Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: It may be that Aristarch 4.5 is as strong or even stronger than Ruffian

Author: Uri Blass

Date: 14:48:35 10/22/04

Go up one level in this thread


On October 22, 2004 at 17:36:43, Uri Blass wrote:

>On October 22, 2004 at 16:17:16, Tord Romstad wrote:
>
>>On October 22, 2004 at 15:31:20, Dann Corbit wrote:
>>
>>>http://wbec-ridderkerk.nl/
>>>See "Premier Div." drop down.
>>>
>>>Elostat output:
>>>
>>>   Program                   Elo    +   -   Games   Score   Av.Op.  Draws
>>> 2 Aristarch 4.50          : 2714   72  81    64    65.6 %   2602   28.1 %
>>> 6 Ruffian 2.1.0           : 2659   80  61    64    57.8 %   2604   40.6 %
>>>
>>>A lot of uncertainty remains, clearly, from the size of the error bars.
>>
>>It seems to me that Ruffian 1.0.5 is still clearly stronger.  Ratings from a few
>>of the most
>>popular public rating lists (notice that in all lists, the error bars are much
>>smaller than the
>>above):
>>
>>                        Rating    +   -  Games   Score   Av.op   Draws
>>UEL:
>> 1 Ruffian 1.0.5          2654   18  23   879    66.2%    2537   27.2%
>> 3 Aristarch 4.50         2602   26  27   507    59.8%    2533   27.6%
>>
>>AEGT:
>> 2 Ruffian 1.0.5          2760   46  54   152    63.8%    2662   25.0%
>> 3 Aristarch 4.50         2749   47  52   152    62.2%    2663   25.7%
>>
>>HvK Nunn list:
>>10 Ruffian 1.0.5          2754   19  24   762    66.0%    2639   29.9%
>>15 Aristarch 4.50         2692   18  25   882    67.1%    2568   24.0%
>>
>>YABRL:
>> 7 Ruffian 1.0.1          2649   16  22  1015    67.7%    2520   27.9%
>>10 Aristarch 4.50         2614   17  19  1117    60.8%    2538   26.6%
>>
>>Looks rather  convincing to me.
>
>Not to me
>I guess that all these lists except AEGT are blitz and the gap in AEGT is the
>smallest.
>
>Ruffian may be better in blitz than Aristarch but not better than it at longer
>time control.
>
>I believe that the rating lists that you  give are based on ponder off games.
>WBEC is based on ponder on games and I read
>the following rating list by Frank quisinsky in the arena forum based on ponder
>on games at 10/40 time control.
>
>4 Aristarch 4.50                 : 2685   54  53   130    58.1 %   2629
>8 Ruffian 2.1.0                  : 2657   58  45   130    53.8 %   2630   33.8 %
>
>
>Uri

Here is another one at slow time control( 80+3)

http://www.digichess.gr/infiniteloop/nil/NIL6_rat.txt

-2     3 Aristarch 4.50                 : 2737  153 115    20    60.0 %   2667
40.0 %
-1     4 Gothmog 1.0 beta 7             : 2720  149 146    20    62.5 %   2632
25.0 %
 0     5 Chinito 3.4c                   : 2712  153 133    20    60.0 %   2641
30.0 %
 0     6 Ruffian 1.0.1                  : 2702   39  56   179    70.9 %   2547
25.7 %


You can see again that Aristarch is better than Ruffian
Aristarch played only 20 games so it is not convincing but it is not surprising
that you cannot find a lot of games in slow time control  and at least the
evidence that I see does not support the claim that Ruffian is better than
Aristarch at slow time control.

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.