Predicting play calls in the National Football League using hidden Markov models

IF 1.9 3区工程技术 Q3 MANAGEMENT

IMA Journal of Management Mathematics Pub Date : 2021-08-01 DOI:10.1093/imaman/dpab005

Marius Ötting

引用次数: 3

Abstract

In recent years, data-driven approaches have become a popular tool in a variety of sports to gain an advantage by, for example, analysing potential strategies of opponents. Whereas the availability of play-by-play or player tracking data in sports such as basketball and baseball has led to an increase of sports analytics studies, equivalent data sets for the National Football League (NFL) were not freely available for a long time. In this contribution, we consider a comprehensive play-by-play NFL dataset provided by www.kaggle.com, comprising 289,191 observations in total, to predict play calls in the NFL using hidden Markov models. The resulting out-of-sample prediction accuracy for the 2018 NFL season is 71.6%, which is similar compared to existing studies on play call predictions in the NFL. In practice, such predictions are helpful for NFL teams, especially for defense coordinators, to make adjustments in real time on the field.

查看原文本刊更多论文

用隐马尔可夫模型预测国家橄榄球联盟的比赛判罚

近年来，数据驱动的方法已成为各种体育运动中的一种流行工具，例如通过分析对手的潜在策略来获得优势。尽管篮球和棒球等体育项目中逐场比赛或球员跟踪数据的可用性导致了体育分析研究的增加，但美国国家橄榄球联盟（NFL）的等效数据集在很长一段时间内都无法免费获得。在这篇文章中，我们考虑了由www.kaggle.com提供的一个全面的逐场比赛NFL数据集，该数据集总共包括289191个观察结果，以使用隐马尔可夫模型预测NFL中的比赛呼叫。2018年NFL赛季的样本外预测准确率为71.6%，与NFL现有的比赛呼叫预测研究类似。在实践中，这样的预测有助于NFL球队，尤其是防守协调员，在球场上实时做出调整。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

IMA Journal of Management Mathematics OPERATIONS RESEARCH & MANAGEMENT SCIENCE-MATHEMATICS, INTERDISCIPLINARY APPLICATIONS

CiteScore

4.70

自引率

17.60%

发文量

审稿时长

>12 weeks

期刊介绍： The mission of this quarterly journal is to publish mathematical research of the highest quality, impact and relevance that can be directly utilised or have demonstrable potential to be employed by managers in profit, not-for-profit, third party and governmental/public organisations to improve their practices. Thus the research must be quantitative and of the highest quality if it is to be published in the journal. Furthermore, the outcome of the research must be ultimately useful for managers. The journal also publishes novel meta-analyses of the literature, reviews of the "state-of-the art" in a manner that provides new insight, and genuine applications of mathematics to real-world problems in the form of case studies. The journal welcomes papers dealing with topics in Operational Research and Management Science, Operations Management, Decision Sciences, Transportation Science, Marketing Science, Analytics, and Financial and Risk Modelling.