Author: Simon Waters
Date: 14:56:02 11/08/00
Go up one level in this thread
The assumption that a chess computers move would be deterministic is likely to be false. Typically the programs are complex and run on complex operating systems, so a machine may make a different move given the same time and same position, as the operating systems eats a few extra milliseconds doing some unforeseen system thing (Mind likes to unload kernel modules after a few minutes). Similarly if it is thinking in your time you would have to make moves at exactly the same rate. That said you may find flaws in an opening book, but unless you actually like the alternative lines the machine may bail out into, playing to exploit this is risky. Advanced Star Chess use to play the black side of the Panov-Botvinnik attack in the Caro-Kann Defense (The one with white Qb3, Qb7, et al as Fischer played in M60MG's). However it had no clue how to play the ending after the queen exchange, and even managed to get itself mated about 8 moves after the exchange in one memorable game. One could do worse than study other GM games against machines, especially where the GM's won, and ponder the machines anti-stonewall lines (It'll have some 8-). I have some collected anticomputer ideas on my web site - but they weren't aimed at GM's (I wouldn't presume). http://www.wretched.demon.co.uk I have a small collection of GM v Computer games that I used in preparing the web site. Your welcome to have a copy, but nothing there that a quick search of any decent Chess database shouldn't throw up. I particularly like the Polgar game on the website - although I'm not sure I can recommend her general approach - other than the playing of exceeding good chess.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.