Author: Kurt Utzinger
Date: 00:14:06 01/06/04
Go up one level in this thread
On January 06, 2004 at 02:49:04, Sedat wrote:
>>The comparison between CM9_Blitz and CM9_Gladiator shows a big difference of
>>12.5 points out of 70 games (half time) and a score of 60.7 % against only
>>42.9 %. It seems as if the Gladiator settings do not work well at the time
>>control 15m+10s. Under these circumstances we abstain from also testing
>>CM9_Gladiator in the second half of the tournament.
>
>
>Hi Kurt,
>
>Thanks for testing CM9 Gladiator ,
>
>You are maybe right about that at 15 min+10 sec - CM9 Gladiator has bad results,
>
>Then clearly we can say for my Ratinglist that The King 3.23 SKR -do not work
>well at the time control: 40 moves in 40 min rest 30 min+10sec and here is the
>proof after 220 games:
>
>---------------------------------------------------------------------------
>
>
>
>Program Elo + - Games Score Av.Op. Draws
>
>1 Shredder 7.04 : 2785 37 37 240 63.3 % 2690 35.0 %
>2 CM9 Gladiator : 2759 43 35 220 55.9 % 2718 35.5 %
>3 Fritz 8 : 2747 44 37 200 57.5 % 2694 35.0 %
>4 Hiarcs 9 : 2743 40 32 240 57.1 % 2693 40.0 %
>5 Junior 8 : 2731 41 34 240 55.2 % 2694 32.9 %
>6 The King 3.23 SKR : 2717 44 32 220 53.9 % 2690 40.5 %
>7 Chess Tiger 15.0 : 2712 43 27 240 52.3 % 2696 47.9 %
>8 Ruffian 1.0.5 : 2685 31 46 220 48.9 % 2693 39.5 %
>9 Aristarch 4.21 : 2658 33 41 240 44.0 % 2700 35.4
>%10 List 504 : 2657 30 41 240 43.8 % 2701 42.5
>%
>11 Rebel 12b9 : 2635 34 39 240 40.4 % 2702 36.7 %
>12 Deep Sjeng 1.5 : 2620 41 41 200 37.8 % 2707 33.5 %
>13 Gandalf 5.1 : 2615 37 36 240 35.0 % 2722 36.7 %
>
>
>Best,
>Sedat Canbaz
Hi Sedat
Well, it's indeed strange what CM9_Gladiator demonstrated
in our "Rapid 15m+10s Tournament". We had really hoped that
all CM9-fans would now have the best overall CM9-setting
with Gladiator. On the other hand, the score of CM9_SKR
in your rating list is not much behind CM9_Gladiator
[score of 55,5 % vs 53,9 %] or in other words correspond
with the usual statistical margin error. After all the
testing with dozens of CM9-settings I do no longer believe
that it is possible to find a setting that plays successful
at all time controls. The same thing applies to our
CM9_Blitz: very strong indeed at tc 10m/g or 15m+10s but
significantly weaker at longer time controls. And as to
CM9_SKR we have only tested at 40'/40+40'/40+40' and not
with the time control used for your rating list.
http://www.utzingerkurt.com/cm9_skr_report.htm
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.