Mac Engines rating tournament

Discussion of computer chess matches and engine tournaments.

Moderator: Ras

User avatar
JuLieN
Posts: 2949
Joined: Mon May 05, 2008 12:16 pm
Location: Bordeaux (France)
Full name: Julien Marcel

Mac Engines rating tournament

Post by JuLieN »

I've started a tournament, using cutechess-cli, to rate the engines available for Mac. The rules are:
- 2mn for each side for a game,
- 64MB Hash tables,
- one core per engine.,
- pools of 20 engines playing a round-robin tournament,
- each engine meets the other engines four times (760 games per pool tournament)
- the games start from an EPD list of opening positions (so all engines that can't set up a game starting from a fen string are eliminated).
- All engines crashing, losing too many games on time, and so on, are eliminated too, replaced with other engines, and the tournament is restarted.
- The pools are simply populated alphabetically.
- The ten best engines of each pool tournament will qualify for next phase.

Here's the result of the first pool tournament:

Code: Select all

Rank Name                        ELO   Games   Score   Draws
   1 Critter-1.6a                527      76     95%      9%
   2 BlackMamba-1.4              349      76     88%      3%
   3 Fruit-2.3.1                 185      76     74%     14%
   4 Crafty-23.6                 185      76     74%     12%
   5 DeepShredder-11             167      76     72%     13%
   6 EXchess-7.11b               162      76     72%     14%
   7 Arasan-16.0                 134      76     68%     16%
   8 Daydreamer-1.75             109      76     65%     28%
   9 Fruit-2.1                    94      76     63%     26%
  10 Fruit-2.0                    51      76     57%     22%
  11 Danasah-5.07                  9      76     51%     16%
  12 Fruit-1.5                   -55      76     42%     16%
  13 Fruit-1.0                   -79      76     39%     12%
  14 DoubleCheck-2.6             -89      76     38%      7%
  15 Amundsen-0.80              -140      76     31%      9%
  16 Bikjump-2.01               -179      76     26%     13%
  17 Chesley-2009               -244      76     20%      5%
  18 Chenard-20130227           -349      76     12%      3%
  19 Chessone-2.01              -461      76      7%      8%
  20 Belofte-0.2.8              -527      76      5%      1%
Link to the PGN of this first pool tournament (760 games).
"The only good bug is a dead bug." (Don Dailey)
[Blog: http://tinyurl.com/predateur ] [Facebook: http://tinyurl.com/fbpredateur ] [MacEngines: http://tinyurl.com/macengines ]
User avatar
JuLieN
Posts: 2949
Joined: Mon May 05, 2008 12:16 pm
Location: Bordeaux (France)
Full name: Julien Marcel

Re: Mac Engines rating tournament

Post by JuLieN »

Results of the second pool tournament:

Code: Select all

Rank Name                        ELO   Games   Score   Draws
   1 Komodo-5.1                  480      76     94%      9%
   2 Komodo-2.03                 300      76     85%     12%
   3 Komodo-2.01                 300      76     85%     17%
   4 Hiarcs-14WCSC               274      76     83%     16%
   5 Komodo-1.3                  251      76     81%     17%
   6 Glaurung-2.2                162      76     72%     17%
   7 Gnuchess-5.50               129      76     68%     22%
   8 Gaviota-0.86                129      76     68%     12%
   9 GambitFruit-1.0b4x           79      76     61%     17%
  10 Hamsters-0.7.1              -14      76     48%     12%
  11 GreKo-9.0                   -65      76     41%     16%
  12 Kiwi-06d                    -69      76     40%     17%
  13 Gaia-3.5                   -104      76     36%     13%
  14 GreKo-10.2                 -109      76     35%     14%
  15 GreKo-9.7                  -109      76     35%     17%
  16 Jazz-r501                  -119      76     34%     14%
  17 Leonidas-8.3               -259      76     18%      3%
  18 Jabba-1.0                  -372      76     11%      0%
  19 KmtChess-1.2.1             -443      76      7%      1%
  20 Ges-134                    -inf      76      0%      0%
Link to the PGN (760 games)

