Houdini 2.0z test with 2700 bullet-games

[pohl4711]

The Houdini 2.0z-Setting in my private LightSpeed-rankinglist (1/3 thinking time of TEB-rankinglist and only 64 MB Hash). 2700 games againts the same opponents like Houdini 2.0c. Disappointing: +9 elo (inside errorbar).

Intel i7-2630QM (SSE42 support, Windows 7 64bit, Quadcore, FritzMark=20.2), 64 MB Hash, 1 core per engine, no ponder, no endgame-bases, no resign. 150 selected opening-positions (all 8 moves deep, from Frank Q.-database)
Elos calculated with bayeselo (fixpoint Robbolito 0.085g3 3000 Elo). LittleBlitzerGUI (gauntlet-mode only, because this GUI chooses opening-positions per random in the round-robin-mode from the PGN-file...)
Time: 20''+250ms Fischerbonus (= 40 seconds per game/engine).

Code: Select all

Rank Name                    Elo    +    - games score oppo. draws 
   1 Houdini 2.0z x64       3102   11   10  2700   63%  3014   38% (Setting „z“ (by Salvador))
   2 Houdini 2.0c x64       3093   10   10  2700   62%  3014   40% 
   3 Houdini 1.5a x64       3074   10   10  2700   59%  3014   42% 
   4 Critter 1.6 x64        3068    9    9  3300   56%  3030   49% 
   5 Strelka 5.5 x64        3060    9    9  3300   55%  3030   50% 
   6 Ivanhoe 46h x64        3024    9    9  3300   49%  3034   50% 
   7 Robbolito 0.10 x64s    3010    9    9  3300   46%  3035   51% 
   8 Komodo 4 x64s          3005   10   10  3300   45%  3035   38% 
   9 Rybka 4.1 x64s         3004    9    9  3300   45%  3035   42% 
  10 Robbolito 0.085g3 x64  3000    9    9  3300   44%  3036   49% 
  11 Deep Saros 3.0 x64     2980    9    9  3300   41%  3038   43% 
  12 Stockfish 2.2.2 x64s   2974    9   10  3300   40%  3038   40% 

(x64=64bit version, x64s=64bit SSE42-version)

If you want to get the games, send me an eMail-adress per PM. I will send the games (PGN-file) as soon as possible...

Greetings – Stefan

[Sedat Canbaz]

The Houdini 2.0z-Setting in my private LightSpeed-rankinglist (1/3 thinking time of TEB-rankinglist and only 64 MB Hash). 2700 games againts the same opponents like Houdini 2.0c. Disappointing: +9 elo (inside errorbar).

Intel i7-2630QM (SSE42 support, Windows 7 64bit, Quadcore, FritzMark=20.2), 64 MB Hash, 1 core per engine, no ponder, no endgame-bases, no resign. 150 selected opening-positions (all 8 moves deep, from Frank Q.-database)
Elos calculated with bayeselo (fixpoint Robbolito 0.085g3 3000 Elo). LittleBlitzerGUI (gauntlet-mode only, because this GUI chooses opening-positions per random in the round-robin-mode from the PGN-file...)
Time: 20''+250ms Fischerbonus (= 40 seconds per game/engine).

Code: Select all

Rank Name                    Elo    +    - games score oppo. draws 
   1 Houdini 2.0z x64       3102   11   10  2700   63%  3014   38% (Setting „z“ (by Salvador))
   2 Houdini 2.0c x64       3093   10   10  2700   62%  3014   40% 
   3 Houdini 1.5a x64       3074   10   10  2700   59%  3014   42% 
   4 Critter 1.6 x64        3068    9    9  3300   56%  3030   49% 
   5 Strelka 5.5 x64        3060    9    9  3300   55%  3030   50% 
   6 Ivanhoe 46h x64        3024    9    9  3300   49%  3034   50% 
   7 Robbolito 0.10 x64s    3010    9    9  3300   46%  3035   51% 
   8 Komodo 4 x64s          3005   10   10  3300   45%  3035   38% 
   9 Rybka 4.1 x64s         3004    9    9  3300   45%  3035   42% 
  10 Robbolito 0.085g3 x64  3000    9    9  3300   44%  3036   49% 
  11 Deep Saros 3.0 x64     2980    9    9  3300   41%  3038   43% 
  12 Stockfish 2.2.2 x64s   2974    9   10  3300   40%  3038   40% 

(x64=64bit version, x64s=64bit SSE42-version)

If you want to get the games, send me an eMail-adress per PM. I will send the games (PGN-file) as soon as possible...

Greetings – Stefan

Dear Stefan,

Good job...

I see in your rating that 2.0z performed 9 Elo better than v2.0c
Interesting,even with ultra fast time control,we see improvement
And when we compare your results with my results,the difference is 11 Elo


I strongly believe the below factors can be one of the reasons
about why we see Elo differences between my rating and yours:
---------------------------------------------------------------------------
-I use different hardware than yours
-SCCT's MP Engines are played with 6 cores
-I use different openings than yours (Perfect 2012b is highly well optimized for such conditions)
-I use different time control;larger hashtables ...
-You have more games played (by 2.0z) than me


Best,
Sedat