As the use of analytics grows in the sports industry, debates about the usefulness of analytical models in sports has also grown. There is no doubt that analytics have impacted the sports industry in many positive ways, but it is an evolving story as analysts seek better models of player/team performance evaluation, forecasting, and decision-making. Communicating new results in these areas requires analysts to connect with organizations and fans by putting the results in context to tell a more complete story. In this work, we give examples from our own work and the work of others showing how to frame analytics within a story. At the same time, we give a brief history of the evolution in the descriptive, predictive, and prescriptive areas of sports analytics. While this work is not meant to be exhaustive, it highlights some of the major issues that analysts face in building useful models in these areas. The paper also represents a decade-long collaboration between academics and sports writers, and we highlight some of the lessons we’ve learned from that collaboration.
Authors: Elizabeth L. Bouzarth, Benjamin C. Grannan, John M. Harris, Kevin Hutson, Peter J. Keating