Bayesian Hierarchical Modeling of Field Goals in the NFL

• Activity Number: 578 - Bayesian Methodologies in Sports Statistics
• Type: Contributed
• Date/Time: Wednesday, July 31, 2019 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Sports
• Author(s): Sudipto Banerjee* and Jay Xu
• Companies: UCLA and University of California, Los Angeles
• Keywords: Bayesian Regression; Field Goals; Hierarchical Models; JAGS; NFL; Variable Selection

Abstract

Kicking field goals play an important role in the sport of American football. Memorable games in recent memory, such as this past season's LA Rams’ NFC championship victory over the New Orleans Saints or the Chicago Bears’ NFC Wild Card Game loss to the Philadelphia Eagles, have ended with the result of a field goal attempt determining the winner and loser of the game. On October 5, 2017, Mike Lopez (@StatsByLopez), Director of Data and Analytics for the NFL, tweeted that “A missing but important NFL study would look at the link between field/stadium and field goal success.” Indeed, the effect of the kicking surface and the football stadium on the probability of a successful field goal has not been thoroughly studied in previous research attempting to model the likelihood of a successful field goal. To this end, we propose a Bayesian hierarchical model to investigate this question and apply our model to real NFL data.