06 December 2010

An Olympiad Bind

As I predicted in Fischer - Najdorf, 1962 Varna, the position in the diagram turned out to be the critical position in the game. The game continued 12.Re1 e5 and, as happens so often in chess, there is a monumental battle hidden beneath those two moves.

Varna Olympiad 1962
Najdorf, Miguel

Fischer, Robert
After 11...d6-d5

Both Kasparov and the computer believe that 12.Bb3 is better than Fischer's 12.Re1. Fischer chose an attacking move over a defensive move, a choice upheld by some fundamental conviction that this was the best way to overwhelm his opponent. The problem with 12.Re1 is that it allows 12...Bxg2, a desperado blow that muddies the waters by exposing White's King to a counterattack. Fischer rejected this move and explained in a comment that

12...Bxg2 13.Kxg2 dxc4 14.Qf3 Nd7 15.Nf5 Rg8+ 16.Kh1 e5 17.Be3 [produces] a winning bind despite the two Pawn deficit.

Kasparov repeated this analysis and added,

However, in the opinion of Dr. Huebner, after 17...Rc8 18.Rad1 Qc7 19.Rd5 Qc6 20.Red1 Rc7 21.Kh2 Qb7, Black's defenses are solid.

In addition [after 12...Bxg2 13.Kxg2 dxc4 14.Qf3], Black has the stronger 14...Ra7! 15.Be3 Rd7 16.Rad1 Qc8, when White may not have sufficient compensation for the sacrificed material. 'Fischer once again overestimates the strength of his attack', Huebner sums up, 'and fails to explore the defensive potential of the enemy position.'

I subjected this analysis to the engine stress test and discovered a number of points.

  • First, the two lines given by Kasparov are the two fundamental defenses at Black's disposition. The move 14...Nd7 brings the Knight into play, retaining the Rook for future action, while the moves 14...Ra7 & 15...Rd7 bring the Rook into play, retaining the Knight. Black has time to do one or the other immediately, but can't do both.

  • Second, at every point in the play, Black has the zwischenzug ...Rg8+. This brings the Rook into play on a natural square and forces the choice on White of escaping the check by Kf1, Kh1, or Kh2; the relative tradeoffs of these three moves are not easy to calculate. Fischer introduced this resource in his analysis with 15...Rg8+ 16.Kh1, while Kasparov / Huebner avoided it. The computer wants to play the Rook move at almost every point where it has the option.

  • Third, Fischer stopped his analysis at 17.Be3, avoiding many of the subtleties in the variation, while Kasparov / Huebner continued. In their first line, at the point where 'Black's defenses are solid', White has the surprising 22.Bh6, where the engine gives White a large plus in its evaluation. In their second line, White has 16.Rac1 instead of 16.Rad1, when Black is still struggling to coordinate his pieces and find safety for his King. I could easily turn Huebner's words on him with, 'Huebner once again underestimates the strength of his attack, and fails to explore the offensive potential of his own position', but that would be introducing an unwarranted generalization.

The point that impressed me the most in the time I spent on this game was that Fischer summed up his position with the phrase 'winning bind', and left it at that. Kasparov / Huebner attempted to show by calculation of variations that Black has defensive resources, and while they succeeded in doing that, they didn't refute Fischer's claim of a 'winning bind'. The word 'bind' is a positional concept that is best tackled by verbal analysis -- explaining 'Why is there a bind?' and 'What can be done to break the bind?' -- rather than by a nonverbal calculation of variations. I once wrote about this in Chess Tutorial : Bobby's Binds, now available only on Archive.org:-

Binds are fairly common in chess. We don't often see them in master games because good players prefer to avoid getting into binds and will even sacrifice a Pawn to steer away from them. To find good examples of binds, we have to look at the notes to master level games. One excellent source for examples is Bobby Fischer's 'My 60 Memorable Games'.

Binds are an area where computer analysis is not always helpful, where the master sees easily what the engine never sees. It is a chess equivalent of not being able to see the forest because of the trees.


Later: One of the posts featured in Chess improvement blog carnival #1!, a good start for a new blog carnival.

No comments: