NBA Game Desk — 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 Game Desk — Model Anchor Calibration

We are building model anchors for NBA game-level betting markets on Kalshi: moneylines, point spreads, totals.

Questions:

1. What team metrics best predict point differential? (Net rating, SRS, Elo, four factors — eFG%, TOV%, ORB%, FTR?)
2. How should pace interact with the total? Two fast teams = higher total?
3. How many games until team offensive/defensive ratings stabilize? k values?
4. How should rest advantage (back-to-back, 3-in-4, rest days) adjust the spread?
5. Home court advantage — magnitude and how it varies by team/arena?
6. Injury impact: how should missing players (especially stars) adjust team rating? RAPTOR/EPM-based replacement?
7. How to model totals: team projected points = (team off rating x opp def rating / league avg) x pace?
8. Travel and schedule density effects?
9. What distributions? Spread=Normal? Total=Normal?
10. How should playoff vs regular season be handled differently?
11. EWMA alpha for team ratings?
12. Give specific numbers, formulas, and implementation-ready recommendations.