Abstract
Two new Bayesian methods for estimating and predicting in-game home team win probabilities in Division I NCAA men's college basketball are proposed. The first method has a prior that adjusts as a function of lead differential and time elapsed. The second is an adjusted version of the first, where the adjustment is a linear combination of the Bayesian estimator with a time-weighted pregame win probability. The proposed methods are compared to existing methods, showing the new methods are competitive with or outperform existing methods for both estimation and prediction. The utility is illustrated via an application to the 2012/2013 through the 2019/2020 NCAA Division I Men's Basketball seasons.
Original language | English (US) |
---|---|
Pages (from-to) | 201-213 |
Number of pages | 13 |
Journal | Journal of Quantitative Analysis in Sports |
Volume | 18 |
Issue number | 3 |
DOIs | |
State | Published - Sep 1 2022 |
Keywords
- Bayesian estimation
- dynamic prior
- in-game probability
- maximum likelihood
- pregame probability
- probability estimation
ASJC Scopus subject areas
- Social Sciences (miscellaneous)
- Decision Sciences (miscellaneous)