Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Just comparing lists...

Author: Sune Fischer

Date: 03:09:04 10/25/02

Go up one level in this thread


On October 25, 2002 at 05:57:35, Uri Blass wrote:

>On October 25, 2002 at 05:50:07, Sune Fischer wrote:
>
>>On October 25, 2002 at 00:10:20, Aloisio Ponti Lopes wrote:
>>
>>I don't like format either, in principle Christophe could make Tigers 15_a, 15_b
>>and 15_c (minor updates of some form) and take the first three spots with,
>>basicly, the same program!
>>
>>It is pretty rediculous, one author - one entry!
>>
>>I think they are doing it like that for users with old versions, they would like
>>to see how much better the new version is, should they go out and buy it or not,
>>that sort of thing. For that you need the comparison with previous versions.
>>
>>Personally I am only interested in "how well does the lastest program X compare
>>to latest program Y". Older versions simply do no longer represent what the
>>author is capable of and are therefore NOT interesting.
>>
>>-S.
>
>They are interesting and there are cases when the older version is better so the
>new program does not represent what the author is capable because he proved that
>he was capable of more than the new version.
>
>Uri

I assume you are thinking of Crafty?
In that case I think Crafty is geared towards human games and not towards
eng-eng matches. Hyatt has added special anti-human knowledge AFAIK, my guess is
that that stuff is counter productive in eng-eng matches, but if the play
against humans is better then who cares about eng-eng matches?

Remember the SSDF measures only eng-eng performance.

I consider eng-eng matches only to be a tool in the development of the engine,
doing well in eng-eng matches is not a goal in itself IMO.

-S.



This page took 0.01 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.