I think I'll only keep the best version when an engine has several versions (one Fruit, one Komodo, etc...), so there will be more variety for stage 2. Same thing for the derivatives.
"The only good bug is a dead bug." (Don Dailey)
[Blog: http://tinyurl.com/predateur ] [Facebook: http://tinyurl.com/fbpredateur ] [MacEngines: http://tinyurl.com/macengines ]
User avatar
JuLieN
Posts: 2949
Joined: Mon May 05, 2008 12:16 pm
Location: Bordeaux (France)
Full name: Julien Marcel

Re: Mac Engines rating tournament

Post by JuLieN »

Results of the third pool tournament:

Code: Select all

Rank Name                        ELO   Games   Score   Draws
   1 Robbolito-0085g3l           398      76     91%     13%
   2 Stockfish-4                 349      76     88%     18%
   3 StingSF-3VE                 309      76     86%     18%
   4 Spark-1.0                   179      76     74%     29%
   5 Protector-1.5.0             173      76     73%     14%
   6 MinkoChess-1.3              140      76     69%     22%
   7 TogaII-1.4.1                114      76     66%     21%
   8 TogaII-3.0                  109      76     65%     14%
   9 Smaug-2.2.1                  27      76     54%     13%
  10 OctoChess-r5178              27      76     54%     21%
  11 TogaII-1.2b2a                18      76     53%     21%
  12 RedQueen-1.1.4               -5      76     49%     12%
  13 Rodent-1.00                 -23      76     47%     20%
  14 Sloppy-0.2.3                -41      76     44%     17%
  15 Myrddin-0.86               -140      76     31%      9%
  16 Rattatechess-Nosferatu     -216      76     22%      5%
  17 Sissa-2.0.0.0              -360      76     11%      7%
  18 Predateur-2.2.1            -360      76     11%      4%
  19 Simon-1.3                  -412      76      9%      4%
  20 Mizar-3.0                  -554      76      4%      3%
Link to the PGN (760 games)

Interestingly, in this single-threads tournament, Robbolito is quite stronger than Stockfish 4!
"The only good bug is a dead bug." (Don Dailey)
[Blog: http://tinyurl.com/predateur ] [Facebook: http://tinyurl.com/fbpredateur ] [MacEngines: http://tinyurl.com/macengines ]
User avatar
JuLieN
Posts: 2949
Joined: Mon May 05, 2008 12:16 pm
Location: Bordeaux (France)
Full name: Julien Marcel

Re: Mac Engines rating tournament

Post by JuLieN »

Results of the fourth (and last) pool tournament:

Code: Select all

Rank Name                        ELO   Games   Score   Draws
   1 IvanHoe-Beta-999966         338      76     88%     12%
   2 Igorrit-0.086v2             328      76     87%     11%
   3 Crab-1.0b                   309      76     86%     16%
   4 TogaReturns-1.1             244      76     80%     13%
   5 TogaToy-1.0                 124      76     67%     21%
   6 Texel-1.02                  114      76     66%     18%
   7 TogaII-3.1.2                104      76     64%     13%
   8 DiscoCheck-4.3               99      76     64%     20%
   9 DeepJunior-13.3              84      76     62%      8%
  10 RainbowSerpent-2.3.1         60      76     59%     12%
  11 DoubleCheck-3.4              51      76     57%     12%
  12 Tucano-2.0                  -18      76     47%     24%
  13 Umko-0.3.0.1                -32      76     45%     17%
  14 Viper-0.1                   -60      76     41%      7%
  15 Vice-1.0                   -216      76     22%      5%
  16 Simplex-098                -216      76     22%      0%
  17 ZetaDva-022                -274      76     17%      3%
  18 Kenny-0.1.1.0              -309      76     14%      5%
  19 Rocinante-2.0              -427      76      8%      0%
  20 APILchess-1.06             -627      76      3%      0%
Link to the PGN (760 games)

The reigning CC World Champion, Deep Junior, wasn't very impressive in this mono-threads tournament...

I now selected the 40 engines that will go to stage 2. The remaining engines, less the duplicate versions and derivative, will also compete in a "second division" rating tournament. :) I like stable original engines, however weak they are.

