Author: Don Dailey
Date: 14:24:25 01/03/98
Go up one level in this thread
>>Here is a thought on the isolated pawn thing. Like any pawn structure >>thing a good program will have different values depending on file >>and rank. For instance a backward 'a' pawn is not nearly as serious >>as a backward d or e pawn. >> >>-- Don > > >I'm not sure I agree with this. Generally an isolated pawn on the edge >is not quite as bad as one in the middle, because it can only be >attacked >from the front/back and one side, rather than both sides. but I've lost >plenty of games from a weak a pawn. (weak = isolated, backward, and >so forth).. > >I think the file is less important than other things about the pawn... >ie >is the file half-open and does the opponent have rooks to attack down >that >file? etc... Bob, We have a lot of different categories for pawn weaknesses. We consider the file they are on, backward, isolated, restrained, on open files with and without major enemy pieces, part of majority, pawn islands and others. Is the file less important? I could only tell you if you gave me a specific example. Larry Kaufman noticed a long time ago that all our bonus and penalties did not apply equally to pawns on flanks. Our current implementation gives a and h pawns about half of the penalty the others get but it depends on which specific weakness you are talking about. This was a big improvement to our program. Also the degree of advancement is important. The famous isolani is an example of a pawn weakness that also is a dynamic strength and often programs play these positions well. Advancement is a very important concept in isolani positions. -- Don
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.