Been diving into NBA data lately and built a logistic regression model to predict game outcomes based on team stats like offensive rating, pace, and defensive efficiency. Backtested it on last season’s games and hit around 68% accuracy on point spread predictions. Anyone else messing with similar models? Curious about what variables you’re prioritizing or if you’re blending in player-level metrics like usage rate or PER.
Using Statistical Models to Predict NBA Game Outcomes
- Thread starter Κουκουβάγια
- Start date