| Division | Number of Teams | Baseline Win % (Intercept) | Win % Gain per +1% HR Rate | R² (Variance Explained) |
|---|---|---|---|---|
| D3 | 391 | 35.1% | 11.9% | 22.8% |
| JUCO | 338 | 21.9% | 8.4% | 32.3% |
| D1 | 295 | 33.6% | 6.7% | 27.0% |
| D2 | 262 | 37.2% | 5.7% | 17.4% |
Statistical Modeling: Quantifying Power
Modeling the Impact of Home Runs on Win Percentage
While the Story tab visually highlights the positive trend between home run frequency and winning, statistical modeling allows us to answer a precise question: How much does a team’s expected win percentage increase for every 1% increase in home run rate?
We fit linear regression models across all four collegiate divisions: \[ \text{Win Percentage} = \beta_0 + \beta_1 \times (\text{HR Rate}) + \epsilon \]
Division-by-Division Model Estimates
Visualizing the Model Slopes
Key Takeaways from the Model
- Consistent Positive Relationship: In every collegiate division, the slope (\(\beta_1\)) is positive and statistically significant (\(p < 0.001\)).
- Division III Sensitivity: Division III teams have the steepest slope (\(\approx +11.9\%\) win rate per \(1\%\) increase in HR rate), meaning power hitting separates top teams from bottom teams even more dramatically at the D3 level.
- Explaining Season Outcomes: Home run rate alone accounts for 20% to 32% of the variance in season win percentage depending on the division, proving that long-ball power is a primary driver of collegiate softball success.