Tuesday, June 17, 2008

How to Defend? Or, How to Write a Chess Book? -- A Review

How to Defend in Chess: Learn from the World Champions
by Colin Crouch
Gambit, 223 pp., £13.99

If I am typical of most chess players, then it is more common for us to wax lyrical about a great player's style than to actually analyze their games methodically and intricately with pieces in hand. Karpov we declare a Python, Tal a Magician, without so much as studying a rook endgame, analyzing a bishop sacrifice. Or we frame the development of chess style as a story about historical discoveries. For instance, we talk about the development of classicism out of romanticism, perhaps charted through a few famous game snippets from Steinitz, separated by decades. Colin Crouch's unusual and interesting book How to Defend in Chess avoids both the crass unchessiness of the former, metaphorical attitude and the deceptive clarity of the latter, fragmentary, unstudied, historical sweep. Yet his book is about the play of two very different World Champions whose styles have always attracted considerable verbiage. The two players are Emmanuel Lasker and Tigran Petrosian, and Crouch is particularly concerned with the different styles with which they successfully handled inferior positions.

All that is not to say Crouch doesn't make overarching points. He sets out his broadest conclusions about the two player's "defensive" styles right from the start in his brief Preface. In the inferior positions Lasker frequently stumbled into, he was capable of creating great complexities that his opponents would typically fail to fathom. Instead they would flail amidst the confusion Lasker conjured, as Lasker himself demonstrated again his greater practical mastery of chaos. Petrosian, meanwhile, had a chess nose capable of scenting the subtlest of dangers over the longest of distances, and a corresponding facility for then preventing potential attacks long before most players would sense that they were on the horizon, says Crouch. Nor does Crouch's writing lack a broad historical context. His first chapter - Principles of Defence - is an interesting discussion that charts defensive theory and practice from before Steinitz to after Nimzoswitch. The most important historical point he makes here amongst many is that the concept of Prophylaxis divides the styles (and, to an extent, eras, one wonders) of the two champions he writes about. Already we are a long way from Lasker "The Psychologist" and "Iron" Tigran.

But the rest of the book is the bulk of the matter: 200 pages consisting of two approximately equal-length chapters - Lasker as Defender, and then Petrosian as Defender. Each chapter in turn consists of ten main games, deeply analyzed and intricately described, along with several supplementary related games and fragments, either briefly analyzed or not analyzed at all. Here amidst Crouch's annotations, there is a lot to praise. There's historical detail, psychological insight, aptly-chosen quotations, often from contemporary sources that must have required considerable research to track-down. There is also a thread of original ideas running through these two chapters, related to Crouch's thoughts about the Geography of the Chess Board which he articulates in the Principles of Defence chapter. There's also an overarching attention to his thesis about the two players differing style in inferior positions. This proves hard to argue with, especially so given Crouch's evident intimacy with the games of the two players, which is itself visible from many interesting examples. For instance, he usefully compares two Petrosian games that featured double-isolated central pawns - the only two, he adds.

But of course the main thing is how Crouch describes the games, and the variations he analyzes. These two dimensions of the book are intricately linked, each informing the other, something all the more impressive given how deep and detailed the variations are. In fact, a casual glance at a few passages might suggest computer spool, but a closer read shows this is not the case. Instead, Crouch has chosen the variations he analyzes on the basis of a very human understanding of the position. I think this is an extremely important principle that every chess author should obey nowadays, and of which Crouch proves himself exemplar. An example of what I mean and why might help here to clarify my point and praise. When I put the following position from Fischer-Petrosian (5th game Candidates Match 1971, p.190 - 197)...

... into the computer programme Crafty, the first move it wants to play is 13.Qe2. This is followed by several minutes of analysis of 13.Nh4. Considerably later, it chooses the superior 13.d5. Were an average annotator of this position to follow Crafty, then, he might arrive at 13.d5 having first considered two rather off-the-boil moves. Instead, Crouch has already spent half a page or so discussing why white should attack in this position on the queenside, and not the kingside. In particular he discusses whether or not white should attack with pieces. This is what Fischer did with 13.Qb3?!, but this soon walked into a ...b5 push that diluted white's space advantage. Instead, then, white should attack on the queenside with pawns, says Crouch. Thus after Fischer's 13.Qb3 in the diagram above, Crouch instead analyzes the moves 13.d5, 13.b4 and 13.a4, all of which follow from his description of the position. This is a fairly simple example of what I'm talking about, but more complex examples can be found in each game featured in the book; an outstanding quality in itself, also suggesting not a moment of laziness in the book's writing.

In fact, if there's anything to criticize in Crouch's analysis, it's that sometimes the wordage could have been edited down a touch. This is especially so in the Petrosian games, which, far more than the Lasker games, feature subtle but nonetheless absolutely clear positional decisions. Take the following position from Spassky-Petrosian (7th game World Championship 1966, p. 172 - 181):

Here, famously, Petrosian played 17...c4!! and after 18.Be2, 18...a6!. Crouch analyzes this sequence across two or so pages, but I am not convinced that it was necessary to do so at such length. The idea is simple: if white now tries to attack with a4-a5, ...b6-b5 will close the queenside, or if he tries to attack with b4-b5, instead ..a6-a5 closes the queenside - and anyway, a white knight on d4 isn't going anywhere, so there is no particular point centralising it, thus no real concession on black's part. All in all, after the sequence 17...c4!! 18.Be2 a6!, black is left free to attack white on the kingside, and the position is more or less strategically won at this point. There are, true, several auxiliary questions surrounding this sequence: does the plan of b4-b5 with Nd4-c6 offer white any convincing attacking chances on the queenside? Could ..c5-c4 have been played earlier, before black castled queenside? Should white have played b4xc5 before black castled queenside and was able to play ..c5-c4? Or with the king in the centre does this exchange favour black? Would various shots involving Bd3-f5, with ideas of e5-e6 if black takes it - or Bh3 defending if he doesn't - have improved for white? These Crouch answers convincingly amidst his long discussion at this point, but sometimes implicitly, sometimes without concision. Still, this is a minor quibble. It is in fact better to have too much writing than not enough. This is because the latter leaves the reader groping in the dark, whilst the former is frequently pedagogically useful, as it allows the reader to mentally trim down the flesh of language and discern for him or herself the fundamental chess skeleton beneath.

