Stefan Pohl Computer Chess

Latest Website-News (2025/03/21): Testrun of Obsidian 15.09 finished. +16 Celo to Obsidian 15.0. Very impressive progress in only 6 weeks (!). The author asked me for a new testrun and provided an avx512 binary. You can download it here

 

Ceres Binary here, Ceres Installation Guide here, Ceres Nets 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)

 

Next testrun: Stockfish 250321

 

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), 2 identical machines. Windows 11 64bit

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: 2025/03/21: Obsidian 15.09 (+16 Celo to Obsidian 15.0)

 

(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 250228: 3853 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 250228 a512    : 3853    4    4 15000    69.6%   3705   48.0%
   2 Stockfish 17 240906      : 3843    4    4 15000    68.2%   3705   49.2%
   3 Torch 3.1 a512           : 3802    4    4 15000    62.7%   3708   48.3%
   4 Obsidian 15.09 a512      : 3773    4    4 15000    58.7%   3710   49.5%
 [   Lc0 0.32dev BT4-6077500  : 3754   15   15  1000 (NN vs SF15.1 testing)]
   5 Berserk 250307 a512      : 3740    4    4 15000    53.9%   3712   50.1%
   6 Alexandria 8.0 a512      : 3730    4    4 15000    52.5%   3713   49.4%
   7 PlentyChess 4.0 a512     : 3715    4    4 15000    50.3%   3714   48.4%
   8 KomodoDragon 3.3 avx2    : 3710    4    4 15000    49.5%   3714   50.0%
   9 Integral 7.0 a512        : 3700    4    4 15000    48.0%   3715   48.4%
  10 Caissa 1.21 a512         : 3688    4    4 15000    46.2%   3716   50.5%
  11 Ethereal 14.38 avx2      : 3657    4    4 15000    41.7%   3718   50.1%
  12 Horsie 1.0 a512          : 3649    4    4 15000    40.7%   3718   49.9%
  13 Viridithas 16.0 a512     : 3646    4    4 15000    40.2%   3719   49.3%
  14 RubiChess 240817 avx2    : 3643    4    4 15000    39.8%   3719   49.4%
  15 Starzix 6.0 a512         : 3638    4    4 15000    39.1%   3719   49.0%
  16 Clover 8.1 avx2          : 3637    4    4 15000    38.9%   3719   50.7%

Games        : 120000 (finished)

White Wins   : 59159 (49.3%)
Black Wins   : 1580   (1.3%)
Draws        : 59261 (49.4%)

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)."

Mention, the errorbar here is +/-5.5 Celo (=11 overall Error), compared to +/-4 Celo in the classical ratinglist above (=1.375 wider errorbar here). But the result-gaps are (at least) 2x up to 2.5x bigger (in average), calculating with gamepairs. So, the statistical reliability of the results is much better in the Gamepair-ratinglist, compared to the classical ratinglist...

CFS means Chance For Superiority in percent, calculated by ORDO.

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

   # PLAYER                   :    Celo  Error   Pairs     W     D     L   (%)  CFS(%)
   1 Stockfish 250228 a512    :    3878     12    7500  5853  1481   166  87.9     100
   2 Stockfish 17 240906      :    3843   ----    7500  5584  1665   251  85.6     100
   3 Torch 3.1 a512           :    3708     11    7500  4366  2405   729  74.2     100
   4 Obsidian 15.09 a512      :    3634     11    7500  3601  2779  1120  66.5     100
   5 Berserk 250307 a512      :    3555     11    7500  2793  3003  1704  57.3     100
   6 Alexandria 8.0 a512      :    3533     11    7500  2569  3047  1884  54.6     100
   7 PlentyChess 4.0 a512     :    3498     11    7500  2183  3172  2145  50.3      97
   8 KomodoDragon 3.3 avx2    :    3490     11    7500  2127  3126  2247  49.2     100
   9 Integral 7.0 a512        :    3463     11    7500  1832  3212  2456  45.8     100
  10 Caissa 1.21 a512         :    3437     11    7500  1603  3174  2723  42.5     100
  11 Ethereal 14.38 avx2      :    3368     11    7500  1091  2921  3488  34.0     100
  12 Viridithas 16.0 a512     :    3347     11    7500   954  2816  3730  31.5      53
  13 Horsie 1.0 a512          :    3347     11    7500   943  2832  3725  31.5      65
  14 RubiChess 240817 avx2    :    3345     11    7500   940  2806  3754  31.2     100
  15 Starzix 6.0 a512         :    3330     11    7500   795  2832  3873  29.5      98
  16 Clover 8.1 avx2          :    3320     11    7500   756  2749  3995  28.4     ---

 

------------------------------------------------------------------- 

--- Number of all Gamepairs          : 60000 
--- Number of drawn Gamepairs overall: 22010 (= 36.68%) 
--- Number of 1:1 drawn Gamepairs    : 10809 (= 18.01%) 
--- Number of 2-draws drawn Gamepairs: 11201 (= 18.67%) 
------------------------------------------------------------------- 

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

Igel 3.6.0, Minic 3.41, Booot 7.3, Black Marlin 9.0, Equisetum 1.0

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"...