NBA Player Props — Model Anchor Calibration Prompt

Context: Our MLB Player Props council (2026-04-03) established the universal framework: per-possession/per-PA rates as foundation, Bayesian shrinkage toward career prior, Log-Odds opponent adjustment, Gaussian copula for correlated combo props, dual-anchor system (sportsbook + model) tracked via Brier scores, confidence tiers. This prompt asks for the SPORT-SPECIFIC parameters to plug into that framework.

NBA Player Props — Model Anchor Calibration

We need fair probabilities for these NBA prop types: points, rebounds, assists, threes, blocks, steals, PRA combo, double-doubles.

Questions:

1. What are the NBA-specific stabilization constants (k values in minutes/possessions) for each prop type? How many games until points/rebounds/assists/threes/blocks/steals stabilize to 50% reliability?
2. Which opponent defensive stats matter most per prop? (opponent DvP for points, opponent rebound rate, etc.)
3. How should pace adjustment work? A player on a fast team (100+ possessions) vs slow matters hugely.
4. How should minutes projection work? Starters vs bench, blowout risk, back-to-back fatigue.
5. What distribution per prop? Points=Normal? Blocks/Steals=ZIP? Threes=NegBin?
6. How should usage rate and usage cascade (when a teammate is OUT) adjust projections?
7. What correlations exist between PTS/REB/AST for the PRA combo copula?
8. NBA-specific regime changes: rest days, trade deadline, playoff rotations, minute restrictions.
9. What EWMA alpha values per NBA prop type?
10. How should per-36-minute rates translate to per-game projections?
11. Give specific numbers, formulas, and implementation-ready recommendations.