Author: Alessandro Scotti
Date: 07:53:15 08/24/04
Go up one level in this thread
On August 24, 2004 at 04:57:02, Tord Romstad wrote: >Making the static eval aware of its limitations offers many interesting >possibilities, and I think there are many valuable and important ideas >waiting to be found by the adventurous programmer here. The basic >idea is to extend in positions where the static eval is likely to be >highly inaccurate, and to reduce in positions where it is likely to >be very accurate (internal node recognizers is an extreme special case). One thing I have played a little with is using the 50-move rule counter as an indicator of whether a line of play is tactically interesting or not. After all if the counter grows then there are no captures and no pawn moves, so it's basically just piece shuffling. At this point one could decide for example to skip the quiesce search at all and using the static evaluator. I kept the code for quite a while actually, but eventually I removed it after a couple of serious blunders. I should try with a SEE though. I also tried to find a correlation between said counter and the probability of a move to cause a cutoff at some depth. After a promising start, eventually I had to put this idea aside too. Anyway, I performed these experiments a few bugs ago so I might try them again in the future...
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.