Tire changes, fresh air, and yellow flags: Challenges in predictive analytics for professional racing

Journal Article (Journal Article)

Our goal is to design a prediction and decision system for real-time use during a professional car race. In designing a knowledge discovery process for racing, we faced several challenges that were overcome only when domain knowledge of racing was carefully infused within statistical modeling techniques. In this article, we describe how we leveraged expert knowledge of the domain to produce a real-time decision system for tire changes within a race. Our forecasts have the potential to impact how racing teams can optimize strategy by making tire-change decisions to benefit their rank position. Our work significantly expands previous research on sports analytics, as it is the only work on analytical methods for within-race prediction and decision making for professional car racing.

Full Text

Duke Authors

Cited Authors

  • Tulabandhula, T; Rudin, C

Published Date

  • June 1, 2014

Published In

Volume / Issue

  • 2 / 2

Start / End Page

  • 97 - 112

Electronic International Standard Serial Number (EISSN)

  • 2167-647X

International Standard Serial Number (ISSN)

  • 2167-6461

Digital Object Identifier (DOI)

  • 10.1089/big.2014.0018

Citation Source

  • Scopus