Author: Will Singleton
Date: 11:37:42 03/20/02
Go up one level in this thread
On March 20, 2002 at 12:59:58, Gian-Carlo Pascutto wrote: >[D]4r1k1/pp2q1p1/2p1P1Q1/3p4/6R1/P7/KPP5/8 w - - 0 17 > >I get a static eval of +3.7 here, which drops a little >bit after searching a minute or so (to +3.2). > >Shredder Paderborn stays around +0.5 - +0.7 > >Fritz 7: about +1.2 to +1.4 > >Crafty 18.13: +0.30 to +0.40 > >Comet B41: rises to about +1.5 > >Hiarcs 7.32: rises to about +1.0 > >ChessTiger 14.0: starts from 0 and goes to +0.40 >(Gambit Tiger very similar) > >Gandalf 5.1: +1.0 to +0.80 > >So, it seems my engine is way overestimating white's attacking >chances here. On the other hand, white's threat is pretty nasty. >I think part of the problem may be that my thing doesn't really >consider how well the defenders are placed. Perhaps a queen >and a rook only don't justify such large scores either. I'm curious >_how_ other people evaluate this. > >-- >GCP This is similar to a position I posted awhile back, in which all the engines I tested scored near even but sjeng was +3. I don't think it involved king-safety, but had queens on the board near endgame. Can't find it. In this position, Amateur says +0.07 static, and about +0.80 after a minute search. Most of my king-safety stuff is washed out by the lack of pieces. Will
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.