a White edge after a few precise moves. Black will be unable to challenge the
standard ‘Advance French’ blockade on the central dark squares.
9.Bd3 Ng6 10.0-0
337
10...Nf4?!
The secondary line, but here I agree with Dreev that White obtains a slight advantage with accurate play.
A) 10...a5 11.b5 Nce7 would be a good idea if Black could guarantee ...Nf4 next, but instead White plays
12.g3² and Black’s knights are left tangled.
B) 10...0-0 is seen next and is my main recommendation.
11.c4 0-0
11...Bc7 12.Qd2 Nxd3 13.Qxd3² Ly, M – Cheng, B Melbourne 2014.
12.Nc3
12.Qd2 wastes time and after 12...Nxd3 13.Qxd3 Ne7 14.Nc3 Bd7!= Black is fine- but he should be careful
not to take on c4 as this gives White the option of Ne4-d6 in future.
12...Bc7
A) 12...a5 13.b5 Ne7 14.Qd2 Nxd3 15.Qxd3²

B) 12...Ne7² Smirnov, P – Kharitonov, A Moscow 2007.
13.Re1 a6!
338
Black should try and be the immovable object rather than the unstoppable force. This move cuts out Nb5 and
discourages b5, as well as preparing Black’s own ...b5.
14.Qc2
A precise engine move.

A) 14.g3 dxc4! 15.Bxc4 Ng6= is awkward to face now that White’s c3-knight is unable to use b5 or a4.
B) 14.Rc1 dxc4 15.Bxc4 Qxd1 16.Rcxd1 b5!„ is another point – Black will develop with ...Bb7 and perhaps
...Ne7-d5.
14...Nxd3 15.Qxd3 Ne7 16.Rad1 Bd7²
339
Position after: 16...Bd7²
Black has not equalised, but he has the two bishops and there is always a good chance of the game opening up,
so some people may nevertheless find this playable. Notably, if White takes thrice on d5 now, there is ...Ba4,
with good compensation in the endgame.
4.3) Main line with 10...0-0 11.Re1 f6 12.exf6 gxf6 13.c4 Nf4 14.cxd5
1.e4 c6 2.d4 d5 3.e5 c5 4.dxc5! e6 5.a3 Nc6 6.Nf3 Bxc5 7.b4 Bb6 8.Bb2 Nge7 9.Bd3 Ng6 10.0-0 0-0!
Position after: 10...0-0!
340
A recommendation I share with fellow Caro author IM Jovanka Houska. (As the observant reader will
remember, when we met I decided to avoid a ‘real Caro’ derby and opted for 2.f4...)
11.Re1 f6!
Else c4 comes under less auspicious circumstances.
12.exf6 gxf6 13.c4™
Black wanted ...e5, which is now not really possible.
13...Nf4!
13...dxc4 14.Bxc4 Qxd1 15.Rxd1 is rather unconvincing for Black, especially after the improvement over
Houska’s analysis: 15...Nge5 16.Nxe5 fxe5 17.Rd2!± (17.Rf1?! e4„)
14.cxd5!?
Position after: 14.cxd5!?
According to my analysis, Black can neutralise the pressure here, but only with great precision.
14...Qxd5 15.Be4 Qxd1 16.Rxd1 e5 17.Nc3 Nd4 18.Na4!
The most challenging.

A) 18.Nd5 Nxf3+ 19.Bxf3 Nxd5 20.Bxd5+ Kg7= Black’s development has been slow but it will now be
completed: ...Rb8 and ...Bf5 come to mind and if White tries Be4 then there is ...Rf7.
B) 18.Nxd4 Bxd4 19.Rd2 Kg7 20.Ne2 Nxe2+ 21.Rxe2 Bxb2 22.Rxb2 f5!= is nothing much.
18...f5™
341
Position after: 18...f5™
19.Bd3!
19.Nxd4 fxe4 20.Nxb6 axb6 21.Nb5 Bg4!= Black gets enough play for his strategic deficiencies.
19...Nxd3
19...Nxf3+ 20.gxf3 Nxd3 21.Rxd3 Bc7 22.Rc1 Rf7 23.Rdc3² will net a pawn, albeit not a very healthy one.
20.Bxd4 Bxd4 21.Nxd4 Bd7!
21...Nf4 22.Nf3 Re8 23.Nc3ƒ Deep into the endgame, Black still faces a string of small issues with
development.
22.Rxd3 Bxa4 23.Rg3+ Kh8 24.Ne6
342
24...Rf6!
24...Rf7 may be inaccurate: 25.Re1 Re8 26.Nc5 Bc6 27.Nd3 Rfe7 28.Rge3 e4 (28...Be4 29.f3 Bxd3 30.Rxd3²)
29.Nf4 with continuing positional pressure.
25.Nc5 Bc6 26.Nd3 f4!
26...Re8 27.Re3 Be4 28.f3 Bxd3 29.Rxd3² once more gives White a slightly more pleasant double-rook ending.
27.Rg5 e4 28.Ne5
343
28...Rg8!
The final accurate move to take play into a drawn rook ending.
28...Re8 29.g3 fxg3 30.hxg3² could still become unpleasant
29.Rxg8+ Kxg8 30.Re1 Re6 31.Nxc6 Rxc6!=
There is not a whole lot more to try.
4.4) Main line with 10...0-0 11.Re1 f6 12.exf6 gxf6 13.c4 Nf4 14.Bf1
1.e4 c6 2.d4 d5 3.e5 c5 4.dxc5! e6 5.a3 Nc6 6.Nf3 Bxc5 7.b4 Bb6 8.Bb2 Nge7 9.Bd3 Ng6 10.0-0 0-0!
11.Re1 f6! 12.exf6 gxf6 13.c4™ 13...Nf4! 14.Bf1!?
Position after: 14.Bf1!?
This is Dreev’s recommendation. I do not consider it to get an edge, but there are some tricks and wrinkles
which are worth knowing. Black needs to make the right choice now.
14...e5!
14...dxc4?! To be avoided. 15.Bxc4 (15.Qxd8?!= did not promise anything in Hansen, E – Houska, J Gibraltar
2015.)
A) 15...Qc7N has the idea of ...Kh8, ...e5 and ...Qg7, developing a kingside attack, but the engine points out
16.Re4!ƒ when White is the one dictating events, since Black is prevented from going ...Qg7 either with or
without ...f5.
B) 15...Qxd1 16.Rxd1 Kg7 17.Bc1!?² Black is forced into the pawn sacrifice ...Nd5, which might be okay for
344
some practical games but which I suspect does not give full compensation.
15.c5!?
A) 15.g3 wins a pawn, but Black’s game is much easier to play. 15...Nh3+ 16.Bxh3 Bxh3
Position after: 16...Bxh3
17.cxd5 (17.Qxd5+ Qxd5 18.cxd5 Ne7 19.Nc3 Rad8 20.Rad1 Rd7= Black has great compensation and
probably the easier game.) 17...Nd4! 18.Bxd4 exd4 19.Qb3 Qd6 Apparently regaining the pawn, but White
can arrange his pieces so as to take on d4 when Black takes on d5. 20.Nbd2 Rad8 21.Re4 Qxd5 22.Nxd4 f5©
Nevertheless, Black gets great compensation.
B) The engine chooses 15.Bc1 dxc4 16.Bxf4 exf4 17.Bxc4+ Kh8 18.Nc3
345
The question is whether the bishop pair constitutes enough compensation for the doubled f-pawns. My answer
is definitely ‘yes’ and I would begin with 18...Bg4!? which not-very-subtly hints that White may have his own
doubled f-pawns soon. The critical continuation now is 19.Nd5! Nd4!? This looks like great fun. (19...Bd4 is
the ‘correct’ move and after 20.Ra2 b5 21.Bb3 Be5= Black has no problems.) 20.Qd3 Bxf3 21.Re7!?
Maximalism. (21.gxf3 Qd7 22.Nxf4 Qg7+ 23.Kh1 Rad8©) 21...Qxe7 22.Nxe7 Be2 23.Qc3 f3∞
Position after: 23...f3∞
With a really confusing position. I suspect White has to play 24.g3 and then after ...Rac8 we obtain that
dreaded material balance, three pieces for a queen. White might be marginally better but I would still take
346
Black. For the more boringly minded among you, there is always 19...Bd4.
15...Bc7 16.Nc3
Questioning Black’s arrangements for the defense of d5.
16...Be6 17.g3
17.Nb5 a6 forces the capture on c7, which is not in White’s interests.
17...Ng6 18.Nb5!
Clearly the most critical.
18...a6 19.Nfd4
19.Nxc7 Qxc7 is nothing; Black can follow up with say ...Qf7 and ...d4 and the bishop pair is useless.
19...Nxd4 20.Nxd4 Bd7
347
21.Nf3
Black has reacted correctly to the challenges and White should probably accept an equal position.
21.Bg2!? is a move that normally would not be considered, because it looks like a position where a piece easily
outweighs three pawns. However it is only after 21...exd4 22.Bxd5+ Kg7 23.Bxb7 Ra7 24.Bg2 Bf5 25.Bxd4
Be5!„ that Black stands any chance of this happening.
21...Bc6=
The position is objectively equal and will provide an interesting proving ground for the relative merits of the
classical and hypermodern strategies...
Based on what we have seen in the last two chapters, both 3...Bf5 and 3...c5 can lead (after good White
play) to Black having to master subtle nuances in heavily-analysed middlegames.
What I would say by way of conclusion is that after 3...c5, the nuances seem to end around move 25, and
leave Black with more winning chances, whereas after 3...Bf5, they can continue deep into the 30’s.
As such, I would say that past a certain level of preparation, 3...c5 once again becomes the way to go,
though of course both are playable.
348
Chapter 5
Annotated games for thematic structures in the advance Caro
Chapter guide
Chapter 5 – Annotated games for thematic structures in the advance Caro
5.1) Petrov-Fernandez, Paleochora 2012
5.2) Svidler-Adhiban, Tata Steel Masters 2018
5.3) Kollars-Can, Rome 2017
In this Part we have seen a great number of different structures, most of them rather static. It occurred to me
that the best way of dealing with these was to find good games illustrating Black’s (or preferably both players’)
plans, rather than worrying about specific move-orders. As such, here are three of the games that I feel
encapsulate important information about the structures in this Part.
The first one is one of my own, played when I was 17 years old. It features, in simple terms, a struggle between
a bishop pair and a tricky pair of knights with central outposts. Black plays against the d4-pawn in slow yet
successful fashion; the extent of his mazy manoeuvres being seen in the diagram above.
The second game builds on the same theme, and it is one I commentated on one evening while at the British
chess league (4NCL). There is an additional complication in that Black has not one but two central outposts, but
a chronically unsafe king. The opening is a rather ambitious construction with ...h6 and ...g5 against the Short
system, and it’s probably not enough for equality, but this being a practical game, the outcome of the opening
battle swings in the balance a few times before coming down firmly on Black’s side, and the nuances are great
fun to explore.
Finally, I have chosen the game Kollars-Can to illustrate the structure with doubled White d-pawns that can
arise in the line 4.h4 h5 5.Bd3. Notwithstanding that he got his move-order a bit wrong, Turkish GM Emre Can
illustrated a really interesting idea in this game, wherein he brought his dark-squared bishop to f6 without any
delay at all and then immediately played for ...c5. Another approach which you could take to this line is
exemplified by Houska in a very recent practical game.
349
5.1) Petrov-Fernandez, Paleochora 2012
1.e4 c6 2.d4 d5 3.e5 Bf5 4.c4?!
Position after: 4.c4?!
This is too quick; White will end up having to cede the d5-square and may not get anything for it in terms of
development or the bishop pair.
4...e6 5.Nc3 h6 6.h3
Another sign that we are departing from the realm of the theoretically important.
6...Ne7 7.Nf3 Be4!?
Trying to take advantage of White’s slow play and get in ...Nf5.

7...Nd7 was obviously fine, as was any typical move for the structure such as ...a6 or ...Ng6.
8.cxd5 Bxd5!?
But this is genuinely interesting.

A) 8...Bxf3 9.Qxf3 cxd5 would have been the most consistent follow-up to ...Be4, and would bring about a
position reminiscent of the Slow Slav (1.d4 d5 2.c4 c6 3.Nf3 Nf6 4.e3 Bg4 5.cxd5 Bxf3 etc.).
B) 8...cxd5!?=
9.Nxd5 Nxd5
350
10.a3
A good ‘culture’ move, preventing ...Bb4+ which would make a trade of bishops likely.
10...Nd7
At this point, I envisaged the entire regrouping that was about to take place, so even though the computer might
insist White is slightly better, my position plays itself. Quite often it is this that we want from our positions.
Furthermore, in this case, the evaluation drops whenever one plays down a line, so maybe the bishop pair
doesn’t even represent an ‘objective’ advantage here.
11.Bc4 Be7 12.0-0 N7b6 13.Bd3 Qd7 14.Bd2
351
14...a6?!
Not a clever order as these things go. If one has a space advantage, then one can usually afford spare moves
(g3, Kg2) while maneouvring, and as if by magic these moves will often turn out to have been necessary. But
with a space disadvantage one should prioritise that which is indispensable. 14...Nc8 would have served the
same purpose while possibly saving a tempo. 15.Ba5 Bd8 16.Bxd8 Qxd8 17.g3 Qb6„
15.Rc1
15.Ba5! Not an immediately obvious move but one which greatly hinders Black’s plans. 15...Nc8 (15...Bd8
16.g3²) 16.g3 Bd8 There isn’t much else to do. 17.Bxd8 Qxd8 18.Nd2² Now White’s play is a lot more fluid
than before.
15...Nc8 16.Qe2
16.Ne1 Bd8 17.f4 Bb6 18.Nf3 Nce7 White’s pawn storm does not arrive in time to interfere with Black’s
regrouping.
16...Bd8 17.b4 Nce7 18.Qe4
352
Position after: 18.Qe4
18...Bb6
Complete! Contrast this with the last diagram.
19.Rc2
19.a4 is suggested by the machine, but I don’t seriously believe White wants to play b5 in most cases. 19...Kf8
Playing to connect rooks. 20.b5 axb5 21.axb5 g6 22.bxc6 bxc6= and White has not actually gained any targets
from opening
the queenside.
19...Nf5
19...Rd8= was more conservative, but I suspected that in response to my move g4 would be played and that it
would be a mistake!
20.g4?!
20.a4 Rd8 was my intention, and now White really should do something about potential discovered attacks on
the d4-pawn. 21.b5 axb5 22.axb5 Nde7 23.bxc6 bxc6„ However, it transpires that White can probably
sacrifice this pawn and get compensation, exactly as in the game Fernandez – Arkell.
20...Nfe7
Now it would not surprise me if Black were actually better.
21.Kg2
353
Position after: 21.Kg2
21...Nc7
Maybe not strictly necessary, but on thematic grounds I wanted to attack d4 right now rather than try and attack
on the kingside. The move did also help set up the ‘ghost’ of ...Qd5, and also discourage White’s b5-push.
A) 21...g5?! 22.h4 looks premature, although Black can maybe initiate tactical mayhem with ...h5!? and then
long castling.
B) 21...h5!? would break with the pattern of the rest of the game, and so this is hard to play, but maybe it was
best. For instance 22.a4 hxg4 23.hxg4 f5‚.
22.a4 Rd8 23.Rc4 Ba7
Looking for ...b5.
24.Rd1 Ng6
354
Now looking for ...Qd5 and piece trades that would make White’s life quite difficult.
25.Qe2?!
25.Rcc1 is all the computer can come up with. 25...Qd5 26.Qxd5 Rxd5 27.Bc4 Rd7= and even with the position
‘objectively level’, this would be a nightmare to play with White.
25...Nd5³
Setting up ...Nb6 ideas.
26.Bxg6?! fxg6 27.Qe4 Ne7
27...g5!? was thematic.
28.Be3 0-0 29.Nd2 g5
355
Suddenly Black is everywhere, but just as suddenly, I developed a fear of my Grandmaster opponent (I was a
low rated IM at the time) and the position rapidly simplified to a draw... How often the stronger player is able to
save points and dignity just due to their opponent’s fear!
30.Nb3 Qd5?
30...Nd5 followed by doubling on the f-file was extremely strong, and looking back close to 6 years later it is
hard to understand not playing on with something like this.
31.Qxd5 Nxd5 32.Nc5 Bxc5 33.bxc5=
356
Position after: 33.bxc5=
Black is still marginally for choice, but as alluded to before, I was fine with a draw...
½-½
5.2) Svidler-Adhiban, Tata Steel Masters 2018
1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nf3 e6 5.Be2 Ne7 6.0-0 h6
As the reader may recall, my first recommendation against the Short System went 6...c5 here and may be seen
in Chapter 2.
7.Nbd2 Bh7 8.Nb3 Nd7 9.a4
9.Bd2!? After detailed examination of the wrinkles in the position, I think this move is of precisely equal value
to the one played in the game. 9...Nf5 10.Rc1
357
Position after: 10.Rc1
10...a5 (10...g5 11.h3 Be7 12.c4² and White obtains essentially what he wants.) 11.a4 (11.c4 is a little bit too
early, as Black so values a knight on d5 that he can undevelop in order to get it. 11...dxc4 12.Rxc4 Ne7=)
11...Be7 Black lacks other moves he can sensibly play. 12.c4 dxc4 13.Bxc4 0-0 14.Qe2 Bb4!?
Position after: 14...Bb4!?
This added nuance resulting from inserting the moves ...a5 and a4 is the only reason Black has any chances of
equality at all. 15.Bxb4 axb4 16.a5ƒ and White’s positional pressure continues, but it is not so bad if Black
keeps calm.
358
9...g5
An interesting novelty and the reason the game was picked for this chapter, but perhaps this signs away the
advantage to White?
A) 9...Nf5 10.a5 Be7 11.c4!² Azarov, S – Kunal, M Abu Dhabi 2016.
B) 9...a5!
It is important to secure the b4-square for Black’s bishop. This is not a wasted tempo, since if White got in a5,
Black would probably need to respond ...a6 sooner or later anyway. 10.Re1 Nf5!?N (10...Nc8 11.Be3 Be7
was playable if a little passive in Rubino, W – Okhotnik, V Avoine 2014) 11.Bd2 Be7 12.c4 0-0= Black
reserves the option of ...Bb4 until White has wasted a further tempo with his queen’s rook or d2-bishop.
10.a5
359
Position after: 10.a5
Now White has a slight advantage.
10...a6
10...Nf5!? Maybe this is the best order, cutting down White’s options.
A) 11.c4 g4 12.Nfd2 h5 13.a6 b6 14.cxd5 cxd5 15.Nb1
Position after: 15.Nb1
leads to a position where in January 2018 I remarked that “Black’s king lacks shelter on either the kingside or
360
queenside, and its home in the centre will soon come under attack.” Indeed, further engine analysis shows that
Black is quite likely objectively scraping by following 15...f6! 16.exf6 Nxf6 17.Bg5 Bd6 18.Nc3 0-0! but his
position is still harder to play.
B) 11.Bd2 g4 12.Ne1 h5 13.Nd3 (13.c4 dxc4 14.Bxc4 b5!„) 13...f6 14.Bf4 Qe7² is also objectively better for
White but Black has clear avenues of counterplay.
C) 11.h3! The point is that without ...g4, Black’s play may well be going nowhere fast, especially since ...c5 is
guaranteed to end badly. 11...a6 12.Bd2 White keeps open the option of g4 and also may be thinking vaguely
about the strategically desirable Ne1-d3 or Qe1/Bb4 plans. 12...Be7 13.Rc1 h5 14.Ne1 Bg6 15.Nd3 Ng7
This position is quite remarkable and a bit of a throwback to the first annotated game of this set.
Black’s plans are all extremely grandiose in nature (think bringing the g7-knight to b5) but there is very little
White can do right here. Nevertheless, with slow plans like Qe1 and Bb4 White will keep a small plus.
11.c3
This is almost definitely inaccurate.

11.Bd2!? White obtains what I consider to be a more flexible version of the play in the last note. 11...Nf5
(11...c5?! 12.Nxc5 Nxc5 13.dxc5 Nc6 14.b4²) 12.c4 (12.h3 is the last note, but if White can do without it then
he should.) 12...Be7 13.Rc1 g4?! 14.Ne1 h5 15.cxd5 cxd5 16.g3±
11...Nf5
361
12.Bd3
A) 12.c4 g4 13.Ne1 dxc4 14.Nd2 b5=

B) 12.g4 Ne7 13.Be3 f6! 14.exf6 Ng6∞
12...g4 13.Ne1 h5 14.Nc2 c5
Another slight inaccuracy, perhaps.

14...f6!?N 15.f4 Ne7! avoids giving White such easy play: Black can combine ideas of ...c5 with his standard
kingside expansion, and the resulting play is very messy indeed. For instance: 16.Bxh7 Rxh7 17.Ne3 f5„.
15.Nxc5 Nxc5?!
This one really gives White a significant advantage.

15...Bxc5! 16.dxc5 Nxc5 17.Nb4 Rc8 may have been easier, and was probably equal.
16.dxc5 Bxc5
362
17.Nd4!
17.Bf4 Nh4!= Black exchanges the bishops without getting tied down by a possible Bg5, and brings his knight
safely to g6.
17...Bxd4?!
17...Nh4!? 18.Be3 Rc8 19.Bxh7 Rxh7 20.Qd2 Rg7² leads to a strategically confusing position where White is
better but has no meaningful pawn breaks.
18.cxd4 Rc8 19.Ra3?!
White wants to lift this rook to the kingside, but this is ultimately too fanciful.
19.Bf4! Nh4 (19...Ne7 20.Bxh7 Rxh7 21.Bg5±) 20.Bxh7 Rxh7 21.Rc1± White has a really significant
advantage.
19...Rc7 20.Bf4 Nh4 21.Qd2
21.Bxh7 Rxh7 22.Qd2 was another idea, trying to do the swap from the previous note with the other rook
instead.
21...Bxd3 22.Bg5 Qc8 23.Rxd3 Nf5
363
Now that the opening phase is well and truly over, I shall just comment that the two dominating minor pieces
merely balance each other out and that Black’s king’s defences are totally watertight. Peter Svidler tried for a
few more moves but got nowhere.
24.Qb4 Kd7 25.Qa4+ Rc6 26.Rc1 Rg8 27.Bf6 Qc7 28.g3 Rc8 29.Rc5 Ke8 30.b4 Qd7 31.Qd1 Rxc5 32.bxc5
Qb5 33.h3
½-½
5.3) Kollars-Can, Rome 2017
1.e4 c6 2.d4 d5 3.e5 Bf5 4.h4 h5 5.Bd3 Bxd3 6.Qxd3 Qa5+ 7.Bd2 Qa6 8.e6 Qxd3 9.exf7+ Kxf7 10.cxd3
364
Position after: 10.cxd3
In principle this is not a position Black should worry too much about: the doubled d-pawns provide him with
quite sufficient compensation for the weaknesses of the g5-square and e7-pawn. However, chess is quite an
ironic game, and one way or another he quite often finds himself having to play ...c5 in order to develop his
counterplay fully.
10...Nd7
10...Nh6! could be a better move-order, delaying the decision about the f8-bishop for a future date. 11.Nf3 Nf5
12.Ne5+ (12.Be3 Nd7 13.Nbd2 e6 14.Bf4 Be7³ left Black in full control in Haug, J – Houska, J Kragero 2018)
12...Ke8 13.Ng6
365
This move can best be understood as buying, for the price of a presently unfavourable-looking minor piece
exchange, a few tempi and the assurance that when Black does free his position, it will not be with any
particular force. 13...Rg8 14.Be3 Nd7 15.Nd2 Kf7 16.Nxf8 Rgxf8 17.Nf3 Rae8 18.Kd2 e5! 19.dxe5 Nxe3
20.fxe3 Nxe5=
11.Nf3 g6?!
A maximalist move, trying to bring the bishop to f6 as a matter of some urgency and not worrying too much
what White does in the meantime.
11...Nh6 is a respectable alternative, and could transpose. As in the next note, White’s best plan of development
could be 12.Bc3 e6 (12...Nf5 13.Nbd2² leads to a bind) 13.Nbd2 Be7 14.Ng5+ Bxg5 15.hxg5
Position after: 15.hxg5
however Black should not get too worried as even if he cannot play ...e5, it is nearly impossible to target his e6-
pawn without use of the g5-square. 15...Nf5„
12.Nc3
12.Ng5+! must obviously be the first move a human considers. I suspect after 12...Ke8 White considered only
13.Ne6 and didn’t see anything profitable to do after ...Kf7. However, here 13.Bc3!
366
Position after: 13.Bc3!
is surprisingly strong, bringing out the second knight to d2 and then f3, and trying to enforce some clamp on the
e5- and e6-squares. 13...Nh6 14.Nd2 Bg7 15.Ndf3 (15.f4!?) 15...Bf6 16.0-0² Black lacks any sort of play and
White is about to double on the e-file. It is only not even worse because we are also left with lingering
questions about what White can do after he has doubled.
12...Nh6
Back on track. Now Ng5+ accomplishes little, but it may still be the best way to play.
367
13.0-0-0
Not a bad move as such, but every move that White leaves the centre the way it is, he risks sliding back into a
worse position, unless he can acheive some kind of bind against the Black e-pawn, which is not the case right
here.
13.Ng5+ Kg8 14.Ne2 Nf5 15.Nf4 White forces a liquidation and runs while the going is good. 15...Bg7!
16.Nxg6 Rh6 17.Ne5 Nxe5 18.dxe5 Bxe5 19.Bc3=
13...Nf5 14.Rde1 Bg7 15.Ne2 Bf6
Reorganisation complete.
16.Bc3
16.Ng5+ Kg8 17.Nf3 might be called for, trying to prevent Black from developing his h8-rook. Note also the
tactic 17...Kg7 18.Nf4!„.
16...Rhc8 17.Nf4 c5
18.dxc5?
A) 18.Kb1!? cxd4 19.Bd2 is a machine suggestion, trying to belatedly get some kind of Ne6 and Nfg5+ plan
going, but Black can nip it in the bud and keep a very comfortable kind of equality: 19...Rc6 20.Nxd5 Nc5=.
B) 18.Ne5+ creates a surprising amount of mess for such a closed position: 18...Bxe5 (18...Ke8 19.Nxd7
Kxd7 20.Kb1 cxd4 21.Bb4 Nxh4 22.Nxd5‚) 19.dxe5 e6 20.Nxd5 exd5 21.e6+ Ke8 22.exd7+ Kxd7 23.g4
hxg4 24.h5
368
24...d4 25.Bd2 Rh8 26.h6 (26.hxg6 Rxh1 27.Rxh1 Ne7 and Black could have a squeeze) 26...Rh7 27.Reg1 g3
28.fxg3 Rah8 29.Rh2 Ne3 30.Rgh1 Ke6= Both sides are so tied up because of the h6-pawn that it is the
incursion of the Black king which forces matters. White will have to take on e3, give up h6 for e3 and agree to
a draw in the rook endgame.
18...d4 19.Bb4 a5!
Unless you saw this and the next Black move together, it would have been difficult to appreciate their impact
369
ahead of time.
20.Ba3 b5! 21.b3 b4 22.Bb2 e5 23.Nh3 Nxc5 24.Kd2 a4
Suddenly White is so uncoordinated that he loses by force. While he tried to wriggle for a further 30 moves, the
outcome was never in serious doubt.
0-1 (56)
It is not often you see a collection of annotated games meshed in with the rest of the text in this way. I hope
the experiment has been successful and that you appreciate the idea (even if not the necessity) of hashing
out the static elements of a structure in a less formal way than a variation tree.
The main theoretical value here is in the third game, where Black needed to make a decision between moves
10-15 about the optimal way to develop his pieces. Two possibilities are outlined, of which the one by
Houska was ‘better’ while the one by Can is ‘purer’.
370
Part III:
4...Nd7 Sidelines
Chapter 1
The 5.Bc4 system
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Bc4
Chapter guide
Chapter 1 – The 5.Bc4 System
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Bc4 Ngf6 6.Ng5 e6 7.Qe2 Nb6 8.Bd3 h6
1.1) Minor lines and 9.Ne4!? Qxd4 10.Nf3 Qd8 11.0-0
1.2) 9.Ne4!? Qxd4 10.Nf3 Qd8 11.Bf4
1.3) 9.N5f3 c5: 10.Be3 and others
1.4) 9.N5f3 c5 10.dxc5 Nbd7: 11.Bf4 and others
1.5) 9.N5f3 c5 10.dxc5 Nbd7 11.b4 a5!?
1.6) 8.Bb3 alternative
5.Bc4 is one of those lines which was popular around the turn of the century, and perhaps a little before as well.
For instance, one book I read about Karpov (I believe it was the one by Matsukevich) went into considerable
detail about possible knight manoeuvres for Black and the precise method by which the 12th world champion
used to equalise against 5.Bc4, both of which were vaguely reminiscent of a later line in this book, i.e. III.4.2.
Notwithstanding the historical precedents, currently the line is unfashionable and it is noteworthy that the ‘less
371
sound’ gambit line right at the start – lines 1.1) and 1.2) – is the most dangerous in the entire chapter. Black
needs to exercise care, particularly with 1.2).
When I was learning about how to play chess seriously, at the end of the ‘00s, one of my fantasies was to catch
people in the ‘second Caro mate trick’, the first being a certain smothered mate on move 6. As it turned out,
nobody ever fell into the trap (if you were wondering, it is mentioned in 1.4)) and when on one occasion my
school-mate attempted to snare a grandmaster with it, the attempt ended with the loss of the bishop-pair for
nothing, followed by tears and recriminations...
It is, however, worth paying attention to how Black avoids the more sophisticated trickery of line 1.5). I have
given a route which is not necessarily best, but which is clean and easy to play even with only partial recall of
this text.
1.1) Minor lines and 9.Ne4!? Qxd4 10.Nf3 Qd8 11.0-0
One of the most natural developing moves at White’s disposal, and Black is well advised to take this seriously.
5...Ngf6 6.Ng5
6.Nxf6+ Nxf6 leaves White with no real independent options. For the most logical 7.c3, see III.3.4.
6...e6 7.Qe2
A) 7.Ne2 is maybe a bit too creative. 7...h6!
372
Black is able to avoid the slightly awkward f4, Nf3-e5 setups in this way. 8.Nf3 (8.Nxe6? fxe6 9.Nf4 Bd6! It
is most precise to encourage White to take back some material. The compensation will then fizzle out by itself
after 10.Nxe6 Qe7 11.0-0 Nf8³) 8...b5! (8...Bd6 9.Bf4 Bxf4 10.Nxf4 The computer proclaims this equal, but
its suggestion allows all kinds of sacrificial ideas, e.g. 10...Qc7 11.Qd2 b5 12.Bxe6‚ ) 9.Bd3 Bd6 10.Bf4
Bxf4 11.Nxf4 Qc7 12.Qd2 a6=
Position after: 12...a6=
With ...c5 to come next unless White plays the positionally disastrous b4; as in analogous positions from the
Catalan and other openings, getting ...c5 in signals that Black has equalised.
373
B) 7.f4?! immediately is no good: 7...b5 (7...c5? 8.f5!±) 8.Bd3 c5 9.Qf3 (9.f5 exf5 10.Qf3 Nb6³) 9...Rb8
10.f5 e5³.
7...Nb6
The theoretical main line. There are not a lot of other ways to avoid Nxf7 sacrifices.
8.Bd3
8.Bb3 is covered later in this chapter.
8...h6
Following 8...c5?! 9.dxc5 Bxc5 10.N1f3 0-0 11.0-0² Black could be fine here, but both of White’s knights play
sensible roles, which is not something that can be said after 8...h6.
9.Ne4!?
This is an interesting gambit line, which it took me a while to neutralise. More common is 9.N5f3.
9...Qxd4!
I’m not often going to recommend taking pawns that give White intangible compensation that is hard to pin
down, but here it is best.
9...Nxe4 10.Qxe4 Qd5 11.Nf3
374
11...c5 This is another valid try to force equality, and may appeal to solid players, but from my perspective
there seems to be a concrete issue regarding the c8 bishop and/or b6 knight. (11...Qxe4+ 12.Bxe4 c5 13.0-0
leaves Black with a problematic light squared bishop.) 12.c3! c4 (12...Bd6 13.dxc5 Qxe4+ 14.Bxe4 Bxc5 15.0-
0² leads to a position where Black still has significant problems completing his development.; 12...cxd4
13.Nxd4 Bc5 14.Qg4 Kf8 15.Qe2ƒ is also quite unpleasant) 13.Qxd5! Since the centre will be closed anyway,
White does it in a manner that ensures Black cannot use the d5-square for his knight. 13...exd5 14.Bc2
White has ensured that if dark squared bishops are traded he stands better for all time. Perhaps a small thing,
375
but enough to put me off this way of playing.
10.Nf3 Qd8
Position after: 10...Qd8
As befits a position like this, we work on the understanding that White has adequate compensation in a vast
variety of ways, but investigate the forcing attempts to get more. The obvious choice is between short castling
right now, and trying to prepare long castling (or at least keep options open) with Bf4 – the latter is seen next.
11.0-0
11.Nxf6+ Qxf6 12.0-0 Bd6 13.c4 Nd7!= We could go further, but Black has an easy enough life; ...e5 can
come soon and then either ...Qe7 & ...Nf6 or just ...Nc5.
11...Be7 12.c4
12.Bf4 transposes to the next subchapter.
12...Qc7!
12...Nbd7 This is too compliant and allows White to make the last of the developing moves that he wants i.e.
Bf4. 13.Nxf6+ Bxf6 (13...Nxf6 14.Rd1²) 14.Bf4².
13.Bd2!
The bishop has to be developed, so why not do it immediately. Now Black has to put in a bit of effort to
untangle, because White has lots of tactics.
A) 13.g3 e5=
376
B) 13.Rd1 White loses a certain degree of flexibility by playing this first. 13...Nbd7 14.Bc2 (14.g3 e5!=)
14...b6 15.g3 Nxe4 16.Qxe4 Bb7 17.Bf4 Qc8=
C) 13.Be3 makes less sense, as White does not attack the e5 pawn with either Bc3 or Rfe1 if Black plays the
logical 13...e5!³.
Position after: 13.Bd2!
13...Nbd7
A) 13...0-0? 14.Nxf6+ Bxf6 15.Qe4²

