Abstract by David Teuscher
MLB Offensive Efficiency
We predict offensive efficiency, measured by on-base plus slugging percentage (OPS), for Major League Baseball players over time. We estimate performance curves for each player using orthonormal quartic polynomials. The performance curves for the players were divided into clusters and the predictive performance of the models in each cluster was analyzed. Data from Lahman’s baseball database from 1871-2017 was used to build the model and then data from the 2018 season was used to test predictive capabilities. The predictive capabilities were tested by calculating a 95% predictive interval on the curve for players and determining whether or not the OPS for the 2018 season was included within the interval. These models provide insight into offensive efficiency of MLB players over the course of their careers and could be used when negotiating salaries and contracts.