Back to Documentation
Model Documentation

NBA Models

Deep-dive into the methodology, edge detection, and operational details for our NBA prediction models.

Available Models

NBA Model Overview

Our NBA models are built on a foundation of comprehensive data collection, advanced statistical modeling, and rigorous backtesting. Key principles include:

  • 1.Per-100 possession normalization - We use pace-adjusted metrics to compare teams and players fairly across different game tempos.
  • 2.Player-specific variance - Each player has unique volatility patterns. We model individual variance rather than applying generic assumptions.
  • 3.Real-time adjustments - Injuries, rest, and matchup factors are incorporated dynamically before picks lock.
  • 4.Four-gate filtering - Every pick must pass EV threshold, projection confidence, odds floor, and breakeven buffer tests.