Stefan Pohl Computer Chess

Latest Website-News (2024/10/20): Testrun of Viridithas 15.0 finished: +45 Celo to Viridithas 14.0. Very nice progress, but Viridithas still has a very small EAS-score...

 

I released an update of my EAS-Tool (now V5.8). No functional change, only cleaner output and a small bugfix. Gaunlet-EAS-tool stays unchangend...Download here

 

Dont forget to take a look at my EAS-Ratinglist (The world's first engine-ratinglist not measuring strength of engines, but engines's style of play)

 

Update: Alexander 2.0 alew setting testrun aborted, after I was imformed about the inacceptable behaviour of the author of Alexander on Stockfish discord ("If you want to ban me for this, this is nazism"). For me, coming from Germany, blaming anyone or anything as "nazism" is completely unacceptable.

 

Next NN-testrun: New Lc0 official release binary 0.31.2 with BT4-6147500pt332. Lets see, if the new binary is better than 0.31.1 (and if it runs stable without disconnects (like the dev-binaries of the last months, which again and again disconnected))

 

 

Stay tuned

UHO-Top15 Engines Ratinglist (+ regular testing of Stockfish Dev-versions)

 

 

Playing conditions:

 

The 15 strongest engines and the latest StockfishDev version are playing 1000 games vs. each opponent: 15000 games per engine are played. A RoundRobin Tournament with 120000 games overall !

 

Hardware: AMD Ryzen 9 7945HX 16-core (32 threads) notebook with 64GB RAM (since 2024/10/08)

Speed: (singlethread, TurboBoost-mode switched off, chess starting position) Stockfish 17: 600 kn/s (when 26 games are running simultaneously)

Hash: 512MB per engine

GUI: Cutechess-cli (GUI ends game, when a 5-piece endgame is on the board, all other games are played until mate or draw by chess-rules (3fold, 50-moves, stalemate))

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

Openings: My UHO_2022_6mvs_+120_+129 openings are used (first 500 lines) (UHO 2022 openings are part of my Anti Draw Openings download-package). 

Ponder, Large Memory Pages & learning: Off

Thinking time: 3min+1sec per game/engine. One testrun (15000 games) takes around 80-85 hours (=3.5 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: 200807 = August, 7, 2020). The used SF compile is the AVX512-compile, which is the fastest on my AMD Ryzen CPU. SF binaries are taken from https://github.com/official-stockfish/Stockfish/releases (except the official SF-release versions, which are taken form the official Stockfish website).

Read the explanation, why I chose my testing-conditions, like I did here

 

Latest update: 2024/10/20: Viridithas 15.0 (+45 Celo to Viridithas 14.0)

 

(Ordo-calculation fixed to Stockfish 17 = 3843 Celo. Mention, there are no human or "realistic" Elo numbers in enginechess, since the engines reached superhuman strength, so I decided to use "Celo" (=Computer Elo) instead)

 

See the tournament cross-table as a picture (done by Fritz18 GUI) here

See the individual statistics of engine-results here

See the Engines Aggressiveness Score Ratinglist here

See the Gamepair-statistics of the UHO-Top15 Ratinglist here

See the full UHO Ratinglist (with EAS-Ratinglist): here

Download the current gamebase, the full-list gamebase and the archive here

See the old SPCC-Ratinglist (full list) and EAS-List from 2020 until 2023/08/28: here

 

(best Stockfish Celo: Stockfish 241004: 3850 Celo, latest full release (Stockfish 17) had 3843 Celo)

(highest EAS-Score since Stockfish 16 release: Stockfish 240519: 203882 EAS-points)

(Read the explanation of the Lc0-entry in the Ratinglist here).

