Loading…
AI Bets
Each week of the NFL season, our model prices a set of player prop markets and publishes the handful where it most disagrees with the market's number — timestamped before kickoff, graded against the closing line, losses kept. Whether the model actually has an edge is an open question one season cannot answer, and this page says so rather than pretending otherwise.
Season underway. No card has been published yet — when one is, it lands here before kickoff and is graded in public.
28 line snapshots archived since Aug 26 — the record starts before the first pick.
Latest capture Oct 2, 2026 · 1,236 book prices · every snapshot is kept, and the archive is what future picks are graded against.
At roughly five picks a week over an 18-week season — about 90 bets — a 55% hit rate carries a 95% confidence interval of roughly 45–65%. The statistical power to distinguish a genuinely excellent 55% model from a break-even one at that sample is about 12%, and a confident win-rate answer needs about 2,250 bets — roughly 25 seasons. One season cannot answer "is the model smarter than the market", and this page will never claim it has.
What a season can show is whether the stated probabilities were calibrated and whether the published numbers beat the close — the two measures that converge fast enough to mean something. The full math is published before the first pick: why one season proves nothing.
BetFantasy.ai is not a sportsbook. We accept no wagers and hold no money. Everything published here is informational analysis and opinion — not financial advice, and not a prediction of any outcome. These are model outputs, the model is often wrong, and the complete record of that is public. Sports betting carries real risk; only ever stake what you can afford to lose. 21+ and only where legal. If gambling stops being fun, call or text the National Problem Gambling Helpline at 1-800-MY-RESET (1-800-697-3738).