B) 13...e5 14.Rfe1²
14.Rad1!
The best rook placement, eyeing the d6-square.

14.Rfe1 b6 15.g3 Nxe4 16.Bxe4 Bb7= is easy enough.
14...b6
Note that ...Nxe4 is also fine, if followed by ...Bf6 and ...b6 in analogous fashion to the text.
14...0-0?! however walks into it just a bit too much – Black lacks ideas after 15.Nxf6+ Nxf6 16.Bc3² when
White, by contrast, has a clear plan – he will be playing Ne5-g4 or Nd2-e4.
15.g3
15.Nxf6+ Bxf6 16.g3 Bb7 17.Bf4 will transpose.
15...Bb7
377
It is most accurate to reserve the right to recapture on f6 with the bishop.
16.Bf4 e5
17.Nxf6+!?
Quite testing, but I think this can be analysed to an endgame draw.

17.Nxe5 Nxe5 18.Nc3 An interesting tactical try. 18...Nfg4!
A) 19.Qxg4 Nxg4 20.Bxc7 0-0 Ironically, the presence of light-squared bishops seems to help Black. For
instance: 21.Be2 (21.Rfe1 Rfe8 22.Bf5 Nf6 23.a3 Bc8=) 21...Nf6 22.Bf3 Rac8 23.Be5 Rfd8=.
B) 19.Bf5 Bc8 20.Bxg4 Bxg4 21.Qxg4 Nxg4 22.Bxc7 0-0
378
We have reached an endgame in which there still exist one or two noteworthy wrinkles, but Black is equal
unless he does something stupid like exchanging his bishop for the knight. 23.Bd6 (23.h3 Nf6 24.a3 Rac8
25.Be5 Rfd8=; 23.Rfe1 Bb4 24.Bd6 Bxc3 25.bxc3 Rfd8=) 23...Rfe8 24.h3!? The best try in my opinion.
(24.Bxe7 Rxe7 25.Rd6 Ne5 26.b3 Nf3+ 27.Kh1 c5=) 24...Nf6 25.Bxe7 Rxe7 26.Rd6 Re6
27.Rxe6 (Following 27.Rfd1 Rxd6 28.Rxd6 Rc8= , ...Kf8-e7 next is not preventable.) 27...fxe6 28.Rd1 Kf7=
White may be able to lay claim to a moral edge based on pawn islands, but for practical purposes that is
deeply unlikely to result in a win.
379
17...Bxf6 18.Bf5
18.Rfe1 leads to a long series of exchanges. 18...0-0-0! (18...0-0 19.Bxh6²) 19.Nxe5 Nxe5 20.Bxe5 Bxe5
21.Qxe5 Qxe5 22.Rxe5 Kb8!
Position after: 22...Kb8!
The maximalist move. A holding approach was also possible with ...g6, meeting Re7 with ...Rhf8 and then
...Rde8. 23.Re7 Rhe8= Black is straightforwardly equal or perhaps even a shade better, unless White takes the
risky route of 24.Rxf7 c5, which is still dynamically balanced.
18...Rd8
380
Position after: 18...Rd8
19.Bxd7+
19.Nxe5 Nxe5 20.Rxd8+ Kxd8 21.Rd1+ Ke8 22.Re1 Kd8=
19...Rxd7 20.Bxe5
20.Nxe5 Re7 21.Rfe1 Qc8 22.Rd7!? 0-0 23.Rxe7 Bxe7=
20...Re7! 21.Bxc7 Rxe2 22.Rfe1 Re6!
22...Rxe1+?! 23.Rxe1+ Kd7 24.Be5! Re8 25.Bxf6 Rxe1+ 26.Nxe1 gxf6 27.Kg2²
Position after: 27.Kg2²
This endgame is unpleasant for Black; here we should mention the classic game Cohn – Rubinstein, St
Petersburg 1909, where the reader could look to get an idea of White’s plan and chances.
23.Rxe6+ fxe6 24.Ne5
24.c5 bxc5 25.Ne5 g5 White’s advantage is in name only.
24...c5
381
This was coming anyway. In this particular case Black forces a draw in an opposite colour bishop ending.
25.Bb8 Bxe5 26.Bxe5 Bf3 27.Rd6 0-0! 28.h3
28.Rxe6 Rd8 29.Rd6 Rxd6 30.Bxd6 Be2=
28...Rf5 29.Bc3 Be2 30.b3 Rf3! 31.Be5 Rd3 32.Rxe6 Kf7=
Position after: 32...Kf7=
382
The rooks come off; both sides mine their opponent’s queensides to the tune of 2 pawns each, and peace results.
1.2) 9.Ne4!? Qxd4 10.Nf3 Qd8 11.Bf4
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Bc4 Ngf6 6.Ng5 e6 7.Qe2 Nb6 8.Bd3 h6 9.Ne4!? Qxd4! 10.Nf3
Qd8 11.Bf4!
Position after: 11.Bf4!
This seems most testing to me, and unlike the previous line there seems no way to quickly diffuse the
compensation. Nevertheless, the position is sufficiently forcing that analysis makes sense.
11...Be7
A) 11...Nbd5?! Favoured in practical tests, but seems a little simplistic, and Black ends up moving his knights
even more than White did. 12.Be5 Nxe4 (12...Be7 13.0-0 0-0 14.c4²) 13.Qxe4 Qa5+ (13...Nf6 14.Qe2 Bd6
15.0-0-0 Bxe5 16.Nxe5² was basically a dream attacking position in Frede, L – Dettmann, G Mecklenburg
VP 1998) 14.c3 Nf6 15.Qf4 Be7 16.0-0 0-0
383
So far following Hoogland, D – Drenchev, P France 2006. Now 17.Rfd1!? improves greatly on 17.b4 and
continues the gentle questioning into exactly what Black is doing with the bishop c8. 17...Bd7 (17...Qb6
18.c4²) 18.Bc2 Be8 19.c4²
B) 11...Nxe4 is similar. 12.Bxe4 Bd6 13.Rd1 Nd5 14.Bxd6 Qxd6 15.0-0 Qc5 16.c4² Moran, G – Mareco, S
Moron 2003.
C) 11...Nfd5!? 12.Be5 forces Black to play some quite ugly chess, for instance 12...f6 13.Bg3 f5
384
14.Ned2 (14.Ne5 h5³) 14...Be7 15.c4 Nb4 16.Bb1 c5 17.Ne5 0-0 18.a3 Nc6 19.Nxc6 bxc6 20.Nf3©
Objectively, this may not be terrible, but White is having the fun.
12.0-0
Obviously, White could have made the last 2 moves in the other order.
A) 12.c4?! Bb4+ 13.Ned2 Nh5= fails to convince, so White must really castle before embarking on c4.
B) 12.Nxf6+ Bxf6
Position after: 12...Bxf6
looks like what White should be doing, but for concrete reasons he has trouble playing both c4 and a dark
squared bishop trade. 13.0-0-0 (13.Be5 Nd7„) 13...Nd5 14.Be5 b5! 15.Bxf6 Qxf6 16.Ne5 a6! Structural
integrity is important. (16...Bb7 17.Bxb5ƒ) 17.Nxc6 Bb7 18.Ne5 (18.Na5 Nf4 19.Bxb5+ Kf8 20.Qd2 axb5
21.Nxb7 Rxa2 22.Qb4+ Kg8³) 18...0-0 19.g3 Qe7=
12...Nfd5!
Better than on the last move for concrete reasons.

A) 12...0-0 13.Nxf6+ Bxf6 14.Qe4² should be a familiar theme by now.
B) 12...Nbd5 type things are just as bad as ever: 13.Be5 Nxe4 14.Bxe4 0-0 15.c4².
13.Be5 f6
385
14.Bg3
14.c4?! fxe5 15.cxd5 exd5 16.Nxe5 0-0 17.Ng3 Bd6 18.Rae1 Bxe5 19.Qxe5 Qf6³
14...f5 15.Ned2
A) 15.Ne5? 0-0µ
B) 15.c4!? fxe4 16.Bxe4 0-0! Black has to give the piece back. 17.cxd5 exd5 18.Bc2

386
18...Bg4! It was critical to avoid Ne5. (18...Nc4 19.Rfe1 Bb4 20.Nh4‚) 19.h3 Bxf3 20.gxf3 Rf6 21.Qd3
Bd6© Dealing maturely with the kingside problems; Black should not be worse.
15...Nb4
Now f5 is protected!
16.Bc4!
Taking the right knight for the bishop.

16.Ne5 0-0 17.c3 Nxd3 18.Nxd3 Qe8= Black is getting organised, for instance 19.c4 Nd7!? 20.Qxe6+ Qf7
with ...Bg5, ...Re8 or ...Nc5 next.
16...Nxc4
A) 16...N4d5 shows only that the machine doesn’t understand chess: 17.h3 (17.Bxd5 cxd5 18.Nd4 Bf6
19.Nxe6 Bxe6 20.Qxe6+ Qe7 21.Qxf5 0-0! was the point, winning back the b2 pawn.) 17...0-0 18.Bb3²
B) 16...Nxc2 17.Rad1 Nd4 18.Nxd4 Qxd4 19.Be5 Qg4 20.Qxg4 fxg4 21.Bxg7 Rh7 22.Bc3² is a desperately
inhuman way to play, but may not be terrible.
17.Nxc4 0-0
387
18.Rad1
18.h3!? f4 19.Bh2 Qc7 20.Nce5 b5 is similar.
18...Nd5
The difference between this line and 16...N4d5 is that Black’s knight will be untouchable on d5. A sample
continuation is...
19.h3 f4 20.Bh2 b5 21.Nce5 Qc7©
388
Position after: 21...Qc7©
White has full compensation due to Black›s disastrous structure, but Black is developing, so White should cash
in sooner rather than later, with a line like
22.Rfe1 Bd6 23.Nd4 Bxe5 24.Qxe5 Qxe5 25.Rxe5 Bd7 26.Nxe6=
1.3) 9.N5f3 c5: 10.Be3 and others
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Bc4 Ngf6 6.Ng5 e6 7.Qe2 Nb6 8.Bd3 h6 9.N5f3
Position after: 9.N5f3
Generally acknowledged as the main move, though it is not completely clear to me why; as we have seen
9.N5e4 is no walk in the park for Black either.
9...c5 10.Be3?!
In recent times no strong players have ventured this (indeed, 5.Bc4 itself is under a theoretical cloud) and there
are 2 good reasons why. We shall also look at some other minor moves.
10.c3 Qc7 Now ...c4 is a serious idea, and White can no longer really complete his development in a way that
doesn’t leave him with an IQP. 11.Ne5 (11.dxc5 Bxc5 12.Ne5?! Bxf2+!³ is a riff on a fairly well known
opening trap idea.) 11...cxd4 12.cxd4 Bb4+ 13.Kf1 Bd7 14.Ngf3 Rc8 15.g3=
389
Position after: 15.g3=
The position is objectively equal, but from a strictly practical point of view Black should be preferred due to his
long term trumps; these were well realised in Petrushin, A – Burmakin, V Krasnodar 1997.
10...Nbd5
Equalising in dynamic fashion.

A) 10...a6!? is quite intuitive and seems to do the job as well.
B) Following 10...Qc7?! 11.Ne5 Bd6 12.Ngf3 Black has tried a great many moves, but none of them quite
convince me. Possibly the quick draw that followed in Svidler, P – Karpov, A Tilburg 1996 had more to do
with respect than anything else.
11.Ne5
11.0-0-0 a6 should transpose as White cannot make do without Ne5.
11...a6 12.Ngf3
390
Position after: 12.Ngf3
12...Nxe3
12...Be7 13.0-0 Qc7 14.Bd2² Black has delayed the trade too long and now it is no longer available.
13.fxe3 Bd6
13...b5? 14.Bg6!!±
14.a4 b6!
A) 14...Bxe5 is way too materialistic; White has a pleasant choice between the obvious gambit and 15.dxe5
Nd7 16.Qd2².
B) 14...0-0 will transpose to the text.
15.0-0 0-0
15...Bb7 16.a5 forces a pawn structure weakness that Black may not wish to make.
16.c3 Bb7=
391
Position after: 16...Bb7=
Both sides have basically completed development. The position resembles a London system where neither side
has any obvious breaks (e4 and b4 for White and ...b5 for Black all have obvious dis-advantages.) A rich
manoeuvring game lies ahead and all three results are possible.
1.4) 9.N5f3 c5 10.dxc5 Nbd7: 11.Bf4 and others
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Bc4 Ngf6 6.Ng5 e6 7.Qe2 Nb6 8.Bd3 h6 9.N5f3 c5 10.dxc5!
Position after: 10.dxc5!
392
Obvious and best.
10...Nbd7!
The most ambitious, aiming for a position with a bishop-pair imbalance, as this will dwarf the pawn structure
imbalance in importance.
10...Bxc5 has been played a great many times for no obvious reason; it is well known by now, surely, that the
Rubinstein French structure after (say) the follow-up 11.Ne5 Nbd7 (11...0-0 12.Ngf3 a6 13.g4‚ Andreikin, D
– Sjugirov, S Moscow 2012) 12.Ngf3 Qc7 13.0-0 0-0 14.Bf4 Bd6 15.Nxd7 Bxd7 16.Bxd6 Qxd6 17.Ne5 is
extremely pleasant for White, especially if he is able to snap off the d7-bishop in the style of Fischer...
11.Bf4
A) 11.c6?! is the mate trap I alluded to in the introduction. 11...bxc6
Position after: 11...bxc6
12.Nd4 Played with a confident smile by one of my team-mates, but the smile and confidence were
misplaced; also, the choice of line was quite literally ill-advised, if you see what I mean... (12.Nd2 Nc5
13.Bc4 Be7 14.Ngf3 0-0=) 12...Nc5!³ In Horton, A – Arkell, K Buxton 2013, the experienced Grandmaster
on the Black side easily dispatched the trappy attempt that had been made on his life by the young White
player. A side note: the White player is now an FM with 2 IM norms as of time of press, as well as one of my
best friends. (After 12...c5?? 13.Nc6 Qc7
393
14.Qxe6+!! delivers mate.)

B) 11.Ne5
B1) 11...Nxe5 may also be fine but the text has been played far more and is easier. Notice especially the
following difference when compared to 8.Bb3: 12.Qxe5 Qa5+ 13.Bd2 Qxc5 14.Qg3 There is no ...Ne4 idea
for Black. White won convincingly in Petrov, M – Arkell, K Penarth 2011.
B2) 11...Nxc5 12.Bb5+ Bd7 13.Nxd7
Position after: 13.Nxd7
394
Now the struggle revolves around the bishop pair.
B2.1) 13...Ncxd7 is what the engine says, but this leads to a less pleasant imbalance in the position if White
takes the necessary steps to conserve his bishop pair. 14.c3 (14.Nf3 a6„) 14...Bd6 15.Nf3 0-0 16.0-0 Nc5
17.Qc2 Qc7 18.g3 (18.Re1 Bf4=) 18...a6 19.Be2 The position is of course objectively fine for Black right
now but the minority attack is not a viable plan and so once White’s had a few more developing moves he
will have momentum on his side.
B2.2) 13...Nfxd7 14.Nf3 a6! Now logical moves from both sides are: 15.Bxd7+ Nxd7 16.0-0 Qc7
(16...Be7 17.Bf4 Qa5 18.Rfd1 Rd8=) 17.b3 Be7 18.Bb2 Bf6 19.Bxf6 Nxf6 20.c4= With a completely
standard Rubinstein French position; here Black is probably better placed to contest the d-file than normal.
C) 11.b4 is most critical, on which more later.
11...Nxc5 12.0-0-0
Position after: 12.0-0-0
12...Be7
Not being afraid of ghosts.

12...Nxd3+?! hands White the d-file. After 13.Rxd3 Qa5 (Or 13...Nd5 14.Nh3!) 14.Kb1 Be7 15.Ne5 0-0
16.Ngf3² Black finds it hard to complete development of his queenside.
13.Ne5
Logically consistent with the previous note.

13.Bb5+ Bd7 14.Ne5 Qa5! 15.Nxd7 Nfxd7 16.Kb1 Rd8=
395
Position after: 16...Rd8=
White keeps the bishop pair for now, but in order to keep it against ...a6 and ...b5 ideas he will have to go
through some contortions that make Black’s queenside attack into an actual thing.
13...Qa5
13...Nxd3+ is again tempting, but bad: 14.Rxd3 Nd5 15.Bg3! b5 16.Nc6 Qd7 17.Nxe7 Qxe7 18.Nf3².
14.Kb1 0-0 15.Ngf3 Bd7!

396
Thus completing queenside development.
16.Nxd7
16.c4 Nxd3 17.Rxd3 Rfd8=
16...Nfxd7 17.Bc4
The only way to keep the bishop pair. White may actually be hoping for it to be chased around.
17.Bb5 Rfd8 and ...a6 is next with tempo.
17...Bf6=
Position after: 17...Bf6=
Black saves decisions about . ..Nb6 or ...a6/...b5 for another day, and in fact now ...Qb6 is also a significant
idea. The position is dynamically equal as White’s bishop pair should, of course, also not be underestimated.
17...Nb6?! 18.Bd2² does not acheive anything in particular for Black, and in fact gives White the option of
putting his own bishop on the long diagonal.
1.5) 9.N5f3 c5 10.dxc5 Nbd7 11.b4 a5!?
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Bc4 Ngf6 6.Ng5 e6 7.Qe2 Nb6 8.Bd3 h6 9.N5f3 c5 10.dxc5!
Nbd7! 11.b4!?
397
Position after: 11.b4!?
Of course, we should always carefully investigate attempts to keep sacrificed pawns! Now Black has two
options, and they are both very much playable, so a coin was tossed. The reader may wish to independently
investigate 11...b6, on which I created long and unnecessarily complicated analysis, if for any reason the text is
not to his taste.
11...a5
11...Nd5 12.Bd2 Qf6 13.Rc1!² is an obvious novelty that someone will play before long; suffice to say that
since 11...b6 and 11...a5 are both fine, we need not do any special work to change my view that White stands
better here.
12.c3 axb4 13.cxb4 b6
Black is clearly a fan of the Noteboom variation, but here there are far more immediate tactics.
14.Nd4 Qc7!?
With the aid of some concrete details Black now obtains clear equality.
A) 14...Nxc5?! 15.Bb5+ Ncd7 16.Bd2 Bb7 17.Ngf3² White’s game is slightly preferable; against careless
play he wins the bishop pair in a fairly open position.
B) 14...Nd5?! This may be better than ...Nxc5, but both are questionable. 15.Nc6!?
398
Position after: 15.Nc6!?
15...Qc7!! Almost allowing mate! (15...Qf6 16.Bb2 Nf4 17.Bxf6 Nxe2 18.Be5 Nxg1 19.Bb2 bxc5 20.b5²)
16.Qxe6+! Ne7! 17.Qe2!! Qxc6 18.Be4 Qa4 19.Bxa8 Qxa8 20.Nf3 bxc5 21.0-0 cxb4 22.Re1² The computer
sees no great problem with Black’s position, but as he is not going to succeed in castling, it seems to me that
eventually the rook will prove superior to the two minor pieces.
15.Nb5
15.Ngf3 bxc5 16.Nxe6!?
Position after: 16.Nxe6!?

