SWCR LIVE: Critter 1.4 x64 and Komodo 4.0 x64

[MM]

Komodo 4.0 x64 = +-0
Critter 1.4 x64 = +24

Code: Select all

Rank Name                            Elo    +    - games score oppo. draws 
   1 Houdini 2.0c x64               3019   18   18  1382   79%  2777   28% 
   2 Houdini 1.5 x64                2999   14   14  2320   78%  2772   29% 
   3 Houdini 1.5 w32                2981   19   18  1240   79%  2757   29% 
   4 Critter 1.4 x64                2980   17   16  1600   78%  2757   32% 
   5 Rybka 4 x64 Exp. 42            2968   18   18  1360   76%  2757   29% 
   6 Rybka 4.1 x64 Exp. 79TD v.1    2963   19   19  1200   79%  2736   26% 
   7 Komodo 3.0 x64                 2960   16   16  1581   75%  2763   33% 
   8 Komodo 4.0 x64                 2960   16   16  1600   76%  2758   32% 
   9 IvanHoe 999946f x64            2960   16   16  1480   73%  2781   34% 
  10 Rybka 4 x64 Exp. 61            2959   21   21  1000   78%  2742   27% 
  11 IvanHoe B46fB x64              2957   17   17  1400   74%  2770   35% 
  12 Critter 1.2 x64                2956   15   15  1783   75%  2761   34% 
  13 Rybka 4.1 x64                  2953   14   14  1941   71%  2791   37% 
  14 Stockfish 2.1.1 JA x64 PHQ     2949   17   16  1440   73%  2770   35% 
  15 Houdini 1.03a x64              2943   21   21  1000   80%  2710   30% 
  16 Fire 2.2 xTreme x64            2943   15   15  1821   73%  2767   37% 
  17 IvanHoe B47cB x64              2941   15   14  1779   71%  2785   39% 
  18 Rybka 4 x64                    2940   18   17  1480   80%  2696   28% 
  19 Fire 1.5 xTreme x64            2936   17   16  1400   73%  2762   36% 
  20 Critter 1.2 w32                2935   18   18  1200   75%  2754   36% 
  21 IvanHoe B49jA x64              2932   19   18  1160   76%  2738   34% 
  22 Stockfish 2.1.1 JA x64         2930   14   14  1981   71%  2772   36% 
  23 Komodo 2.03 JA x64             2927   17   16  1480   73%  2749   31% 
  24 RobboLito 0.09 x64             2923   16   15  1640   71%  2753   36% 
  25 IvanHoe B52aC x64              2918   20   20  1000   76%  2726   34% 
  26 Stockfish 1.8.0 JA x64         2906   19   19  1160   75%  2711   32% 
  27 Stockfish 2.0.1 JA x64         2906   18   18  1120   69%  2767   38% 
  28 Stockfish 1.9.1 JA x64         2904   18   18  1240   72%  2739   34% 
  29 Fire 2.2 xTreme w32            2904   17   17  1240   70%  2759   40% 
  30 Fire 1.31 x64                  2903   19   19  1040   73%  2733   37% 

Thank you

Best Regards

[Graham Banks]

Komodo 4.0 x64 = +-0

In CCRL 40/40 testing, there is no gain for Komodo 4 x64 over Komodo 3 either.

[natasha]

[beram]

Komodo 4.0 x64 = +-0

In CCRL 40/40 testing, there is no gain for Komodo 4 x64 over Komodo 3 either.

well, but at least Komodo does scale better than Houdini

[Frank Quisinsky]

Hi Graham,

I tested vs. 39 opponents.
Have a look in the results vs. Rybka 4 Exp. 42 x64 or vs. Stockfish 2.1.1 JA PHQ x64. Also Komodo lost some points vs. Robbolito 0.9 x64.

With other words:
I tested vs. much more strong opponents others have in the list. Could be a reason for the results. Means if you delete the results vs. PHQ, EXP 42 or Robbo Komodo have +6.

That is also a reason for myself that SWCR ended. Endless story with testing since so many years. A good list can produce a result +- 20 ELO I think, not important how many games played.

You can make the following experiment:
Delete 10x40 games Komodo have problems. So you have 1.200 games with a complete other results.

So I hope with a new activity I have in my brain to find out more about engines and playing style. I think all the time SWCR is still running that playing style, strengths and weeknesses are much more important as ELO pur.

In case of Komodo it's nice to see that with more time Komodo have weaker results. Only one available results from CEGT will give us an other view.

Komodo fans have to look on CEGT, Rybka settings fans have to look on SWCR, Junior fans have to look on SWCR other fans have to look on CCRL, do you understand what I mean in my bad English?

CCRL, CEGT, IPON and all other results are important, absolutly important but we should not lost the view on the playing style of an engines and programmers should try to give an engine an own face. So many strong engines plays the same chess ... this should be changes and programmers should more work on own ideas, new ideas.

A good example is Komodo with his strengths!
So the Komodo result in ELO in SWCR isn't really important.

Best
Frank