Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Latista 1.3 released

Author: Tord Romstad

Date: 07:52:21 07/03/05

Go up one level in this thread


On July 03, 2005 at 01:28:56, Eric Oldre wrote:

>King Safety is the part of latista's evaluation that I'm
>most worried about. Sometimes it gives a really high score
>when it shouldn't, and also the low scores when it's own
>king is in deep trouble.
>
>looking at the static evaluations of each position of the
>game, In this game it seemed to do a decent job of properly
>evaluating the attack. but maybe others have an opinion too.
>
>Here Latista gave a score of 46 for the whites king attack
>[d]r1bk1r2/ppp1nBb1/3q2Qp/8/3n4/2B5/PPP2PPP/RN2R1K1 w - - 0 25

Glaurung 0.2.4 with the default king safety settings gives 87
centipawns for white's attack.

>The highest the evaluation ever came was in this position where
>white's king attack reached 72 centipawns.
>
>[d]r7/p2k4/1p2bq2/1N1p1r2/8/4Q3/PP1R1PPP/4R1K1 w - - 2 38

And here, Glaurung gives 1.53 centipawns for white's king attack.

Tord



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.