PyMC Labs blog quantifies college football rankings with Bayesian uncertainty over 929 games
twiecki · x · 2026-10-28
PyMC Labs published a new post by Chris Fonnesbeck, "Rank and Uncertainty," arguing that typical rankings only give an ordered list. Instead, a Bayesian model in PyMC estimates how far apart two college football teams really are — and how confident we can be — validated on 929 games from the 2025 season, with posterior distributions showing overlap between teams like Notre Dame, Ohio State, Miami and Alabama.
More from Research
- Andrew Critch's dubious credences: P≠NP at 80%, Riemann hypothesis at just 60% — AndrewCritchPhD · 2026-10-28
- AI-driven Indian biotech PopVax, backed by $20M from Vitalik's Balvi fund, enters human trials with fridge-stable broad-spectrum Covid vaccine — chaitjo · 2026-10-28
- Cognition launches FrontierCode 2.0, a coding benchmark testing judgment beyond the spec — DevinAI · 2026-10-28
- 60% of arXiv text is now AI-written, with fully AI-authored papers at 25% — wandedob · 2026-10-28
- Do Neural Networks Stumble Into Intelligence by Chance? A Reddit Deep-Dive on Architectural Determinism — Lost-Quiet-2021 · 2026-10-28
- Princeton posts its AI alignment and safety seminar talk videos online — HazanPrinceton · 2026-10-28