Author: Sune Fischer
Date: 12:30:58 09/14/04
Go up one level in this thread
On September 14, 2004 at 04:58:30, Martin Slowik wrote: >Hello Sune, > >I found your discussion with Bob very interesting and enlighting, even more so >(but not only :)) because my approach to chess programs seems to be similar to >yours. We rule. :) >A couple of weeks ago I tested all my engines in a double round robin tournament >starting with a line of the Leningrad Dutch (1.d4 f5 2.c4 Nf6 3.g3 g6 4.Bg2 Bg7 >5.Nc3 d6 6.Nf3 0-0 7.0-0 Qe8 8.d5 a5). > >First of all it was fun to watch those programs struggling in the same murky >positions I usually find in myself. :) But then I obviously wanted to know which >program to trust most when analysing those positions. Do you think you can guess >which program won that tiny experiment? Well, to make it short: 'twas Shredder >(followed by Hiarcs), great surprise, hm...? ;) It wouldn't be surprising if the best tournament engine was also the best analysis engine. I'd expect a very strong correlation :) >Looking closer at the positions revealed that in many of the games the endgame >abilities of the participants played a crucial role, to say the least. > >So may I ask you, how do you test the middlegame abilities of your candidates? That's a tough question, I don't have a very good recipe for that. I think the endgame is easier to analyze. >And, is there perhaps a web page of yours with some test results? Sorry there is no webpage. I don't think it would be very interesting because I use mostly older versions of other engines. If I update the opponents to newer versions then I can no longer compare the results with old versions of frenzee. -S. >Best, >Martin
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.