Here are the engines that qualified and will compete in the first league. I kept only one engine of each family, and I interlaced the groups as much as possible :
League 1, group A: Arasan-16.0, Crab-1.0b, Critter-1.6a, Danasah-5.07, DeepJunior-13.3, DeepShredder-11, Fruit-2.3.1, Gaia-3.5, Gaviota-0.86, Glaurung-2.2, GreKo-9.0, IvanHoe-Beta-999966, Komodo-5.1, Myrddin-0.86, OctoChess-r5178, Protector-1.5.0, Rodent-1.00, Stockfish-4, Texel-1.02, Umko-0.3.0.1
League 1, group B: Amundsen-0.80, BlackMamba-1.4, Crafty-23.6, Daydreamer-1.75, DiscoCheck-4.3, EXchess-7.11b, Gnuchess-5.50, Hamsters-0.7.1, Hiarcs-14WCSC, Igorrit-0.086v2, Jazz-r501, Kiwi-06d, MinkoChess-1.3, RedQueen-1.1.4, Robbolito-0085g3l, Sloppy-0.2.3, Spark-1.0, TogaReturns-1.1, Tucano-2.0, Viper-0.1
And here are the engines that will play in League 2. As I had only 18 remaining ones, I chose to add two engines that already have a family member in league 1 (Fruit 1.0 and Greko 9.7):
League 2: APILchess-1.06, Belofte-0.2.8, Bikjump-2.01, Chenard-20130227, Chesley-2009, Chessone-2.01, Fruit-1.0, Ges-134, GreKo-9.7, Kenny-0.1.1.0, KmtChess-1.2.1, Mizar-3.0, Predateur-2.2.1, Rattatechess-Nosferatu, Rocinante-2.0, Simon-1.3, Simplex-098, Sissa-2.0.0.0, Vice-1.0, ZetaDva-022
"The only good bug is a dead bug." (Don Dailey)
[Blog: http://tinyurl.com/predateur ] [Facebook: http://tinyurl.com/fbpredateur ] [MacEngines: http://tinyurl.com/macengines ]
User avatar
JuLieN
Posts: 2949
Joined: Mon May 05, 2008 12:16 pm
Location: Bordeaux (France)
Full name: Julien Marcel

Re: Mac Engines rating tournament

Post by JuLieN »

Results of the "league 2" tournament:

Code: Select all

Rank Name                        ELO   Games   Score   Draws
   1 TogaII-1.2b2a               872      76     99%      1%
   2 Fruit-1.0                   309      76     86%      8%
   3 Fruit-1.5                   291      76     84%     11%
   4 GreKo-9.7                   230      76     79%     13%
   5 Rattatechess-Nosferatu      162      76     72%     20%
   6 Chesley-2009                134      76     68%      5%
   7 Bikjump-2.01                114      76     66%     11%
   8 ZetaDva-022                  79      76     61%      4%
   9 Vice-1.0                     37      76     55%     16%
  10 Simplex-098                  37      76     55%      8%
  11 Kenny-0.1.1.0               -41      76     44%     12%
  12 Chenard-20130227            -65      76     41%     11%
  13 Sissa-2.0.0.0               -94      76     37%     11%
  14 Predateur-2.2.1             -94      76     37%      8%
  15 Mizar-3.0                  -119      76     34%      7%
  16 Rocinante-2.0              -140      76     31%      4%
  17 Simon-1.3                  -167      76     28%      8%
  18 APILchess-1.06             -300      76     15%      4%
  19 Ges-134                    -412      76      9%      7%
  20 Belofte-0.2.8              -inf      76      0%      0%
Link to the PGN (760 games)

As ChessOne was playing some illegal moves and KMTChess crashed several times, I also had to replace them by place holders (Fruit 1.5 and Toga II-1.2ba).
"The only good bug is a dead bug." (Don Dailey)
[Blog: http://tinyurl.com/predateur ] [Facebook: http://tinyurl.com/fbpredateur ] [MacEngines: http://tinyurl.com/macengines ]
User avatar
JuLieN
Posts: 2949
Joined: Mon May 05, 2008 12:16 pm
Location: Bordeaux (France)
Full name: Julien Marcel

Re: Mac Engines rating tournament

Post by JuLieN »

Results of League 1, group A:

