Author: Jon Dart
Date: 20:42:48 11/03/05
Go up one level in this thread
On November 03, 2005 at 15:02:13, Robert Hyatt wrote: >I am not sure this is an "eval test". A static eval by crafty says -2.7 on the >latest version. But tactics show that black's pawns are too far advanced and >are going to mostly disappear and get blockaded while white's file control is >hard to deal with... > >This looks more like searching deeply enough to see the resulting positions, >rather than evaluating what the resulting positions _might_ look like later... True enough, but my experimental version had a +4 pawn bonus or so for the advanced pawns, which is too much I think (the released version has less), so that was wacky. (A smart program might take into account White's blocking chances). Even leaving that aside, I can easily get big score swings depending on how I rate the White and Black king attack scoring. Static eval interacts with search, among other reasons because you can cutoff search if your scores get too big without some good reason. --Jon
This page took 0.02 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.