Datasets › Maia KDD Testing Set

Maia KDD Testing Set

Introduced by Reid McIlroy-Young et al. in Learning Models of Individual Behavior in Chess23 Aug 2020 archive 2025-07-28

These are the games used for testing models in the paper Aligning Superhuman AI with Human Behavior Chess as a Model System. The games have been converted from pgn to csv and the per move and per game information has been extracted. If you wish to see the original game files download the December 2019 standard games database release from Lichess. The game_id column matches that used by lichess. The games were selected to have both players of similar Elo Rating and for each divisions of 100 from 1000 to 25000, i.e 1000, 1100, ... 2500 there are 10,000 games. The meaning of each column should be straightforward. The per move information tends to be from the perspective of the active player, although the cp is not, instead the cp_rel columns is the relative centipawn value.

Benchmarks archive 2025-07-28

No leaderboard in the archive resolves to this dataset.

Papers archive 2025-07-28

No paper in the archive has a leaderboard row on this dataset; the archive counts 1 paper for it but never published that list.

Dataset loaders archive 2025-07-28

No loader listed in the archive.

Tasks archive 2025-07-28

No task tagged in the archive.

License archive 2025-07-28

No licence recorded in the archive. Absence here is not a statement about the dataset's terms.

Modalities archive 2025-07-28

No modality tagged.

Languages archive 2025-07-28

No language tagged.

Variants archive 2025-07-28

  • Maia KDD Testing Set

1 variant name, as the archive lists them.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections