Season 1 Evaluation in Progress
Leaderboard Coming Soon
Standardized swap-opening evaluations across leading frontier foundation models and specialized neural game engines are underway. Official Season 1 standings will be published upon completion.
Track 1: Rule Adherence
Measuring illegal move rate and spatial hallucination in the 90-intersection discrete board grid.
Track 2: Tactical Endgames
Checkmate detection, fork defense, and counter-tactics accuracy on standardized tactical benchmark sets.
Track 3: Live Elo Arena
Swap-opening round-robin tournaments fitted via Bradley-Terry maximum likelihood estimation.