Author: K. Burcham
Date: 06:46:30 12/20/03
Go up one level in this thread
1.we know the GM will try to play for winning position out of opening. 2.we know that sometime the GM has studied the opening book of the program. 3.we know that sometime the GM has studied certain moves in middlegame that the program will play, that gives the GM advantage. 4.we know that in closed position with walls of pawns, programs seem to have no plan. In other words if the GM plays move 14 out of book, and he sees he has no advantage, he will worry about possibility of a loss. if the GM has not studied the program and/or cannot find a position that he thinks he can get the program to play so as to gain advantage, then he will worry about the possibility of a loss. if the program exchanges pawns in opening, opens files instead of advancing, GM will consider this a disadvantage for him and will worry about a loss. if these things happen, GM will go to a draw mode and try to get program to play into a draw. If we eliminate these problems, then I feel several free chess engines can win against GM. I know that some will say but these problems are what chess is all about. this is true, but I think these problems can be improved in code. it would be interesting to see ICC records for free chess engines against GM. I know that GM prefer to play programs at 3000 rating at ICC. maybe other programs can win against GM at ICC but rating is too low, such as 2500. kburcham
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.