Estimating the Impact of Automated Umpiring in Baseball via Monte Carlo Simulation

Abstract

The MLB (Major League Baseball) has made multiple changes to the game of baseball recently to enhance the viewing experience for fans. One viable idea that has been tossed around for multiple years has been the implementation of an automated umpiring system. The MLB has the technology to utilize such a system using Trackman technology however most MLB teams have expressed opposition to the idea. Using an automated system would get rid of human mistakes that umpires make due to the high-speeds of MLB pitches and other challenges.

We present a method to estimate the impact of automated umpiring given MLB pitch data. We define a novel pipeline for simulating the statistical changes in MLB games following the correction of umpire mistakes. This pipeline uses historical game data to guide our estimations and then compares our findings to the baseline real game statistics. We finally use this pipeline to analyze the changes that an automated umping model would bring on average to the MLB game.