Code: Select all

Rank Name                        ELO   Games   Score   Draws
   1 Stockfish-4                 372      76     89%     16%
   2 Critter-1.6a                300      76     85%     22%
   3 IvanHoe-999946f             291      76     84%     24%
   4 IvanHoe-Beta-999966         291      76     84%     24%
   5 Komodo-5.1                  282      76     84%     14%
   6 Protector-1.5.0             124      76     67%     16%
   7 Texel-1.02                   69      76     60%     22%
   8 Fruit-2.3.1                  14      76     52%     20%
   9 Glaurung-2.2                  5      76     51%     14%
  10 DeepShredder-11               5      76     51%     20%
  11 Gaviota-0.86                 -9      76     49%     21%
  12 DeepJunior-13.3             -37      76     45%     13%
  13 OctoChess-r5178             -69      76     40%     20%
  14 Arasan-16.0                 -74      76     39%     21%
  15 Rodent-1.00                 -84      76     38%     18%
  16 Umko-0.3.0.1               -167      76     28%     18%
  17 GreKo-9.0                  -274      76     17%     13%
  18 Gaia-3.5                   -309      76     14%      8%
  19 Myrddin-0.86               -318      76     14%     12%
  20 Rattatechess-Nosferatu     -398      76      9%      3%
Link to the PGN (760 games)

Stockfish 4 was imperial. Deep Junior 13.3 looks like a let down, but it is apparently suffering from two things:
- the mono-threaded nature of the tournament,
- the fast cadence: it lost many games on time (!)

If you wonder where Danasah is, it crashed so I had to replace it with Rattatechess, first "different" engine of the League 2 tournament.

Gaviota is NOT disqualified, because I'll keep only one Ivanhoe (999946f) for the final.

The aging Fruit is still a formidable opponent. Impressive.
"The only good bug is a dead bug." (Don Dailey)
[Blog: http://tinyurl.com/predateur ] [Facebook: http://tinyurl.com/fbpredateur ] [MacEngines: http://tinyurl.com/macengines ]
User avatar
JuLieN
Posts: 2949
Joined: Mon May 05, 2008 12:16 pm
Location: Bordeaux (France)
Full name: Julien Marcel

Re: Mac Engines rating tournament

Post by JuLieN »

Results of League 1, group B:

Code: Select all

Rank Name                        ELO   Games   Score   Draws
   1 Robbolito-0085g3l           300      76     85%     22%
   2 Igorrit-0.086v2             259      76     82%     18%
   3 BlackMamba-1.4              244      76     80%     13%
   4 TogaReturns-1.1             230      76     79%     24%
   5 Hiarcs-14WCSC               191      76     75%     21%
   6 Spark-1.0                   173      76     73%     20%
   7 Hiarcs-13.1                 104      76     64%     24%
   8 MinkoChess-1.3               37      76     55%     29%
   9 Crafty-23.6                  23      76     53%     20%
  10 DiscoCheck-4.3                5      76     51%     20%
  11 EXchess-7.11b                 0      76     50%     21%
  12 Arasan-16.1                 -69      76     40%     22%
  13 Daydreamer-1.75             -89      76     38%     22%
  14 RedQueen-1.1.4              -99      76     36%     20%
  15 Sloppy-0.2.3               -114      76     34%     24%
  16 Tucano-2.0                 -156      76     29%     18%
  17 Hamsters-0.7.1             -179      76     26%     18%
  18 Viper-0.1                  -203      76     24%     18%
  19 Kiwi-06d                   -251      76     19%     17%
  20 Amundsen-0.80              -461      76      7%      3%
Link to the PGN (760 games)

GNU CHess and Jazz crashed and were replaced with HIARCS 13.1 and Arasan 16.1. Pity for GNU Chess that crashed at mid-tournament, when it was at the tenth place, qualified for the final...

Now that both League 1 groups played, here's the list of engines qualified for the final :
Qualified: BlackMamba-1.4, Crafty-23.6, Critter-1.6a, DeepShredder-11, DiscoCheck-4.3, EXchess-7.11b, Fruit-2.3.1, Gaviota-0.86, Glaurung-2.2, Hiarcs-14WCSC, Igorrit-0.086v2, IvanHoe-999946f, Komodo-5.1, MinkoChess-1.3, Protector-1.5.0, Robbolito-0085g3l, Spark-1.0, Stockfish-4, Texel-1.02, TogaReturns-1.1
I will also play a semi-final tournament, with the following engines :
semi-finalists: Amundsen-0.80, Arasan-16.1, Daydreamer-1.75, DeepJunior-13.3, Gaia-3.5, GreKo-9.0, Hamsters-0.7.1, Hiarcs-11.1, Hiarcs-12.1, Kiwi-06d, Myrddin-0.86, OctoChess-r5178, Rattatechess-Nosferatu, RedQueen-1.1.4, Rodent-1.00, Sloppy-0.2.3, Tucano-2.0, Tucano-3.0, Umko-0.3.0.1, Viper-0.1
(To replace the engines that were removed, I add Tucano 3.0 and Hiarcs 11.1 and 12.1.)
"The only good bug is a dead bug." (Don Dailey)
[Blog: http://tinyurl.com/predateur ] [Facebook: http://tinyurl.com/fbpredateur ] [MacEngines: http://tinyurl.com/macengines ]
User avatar
JuLieN
Posts: 2949
Joined: Mon May 05, 2008 12:16 pm
Location: Bordeaux (France)
Full name: Julien Marcel

