Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: SSDF ... a question

Author: Dann Corbit

Date: 13:24:32 09/08/03

Go up one level in this thread


On September 08, 2003 at 16:19:06, Bruce A Rhoades wrote:

>"In other words, Fritz 5.32 will always play the same on a P90 so we can
>reproduce the results.
>
>What do you expect the results to mean?"
>
>    Perhaps that's what I don't understand. The SSDF listing appears to indicate
>for instance:
>
>    Fritz 6 running on xyz pentium 1 cpu
>    Shredder 7 running on wxy pentium 111 cpu
>
>    etc.
>
>    If the programs are not running on PC's with the same
>configuration/capabilities how can the SSDF list makes sense?

Because there is no other way to do it that makes sense.

For instance, the above rating would tell you a rough expectation for:
Fritz 6 running on xyz pentium 1 cpu running against Shredder 7 running on wxy
pentium 111 cpu using an autoplayer.

Nothing more, nothing less.

Will we throw out 99% of the useful data when a new CPU comes out?  That is
simply absurd.

Every strength rating is a combination of program+hardware+timecontrol+playmode
along with any other relevant conditions.  You cannot extrapolate it to any
other condition set (except as a non-mathematical but perhaps reasonable guess).

The programs with the most games played produce the most accurate data for
calibrations.



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.