Skip to main content

Data from: Hit probability as a function of foul-ball accumulation

Cite this dataset

Howard, Jeffrey N. (2018). Data from: Hit probability as a function of foul-ball accumulation [Dataset]. Dryad.


In a day and age where professional baseball has implemented the use of countdown timers between innings, and restricted batter movement away from the batter’s box during at-bats–all in an effort to truncate drawn-out games—the foul ball at present, remains untouched. Despite some calls by fans for a rule change on foul balls so as to curb the amount of time they expend in a game, the foul ball has an interesting tale to tell when one actually looks at the data. Foul balls are ubiquitously portrayed as devalued events that merely slow down baseball games. However, despite its ‘do-over’ game-dragging reputation, the foul ball event can provide potential insight into the fatigue status of a pitcher, while also functioning as a predictor of hitter success. This paper investigates historical Retrosheet Major League Baseball (MLB) foul ball event data from 1945 through 2015 so as to analyze the historical nature of foul ball occurrence with respect to two states: [1] foul-ball accumulation when foul ball strikes occurred ‘inside-the-count’ (ITC); and [2] foul ball accumulation when foul balls only occurred ‘outside the count’ (OTC) and thus did not count as strikes. It is hypothesized that the foul ball as an event can predict hitting success, but that the probability of such success is dependent upon when foul balls occur with respect to the ball-strike count, and how many foul ball events accumulate during an at-bat.

Usage notes