Chess Engines Diary

Very large test tournament - 1200 games won by the Kookaburra 3.00 chess engine. In 2nd place for the JigSaw 5.9 engine. 3rd place for the Marauders 3.6 chess engine. GPU NVIDIA  RTX 3060 ti. GUI-Banksia, For Every Engine only 1 CPU and 512 Hashtables, Book - 4 GM mvs book,  6TBs. 🕒🕒Time 3'+3"   1200 (!!!) tournaments games download   💾 All games CEDR  381.115 games download  01.07.2024 (3'+3")   @chessenginesdiary P Engine Score +/-/= 1: Kookaburra 3.00 78.0 / 150 (+7 -1 =142) 2: JigSaw 5.9 77.5 / 150 (+6 -1 =143) 3: Marauders 3.6 77.0 / 150 (+5 -1 =144) 4: Killfish 150624 76.5 / 150 (+3 -0 =147) 5: Tactical Titan 76.5 / 150 (+4 -1 =145) 6: Yuli GM Pro 16 76.0 / 1...

Small test tournament of the strongest chess engines. Three engines achieved the same result: Kookaburra 3.00, Stockfish, dev-20240723 and CorChess 4.5 dev-20240724 .   All CEDR games  381.115 download    💾  220 games from the tournament download @chessenginesdiary Country - Poland, City -  Malbork   🕓 Time 3'+3"  💻HP Pavilion i5-1035G1 8GB RAM 🖬 GUI-Fritz 17 P Engine Score +/-/= 1: Kookaburra 3.00 21.5 / 40 (+3 -0 =37) 2: Stockfish, dev-20240723 21.5 / 40 (+3 -0 =37) 3: CorChess 4.5 dev-20240724 21.5 / 40 (+3 -0 =37) 4: HypnoS dev-20240727 21.0 / 40 (+2 -0 =38) 5: Stockfish 16.1 21.0 / 40 (+4 -2 =34) 6: Fisherov chess monk 1.2 20.5 / 40 (+1 -0 =39) 7: SF-PRO 22.07.2024 20.5 / 40 (+1 -0 =39) 8: Stockfish dev-20240723 20.0 / 4...

Catto - UCI chess engine Author: Phu Minh  Rating CEDR=2316 New versions of the Catto engine are released every few moments, so there is no time for thorough testing. Maybe the author should reduce the frequency of releasing new builds to fine-tune them?  What do we currently have? The engine Negamax search algorithm with Alpha-Beta pruning. Move ordering: MVV-LVA heuristic. Killer heuristic. History heuristic. Countermove heuristic. Checkmate and stalemate detection. Quiescence search. Evalution: PeSTO evaluation. Pawn structure. UCI. v.0.12.1: Minor hash optimization. Catto 0.12.1 download

 As many as 33 chess engines and over 1000 games were played in the next test tournament. The winner was the Yuliana 5.0 engine (currently available only to our testers). Places 2-3 were shared by the Stockfish dev-20240723 and Stockfish 16.1 engines. More tests soon, we are also preparing a new ranking list for August 1, 2024. All CEDR games  381.115 download    💾  1056 (!!!) games from the tournament download @chessenginesdiary Country - Poland, City -  Malbork   🕓 Time 3'+3"  💻HP Pavilion i5-1035G1 8GB RAM 🖬 GUI-Banksia Tech table: Engine KN/mov NPS dep/mov time/mov mov/game time/game fails Aramis 1.3.0 1364 264585 14.6 5.2 41.2 212.2 Arcanum 2.2 4037 691778 13.7 5.8 44.8 261.4 Bagatur 5.1a 158 21966 13.0 7.2 31.7 228.5 Brainlearn 28.1 2283 404630 29.2 5.6 48.5 273.5 Bread Engine 1.0.0 4870 804914 11.1 6.1 43.4 262.6 Caissa 1.20 5503 1012612 29.8 5.4 52.0 28...

Mufasa - chess engine written in C++ with the help of CMake and GTest framework. Author - Alikhan Sirgaliyev License Mufasa uses [GNU GPLv3] (https://choosealicense.com/licenses/gpl-3.0/), which means you can do anything with the project, except distributing closed source versions. v.0.2.1: Adjustments to resolve time management issues and compilation errors Mufasa 0.2.1 download

  Bread Engine   is a chess engine written in c++. I started working on it in 2021, and only just finished. There is still a lot of room for improvement, but the engine is quite strong (for humans, at least). It uses NNUE (efficiently updatable neural network) to evaluate positions, as well as minimax search. Bread engine does not have a GUI built in, however it supports the uci protocol, you can therefore run it on any chess GUI. Author: Nonlinear2 Improved UCI commands, neural network is now always embedded in the executable.  Bread engine is now a single executable, that you can download and run directly. v.1.1.0: Bread Engine 1.1.0 is around 130 elo stronger than version 1.0.0 in short time controls, and 65 elo stronger in long time controls. Additions: Mate scores Internal iterative reductions uci information for tablebase moves history heuristic razoring static exchange evaluation pruning and... there is now a nonsense option, which makes Bread bongcloud occasionall...

