CBB Engine Highlights
Player-Level Simulation
Once the season's EvanMiya player ratings are loaded, the player-level Savant Ultra engine is eligible to run the slate again: it models 5-on-5 player interactions using individual efficiency ratings, shooting profiles, and defensive matchup data. Each player's contribution is simulated rather than assumed from team averages.
EvanMiya BPR Integration
Bayesian Performance Ratings from EvanMiya provide the most accurate college basketball player-level metrics available. When loaded, these ratings power the player-level engine's efficiency estimates and matchup projections.
ATS Regime Analysis
Adaptive bucket filters analyze which spread ranges and game contexts are currently profitable. The system learns from resolved picks to gate recommendations through profitability regimes, avoiding historically unprofitable sub-niches.
March Madness Ready
Tournament-specific adjustments account for neutral-site play, single-elimination pressure, and matchup novelty. The engine handles conference tournaments and the NCAA Tournament with specialized calibration.