Author: Stuart Cracraft
Date: 13:05:54 06/18/99
Go up one level in this thread
On June 17, 1999 at 10:10:43, KarinsDad wrote: >>Recommendation #3: Set up a pawn structure hash table. Since pawns rarely move >in the game (and in the search <40% on average), it is probably better to >calculate pawn advantages and deficiencies once per given pawn structure as >opposed to calculating it every time in your evaluation. > Does not work. The pawn structure hash table is independent of anything except pawn structure. If you try to fold in some representation into that score for the given pawn structure of the board representing the score increase or decrease based on square control for any other piece except pawns, it will have no bearing on all other positions where there is a different arrangement of pieces but the same arrangement of pawns. That's why backward pawns can be calculated and put into the pawn structure hash table but no score corresponding to control of the squares in front of the backward pawn nor the backward pawn itself (i.e. control of). Non-pawn-related factors have to be calculated after the pawn structure hash data structure's score is either retrieved successfully or calculated for the first time.
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.