  The first public release of Integral Author - Aron Petkovski v3: A lot of improvements have been made to the search as well as the evaluation. Syzygy tablebase support has been added, and Integral now has multithreaded search using the Lazy SMP technique. Progression Test +381 elo at STC over Integral v2 Elo   | 381.51 +- 27.65 (95%) Conf  | 8.0+0.08s Threads=1 Hash=16MB Games | N: 1014 W: 825 L: 14 D: 175 Penta | [0, 2, 30, 137, 338] https://chess.aronpetkovski.com/test/2268/ Integral v1 results: Drosophila 1.6 JA 7/7 +7 7 Games Cyrano 0.6b17 JA 6.5/7 +6 7 Games Odonata 0.9.0 1.5/7 -4 7 Games Obsidian 12.29 0/7 -7 7 Games Bagatur 5.0 4/5 +3 5 Games Elixir 1.0 4/5 +3 5 Games Lime 66 JA 4/5 +3 5 Games DoctorB 1.2.1 JA 4/5 +3 5 Games Princhess 0.17.0 2/5 -1 5 Games Vault 0.4.1 JA 1.5/5 -2 5 Games Svart 6 JA 1/5 -3 5 Games Stash 35.26 JA 0.5/5 -4 5 Games Wasp 7.00 0/5 -5 5 Games Beast 15 MPV 0/4 -4 4 Games PlentyChess 2.0.0 0/4 -4 4 Games Stockfish 16.1 0/4 -4 4 Games...

Calvin Chess - java chess engine v.4.0.0: NNUE finally comes to Calvin! And brings with it ~400 elo (at least in self-play). Calvin's hand-crafted evaluation function has been fully removed and replaced with an Efficiently Updated Neural Network (NNUE). Architecture is (768->256)x2->1. The net was trained on 250 million positions taken from the Leela Chess Zero dataset, that I re-scored using Calvin's own search and HCE. For training I used the excellent bullet NNUE trainer. (I started off by writing my own trainer, but I couldn't get it anywhere near the speed of bullet). Other minor features include: ⚡ Minor move generation optimisations (#21) ⚡ Add age to TT replacement scheme (#23) ⚡ Add static eval to TT hash entry (#27) ⚖️ Add training data scorer to re-score Leela data for NNUE training (#44) 🔍 Extend search 1 ply when entering a pawn endgame (#22) 🐛 Fix bug where killer ply > max ply (#43) Score of Calvin DEV vs Calvin: 436 - 19 - 45  [0.917] 500 Elo dif...

The first test of the new version of the Caissa 1.20 engine is here. It lost all mini-matches with engines based on the Stockfish source code. Caissa drew with Lc0 0.31.0 and Berserk 13 , and tied with the previous version of Caissa 1.19 . Overall result +3 -14 =83 with very strong chess engines. All CEDR games  381.115 download    💾  100 games from the tournament download @chessenginesdiary Country - Poland, City -  Malbork   🕓 Time 3'+3"  💻Asus Zenbook, AMD Ryzen 9 6900HX, RAM 32,0 GB 🖬 GUI-Fritz 17 P Engine Score Caissa 1.2 +/-/= 1: Caissa 1.20 44.5 / 100 XXXXXXXX (+3 -14 =83) 2: JigSaw 5.9 7.0 / 10 =====1111= (+4 -0 =6) 3: Yuliana 5.0 6.5 / 10 ===1=1===1 (+3 -0 =7) 4: CorChess 4.5 dev-20240724 6.0 / 10 ======11== (+2 -0 =8) 5: Obsidian 13.0 6.0 / 10 ======1==...

The strongest new chess engines were tested in another tournament. The winner was the SF-PRO 22.07.2024 engine (available only to our testers for now). Places 2 to 4 were taken by the Beast 15 MPV , Deep SOLIDOR and Stockfish 20240723 Ivec engines. We invite you to visit our website, next tournaments coming soon. All CEDR games  381.115 download    💾  364 games from the tournament download @chessenginesdiary Country - Poland, City -  Malbork   🕓 Time 3'+3"  💻Asus Zenbook, AMD Ryzen 9 6900HX, RAM 32,0 GB 🖬 GUI-Banksia Tech table: Engine KN/mov NPS dep/mov time/mov mov/game time/game fails Beast 15 MPV 10869 2102255 47.1 5.2 70.4 363.8 Caissa 1.19 24013 4594419 40.6 5.2 72.6 379.6 Clarity 7.2.0 6535 1276771 35.6 5.1 71.3 364.7 Clover 7.0 23742 4681286 40.1 5.1 79.7 404.4 CorChess 4.5 dev-20240724 11170 2035884 41.3 5.5 65.9 361.7 Deep SOLIDOR AI 9122 1659887 31.5 5.5 65.7 360.8 Kookaburra 3.00 11635 2180127 50.7 5.3 66.4 ...

Stockfish 2024072820  - UCI chess engine, compiled by  Shawn Xu Rating Chess Engines Diary CEDR=3766 Stockfish 2024072820  Shawn Xu  (all versions) download Timestamp: 1722198170 Remove Killers The removal of killers on line 1774 resulted in a substantial decrease in pre-LMR history average, so a negative history fill is applied to counter it. Passed Non-regression STC (vs #5513): LLR: 2.95 (-2.94,2.94) <-1.75,0.25> Total: 21984 W: 5886 L: 5645 D: 10453 Elo +3.81 Ptnml(0-2): 80, 2492, 5628, 2691, 101 https://tests.stockfishchess.org/tests/view/66a095894ff211be9d4ecb9d Passed Non-regression LTC (vs #5513): LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 95430 W: 24141 L: 23995 D: 47294 Elo +0.53 Ptnml(0-2): 97, 10537, 26298, 10689, 94 https://tests.stockfishchess.org/tests/view/66a11c8d4ff211be9d4ecbf8 closes https://github.com/official-stockfish/Stockfish/pull/5517 Bench: 1660869 Stockfish 20240521 - results: Sun 14 3.5/6 +1 6 Games Raid 3.6 TR 3/6 +0 6 ...