Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Voting...

Author: Don Prohaska

Date: 20:55:58 04/08/98

Go up one level in this thread


As been idicated somewhere before, the SSDF doesn't test a Chess
Engine's strength, but a Chess Programs system's strength. The problem
with the SSDF is there is a mixture. A strong engine with a weak opening
book (or whatever else there is) will most likely show poor results. I
don't know what all the fuss is about. As far as I can see, the SSDF is
nothing more than a guide. All the top programs are pretty close in
strength. I'd pick Rebel9 and FRitz5 because they are both very strong
and give lots of extras even if some other programs scored higher.
Different people have different needs. Let SSDF have two sets of tests,
one for the engines and one for the systems. As far as voting goes, I
don't see any harm. Another, "what's the fuss!"

On April 08, 1998 at 21:50:28, Steven Schwartz wrote:

>On April 08, 1998 at 19:06:14, Moritz Berger wrote:
>
>>2nd thought: Why not have a discussion here first about the topic of the
>>poll?
>>
>>Maybe find out first how important the SSDF results are to our
>>community? i.e. the question might be:
>>
>>How important for you is the playing strength 'feature' when you buy a
>>new chess program? (scale lowest to highest from 1 to 10)
>>
>>How important are the SSDF ratings for your evaluation of a programs
>>playing strength? (1-10)
>>
>>How important are chess server ratings of commercial programs in your
>>decision? (1-10)
>>
>>etc.
>>
>>
>>Just an idea about some 'harmless' questions to ask ...
>>
>>Moritz
>
>These are fine questions, but to try to determine what
>questions were of interest to the CCC members, I posted
>on April 1st here on CCC the following...
>
>"We would like some input from the members as to
>questions that might appear on the ballot. If you have
>any questions that you would like submitted to the
>membership for a vote, please respond here or by
>private email (yourmove@icdchess.com)  Please keep the
>questions focused on the issue of computer chess. We
>cannot promise that your questions will be used, but
>they certainly will be considered."
>
>As it turned out, we received many suggestions, but
>the Fritz issue was the one most often mentioned. We are
>not etched in stone on the questions that will be posed,
>but it appears that is what the members wanted.
>
>Our plan was to have a single question at first to
>check the polling system's reliability and then post
>multiple questions. There should be enough questions to
>go around eventually, but we had to start somewhere.
>
>It is clear to me that in the world of computer chess,
>there is never much chance of getting two people to
>agree on anything. Much of it appears to me to be cultural,
>but we have to start somewhere; are you willing to give this
>project a chance?
>- Steve



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.