Crash gambling: anatomy of an online gambling category — definition, documented mechanisms and a one-million-round measurement

The defining study of the crash-game category: a three-condition definition (a rising multiplier; a crash point drawn at random and fixed before the round; a single player decision — cash out before the stop), the crash-point determination mechanisms documented by the developers themselves (Bustabit's pre-generated SHA-256 hash chain salted with a future Bitcoin block; Spribe Aviator's server-seed + player-seed provably fair system; certified generators and studio seed systems at InOut Games and SmartSoft), and a one-million-round measurement of the documented distribution P(bust >= x) = RTP / x at the publisher-stated RTP of 98% (Scott, Sher & Daigle, UNLV Gaming Research & Review Journal, 2026). Headline results, recomputed independently on 29 October 2026 from the deposited 1,000,000-round corpus of Study 01 (DOI 10.5281/zenodo.23597144): 3.00% of rounds stop at 1.00x; median crash point 1.95x; 48.80% of rounds reach 2.00x and 0.97% reach 100.00x, tracking the theoretical 0.98 / x law within 0.2 percentage points at all thirteen measured thresholds; at fixed cash-out targets from 1.01x to 10.00x the realised RTP stays between 97.58% and 97.97% — the cash-out target chooses the variance of play, not its expectation. The deposit carries the technical report (PDF) and this study's derived dataset (CSV: bust distribution, survival measured against theory, outcomes by fixed cash-out target). The figures describe the documented model, not rounds measured at an operator. No purchase, real-money play or unpublished operator data enters the study. The study's article home: https://chickenroadproductions.com/anatomie-du-jeu-crash/ (French) and https://chickenroadproductions.com/en/crash-gambling-anatomy/ (English).