Author: Uri Blass
Date: 00:28:05 11/17/03
Go up one level in this thread
On November 17, 2003 at 00:32:40, Terry McCracken wrote: >On November 16, 2003 at 17:43:10, Hristo wrote: > >>The computer played badly ... really bad ... >> >>Regards, >>Hristo > >Actually the computer played well....well within the parameters of it's >"understanding" of the game:) > >This is something programmers have to solve, and I believe they will, someday. >This game was revealing, and it will help computer chess in the long run. >This game was great!;-) > >Regards, > Terry I looked at the evaluation of Fritz and shredder and I found that they evaluated the position some moves before kasparov resigned as almost equal. I do not understand why the top programs cannot understand something that even my stupid program can understand. Some facts that should be easy for chess programs to see few moves before Fritz resigned: 1)White was pawn up. 2)No square near the white king was controlled by the enemy pieces so there is no logical reason for a big king safety score against white. 3)White had advanced protected passed pawn at b6. I think that all these factors should tell chess programs to give clear advantage for white but for some reason they do not do it and I do not understand the reason for it. I am interested to know what is the reason for the bad evaluation of top chess programs that even a stupid bean counter seems to be smarter than them. Movei can see scores above +1 from move 35 At move 35 it needs to search for some minutes to see above +1 score and I did not check previous moves. Even before capturing a6 it is clear for it that white has a big advantage. White has a clear positional advantage thanks to protected passed pawn at b6 and the black isolated pawn at a6. Uri
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.