2020/03/24 by Marius Ötting, Ötting, Marius · 1 citation
Economics, Econometrics and Finance · Computer Science · #Sports Analytics and Performance #Time Series Analysis and Forecasting #Anomaly Detection Techniques and Applications
paper · pdf · doi:10.48550/arxiv.2003.10791
In recent years, data-driven approaches have become a popular tool in a\nvariety of sports to gain an advantage by, e.g., analysing potential strategies\nof opponents. Whereas the availability of play-by-play or player tracking data\nin sports such as basketball and baseball has led to an increase of sports\nanalytics studies, equivalent datasets for the National Football League (NFL)\nwere not freely available for a long time. In this contribution, we consider a\ncomprehensive play-by-play NFL dataset provided by www.kaggle.com, comprising\n289,191 observations in total, to predict play calls in the NFL using hidden\nMarkov models. The resulting out-of-sample prediction accuracy for the 2018 NFL\nseason is 71.5%, which is substantially higher compared to similar studies on\nplay call predictions in the NFL.\n