399
This is a highly speculative sacrifice, but worth noticing all the same. 16...fxe6
A) 17.0-0 Bd6 (17...Bb7?! 18.Qxe6+ Kd8 19.Rd1©) 18.Bg6+ Ke7µ
B) 17.Bg6+ Kd8 18.0-0 Bb7µ
C) 17.Qxe6+ Kd8 18.0-0 Bd6 (18...Qb6 19.Qf7! Bb7 20.Rd1 Bd5 21.Qg6© keeps the f8 bishop on a bit of a
leash) 19.Qb3 Bb7 20.Rd1 Rc8³ White naturally continues to have a fair amount of compensation, but nothing
like what he hoped for.
15...Qb7
16.Nf3
16.Bf4 Qxg2 17.f3 Qxh1 18.Qxe6+ is an entertaining perpetual check (Bd6+/Bc7+), and it could be that by
now White should head for this.
16...bxc5 17.0-0
17.Bf4 c4! 18.Nc7+ gives Black two nice options. 18...Kd8 (18...Qxc7!? 19.Bxc7 Bxb4+ 20.Nd2 cxd3
21.Qxd3 Nd5© reverses the roles: now Black has sacrificed material and will be successful in finding
compensation for it.) 19.Bxc4
400
19...Bxb4+ (19...Nh5 unnecessarily lets White castle at the cost of a piece: 20.Nxe6+ fxe6 21.Bd2‚) 20.Kf1
Nh5 21.Be5 Nxe5 22.Nxe5 Qxc7 23.Qxh5 Ke8„
After a long forced sequence, we arrive at a position where the chief defining feature is the great weakness of
both kings, but underlying it is superior development and pawn structure for Black.
17...Be7!
17...cxb4 Black has no need to go into this, and the concrete problem I would point out is 18.Bb2 Be7 19.Bxf6
Nxf6 20.Rac1 0-0 21.Rc7 Qb6 22.Rxe7 Nd5 23.Ne5!‚.
18.Bb2 0-0 19.Bxf6 Nxf6 20.Rfc1
401
Position after: 20.Rfc1
White is really scraping the barrel here, and Black just needs a few more moves.
20...c4! 21.Bxc4 Bxb4 22.Ne5 Qe4!=
Black has completely neutralised the tricks and either side can outplay the other from here.
1.6) 8.Bb3 alternative
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Bc4 Ngf6 6.Ng5 e6 7.Qe2 Nb6 8.Bb3
402
Position after: 8.Bb3
8...h6
8...c5?! A move Black really, really wants to play. It may well be playable, but gives White an easy route to the
nice, Rubinstein-French setups we have talked so much about, or else a positional bind. 9.N1f3 cxd4 (9...h6
10.dxc5 Bxc5 11.Ne4 Nxe4 12.Qxe4 0-0 13.0-0²) 10.0-0 (10.Ne5 Bb4+ 11.Bd2 Bxd2+ 12.Qxd2 0-0=) 10...h6
(10...Be7 11.Rd1 is painful to even look at, even if it’s only slightly worse for Black) 11.Ne4 a5 12.a4 Nbd7
Position after: 12...Nbd7
13.Nxd4 (13.Rd1 Nxe4 14.Qxe4 Nc5= was the point of inserting ...a5 a4: now White won’t be getting the half-
open a-file in return for the bishop pair.) 13...Nxe4 14.Qxe4 Nc5 15.Qe5 Be7! (15...Qd6 16.Qe3!² is computer
chess at its finest; 15...Nxb3 16.Nxb3²) 16.Be3 Nxb3 17.Nxb3 0-0 Here Black’s disadvantage isn’t quite as
clear-cut but it still exists, especially if White exchanges dark-squared bishops. 18.Rfd1 Qe8 19.Bc5²
9.N5f3
9.Ne4 Qxd4 is nowhere near as convincing as in the 8.Bd3 lines, because the e4-knight is now threatened.
9...c5
Most natural to me, and by a slim margin the most played move.
9...a5 is the other main move, and has been played by all the big names of the Caro – Kann, but I don’t trust it.
It seems to me that c3 is a move White probably wanted anyway, and that a subsequent ...c5, if met with an
exchange and Be3, would serve only to show the weaknesses in Black’s camp. 10.c3
403
10...a4 (10...c5 11.Bc2!? Nbd7 12.Bf4 b6 13.0-0-0 White may not be better, but his plans are easy to execute,
and additionally Black’s queenside pawn pushes may have only served to make Ba4 a possibility.) 11.Bc2
A) For 11...c5 12.dxc5 Bxc5 13.Ne5 Nbd7 14.Ngf3 Qc7 15.0-0² see Adams, M – Speelman, J London 1995.
B) 11...Qd5 could be my choice if I ever reached this position as Black; the play is somewhat original and
here at least I was not able to find a way to obtain a position I preferred for White. 12.h4!? Bd6 13.Nh3 c5„
However, this is insufficiently solid or well-defined to be a reasonable re-commendation for a repertoire.
C) 11...Nbd5 12.Bd2 Qb6 13.Rb1 c5 14.dxc5 Bxc5 15.Ne5² With natural play White had diffused all the
tricks associated with Black’s early a-pawn thrust in Anand, V – Epishin, V Biel 1993.
404
10.dxc5
A) 10.Be3 is every bit as toothless as in the 8.Bd3 system. 10...Nfd5 (10...Nbd5!? 11.0-0-0 Qc7 12.Kb1 a6
13.c4 Nxe3 14.fxe3 Be7 15.Ne5 0-0 16.Ngf3 b5 Sax, G – Adorjan, A Hungary 1993) 11.0-0-0 (11.c3 Nxe3
12.fxe3 Bd6=) 11...c4!? 12.Bxc4 Nxc4 13.Qxc4 a6©
B) 10.c3 Qc7! 11.dxc5 (11.Ne5 cxd4 12.cxd4 Bb4+= Schmid, L – Petrosian, T Hamburg 1960) 11...Bxc5
12.Be3 (12.Ne5 may not be terrible here but after 12...Bxf2+ 13.Kxf2 Qxe5 14.Qxe5 Ng4+ 15.Ke2 Nxe5
16.Nf3 Nxf3 17.gxf3© the compensation seems only barely adequate if that.) 12...0-0 13.Bxc5 Qxc5 14.Ne5
Bd7
For once, I am not objecting to a Rubinstein type setup, because Black has a huge amount of development and
the threat of ...Bb5.
15.a4 Bc6 16.Ngf3 Nbd7 (16...Be4!?) 17.Nxc6 bxc6=
10...Nbd7 11.Ne5
11.Bf4!? Nxc5 (11...Bxc5 12.0-0-0 with Ne5 to come looks promising) 12.Rd1 Qa5+ 13.c3 Nxb3 14.axb3
Qf5! Else White’s dark-squared control might have come to something. 15.Ne5! (15.Bg3 Qe4=; 15.Bd6 Bxd6
16.Rxd6 0-0³) 15...Qxf4 16.Qb5+
405
Position after: 16.Qb5+
16...Bd7! (16...Ke7 17.Ne2 Qf5 18.Ng3 Qf4 19.0-0±) 17.Nxd7 0-0-0 18.Ne2 Qc7 19.Nxf6 Rxd1+ 20.Kxd1
gxf6 21.Qe8+ Qd8+ 22.Qxd8+ Kxd8 23.b4 f5 After more or less forced play, we get a level endgame – though
as always there is still a lot to play for.
11...Nxe5
11...Nxc5 12.Ngf3 Be7 13.0-0 seems far less promising than analogous positions from the Bd3 system,
because when Black takes the b3-bishop he cedes a half-open file.
12.Qxe5 Qa5+
406
Black heads for an endgame.
13.Bd2
13.c3 Bxc5! (13...Qxc5 14.Qg3!? could be ever so slightly annoying, as in Petrov – Arkell) 14.Qg3 0-0!
15.Bxh6! Bxf2+! 16.Qxf2 Ng4=
13...Qxc5 14.Nf3 Bd6!N
14...Bd7 15.Be3 Qxe5 16.Nxe5² Simacek, P – Smolkov, I Pardubice 2015.
15.Qxc5 Bxc5=
Position after: 15...Bxc5=
If anything, Black has slightly more ideas in this position. White’s hand is being forced by the presence of a
Ne4 motif, so maybe Be3 is called for but then the position will be totally equal.
CONCLUSIONS TO CHAPTER 1
Like the rest of the Chapter 1’s in this book, the present one was light on criticality, but nevertheless
important to have read and understood, unlike some of the other chapters which Black may choose to skip. I
would not be surprised at all, for instance, if the gambit line makes something of a revival.
407
Chapter 2
Minor 5th moves
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.--
Chapter guide
Chapter 2 – Minor 5th moves
A. 1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Nf3 Ngf6
2.1) 6.Nc3 e6?!
2.2) 6.Nc3 Nb6!
2.3) 6.Bd3 (Polgar-Anand, San Luis 2005)
2.4) 6.Qe2!?
B. 1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Qe2 Ndf6 6.c3 Nxe4 7.Qxe4 Nf6
2.5) 8.Qc2 Qd5?!
2.6) 8.Qc2 Bg4 9.Ne2 g6! 10.Ng3
2.7) 8.Qc2 Bg4 9.Ne2 g6! 10.Nf4
With the choice of title above it was my intention to convey the idea that across this chapter and the next, there
are a vast number of possible transpositions. Thankfully, though, Black’s moves are essentially unchanged,
except that he should be careful not to fall for smothered mate.
That being said, this chapter and the next were the hardest for me to write. Probably that is because I play some
of these myself from the White side, as you might be able to tell from the greater amount of detail I give on
these lines compared to some other authors!
In 2.1) and 2.2), we explore first the idea behind 6.Nc3, and then the detail of how it should be disarmed. I have
not seen fit to do the same for 6.Bd3, as I consider that move to require far less subtlety to equalise against, and
408
indeed for one of only a handful of times in the whole book, I have embedded an entire ‘modern classic’ game
into the text, so exemplary is Black’s handling of how to play against 6.Bd3.
Line 2.4) is an interesting, very minor, hybrid attempt that I think might deserve more attention in the near
future, especially for White players; the resulting position is dynamically balanced. It segues nicely into the
following theme, which is lines with an earlier c3-push from White.
For the final three lines of the chapter, then, we examine 5.Qe2, which is on one level a crude attempt to set up
a mate trick, and on another a fairly sophisticated move aiming to play against the e5-square and the
development of Black’s bishops (in simple terms: Black must either fianchetto his king’s bishop, or cede the
bishop pair, or shut his c8-bishop in with ...e6) and none of the above is particularly appealing. The end result is
a deceptively sharp struggle where Black usually tries to play ...e5 in one. He is ultimately successful in
equalising, but not before some subtleties.
2.1) 5.Nf3 Ngf6 6.Nc3 e6?!
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Nf3 Ngf6 6.Nc3!?
Position after: 6.Nc3!?
Aiming for a slight plus based on the misplaced d7-knight. This is probably what happens, unless Black moves
it right away. We shall investigate briefly 6...e6?! before moving on to the best move 6...Nb6.
6...e6?!
6...Nd5 7.Nxd5 cxd5 8.Bd3 g6² eventually led to a long endgame win for Black in McPhillips, J – Fernandez,
D Leeds 2015, but I would not try and deceive anyone into thinking Black is close to equalising like this.
7.g3
Targeted against eventual ...c6-c5.

409
7...b5!?
Trying for Meran type play.

A) 7...Bd6 is terribly passive. As this is not a White book as such, I shall limit myself to the comment that my
analysis indicates a White plus after 8.Bg2².
B) 7...Be7 is likely also too passive after 8.Bg2 0-0 9.0-0².
8.Bg2 b4
9.Na4!?
9.Ne2 Bb7 10.0-0 c5 11.a3 bxa3 12.b3 cxd4 13.Nexd4=
9...Ba6
9...Bb7 10.a3 bxa3 11.b3 (11.0-0!?) 11...c5 12.Bxa3 Qa5+ 13.c3 cxd4 14.Bxf8 Nxf8 15.Qxd4 N8d7 16.0-0 0-0
17.Rfd1 is maybe slightly more pleasant for White.
10.a3 bxa3 11.b3
410
11...Qa5+
Black continues trying to disrupt c4/0-0.

11...Qc7 12.c4 Bb4+ 13.Bd2 Be7 14.0-0 0-0 15.Re1²
12.Qd2 Qb5 13.Bf1 Qb7 14.Bxa6 Qxa6 15.Bxa3 Bxa3 16.Rxa3 0-0 17.c4
It is not clear how to evaluate this position. White has a better structure but less development. Nevertheless, I
411
could not find a convincing move to recommend to Black now that he is powerless to stop castling, so we
terminate here with a note that the line could be considered if Black finds a continuation, and move on.
2.2) 5.Nf3 Ngf6 6.Nc3 Nb6!
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Nf3 Ngf6 6.Nc3 Nb6!
Position after: 6...Nb6!
Conceptually, the move that easily makes most sense. Black’s equalising plan is, as usual, based upon ...Bg4.
7.h3
A) After 7.Bd3 Bg4!= Black equalises in fairly standard fashion; h3 is met by ...Bh5-g6.
B) 7.Bg5 might appear to discourage moves that allow the f-pawns to be doubled, but Black is actually
completely fine with this and his piece deployment makes sense after 7...Bg4 8.h3 (8.Bxf6 exf6 9.Be2 Be7
10.0-0 0-0 may even be better for Black since he can avoid swapping his Bg4, by retreating it to e6.) 8...Bh5
(8...Bxf3 9.Qxf3 Qxd4© might not be something Black wants to get into.) 9.Bxf6 exf6 10.Bd3 Be7 11.0-0 0-0
12.Re1 Re8=
412
Position after: 12...Re8=
In principle White has a static advantage, which grows with the correct trades; however, getting his Nc3 to
play a reasonable role is more difficult than it might originally appear. If White had the ability to play, say, c3
and Nc3-g3 in one move, he would be easily for choice.
C) 7.a4 a5! 8.Bg5 Bg4= This is certainly not a better version for White than the one without the pawn moves.
D) 7.Be2 Bg4 8.0-0 e6 9.h3 Bf5 10.Nh4 Bg6 11.Bf3 Bd6 12.Qe2 0-0 13.Bg5 Qc7 14.Rad1 Nbd5 15.Nxd5
cxd5 16.Nxg6 hxg6= Yu, Y – Ding, L Beijing 2012.
7...Bf5
Now, otherwise White will play Bd3, restricting the development of exactly this piece.
8.Bd3 Bxd3 9.Qxd3
413
Position after: 9.Qxd3
9...e6
The most natural.

9...h6 Now White plays without developing his c1 bishop for now, and is slightly better. 10.Ne5 e6 11.0-0²
With Qg3 to come after either bishop development.
10.0-0 Be7
So that tempi are not gained by a later Ne4.

A) 10...h6 11.Ne5 transposes to the previous note.
B) 10...Bd6 11.Bg5 h6 (11...0-0 12.Ne4 Be7 13.Bxf6 Bxf6 14.c4 leads to a type of position I like the White
side of, with the knight pair and slightly more space.) 12.Bxf6 Qxf6 13.Ne4 Qe7 14.c4 Bc7 15.Rad1 0-0
16.Rfe1²
C) 10...Qc7 11.Bd2 Bd6 12.Rfe1 Bf4 13.Bxf4 Qxf4 14.Ne2 Qe4 15.Qd2² Kistler, D – Campbell, J USA
1985
11.Ne2 0-0 12.c4
414
12...Nbd7
12...c5 doesn’t solve Black’s problems. 13.Bf4 cxd4 14.Nexd4 Rc8 15.b3² The b6-knight is misplaced.
13.Bf4
We reach a standard Caro-Kann type position, where Black suffers from a slight lack of space, but cannot
actually be said to be worse.
13...a5 14.Rad1
Natural.
14.a4?! As in an Inarkiev – Ding game. 14...Nc5! 15.Qe3 Na6 It is instructive how Black develops counterplay
of full value using just the b4 square and environs. 16.Rfd1 Nb4 17.Rd2 b5 18.b3 bxc4 19.bxc4 Qb6 20.Nc3
Rad8 21.Rad1
415
21...Nd7?! (21...Qa6N 22.Ne5 Nd7 23.Qg3 Nxe5 24.Bxe5 f6=) 22.Ne4!?N (22.Qe2 Rfe8 23.Bg5= Inarkiev, E
– Ding, L Moscow 2014) 22...Qa6 23.Bc7!? Rc8 24.Bd6 By this tactical method, White defends c4 without
playing Qe2, and even gains a tempo in the process. 24...Bxd6 25.Nxd6 Rcd8 26.Ng5 Nf6 (26...h6?!
27.Ngxf7! and Black is in trouble.) 27.Nde4 Nd7 28.Qb3 h6 29.Nf3 Qb6 30.Ne5²
14...a4 15.Nc3
The most constructive try, aiming for a d5 push.

416
15...Qa5
Trying for some plan like ...a3, ...Bb4.

15...a3 16.b3 Qa5 17.Ne1! is better for White than the similar 15...Qa5 16.Rfe1 a3, because now a knight
comes to c2 quickly and harrasses the a3 pawn in case of a trade of dark squared bishops.
16.a3
A) 16.d5 is a flash in the pan: after everything comes off Black should attain good play. 16...cxd5 17.cxd5
Nxd5 18.Nxd5 exd5 19.Qxd5 Qxd5 20.Rxd5 Nc5 21.Re1 (21.Be3 Rfc8=) 21...Ne6 22.Rb5 Bc5 23.Be3 b6=
B) 16.Rfe1 a3 Missed in the analysis of Evgeny Postny. Black seems to be just about fine. 17.b3 Bb4 18.Bd2
Rfd8=
16...Rfd8
Position after: 16...Rfd8
The manoeuvring continues, but Black has halted the d5 plan in time, and can be reasonably confident about the
future. There are ideas of ...Nc5-b3, or ...Nf8-g6 next.
2.3) 5.Nf3 Ngf6 6.Bd3 (Polgar-Anand, 2005)
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Nf3 Ngf6
417
Position after: 5...Ngf6
6.Bd3?!
6.Bd3 is indeed a minor line nowadays, but it was well-trodden before, and discarded for very good reasons. I
now claim author’s prerogative and showcase a very nice game from Anand.
6.Nxf6+ Nxf6
A) 7.Bc4 is the move people play if they really don’t get opening subtleties and just want to develop their
pieces. Here, by contrast to say the same move after 5.c3, Black is able to develop his light squared bishop,
just not to g4! 7...Bf5 8.0-0 e6 9.c3 Be7 10.Re1 0-0= Here, despite my usual good eye for subtlety, I am
inclined to agree with the judgement of Messrs. Petrov and Arkell, who agreed a draw here in 2016.
418
Position after: 10...0-0=
B) 7.Ne5!? Nd7! After ...Bf5, 8.c3 is oddly annoying. 8.Nd3 (8.f4 Nxe5 9.fxe5 Bf5=) 8...g6!
For once I am fine to fianchetto the bishop, because it obviously lends support to ...e5. This is unstoppable
unless White plays f4, since Black can always bring four pieces to the defence of e5 after castling. Now in
Mestel, A – Miles, A London 1982 Black equalised as follows: 9.c3 Bg7 10.Be2 e5 11.dxe5 Nxe5 12.Nxe5
Qxd1+ 13.Bxd1 Bxe5 14.Bh6 Be6 15.Bb3 Bxb3 16.axb3 Kd7!= and indeed a purist could argue that he has
the better bishop.
6...Nxe4 7.Bxe4 Nf6 8.Bd3 Bg4=
419
Position after: 8...Bg4=
Black has completely equalised. The remainder of the game is included for general chess education purposes!
9.Be3
9.c3 e6 10.h3 Bh5 11.Qe2 (After the possibly better 11.0-0 Bd6= Black has vague ideas of ...Qc7 and then
...Bf4.) 11...Be7 12.Bd2 a5!
Discouraging long castling. 13.a3 (13.0-0-0 a4 14.a3 Bg6ƒ gives a significant amount of pressure on the light
420
squares.) 13...Qd5 14.g4 Bg6 15.Bxg6 hxg6 16.c4 Qe4³ Velimirovic – Smagin, Zenica 1987; White’s d2
bishop is still miserable and Black has reasonable winning chances in the ending.
9...e6 10.c3 Bd6 11.h3 Bh5 12.Qe2 Qa5!
Position after: 12...Qa5!
Prophylaxis against 0-0-0 and preparing to go to d5.
13.a4?!
Essentially admitting to not knowing what is going on. White is trying to long castle anyway and then start an
attack! While on the previous few moves White could have castled short and been fine, now he is decidedly
worse.
13...0-0 14.Qc2 Bxf3 15.gxf3 Qh5 16.0-0-0 Nd5!
This knight move is good on general grounds, and with it Black keeps in reserve the idea of exchanging the
Be3, either for the knight itself, or with ...Bf4, which is now supported.
17.Kb1 b5 18.Rdg1 f6 19.axb5 cxb5 20.Bc1 Rab8
Engines on good depth now see that Black’s attack will land first.
421
Position after: 20...Rab8
21.Qe2 Rfe8 22.Qe4
22.Rg4!? f5 23.Rg5 Qh4³ keeps the damage smaller.
22...Kh8! 23.h4
23.Qg4 Qf7 with ...f5 to come does not gain a tempo at all.
23...f5! 24.Qe2 Qf7
In any case as in the previous note.
25.Rg2 Bf4
25...b4 26.c4 Qd7 is the engine recommendation, but there is absolutely no need for this.
26.Rhg1
422
Position after: 26.Rhg1
26...Rg8
26...Bxc1? 27.Rxg7! Qf6 28.Kxc1 causes a horrendous mess that Black need not get into, even if he still has
decent attacking chances with ...b4.
27.Be3 Qd7 28.Qd2 Bd6 29.Bc2 Qb7 30.Bg5 b4 31.c4 b3! 32.Bd3 Bb4 33.Qe2 Qa6 34.Bh6
Position after: 34.Bh6
34...Nc3+! 35.bxc3 Bxc3 36.Kc1 Qa3+ 37.Kd1 Qa1+ 38.Bc1 b2 39.Qe3 Bxd4 40.Qd2 bxc1=Q+ 41.Qxc1
423
Qa2
White resigned here in Polgar, J – Anand, V San Luis 2005.
2.4) 5.Nf3 Ngf6 6.Qe2!?
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Nf3 Ngf6 6.Qe2!?
This move is not completely stupid, and it reaches a quite original position nearly by force. There are obviously
transpositions from 5.Qe2.
6...Nxe4
6...e6 7.Nxf6+ Nxf6 8.g3 gives White a nice comfortable position.
7.Qxe4 Nf6 8.Qd3
There is no c2 square available, but on the plus side White has Qb3 ideas.
8...Bg4!?
8...g6 9.g3 Bg7 10.Bg2 gives White his desired development: 10...Bf5 11.Qe2 0-0 12.0-0².
9.Ne5!
Other choices on the last 2 moves allow the bishop to reach its favoured post on f5 without playing ...g6.
9.c3 e6 10.Ne5 Bf5=
424
9...Bh5
9...Be6 is another valid way to play, aiming for a future ...c5 since White is not really able to fianchetto, but it
looks quite artificial. 10.Be2 g6 11.c3 Bg7 12.0-0 0-0 13.Re1 Rc8 14.Bf3²
10.c3 e6
11.Qh3!
White has a few more tactics in reserve to stop ...Bd6, ...0-0, ...c5.
A) 11.g3 Bd6=
B) 11.Qg3!? Bd6 12.Be2 0-0 13.Bxh5 Nxh5 14.Qf3 Nf6 15.Bg5 Be7=
11...Bd6! 12.g4!?
12.Bg5 runs into a string of tactics: 12...Bxe5! 13.dxe5 Qd5 14.Bxf6 Qe4+ 15.Qe3 Qxe3+ 16.fxe3 gxf6
17.exf6 Rg8 18.h3!? Rg3 19.Kf2 Rg6=.
12...Bxe5 13.gxh5 Bc7„
The play is fluid and objectively fine for Black, with his development advantage and compact structure to
balance out the bishop pair, but White has also gotten what he wanted.
425
Position after: 13...Bc7„
2.5) 5.Qe2 & 8.Qc2 Qd5?!
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Qe2!?
This is one of the rare sidelines that I feel may have been unfairly marginalised thanks to its reputation as
mostly a trick (Nd6 mate, if you were wondering!).
5...Ndf6
426
The conventional wisdom is that since ...Ngf6 has been stopped, Black should try and exchange the e4-knight
by this method instead.
6.c3 Nxe4
6...Qd5!? 7.f3 Bf5 8.Ng3 Bg6 9.Nh3 Qd7 10.Nf4 Nd5 11.h4² Tan, W – Fernandez, D, Ho Chi Minh City
2011.
7.Qxe4 Nf6 8.Qc2
Position after: 8.Qc2
By one method or another, White will be able to get this position if he wishes, against any 4...Nd7 repertoire
which makes sense. If we had chosen 5...Qc7 as our repertoire choice, White could instead play with 5.c3 and
reach it anyway.
8...Qd5
As quite often in this part of the book, we have an instructive choice between ...Qd5 and ...Bg4, and we should
go into why I’m not recommending ...Qd5.
8...g6 as usual is not a serious challenge for equality: 9.Nf3 Bg7 and now 10.Bc4² here might not be such a
terrible idea even if it does compromise White’s development ideals slightly.
9.Nf3!
The main move for a reason.

9.Ne2!? Could be a better knight placement for the ending in some abstract theoretical sense, but invites ...e5.
9...Bf5?! (9...e5!N improves: 10.dxe5 Qxe5 11.Bf4 Qe6= In the battle of obscure queen manoeuvres to hinder
the opponent’s development, Black has clearly come up trumps.) 10.Qb3 Qxb3 11.axb3
427
Position after: 11.axb3
In my database, this position has been reached three times, and in each the White player was Romanian IM Alin
Ardeleanu. To illustrate the problems:
A) 11...Nd5 12.b4 e6 13.Ng3N (13.Nf4 Bd6 14.Nd3= Ardeleanu, A – Dobre, C Olanesti 2013) 13...Bg6
14.f3²
B) 11...e6 12.b4 Bd3 (12...Bd6 13.f3 0-0 14.Bf4 Bxf4 15.Nxf4² Jablecnik, G – Gillet, R Dresden 2000 (by
transposition)) 13.Nf4 Bxf1 14.Kxf1 Bd6 15.Nd3² Ardeleanu, A – Dodic, D Paracin 2014.
9...Bf5 10.Qb3 Qxb3 11.axb3 e6 12.b4
428
This really looks like no fun for Black.
12...Bd6 13.Be2ƒ
Ardeleanu, A – Pisu, R, Arad 2016.

13.b5!?N 13...0-0 14.bxc6 bxc6 15.Ba6²
2.6) 5.Qe2 & 8.Qc2 Bg4 9.Ne2 g6! 10.Ng3
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Qe2 Ndf6 6.c3 Nxe4 7.Qxe4 Nf6 8.Qc2 Bg4
The main move, and a lot of instructive practical tests have happened.
9.Ne2!?
9.Bd3 tries to play on ever smaller wrinkles, but they should not amount to much. 9...Bh5! (9...e6 10.Ne2 Bd6
is also possible as in Chan, Y – Paciencia, E Singapore 2008, though maybe now 11.Ng3N gets somewhere.)
10.Ne2 Bg6
429
11.Nf4!? (11.Bg5 e6= was even easier equality for Black in my then best draw: Fernandez, D – Seirawan, Y
Vlissingen 2011) 11...Bxd3 (11...e6 12.Nxg6 hxg6 13.g3 gets the standard imbalance) 12.Nxd3 e6 13.Bf4
Be7= The position is essentially equal. Note, however, that White would have an actual edge in the event that
Black had had to fianchetto instead.
9...g6!
Clearly the most logical, responding to changing circumstances.

A) 9...e6 is a little stubborn: 10.Ng3 Qd5 11.f3 Bh5 12.Nxh5² Lutikov, A – Petrosian, T Leningrad 1960.
B) 9...Bh5 10.Nf4 Bg6 11.Nxg6 hxg6 12.Be2 e6 13.0-0 Bd6 14.g3²
10.Ng3
430
The first of two moves that make Black develop in particular orders lest he suddenly find himself worse.
10.h3 Bf5 11.Qb3 Qb6=
10...Bg7
The most natural way to follow up 9...g6.

10...Qd5 is an engine suggestion, but as so often in these lines, Qb3 at some point will now be good. 11.f3 is
likely the best way to play, creating play on the dark squares: (11.Bf4!?; 11.Qb3!? Restricting Black’s choice of
recapture on d5 when compared to the version with f3 Bd7 included.) 11...Bd7 12.Qb3 b5 13.Qxd5 cxd5²
White has a static edge based on the central dark squares.
11.Bc4 0-0 12.0-0 Qc7
431
This move and ...Rae8 intending ...e5 is the simplest thing to recommend. Any other deployment of pieces is
susceptible to some kind of disruptive action.
13.Re1 Rae8! 14.Bg5 e5 15.dxe5 Rxe5 16.Rxe5 Qxe5 17.Qd2 Re8 18.h3 Bc8 19.Rd1=
Position after: 19.Rd1=
White has milked his slightly superior co-ordination to the maximum, but it has nevertheless fizzled out.
2.7) 5.Qe2 & 8.Qc2 Bg4 9.Ne2 g6! 10.Nf4

432
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Qe2 Ndf6 6.c3 Nxe4 7.Qxe4 Nf6 8.Qc2 Bg4 9.Ne2!? g6! 10.Nf4
This is an interesting try, and both sides engage in some quite sophisticated tempo play. I don’t think there is a
straightforward way to remember where Black’s light-squared bishop must go to maintain the balance in what
follows.
10...Bg7
Now I shall take a look at what I consider to be White’s two most interesting tries, which both attempt to make
nuanced use of the d3 square rather than just plonking a bishop down there.
11.h3!?
11.Be2 Bf5! (11...Bd7 12.0-0 0-0 13.Qb3² is hard to play for Black, whose dark squares are not looking
healthy. Next can be: Nd3, Bf4/g5, Rad1, Rfe1, and if ...a5 then always a4.) 12.Qb3 Qb6 13.0-0 0-0= Black
seems to be just on time with ...c5 or ...e5.
11...Bd7!
It is best not to ‘gain a tempo’ now.

