Search papers, labs, and topics across Lattice.
This paper introduces Ailed, a chess engine that simulates human-like behavioral variability by modulating move probability distributions based on a dynamic "psyche" score. The psyche score, ranging from -100 to +100, is recomputed after each move based on positional factors and fed into an audio-inspired signal chain that alters the move probabilities generated by an underlying chess engine. Experiments against Maia2-1100 demonstrate that Ailed exhibits a 20-25 percentage point spread in top-move agreement between stressed and overconfident states, and a significant drop in competitive score under stress, suggesting that the signal chain effectively induces human-like tilt and overconfidence.
Forget perfect play: Ailed makes chess engines blunder like humans under pressure by dynamically distorting move probabilities with an audio-inspired "psyche" model.
Chess engines passed human strength years ago, but they still don't play like humans. A grandmaster under clock pressure blunders in ways a club player on a hot streak never would. Conventional engines capture none of this. This paper proposes a personality x psyche decomposition to produce behavioral variability in chess play, drawing on patterns observed in human games. Personality is static -- a preset that pins down the engine's character. Psyche is dynamic -- a bounded scalar \psi_t \in [-100, +100], recomputed from five positional factors after every move. These two components feed into an audio-inspired signal chain (noise gate, compressor/expander, five-band equalizer, saturation limiter) that reshapes move probability distributions on the fly. The chain doesn't care what engine sits behind it: any system that outputs move probabilities will do. It needs no search and carries no state beyond \psi_t. I test the framework across 12,414 games against Maia2-1100, feeding it two probability sources that differ by ~2,800x in training data. Both show the same monotonic gradient in top-move agreement (~20-25 pp spread from stress to overconfidence), which tells us the behavioral variation comes from the signal chain, not from the model underneath. When the psyche runs overconfident, the chain mostly gets out of the way (66% agreement with vanilla Maia2). Under stress, the competitive score falls from 50.8% to 30.1%. The patterns are reminiscent of tilt and overconfidence as described in human play, but I should be upfront: this study includes no human-subject validation.
