Author: Kurt Utzinger
Date: 03:40:42 02/17/05
Go up one level in this thread
On February 16, 2005 at 18:33:05, stuart taylor wrote:
>On February 16, 2005 at 13:48:03, Kurt Utzinger wrote:
>
>>On February 16, 2005 at 13:36:53, Graham Banks wrote:
>>
>>>On February 16, 2005 at 06:57:48, Kurt Utzinger wrote:
>>>
>>>>Test: Shredder 9 UCI by Stefan Meyer-Kahlen
>>>>Time control: 40 min/40 moves
>>>>Hardware: Ath 1.3/64 MB hash
>>>>ponder=off, 3-4 men EGTB, Fritz8-GUI
>>>>book=5moves.ctg, normal, no learning
>>>>Details and games download at:
>>>>[http://www.utzingerk.com/shredder9_test.htm]
>>>>
>>>>
>>>> Program Score % Av.Op. Elo + -
>>>>
>>>> 1 Shredder 9 UCI : 89.0/150 59.3 2467 2533 50 44
>>>> 2 Junior 9 : 23.0/ 50 46.0 2533 2505 75 96
>>>> 3 Fritz 8 : 19.5/ 50 39.0 2533 2455 75 88
>>>> 4 Chess Tiger 15.0 : 18.5/ 50 37.0 2533 2441 76 85
>>>
>>>
>>>Hi Kurt,
>>>
>>>will you also run 50 games against Shredder 8?
>>>
>>>Graham.
>>
>>
>> Hi Graham
>> No, we are not going to do this. It has always been our
>> policy to never let play new vs old version of a chess
>> program. In our opinion this does not make much sense.
>> The next matches will run vs Gandalf 6, Hiarcs 9, CM9_T05,
>> Ruffian 2.1.0, Aristarch 4.50, List 5.12, Fruit 2.0
>> Kurt
>> Kurt
>
>I cannot understand why this does not make much sense.
>
>The above programs, Shredder 9 will certainly beat handily, because even
>Shredder 8 and Shredder 7.04 do, so what are we proving?
>S.Taylor
Hi Stuart
Almost no doubt: Shredder 9 UCI will beat all the
above-mentioned coming matches. We are thus proving
that Stefan Meyer-Kahlen's engine will be clearly
the best chess program of the world. Why we are
never testing new against old version(s) of the same
program is simply a matter of taste and due to the
fact that the increase of strength can't be measured
in such matches. I have posted this often enough:
a special CM8-setting got an average score of 85 %
against default and other CM8-settings, but the same
setting did much worse than other settings vs
non CM-chess-programs. This experience was the main
reason why I think it does not make (much) sense to
test new vs old version of the same program. And further-
more: the latest version is usually stronger than its
predecessor.
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.