11...Bf5 is the obvious alternative, but White’s next gains back the time. 12.Bd3 e6 (12...Bxd3 13.Nxd3 0-0
14.0-0²) 13.Bxf5 exf5 (13...gxf5 14.g4!‚) 14.Qe2+ Qe7 15.Qxe7+ Kxe7 16.Nd3 Kd7 17.f3 Nd5 18.Kf2²
433
Position after: 18.Kf2²
Black will manage to get some counterplay, but the difficulty of ever arranging ...c5 means that in the long run
he will be defending.
12.Bc4 0-0 13.0-0 Bf5!
This move juggles Black’s ...b5 and ...e5 ideas in a unique way, and the question posed to White’s queen does
not have an answer that avoids giving concessions.
434
14.Qe2!
A) 14.Bd3 e6! The point is that now even if Black needs to take back with the e-pawn, his king can no longer
be embarrassed in the centre. 15.Bxf5 exf5 16.Nd3 Qd5=
B) Following 14.Qb3 Qc7 15.Re1 Rae8= White will not be in time to prevent ...e5.
C) 14.Qd1 is even more subtle, but Black is active enough after 14...b5 15.Be2 Qd6 16.Nd3 Nd5= that even
potentially ceding the bishop pair in future is not a big issue.
14...b5 15.Bb3 a5
Black does not insist on ...e5, but begins playing for ...a4 and/or ...b4. Now, very little is forced from here, so
we shall confine ourselves to a few more general comments.
16.a3
Hopefully it should be obvious from the other sample lines that White should avoid Black’s ...b4, and if
possible also avoid retreating his bishop to d1.
A) 16.g4 Bc8 is not something to be afraid of; the bishop will be resurrected on the a6-f1 diagonal. Black just
needs to not watch his pawns too carefully! For instance, the following pawn sacrifices are all sound. 17.Re1
a4 18.Bc2 b4!
435
19.Qc4 (19.Qxe7?! Qxe7 20.Rxe7 g5 21.Nh5 Nxh5 22.gxh5 h6„ threatens both ...Bxh3 and ...a3, and
White’s extra pawn is essentially a mirage.) 19...bxc3 20.bxc3 Qa5! 21.Rxe7 (21.Qxc6?! Bb7!ƒ) 21...Ba6
22.Qb4 Qxb4 23.cxb4 g5 24.Nh5 Nxh5 25.gxh5 Bxd4 26.Rb1 h6= Black has a respectable measure of
control.
B) 16.a4 is not the way to try and neutralise Black’s queenside play: 16...b4 17.Re1 Rb8! 18.Bc4 bxc3 19.bxc3
Ne4„ may even be easier for Black to play thanks to his control of the b-file.
C) 16.Re1 a4 17.Bd1 Re8³ is very nice for Black.
16...a4 17.Ba2 h5 18.Nd3 Bxd3 19.Qxd3 Nd5=
436
Position after: 19...Nd5=
Black has a perfect ‘light squares only’ setup. He should attempt to play ...Qd7 and then prepare either ...e5 or
...c5 with minimal delay, and this is relatively easy to do. The position is equal.
The hardest question in this chapter (and also the next one, but we’ll get there) was how and when Black
should develop his c8-bishop. Shutting it in in line 2.1) led only to misery, leaving it at home but with the
door open in line 2.5) wasn’t good either, and in lines 2.6) and 2.7), it proved necessary to develop this
bishop and then bring it back again a couple of moves later.
With no hard-and-fast rules to obey, these White sidelines risk confusing the principle-oriented Black
player. There are two possibilities: memorise this entire chapter up to move 14, or find some mnemonic that
works for you (Houska swears by her mnemonics in her book and also her training sessions!)
437
Chapter 3
More challenging lines after 5.c3
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.c3
Chapter guide
Chapter 3 – More challenging lines after 5.c3
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.c3 Ngf6
3.1) 6.Bd3 Nxe4 7.Bxe4 Nf6 8.Bd3 Qd5!
3.2) 6.Bd3 Nxe4 7.Bxe4 Nf6 8.Bc2 Qd5?!
3.3) 6.Bd3 Nxe4 7.Bxe4 Nf6 8.Bc2 Bg4!
3.4) 6.Nxf6+ Nxf6 7.Bc4
Here we proceed even further with our discussion of hidden subtleties in overlooked corners of the c3-complex.
Once more, Black has choices between ...Qd5 and ...Bg4, and in this case, the answer varies even within the
chapter, due to nothing more than which piece White is able to bring to b3. Black equalises with good play, but
the key is to know something about the line at all, rather than overlooking it in a simplistic way.
The main point of the chapter comes at the end: in line 3.4) we investigate the hybrid system between c3 and
Bc4, which forces a concession from Black straightaway. Different players have chosen to give it in different
ways, but I have made a clear choice (7...Qc7) and explained why I believe it to be the only one that equalises.
3.1) 6.Bd3 Nxe4 7.Bxe4 Nf6 8.Bd3 Qd5!
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.c3
438
5...Ngf6 6.Bd3
6.Nxf6+ transposes to almost every other line in this part. The lone independent try might be: 6...Nxf6 7.g3!?
Bf5! If Black has a free choice, he should prefer this over ...Bg4. He has equalised, e.g. 8.Bg2 e6 9.Nf3 h6=.
6...Nxe4 7.Bxe4 Nf6
Now White has two very subtly different bishop retreats. The main point of 5.c3 is that White may know more
of the differences than Black.
8.Bd3
8.Bf3 Bf5 9.Ne2 e6= fails completely to disrupt Black’s co-ordination.
8...Qd5!
8...Bg4?!
439
Position after: 8...Bg4?!
After Bd3 rather than Bc2 it could well be this move which is dubious. 9.Ne2 Tried by me on one occasion, and
I think White wins the bishop pair under reasonable circumstances. (9.Nf3 Qd5 10.Be2 e6 11.h3 Bh5=)
A) 9...Bxe2 10.Qxe2 e6 11.f4!N (11.0-0 Be7 12.b3 0-0 13.Bb2 Qc7= was the continuation of Fernandez, D –
Arkell, K England 2014, as seen in the introduction.) 11...Be7 12.f5 exf5 13.Bxf5 0-0 14.0-0 Re8 15.Qf3²
B) 9...Bh5 A try analogous to White’s play in the next line, however here the bishop pair is lost to 10.Qb3
Qd7 11.Nf4 Bg6 12.Nxg6 (12.h4!? e6 13.h5 Bxd3 14.Nxd3 h6 15.Bf4 Be7 initially looks good for White,
but unless he can organise an actual kingside attack, Black will slowly unravel and may even stand better in
the long run.) 12...hxg6 13.0-0 e6 Transposing to 9...e6.
C) 9...e6 10.Qb3!
440
Position after: 10.Qb3!
10...Qd7 (10...Qc7 11.Bf4 is not necessary to allow; 10...Qb6 11.Ng3²) 11.0-0 (11.Ng3 e5 12.0-0 exd4
13.Bc4 0-0-0„ I doubt White gets anything here as he is not developed enough.) 11...Bh5 The best way to try
and equalise. Black can’t yet develop his f8-bishop because of Ng3. (11...Be7 12.Ng3 e5 13.Bc4²; 11...Bd6
12.Ng3²) 12.Nf4 Bg6 13.Nxg6 hxg6 14.Re1 Bd6 15.g3
Position after: 15.g3
There is little point continuing this line, as it should be clear that White’s kingside is robust enough to
withstand doubled rooks on the h-file, and that ultimately the bishop pair should tell.
441
9.Qf3
9.Nf3? Bg4=
9...g6!
Here White does not have the option of Bb3, and there is no real try for an advantage that I can see.
9...e6 10.Qg3 once again seems strong. 10...c5 11.Be3ƒ White is developing with Nf3 or Ne2, and has the idea
of Bb5+ followed by c4 in reserve.
10.Bd2
10.Qxd5 Nxd5 11.Bd2 Bf5=
10...Bf5!
Saving Black even the tiniest amount of trouble.

10...Bg7 Interestingly, the main point of ...g6 is to exchange light-squared bishops, not even to fianchetto the
dark squared one. 11.Qxd5 Nxd5 (11...cxd5 12.f3 0-0 13.Ne2 Bd7 14.Kf2 White seems to have slightly easier
play due to better bishops.) 12.Nf3
442
12...Bg4 Now almost necessary on a sort of micro-positional level, but good enough for an essentially equal
position anyway. (12...0-0 13.h3! Keeps the c8-bishop at home; as in earlier lines this guarantees White
relatively easy play. 13...Bf5 14.Bxf5 gxf5 15.g4²) 13.Ne5 Bf5 14.Bxf5 gxf5 15.Nd3 0-0-0 16.0-0-0 e6 17.h3
h5 18.g3= White may have slightly more ideas but the position is objectively dead level.
11.Qxd5
11.Bxf5 Qxf5 is an extra option, and probably a good one
11...Nxd5=
443
Position after: 11...Nxd5=
The trade of light squared bishops will be next, after which Black will develop his dark squared bishop to d6
and begin considering the timing of either ...c5 or a minority-attack-type construction with ...b5, ...a5-...a4. This
is enough to counterbalance White’s slightly better control of the centre.
3.2) 6.Bd3 Nxe4 7.Bxe4 Nf6 8.Bc2 Qd5?!
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.c3 Ngf6 6.Bd3 Nxe4 7.Bxe4 Nf6 8.Bc2!
Position after: 8.Bc2!
This seems to be the better retreat.
8...Qd5?!
One of the main points of 5.c3 is that the same two Black moves make most sense against both bishop retreats,
but the orders of preference are opposed. It is of some importance to understand why this move isn’t great.
9.Qf3
9.Nf3?! Bg4 fully equalises, and might be practically easier for Black, since Black forces a pawn weakness
which can be fixed.
9...e6
Enough trades have happened that this is a palatable idea.

A) 9...Bg4 10.Qxd5 Nxd5 (10...cxd5 11.h3 Bh5 12.Ne2 Bxe2 13.Kxe2²) 11.Ne2 g6 (11...Bxe2 12.Kxe2²;
11...e6 12.Ng3²) 12.h3 Bf5 13.Bxf5 gxf5 14.g4²
444
B) 9...Qxf3 10.Nxf3 Bg4 11.Ne5 Bh5 12.f3²
C) 9...g6 10.Bb3! A fairly important difference between the 2 bishop retreats. 10...Qxf3 11.Nxf3 Bg7 12.0-0
a5 (12...0-0 13.Re1 Re8 14.h3² Black has no chance to play ...b6 without ...e6, and if he must play ... e6, then
why bother fianchettoing the bishop to g7 in the first place?) 13.a4² Black faces issues with developing the
bishop on c8 stably.
10.Qg3!
Position after: 10.Qg3!
10.Bb3 Qxf3 11.Nxf3 b6 12.Bf4 Bb7 13.0-0 Be7 is a sample line but probably White gets nothing from Bb3 if
Black has played ...e6 rather than ...g6.
10...b6!
Only like this. Black causes disruption of his own by playing ...Ba6. This is the drawback of White’s bishop
placement!
A) 10...Bd6 11.Qxg7 Rg8 12.Qxf6 Qxg2 13.Qf3 fails miserably.
B) 10...Be7 11.Nf3 0-0 12.0-0² Black constantly has Bh6 hanging over his head, and Re1-e5 is hardly
innocuous either.
C) 10...g6 11.Nf3² and Black is not in time to cause co-ordination problems.
11.Nf3
11.Bb3!? leads to play that is unstereotyped enough for the most exacting taste: 11...Qb5 12.Bd1 Ba6 13.Be2
Qa4=.
11...Ba6 12.Ne5
445
12...g6
12...Nd7 13.Nxd7 Qxd7 14.a4!? Qd5 (14...g6 15.Qe5 Rg8 16.a5 Bd6 17.Qe3² White is better, because he can
still play Bd3 and castle!) 15.a5 bxa5 16.Qe5²
13.Bd1!?²
An engine move, highlighting the issue with Black’s ...b6. The creativity continues from both sides, but the
feeling is that Black’s situation is more precarious, so I am not recommending Black play this way.
3.3) 6.Bd3 Nxe4 7.Bxe4 Nf6 8.Bc2 Bg4!
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.c3 Ngf6 6.Bd3 Nxe4 7.Bxe4 Nf6 8.Bc2! Bg4!
446
Position after: 8...Bg4!
9.Ne2 Bh5!
The second not-immediately-obvious move in a row. We see a drawback of Bc2 compared to Bd3: there is no
longer a Qb3 idea.
9...e6 10.f3 Bh5 11.Nf4 Bg6 12.Nxg6 hxg6 13.g3
Position after: 13.g3
might be okay, but the bishop pair is an under-rated force, particularly here since White has played c3 rather
than c4, and Black doesn’t have to allow this.
447
10.0-0 e6
There are still a few more careful moves to make before we can assert equality.
11.Bg5 Be7
A) 11...h6 is generally a bad idea since when White lifts the pin on his knight, Black will have the unenviable
choice between taking it, and allowing Ng3/f4. 12.Bh4²
B) 11...Bg6 12.f4!?ƒ
12.Re1
12.f4 Nd5=
12...0-0 13.f3
13...Bg6!
As if on autopilot, Black plays ... Bg6 when Nf4 was threatened, and not before.
14.Nf4 Bxc2 15.Qxc2 Nd5!?
Best since Black wants material symmetry.

15...h6 16.Bxf6 Bxf6 17.Nd3 could even yet be a little something, as Black has no breaks and the knight is
better placed.
16.Bxe7 Nxe7 17.Nd3 Qc7=
448
Position after: 17...Qc7=
White can’t stop all of ...b6/...c5, ...Ng6-f4, and ...Ng6/...Rfe8/...e5.
3.4) 6.Nxf6+ Nxf6 7.Bc4
We have already seen this move, but now we look at the hybrid of this with c3.
449
5...Ngf6
5...Nb6 is of course playable, but represents extra work, since White could have obtained the position after
move 7 with a 5.c3 order instead; and could also have played 5.Ng5 Ngf6 6.Bc4 to reach White’s other line.
6.Nxf6+ Nxf6 7.c3 Qc7!
7...g6 Played by Ding Liren and Victor Bologan, but it does not seem as good, mostly because it develops the
bishop to a closed diagonal. 8.Qb3 (8.Nf3 will almost inevitably lead to standard bishop-pair-advantage
positions) 8...Nd5 9.Nf3 Bg7 10.0-0 0-0 11.Re1² Some subtleties are required, but I think White can prove an
edge here.
8.Qb3
Otherwise it is not clear why White has gone to all this trouble.
A) 8.Nf3 Bf5 9.0-0 h6 (9...e6 10.Nh4 Bg4 11.f3 Bh5 is perfectly fine, but Black does not even need to let
White have control over the minor-piece balance.) 10.Nh4 (10.Ne5 e6=) 10...Be4 (10...Bh7 11.f4²) 11.f3
Bd5! 12.Bd3
450
12...0-0-0! 13.f4 (13.c4? Be6 14.Be3 g5 15.Nf5 Bxf5 16.Bxf5+ e6 17.Bc2 c5ƒ) 13...e6= With ...c5 to come
next, White’s Nh4 is definitely far more misplaced than Black’s d5-bishop. Thus concludes White’ forcing
play.
B) 8.Qf3!? Now, Black seems to have no better choice than ceding the bishop pair to his opponent. 8...Bg4
(8...g6 9.Bf4²; 8...e5 9.Qg3²) 9.Qg3 Qxg3 10.hxg3 e6 11.f3 Bf5 12.g4 Bg6 13.Ne2
So far the play after 8.Qf3 has been more or less what we could call “weakly forced”. The bishop pair will be
451
lost; Black just has to find the best way to give it.
B1) 13...Be7 14.Nf4 0-0-0 (14...Kd7 15.Be2!?² seems to continue asking searching questions.) 15.Kf2 Nd5
16.Nxg6 hxg6 17.Bd2²
B2) 13...Bd6! Attacking the e5 and f4 squares. 14.Nf4 Kd7 (14...Bc2 15.Bb3 Bxb3 16.axb3; 14...0-0-0
15.Bd2 Nd5 16.Nxg6 hxg6 17.Ke2= could be okay too, but I prefer ...Kd7.) 15.Kf2 (15.Be2 Nd5 16.Nd3
h5= Black is on time since Ne5 isn’t a problem now.) 15...Nd5 The other point of ...Bd6 is to challenge the
Nf4, so that White can’t take his own sweet time over winning the bishop pair. 16.Nxg6 (16.g3 Nxf4
17.Bxf4 Bxf4 18.gxf4 h5=) 16...hxg6 17.Bd2 b5 18.Bd3 Bf4=
8...e6
Having extracted a concession from White (his queen position) Black is content to play this way.
9.Nf3
Trying to play some kind of micro-game surrounding the development of the dark-squared bishops.
9...Bd6!
Important to get this right and not be afraid of ghosts.

A) 9...Be7?! 10.0-0 0-0 11.Ne5 The claim is that after Bf4 next, Black will need to play ...Bd6 anyway.
B) 9...b6?! Can easily transpose; Black should be extremely cautious about doing anything else though. His
independent options connected with ...Bb7 are suspect to say the least. 10.0-0 Bb7?! 11.Re1
452
11...Bd6 (11...0-0-0 12.a4 a5 13.Ne5 Kb8 14.Bf4 Bd6 15.Qd1‚ White’s aggressive ideas are very clear: b4
and Qe2/Ba6; 11...Be7 12.Ne5 sets up a large volume of tactics based on Nxf7, and Black must lose time.)
12.Bxe6 Just the fact of having allowed this should be enough to put off Black players from the line, and you
can take my word for it that great precision is now required to draw.
10.Bg5 0-0
Refusing to be bothered by the downside of having developed the bishop to d6. It seems to me that Black
453
obtains full dynamic equality like this: his plan of ...b6, ...Bb7, ...c5 is easy.
11.0-0 b6 12.Bd3 Bb7
...c5 cannot be stopped and all White can do now is think of how to try and make dynamic threats on the
kingside.
13.Rfe1
13.Qc2 h6 14.Bxf6 gxf6=
13...c5
13...h6 14.Bxf6 gxf6 15.c4 was similar enough in Korneev, O – Rasmussen, A Brescia 2009 though I would
prefer to play for ...c5 with Black.
14.Bxf6 gxf6 15.dxc5 Bxc5
15...bxc5 16.Be4 leaves open the possibility of losing in a long minor piece ending.
16.Be4 Kh8„
Position after: 16...Kh8„
Black will obtain compensation of full value for his broken structure, in the form of g-file play. As usual for
such positions, he should be careful not to trade queens. The position is dynamically equal.
454
Once more the question arose multiple times of when and where to develop that c8-bishop. It is easy, then,
to sympathise with Keith Arkell, whose modus operan-di appears to be to exchange it off for another minor
piece – exactly which one is not relevant – at the earliest opportunity. However, with accurate play
(9...Bh5! in line 3.3) it is possible to ensure that the right trade occurs (light square bishop for light square
bishop).
In some ways, line 3.4) is easier, as we play ...e6 rather quickly, and the decision is made for us. White has
some tries for an advantage, but nothing too serious.
455
Chapter 4
Surprise weapons for Black after 5.Ng5
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Ng5
Chapter guide
Chapter 4 – Surprise weapons for Black after 5.Ng5
4.1) 5...h6!?
4.2) 5...Nb6
4.3) 5...Ngf6 6.Bd3 e5 7.Bc4 recommendation 1
4.5) 5...Ngf6 6.Bd3 e5 7.N1f3 exd4 8.Bc4 Nd5 9.Qxd4 h6 10.Ne4
4.6) 5...Ngf6 6.Bd3 e5 7.N1f3 exd4 8.Bc4 Nd5 9.Qxd4 h6 10.0-0!N
This chapter is something of an ‘extra’, like the final chapter of Parts II and IV (though not I). As such, feel free
to skip it if you’re on your first read or if you’re just after a bare-bones repertoire. The focus in the first two
lines is exchanging off at least one of Black’s bishops in creative ways, and obtaining slightly worse yet solid
positions at the end. However, when Caro players have less bishops than their opponents, they never have
problems thinking of things to do, so it may be that these lines could suit some people.
In 4.3) through 4.6) we take a detailed look at 6...e5, which is a sideline played a few times by Ding Liren. It
requires the strong piece sacrifice 10.0-0! in the final line of the chapter to firmly put paid to its equalising
ambitions, though in one way or another the system continues to be somehow playable.
456
4.1) 5...h6!?
5...h6!?
This move is frequently listed as an ‘opening trap’ together with its ‘friends’, 5...Ngf6 6.Bd3 h6 and the
infamous 5...Ngf6 6.Bd3 e6 7.N1f3 h6 of Deep Blue fame.
6.Ne6
Otherwise there was no point playing 5.Ng5!
6...Qb6
6...Qa5+?! 7.Bd2 Qb6 would be good if Black actually had a serious plan to take on b2 or d4, but is not, for
instance 8.Bd3! Ngf6 9.Nxf8 Nxf8 10.Ne2 Qxb2?! 11.Rb1 Qxa2 12.c4 and Black, while two pawns up,
struggles terribly with development.
7.Nxf8 Nxf8 8.c3!
The move that best constricts the Black light-squared bishop.

A) 8.Nf3 Nf6 and now White should play c3, reverting to the main line, rather than 9.Bd3 Bf5! when Black
exchanges a pair of bishops; this is the tactic underlying the whole system.
B) 8.Be2 Nf6 9.Nf3 Bf5 is of course not terrible, but is a little bit modest for White. Next will come ...Ng6,
...Rd8, ...e6, ...0-0.
457
8...Bf5
Position after: 8...Bf5
9.Nf3
9.Bc4 is also possible as an independent try to develop the knight to e2 and attempt f4-f5. 9...Nf6 10.Ne2 N8d7
11.0-0 0-0 12.Ng3 Bh7! Making use of 5...h6. Now a sample line could be (12...Bg6 13.f4 e6 14.f5²) 13.a4
(13.f4 c5=) 13...a5 14.b3 Rfe8
Position after: 14...Rfe8
458
when Black is solid, and thinking about ...c5 and ...e5, but objectively slightly worse as he has no obvious
counterweight to the bishop pair.
9...Nf6
9...Nd7 could be a small move-order wrinkle, aiming to control the e5 square and also not committing yet to
...Nf6 rather than ...Ne7. 10.Be2
A) 10...Ngf6 11.0-0 0-0 12.Re1 with a transposition.

B) 10...e6 is a distinct idea, but 11.Qb3! (11.0-0 Ne7 12.a4 Rd8 13.b3 0-0 14.Ba3 Rfe8 for instance lets Black
get his intended setup, with ideas of ...c5 or ...e5 in the near future, and the possibility of ...Ng6-f4.) 11...Ne7
12.Qa3² White forces ...c5 prior to castling and is therefore better.
10.Be2N
A) 10.Bd3 is bizarrely favoured by my engine but after 10...Bxd3 11.Qxd3 Rd8 12.0-0 Ng6 for instance,
Black is completely fine.
B) 10.Ne5
459
...is an interesting move, leading in some cases to positions where neither side has any kind of proper
development, but after some consideration I ditched these in favour of 10...c5 which is safest and now either
11.Bc4 Be6! or 11.dxc5 Qxc5 12.Bf4 a6 (12...Ne6 13.Qa4+) 13.Be2 Ne6 14.Bg3 0-0 is fine for Black.
C) 10.Bc4 was White’s choice in a Wei Yi – Ding Liren game so it deserves to be taken seriously. 10...Ng6
11.0-0 0-0 12.Qe2 Rad8 13.Ne5!
Position after: 13.Ne5!
13...e6 14.Nxg6 Bxg6 15.Bf4 Nd5 16.Be5 was the game continuation, and along the lines of our main line
White has the bishop pair, while Black has an extremely robust position, and even went on to win.
460
10...N8d7
Position after: 10...N8d7
In principle the knight stands better here than on g6, so as to support two central breaks rather than 1, and so as
not to constrict the retreats of the f5-bishop.
11.0-0 0-0 12.Re1 Rad8
With the idea of ...e5 next.
13.Bf1!
Avoiding ...e5 and avoiding creating a target for ...c5.

A) 13.Qb3 Qc7 14.Bf1 Rfe8
461
Position after: 14...Rfe8
Defending the e-pawn and now Black is prepared to play with ...c5, ...b6 and generate some approximation of
a Bogo – Indian structure. 15.g3 (15.Be3 Ng4 16.Bd2 c5=; 15.Ne5 is positionally desirable but with the
queen moves inserted the rook recapture is not possible. 15...Nxe5 16.dxe5 Ne4= Either going to c5-d3 or
d2.) 15...c5 (15...e5 16.dxe5 Ng4= is recommended by the engine but I am suspicious, as Black may need to
play ...g5 in order to comfortably regain the pawn) 16.Bf4 Qc8 17.Rad1 b6 18.Bg2 Be4=
B) 13.h3 e5 equalises and showcases Black’s idea. Now for instance 14.Qb3 (14.Bf1 exd4 15.Nxd4 Bh7=)
14...exd4 15.Nxd4 Bg6 16.Qxb6 (16.Bf4 Nc5„) 16...axb6 when Black gets active just fast enough (...Nc5-d3
ideas among others) to compensate for the bishop pair.
13...Rfe8 14.Ne5
462
Now the rook recapture can be done (contrast the first note to 13.Bf1).
14...Nxe5 15.Rxe5 e6
Black is very solid but the feeling is that he is probably a little worse, at least from a human point of view.
Nevertheless, this line gives a perfectly playable and stereotypically Caro-Kann position, similar in feel to Wei
Yi – Ding Liren from earlier.
4.2) 5...Nb6
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Ng5 Nb6
463
The second deviation. With this move, Black refines the approach of the 5...h6 line, seeking to play ...h6 only
on the following turn. White now has 2 continuations that make meaningful efforts to stop 6...h6.
6.N1f3
As seen in games Chandler – Adams and v.d. Wiel – Karpov.

