Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Hiarcs 7.32 smashing Fritz 5 at 40/240 (games 5-7)

Author: Pete R.

Date: 16:11:41 08/19/99

Go up one level in this thread


On August 19, 1999 at 18:14:06, Mark Young wrote:

>Hiarcs 7.32 PII 450, 200mb Hash
>
>Fritz 5.00/16bit PII 450, 200mb Hash

Are you going to test against Fritz 5.32 as well?  I think it's significantly
faster than the 16 bit version.

I've been using both Fritz 5.32 and Hiarcs 7.32 to help analyze the Kasparov vs.
the World game.  While Hiarcs is slow compared to Fritz, after extensive use of
both programs on a large variety of positions, I trust Hiarcs' evaluation more.
Fritz has the particularly annoying quirk that seems to be invoked in positions
with king check (I can supply a sample position if anyone cares). Fritz will
give, say, a 0.05 eval for the checking move, then you make the recommended move
for the king to get out of check, and all of a sudden the eval is 1.15 or
something.  This is extremely annoying when you are trying to find the best
lines from a given position.  In fairness on other positions Fritz has been more
helpful, even doing a better job with some endgames without even using
tablebases.  But it is susceptible to these sudden changes in eval in a way that
Crafty and Hiarcs are not.



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.