Author: Tony Werten
Date: 04:38:55 04/23/01
Go up one level in this thread
On April 22, 2001 at 22:28:27, William Bryant wrote: >I'm looking for suggestions or comments about handling passed pawns in the >program. > >The specific area of trouble is when doubled pawns are passed. > >At present I only extend on the most advanced of the doubled passed pawns >and only evaluate the most advanced of doubled passed pawns as a passed pawn. >The other pawn is simply a doubled pawn. > >When testing this on certain test suites, the version that considers both >doubled pawns as passers did a little better, which is counter-intuitive. It >needs a lot more testing, but I'm look for comments or previously gained >experience. You have found the danger of testpositions. What you are doing is logical. Check your code if it is doing what it's supposed to do and ignore the testresults. If you "correct" your program to the testresults, you will have a nice testposition solver but a lousy chess playing program. You should use testpositions only to see if you broke something after a change. cheers, Tony > >William >wbryant@ix.netcom.com
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.