Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: American Chess Programs vs European Chess Programs

Author: Tord Romstad

Date: 12:45:54 01/11/04

Go up one level in this thread


On January 11, 2004 at 14:40:00, Anthony Cozzie wrote:

>On January 11, 2004 at 10:01:11, Tord Romstad wrote:
>
>>On January 10, 2004 at 23:36:19, Anthony Cozzie wrote:
>>
>>>Don't forget Zappa.  Its probably about 50 elo weaker than Arasan.
>>
>>This seems very unlikely to me -- I think that Zappa must be much better than
>>Arasan.
>>A couple of days ago, you wrote that Zappa was a bit stronger than Gothmog.  I
>>played
>>a blitz match between Gothmog 0.4.5 (which was released about an hour ago) and
>>Arasan
>>7.2 (not the most recent version, but still) last night, and Gothmog won by
>>68.5-31.5.
>>
>>Gothmog 0.4.5 is clearly stronger than 0.4.3 (which, I suppose, is the one you
>>have), but
>>it is not hundreds of points stronger.  My guess is that Zappa is at least 150
>>points stronger
>>than the latest version of Arasan.
>>
>>Tord
>>
>
>Actually I was running against 0.4.1.

That explains a lot.  Version 0.4.3 is a rather bit improvement compared to
0.4.1, and the
new 0.4.5 seems to be a big step upwards from 0.4.5.  The improvements consist
mostly
of bug fixes in the evaluation function and much more aggressive dubious forward
pruning.
;-)

>That was a special version of Zappa, though, which performed poorly overall - it
>just happened to do very well against Gothmog.

Yes, I understood that.  But you also wrote that the normal version of Zappa
usually scored
slightly more than 50% against Gothmog.

>Zappa is still at least 100 elo weaker than Crafty.

That's not bad at all!  I haven't tested against Crafty for a long time, but I'm
fairly sure
Gothmog is at least 250 points behind Crafty on a single-processor 32 bit
machine.

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.