I have been an avid baseball fan since I was four years old, so when I came across this recent study, I knew I had to share it.
A New Approach at the Plate…
A recent article published in the Journal of Quantitative Analysis in Sports introduces a new mathematical model to evaluate a batter’s plate discipline in Major League Baseball.
Traditionally, statistical analyses of a player’s at-bat encourage them to swing at any pitch in the strike zone.
However…
Ryan Lee, a statistics professor at the University of Wisconsin-Madison, led a new study that deviates from traditional statistical analyses of players’ batting discipline. Unlike previous models, this model takes into account three major factors: the history of the player and the umpire, pitch location, and game context. How cool!
The study shows that just because a pitch is in the strike zone does not always mean a batter should swing. For example, say three players are on base, and the pitcher throws the ball to the corner of the strike zone. Based on the batter’s history, he will likely hit a pop-fly, and the inning will end. Additionally, based on the umpire’s strike zone history, the ball won’t be called a strike despite being in the zone. Previous models would tell the batter to swing; however, this model, examining new factors, would advise the player not to swing. A game-changing way to analyze baseball!
It just gets better…
The study uses Bayesian nonparametric models and regression trees - an approach that takes into account multiple probabilities and updates them as a situation unfolds. The result? A complete analysis of the chance a pitch is called a strike, a batter swings, and if runs can score following each pitch.
Three steps are used in the model: probability estimation, combining those estimates, and then determining the optimal decision for the player.
All of this is used to answer the question: Will swinging increase the odds of scoring runs?
Another Win for Applied Math
A change in the way we calculate and use baseball analytics has potential to not only change how we teach mechanics to players but also how we view the game.
These new models have the ability to make a standardized strike zone (shown on TV) a thing of the past, and instead use statistics and math to develop a new strike zone relative to each individual player and umpire.
Math: the key to revolutionizing both sports strategy and how we watch games.