Re: Mac Engines rating tournament

Post by JuLieN »

Results of the semi finals. For some reason, cutechess didn't display the whole results! :shock: So here's ELOStat's cross-table instead (starting ELO: 2300):

Code: Select all

    Program                            Score     %    Av.Op.  Elo    +   -    Draws

  1 Fire-xTreme-2.2                :  70.5/ 76  92.8   2275   2718  115  97   14.5 %
  2 Hiarcs-12.1                    :  62.5/ 76  82.2   2283   2550   80  76   25.0 %
  3 Hiarcs-11.1                    :  57.0/ 76  75.0   2287   2478   75  72   26.3 %
  4 OctoChess-r5178                :  50.5/ 76  66.4   2291   2410   76  75   17.1 %
  5 RedQueen-1.1.4                 :  50.0/ 76  65.8   2291   2405   77  75   15.8 %
  6 DeepJunior-13.3                :  48.0/ 76  63.2   2292   2386   76  74   15.8 %
  7 Arasan-16.1                    :  44.5/ 76  58.6   2294   2354   75  74   14.5 %
  8 Rodent-1.00                    :  44.5/ 76  58.6   2294   2354   75  74   14.5 %
  9 Daydreamer-1.75                :  41.5/ 76  54.6   2295   2327   66  66   30.3 %
 10 Tucano-3.0                     :  41.5/ 76  54.6   2295   2327   70  70   22.4 %
 11 Tucano-2.0                     :  39.0/ 76  51.3   2296   2305   65  65   31.6 %
 12 Sloppy-0.2.3                   :  36.5/ 76  48.0   2297   2284   71  71   19.7 %
 13 Hamsters-0.7.1                 :  33.0/ 76  43.4   2299   2253   70  71   21.1 %
 14 Viper-0.1                      :  31.0/ 76  40.8   2300   2235   73  74   15.8 %
 15 GreKo-9.0                      :  28.0/ 76  36.8   2301   2208   73  74   18.4 %
 16 Kiwi-06d                       :  20.5/ 76  27.0   2305   2132   79  82   14.5 %
 17 Myrddin-0.86                   :  19.0/ 76  25.0   2306   2115   85  89    7.9 %
 18 Rattatechess-Nosferatu         :  15.0/ 76  19.7   2309   2065   89  94   10.5 %
 19 Amundsen-0.80                  :  14.5/ 76  19.1   2309   2058   88  94   11.8 %
 20 Gaia-3.5                       :  13.0/ 76  17.1   2310   2036   90  96   13.2 %
Link to the PGN (760 games)

I replaced Umko, that crashed, with Fire xTreme, that I had just compiled, and it did really well. So if an engine fails in the finale, Fire will take its spot.
"The only good bug is a dead bug." (Don Dailey)
[Blog: http://tinyurl.com/predateur ] [Facebook: http://tinyurl.com/fbpredateur ] [MacEngines: http://tinyurl.com/macengines ]
User avatar
Ajedrecista
Posts: 2189
Joined: Wed Jul 13, 2011 9:04 pm
Location: Madrid, Spain.

Re: Mac engines rating tournament.

