Computer Chess Club Archives




Subject: Re: Tactical monsters: Fritz5.32 and Goliath 2.05

Author: Shep

Date: 01:58:10 08/04/99

Go up one level in this thread

On August 03, 1999 at 14:22:26, James Robertson wrote:

>On August 03, 1999 at 03:23:49, Jouni Uski wrote:
>>I K6-2/450 both have average NPS around 450 000, but peaks up to 900 000.
>>After running about 500 test positions it's still impossible to tell which one
>>is better tactically: both were best in 3 test suites from 6. Some examples:
>>        Fritz      Goliath
>>BS2830  22/27      15/27
>>BT2630  24/30      28/30
>>ECMF5   157/216    164/216    (1 min/pos)
>I think Fritz is. Goliath is a 'testsuite' monster, but in actual play doesn't
>do that well; I have played many games between Crafty and LGG, and often Crafty
>sees tactical wins one or sometimes even two(!) whole moves ahead.

I suppose LGG has some "weird" search extensions that work pretty well in
tactical test positions but wreak complete havoc in ordinary play (I have some
ideas in that direction myself which would boost results in test suites but are
probably complete nonsense in 99% of the situations).


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.