     Program                    Celo    +    - Games    Score   Av.Op. Draws

   1 Stockfish 241012 a512    : 3849    4    4 15000    70.9%   3687   48.2%
   2 Stockfish 17 240906      : 3843    4    4 15000    70.2%   3688   48.7%
   3 Torch 3 popavx2          : 3807    4    4 15000    65.5%   3690   48.6%
 [   Lc0 0.32dev BT4-6077500  : 3762   15   15  1000 (NN vs SF15.1 testing)]
   4 Berserk 13 avx2          : 3734    4    4 15000    55.5%   3695   50.6%
   5 KomodoDragon 3.3 avx2    : 3729    4    4 15000    54.6%   3695   51.2%
   6 Obsidian 13.08 avx2      : 3727    4    4 15000    54.4%   3695   51.3%
   7 Caissa 1.20 avx2         : 3686    4    4 15000    48.4%   3698   51.0%
   8 PlentyChess 2.1 avx2     : 3684    4    4 15000    48.2%   3698   51.9%
   9 Ethereal 14.38 avx2      : 3683    4    4 15000    48.0%   3698   50.5%
  10 RubiChess 240817 avx2    : 3669    4    4 15000    46.0%   3699   50.0%
  11 Viridithas 15.0 a512     : 3661    4    4 15000    44.8%   3700   50.9%
  12 Alexandria 7.0 avx2      : 3654    4    4 15000    43.8%   3700   51.3%
  13 Clover 8.0 avx2          : 3644    4    4 15000    42.4%   3701   51.1%
  14 Lizard 11.0 avx2         : 3601    4    4 15000    36.4%   3704   48.5%
  15 Motor 0.70 avx2          : 3598    4    4 15000    35.9%   3704   51.0%
  16 Titan 1.1 avx2           : 3591    4    4 15000    35.0%   3705   49.6%

Games        : 120000 (finished)

White Wins   : 57810 (48.2 %)
Black Wins   : 1878   (1.6 %)
Draws        : 60312 (50.3 %)

Below the UHO-Top15 Ratinglist, recalculated with Gamepairs (using my Gamepair Rescoring Tool V1.5), Ordo-calculation fixed to Stockfish 17 = 3843 Celo, realizing Vondele's (Stockfish Maintainer) idea: "Thinking uniquely in game pairs makes sense with the biased openings used these days. While pentanomial makes  sense it is a bit complicated so we could simplify and score game pairs only (not games) as W-L-D (a traditional score of 2-0, or 1.5-0.5 is just a W)."

See the head-to-head gamepair-statistics of the UHO-Top15 Ratinglist here

   # PLAYER                   :    Celo  Error   Pairs     W     D     L   (%)  CFS(%)
   1 Stockfish 241012 a512    :    3867     13    7500  6162  1202   136  90.2     100
   2 Stockfish 17 240906      :    3843   ----    7500  6031  1267   202  88.9     100
   3 Torch 3 popavx2          :    3702     12    7500  5026  1853   621  79.4     100
   4 Berserk 13 avx2          :    3503     12    7500  3194  2648  1658  60.2      98
   5 KomodoDragon 3.3 avx2    :    3493     12    7500  3014  2842  1644  59.1      97
   6 Obsidian 13.08 avx2      :    3484     12    7500  2960  2808  1732  58.2     100
   7 Caissa 1.20 avx2         :    3389     12    7500  2053  2921  2526  46.8      87
   8 PlentyChess 2.1 avx2     :    3384     12    7500  1932  3076  2492  46.3      63
   9 Ethereal 14.38 avx2      :    3382     13    7500  1982  2951  2567  46.1     100
  10 RubiChess 240817 avx2    :    3356     12    7500  1769  2902  2829  42.9     100
  11 Viridithas 15.0 a512     :    3333     12    7500  1565  2897  3038  40.2     100
  12 Alexandria 7.0 avx2      :    3316     12    7500  1408  2897  3195  38.1     100
  13 Clover 8.0 avx2          :    3290     12    7500  1202  2856  3442  35.1     100
  14 Lizard 11.0 avx2         :    3184     13    7500   578  2373  4549  23.5      73
  15 Motor 0.70 avx2          :    3181     12    7500   567  2351  4582  23.2     100
  16 Titan 1.1 avx2           :    3166     13    7500   487  2296  4717  21.8     ---

 

------------------------------------------------------------------- 
--- Number of all Gamepairs          : 60000 
--- Number of drawn Gamepairs overall: 20070 (= 33.45%) 
--- Number of 1:1 drawn Gamepairs    : 9176  (= 15.29%) 
--- Number of 2-draws drawn Gamepairs: 10894 (= 18.16%) 
------------------------------------------------------------------- 

Aborted testruns, because the tested engine was too weak to enter the UHO-Top15 Ratinglist:

Minic 3.41, Booot 7.3, Black Marlin 9.0, Equisetum 1.0, Stormphrax 5

Below you find a diagram of the progress of Stockfish in my tests since February 2023. 

And below that diagram, the older diagrams.

 

You can save the diagrams on your PC with mouseclick (right button) and then choose "save image"...