Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Fun Rating List including Ruffian 2.1.0

Author: Uri Blass

Date: 13:05:54 03/12/04

Go up one level in this thread


On March 12, 2004 at 16:02:18, Chessfun wrote:

>On March 12, 2004 at 15:00:31, Uri Blass wrote:
>
>>On March 12, 2004 at 14:22:29, Kurt Utzinger wrote:
>>
>>>Hi Sarah
>>>A fantastic performance of Rufian 2.1.0
>>>Kurt
>>
>>
>>11 Ruffian 2.1.0         : 2623   62  46   114    53.9 %   2596   37.7 %
>>17 Ruffian 2.0.0         : 2602   37  56   150    49.0 %   2609   40.7 %
>>21 Ruffian 1.0.1         : 2583   46  32   220    51.4 %   2573   38.2 %
>>
>>statistical error is bigger than the 40 elo
>>
>
>As I'm sure you're aware Uri bringing down the statistical error requires
>thousands of games, for example:
>
>On this same list is good old Fritz 6. And even after 1224 games still shows
>statistically that it could be 19 points stronger or 14 points weaker than it's
>current rating.
>30 Fritz 6               : 2543   19  14  1224    53.0 %   2523   37.5 %
>
>All I'm saying is playing another 200 games with this engine will not resolve
>that issue for you.
>
>Sarah.

I do not blame you of not playing enough games and I only responded to the
fantastic result claim.

If the result is really fantastic then there is no doubt that there is an
improvement.
If there is only small improvement I cannot define the result as fantastic.

Uri



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.