Computer beats book, again
Submitted by
on Sun, 04/19/2009 at 7:55am.
Another in the continuing series of the computer disagreeing with GM book analysis.
Soltis, The Wisest Things..., p. 13, Karpov - Polugayevsky, QF6, 1974
First of all, the GMs in the press room were right to prefer 27. Rxf6 over Karpov's 27. Bf4.
But more importantly, Soltis gives 33. Nf5 an exclamation point, where the computer gives it a ?. Oh, it wins (+2.67). But 33. Nxf7! is crushing (+12.05).
It might be fair to give an ! to a move which "still wins" but which wins in a beautiful or instructive way, I guess. But GM Andy Soltis could have put it in the computer and found 33. Nxf7 in 5 seconds.
The silicon beast humbles us all.
Oh, and don't let this put you off buying and studying Andy Soltis' The Wisest Things Ever Said About Chess. It's still an instructive and helpful book.