6.Bd3 g6 White lacks things other than N1f3 to do. An idea is 7.h3 (7.N5f3 Bg7 8.c3 Nf6 9.Ne5 s coherent but
slow, e.g. 9...Qd5 10.Ngf3 Ng4!=) 7...Nf6!? 8.N1f3 h6 (8...Bg7 9.0-0 0-0 would transpose to the first note to
White’s 9th.) 9.Ne4 Nxe4 10.Bxe4 Bg7 11.0-0 Nc4„
464
Position after: 11...Nc4„
Here we have very stereotyped play for this line; Black wants to play ...Nd6 and ...Bf5.
6...g6!?
Renewing the threat of ...h6 by providing the Black king with some cover on g7.
7.Bd3
A) 7.h4 was chosen by Speelman’s opponent but here 7...h6 8.Ne4 Bg7 looks fine to me.
Black will follow up with ...Bg4 unless White probes new weaknesses with 9.Nc5 after which the rather
entertaining line might continue 9...Nd7! 10.Nd3 Ngf6 11.Be2 c5 12.dxc5 Nxc5 13.Nxc5 Qa5+ 14.c3 Qxc5
15.Be3 Qa5=.
B) 7.Ne5?! f6 8.Ngf7 (8.Nef7 Qd5 9.Nxh8 fxg5 and it transpires that the knight does not escape.) 8...Qd5!
9.Nxh8 (9.Nxg6 Kxf7 10.Nxh8+ Kg7!) 9...fxe5µ
7...Bg7 8.0-0 Nf6
465
The traditional method of playing this line is with 8...Nh6, aiming for creative knight play with ...Nf5-d6. It so
happens, though, that if Black feels the need he can route the knight via f6-e8 and get to d6 anyway.
9.c3!
As with 5...h6, White’s best try for an advantage consists of this structurally improving move.
A) 9.h3 prevents the Arkell-esque idea of ...Bg4, as well as future challenges to an e5 knight involving ...Ng4,
but is not all that critical. 9...0-0 10.c3 (10.Re1 Bf5 11.Bxf5 gxf5 12.Ne5 e6 and Black has accumulated the
long term pluses of the position.) 10...h6 11.Ne4 Nxe4 12.Bxe4 Nc4 13.Re1 Nd6 14.Bd3 Now Black’s best
is: 14...Bf5 15.Bxf5 Nxf5² with an insanely solid position.
B) 9.Ne5 The two players who reached this position with White up til now have chosen 9.Ne5. 9...0-0
B1) 10.Re1 was the choice of Susan Lalic, thinking that the d4 pawn was protected, but the problem is
10...Qxd4 11.Ngxf7 (11.Nexf7 Ng4! with ...h6 to come) 11...Ng4 12.Nh6+! Nxh6 13.Bc4+ Qxc4 14.Nxc4
Nxc4
466
Black has 3 pieces for a queen and there aren’t any immediate tactics, for instance 15.Bxh6 (15.Rxe7 Nf5
16.Re4 Nxb2 17.Bxb2 Bxb2 18.Rb1 Bf6©) 15...Bxh6 16.Qe2 Nb6 17.Qxe7 Bg7 18.c3 Bf6 19.Qa3 Bf5©
B2) 10.c3 While this looks very similar to the main line, White will lose time later due to the position of his
knight. 10...h6 11.Ne4 Nxe4 12.Bxe4 Nd7 13.Nd3 (13.Nf3 c5=) 13...Nf6 14.Bf3 Bf5=
9...0-0 10.Re1
467
The point is that if Black’s best bet with his bishop is Bf5, White may as well save on h3.
10...Bf5
10...Nbd5 Gearing up for other exchanging operations such as ...Nc7-e6. 11.h3 h6 12.Ne4 Nxe4 13.Bxe4 Bf5
14.Bxf5 gxf5 15.c4 Nf6 16.Bf4 e6².
11.Bxf5 gxf5 12.Ne5
Only now does White play Ne5, and he is also in time to avoid ...h6 forcing any particular concession.
12...e6 13.a4 Nbd7²
Position after: 13...Nbd7²
As usual for these deviations, White is probably better but Black is solid and holds certain long-term trumps.
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Ng5 Ngf6 6.Bd3 e5
468
This is the third deviation, which has been seen in some quite high-profile encounters and to which White must
react fairly accurately. Indeed, the logic is impeccable: why spend so much time preparing ...c5 on move 12 or
something when you can just go ...e5 in one and free your position this way?!
7.Bc4?!
Not like this! 7.N1f3 is better and seen later.
7...Nd5 8.N1f3 h6!
Provocative, but seemingly not punishable and so likely best.
9.Ne4 f5 10.Nc3
A) 10.Bxd5!? cxd5 11.Nc3 e4 12.Ne5
469
12...Nxe5 (12...Nf6 13.Ng6 Rg8 14.0-0 Be6 does not create a good impression, though it could be playable)
13.dxe5 Bb4 (13...Be6 is the pawn sac recommended by the engine, but I don’t think there is compensation,
say 14.Qh5+ Bf7 15.Qxf5 Qc8 16.Qxc8+ Rxc8 17.Be3 Bb4 18.Bd4²) 14.Qh5+ Kf8 15.0-0 Bxc3 16.bxc3
Be6
We reach an interesting position; White may be for choice in a practical game due to his attacking chances,
but objectively Black is fine right now.
470
B) 10.Ng3 e4 was played in some practical games, but here White’s knight sacrifices don’t seem adequate.
10...Nxc3
10...e4!? is given next and is my main recommendation.
11.bxc3 e4 12.0-0!
A piece sacrifice, but at the same time the only sensible move. Black’s position looks a little scary here, but he
should be fine.
12...g6
12...b5 asked too much from the Black position and wound up worse after 13.Bb3 Bb7 14.Re1 Qf6 15.Ne5±
Torma, R – Koczo, K Budapest 2013.
13.Re1 Bd6 14.Bb3 Kf8
Black completes castling by hand and his position seems playable.
15.Ne5
White will not punish his opponent’s excesses by retreating the knight.
A) 15.d5 c5 16.Nd2 Be5=
B) 15.Nd2 Kg7=
15...Nxe5 16.dxe5 Bc7
471
17.Ba3+
17.Qxd8+ Bxd8 18.f3 is another try to hold up Black’s development; now say 18...Ba5 19.Bb2 Kg7! is one
way of untangling, leaving the e4 pawn on the basis that 20.fxe4 f4! would isolate White’s second batch of
doubled pawns and leave him in trouble after a subsequent ...Re8.
17...Kg7 18.Bd6 Re8! 19.Qd4!
19.c4 Bd7
472
20.c5 (20.Qd4 Bxd6 21.exd6+ Qf6=) 20...Rxe5 21.Bxe5+ Bxe5© White is no longer better and is in fact at risk
of becoming worse.
19...Bb6 20.Qc4
20.Qd2 Qg5 and the Black central pawns begin to generate some unpleasant offensive ideas.
20...Be6
The opening is very much over as both sides are completing development on the next move. This seems like a
fine outcome for Black from the opening.
21.Qe2 Qd7 22.Rad1 Qf7„
The computer still likes white, but is not producing concrete reasons, and meanwhile he has a fairly bad
structure as these things go.
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Ng5 Ngf6 6.Bd3 e5 7.Bc4?! Nd5 8.N1f3 h6! 9.Ne4 f5 10.Nc3
e4!?
473
Position after: 10...e4!?
An even more ambitious attempt, given for completeness.
11.Ne5
11.0-0 doesn’t work this time: 11...exf3 12.Re1+ Kf7 13.Nxd5 cxd5 14.Bxd5+ Kg6 15.Qxf3 Nf6µ.
11...Nxe5 12.dxe5 Bb4 13.0-0
13.Qh5+ can be played first, but it is only a move-order thing; the attempt at independent play with 13...Kf8
will be met by 14.Bd2 Bxc3 15.Bxc3 (15.bxc3!? Qe7 16.0-0 Qxe5 17.Rad1 Be6 18.Bb3©) 15...Qe8 which
either forces the queens off or takes the bishop pair.
474
Position after: 15...Qe8
16.Qh3 (16.Qd1 Nxc3 17.Qd6+ Qe7 18.bxc3 Qxd6 19.exd6 Ke8 20.Bb3 Kd7 21.0-0-0 a5=; 16.Qxe8+ Kxe8
17.0-0-0 Be6 18.Bd2 Kf7 and Black is solid.) 16...f4 17.Bxd5 cxd5 18.Bb4+ Kg8 19.Qc3 Bf5„
13...Bxc3 14.Qh5+
14.bxc3 0-0=
14...Kf8 15.bxc3 Be6

475
16.f3
16.Qh3 indicates that White would not like to force matters immediately and is instead fine to play the long
game; now 16...Kg8 17.Bb3 Kh7 can be played, when Black has completed development and seems solid.
16...Qe8
Black either forces the queens off or advances the central pawn wedge.
17.Qh4
17.Qh3 f4 18.Qh4 g5³
17...Qe7 18.Qxe7+
A) 18.Qf2 e3 19.Qe2 f4µ

B) 18.Qg3?? Qc5+–+
18...Kxe7 19.Ba3+ Kf7!
Position after: 19...Kf7!
Giving up a pawn to quell the bishop pair.
20.fxe4 Ne3 21.Bxe6+ Kxe6 22.exf5+ Nxf5©
White may press in the ending, but not without risk!
4.5) 5...Ngf6 6.Bd3 e5 7.N1f3 exd4 8.Bc4 Nd5 9.Qxd4 h6 10.Ne4
476
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Ng5 Ngf6 6.Bd3 e5 7.N1f3!
Position after: 7.N1f3!
The second main move, and a much better choice.
7...exd4
A) 7...h6 8.Ne6±
B) 7...Bd6?! 8.0-0! 0-0 9.dxe5 Nxe5 10.Nxe5 Bxe5 11.Nxh7 Wins either the bishop pair or a pawn for which
there is minimal compensation.
8.Bc4 Nd5 9.Qxd4
9.0-0?! lets Black develop his king’s bishop: 9...Be7 10.Re1 0-0=.
9...h6
9...N7f6?! Unfortunately, this move deserves to be written off, and I shall demonstrate why. 10.Ne5 Qa5+
11.Bd2 Bc5
477
(11...Qc5 might preserve a little bit of the mess, but none of it is ever in Black’s favour. 12.Qd3 Be7
13.Ngxf7±)
A) 12.Nxc6? Bxd4 (12...Qb6!? is even more complicated) 13.Nxa5 0-0 14.0-0 h6 15.Nf3 Bxb2 16.Rab1 Bc3
17.Nxb7 Be6©
B) 12.Bxa5 does not prove anything either, for instance 12...Bxd4 13.Ngxf7 0-0 14.0-0-0 Bxf2 15.Nd6 Bc5=
C) 12.Qh4! Qc7
478
13.0-0!! (13.0-0-0 0-0 14.Rhe1 h6 15.Ngf3 Re8 led to reasonable play for Black in Balogh, C – Ding, L Saint
Louis 2012; 13.Nexf7 0-0µ) 13...0-0 14.Bd3 h6 15.Ngf3 Bd6 (15...Re8 16.Rae1 Bf8 does not lose
immediately to sacrifices, but the position is still very bad, e.g. 17.c4 Nb4 18.Bb1±) 16.Rae1± The perfect
duo of ideas- c4 and Bxh6 – make Black’s position essentially indefensible.
Now we have a split.
10.Ne4
This leads to play in which Black should probably at some stage give up a pawn in exchange for the bishop
pair, but should not be particularly afraid of this.
A) 10.Nxf7 Kxf7 11.Bxd5+ cxd5 12.Qxd5+ Kg6
...looks extremely scary, but White does not seem to have more than a draw; for instance 13.0-0 Qf6 14.Re1
(14.Be3 Kh7 and ...Nb6 will consolidate) 14...Nb6 15.Qd3+ Kf7 16.Ne5+ Kg8 17.Ng6 Bf5 18.Qb3+ Kh7
19.Nxh8 Kxh8³
B) 10.0-0 is seen next.
10...Nc5!
479
11.0-0
A) 11.Nxc5 Qa5+ 12.c3 Qxc5= and it is hard to come up with anything very convincing for White.
B) 11.Be3 Gaining tempo on the Nc5 in a beautifully indirect way. 11...Ne6 12.Qe5 (12.Qd2 Nxe3 13.Qxe3
Be7=) 12...Qa5+ 13.c3 (13.Bd2 Qc7 14.Bxd5 cxd5 15.Qxd5 Be7 16.0-0-0 0-0 and Black seems to have
compensation, though this could be developed further.) 13...Be7 14.0-0 0-0
Now, it is helpful but certainly not necessary that placing either rook on d1 leaves White open to tactics based
480
on queen sacrifices... 15.Rfd1 (15.Rfe1 Rd8 16.Rad1 Qc7 When the game goes on, but Black has solved his
development problems.) 15...Nxe3!! (15...Qc7 type pawn sacrifices offer a lot less compensation now, for
instance 16.Bxd5 Qxe5 17.Nxe5 cxd5 18.Rxd5 Nc7 19.Rdd1 b6 20.Nd6 Ba6 21.Nf5±) 16.Qxa5 Nxc4
17.Qf5 (17.Qa4 Nxb2 18.Qc2 Nxd1 19.Rxd1 Rd8 20.Rxd8+ Nxd8=) 17...Nd4
18.Nxd4 (18.Qh5?? Ne2+ 19.Kh1 Nf4–+) 18...Bxf5 19.Nxf5 Rae8=
11...Ne6 12.Qd3
12.Qe5 Qc7 13.Bxd5 cxd5 14.Qxd5 Be7 15.Be3 0-0 16.c3 Rd8©
12...Be7 13.Rd1 0-0
481
Here White can take on d5 in a diverse variety of ways but in all cases Black obtains the bishop pair, and they
are well enough deployed to compensate to a great extent by pressurising White’s pawns.
14.Bxd5
14.Qb3 b5 15.Bxd5 cxd5 16.Rxd5 Qc7©
14...cxd5 15.Qxd5 Qxd5 16.Rxd5 b6 17.Nd6 Ba6©
Position after: 17...Ba6©

482
Black obtains full compensation.
4.6) 5...Ngf6 6.Bd3 e5 7.N1f3 exd4 8.Bc4 Nd5 9.Qxd4 h6 10.0-0!N
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Ng5 Ngf6 6.Bd3 e5 7.N1f3! exd4 8.Bc4 Nd5 9.Qxd4 h6 10.0-
0!N
Position after: 10.0-0!N
Simple and good – White gets the bishop pair in a position where it is most useful. I think this is necessary to
claim an edge.
10...N7f6 11.Ne4
Now, and not before. The tempi work out just right for White to take the bishop pair without trouble.
11...Be6!?
Maximalism from Black, which is met by the same from White.

11...Be7 A secondary try which requires some finesse from White as well. 12.Nxf6+ Bxf6 13.Re1+
483
Position after: 13.Re1+
13...Kf8 (13...Be6 14.Qe4 Qc7 15.Nd4 Bxd4 16.Qxd4 0-0 17.b3 Here the bishop pair is very relevant, and
Black’s position does not have quite the usual resilience. Nevertheless there is still some solidity, e.g. 17...Rfe8
18.Bb2 f6²) 14.Ne5! (14.Qd3 g6 promises nothing since Black will get in ...Kg7 and ...Bf5 soon. At the most
White can take a pawn: 15.Rd1 Kg7 16.Bxd5 Qxd5 17.Qxd5 cxd5 18.Rxd5 but Black has full compensation
based on the bishop pair. 18...Be6©) 14...Be6 (14...g6 15.Bxd5! Seizing the moment when Black finds himself
unable to take back with the queen. 15...cxd5 16.Be3 Kg7 17.Qf4²) 15.Bd2²
Position after: 15.Bd2²
Nothing dramatic is about to happen, and Black can keep a playable game with, if nothing else, ...Nb6 entering
484
an ending with a knight and bishop against two bishops, but there is no question that White is better; after all,
he has completed development.
12.Re1!
This is maximalism. White could obtain the bishop pair and a slight advantage in two ways, though neither
would really be anything to write home about as Black would be solid.
A) 12.Nxf6+ This looks a lot better than it is.
In reality White is not well placed to exploit either the bishop pair or the doubled f-pawns, and he can only
choose one. 12...Qxf6 13.Qxf6 gxf6
14.Nd4 (14.Bd2 0-0-0 15.Rfe1 Rg8„ Black gets some counterplay on the g-file and his central control is also
good. White is a long way from making effective use of the f4 square – which would be done by c4 and Ne2-
f4.) 14...0-0-0 15.Nxe6 fxe6 16.Bd2 Kc7² White is of course better, but the d5-knight is not trivial to
displace, and Black has in mind a scheme of development including ...e5, ...h5 and ...Bh6, which would fully
equalise.
B) 12.Nc5 Bxc5 13.Qxc5 Ne4 14.Qd4 Nef6
485
Position after: 14...Nef6
15.Bd2 (15.b3 0-0 16.Bb2 Bf5 17.Rac1 Re8 does not impress for White) 15...0-0 Now, while White remains
better, the contours of Black’s possible counterplay become clear; develop, wait for c4, then regroup on the
dark squares. For instance 16.Rfe1 Bg4 17.Ne5 Bf5 18.Bb3 Re8 19.c4 Nc7 20.Qxd8 Raxd8 21.Ba5 Rc8 22.f3
Na6².
12...Be7 13.Nc5 Bc8™
Position after: 13...Bc8™
14.Bxd5
486
The only way to cause Black serious problems with co-ordination (here, the immediate problem is how to
castle).
14...Qxd5
14...cxd5 15.Qa4+ Kf8 16.Nd3± Black will need 2 more tempi to even get his king safe, and White will use
those to play Be3-d4, cuing up a doubling of rooks or a Nfe5/Nf4 type swing.
15.Qe3
15.Qb4!? does not lead to an edge: 15...b6 16.Nd3 (16.Na6 0-0!=) 16...c5 17.Qc3 Bf5 18.Bf4 (18.Re5 Qd7
19.Qe1 Bxd3 20.cxd3 Nd5=) 18...Be4! 19.Nd2 0-0 20.Nxe4 Nxe4 21.Qe5 Qxe5 22.Nxe5 Ng5=.
15...Qd6
Position after: 15...Qd6
16.Nb3!
The only move that continues to derail Black’s castling attempts.

A) 16.Bd2 Nd5=
B) 16.Ne4 Nxe4 17.Qxe4 Be6 18.Nd4 Rd8 19.Nxe6 (19.Be3 Bd5 20.Qd3 Qg6!=) 19...Qxe6 20.Qxe6 fxe6=
16...Bf5!
16...Nd5 17.Qe2! (17.Qe4 lets Black off a bit after 17...c5! at which point we can terminate the analysis since
this is not good play for either side; suffice it to say Black is probably fine.) 17...Qe6 (17...c5? now loses
straightforwardly: 18.c4 Nb4 19.Bf4 Qe6 20.Qd2 Qf6 21.Rxe7++–) 18.Qf1 (18.Bd2 Qxe2 19.Rxe2 Nc7
20.Nfd4² although the bishop pair should not be underestimated!) 18...Qg6 (18...Qf6 19.Bd2 b5 20.c4±)
487
19.Nbd4 Nc7 20.Bd2 Ne6 21.Ne5± Black faces severe problems and he has still not gotten to castle.
17.Qe2
A) 17.Nfd4?! Bh7= White has lifted his surveillance of e1, so Black threatens to castle next for tactical
reasons. 18.g3 Kf8 19.Qe2 Re8 20.Bf4 Qd7„
B) 17.Bd2 Nd5 18.Qd4 Bxc2 19.Qxg7 0-0-0©
C) 17.c4 There are only so many constructive moves White can make that don’t simply hand over material for
nothing. 17...Bh7!
Position after: 17...Bh7!
18.Qc5! (18.Bd2 Ne4 19.Rad1 0-0=; 18.h3 Ne4„) 18...Qxc5 19.Nxc5 b6 White now has to follow an
extremely precise path to stop Black from just playing ...Ne4 and ...0-0. 20.Na6 Rd8 21.Bf4 Ne4 22.Ne5 Rc8
23.c5!²
D) 17.Nbd4 Be4 18.Nd2 (18.c4 0-0=) 18...Bh7 19.Nc4 Qc7 is cer tainly forcing, but it is not clear to me that
the things it forced were concessions.
17...Qe6
488
18.Bf4
18.Qd1 Rd8! seems to essentially equalise.
18...Qxe2 19.Rxe2 Kf8 20.Rae1²
Black’s development woes continue, though he is not losing material. To conclude, either this or 11...Be7
seems essentially playable, but Black is worse if White knows his stuff.
There is never an awful lot to say about ‘surprise weapons’ in a repertoire book, as both you and I know
they are less sound, or less equal, or less safe, than anything else.
These particular sidelines are no exception and I will confine myself to commentIng that they are all
perfectly playable, but against good play, there is no equality. A shame, considering the promise that 6...e5
seemed to have. The critical thing to do if you are playing this is to have a plan against White’s 10.0-0!
which must surely enter mainstream practice very soon.
489
Part IV:
4...Nd7 Main line
Chapter 1
An interesting attempt (11.a4!?)
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Ng5 Ngf6 6.Bd3 e6 7.N1f3 Bd6 8.Qe2 h6 9.Ne4 Nxe4 10.Qxe4
Qc7 11.a4!?
Chapter guide
Chapter 1 – An interesting attempt (11.a4!?)
1.e4 c6 2.d4 d5 3. Nc3 dxe4 4. Nxe4 Nd7 5.Ng5 Ngf6 6.Bd3 e6 7.Ngf3 Bd6 8.Qe2 h6 9.Ne4 Nxe4 10.Qxe4
Qc7 11.a4!?
1.1) 11... b6?!
1.2) 11...c5!
This is one of the most important of the lines I am genuinely recommending White take a closer look at in these
pages. It aims to play on the dark squares and also against the c8-bishop, without an excess of theoretical
baggage. Following the standard plan doesn’t really work for Black now, as we see in the first branch. Then in
1.2), we follow Black’s objectively best path, which nevertheless gives White scope for imaginative play, albeit
not to his advantage after Black takes on f3 at the correct moment.
1.1) 11... b6?!
1.e4 c6 2.Nc3 d5 3.d4 dxe4 4.Nxe4 Nd7 5.Ng5 Ngf6 6.Bd3 e6 7.N1f3 Bd6 8.Qe2 h6 9.Ne4 Nxe4 10.Qxe4
490
Qc7 11.a4!?
11...b6?!
Not a terrible move, and maybe suitable for occasional use, but as we shall see, I think White gets an advantage.
Nevertheless, we should investigate this move first, because if it can be played, then 11.a4 is not great.
Furthermore, we need to motivate the decision we shall ultimately take, which leaves the c8-bishop at home for
some considerable time.
A) 11...Nf6?! 12.Qh4 Now, the Karpovian ...Ke7 makes no sense (see by contrast the 10...Nf6 line in the
Appendix to this part) and a4 was reasonably useful, either in stopping Black’s fianchetto or in preparing b3,
Ba3. (12.Qe2?! would admit that a4 was a waste of time.) 12...Bd7 The only independent type of development
left for Black. (12...b6 transposes back to 11...b6 12.a5 Nf6) 13.0-0 0-0-0 14.Ne5² It seems to me that this
pawn sacrifice should not be accepted, but regardless of whether it is or not, White stands slightly better.
B) 11...c5 seems best and will be covered in the next branch.
12.a5 Nf6
12...b5 13.Bxb5² is initially a tempting sacrifice for Black, but it turns out that even with his queen in the
corner White can get away with returning only a pawn (either directly on c2 or else with Ne5) and stands better.
13.Qh4
13.Qe2?! As this is not actually a White book, and not my recommended counter for Black, I shall content
myself with commenting that after 13...Rb8! Black is able to maintain the balance with mathematical precision.
13...b5
491
13...Rb8 14.axb6 axb6 15.0-0 Nd5 16.Ne5²
14.Bd2!
14.0-0 a6 is totally equal since Black forces through ...c5 early and without making concessions in the centre.
14...Nd5
After other moves like ...a6 White is straightforwardly completing his bind with b4.
15.0-0
15.c3?! The more natural move order to obtain the standard dark-squared bind, but Black is on time to free his
position with two dynamic pawn breaks. 15...0-0 16.0-0 (16.b4 e5!=) 16...c5! To play this way, Black has to not
only be happy with the main line, but also see a tricky resource (18.Qe4 Nf6.) 17.dxc5 (17.Bxb5 c4„ is
tactically inconvenient for White) 17...Bxc5
492
18.Bxb5 (18.Qe4!? Nf6!! 19.Qxa8 Bb7 20.Bf4 Qe7 21.Qxf8+ Kxf8 22.Be2 The material is imbalanced,
leaving plenty of room for either side to outplay the other- there isn’t really a White edge here.) 18...Rb8 19.c4
(19.Qa4 a6³) 19...Be7! (19...e5 20.a6 and it turns out that none of Black’s queen-trapping motifs quite bring
success.) 20.Qe4 a6 21.Ba4 (21.cxd5 exd5 22.Qd3 Rxb5=) 21...Nf6 22.Qd4 e5!= Obtaining a variety of three-
results positions where Black regains one of his pawns and gets decent compensation for the other.
15...a6
15...0-0 16.Rfc1! (16.c3 c5 is the previous note.) 16...a6 (16...c5 17.dxc5 Bxc5 18.c4±) 17.c3 Qd8 (17...Nf4
18.Bxf4 Bxf4 19.Re1²; 17...Bb7 18.b4²) 18.Qxd8 Rxd8 19.b4
493
So White has got his bind but it is not a great version. 19...f6 20.Re1 g5 21.g3 Black might need to look in
places like this if 11...c5 doesn’t work out, but the position still looks slightly unpleasant.
16.c3 c5
Once more the only move to avoid the familiar b4 push.
17.dxc5 Bxc5
17...Qxc5 18.Qg4²
18.Rfc1!
Position after: 18.Rfc1!
18...Be7
A) 18...0-0 19.c4 bxc4 20.Rxc4± Black has to worry about Bxh6 ideas, and seems not to have a fully adequate
solution.
B) 18...Nf6 19.b4 Be7 20.c4 bxc4 21.Rxc4 Qd6 22.Rac1 0-0 23.Rd4 Nd5 24.Qe4²
19.Qg4 Bf6 20.c4 bxc4 21.Bxc4 Qd6 22.Bxd5 Qxd5 23.Bb4²
494
Position after: 23.Bb4²
Black continues to face development issues in a fairly open position. This is not even one of the cases where if
he survives the immediate attack he wins; he would have to additionally trade two pairs of major pieces before
that becomes even slightly true.
1.2) 11...c5!
Qc7 11.a4!? c5!