Post by Ajedrecista »

Hello Julien:
JuLieN wrote:Results of the semi finals. For some reason, cutechess didn't display the whole results! :shock: So here's ELOStat's cross-table instead (starting ELO: 2300):

Code: Select all

    Program                            Score     %    Av.Op.  Elo    +   -    Draws

  1 Fire-xTreme-2.2                :  70.5/ 76  92.8   2275   2718  115  97   14.5 %
  2 Hiarcs-12.1                    :  62.5/ 76  82.2   2283   2550   80  76   25.0 %
  3 Hiarcs-11.1                    :  57.0/ 76  75.0   2287   2478   75  72   26.3 %
  4 OctoChess-r5178                :  50.5/ 76  66.4   2291   2410   76  75   17.1 %
  5 RedQueen-1.1.4                 :  50.0/ 76  65.8   2291   2405   77  75   15.8 %
  6 DeepJunior-13.3                :  48.0/ 76  63.2   2292   2386   76  74   15.8 %
  7 Arasan-16.1                    :  44.5/ 76  58.6   2294   2354   75  74   14.5 %
  8 Rodent-1.00                    :  44.5/ 76  58.6   2294   2354   75  74   14.5 %
  9 Daydreamer-1.75                :  41.5/ 76  54.6   2295   2327   66  66   30.3 %
 10 Tucano-3.0                     :  41.5/ 76  54.6   2295   2327   70  70   22.4 %
 11 Tucano-2.0                     :  39.0/ 76  51.3   2296   2305   65  65   31.6 %
 12 Sloppy-0.2.3                   :  36.5/ 76  48.0   2297   2284   71  71   19.7 %
 13 Hamsters-0.7.1                 :  33.0/ 76  43.4   2299   2253   70  71   21.1 %
 14 Viper-0.1                      :  31.0/ 76  40.8   2300   2235   73  74   15.8 %
 15 GreKo-9.0                      :  28.0/ 76  36.8   2301   2208   73  74   18.4 %
 16 Kiwi-06d                       :  20.5/ 76  27.0   2305   2132   79  82   14.5 %
 17 Myrddin-0.86                   :  19.0/ 76  25.0   2306   2115   85  89    7.9 %
 18 Rattatechess-Nosferatu         :  15.0/ 76  19.7   2309   2065   89  94   10.5 %
 19 Amundsen-0.80                  :  14.5/ 76  19.1   2309   2058   88  94   11.8 %
 20 Gaia-3.5                       :  13.0/ 76  17.1   2310   2036   90  96   13.2 %
Link to the PGN (760 games)

I replaced Umko, that crashed, with Fire xTreme, that I had just compiled, and it did really well. So if an engine fails in the finale, Fire will take its spot.
Thanks for the tournament.

I ran my own rating programme for Round Robin tournaments and I got the following:

Code: Select all

Round Robin with 20 engines and     76 games per engine.
Total number of games:       760 games.
 
 Engines:     Performance:     Score:
 
Engine 01:      2716.55       92.76 %
Engine 02:      2548.95       82.24 %
Engine 03:      2477.55       75.00 %
Engine 04:      2409.20       66.45 %
Engine 05:      2404.37       65.79 %
Engine 06:      2385.45       63.16 %
Engine 07:      2353.61       58.55 %
Engine 08:      2353.61       58.55 %
Engine 09:      2327.15       54.61 %
Engine 10:      2327.15       54.61 %
Engine 11:      2305.41       51.32 %
Engine 12:      2283.75       48.03 %
Engine 13:      2253.18       43.42 %
Engine 14:      2235.41       40.79 %
Engine 15:      2208.04       36.84 %
Engine 16:      2132.84       26.97 %
Engine 17:      2115.94       25.00 %
Engine 18:      2065.88       19.74 %
Engine 19:      2058.95       19.08 %
Engine 20:      2037.02       17.11 %
 
Mean of ratings:  2300.00 Elo.
Almost the same! I always get very similar results to EloSTAT but I think that this time was the best one. My rating programme is VERY simple. I do the following for a number of engines greater than two:

Code: Select all

[...]

sum_delta = 0d0  ! Initialization.
do i = 1, engines
  delta(i) = 4d2*log10(points(i)/(games_per_engine - points(i)))  ! By definition.
  score(i) = points(i)/games_per_engine
  sum_delta = sum_delta + delta(i)
end do
average_delta = sum_delta/engines
do i = 1, engines
  average_delta_of_opponents(i) = (average_delta - delta(i))/(engines - 1)
  rating(i) = average_delta_of_opponents(i) + delta(i) - average_delta + mean_of_ratings
end do

[...]
It is Fortran 95 code (a true oldie!). My constant mean_of_ratings = 2300 this time (it is completely equivalent to starting Elo of EloSTAT). I do not use prior, drawelo.. not even the results of each 1 vs. 1 engine match. Only the final number of points, so there is no need of the PGN (I find it useful when people post the results of a Round Robin (the amount of points won by each engine) without providing the PGN and without sharing a rating list for the tournament). Few lines of code can do a nice job in comparison with EloSTAT, although my tool is restricted to RR.

