Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: RK 2004 (90m+30s) Round 90 (of 140) 360 games of 560

Author: Kurt Utzinger

Date: 01:40:44 04/30/04

Go up one level in this thread


On April 30, 2004 at 03:53:09, Graham Banks wrote:

>On April 30, 2004 at 03:41:50, Kurt Utzinger wrote:
>
>>On April 30, 2004 at 03:17:16, Jouni Uski wrote:
>>
>>>On April 30, 2004 at 02:32:48, Kurt Utzinger wrote:
>>>
>>>>On April 30, 2004 at 01:31:55, Jouni Uski wrote:
>>>>
>>>>>Ruffian is improving slowly! And if tested 2.0.2 would score better.
>>>>>
>>>>>Jouni
>>>>
>>>>     Hi Jouni
>>>>     We have not (yet) tested Ruffian 2.0.2 but should be
>>>>     interested if you have some evidence for your opinion.
>>>>     Other people argue that v2.1.0 would definitely be
>>>>     the strongest version.
>>>>     Kurt
>>>
>>>There seems to be enough evidence based on >1000 games:
>>>
>>>Chessfun:
>>>
>>>15 Ruffian 2.0.2                  : 2606   44  34   210    56.2 %   2563   40.0
>>>%
>>>19 Ruffian 2.1.0                  : 2595   51  34   180    51.4 %   2585   40.6
>>>%
>>>
>>>Michael 42:
>>>
>>>13 Ruffian 2.0.0                  : 2664   20  24   646    43.3 %   2710   36.2
>>>%
>>>15 Ruffian 2.1.0                  : 2654   23  27   532    43.0 %   2703   35.9
>>>%
>>>
>>>Frank Q:
>>>
>>>Ruffian 2.0.2 against Shredder 8:    15.5 - 24.5
>>>Ruffian 2.1.0 against Shredder 8:    09.5 - 30.5
>>>
>>>These are mostly ponder of but I don't think it matter ANY!
>>>
>>>Jouni
>>
>>
>>      Hi Jouni
>>      Many thanks for these additional information.
>>      Kurt
>
>
>These might all be at one particular time control (?) and the difference in
>rating is minimal anyway.
>It could very well be that Ruffian 2.1.0 is superior at longer time controls
>although initial SSDF ratings don't support this. However I haven't looked at
>the margin of error.
>The fact that the Ruffian programmer preferred Ruffian 2.1.0 at Leiden seems to
>suggest that he feels this version is stronger.
>Graham.


     Hi Graham
     You may be right. And do we not forget the book question.
     In our "RK 2004 Tournament" http://www.utzingerkurt.com/rk2004.htm
     Ruffian 2.1.0 is using the Leiden book which is restricted to play
     1.d4 as the first move. And this could perhaps have a negative
     influence on the results of Ruffian 2.1.0.
     Kurt



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.