Book review: Danger in chess - Avni

[This review first appeared in KingPin]

Danger in Chess: How to avoid making blunders

Amatzia Avni (121pp, Cadogan Chess, 1994) [[sterling]]9.99
ISBN 1-85744-057-9

  This is a great little book, from an Israeli psychologist and chess-player, on a subject that must be on every chess-player's list of New Year Resolutions: I will not overlook pieces en prise, I will not miss a mate in two... Avni takes a brisk look around at blunders and other 'failures in the sensing of danger', that is, sins of omission as well as commission.

  Forgive me if you've seen this particular sin of omission before, but it was a perfect recent illustration of Avni's themes:

Kasparov - Karpov, Linares, 1994. 1. e4 c6 2. d4 d5 3. Nd2 dxe4 4. Nxe4 Nd7 5. Bc4 Ngf6 6. Ng5 e6 7. Qe2 Nb6 8. Bb3 h6 9. N5f3 a5 10. c3 c5 11. a3 Qc7 12. Ne5 cxd4 13. cxd4

Now Black played 13... a4 missing 13... Bxa3 winning a pawn. If 14. bxa3 Black wins the exchange by 14...Qc3+; and if 14. Rxa3 Black recovers the piece by 14...Qxc1+.

  These two ideas are very simple, and if we are alerted to the possibility of a tactical blow here, I imagine most of us can look for and find it. But without suspecting danger, the two strongest players in the world missed it!

  As well as welcoming its subject matter (not another Solve all your opening problems forever title from Batsogan Digest), I also welcomed its approach, its use of examples and exercises, and his recommendations which attempt to remedy the ills it diagnoses.

  Avni has a determinedly common-sense approach to his topic and avoids psychological jargon (for example, Krogius' book talks about "failures of selective attention", not a form of words I found easy to get on with). He has collected together dozens of examples of chess blunder and oversights and has assigned them to three main classes according to the suspected cause of error: the opponent (e.g. when the opponent plays incomprehensibly), the position (e.g. when the situation looks familiar) and more general failures of thinking (e.g. neglecting defence while attacking). He also talks interestingly of integrating cues from different sources. This catalogue may sound like Abrahams' 'varieties of error', but Avni is much more clear and robust, less anecdotal, than Abrahams' tour.

  Each category of error, and each point he makes throughout, is illustrated with two or three examples from actual games, each with a diagram. This makes much of the book possible to read without a board. Moreover, Avni does not intend to allow you to read passively, but continually presses you to apply yourself to problems and exercises.

  Avni has done his homework very well: he has clearly trawled the chess press for good examples but has also read widely among the various books and papers that have been published on chess psychology. The book is packed with titbits from journals and books, as well as Avni's own chatty but pointed observations. From his collation I was interested in de Groot's notion of discrepancy, which was new to me: de Groot wondered 'when does a chessplayer look for an improvement to his intended move?' and concluded that it is when the result of analysis is at odds with the player's expectations - "the discrepancy being a signal that something is probably amiss" [1] (author's italics).

  [I don't know how far this explains the K-K incident: was there no discrepancy for the players? Or, for a GM Caro-Kann player, was a3 just the sort of move that White shouldn't be allowed to get away with? Nonetheless, this idea of discrepancy has given me food for thought in looking at my games.]

  The book concludes with a series of about a dozen recommendations for players and coaches which I found very helpful. It is so easy to recognise a chess disaster, so hard to prevent the next one. For example, to Avni and his colleagues we owe the discovery that strong, experienced chess players score higher for paranoia on the MMPI[2] than do weaker players and non-players. His recommendation: adopt a paranoid approach (at least at the chessboard). An example of how this might be done: rather than comfort yourself with the thought that you cannot possibly lose, think instead how could I possibly lose this game? Search for a possible catastrophe.

  This advice is, as ever, illustrated with examples: the one which made me wince was from Nigel Short, whose paranoia must have been running particularly low against Belyavsky at Linares in 1992:

1 Nd5 f6+, 2 Ke6?? Bc8 mate...

  A couple of stray thoughts:

  1. His evidence for his suggestion that (amongst other habits) we should cultivate paranoia this is essentially negative: he gives further examples of failures of suspicion which led to instant losses by strong players. Perhaps, since disease is so often the route by which we can gain insight into normal functioning of the body, this is quite right, but I am dogged by the suspicion that these are actually examples of unusual and unsuccessful chess thinking, and we know little of the usual kind. I know of a few examples of normal or more successful chess thinking: I remember some old articles by Simon Webb in the Chess magazine of the late 1970s under the title 'How do chessplayers think?' which tried to capture (live) some chess thinking, the fascinating (if retrospective) comments that were the appeal of the BBC Master Game series, and a couple of odd comments by annotators (like Fischer's "...Trifunovich seemed too quiet all of a sudden" [3] ). Perhaps readers know more. But my judgement is that, even in chess, which seems to me an ideal arena for research, we know little about how we go about chess decisions, what makes a difference between weak and strong players, and how to improve. This, really, is what makes a book like Avni's so interesting.

  2. A technique that Avni uses more than once is to give you similar positions and ask you to puzzle out which one has the catch. This is very different to the usual type of calculative exercise that I have seen (as in the Winning Combinative Play or Find the Winning Continuation columns of magazines). Knowing there is a problem which has a solution, I can sit and puzzle out the answers to most of these up to, say, five moves deep[4]. And yet, like many club players, my games are pocked with oversights two moves deep. Even some of the games I survive can be busted in short order by my modest computer. Now, if books on tactics had half their positions with only tempting but bogus tries, or a third of positions which were promising but which had no tactical solution yet available, this would be a much sterner and more realistic test. Does any publisher wish to take this on?

  I had only a couple of minor quibbles, but the conscientious reviewer feels obliged to note them:

  1. There are of course many different things to consider when choosing a move, and unless you find yourself with time to spare in each game (not true for me, I'm afraid) you may have to do something else less often or more quickly if you are to become more self-critical or more chessically paranoid. At the moment I fear that after reading Avni I will only make judgements more accurately while calculating less deeply, unless I can find a way to analyse faster. Avni does not try to address this issue, but rather treats the topic of developing a sense danger as a separate add-on to whatever you do normally.

  2. I did wonder if the number of pages per pound is good value. As chess publishing has become more computerised, with great benefits in speed and accuracy, I had hoped that some savings might trickle down to the purchaser. I think that chess books have been getting worse value rather than better, but perhaps this is merely my chessplayer's paranoia!

  3. There were some recommendations I expected to see and didn't! They are: write your move down first and check it (which I owe to Simon Webb), and when analysing, always look one move further for at your opponent's possibilities at the 'end' of a combination (Dvoretsky).

[See also An important note about Blumenfeld's rule -- DR]


[1] Thought and Choice in Chess, 1965.

[2] One jargon term Avni slips in without explanation: the Minnesota Multiphasic Personality Inventory.

[3] My Sixty Memorable Games. Faber. (p.203).

[4] I can even survive some of the examples from the section on 'Traps' in Livshits' Test your Chess IQ: Grandmaster Challenge (Cadogan, 1993), an otherwise excellent book, because firstly, the fact that they are in a section on "traps" turns the paranoia dial up to 11, and secondly, he offers give-away comments like "White was hoping for 1...Bxf4, on which he had prepared a counter-blow. What was it?".