495
12.0-0 Nf6
12...c4?! 13.Be2 0-0 (13...Nf6?! 14.Qh4 b6 15.Ne5² White wins the bishop pair in a position where it matters a
fair amount.) 14.c3² It seems to me Black should not be happy with this, as his c4 pawn is quite vulnerable to
later Ne5 or Nd2 ideas.
13.Qh4 b6
Totally consistent play from Black. Now, he would be thrilled to see Bb5+, as this encourages ...Ke7 – in
principle not something Black minds doing at all, especially for free.
14.a5
14.dxc5 bxc5 15.Nd2 is an alternate way of playing, claiming that Black’s queenside structure is bad.
15...Nd5! (15...Bb7 16.Nc4±; 15...Bf4 16.Ne4! Be5 17.Rd1 a6 18.Nd2!²) 16.Nc4 Bf4 17.Bxf4 Qxf4
(17...Nxf4 18.Rfd1 0-0 19.Nd6!²) 18.Qxf4 Nxf4= It could be that White can somehow claim an advantage in
this endgame, but my further analysis has indicated that all the minor piece trades are okay for Black and he has
the b-file for counter-attack, so he should not be worried.
14...Bb7 15.b3
496
15...Rd8
Giving White the positional trumps he only vaguely wants, rather than the attack-on-uncastled-king which he
really does.
A) 15...Bxf3 16.gxf3 Ke7 looks suspicious due to 17.axb6 axb6 18.Rxa8 Rxa8 19.d5² and White gets to
wreck Black’s structure as well.
B) 15...Ke7!? This is the move that seems to want to be played, connecting rooks in an alternative way to
avoid either a queenside attack or Bxh6 tricks. I have not cared to drill down to an objective advantage for
White, but one should exist, and even if it doesn’t, White is having a lot more fun for little or no memory
work. (For similar reasons, later in this part we have avoided exclusively recommending 14.Re1 Nf6 15.Qh3
Bb4.) 16.Qh3 Rhd8 17.Bb2 Kf8 (It goes without saying that taking a pawn with 17...Bxf3 18.Qxf3 Bxh2+
19.Kh1 Bd6 is not clever, not least because of the a-file pin. 20.axb6 Qxb6 21.d5!±) 18.dxc5 bxc5 19.Rfe1
497
Position after: 19.Rfe1
B1) 19...Nd5 leads to a forced line: 20.Ne5 Nf4 21.Qg4 Nxd3 22.Nxd3 f6 23.Qxe6 Bxh2+ 24.Kh1 Bd6 at
the end of which the computer’s cheery evaluations for Black quickly give way to pessimism (the Qf5-h7
route, the a6-push and the weakness of c5 combine to be worth more than the bishop pair).
B2) 19...Bxf3 Neither of Black’s moves is all that convincing, which is why I don’t recommend 15...Ke7.
20.Qxf3 Bxh2+ 21.Kh1 Bf4© White has interesting and long-lasting compensation for a largely irrelevant
pawn. His improvements are easy to find whereas Black’s plans are artificial; one of the more obvious ones
is ...Kg8, ...Ne8, ...Bg5-f6 but by this time White could have a rook on c4, a queen on b7 or any of a number
of other ideas.
16.axb6 axb6 17.dxc5 bxc5
498
Black is still not castling next, is he? Well, yes and no. By contrast to 15...Ke7 Black still has that option, and
White must make some concession in order to take it away. If he does not, Black can play, in that order, ...Bxf3,
...Be5 and ...0-0.
18.Bd2
A) 18.Ra7 Ra8 19.Rxa8+ Bxa8 20.Rd1 Ke7 is now fine, because the rook trade was favourable and another
seems not far off. 21.Qh3 Rd8=
B) 18.Bb2 Bxf3 19.gxf3 Be5! The point, taking the sting out of White’s bishop pair after its most natural
deployment. 20.Bxe5 Qxe5 21.Rfe1 (21.Qg3 Qb8! 22.Ra5 Nh5©) 21...Qc3 22.Qa4+ Kf8 23.Qa5 Qxa5
24.Rxa5 Rc8= Black will develop with ...g5 and ...Kg7.
18...Bxf3!
Anyway! Based on a very curious and concrete idea where the Black queen, from its posts on the b-file,
simultaneously stops the Ba5 idea, prevents Bb5+, and attacks f4 and f3 respectively. Notably, if White’s
bishop were on c1 this might not be playable, due to ideas of Ba3 against ...Bf4.
In the interests of honesty, I would have indicated if I could see promising directions for White from here, but I
can’t, so it seems this precise continuation equalises.
A) 18...Ra8 19.Rxa8+ Bxa8 20.Ra1‚
B) 18...Qe7 is the engine suggestion, but it is a worse version: sooner or later Black must take on f3. 19.Rfe1
Bxf3 (19...0-0 20.Bxh6+– is terminal) 20.gxf3 Nd5 21.Qe4²
19.gxf3 Qb7
499
20.Qh3
A) 20.Kg2 g5 Black’s king trivially gets to safety. 21.Qa4+ Kf8 22.Qa5 Be7= To be followed with ...Kg7,
...Nd5 etc. The two bishops have some impact but Black’s king is not in real danger.
B) 20.Qa4+ Kf8 21.Qa5 Rb8= is similar.
20...Qb8!
Trying to draw by ‘perpetual bishop’... White can stop this, but can’t stop equality.
20...0-0 21.Kh1±
21.Ra4 Nd5 22.Re1 Kf8 23.Ree4 Nb4!=
500
Position after: 23...Nb4!=
The position is balanced – ‘equal’ would probably not be fair to say.
As hinted in the chapter introduction, this line deserves much closer attention from White players. It was
not easy to find the series of moves culminating in 18...Bxf3! which seem at the present time to give Black
equality, and I imagine it would be even harder over the board, so the reader is strongly advised to
memorise that much at least.
To get an idea of the dangers of the structure, I recommend a look at the game McShane-Howell, British
Championships 2017, which I was privileged to be able to comment upon live. Do not be put off by the fact
that it began 1...e6.
501
Chapter 2
The main line with 14...Nf6 (and a reason for rejecting it)
Qc7 11.0-0 b6 12.Qg4 Kf8 13.b3 Bb7
Chapter guide
Chapter 2 – The main line with 14...Nf6 (and a reason for rejecting it)
Qc7 11.0-0 b6 12.Qg4 Kf8 13.b3 Bb7
2.1) Minor lines and 14.Bb2 Nf6 15.Qh3 Nd5 16.g3 Nb4?!
2.2) 14.Bb2 Nf6 15.Qh3 Nd5 16.g3 c5!
2.3) 14.Re1 Nf6 15.Qh3 Bb4 16.Rd1
2.4) 14.Re1 Nf6 15.Qh3 Bb4 16.Re2
2.5) 14.Re1 Nf6 15.Qh3 Bb4 16.Bb2!?
We have reached a significant branch as far as White’s options are concerned. Early on in the theory of this
variation, 14.Bb2 was played automatically; after all White has just played 14.b3. In line 2.1) we take a look at
why and how this move can be successful; some very strong Black players have played in this misguided way,
not because they were stupid but because they lacked modern engine analysis. That being said, Black can
achieve equality without too much trouble after 16...c5, which is covered in line 2.2).
The major refinement occurred in 2013-4, when Indian GM Parimarjan Negi began publishing his 1.e4
repertoire series. After that, people began taking the move-order 14.Re1 much more seriously, to the point
where I now consider it the main line. However, I think the present work could be the first that has actively
502
made a point of acknowledging the main thrust of his work, which is the exchange sacrifice in 2.5).
I haven’t the heart to cut out 14...Nf6 from the book completely, because it’s not been refuted- White’s not
proved any sort of edge. Therefore, I have analysed the two ‘normal’ replies 16.Rd1 (2.3) and 16.Re2 (2.4),
before going into a fair amount of detail regarding what I don’t like about the Black position after Negi’s
16.Bb2. This detail, for what it is worth, is probably sufficient for you to make an informed choice in favour of
14...Nf6, if that’s what you choose.
2.1) Minor lines and 14.Bb2 Nf6 15.Qh3 Nd5 16.g3 Nb4?!
Qc7 11.0-0
This is the main line, played an order of magnitude more often than 11.a4 and so on put together.
11...b6 12.Qg4
12.Re1 Bb7 leads to few independent possibilities; one worth mentioning is 13.Ne5!?
A) 13...c5?! leads to some slightly worrying positions for Black after 14.Qe2: 14...0-0 15.Bxh6!? obliges
Black to liquidate into an ending where he is worse on pawn-islands grounds; also 14...cxd4 15.Nxf7 leads to
a strong attack.
B) 13...Bxe5! 14.dxe5 Nc5 15.Qg4 Nxd3 16.cxd3 0-0-0= Black will obtain good play on the d-file and long
diagonal.
12...Kf8!
503
Position after: 12...Kf8!
There are only two sensible ways to handle the threat to g7, and the other one, 12...g5 is terribly dangerous and
unreliable; see the Appendix to this part.
13.b3
13.Re1 Bb7 could be played first but avoiding the dark squared bishop fianchetto entirely does not seem clever.
14.Bd2 (14.c4 c5=) 14...Rd8 15.Rad1 c5 16.dxc5 bxc5= Svidler, P – Morozevich, A Mexico City 2007.
13...Bb7
Reaching, although it is not immediately obvious, a critical juncture for White.
14.Bb2
A less sophisticated move, and it seems that with some care Black equalises. 14.Re1 is examined later in this
chapter and the next.
14...Nf6 15.Qh3 Nd5
Black threatens Nf4, which would make White give up the wrong bishop for the knight. Now g3 is most
common, and understandably so.
16.g3
16.Ne5 Nf4 17.Qf3 Nxd3= for instance promises White absolutely nothing. Black will, in stereotyped fashion,
take on e5 when necessary, play ...Rad8, ...c5, ...Kg8-h7, look at his opponent and shrug...
16...Nb4?
504
Position after: 16...Nb4?
This line should be treated as a ‘what not to do’: this move has been somewhat popular, and is the most natural
in the position (as it seems to exchange off the d3-bishop) but we are going to show in an instructive way what
is wrong with it and why Black players should find other methods.
17.Be4
17.Rfd1 Nxd3 18.Rxd3 Rd8=
17...f5 18.Ng5
The most aggressive, and having investigated this in some depth I think Black should be worried.
18.Nh4?! Qf7 19.a3 Nd5 20.Bxd5 cxd5 21.Nf3©
505
Position after: 21.Nf3©
White›s long-term bind on the central dark squares is to some extent balanced out by the half-open c-file and
bishop pair.
18...Re8
18...Qe7?! 19.Nxe6+ Qxe6 20.Bxf5 Qf7
Position after: 20...Qf7
21.Rae1 This looks extremely worrying and cloud engines on high depth are pessimistic about Black’s chances.
506
(21.c4!? also led to rich play in Sutovsky – Antoniewski) 21...Re8 Maybe best, having examined the
alternatives, but 22.Re6 Rxe6 23.Bxe6 is simple and good. 23...Qe7 (23...Qf6 24.Bc8!± is a cute idea) 24.d5
Ke8 25.dxc6 Nxc6 26.Rd1‚ The next idea is Bc8 or Qh5+ and Black is in a lot of trouble.
19.a3!
19.Nxe6+?! was the only move I considered in my original analysis, because it seemed White had to do
something about his attacked pieces, but of course the counter-attacking text move is also valid. 19...Rxe6
20.Bxf5 Rf6 21.Rae1 Nd5
Black is holding in all lines:

A) 22.Bd7 Ne7 23.d5 (23.c4 Bc8„ shows another defensive idea) 23...cxd5 24.Bxf6 gxf6 25.Re2 d4„
B) 22.c4 leads also to exchange sacrifices and play of similar character: 22...Ne7 23.Be6 h5! 24.c5 (24.d5
cxd5 25.Bxf6 gxf6³) 24...bxc5 25.dxc5 Bxc5 26.Rd1 Bd6 27.Bxf6 gxf6„
19...Nd5
507
20.Bd3!
A good wrinkle to know, as this is maybe a bit less natural than Bg2.
20.Bg2 Kg8! 21.Nf3 h5 Black will save himself due to awkward position of the White queen. 22.Qh4 Be7
23.Ng5 Qd8 24.Bc1 Rh6 25.Re1 Bf6„ The pin is very annoying indeed.
20...Kg8™
20...c5 21.Rae1+– is terminal.
21.Nf3 c5™
Now too there isn’t really a choice as White was about to play c4-c5 himself.
22.dxc5
22.Rad1?! Qc6 lets too much develop along the long diagonal.
22...Bxc5 23.b4
508
23...Nf6
23...Bd6 24.Nd4 is much the same, with e6 looking extremely tender.
24.Bxf6!
Cleanest.
24.Ne5?! By far the engine favourite, and so we proceed down a fun line against it, but White can play a lot
more simply. 24...Bd6 25.Rae1 h5 26.Bb5!? (26.Nc4 h4 27.Nxd6 Qxd6 28.f3 hxg3∞) 26...Rd8! 27.Bc4
509
27...Bd5! 28.Bxd5 Nxd5 29.Nf3 (29.Ng6 highlights the key reason why 26...Rd8 is picked over 26...Rc8: after
29...Rh6 30.Rxe6 Qf7 the rook defends the bishop!; 29.c4 Bxe5 30.Rxe5 Qxc4= seems solid for Black)
29...Qxc2 30.Rxe6 Qxb2 31.Qxf5 Qb3!„ White will regain his piece with tricks, but have no win as such, and
there will follow a long ending with queen and knight each and equal pawns...
24...Bxf3
24...gxf6 25.Be2²
25.Bb2!
25.bxc5 gxf6 26.cxb6 Qxb6 and similar lines might look bad, but in fact Black has passable coordination here
once he plays ...Kg7 and possibly ...Rb8.
25...Bd6 26.Rfe1²
White has found a simple path to a sizeable advantage.
Position after: 26.Rfe1²
2.2) 14.Bb2 Nf6 15.Qh3 Nd5 16.g3 c5!
Qc7 11.0-0 b6 12.Qg4 Kf8 13.b3 Bb7 14.Bb2 Nf6 15.Qh3 Nd5 16.g3 c5
510
Deservedly the main line, and the move that most clearly highlights what makes 14.Bb2 less critical.
17.dxc5
Making the most of the little wrinkles of the position.

17.c4 is not as good here as in the analogous line with 14.Re1 because there is no option of Be4, but the tactics
repay investigation. 17...Nb4 18.Be2 Qc6!
511
Black lines up various ideas along the long diagonal and seems completely fine. White is now well advised to
take the queens off.
A) 19.Rad1?! This move only barely keeps equality. 19...cxd4! 20.Bxd4 (20.Rxd4? Be5 21.Rd2 Bxb2 22.Rxb2
Nd3!µ) 20...Qe4 21.Rd2 (21.Be3 Nxa2! 22.Rxd6 Nc3 is another of Black’s tricks.) 21...Nc6 22.Bc3 Ne5!=
Position after: 22...Ne5!=
The position is practically much easier to play as Black, even if the machine equalises with White.
B) 19.Qg4 h5! 20.Qg5 (Drastic action was possible: 20.Qxg7+! Kxg7 21.d5+ f6 22.dxc6 Nxc6= This equal
ending was reached in Wei, Y – Barbosa, O Tagaytay City 2013.) 20...cxd4 21.Bxd4 e5! 22.Bc3 (22.Bxe5?
Bxe5 23.Qxe5 Re8µ) 22...h4! Threatening nothing more sophisticated than ...hxg3, and if hxg3 then Qxf3
with mate. 23.Rfd1
512
Position after: 23.Rfd1
Setting up Rxd6, and it transpires that White must play this, even as an exchange sacrifice. After 23...Nc2
24.Rxd6™ (24.Rac1 hxg3 25.Qxg3 Nd4–+) 24...Qxd6 25.Rd1 (25.Bxe5 Qh6!³) 25...Qe7 26.Qxe5 (26.Bxe5
Qxg5 27.Nxg5 Re8³) 26...f6 27.Qxe7+ Kxe7 28.Nxh4 Be4© the position is in a state of dynamic equality;
either side could outplay the other.
17...Bxc5
17...Qxc5?! This is an additional option which could be investigated. My opinion is that after the forced (on
Black’s side) line 18.Be4!
513
Position after: 18.Be4!
18...Nf4! (18...Rb8 19.Rfe1 Nf6 20.Bxb7 Rxb7 21.Rad1±) 19.Bxg7+ Kxg7 20.Qg4+ Kf8 21.Bxb7 Rb8 22.Be4
h5 23.Qh4 Ne2+ 24.Kh1 Be7 25.Qh3 White stands better. This is trivial in case Black does not retrieve his
pawn...
Position after: 25.Qh3
...so he must continue with ...Bf6. 25...Bf6! (25...h4 26.Rae1 Nc3 27.gxh4²) 26.Rae1 Nc3 27.Nd2! Nxa2
28.f4² White has better coordination and hence attacking chances, but it could happen that Black finds a precise
way to reach definitive equality.
18.Ne5
18.Be4N Here too White can do this, but the subtlety is that Black has defended the b7 bishop now. 18...Nf4
19.Qg4 Nh3+!
514
Position after: 19...Nh3+!
20.Kg2 (20.Kh1 f5!³) 20...Ng5! 21.Nxg5 hxg5 22.Qf3 Bxe4 23.Qxe4 Re8= Black will continue with, if
permitted, ...e5 and ...f6 to blunt his opponent’s bishop, before thinking about the positional desirability of a
queen swap.
18...Nf6
19.Qh4!?
515
19.Rae1 is more heavily played but probably less critical. The response is by now well known to Black
theoreticians. 19...Rd8
A) 20.Bg6?! A surprising number of White players have chosen this, despite the high-level game which
continued 20...Ba8! and suddenly the long diagonal is critically weak and White has to scramble. 21.Bh5
Nxh5 22.Qxh5 Qb7 23.Re4 Kg8 24.Rfe1 Rd2µ Sutovsky, E – Svidler, P Novi Sad 2009.
B) 20.Qh4 Still the best move, although it doesn’t ask as many questions as on the previous turn. 20...Rd5
21.b4 Bd6 22.Qc4 Ke7 23.Re2 Rc8 24.Qxc7+ Bxc7 25.Rfe1 Ganguly, S – Rasmussen, A Copenhagen 2010.
19...Kg8!
The best order.

19...Rd8?! 20.Qf4?! The queen contributes little here other than setting up an extremely cheap trick. (20.Rad1
Rd5 21.Rfe1²) 20...Kg8 21.b4 Bd6 Kokarev, D – Berczes, D Rijeka 2010.
20.Rad1
20...h5!=
Black has good play stemming from his kingside ideas like ...Ng4 or ...Be7.
2.3) 14.Re1 Nf6 15.Qh3 Bb4 16.Rd1
Qc7 11.0-0 b6 12.Qg4 Kf8 13.b3 Bb7 14.Re1!? Nf6
516
15.Qh3
White almost never moves the queen to h4 in these lines, because it is vitally important to recapture on f3 with
a piece if ...Bxf3 occurs. 15.Qh4 c5 16.dxc5 (16.Ne5 cxd4 17.Qxd4 Qc5 18.Bb2 Qxd4 19.Bxd4 Ke7=; 16.Bb2
Bxf3 17.gxf3 Rd8!?„) 16...Qxc5 17.Be3 Qh5=
15...Bb4
A) 15...Rd8 may now be a slightly inaccurate move order due to 16.c4 when after ...c5 dxc5 Black does not
have the knight recapture that he would like.
B) 15...Nd5?! 16.c4 was Negi’s point; now White is ready to take back on f4 with a bishop he doesn’t mind
losing, and then dominate the board with a c5 push. 16...Nf4 17.Bxf4 Bxf4 18.c5²
517
Position after: 18.c5²
Black nevertheless went on to outplay his opponent in Jackson, J – Hawkins, J Hastings 2014.
16.Rd1?!
This is a bit compliant in that it moves off the e6-pawn. What white would like with this move is for his
opponent to considerately open the d-file for him. We examine the more challenging alternatives later: Re2 and
Bb2!?.
16.Bd2?! is even more tame. 16...Bxd2 17.Nxd2 Rd8 18.Nf3

518
A) 18...h5!? White’s problem now is that he doesn’t have a way to easily defend the d-pawn. Even though it is
not attacked. This is because Black wants to play ...h4!? next and maybe even develop his rook to the 5th
rank... 19.Rad1 (19.Qh4 c5³) 19...h4„ The position is complicated, and worth a look if you have time and
want a way to try and win rather than equalise after the (inferior) 16.Bd2, but there is no point my going into
it.
B) 18...c5 19.dxc5 (19.Ne5 g5!„ Black plays ...Kg7 next and gets coordinated.) 19...bxc5 20.Rad1 g6 21.Ne5
Kg7= White still has his c4-motif going on and better central control, but Black has gotten coordinated and
has the better bishop.
16...Bc3 17.Rb1
Position after: 17.Rb1
17...Kg8!
17...c5?! This combines badly with ...Bb4, even though Black has managed to get an extra concession out of his
opponent. 18.dxc5 bxc5 19.Bc4²
519
Position after: 19.Bc4²
White has excellent light-squared control, which can be easily augmented by a well chosen discovery along the
third rank. Also, if Black is not careful Qg3 will force the queens off, leading to a slightly worse ending for him
based on pawn islands. This is why the computer wants to play ...Ne4, but there then follows Ng5 and Black is
getting strung up...
18.Bc4!
A) 18.Ba3 is the only other move that vaguely makes sense from a development perspective, but this enables
Black to gain tempi off it, e.g. 18...c5 19.dxc5 bxc5 20.Bc4 Qa5 21.Bb2 Bxb2 22.Rxb2 Rd8=
B) 18.b4?! cedes the a-file to all intents and purposes: 18...a5=.
18...Rd8
520
White’s position does look nice but he lacks a next move.
19.Ba3
A) 19.Rd3 Bb4 just suffocates the c4-bishop, causing it to sacrifice itself on e6, but after 20.Ne5 b5= there
isn’t a win, only perpetual check.
B) 19.Ne5!? is enterprising. 19...Bxd4 20.Bf4 (20.Nxf7 Kxf7 21.Qxe6+ Kf8µ) 20...Ne4 The counter-attack
arrives just in time. (20...Nd5 21.Rxd4 Nxf4 22.Rxd8+ Qxd8 23.Qf3²) 21.Rxd4 Rxd4 22.Ng6 Qd8 23.Nxh8
Rd1+ ...Kxh8 should also be fine here. 24.Rxd1 Qxd1+ 25.Bf1 Qe1=
19...b5!?
Leading to a collection of elegant drawing lines, but the alternative is fine too and simpler.
19...Ne4 A human continuation, because as often as possible we should try to play moves that we understand!
20.Bxe6 (20.Ne5!? Bxd4 21.Rxd4 Rxd4 22.Nxf7 Rxc4! 23.Nxh8 Rd4=) 20...Ng5 21.Nxg5 hxg5
521
Position after: 21...hxg5
22.Bxf7+™ 22...Kxf7 23.Qf5+ Kg8 24.Qe6+ Qf7™ 25.Qe3 Qf4 26.Qe6+=
20.Ne5 b4!
21.Rd3
21.Bxe6 fxe6 22.Qxe6+ Kh7=
522
21...bxa3! 22.Rxc3!
22.Bxe6 fxe6 23.Qxe6+ Kh7 is still just a draw.
22...Rxd4 23.Re1 c5!
The bishop has to play a role in defense, and ...Bc8 looks terribly passive.
24.Rce3!?
24.Nxf7 Kxf7 25.Qxe6+ Kf8 26.Rce3 Bd5 27.Qf5 Bxc4 28.Re8+ Kf7 29.Rxh8 Be2!
With a supposed 0.00 position that nevertheless looks very unbalanced, and not in a bad way for Black.
24...Bd5 25.Bd3
25.c3 Rd2³
25...g5
523
White has reasonable short-term compensation, which means it only makes sense to examine the forcing
continuations. Right here, the motif White needs to play on is trapping the d4-rook.
26.g3
26.f3?! Kg7 27.c3 Rf4 28.g3 (28.Qg3 is the engine suggestion, but White’s play is visibly petering out, for
instance 28...Rd8 29.Qf2 Be4!³) 28...Rxf3! 29.Nxf3 Bxf3 30.Be2 Bb7‚ With two pawns and good pressure
against White’s king in return for the exchange.
26...Ba8! 27.Nxf7™
27.c3 Rd8³
27...Qxf7 28.Rxe6 Qb7 29.f3 Qxf3 30.Rf1!=
524
Position after: 30.Rf1!=
Forcing Black to give a perpetual check, as the f1-rook makes use of a ‘peekaboo’ motif...
2.4) 14.Re1 Nf6 15.Qh3 Bb4 16.Re2
Qc7 11.0-0 b6 12.Qg4 Kf8 13.b3 Bb7 14.Re1!? Nf6 15.Qh3 Bb4 16.Re2
This was chosen in the only practical example to date, but it does not set too many problems for the prepared
525
player.
16...Bc3
Also playable is 16...Rd8, but for space reasons I omit the analysis of that.
17.Rb1 c5
Now this is best, since White is not in a position to use the opening d-file.
18.dxc5
18.Ng5!? is flashy, but easily enough defused:

A) 18...Rd8 19.Bf4!? Qd7 (19...Qxf4 20.Nxf7) 20.dxc5 bxc5 21.Be5 Nd5=
B) 18...Bxd4 19.Bg6
Position after: 19.Bg6
19...hxg5! 20.Qxh8+ Ke7 21.Qxg7 Qc6 22.Qxf7+ Kd8 23.Be4 (23.Qf8+=) 23...Nxe4 24.Be3 Nc3=
18...bxc5 19.Bc4
19.Bd2?! Bxd2 20.Rxd2 Rd8 is as usual nothing: Black may continue ...g6 and ...Kg7.
19...Rd8!?N
This move makes loads of sense as White does not have an immediately appealing way to stop the back-rank
threat.
19...Nd5 was played in the practical example, and can lead to some thrilling chess.
526
A) 20.Bd2 Nf4=
B) 20.Bxd5 Bxd5 and now by contrast to the case after 16.Rd1 Black is the one dictating tempo, e.g. 21.Bb2
Bxb2 22.Rxb2 Qf4 23.Ne1 c4=
C) 20.Ng5 Bd4
C1) 21.Bd2!?N 21...g6! 21...Nf4 meets a deep tactical refutation which I shall leave you to find yourself.
22.Nxe6+ (22.Nf3 Nf4=) 22...fxe6 23.Rbe1 Qf7! 24.c3 Bf6 25.Rxe6 Rd8 White has full compensation but
not more.
C2) 21.Bxd5 Bxd5 22.c3
527
22...Bf6 A very obvious improvement. (22...Be5? 23.Nxf7!± won a pawn in Sethuraman, S – Rasmussen, A
Hastings 2017) 23.Ne4 (23.c4 Bb7 24.Ne4 is the kind of thing I tend not to like, but concretely here there is
24...Rd8! and Black is doing at least fine) 23...Rd8! 24.Ba3 (24.Nxf6?? Be4! is a required tactic without
which Black’s construct isn’t valid.) 24...Bxe4 25.Rxe4 Kg8= Black gets coordinated with e.g. ...Kh7 and
...Rd5; the weakness of the c-pawn is compensated for by the awkwardness of White’s bishop, as long as
Black takes care to keep the queens on.
D) 20.Ne1!?N is a machine suggestion, of course. I simply cannot resist giving a line now. 20...Qa5 Black
takes care to avoid setups where White gets in a c4 blockade. (20...Bf6 21.Bxd5 Bxd5 22.c4 Bb7 23.Bb2² The
machine proudly proclaims equality here, but for me White is somewhat better already.) 21.Qg3! To attack
the g7-square, and also... the b8-square?! (21.Nd3 Qxa2µ; 21.Bxd5 Bxd5 22.Bb2 Bxb2 23.Rxb2 c4= provides
another illustration of a recurrent theme.) 21...Rc8!
Position after: 21...Rc8!
Neither of the captures works. 22.Bb2!? (22.a4 Ne7!? is an inspired re- routing of the knight: 23.Nd3 Kg8
24.Qh3 Bd5! 25.Bxd5 Nxd5= and White may have nothing better than Rxe6 forcing a draw.) 22...Bxb2
23.Bxd5 Bxd5 24.Rxb2 c4! Black absolutely has to stop c4, which brings about a crushing bind. (24...Qc7
25.c4² leads to endings where White wins c5.) 25.bxc4 Bxc4 26.Rb8 This is critical and what I was referring
to when I said Qg3 attacked the b8-square! 26...Rxb8 27.Qxb8+ Ke7 28.Qb7+ Kf6 29.Re3 (29.Re4 is
similar.) 29...Rd8!„
528
Position after: 29...Rd8!„
Black isn’t bothered about Rf3+, and calmly begins his defence with ...Rd5. The position is objectively level
and if Black can survive the attack, he wins.
20.Ne1
A) 20.Be3 Nd5=
B) 20.Ba3 Qa5 21.Ng5 Bd5!=
20...Be5
529
21.Ba3!?
21.Nd3 Be4 defines the position a bit too much, and Black can operate comfortably. 22.Be3 (22.Ba3 Bxd3
23.Bxd3 Nd5! 24.Ree1 Nb4 25.Be4 Qa5=) 22...Bd6 23.Qh4 Kg8=
21...Be4 22.Bd3 Bb7=
The computer believes this is equal and that White should play Bc4, repeating moves. I see no reason to
disagree.
2.5) 14.Re1 Nf6 15.Qh3 Bb4 16.Bb2!?
Qc7 11.0-0 b6 12.Qg4 Kf8 13.b3 Bb7 14.Re1!? Nf6 15.Qh3 Bb4 16.Bb2!? Bxe1 17.Rxe1 Rd8
At least on a human level, Black has to start thinking about ...c5 and ...Rxd3, then ...Kg8-h7 and digging his
rook out of the corner. Any of the other ‘amorphous’ moves he makes do not materially change the position, but
in the meantime it is harder for him to improve his position than for his opponent.
18.Ne5
Most natural and played in my source game. Here White’s choice is slightly limited by the Black idea of ...Qf4-
g4, but there is still a major unforcing option which occurs to one after a little thought.
A) 18.Qh4N 18...Kg8 19.c4!? (19.Ne5 transposes to the text) 19...Qe7
530
Black contemplates ...g5 and will meet Ne5 with ...c5. This position is impossible to analyse, though by the
same token I would be shocked if White could prove an advantage. Here I shall leave you to your own
devices, since even if I analysed 100 such lines, within a week there would be another 100 on people’s
computers.
B) 18.c4?! Qf4!³
18...Kg8
18...c5? is inadvisable: 19.Bg6+–.
19.Qh4
White will realise sooner or later that Qh4 is indispensable, whereas Re3 may not be. 19.Re3 c5 20.dxc5 bxc5
21.Qg3 (21.Rg3 Be4=; Note that 21.Qh4 transposes to the text.) 21...Nd5 22.Rf3 f6„
19...c5 20.dxc5 bxc5
531
21.Re3
21.Nc4!? can probably be neutralised. 21...Qe7 22.f4! Nd5 23.Qg4!? (23.Qf2 Nb4=; 23.Qg3 f6 24.Ne3 Nxe3
25.Qxe3 Kf7 26.Ba3! White’s pressure will gain pace with moves like f5 or Bc4 so now is a great time to
change the game and play: 26...Rxd3!=) 23...Qf8 Black would really prefer not to play this, but he is cueing up
...Nb4 and he has a counter-trick in the forced line that follows. 24.f5 (24.a3 Nb6= gets ready to simplify, with
an exchange sacrifice if necessary.) 24...exf5 25.Bxf5 Bc8! 26.Bxc8 Rxc8 27.Bxg7
Position after: 27.Bxg7
532
27...h5! This is the counter-trick. 28.Qd7 Rd8 29.Bxf8 Rxd7 30.Bxc5 Kh7© Black should not be worse and he
position has been clarified somewhat.
21...g5!N
Not the only move, but has the advantage of being forcing, and I think I have found concrete equality for Black.
21...Rxd3? was a terrible decision by me in the blitz game curtains-batmanking, ICC 2017. 22.Nxd3 Nd5
23.Rg3+–
22.Qa4
A) 22.Qc4 Nd5 23.Rg3! (23.Re1 Qa5!³) 23...f6„
Position after: 23...f6„
In this and many similar positions, Black plays with fire, although I’ve still not seen a White path to an
advantage. 24.f4 (24.Ng6 Rh7 25.h4 Rg7 26.hxg5 hxg5 27.Qg4 Qf7©) 24...fxe5 25.Bxe5 Qa5 26.c3 Qa3
27.fxg5 Qc1+ 28.Bf1 hxg5 29.Bxh8 (29.Qg4 Kf7!!³) 29...Kxh8 30.Rh3+ Kg7 31.Qe4 Ba6!= There isn’t a
mate!
533
Position after: 31...Ba6!=
B) 22.Qh3 Rd4! 23.Re1 Kg7 24.f4 gxf4 25.Bxd4 cxd4 26.Qh4 Nd5= The defensive moves ...f5 and ...Qe7 are
just close enough at hand.
22...Nd5 23.Re1 Nb4
As normal, Black tries for this exchange.
24.a3
534
Trying to use the Nxd3-tempo.
24.Be2 can be neutralised by: 24...Bc6!. Much more important than snatching material, which is what the
computer wants. 25.Nxc6 Qxc6 26.Qxc6 Nxc6 27.Bxh8 Kxh8 28.Rd1 Rxd1+ 29.Bxd1=
Position after: 29.Bxd1=
This endgame should be drawn with a little care.
24...Nxd3 25.Nxd3 Bc6 26.Qc4 Bd5 27.Qc3 Rh7 28.Nxc5 e5=
Position after: 28...e5=
535
It appears that the system with 14.Bb2 deserves to be replaced by 14.Re1 – it is simply too easy for Black
to equalise as in line 2.2). I recommend taking a brief look at that analysis, and not more.
Regarding 14.Re1, just like the Short system and Two Knights system earlier in the book, I have offered
Black a choice of options.
The 14...Nf6 repertoire would have been trivial to write (note 19...Ne4! in line 2.3) and 19...Rd8 in line
2.4, both of which equalise without problems) if not for the final subchapter, and here, for maybe the first
time in my life, both I and my computer are stumped. I have tried to offer some guidance, and analyse those
lines which can be worked out to a forcing conclusion, but while White has no objective advantage, the
positions are amorphous enough that in order to play them confidently you should have first done your own
homework too.
536
Chapter 3
The main line with Ding’s improvement 14...Rd8!
Qc7 11.0-0 b6 12.Qg4 Kf8 13.b3 Bb7 14.Re1 Rd8!
Chapter guide
Chapter 3 – The main line with Ding’s improvement 14...Rd8!
Qc7 11.0-0 b6 12.Qg4 Kf8 13. b3 Bb7 14. Re1 Rd8! 15.Bb2 Nf6 16.Qh3 Nd5 17.g3 Bb4 18.c4 Nf6
3.1) 19.Re3 and 19.Red1 minor lines
3.2) 19.Red1 c5!?: main line with 26...Rh5!?
3.3) 19.Red1 Be7
3.4) 19.Red1 Be7: main line with 23.d5!
Notwithstanding that 14...Nf6 has not been refuted or (objectively) even dented by recent developments, I find
a much more reliable choice after 14.Re1 to be Ding Liren’s 14...Rd8, which I have taken great trouble to
analyse here in great detail.
The main line proceeds 15.Bb2 Nf6 16.Qh3 Nd5 17.g3 Bb4 18.c4 Nf6 and now White has to choose between
rook moves. It is again possible to sacrifice an ex-change, but that is less convincing here than in the previous
chapter.
The main move is 19.Red1, examined for most of this chapter after some initial notes about alternatives in 3.1).
Now Black has two continuations, which are sty-listically quite distinct, and on which I have spent equal
amounts of effort.
537
You might feel that I could have streamlined this book so much by giving Black fewer alternatives, or that the
mere presence of multiple choices indicates some lack of faith in my own analysis. In reality, neither is true. I
feel a sense of duty to provide different styles of player with weapons that they feel comfortable using. With
that in mind, 19...c5!? caters to the mess-loving crowd. Material balance is thrown out of the window and while
the exchange sacrifice of 3.1) is likely not completely correct, it is nevertheless interesting, and Black’s play
can easily be improved as in 3.2).
As always, I give a solid alternative, which in this case is the novelty 19...Be7. Now White’s critical move is
20.Ne5, and play proceeds to the final diagram above, where White can choose between a violent perpetual
check line (3.3) and a drawn endgame (3.4.)
3.1) 19.Re3 and 19.Red1 minor lines
Qc7 11.0-0 b6 12.Qg4 Kf8 13.b3 Bb7 14.Re1 Rd8!?
Position after: 14...Rd8!?
15.Bb2
15.c4 c5= is fully equal since Black has the desired recapture on c5 available.
15...Nf6 16.Qh3
16.Qh4 has another weakness in addition to that already mentioned: it cues up two tempo gainers (...Be7 or
...g5) compared to one for Qh3 (...Nf4). 16...Nd5 17.Ne5 g5!=
16...Nd5 17.g3 Bb4!? 18.c4
18.Ne5!? should be considered, if we have learned anything at all from the last chapter. 18...Bxe1 19.Rxe1 Re8
538
However, to my mind, in this version Black is doing better, since White’s queen lacks the same degree of
control and ...c5 is harder to stop. For instance: 20.Qh4 (20.c4 Nf6 is the same as the 19.Ne5 note.) 20...Kg8
21.a3 Trying to get by without c4, but this seems likely to be unsuccessful. 21...Nf6 22.Qf4 c5„
18...Nf6 19.Red1
A) 19.Re3 I can’t think of any a priori reason why the rook should be worse here than on d1, except that it is
more vulnerable to tempo gainers. There are 3 relevant ones: ...Ng4, ...Nxd5, and ...cxd4. 19...h5! 20.Ng5
White can of course play other moves, but the knight moves are the main ones which scream out to be played
against ...h5. (20.Ne5 c5 21.d5 exd5 22.cxd5 Nxd5!„ This comes with tempo, which is important.) 20...c5
539
21.Bg6!? As is almost inevitable for such lines, the result is a perpetual check with correct play. (The same
goes for 21.Qh4 Bd2; but not for White’s thematic reply 21.d5, which fails since when Black recaptures on f6
it will gain tempo.) 21...Qc6! 22.d5! (22.Rf3 Rd7 leaves all White’s stuff hanging and he has no trick)
22...exd5 23.Nxf7 d4 24.f3! (24.Be4!? Qd7!³) 24...dxe3 25.Nxd8 Qxf3 26.Nxb7 Qf2+ 27.Kh1 Ng4 There
was also ...Ke7 to play on, but chances are balanced anyway. 28.Qh4™ 28...Qf3+=.
B) 19.Ne5 Bxe1 20.Rxe1 Re8
21.c5 (21.b4 intending to meet with the thematic capture followed by b5, is also possible but then Black can
play the equally thematic 21...b5 and obtain the same structure anyway.) 21...Nd7 22.Nxd7+ Qxd7 Here
White has some compensation but he will have to struggle to show that it’s sufficient.
19...c5!?
540
Position after: 19...c5!?
This line is probably theoretically more valuable than 17...c5, which has been relegated to the final chapter. I
have been unable to form a definite assessment here despite long analysis, and Black has some interesting
directions to research, but the feeling is nevertheless that White should be better *somehow*. To play either,
Black needs to be comfortable with having doubled f-pawns, and nothing else, in front of his king (in exchange
for the bishop pair and some material). My main recommendation is therefore 19...Be7, as in the rest of this
chapter.
20.d5!
This is of course again most critical in response to ...c5.

20.Be2 cxd4 21.Bxd4 Ke7! does not really give Black much trouble: the plan now is to double on the d-file and
trade rooks when possible.
20...exd5 21.cxd5
21.Bxf6 makes limited sense in this context, as White’s dark-squared bishop is hugely important.
21...Rxd5
Black’s next idea is of course ...Qc6 followed by some drama on the long diagonal. White has two sensible
choices: redeploy the queen, or reinforce the diagonal immediately.
Against 21...Bxd5?! White goes for an ending with 22.Qf5 (22.Nh4 Bb7 is another story, and my view is that
it’s not as critical) 22...Be6 23.Qe5
541
This has been extensively analysed by cloud engines on high depth and the conclusions are extremely
discouraging for Black, whose Bb4 plays no part in proceedings.
22.Qf1!
Now Black has to use his immediate tactics in a highly concrete way in order to obtain a playable position. We
take it as an article of faith that if the move pair a3 Ba5 occurs, Black’s dark-squared bishop will be well and
truly sidelines, so if White is about to play a3, Black must play ...c4.
22.Be2
542
Recommended in the literature, and the idea has a certain appeal. White wants to trade a pair of rooks and get
his other rook to d7. However I do not think this claims an edge.
A) 22...c4?! is one idea, with the point being to actually avoid the rook swap, but perfectly natural play
suffices for a white edge. 23.Bxf6! (23.bxc4?! Ra5!∞) 23...gxf6 24.Bxc4 Rd6 25.Qh5±
B) 22...Kg8?! is what people have overwhelmingly looked at – this is the right idea but needs a bit of
sophistry adding. 23.Bxf6 gxf6 24.Rxd5 Bxd5 25.Qf5² Black’s position does not look encouraging,
particularly after his light-squared bishop is traded off for one of White’s minor pieces.
C) 22...Qe7!
C1) 23.Rxd5 is the initial machine suggestion, but it calms down a bit after natural moves. 23...Nxd5 24.Bc4
Bc3! 25.Bxc3 Nxc3 26.Re1 Ne4=
Position after: 26...Ne4=
I did further analysis here, but suffice to say White’s tactics fizzle out after Black meets Qf5 or Qg4 with
...g6 and ...g5 respectively, with ...Kg7 to follow. Do note also that Nd2 is met by ...Ng5!
C2) 23.Qf1 Nevertheless. 23...Rxd1 24.Rxd1 c4!
543
Carving out the c5 square and thus a role for the dark squared bishop. 25.bxc4 Kg8!= With this precise,
machine-esque continuation I be lieve Black nevertheless equalises. It will be difficult to prevent ...Kh7 and
...Rd8.
22...Qc6!
23.Be2 Rf5! 24.Rd3!
544
24.Rd8+ Ke7! 25.Rxh8 (25.Rd3 Rd8³) 25...Rxf3 With a choice of perpetual check lines for White.
24...c4! 25.bxc4 Kg8!
Black seems to be getting organised. But it makes sense to continue the line, because there is still forcing play.
26.Qh3!
We are going to declare a late branch here, because the position is quite critical. 26.Qb1?! is beautifully
artificial, even if the idea of setting up two x-ray attacks on b4 and f5 is somewhat original and reminiscent of
the children’s game ‘connect 4’. It is met by 26...g6= ; Black plays ...Bc5 next and it becomes clear that the
attack on b4 was not real.
26...g6?!
Position after: 26...g6?!
Almost recommendable. It is immensely frustrating that this runs into a move-order wrinkle. At first I
considered that it does not matter whether White checks on d8 before playing g4 to alleviate the long-diagonal
pressure, but then an obvious diference appeared. Nevertheless, the line is instructive as a demonstration of
Black’s exchange sacrifice possibilities. For 26...Rh5!? – see the next subchapter.
27.g4!
27.Rd8+? Initially I analysed only this ‘order’. 27...Bf8 28.g4™ 28...Ra5!
545
Position after: 28...Ra5!
Now White would like to make it so that Black’s capture of the a-pawn does not gain tempo.
A) 29.Qh4 Ne8!! 30.Bxh8 Kxh8© Black›s position rests on the presence of tricks like ...Re5. 31.Rad1
(31.Qg3 Bd6) 31...Re5 32.Kf1 Rxe2 33.Nd4 Qh1+=
B) 29.Rad1? Rxa2 30.R1d6 Qxd6 31.Rxd6 Rxb2 32.Rxf6 Rxe2=
C) 29.a3?! h5 (There is also 29...Kh7). 30.Rad1 White does not have infinite time to unravel; Black wanted
...Rh7. 30...Kg7! Threatening ...hxg4, and arriving by force at a messy position a piece down. 31.Qg3 Bxa3!
Position after: 31...Bxa3!
546
The reason to prefer 29.a4 becomes clear! 32.Bxf6+! (32.Rxh8 additionally gives Black 32...Bxb2∞)
32...Qxf6 33.Rxh8 Kxh8 34.Qb8+ Kg7 35.Qxb7 hxg4 36.Nd2 Qf4∞
D) 29.a4! Note that on a4 this pawn is not attacked by the f8-bishop. 29...Kh7 (29...h5 30.Rad1 now doesn’t
work since Black never has ...Bxa3 with tempo!) 30.g5! (After 30.Bc3 Bg7 31.Rxh8+ Kxh8 32.Bxa5 bxa5
Position after: 32...bxa5
Black gets excellent colour-complex based compensation.) 30...Nh5! Now there follow a series of only moves
and then a choice of perpetual-check lines. 31.Bxh8 Nf4 32.Qd7 Nxe2+ 33.Kf1 Qxf3 34.Rxf8 Bd5!!=
27...Ra5
27...Rf4 28.Rd8+ Bf8 29.Rad1!²
547
Position after: 29.Rad1!²
This is the problem; concretely none of the things Black can do about the impending g5 thrust are any good,
and he needs machine precision in order to stay only slightly worse.
28.c5!
Position after: 28.c5!
Black is not the only one who can sacrifice his c-pawn! Here White makes use of some kind of Novotny theme.
28...Rxc5 29.Rd8+ Kg7 30.Rxh8 Kxh8 31.Qxh6+ Kg8 32.Rd1 Bc3 33.Ba3!
548
33.Rd8+ Ne8 34.Qe3 Bxb2 35.Rxe8+ Kg7 36.Qe7
This looks very scary, but the computer doesn’t even blink, and produces 36...Rc1+ 37.Kg2 Ba3=.
33...Nd7! 34.Bxc5 bxc5²
Black has considerable compensation
Position after: 34...bxc5²
for the exchange, and plenty of scope to outplay his opponent and win, but objectively it is not enough. Hence
549
we turn in the next subchapter to a late improvement of his.
3.2) 19.Red1 c5!? : main line with 26...Rh5!?
Qc7 11.0-0 b6 12.Qg4 Kf8 13.b3 Bb7 14.Re1 Rd8!? 15.Bb2 Nf6 16.Qh3 Nd5 17.g3 Bb4!? 18.c4 Nf6
19.Red1 c5!? 20.d5! exd5 21.cxd5 Rxd5 22.Qf1! Qc6! 23.Be2 Rf5! 24.Rd3! c4! 25.bxc4 Kg8! 26.Qh3! Rh5!?
Position after: 26...Rh5!?
The better of Black’s two real options, and we shall try and provide a conclusive and favourable judgement at
last.
27.Rd8+
27.Qg2 Kh7= and Black has developed.
27...Bf8
27...Kh7?!
550
Position after: 27...Kh7?!
Bad for a beautiful reason. 28.Qxh5!! One of the nicest moves in the whole book. 28...Rxd8! 29.Ng5+ Kh8!
30.Nxf7+ Kg8 31.Bf3 Nxh5 32.Bxc6 Kxf7 33.Bxb7 Nf6 34.Rb1² Black is certainly active, and can take some
of White’s trumps off him, but White will keep either the bishop pair or the pawn, and hence press without risk.
28.Qg2!
28.Qf1 Rf5„ reminds White of his long-diagonal problems.
28...Qe4!?
The beginning of some bizarre geometrical motifs – there does not seem to be another way for Black to even try
and develop. I briefly tried to make ...Nh7 work, but if Black is preparing until move 28 only to get a kingside
like that, he has to question his life choices.
29.Re1 Qe7 30.Rd3!
For the second time White comes to this square to overprotect f3.
A) After 30.Rdd1 Qe8= Black tries to reinforce the pin from a8.
B) 30.Rb8 sets up various exchange sacrifice ideas, but 30...Ra5! keeps the balance. For instance: 31.Qf1 Qc7
32.Be5 Rxe5 33.Rxb7 Qxb7 34.Nxe5 Qc7 35.Bf3 Bd6=.
30...Ra5!
Black is nevertheless able to equalise this way. White has no really forcing continuations and play might
continue (for instance)...
31.Re3
551
31.a3 Qc7 32.Bxf6 gxf6= and Black has unravelled.
31...Qd8 32.a3 h5 33.Rd1 Qc8 34.Qf1 Rh6„
Position after: 34...Rh6„
Both sides have finally completed development and a very tense late-middlegame phase will now follow.
3.3) 19.Red1 Be7
Qc7 11.0-0 b6 12.Qg4 Kf8 13.b3 Bb7 14.Re1 Rd8!? 15.Bb2 Nf6 16.Qh3 Nd5 17.g3 Bb4 18.c4 Nf6 19.Red1
Be7N
552
Position after: 19...Be7N
Officially my main recommendation, though I have evaluated both the lines in this chapter as giving equality. It
definitely helps to have some variety! Here Black still plays for ...c5, but does so with the bishop behind it and
the f6-point reinforced.
The reason for preferring the text over ...c5 is that it is easier to analyse, and I can be more confident of my
conclusion that Black should reach a position where White’s advantage is entirely symbolic. Not to mention, it
is not given in any work that I know of.
20.Ne5
Most sensible, given that there wasn’t yet a logical way to deploy the a1-rook.
A) 20.c5 can probably just be gobbled. 20...bxc5 21.dxc5 Bxc5 22.Rac1 Be7 23.Bc4 c5 24.Ne5 Bd6=
B) 20.Qf1!? The favourite of cloud engines on high depth; it is as if some 10 moves too late they realise the
White queen belongs on e2! 20...h5! Concretely after 20...c5 21.d5 Black’s game does not work out. 21.Qe2
(21.h4 c5 22.d5 exd5 23.cxd5 Bxd5! 24.Qe2 Ng4ƒ Now Black has some targets he can use to gain
counterplay on the kingside, and furthermore White’s f3-knight looks a bit lost.) 21...h4! The gained tempo is
worth the pawn. (21...c5?! fails to convince, for the usual reason: 22.d5 exd5 23.cxd5 Nxd5 24.Bc4² Black
lacks moves and White does not: he can think about doubling, or Qe4, or Ne5.) 22.Nxh4 c5
553
23.d5 (23.Bc2 If White tries to avoid giving back the pawn, there are some novel possibilities. 23...Rxh4!?
24.gxh4 Qf4©) 23...exd5 24.cxd5 Nxd5 Now White suffers from not having Ne5 available. 25.Nf5 (25.Nf3
Qc8!? 26.Bc4 Qg4„) 25...Bf6 26.Bxf6 Nxf6 White could be said to have a development advantage, but I’m
not able to find a way of using it. Critically, Black avoids the rook trade if he can. 27.Bb5 (27.Bc4 Re8=)
27...Bd5! 28.f3 (28.Ne3 Rh5=; 28.Rd2 g6 29.Ne3 Kg7=) 28...Rh5 29.g4 Rh3 30.Qg2 Rh8 This is – would you
believe – a machine line, but even in this line Black will manage to deploy his h-rook well, e.g.
31.Ne3 Qe5 32.Qe2 g6= …...Kg7.

554
C) 20.Rac1 c5= now functions as normal and equalises.
20...Nd7!?
Will lead to some very welcome exchanges and a solid enough position.
20...c5?! Extremely risky and unless the forced variations can be improved drastically, this is not
recommendable. 21.dxc5 Bxc5 22.Bg6!±
21.f4
Critical, of course.
A) 21.Re1 Nxe5 22.Rxe5 Bf6³
B) Black is able to repulse immediate attacks such as this one: 21.Nxf7 Kxf7 22.d5 cxd5 23.Re1 Nf8³.
C) 21.Be4 This leads to mass simplifications. Initially, I didn’t believe White had anything better than this,
whether practically or objectively. 21...Nxe5 22.dxe5 g6
23.Qg2 The queen needs to play some role. (There are also ideas like 23.Qg4 c5 24.Rd6 which should be
calmly met by contesting the d-file within the allowed space: 24...Kg7 25.Qf3 Bxe4 26.Qxe4 Qc8! This now
threatens Bxd6 and f6. 27.Qf3 Rd7 28.Rad1 Rhd8 29.h4 Bxd6 30.exd6+ Kh7©) 23...Kg7 24.Qf3 There is not
much left in the position: here a sample line leading to essentially equality is 24...Rxd1+ 25.Rxd1 Rb8 26.h4
h5 27.Kg2 c5 28.Bxb7 Qxb7 29.Qxb7 Rxb7 White has a purely nominal advantage. By the time he arranges
any activity at all, the rooks will be off and Black will have a fortress.
21...Nxe5
555
22.fxe5
22.dxe5?! Bc5+ (22...b5 might be a way to play more enterprisingly: 23.Qf1 g6∞) 23.Kf1 Bd4= This is most
principled play by Black: he exchanges his dark squared bishop and then plays c5.
22...c5
Entering a somewhat messy position where two White moves suggest themselves. The first, and less critical,
is...
23.Rf1
...aiming for a quick attack through the centre. The more critical 23.d5 is seen in the final subchapter.
23...cxd4 24.Qxe6
A) 24.Qh5 Bg5³
B) 24.Bg6 Qc6 25.Rxf7+ Kg8 26.Rxe7 (26.Raf1 d3™) 26...Qh1+ 27.Kf2 Qf3+ 28.Ke1 Qe3+=
24...Bg5
556
25.Rae1
Not just stopping Be3 but also defending the e-pawn.

