Author: Severi Salminen
Date: 06:27:58 07/14/00
Go up one level in this thread
Hi! >I read some of the posts on this question and sit back and laugh. Any real >chess player who has played many games of chess with computers KNOWS THERE IS A >BIG DIFFERENCE between chess programs with respect to human style. Talk of >scientific proof of this is as silly as trying to prove that the great painters >created works of art. THE COMPUTER STYLE IS TO PLAY TO OPEN THE POSITION AT >EVERY OPPORTUNITY AND WIN BY BRUTE FORCE TACTICS ALONE! There is no attention >paid to bad bishops, knight outposts, pawn structure chess, etc. Hmm. Have you EVER looked at source code of (for example) Crafty? You'll find that the top chess programs consider the things you mentioned + much more. They DO consider pawn structures, pawn majorities, positioning of pieces... It is true that open positions are better for chess programs. But there is much more than just combinations leading to material advantage you have to consider in chess program's evaluation function. >Any class A or >better player can spot the difference. Hiarcs 7.32 is more human-like than >Junior 6, Junior 6 is more human-like than Fritz 6, Fritz 6 is more human-like >than Fritz 5, Fritz 5 is more human-like than Fritz 4. And what do you base these comparsions on?? The version number? >It's great to have the programmers giving input on this site, but they should >confine their comments to what they know, programming issues! When they start >talking about chess, they look as silly as I would if I started giving posts >about computer chess programming. 100% true, one should not comment on topics they don't know anything about... Severi
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.