And anyway, Crouch's discursive, inquisitive writing style is ideally suited to talking through Lasker's games. Sometimes in these we even find positions where it looks like some one has just dropped a bunch of pieces on the board at random, and Crouch does a good job of disentangling the disorientating visual impression these positions make. However, it is important to note that Crouch never pretends there is anything simple about any of these games - he never talks them into triteness. In staying true to their difficulty, Crouch also assumes a degree of chess competence on the part of the reader which makes this book unsuitable for less-strong players. For instance, above I briefly mentioned double-isolated central pawns. Crouch does not pause to explain the basics of when and why these are not necessarily weak. If this general understanding is not already and automatically yours - if, say, you read the phrase double-isolated central pawns above and shuddered, or assumed Petrosian won against them rather than with them - then Crouch's discussion of Petrosian's double-isolated pawns in these games will be pitched at least one notch above chess knowledge you lack. For the record, there is one game in the book I still feel like I don't have much of a grasp on; my grade next year will probably be in the mid-180s (2175 FIDE Elo) although I had a good season and so my true strength is probably a bit below that.

There is another way to talk about this book: as a games collection, irrespective of your interest in defence or the style of the two Champions. In this way we see how unusual and interesting the book really is. Certainly it is no ordinary games collection. Only a few of the games would find their way into "Best Ofs", one or two into "Greatests", but most would not. This attractive, rare quality makes the games far more typical of ordinary player's games than those in most games collections. There are rough and tumble ganes, games where all plans get stifled, games that are practical and imperfect, games where evaluations yo-yo, and so on. A large majority of the games see Lasker or Petrosian handling the black pieces too, so the book might be of especial interest to those who struggle more on that side of the board than on the white side. Perhaps one downside is that for all their prophylactic accuracy, one or two of the Petrosian games are really quite dull, where "nothing happens" or at least so it seems. ("They say my chess games should be more interesting. I could be more interesting - and also lose." – Petrosian) But the Lasker games are invariably interesting throughout.

One thing as a games collection this book certainly is not, is a "How to" book. In fact several of these games are not even really defensive in nature, something John Watson's mostly-laudatory review details further. Instead it is about how Lasker and Petrosian handled inferior positions with different styles. Likewise it is misleadingly subtitled: the book is not directly concerned with teaching the reader something, although of course learning from it is not impossible, but teaching is not this book's central concern. Anyway it is a shame that such an intrinsically interesting book with so much self-evident integrity should need to be packaged in such a way, but this shouldn't put potential readers off.

Finally, here are two of my favourite games from the book for you to play through. If you find yourself intrigued too, then I recommend Crouch's book as a good place to understand them better:

PS. The Gambit website for this book also contains a PDF sample from Crouch's work to download.


Anonymous said...

Thanks Tom,

That's a thoughtful review and Crouch's book is now on my shopping list.

There's another review here.

What is meant by 'geography of the chessboard'?


Tom Chivers said...

His thoughts on the geography of the chess board are about how different sectors of the board interact with each other and how to conceptualize this. It's interesting, but perhaps more esoteric than enlightening. Or perhaps I'm not strong enough to judge. For instance in this game of Petrosian's as black v Duckstein from Varna 1962...

1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Bf5 5. Ng3 Bg6 6. Nf3 Nd7 7. Bd3 e6 8. O-O Qc7 9. c4 O-O-O 10. Bxg6 hxg6 11. Qa4 Kb8 12. b4 Nh6 13. Qb3 Nf5 14. a4 e5 15. dxe5 Nxe5 16. Nxe5 Qxe5 17. Bb2 Qc7 18. c5 a5 19. Rad1 Rxd1 20. Rxd1 Rh4 21. bxa5 Bxc5 22. a6 b6 23. Re1 Ka7 24. Be5 Qd7 25. Ne4 Bd4 26. g3 Bxe5 27. gxh4 Nd4 28. Qd1 Qd5 29. Re3 Nf5 30.Re1 Nd4 31. Qd3 f5 32. Ng5 c5 33. Re3 c4 34. Qd1 Kxa6 35. Ra3 Bf6 36. h3 f4 37. Qg4 Ka5 38. Nf3 Kb4 39. Nxd4 Kxa3 40. Nc2+ Kxa4 0-1

... Crouch emphasizes Petrosian's correct use of the centre as a springboard for action rather than as permanent residency for his pieces, pointing out how black vacates and occupies it at various points.

Anonymous said...

I think Colin told me he didn't use a computer to write/check this book. An impressive piece of computerless work. I wonder if the re-release has been computer checked. Kasparov quoted Crouch's analysis a few times I think in the Lasker section of his work. Colin's analysis came out pretty well from this grilling.

Tom Chivers said...

Perhaps the book's editor Graham Burgess would have computer-checked them though, as a matter of course . . . ?