Introduction

The question is whether score margin differentials in basketball are highly explainable, moderately explainable, or difficult to explain. Based on the regression output, I would classify them as highly explainable at the team level.

The models do not use score margin directly. They use Offensive Rating and Defensive Rating, which are the two main pieces behind margin. A team’s margin is mostly built from how efficiently it scores and how efficiently it prevents the other team from scoring.

Model Summary

Regression Model Summary
Model Outcome Factors R Squared Percent Explained
Offense Offensive Rating eFG%, TOV%, ORB%, FT/FGA 0.972 97.2%
Defense Defensive Rating eFG%, TOV%, DRB%, FT/FGA 0.994 99.4%

The offensive model had an R-squared of 0.972, meaning the four offensive factors explained about 97.2% of the variation in Offensive Rating.

The defensive model had an R-squared of 0.994, meaning the defensive four factors explained about 99.4% of the variation in Defensive Rating.

R-Squared Comparison

R-Squared Comparison
Side Regression R Squared Percent Explained Interpretation
Offense ORtg on offensive four factors 0.972 97.2% Very strong offensive fit
Defense DRtg on defensive four factors 0.994 99.4% Very strong defensive fit

These numbers are too strong to call moderate. The four factors explained almost all of the variation in the offensive and defensive efficiency numbers that drive scoring margin.

Interpreting Scoring Margin

I would not say this proves every individual game margin is easy to predict. Basketball still has randomness. A team can get hot, miss free throws, turn it over late, or have one bad stretch that changes the final score.

Over a full season, though, that noise starts to even out. The R-squared values show that team efficiency is strongly tied to repeatable basketball factors like shooting, turnovers, rebounding, and free throws.

Based on these results, I would classify basketball scoring margin as highly explainable at the team-season level. The model does not directly regress scoring margin, but it explains the main ingredients behind it extremely well.