School navigation

Mathematical Sciences

Explanation versus Prediction in Sports

Date: 3:30pm - 4:30pm PST December 4, 2013 Location: John Howard 132

John Howard 132

Explanation versus Prediction in Sports:
Some uses of Linear Regression and the Binomial Distribution in Major League Soccer and Baseball

Matthias Kullowatz, LC Alum ‘09
Lecturer in Statistics
Portland State University

The world of sports is littered with data. In fact, professional sports in the United States may produce more data than just about any other industry. The problem: How do we arrive at meaningful conclusions from small sample sizes in an environment where experimental design is a fantasy? In this talk, I strive to explain the important difference between using statistics to explain what happened, and using statistics to help predict outcomes in the future. Through examples from both baseball and soccer, I will demonstrate some, perhaps, unintuitive results and tie in some philosophy about what’s actually going on in the field of play.