Joel Sokol is founding director of Georgia Tech’s interdisciplinary Master of Science in Analytics degree, and Fouts Family Associate Professor in the Stewart School of ISyE. His research focuses on developing models and methods that provide new insights and make previously-intractable problems solvable. Dr. Sokol’s research has won the EURO Management Science Strategic Innovation Prize, been successfully used by business and industry, and led to appearances on CNN and ESPN. As a teacher, Dr. Sokol has won recognition from the NAE, IIE, INFORMS, and EURO, served two terms as INFORMS VP of Education, and has been awarded Georgia Tech’s highest awards for teaching and student impact.
In addition, some sports just have more variability than others. For example, compare college football and college basketball. Both have pretty high variances between the betting line and the overall outcome, but football’s is higher despite overall scores being much lower. That means predicting the outcome of a college football game is harder than predicting the outcome of a college basketball game – for all the talk of upsets in “March Madness,” the college bowl season has a much higher fraction of upsets. We just don’t notice because most of the bowl games are pretty meaningless.
Some physiological components of sports are actually well studied. Examples include things like optimizing the precise running style of a sprinter, the hydrodynamic properties of a swimmer’s suit, and measuring and trying to control the heart rate of shooters in archery or a biathlon. Even in baseball, individual aspects, such as optimizing pitching motions and batters’ swings, have been worked on. But analytics tying those physiological models (e.g., “How do you optimize the torque a batter gets when swinging?”) to performance models (e.g., “What’s the potential increased value of this batter’s offensive contribution?”) haven’t been investigated that much.