Stefan Pohl Computer Chess

Latest Website-News (2025/05/14): Testrun of Patricia 250510 avx512 (dev) finished. Binaries by Jim Ablett, download here. The result is impressive, because there are clear improvements in Celo and EAS at the same time (higher EAS-scoring normally lowers Celo-strength...). Patricia is on the same EAS-level as CSTal 2.1 EAS, but plays nearly +100 Celo stronger (!)

 

Worst-10 of my full UHO ratinglist (Celo):

     Program                     Celo    +    - Games    Score   Av.Op. Draws

 114 Stockfish final HCE       : 3468    3    3 24000    44.1%   3513   38.6%
 115 Slow Chess 2.9 avx2       : 3448    4    4 24000    42.2%   3508   42.7%
 116 Cerberus 21124081r81      : 3447    4    4 14000    39.9%   3520   37.4%
 117 Patricia 250510 a512      : 3440    4    4 14000    39.0%   3520   43.0% (NEW)
 118 Rebel Extreme avx2        : 3402    5    5 14000    34.0%   3520   39.3%
 119 Patricia 4.0 avx2         : 3381    4    4 14000    31.4%   3520   32.2% (OLD)
 120 Revenge 1.0 avx2          : 3356    5    5 15000    18.3%   3628   30.3%
 121 CSTal 2.1 EAS             : 3343    5    5 14000    26.9%   3520   29.6%
 122 Monty 241220 a512         : 3321    5    5 14000    24.6%   3520   35.4%
 123 Patricia 3.1 avx2         : 3209    6    6 14000    14.7%   3520   21.1%

Top-10 of my full EAS UHO ratinglist:

                                 bad  avg.win 
Rank  EAS-Score  sacs   shorts  draws  moves  Engine/player 
-------------------------------------------------------------------
   1    432481  51.61%  38.03%  05.49%   66   Patricia 3.1 avx2  
   2    381256  50.51%  36.25%  06.08%   67   Patricia 250510 a512 (NEW) 
   3    379711  50.62%  36.92%  07.20%   68   CSTal 2.1 EAS  
   4    340432  46.96%  29.39%  03.46%   70   Patricia 4.0 avx2 (OLD) 
   5    315346  43.56%  27.63%  04.21%   68   Cerberus 21124081r81  
   6    253682  37.88%  23.88%  08.79%   75   Rebel Extreme avx2  
   7    221083  32.46%  23.04%  09.25%   71   Velvet 8.1.0 risky  
   8    204351  34.75%  22.22%  10.27%   73   Velvet 8.0.0 risky  
   9    200919  31.18%  29.46%  17.09%   71   Revenge 1.0 avx2  
  10    196748  21.29%  26.40%  09.17%   70   Stockfish 16 230630  

 

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 testruns: Horsie 1.1 followed by latest Stockfish-dev.

 

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: 2025/05/09: Obsidian 15.19 (-2 Celo to Obsidian 15.14)

 

(Ordo-calculation fixed to Stockfish 17.1 = 3854 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 250321: 3858 Celo, latest full release (Stockfish 17.1) had 3854 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 250427 a512    : 3856    4    4 15000    69.3%   3709   47.9%
   2 Stockfish 17.1 250330    : 3854    4    4 15000    69.1%   3709   48.3%
   3 Torch 3.1 a512           : 3800    4    4 15000    61.9%   3713   48.6%
   4 Obsidian 15.19 a512      : 3774    4    4 15000    58.2%   3715   48.6%
 [   Lc0 0.32dev BT4-6077500  : 3753   15   15  1000 (NN vs SF15.1 testing)]
   5 Berserk 250307 a512      : 3739    4    4 15000    53.1%   3717   50.1%
   6 PlentyChess 5.0 a512     : 3737    4    4 15000    52.9%   3717   49.0%
   7 Alexandria 8.0 a512      : 3729    4    4 15000    51.7%   3718   49.8%
   8 KomodoDragon 3.3 avx2    : 3707    4    4 15000    48.4%   3719   50.2%
   9 Caissa 1.22 a512         : 3702    4    4 15000    47.6%   3719   50.9%
  10 Integral 7.0 a512        : 3698    4    4 15000    47.1%   3720   48.4%
  11 Viridithas 17.0 a512     : 3677    4    4 15000    44.0%   3721   50.6%
  12 Ethereal 14.38 avx2      : 3656    4    4 15000    41.0%   3722   50.0%
  13 Horsie 1.0 a512          : 3648    4    4 15000    39.9%   3723   49.7%
  14 RubiChess 240817 avx2    : 3641    4    4 15000    39.0%   3723   49.1%
  15 Starzix 6.0 a512         : 3638    4    4 15000    38.5%   3724   49.2%
  16 Clover 8.1 avx2          : 3637    4    4 15000    38.3%   3724   50.8%

 

Games        : 120000 (finished)

White Wins   : 59132 (49.3%)
Black Wins   : 1532   (1.3%)
Draws        : 59336 (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 250427 a512    :    3861     12    7500  5799  1523   178  87.5      88
   2 Stockfish 17.1 250330    :    3854   ----    7500  5739  1574   187  87.0     100
   3 Torch 3.1 a512           :    3681     11    7500  4207  2476   817  72.6     100
   4 Obsidian 15.19 a512      :    3616     12    7500  3568  2731  1201  65.8     100
   5 Berserk 250307 a512      :    3532     11    7500  2641  3096  1763  55.9      83
   6 PlentyChess 5.0 a512     :    3528     11    7500  2580  3137  1783  55.3     100
   7 Alexandria 8.0 a512      :    3509     11    7500  2435  3074  1991  53.0     100
   8 KomodoDragon 3.3 avx2    :    3463     11    7500  1972  3133  2395  47.2     100
   9 Caissa 1.22 a512         :    3448     11    7500  1841  3106  2553  45.3      98
  10 Integral 7.0 a512        :    3439     11    7500  1729  3161  2610  44.1     100
  11 Viridithas 17.0 a512     :    3397     11    7500  1385  3052  3063  38.8     100
  12 Ethereal 14.38 avx2      :    3346     12    7500  1005  2895  3600  32.7     100
  13 Horsie 1.0 a512          :    3323     12    7500   857  2774  3869  29.9      65
  14 RubiChess 240817 avx2    :    3321     12    7500   846  2765  3889  29.7     100
  15 Starzix 6.0 a512         :    3307     11    7500   752  2724  4024  28.2      99
  16 Clover 8.1 avx2          :    3298     11    7500   712  2643  4145  27.1     ---

 

------------------------------------------------------------------- 
--- Number of all Gamepairs          : 60000 
--- Number of drawn Gamepairs overall: 21932 (= 36.55%) 
--- Number of 1:1 drawn Gamepairs    : 10739 (= 17.90%) 
--- Number of 2-draws drawn Gamepairs: 11193 (= 18.65%) 
------------------------------------------------------------------- 

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

Renegade 1.2.0, 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"...