Featured
Live
Seats limited
Nov 5–6 from 2:00pm ET - 4:30pm ET each day · Hierarchical Models in PyMC
Walk away with a working multi-level model, posterior checks, and a reusable sports notebook.
Schedule: 2 live sessions (~5h total), times TBD
Format: Short · Live · Hands-on · Discord Q&A · Cloud lab included · Recordings
Level: Comfortable with Python + basic Bayes; PyMC newbies welcome
Early bird through Oct 17: $259
Standard: $299
Team pack (3 seats): $750
Pilot pricing. Alumni receive a discount toward Alex’s 6-week flagship.
Syllabus at a glance
What you'll build
- Partial pooling models with domain-informed priors
- Posterior checks & diagnostics coaches understand
- Model comparison frameworks (LOO, PPC) & visuals
- Reusable, annotated sports notebook (PyMC + ArviZ)
Session 1 — Bayesian Models Fundamentals (2.5h)
- MLB case: pitcher performance variability
- Overdispersion with Beta-Binomial
- Fixed-talent vs. game-to-game variation
- Predictive performance: LOO & posterior predictive checks
- Sequential updates & ranking under uncertainty
Session 2 — Hierarchical Models Fundamentals (2.5h)
- Pooled vs. unpooled vs. hierarchical
- Partial pooling & regularization to the mean
- Build in Bambi & PyMC; centered vs. non-centered
- Posterior analysis & ArviZ visualization
Prereqs & materials
- Python + basic Bayes (PyMC/Bambi experience helpful, not required)
- Provided: MLB datasets, notebooks, exercise solutions, templates