Welcome to the Challenging Chess Engine Test!
CCET is a test suite of 160 chess studies where the task is to find a single best move.
It is designed to be difficult to solve on modern hardware in the year 2025 using state of the art chess engines.
You can download the test suite and view the current rankings.
Downloads: 🔽 pgn 🔽 epd 🔽 cbh 🔽cbv 🔽All formats
🏆Standardized Ranking
Goal: Comparison of engines under the same conditions
Requirements:
the machine used for the test has at least 8 physical cpu cores
8 Threads (used for calculation by the chess engine)
8 GB (8000MB) hash table size
180 secs per test position (classical tournament conditions)
no 'early exit' to speed up the test which can lead to less reliable results
NO endgame table bases (otherwise many of the positions are no test at all)
🏆Open Ranking
Goal: Maximum performance (hardware+software) and research
Requirements:
no 'early exit' to speed up the test which can lead to less reliable results
NO endgame table bases
If you have any feedback or questions please write to chesstestsuite@gmail.com .
Details about the test suite
Objective: 160 correct, difficult, versatile and at least to a large extent new test positions with unique solutions
160 * 3m = 8h ---> suitable for an overnight test
exactly one game theoretical best move (i.e. one move wins and other moves don't win or one move draws and other moves lose) in every position
extensively computer-checked solutions
very challenging: positions have been tested for difficulty using top engines like Stockfish, Obsidian, PlentyChess and others
Positions 1 to 46: best positions of previous test suites
Positions 47 to 80: best positions of published studies and problems (which have so far not been included in a test suite)
Positions 81 to 160: automatically generated and completely new endgame test positions; provable by endgame table bases; sorted by difficulty
more even chances for special solvers (like Crystal, Sting...) and general purpose engines (like Stockfish, Torch...)
Positions 1 to 80: (much) better chances for special solvers because of motifs like Zugzwang
Positions 81 to 160: better chances for general purpose engines (noticeable with large computing effort)
