Abstract
Slugging percentage has been a measure of a player's offensive production in baseball for generations. The traditional weighting of hits (one for single, two for double, three for triple, four for home runs) was a natural and reasonable way to define this statistic when it was introduced. However these weights do not fairly reflect the actual value of each type of hit. By studying major league baseball games using Markov chains and expected run probabilities, we derive weights that are more reflective of the actual value of each type of hit and call the resulting measure wSLG, for weighted slugging. We examine the relative weights under two different models for how baserunners advance on hits, and we suggest ways that teams could 'personalize' this metric to their particular team tendencies.
Keywords: Slugging percentage, Markov Chains in baseball, Expected run probabilities, Batting performance metrics, Weighted linear combinations
How to Cite:
Cleary, R., Miller, S., Cardonick, A., Carso, M., Eckerle, J., Prout, E. & Ruschil, E., (2025) “wSLG: Redefining the Relative Values of Various Hits in Baseball”, Maths and Sports 7(1). doi: https://doi.org/10.5149/ms.1266
Downloads:
Download cleary_25
View PDF
49 Views
6 Downloads