Computer Chess Club Archives


Search

Terms

Messages

Subject: Some Results from test suites

Author: Joshua Lee

Date: 10:48:51 05/25/00

Go up one level in this thread


From the stuff i had tried Hiarcs played a master move around 8 or 9 ply and for
example results varied with endgame tests it would get a bunch right (Anemis 2 i
think)where the rating of the player was 2500+ but then i had the idea to start
at the problems where the humans where loest rated and it had difficulty at
around 2330 or so  this would correspond with results from that positional test
from SSDF
forgot the guys name. One problem is from a Reshevsky - Smyslov game where
Reshevsky played f4, Hiarcs liked c4 for 10 minutes but then at 10:02 it picked
f4
. at the time Reshevsky was 2415 (1991??) and smyslov 2530. so does it play
around 2415 or does it play 2930?  of course not ... Yet another result was a
move of a player rated 2335 that took 7 minutes to find a move better than the
one played and this is on an 800Mhz Athlon clocked to 880!!  I want to say
programs at 3 minutes a move are over 2200 but by how much i don't know. But
then again it was looking around 23 minutes for the opponent of Kotov and it
couldn't find anything much better than his 2160 rated opponent.... it would've
taken 13-15 ply i think.

Lastly LCT II results for Hiarcs are wrong it is definately 2690 on my machine
but i guarantee anybody that POS 14 Spassky Aftonomov  isn't solved for
23Minutes!!!
Spassky at the time was probably 2600+



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.