Baseball is full of cutting-edge stats that try to predict pitcher performance. But one centuries-old gauge is just as telling: the barometer.
ESPN The Magazine recently featured the research of Dr. Rodney Paul, Falk College Professor of Sport Management, and the SU Baseball Statistics and Sabermetrics Club. In the article, reporter Peter Keating talks about Paul’s research and the fascinating discovery that the physics of air density doesn’t just affect the speed of pitching but also pitch selection. The research showed that pitchers alter their pitch selection based upon air density (altitude, temperature, humidity, barometric pressure), with pitchers throwing more breaking pitches when air density is low and fewer breaking pitches when air density is high. Furthermore, when comparing pitcher performance to expectations (based on betting market odds), pitchers that threw a high percentage of breaking pitches outperformed expectations on low air density days and pitchers who primarily throw fastballs outperformed expectations on high air density days.
Keating suggests that the results of the SPM club’s research may affect how teams play in the future based on weather and altitude.
Dr. Paul and SPM students Matt Filippi, Greg Ackerman, and Zack Albright were asked to present the research at the 2014 MIT Sloan Sports Analytics Conference in Boston. Their research paper, “The Effects of Atmospheric Conditions on Pitchers,” was one of 16 papers selected (8 for presentation and 8 for posters) from more than 300 submissions worldwide.
The paper originated from a detailed data set on baseball from an earlier study Paul collaborated on with co-authors Andrew Weinbach (Coastal Carolina University) and Jeff Gurney (University of South Carolina). The student researchers used that data set as a starting point, and the Falk College’s Baseball Statistics Club added data on individual pitchers and the pitches they throw for each game (PitchF/X data). From there, the group tested the impact of air density on pitch selection and performance compared to expectations, according to their research, and submitted it in a competitive process for publication at the MIT Sloan Sports Analytics conference.