Please keep up your good work! :)

Regards from Spain.

Ajedrecista.
User avatar
JuLieN
Posts: 2949
Joined: Mon May 05, 2008 12:16 pm
Location: Bordeaux (France)
Full name: Julien Marcel

Re: Mac engines rating tournament.

Post by JuLieN »

Thanks Jesus, interesting results. :) Was this a mac engines tournament as well ? What were the engines in your tournament ?

The final is being played right now (results tomorrow), but I can give a sneak peak of the top 30 in my ratings list. As Fire is leading it without having met any of the leaders, I'll have to get it play a gauntlet against them after the final.

Code: Select all

    Program                          Elo    +   -   Games   Score   Av.Op.  Draws

  1 Fire-xTreme-2.2                : 2758  115  97    76    92.8 %   2315   14.5 %
  2 Stockfish-4                    : 2741   69  65   152    88.8 %   2381   17.1 %
  3 Robbolito-0085g3l              : 2721   67  64   152    87.8 %   2377   17.8 %
  4 IvanHoe-999946f                : 2717   83  79    76    84.2 %   2427   23.7 %
  5 Critter-1.6a                   : 2710   72  68   152    90.1 %   2326   15.8 %
  6 Komodo-5.1                     : 2689   76  72   152    88.8 %   2329   11.8 %
  7 IvanHoe-Beta-999966            : 2680   65  62   152    85.9 %   2367   17.8 %
  8 Igorrit-0.086v2                : 2654   67  64   152    84.2 %   2363   14.5 %
  9 StingSF-3VE                    : 2651   92  87    76    85.5 %   2342   18.4 %
 10 Crab-1.0b                      : 2617   96  91    76    85.5 %   2309   15.8 %
 11 BlackMamba-1.4                 : 2617   73  69   152    84.2 %   2326    7.9 %
 12 TogaReturns-1.1                : 2602   59  58   152    79.6 %   2366   18.4 %
 13 Hiarcs-12.1                    : 2590   80  76    76    82.2 %   2323   25.0 %
 14 Spark-1.0                      : 2562   53  52   152    73.4 %   2386   24.3 %
 15 Hiarcs-14WCSC                  : 2560   59  57   152    78.9 %   2331   18.4 %
 16 Protector-1.5.0                : 2540   56  54   152    70.1 %   2392   15.1 %
 17 Komodo-2.03                    : 2538  101  94    76    84.9 %   2238   11.8 %
 18 Komodo-2.01                    : 2538   93  88    76    84.9 %   2238   17.1 %
 19 Hiarcs-13.1                    : 2530   72  71    76    64.5 %   2426   23.7 %
 20 Hiarcs-11.1                    : 2518   75  72    76    75.0 %   2327   26.3 %
 21 Komodo-1.3                     : 2492   88  84    76    80.9 %   2241   17.1 %
 22 MinkoChess-1.3                 : 2477   49  49   152    62.2 %   2390   25.7 %
 23 Texel-1.02                     : 2469   51  51   152    62.8 %   2378   20.4 %
 24 TogaII-1.4.1                   : 2466   74  72    76    65.8 %   2352   21.1 %
 25 TogaII-3.0                     : 2461   77  75    76    65.1 %   2352   14.5 %
 26 TogaToy-1.0                    : 2442   74  73    76    67.1 %   2318   21.1 %
 27 Fruit-2.3.1                    : 2434   52  52   152    63.2 %   2341   17.1 %
 28 Crafty-23.6                    : 2434   53  52   152    63.8 %   2336   15.8 %
 29 DiscoCheck-4.3                 : 2426   50  50   152    57.2 %   2375   19.7 %
 30 DeepShredder-11                : 2423   52  52   152    61.5 %   2341   16.4 %
(Starting Elo : 2300).
"The only good bug is a dead bug." (Don Dailey)
[Blog: http://tinyurl.com/predateur ] [Facebook: http://tinyurl.com/fbpredateur ] [MacEngines: http://tinyurl.com/macengines ]