This paper aims to develop an interpretable machine learning model to predict plays (pass versus rush) in the National Football League that will be useful for players and coaches in real time. Using data from the 2013-2014 to 2016-2017 NFL regular seasons, which included 1034 games and 130,344 pass/rush plays, we first develop and compare several machine learning models to determine the maximum possible prediction accuracy. The best performing model, a neural network, achieves a prediction accuracy of 75.3%, which is competitive with the state-of-the-art methods applied to other datasets. Then, we search over a family of simple decision tree models to identify one that captures 86% of the prediction accuracy of the neural network yet can be easily memorized and implemented in an actual game. We extend the analysis to building decision tree models tailored for each of the 32 NFL teams, obtaining accuracies ranging from 64.7% to 82.5%. Overall, our decision tree models can be a useful tool for coaches and players to improve their chances of stopping an offensive play.

This article applies a method we term "predictive clustering" to cluster neighborhoods. Much of the literature in this direction is based on groupings built using intrinsic characteristics of each observation. Our approach departs from this framework by delineating clusters based on how the neighborhood's features respond to a particular outcome of interest (e.g., income change). To do so, we leverage a classification and regression via integer optimization (CRIO) method that groups neighborhoods according to their predictive characteristics and consistently outperforms traditional clustering methods along several metrics. The CRIO methodology contributes a novel methodological and conceptual capability to the literature on neighborhood dynamics that can provide useful insights for policymaking.