Stefan Pohl Computer Chess

Home of famous UHO openings and EAS Ratinglist

Patricia - the EAS ChessEngine

The most aggressive superhuman engine ever created

"Patricia is brilliant" - Ed Schroeder (Computerchess legend, author of Rebel engine and famous chesscomputers (Mephisto MM IV, Polgar, Milano, Risc and many more!))

"I am checking games with Patricia engine. It's totally crazy! I haven't seen any engine play like this" - GM Vidit Gujrathi (world ranking #26, Indian Chess Olympiad gold medalist in 2024), 2025/07/27 on Lc0-discord

Patricia by Adam Kulju is the first engine besides Ed Schroeder's Rebel EAS, using my EAS-Tool (= Engine Aggressiveness Statistics Tool) to make the engine playing more aggressive:

It's worth noting that Patricia is still well into superhuman territory, with an estimated CCRL elo of 3500; she can put up a solid fight against top engines, and can crush any human or weak engine that dares to oppose her.

If you think Patricia is a cool project, please spread the word of it! There are lots of people interested in an aggressive engine, whether it be for sparring or for other purposes, and greater awareness of what should be the undisputed queen of style would make their wish come true.

Since the 1950s, the only goal of computerchess was gaining Elo. But in these days of superstrong engines beyond 3700+ Elo, IMHO it makes a lot of sense, to make engines playing more spectacular, aggressive and interesting, instead of just gaining more and more Elo... Since my EAS-Tool was made, it is possible for the first time, to measure the aggressiveness of engines. And so, using the EAS-Tool, to make an engine playing more aggressive, is the next logical step of development in computerchess.

2025/08/05 Patricia 5.0 using 3 nets (1 super-aggressive net, when playing with reduced strength vs. humans), 1 aggressive net for normal play and 1 net for the endgame and when Patricia is worse) is available. Jim Ablett made very fast binaries of Patricia code from github. Download here

In my testrun, Patricia 5.0 played +84 Celo stronger than Patricia 4.0 (strength is now somewhere around Stockfish 11) and got an EAS-score of 374208 (around +34000 better than Patricia 4.0)

My small Gambit Opening-book (using J.Noomens Gambit-files) for Patricia is here

ArenaGUI (Windows) with Patricia 5.0 installed and ready for playing here

The author of Patricia (Adam Kulju) wrote this:

"Patricia is, per the EAS tool, unquestionably the most aggressive superhuman chess engine ever created (and the most aggressive engine of any kind that I currently know of).

The metric that Patricia's aggression claims are based off of is Stefan Pohl's EAS-Tool, which is the most well known and well regarded tool for determining the aggressiveness of chess engines. It looks at a combination of factors, such as sacrifice rate, short win rate, and unnecessary draw rate, and outputs a score that captures how "exciting" an engine tends to play.
A huge shoutout to Stefan Pohl. His EAS Tool works wonderfully, makes properly and objectively testing for increase aggression possible, and is the measure by which Patricia development progressed. He was also very invested in Patricia's development and explained many features of the EAS tool to me so that I had a better understanding of what was a sacrifice and what wasn't. He has put a lot of time into quantifying style, and I am happy to have Patricia be a proof-of concept and culmination of his ideas."

I did a testrun of Patricia 5.0 vs. 14 weaker engines of my full UHO ratinglist. Mention, weaker engines means here 3500+ Celo, so these engine are way too strong for Patricia 5.0. You can see the result in the Full UHO Ratinglist

Here the bottom of the Full UHO Ratinglist:

Program Celo + - Games Score Av.Op. Draws

132 Stockfish final HCE : 3470 3 3 26000 44.6% 3510 38.9% 133 Patricia 5.0 avx512 : 3465 4 4 14000 42.4% 3520 44.9% 134 Slow Chess 2.9 avx2 : 3449 3 3 26000 42.5% 3506 43.3% 135 Cerberus 21124081r81 : 3447 4 4 14000 39.9% 3520 37.4% 136 Patricia 250510 a512 : 3441 5 5 14000 39.0% 3520 43.0% 137 Rebel Extreme avx2 : 3402 5 5 14000 34.0% 3520 39.3% 138 Patricia 4.0 avx2 : 3381 5 5 14000 31.4% 3520 32.2% 139 Revenge 1.0 avx2 : 3356 5 5 15000 18.3% 3628 30.3% 140 CSTal 2.1 EAS : 3343 5 5 14000 26.9% 3520 29.6% 141 Monty 241220 a512 : 3321 5 5 14000 24.6% 3520 35.4% 142 Patricia 3.1 avx2 : 3210 6 6 14000 14.7% 3520 21.1%

And now, see the top of the EAS full Ratinglist (calculated out of the games played for the Full 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 3 379711 50.62% 36.92% 07.20% 68 CSTal 2.1 EAS 4 374208 50.79% 36.98% 05.85% 67 Patricia 5.0 avx512 5 350375 48.39% 32.33% 05.82% 67 Cerberus 21125081r9b 6 340432 46.96% 29.39% 03.46% 70 Patricia 4.0 avx2 7 315346 43.56% 27.63% 04.21% 68 Cerberus 21124081r81 8 253682 37.88% 23.88% 08.79% 75 Rebel Extreme avx2 9 221083 32.46% 23.04% 09.25% 71 Velvet 8.1.0 risky 10 204351 34.75% 22.22% 10.27% 73 Velvet 8.0.0 risky 11 200919 31.18% 29.46% 17.09% 71 Revenge 1.0 avx2 12 196748 21.29% 26.40% 09.17% 70 Stockfish 16 230630

That there is an EAS regression to Patricia 3.1 is no surprise for me: Gaining +255 Celo must lead to such a regression. But IMHO the EAS-Score is still very, very impressive: Better than all other aggressive playing engines and way better than Stockfish 16. And Stockfish 17.1 has only 158647 EAS-points, so Patricia 5 has more than twice as many points as the latest Stockfish. Awesome!

Here 20 impressive wins of Patricia 5.0 in this testrun in the pgn-viewer (engines started thinking at move 7).

If you can not see the chessboard, check, that your browser has Javascript activated or if an AdBlocker is the problem.