Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: SSDF Rating list

Author: enrico carrisco

Date: 19:02:12 12/27/01

Go up one level in this thread


On December 27, 2001 at 19:12:02, Gian-Carlo Pascutto wrote:

>On December 27, 2001 at 19:06:38, Mark Young wrote:
>
>
>>What do you mean.... Just noting that SSDF now shows ChessTiger 14 to be the
>>strongest program.
>
>The SSDF gives ChessTiger the highest rating. However, looking at
>the error bars, we easily see that does not necessarily mean it
>is the strongest program.
>
>As I said before, if anything, the list shows they are of equal
>strength.
>--
>GCP

Exactly why I've always felt that for the majority of viewers, the SSDF list is
dangerous.  A new name appears at the top of the list or another moves to the
top of the list with a +4 rating and gross generalizations are spewed.  The list
is a great reference and very interesting for the games it produces, but many
view the "scoreboard" as a tell-all.

Dan -- breakout the multivariate analysis of variance -- we'll get to the bottom
of this. :)

-elc.



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.