A physical fingerprint is something you are. How you play chess is something you do — a behavioural signature. This binds them: your game database becomes a glyph shaped like a fingerprint, and a short, timed game verifies it's really you. An intentionally overfit per-user template; the live game resists replay. (Second factor only — never a password.)
The ridge flow, gaps and two-tone data-zones are derived deterministically from 's games — same database, same glyph. The coloured minutiae are the style anchors (one per fingerprint dimension). Play new games and it re-forms.
MVP on a 12-game toy corpus — numbers are illustrative, not production. Real deployment needs many games/user, an ROC-tuned threshold, and liveness (a fresh timed opponent) to beat replay. But the mechanism is real: a person's move choices separate them from impostors, and the shorter/faster the game the more it must lean on opening style — where identity is loudest.
← Cognitive Fingerprints · the same style model, now as an authenticator.