Abstract

The spring 2019 edition of Stats 50 (cross-listed as MCS 100) will be bent towards data exploration. The overlying question of this course is how to leverage all the data available in sports to provide professional sports players with meaningful and useful insights on their own performance. We will most notably focus on the evaluation of individual and collective performance (covering regression to the mean or regularized adjusted plus-minus), adaptive in-game strategies (who should shoot the ball?), but also fairness in sports (how to make sure that all participants have equal chances?). A few guests speakers will shed light on more specific problems encountered in professional sport.

We will review linear regression early in the course and use the statistical programming language R to implement the ideas discussed. We recommend that you have taken an introductory statistics course prior to this one, but this is not mandatory. Exposure to linear algebra and basic probability theory is however strongly advised.


Goals of the course

By the end of the course, the hope is that you have learned how to:

  • understand and interpret advanced statistics reported in the media;
  • evaluate individual players in terms of how their performance leads to team success;
  • interpret players' statistics in small sample sizes and against varying opponent quality; and
  • use data to inform in-game strategic decisions in a variety of sports.