| Characteristic | Beta | 95% CI1 |
|---|---|---|
| spread | 0.38 | 0.32, 0.44 |
| 1 CI = Credible Interval | ||
Models
Model Definition
A generalized Bayesian linear model using the stan_glm() function was used, where the dependent variable is the margin of victory and the independent variable is the point spread (absolute). This model was the byproduct of 8,206 observations.
\[y_i = \beta_0 + \beta_1 x_{1,i} + \epsilon_i\]
with \(y = victory\_margin\), \(x_1 = spread\), and \(\epsilon_i \sim N(0, \sigma^2)\).