25.Bc1?! only performs one of the above tasks: 25...Kg8! 26.Qf5 g6 27.Qf2 (27.Qh3 Re8! This is problematic
due to the tactical detail 28.Bf4 Rxe5µ) 27...f5!!
Position after: 27...f5!!
28.Bxg5 (28.exf6 Qc6–+) 28...Qc6 29.Qf3 Qxf3 30.Rxf3 hxg5 31.Rf2 Kf7³ Black continues with ideas like
g4, Ke6, a5, maybe doubling on the h-file, or maybe Rc8-c5, or maybe Be4.
557
25...Bc8 26.Ba3+
26.Qg6 Kg8 27.Qe4 h5„ A computer move to begin Black’s counter-attack.
26...Kg8
This is almost a rather novel queen trap, but sadly there is one way out.
27.Bd6™ 27...Bxe6 28.Bxc7 Rd7 29.Bd6 h5 30.h4
Else Black plays h4.
30...Bd2 31.Re2 Be3+ 32.Kg2 g6=
Black unravels and is not worse.
558
Position after: 32...g6=
3.4) 19.Red1 Be7: main line with 23.d5!
Qc7 11.0-0 b6 12.Qg4 Kf8 13.b3 Bb7 14.Re1 Rd8!? 15.Bb2 Nf6 16.Qh3 Nd5 17.g3 Bb4 18.c4 Nf6 19.Red1
Be7 20.Ne5 Nd7 21.f4 Nxe5 22.fxe5 c5 23.d5!
Position after: 23.d5!
23...exd5 24.cxd5!
559
This and similar variations were the subject of intense debate during a training camp I gave. Typically, we
reached queen endings where defence and attack were both very intricate.
A) 24.e6 Bg5∞ White’s king is no safer than Black’s.
B) 24.Rf1 This does not seem to set Black as many problems. 24...d4!
Position after: 24...d4!
B1) 25.Rxf7+ Kxf7 26.Qf5+ (26.Rf1+ Bf6 27.exf6 g6 White has nothing better than a draw.) 26...Bf6
27.exf6 Kg8 28.fxg7 Qxg7 29.Rf1 White keeps the game going. There was perpetual with Qe6-g4 instead.
29...Bc8! 30.Qf3 Qe7 31.Bc1 Bh3∞
B2) 25.e6?! Bf6 shuts down White’s attack.
B3) 25.Bg6 Qc6! (25...Bg5!? is an interesting queen sacrifice not mentioned by my engine: 26.Rxf7+ Qxf7
27.Bxf7 Be3+ 28.Kf1 Kxf7 and it slowly realises Black is fine) 26.Rxf7+ Kg8 27.e6 h5
560
In the absence of immediate tactics for White, Black gets to activate with Rh6. 28.Rxe7 (28.Raf1?! Rh6‚)
28...Qh1+ 29.Kf2 Qf3+ 30.Ke1 Qe3+= This is a draw since trying to escape with Kd1 leads to death after
...Bf3+.
24...Bxd5 25.Bc4 Bxc4 26.bxc4 Qc6!
It is absolutely critical to avoid the immediate e6 or Rd5 moves. Now it seems to me the only logical
continuation is...
561
27.Rxd8+ Bxd8 28.Rd1 Be7!
I think this saves Black some suffering in objective terms. 28...Bg5 leads to some fascinating endgames with
three pawns for a piece, but I don’t quite believe Black’s game there.
29.Rd7
29.Qd7 Qe6 30.Qxa7 Kg8
29...Qe6 30.Qxe6 fxe6 31.Rxa7 Ke8!
Position after: 31...Ke8!
The idea is once again (see 21.Be4) to make a draw with a nominally worse bishop: ...Rf8-f7, ...Bd8.
32.Kf2
A) 32.Ra6 Bd8=
B) 32.a4 Rf8 33.Kg2 Rf7 34.Bc3 Bd8=
C) 32.Rb7 Kf7! 33.Kf2 Ra8=
32...Rf8+ 33.Ke2 Rf7 34.h3 Bd8 35.Rxf7 Kxf7=
Not, perhaps, the world’s most thrilling draw, but a sound one.
This was the most substantial chapter in the book, if not the longest, and with good reason: 14...Rd8 is the
562
main reason I started to write the book in the first place and I am proud of the analysis in this chapter,
some of which was made in collaboration with other people during a training camp at my house in 2017.
As mentioned in the introduction, I have given Black a stylistic choice. The first line 3.1) serves mostly to
cover the sidelines – as we saw, 26...g6 was interesting but ultimately fell short, though it nevertheless
occupies an important place in the narrative surrounding my development of this whole variation.
Within the first stylistic choice that Black could have made, line 3.2) is also found. It is in lines like this that
one can really win games by preparation alone, and I recommend that if you decide to play 19...c5, you
may as well go the whole hog and memorise the whole line more or less to its conclusion, as White has
numerous ways to go very wrong indeed.
The other choice is the natural 19...Be7 (at least, it is natural to me now, having looked at this position and
its successors for something like 100 hours). In this case, Black need not memorise as much – a further 5
moves or so would do, since that’s enough to know whether White is going for the perpetual check line
(where he can easily get mated) or the endgame line.
In the endgame line, it is possible to look at 28...Bg5 instead of ...Be7, but as you already know from the
move number, it is pointless to even think about that unless the line catches on.
563
Chapter 4
Sightseeing along the Yellow Brick Road
Chapter guide
Chapter 4 – Sightseeing along the Yellow Brick Road
4.1) 10...Nf6 11.Qe2 0-0?!
4.2) 10...Nf6 11.Qe2 b6
4.3) 10...Qc7 11.0-0 b6 12.Qg4 g5?!
4.4) 10...Qc7 11.0-0 b6 12.Qg4 Kf8 13. b3 Bb7 14. Re1 Rd8! 15.Bb2 Nf6 16.Qh3 Nd5 17.g3 c5!?
As with the final chapter of the previous 2 Parts, I have decided to give the reader a bit of breadth in terms of
what is covered. In this case, we look at lines which have fallen by the wayside as theoretical progress has
overwhelmed them. In all cases, there are some subtleties which are worth noting before we consign them to the
great dustbin of history- the lines have cultural value and the ideas are applicable across a range of other
variations.
4.1) and 4.2) deal with 10...Nf6, the main move in Karpov’s time. On the following move Black can follow up
in two ways, and both are dealt with briskly, though it is worth noting that the final position of 4.1) is not so
abjectly terrible for Black.
4.3) deals with the famous 12...g5 from Anand-Bareev, which I tried out once my-self, though without success.
Two lines are worth noting here: the main line, and the Golubov game given in the notes. Some hitherto
unknown subtleties are re-quired to bury both for good, and while these continue to be unknown, Black may
consider using 12...g5 as a surprise weapon.
Finally, 4.4) deals with an extra option from the previous chapter, which was dis-carded in Negi’s book,
564
perhaps slightly prematurely.
4.1) 10...Nf6 11.Qe2 0-0?!
Nf6!?
Position after: 10...Nf6!?
11.Qe2
11.Qh4 is of course famously met by Karpov’s 11...Ke7! which opens up the back rank defence of h8, thus
threatening ...g5. White’s next is almost forced. 12.Ne5 Bxe5 13.dxe5 Qa5+
565
If Black didn’t take this pawn, his game would make no sense. Now White is best advised to offer a second.
A) 14.Bd2 Qxe5+ 15.Be3 b6 16.0-0-0 (16.0-0 c5 and ...Qh5 should equalise) 16...Rd8 17.Rhe1 Qa5=
Golubev, M – Drozdovskij, Y Odessa 2010.
B) 14.c3 Qxe5+ 15.Be3 b6 16.0-0-0 (16.0-0N I believe c3 may actually go better with short castling. 16...c5
17.Rfe1 g5 18.Qh3 Bb7„) 16...g5 17.Qa4 c5 18.Rhe1 Bd7 19.Qa3 Rhd8 20.g3 (20.b4!? cxb4 21.Qxb4+ Qd6
22.Qxd6+ Kxd6 23.Be4+ Nd5= was more testing) 20...Qc7= Black had acheived a measure of solidity in
Kamsky, G – Karpov, A Dortmund 1993.
11...0-0
To save our sanity, we shall explain just once the attacking mechanism against early short castling from Black;
and let this explain also 11...b6 12.Bd2 0-0, 11...Qc7 12.Bd2 0-0 et cetera.
A) 11...Qc7 12.Bd2 does not really lead to independent possibilities.
B) 11...b6 is the subject of the next subchapter.
12.Bd2 Qc7 13.0-0-0
566
Position after: 13.0-0-0
The computer evaluates the position as level but White scores something like 80%. This is because the fruition
of the attacking construct Ne5, g4, h4, g5 is just beyond its horizon. Black can avoid being mated, but cannot
avoid being significantly worse.
13...c5 14.g4!?
14.dxc5?! is too helpful, even if objectively the two moves lead to the same evaluation at the end of their lines.
14...Qxc5 15.g4! e5!
Position after: 15...e5!

567
Black tries to develop his c8-bishop, ideally to g4 but also to f5 in some lines. 16.Be3 (16.g5 e4 is complicated,
but Black seems fine; 16.Rhg1 might ap-pear to solve the problem, but Black has 16...e4 17.Bxe4 Re8=)
16...Qc6 17.g5 hxg5 18.Bb5 (After 18.Nxg5 there was ...e4 and ...Bf5.) 18...Qc7 19.Bxg5 Qe7!
Position after: 19...Qe7!
20.Rhg1 (20.Kb1 Bf5„; 20.Bd3 e4! was Black’s point; 20.Nd2 Bf5 21.Bd3 Bxd3 22.Qxd3 Qe6 23.Rhg1²
White still has some attack, but Black has exchanged a further piece and completed development.) 20...e4
21.Bxf6 Qxf6 22.Qxe4 Rd8²
Position after: 22...Rd8²
568
The discussion currently being had, about whether the bishop pair is enough compensation for a pawn (no, but
close) is far more to Black’s taste than the alternative one of whether his king would end up as lunch.
14...cxd4
A) 14...Nd5?! 15.g5 Nf4 16.Bxf4 Bxf4+ 17.Kb1 hxg5 18.Rdg1 is close to being curtains, as the computer
realises on high depth.
B) 14...Bf4 15.Rhg1 cxd4 16.g5 (16.h4!±) 16...hxg5 17.Nxg5‚ Fedorchuk, S – Koziak, V La Roda 2012.
15.h4
15...Nd5
15...Nxg4 16.Nxd4 f5 17.Nb5 Qe7 18.Nxd6 Qxd6 19.f3±
16.Nxd4 Nf4 17.Bxf4 Bxf4+ 18.Kb1 Bd7
18...e5 19.Nb5 Qb6 20.Bc4± and for concrete reasons Black finds it impossible to complete his development.
19.g5 e5 20.Nf5 Bxf5 21.Bxf5²
569
Position after: 21.Bxf5²
After ...Qb6 (say) Black is surviving; in the next few moves he can challenge for the d-file and think about
...g6, ...Kg7, ...Rh8 operations, however one does not do preparation as Black, just in order to reach ² positions
as a best case.
4.2) 10...Nf6 11.Qe2 b6
Nf6 11.Qe2 b6

570
This seems better than short castling for Black, but the outlook here is also grim. If Black plays down the
critical main line, he ends up dead, so he should bail out at some point into a ² position.
12.Bd2
White’s short castling plans do not appear to offer anything.
12...Bb7 13.0-0-0 Qc7 14.Ne5
This is clearly most critical and I have been unable to find a way for Black to really play here.
14...c5
14...0-0-0 15.f4² and whether Black admits it or not, White has most of what he wants.
15.Bb5+ Ke7 16.dxc5
16.Nc4 Bf4=
16...Qxc5
Position after: 16...Qxc5
17.a3!
A) 17.f4 Rac8 18.Be3 (18.Bd3 Qd5= is a standard Caro pawn fork) 18...Qc7 19.Bd4 Rhd8= Black arranges
...Kf8 and ...Bc5 next and cannot be worse.
B) 17.Bc6?! is tricky, but probably not good: 17...Rac8! 18.Be3™ 18...Qxe5 19.Bxb7 Ba3! 20.Rd4 Rcd8
21.Rhd1 Rxd4 22.Rxd4 Bc5 23.Rd1 Bxe3+ 24.fxe3 Rd8³.
571
17...Qc7 18.f4!
A) 18.Bc3 Nd5 19.Rxd5 Bxd5 20.Rd1 a6!³

B) 18.Bf4 Rhd8 19.Rd4 Kf8 20.Kb1
Position after: 20.Kb1
(20.Rhd1 Kg8„) 20...Kg8! The alternatives were a little less convincing: (20...Nd5 21.Bg3 a6 22.Rc4 Qe7
23.Nc6² Kamsky, G – Karpov, A Elista 1996; 20...a6? Bad because...it leaves b6 unprotected. 21.Rxd6!!
Rxd6 22.Nc4± Ye, J – Schlosser, P Groningen 1996) 21.Rhd1 a6 One way or another, Bf4 leads to this
position. White has to decide where to retreat his bishop. 22.Rc4 (22.Bd3 Nd5 23.Bg3 b5=; 22.Bc4 Bd5
23.Bg3 b5=) 22...Qe7 23.Nc6 Bxc6 24.Bxc6 Rac8 25.Rxd6 Rxd6 26.Bxd6 Qxd6=
18...Rac8
Careful analysis shows that ...Rhc8 (as played in an obscure game from Sunny Beach in Bulgaria) walks into a
strong gambit. The convenience of castling by hand one move earlier is not worth the hassle of poor rook
placement.
572
Position after: 18...Rac8
19.Bd3! Rhd8 20.Kb1!
20.Rhe1 Kf8 21.g4 Ba6!‚ is the kind of thing White wants to avoid, and it is worth taking a tempo out to
ensure that Qxc2 is not mate.
20...Kf8
So Black has acheived basically what he wants, but the knight on e5 may still exert a cramping influence. It
turns out this effect is larger than anyone could have imagined.
21.Rhf1
21.Rhg1!? is a logical novelty to look at.
21...Kg8
573
22.Bc1!
This is the engine suggestion to preface g4. Now Black is more or less toast.
A) 22.g4 was seen in the only practical test, and there followed completely natural attack and counter-attack
from both sides. 22...Bxe5 23.fxe5 Nd7 24.g5 Nc5 25.g6 fxg6 26.Bxg6 Ba6 27.c4 Nb3
28.Bf7+ Kh8 29.Bxh6 Rxd1+ 30.Qxd1 Qxc4 31.Qh5 Qe4+ 32.Ka2 Bxf1?? (32...gxh6™ 33.Qxh6+ Qh7
34.Qxe6 Nd2 35.Qf6+ Qg7 36.Rf4 Qxf6 37.exf6 Rc4!=) 33.Bg6+– Timoshenko, G – Kacheishvili, G New
574
York 1999.
B) 22.g3!? White’s sneaky idea is that the obvious queenside pawn pushes leave non-obvious defects in
Black’s counterplay if he proceeds to model it after Kacheishvili’s. 22...Bf8 (22...a5 fails since in a critical
position White can play Qf2 and hit b6!; 22...b5 23.g4 Bxe5 24.fxe5 Nd7 25.g5±; 22...Nd7 23.Nxd7 Qxd7
24.f5‚) 23.Rfe1 Defending... the e4 square. (23.g4? Rxd3!–+; 23.Bb4 Bxb4 24.axb4=) 23...Bd6
Nevertheless trying to play ...Bxe5 and ...Nd7. 24.Bb4 There wasn’t actually much else to do. (24.Ng4 Nxg4
25.Qxg4 Bf8=) 24...Bxb4 Now for instance: 25.axb4 a5 26.bxa5 bxa5 27.c4 Qc5= White’s position may be
slightly more comfortable due to having the better majority, but the position is equal as long as queens stay
on.
22...Bf8 23.g4 Rxd3!?
575
Position after: 23...Rxd3!?
24.cxd3!
24.Nxd3 Nxg4 25.c3 Nf6 26.Ne5 Bd5©
24...Nd7 25.h4!
A) 25.g5 hxg5 26.Nxd7 Qxd7 27.fxg5 e5©

B) 25.Rde1² appears adequate for neutralising some of Black’s compensation too.
25...Nxe5 26.fxe5 Qd7 27.g5
White has to not worry too much about the queenside. 27.Bd2? Qb5©
27...Qb5 28.Rd2 Bd5 29.Rc2 Rxc2
576
30.Qxc2
30.Kxc2 Qa4+ 31.Kd2 Qxh4 32.gxh6² is similar but a bit less believable in a human game.
30...hxg5 31.a4! Qb4 32.hxg5 Qd4
Now the last computer continuation White needs is:
33.Bd2 Qxe5 34.g6±
577
Position after: 34.g6±
Black is probably not getting out of this one in one piece. The bishop on d5 is a ‹paper tiger›.
4.3) 10...Qc7 11.0-0 b6 12.Qg4 g5?!
Qc7 11.0-0 b6 12.Qg4 g5?! 13.Qh3! Rg8 14.Re1
14...Bf8
The best defence.

A) 14...Bb7?? 15.Rxe6+ fxe6 16.Qxe6+ Kf8 17.Qxh6++– was brutally convincing in Merry, A – Fernandez,
D England 2015.
B) 14...g4!? Highly critical, and White needs to have a few resources up his sleeve, stretching, it seems, to
move 27. However, this is not best. 15.Qxh6 gxf3 16.Rxe6+ fxe6 17.Qxe6+ Be7 18.Qxg8+ Nf8 This has
always, I think, been assumed to be straightforwardly winning for White, but in February 2017 the Russian
player Saveliy Golubov took it for a spin, and things are not as simple as they appear.
578
B1) 19.Bc4?! Qd6 20.Qf7+ Kd8 21.Qxf3 Bd7 (21...Qf6!? with ...Be6 to follow might also come very close
to equalising) 22.c3 Ng6 23.Bd2 Kc7 Black drew this with surprisingly little fuss in Paravyan, D – Golubov,
S Moscow 2017.
B2) 19.Bg6+! A necessary wrinkle. 19...Kd7 20.Bf5+ Ke8 21.Be4! Of course, the move repetition was to
make Black lose castling rights. (21.Bh6?! Qd6„ is not straightforward and Black is fully in the game)
21...fxg2 22.Bg5 Qd6 (After 22...Bxg5 23.Qxg5 Be6 24.c4! Rd8 25.d5 Black has nothing better than giving
up a piece and entering an almost certainly lost ending a pawn down.) 23.d5! It was important to stop ...Be6.
(23.Re1? Be6 24.Qg7 Bxg5 25.Qxg5 Kd7∞) 23...cxd5 (23...Bxg5 24.Qxg5 c5 leads to another line White
might need to know if facing a gambler: 25.Re1 Kd7 26.Bf5+ Kc7 27.Re7+ Bd7 28.Bxd7 Nxd7 29.Re6!
Qf8 30.d6+ Kc6 31.Qg4!±) 24.Bxe7 Qxe7
579
B2.1) 25.Bxd5? leads to a position White could lose: 25...Be6! 26.Bc6+ (26.Qxg2 Bxd5 27.Qxd5 Rc8³)
26...Kd8 27.Rd1+ Kc7 28.Qg3+ Kxc6 29.Qxg2+ Kb5 30.Qxa8 Qg5+ 31.Qg2 Qh5„.
B2.2) 25.Qxd5! Rb8 26.Re1 Bd7 27.Re3!± Computer lines, which we shall not trouble ourselves further
with, indicate that White ends up with a rook and two pawns for two pieces, and a continuing attack. (The
alternative – winning the queen with 27.Bg6 – is less than fully convincing.)
15.Qf5
Position after: 15.Qf5

580
15...Rg7!?N
This improves considerably, and it could be that some very motivated analyst finds a path to playability for
Black here.
15...Bg7 16.h4 Kf8 17.Qh3± Anand, V – Bologan, V Dortmund 2003.
16.Bc4!
A) 16.c4 Bb7 17.d5 cxd5 18.cxd5 0-0-0! 19.dxe6 fxe6 20.Rxe6 Kb8∞
B) 16.a4!? Bb7 17.a5 0-0-0 18.axb6 axb6 19.Qe4 Kb8² is certainly better for White, but not as
straightforward as it might seem, and actual wins on the queenside prove elusive if Black avoids bishop trades
and focuses on playing ...g4.
16...Be7 17.Qh3
Keeping it simple.
17.Bxe6!? fxe6 18.Qxe6 Nf8 19.Qxh6 Kf7 20.Bxg5! (20.Nxg5+ Kg8©; 20.h4 gxh4 21.Bf4 Qb7∞) 20...Rh7
21.Rxe7+ Qxe7
22.Qxh7+ (22.Qxc6 Qb7 isn’t so clear.) 22...Nxh7 23.Bxe7² This is better for White but somewhat difficult to
handle in practice.
17...Rg6 18.Ne5™ 18...Nxe5 19.dxe5 Bb7
581
20.Bd3
20.a4 0-0-0 21.a5 g4 22.Qb3 h5 23.Be3 c5„ is now too slow for White.
20...Rg8 21.Qxh6²
Black can castle queenside and obtain some kind of basic compensation for his pawn, but objectively his
kingside play should not come to much if White plays sensibly e.g. queen to g3 and bishop to e4.
4.4) 10...Qc7 11.0-0 b6 12.Qg4 Kf8 13. b3 Bb7 14. Re1 Rd8! 15.Bb2 Nf6 16.Qh3 Nd5 17.g3
c5!?
Qc7 11.0-0 b6 12.Qg4 Kf8 13.b3 Bb7 14.Re1 Rd8!? 15.Bb2 Nf6 16.Qh3 Nd5 17.g3 c5?!
582
Position after: 17...c5?!
The last deviation from this Part’s lines that I consider to have any merit at all for Black, and that not on
objective grounds but purely subjective ones.
18.c4!
18.dxc5 leads to nothing: 18...Bxc5 19.Be4

A) 19...Nf6 20.Bxf6 (20.Bxb7!?N 20...Qxb7 21.Qg2 There is potential for White to play a long game here
and be better.) 20...gxf6 21.Qh4 Be7 22.Bxb7 Qxb7 23.Qf4 Kg7 24.Qg4+ Kh7= Here the game Giri – Ding,
Antalya 2013 was agreed drawn, though maybe Black could consider playing on at a lower level.
B) 19...Ne3!?N
583
Position after: 19...Ne3!?N
This is more reliable than Ding Liren’s choice. 20.b4 Bxe4 21.bxc5 Bxf3 22.Rxe3 Bb7 Black gets
counterchances along the long diagonal.
18...Nf6
18...Nb4? 19.Be4 cxd4 20.Nxd4 Bc5 21.Rad1 This arises by force from 18...Nb4 and the best Black seems to
be able to manage from here is a somewhat unpleasant ending. This was a main line in Negi’s white repertoire
from 2013.
19.d5!
584
Position after: 19.d5!
19...exd5 20.Bxf6 gxf6 21.cxd5
White is clearly better now to the human eye. We continue the line, as there are some subtleties to come.
21...h5!
Black limits the number of light squares that his opponent has under real control. Now there are all kinds of
random moves that lead to a positional bind for White, though not, dare I say, one without significant chances
of a turnaround, especially if Black takes the pawn. For instance: Bb5, Bf1, Qf1, etc. We shall take those as
given and investigate just those moves that either try and keep the pawn, or try for an immediate attack.
22.Rad1!
It turns out that the two moves which attempt to keep the pawn have concrete drawbacks.
A) 22.Qf5 Be7 Setting up the idea of Be4 Bc8! and so to defend the pawn it becomes necessary to reply
23.Bc4 but now Black frees his game with 23...b5!
585
24.Bxb5 Rxd5 25.Qh3 c4! 26.Bxc4 Qc6 27.Bxd5 Qxd5 28.Re3 Bc5 29.Rc3 Bb4= Black forces a draw by
repetition; there were other options to play on instead of 25...c4 but this is good enough.
B) 22.Bc4 looks sensible, but in fact this loses White’s advantage. The best reply is 22...Qc8!.
Now there are interesting forced lines assuming White dodges the queen swap – and if he agrees to it, then
only Black can really be playing for the win. 23.Qf1 (23.Qh4 Qf5 24.Re3 Kg7= Black will either take the
queens off next with ...Qg4, or in some cases take the pawn with ...Bxd5. If neither appeals he can always
clog the centre with ...Be5.) 23...a6 24.a4 h4!? It transpires that the exchange sacrifice is actually playable
586
here! 25.Nxh4 Rxh4! 26.gxh4 b5 27.axb5 axb5 28.Bxb5 (28.Bd3 Bxd5 29.Be4 Qg4+ 30.Qg2 Qxg2+
31.Kxg2 Bxb3©) 28...Bxd5 29.h3 Other moves are convincingly met by ...Qf5, cueing up ...Qf3 as well as
...Qf4. (29.Be2!? Qf5 30.Ra4 Bf4 31.h3 Qe5!‚ White had apparently covered all his bases, but the battery
works the other way round too!; 29.Qe2? Qf5; 29.f3?! Qf5 30.Qd3 Qh3 31.Qe2 c4‚) 29...Qf5 30.Qd3™
Position after: 30.Qd3™
30...Bh2+ 31.Kf1 Bg2+ 32.Kxg2 Rxd3 33.Bxd3 Qxd3 34.Kxh2 Qxb3=
22...Rh6

587
Controlling g6 and thinking about ...B(Q)c8, ...f5, ...Rf6 possibilities. It was not yet possible to take on d5.
23.Be4!
The best of White’s available bishop moves, defending the d-pawn twice rather than once. I believe only this
move guarantees an advantage. And yet, paradoxically, the way in which it does so involves sacrificing the d5-
pawn.
A) 23.Nh4 Be5! and now 24.Bc4 Bc8 25.Qf1 Bc3 26.Re2 f5„ for instance would make some sense of
Black’s construction.
B) 23.Qf5 Defending the pawn with the queen runs into 23...Bc8 24.Qe4 f5!ƒ.
23...Qc8
Again restricting White’s choice.
24.Bf5!
24.Qf1 f5 25.Bc2 f4 26.Nh4 Qg4= The b7-bishop is not good at all, but Black’s position makes a modicum of
sense and he is not being mated. Thus, he will be able to continue with moves like ...Bc8, ...Qg5, ...Rf6 and
begin enjoying his life.
24...Qa8 25.Bd7!
A) 25.Nh4 can be considered as well, but one strong move is enough.

B) 25.Be4? Qc8 repeats.
25...Bxd5 26.Nh4! Be5!

588
In my original version of this file, I gave as advantageous a much worse line for White and assessed the present
position as ‘counterplay’. Clearly, White has to show remarkable precision here or else he is a pawn down, but
equally clearly, the evaluations were the wrong way round...
27.Qf5! Ke7! 28.f4
28.Rxe5+!? fxe5 29.Qg5+ Kxd7 30.Qxh6²
28...Be6 29.fxe5 Bxf5 30.Nxf5+ Kf8 31.Nxh6 Qb7 32.Bh3 Rxd1 33.Rxd1 fxe5±
Position after: 33...fxe5±
The dust has cleared and White has a clearly better endgame, though one not without certain challenges.
There are some surprise possibilities in here, though they are all a bit ropey, except the last line. It is, for
instance, still just about possible to build your repertoire around 10...Nf6, but you can expect it to be
mercilessly targeted and to end the opening phase of each game noticeably worse.
That being said, all these lines were important in the evolution of the repertoire you see presented to you
today in the main bits of this book, and it seems only right to give them a proper burial, so as not to let the
White players think they can refute our antecedents any old how. (Even the supposedly terrible 12...g5
requires nuances right up to move 26 to finally put it to sleep.)
I am not usually a sentimental person, but to me that seems the perfect note on which to end the book. I
hope you have enjoyed the ride, and that I have been a good driver.
589

Antic D Maksimovic B - The Modern French - New in Chess 2012

Încărcat de

Informații document

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Antic D Maksimovic B - The Modern French - New in Chess 2012

Încărcat de

Drepturi de autor:

Formate disponibile

1

Examples: anything by Boris Avrukh, Nikolaos Ntirlis, or Vassilios Kotronias.

Examples: anything by Negi Parimarjan or Milos Pavlovic.

16. Rac1 Qb8

1.1) 2.Ne2 and others

Position after: 3...dxe4

Position after: 4.Ne5

Position after: 3.e5

4...cxd4?! 5.Nxd4 Nc6 6.c4 e6 7.Nc3 Bc5 (7...Nge7?!

To my mind, the only thing one has to remember about 2.Ne2.

7...Bd7 8.Qh5 Qb8 9.Bf4 Nd3+ 10.Bxd3 Qxf4=

Black had equalised in Sedina, E – Botta, G Lugano 2011.

Position after: 2...d5

The famous Two Knights system.

Black is faced with a slightly awkward choice.

8.d3 Ng5 9.Bxg5

1.3) 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.Be2 Bf5!

1.e4 c6 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.Be2!? Bf5!

Position after: 5...Bf5!

Position after: 8...b6

Position after: 6...Bd7

Position after: 8...c5

Black has a good French-type position.

12...Nc6!? 13.cxd5 (13.dxc5!? d4 14.Nd2‰) 13...exd5 14.Be3

Position after: 17...Qd8!

Position after: 12...dxc4

Unexpectedly bringing about a Paulsen Sicilian setup.

‘Simple and safe.’

Position after: 16...Nd7

A) 17.Nc3 Bc5= is absolutely nothing.

Anyway! This is quite a significant point.

19.Bxc5 Nxc5 20.Qxc5

Else just ...Qb6 with equality.

Position after: 20...Bxg2

21...Qxc8 22.Kxg2 Qc5 23.Nc3 Qxe5∞

1.4) An extra option!

Position after: 15...Be7

16.Nc3 0-0 17.Qg4 Nd7

Position after: 19.Rad1

Position after: 24.Qf3!

Position after: 18.Rad1

Position after: 18...Qb8!

Threatening to take on e5 and not being afraid of ghosts.

Position after: 19...Kh8

A) 20.Nxd7 Bxd7 21.Bd3 Rd8 22.Qe4 g6=

Position after: 25...Rxg8

Position after: 20...e5

22...Nc5 23.Rad1 Qe7 24.Rfe1!

Position after: 24...a5!

Else Black may actually be in some trouble.

Position after: 26...Nf6

28...Rxd8 29.Rxe5 Qc7!

30...Bd7 31.Qxa5 Qxa5 32.bxa5 Rc8 33.Bxf6 Rxc4 34.Bd8 Ra4=

The extra pawn falls, leaving a drawn endgame.

1.5) 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.d4

1.e4 c6 2.Nc3 d5 3.Nf3 Nf6 4.e5 Ne4 5.d4

Position after: 5.d4

5...Nxc3 6.bxc3 c5!?

Given the stamp of approval by Khenkin and Rasmussen among others.

Position after: 15.Nf4ƒ `

Position after: 10...b5!

Trying to keep the central tension alive.

8...cxd4?! This is wrong, but instructively so. 9.Nxd4

Position after: 10.0-0!