Game 5 Prediction · NBA Finals 2026
The Bayesian model identifies two pace-neutral factors critical to San Antonio's survival: effective field goal percentage (eFG%) and turnover management.
The Reality
San Antonio is 3-1 down and must win Game 5 at home. The model puts their series comeback probability at 5.3% based on current Finals performance. But if they win G5, odds improve to 9.6% to 18.4%.
The MCMC sampler ran 4 parallel chains. The Gelman-Rubin statistic (Rhat) measures convergence. All parameters show Rhat < 1.05 — the posterior estimates are reliable.
Green bars help, red bars hurt. eFG% is overwhelmingly dominant (β=+1.69, 100% confidence). Turnover % is the second-strongest signal but negative — keeping turnovers low is critical.
These bars show the probability of winning the entire series (need 3 straight: G5H, G6A, G7H) under three performance scenarios. Red = Finals-level. Gray = improvement. Blue-gray = playoff-average.
1. Effective FG% (eFG%) — Critical
Finals avg (G1-G4)
53.4%
Playoff win avg
57.1%
eFG% = (FG + 0.5×3P) / FGA. The single strongest predictor of wins.
2. Turnover % (TOV%) — Critical
Finals avg (G1-G4)
14.7%
Playoff win avg
12.1%
TOV% = TOV / (FGA + 0.44×FTA + TOV). Reducing turnovers is essential.
Game 5 Implied Odds
+260
Based on 27.9% win probability (Finals avg).
If Spurs Win G5 (get to 3-2)
+420
Series odds improve to 9.6% to 18.4%.
Data: 25 games (4 Finals, 3 pre-Finals NYK, 18 other playoffs)
Features: eFG%, TOV%, ORB%, FT Rate, Home/Away — all pace-neutral.
Sampler: 4 parallel MCMC chains. All parameters converged (Rhat < 1.05).
Key insight: NYK Finals base rate is 33.6% vs 68.7% playoff average. Beating the Knicks is categorically harder.