Investigating the coefficient of restitution of major league baseballs

Abstract

In the middle of the 2015 Major League Baseball season there was a sudden increase in home run rate. Such a sudden league-wide change is difficult to explain, and many baseball pundits suggested that a "juiced" or altered baseball was the culprit. In this thesis, large scale coefficient of restitution (COR) testing is performed on baseballs from before (2014) and after (2016) the alleged change. No evidence is found to suggest that the baseball was altered between 2014 and 2016. The investigation nevertheless revealed telling properties of the coefficient of restitution. First, it was largely found that baseballs from both years fell within the COR specifications of Major League Baseball (between 0.514 and 0.578). Second, the COR is revealed to be a noisy quantity in that the same baseball can exhibit differences in COR on the order of 0.01. Third, most of the variation between baseballs' COR comes from being manufactured on different days in the plant and thus being packaged in different boxes. Additionally, an unexpected fatigue in the strike plate used in the test apparatus is found to have affected the COR measurements for the first 200 trials.