Stefan Pohl Computer Chess

Latest Website-News (2020/07/05): NN-testrun of Lc0 0.25.1 sv-1810 finished (old 20x256 experimental net) - very impressive result. See the result and download the games in the "NN vs SF testing"- section. Next NN-testrun: Lc0 0.25.1 Leelenstein 15 with Kayra4 parameter setting.

AB-testrun of Ethereal 12.25 finished: +23 Elo to Ethereal 12.00.

Next testrun: StockFiNN 0.2 (Stockfish NN with bigger net by Josh (for patreons of Josh, only, so a commercial engine at the moment)).

Stay tuned.

Stockfish testing

Playing conditions:

Hardware: i7-6700HQ 2.6GHz Notebook (Skylake CPU), Windows 10 64bit, 8GB RAM

Fritzmark: singlethread: 5.3 / 2521 (all engines running on one thread, only), average meganodes/s displayed by LittleBlitzerGUI: Houdini: 2.6 mn/s, Stockfish: 2.2 mn/s, Komodo: 2.0 mn/s

Hash: 512MB per engine

GUI: Since 19/09/11: Cutechess-cli (GUI ends game, when a 5-piece endgame is on the board), before: LittleBlitzerGUI (draw at 170 moves, resign at -700cp)

Tablebases: None for engines, 5 Syzygy for cutechess-cli

Openings: HERT_500 testset (by Thomas Zipproth) (download the file at the "Download & Links"-section or here)

Ponder, Large Memory Pages & learning: Off

Thinking time: 180''+1000ms (= 3'+1'') per game/engine (average game-duration: around  7.5 minutes). One 5000 games-testrun takes about 7 days.The version-numbers of the Stockfish engines are the date of the latest patch, which was included in the Stockfish sourcecode, not the release-date of the engine-file, written backwards (year,month,day))(example: 170526 = May, 26, 2017). Since July, 2018 I use the abrok-compiles of Stockfish again (http://abrok.eu/stockfish), because they are now much faster than before - now only 1.3% slower than BrainFish-compiles. So, there is no reason anymore to not use these "official" development-compiles.

Download BrainFish (and the Cerebellum-Libraries): here

Each Stockfish-version plays 1000 games versus Komodo 14, Houdini 6, Fire 7.1, Xiphos 0.6, Ethereal 12.25. All engines are running with default-settings.

To avoid distortions in the Ordo Elo-calculation, from now, only 2x Stockfish (latest official release + the latest version) and 1x asmFish and 1x Brainfish are stored in the gamebase (all older engine-versions games will be deleted, every time, when a new version was tested). Stockfish, asmFish and BrainFish older Elo-results can still be seen in the Elo-diagrams below. BrainFish plays always with the latest Cerebellum-Libraries of course, because otherwise BrainFish = Stockfish.

Latest update: 2020/07/05: Ethereal 12.25 (+23 Elo to Ethereal 12.00)

(Ordo-calculation fixed to Stockfish 11 = 3554 Elo)

See the individual statistics of engine-results here

See the ORDO-rating of the archive-gamebase since 2020 here

Download the current gamebase here

Download the archive-gamebase since 2020 here

     Program                      Elo    +    -   Games   Score   Av.Op.  Draws

   1 BrainFish-2 200329 bmi2    : 3620    9    9  5000    82.4 %   3342   33.6 %
   2 Stockfish 200621 bmi2      : 3578    8    8  5000    78.4 %   3344   39.9 %
   3 Stockfish 200601 bmi2      : 3578    9    9  5000    78.4 %   3344   40.4 %
   4 Stockfish 11 200118        : 3554    6    6 13000    78.2 %   3322   39.7 %
   5 Stockfish 10 181129        : 3507    5    5 15000    78.5 %   3270   37.7 %
   6 Stockfish 9 180201         : 3458    8    8  5000    74.9 %   3255   41.7 %
   7 Komodo 14 bmi2             : 3432    6    6  9000    52.7 %   3412   56.8 %
   8 Houdini 6 pext             : 3425    3    3 31000    62.4 %   3328   48.4 %
   9 Komodo 13.3 bmi2           : 3421    6    6  9000    58.5 %   3358   49.2 %
  10 Komodo 13.1 bmi2           : 3408    5    5 11000    62.0 %   3316   48.8 %
  11 Komodo 12.3 bmi2           : 3395    7    7  7000    62.7 %   3296   49.4 %
  12 SF nnue 200609             : 3386    7    7  5000    55.7 %   3344   53.1 %
  13 Ethereal 12.25 pext        : 3325    7    7  5000    41.0 %   3395   56.8 %
  14 Komodo 14 MCTS             : 3323    7    7  5000    44.4 %   3369   53.4 %
  15 Ethereal 12.00 pext        : 3302    5    5 13000    36.7 %   3411   46.2 %
  16 Ethereal 11.75 pext        : 3291    6    6  9000    39.3 %   3375   53.2 %
  17 Xiphos 0.6 bmi2            : 3282    4    4 18000    36.0 %   3396   50.1 %
  18 Fire 7.1 popc              : 3281    3    3 31000    43.9 %   3333   51.8 %
  19 Xiphos 0.5.6 bmi2          : 3271    6    6  7000    41.2 %   3339   54.6 %
  20 Ethereal 11.53 pext        : 3263    7    7  7000    42.2 %   3325   53.4 %
  21 Komodo 12.3 MCTS           : 3258    7    7  7000    42.7 %   3316   46.3 %
  22 Ethereal 11.25 pext        : 3254    7    7  6000    38.4 %   3344   51.0 %
  23 rofChade 2.3 bmi2          : 3240    7    7  6000    32.6 %   3377   47.4 %
  24 Booot 6.4 popc             : 3227    7    7  6000    31.1 %   3377   46.5 %
  25 Schooner 2.2 popc          : 3224    7    7  6000    31.3 %   3373   50.3 %
  26 Laser 1.7 bmi2             : 3201    7    7  6000    30.8 %   3353   45.8 %
  27 Fizbo 2 bmi2               : 3196    8    8  5000    36.0 %   3308   39.0 %
  28 Fritz 17                   : 3195    7    7  6000    29.4 %   3359   44.2 %
  29 Shredder 13 x64            : 3192    8    8  6000    31.9 %   3341   42.6 %
  30 Defenchess 2.2 popc        : 3188    8    8  5000    26.6 %   3377   41.8 %
  31 Booot 6.3.1 popc           : 3182    8    8  5000    34.0 %   3310   44.1 %
  32 Andscacs 0.95 popc         : 3150    9    9  5000    23.1 %   3373   35.4 %

The version-numbers (180622 for example) of the engines are the date of the latest patch, which was included in the Stockfish sourcecode, not the release-date of the engine-file. Especially the asmFish-engines are often released much later!!

Below you find a diagram of the progress of Stockfish in my tests since the end of 2018. 

And below that diagram, the older diagrams.

You can save the diagrams (as a JPG-picture (in originial size)) on your PC with mouseclick (right button) and then choose "save image"...

The Elo-ratings of older Stockfish dev-versions in the Ordo-calculation can be a little different to the Elo-"dots" in the diagram, because the results/games of new Stockfish dev-versions - when getting part of the Ordo-calculation - can change the Elo-ratings of the opponent engines and that can change the Elo-ratings of older Stockfish dev-versions (in the Ordo-calculation / ratinglist, but not in the diagram, where all Elo-"dots" are the rating of one Stockfish dev-version at the moment, when the testrun of that Stockfish dev-version was finished).