Introducing Discipline Score Based on League Overall Swinging Probability

Wuhuan Deng, Scott Nestler · 2025

Read Introducing Discipline Score Based on League Overall Swinging Probability by Wuhuan Deng, Scott Nestler (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-19672.txt.

Introduction

Wuhuan Deng and Scott Nestler introduce new baseball metrics for plate discipline, a hitter's skill at judging which pitches to attack. A disciplined hitter swings at good pitches and lays off balls outside the zone, but most existing patience measures do not capture the taking side of that skill well. Their Discipline Score, and an adjusted version of it, judge a batter only on pitches that are balls. Comparing the batter's choices with what the league as a whole would be predicted to do in the same situation.

Scoring against the league's overall swinging probability makes the measure relative: a hitter gains credit for restraint that peers would not have shown. The metrics aim to separate true pitch judgement from mere passivity, giving analysts and teams a cleaner read on a skill that drives walks, pitch counts, and long at-bats.

Catalog record

CREATORWuhuan Deng, Scott Nestler
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE18 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2511-19672

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Introducing Discipline Score Based on League Overall Swinging Probability Wuhuan Deng, Scott Nestler Abstract Plate discipline is an important feature of a hitter’s success. Hitters who are able to recognize good pitches to swing at and balls to take are generally recognized as disciplined hitters. Although there are some metrics that can provide insight into the patience of a hitter, most do not capture the ability of a batter to take balls. In this research, we introduce two new metrics, Discipline Score (DS) and Adjusted Discipline Score(ADS), which evaluate batters’ discipline when the pitch is a ball compared with the predicted tendencies of all batters in the league. K eywords Baseball, Clustering Method, Probability, Discipline 1 Introduction 1.1 Motivation Plate discipline is one of the key components of consistent offense. Generally, a hitter with good plate discipline is able to generate more walks and fewer strikeouts. Vock and Vock found that improved plate discipline increases batting average, on-base percentage, and slugging percentage [ 1 ] . When a batter gets a walk, it not only means that he is on base, but it also causes the pitcher to use at least 4 (of a limited number of) pitches. When we look at Major League Baseball (MLB) today, players who can generate many walks are all big stars in the league. Therefore, both the front office and the coaches should be able to recognize players who have great discipline, and this could help them make personnel decisions during free agency and trades. The 5 players who walked the most in 2024 are shown in Table 1 . Player Number of Walks Aaron Judge 133 Juan Soto 129 Kyle Schwarber 106 Shohei Ohtani 81 Ian Happ 80 Table 1: 5 players who generated most walks in 2024 season. On top of that, Run expectancy (RE) models describe how many runs a team is expected to score given the current game state of outs and baserunners. Recent work [ 8 ] represents RE as a Markov process, where each at-bat changes the game state through probabilistic transitions with runs as rewards. Since plate discipline affects this transition, it directly shapes the expected run output of a team. This is also why we need metrics to identify disciplined hitters. 1.2 Goal Our goal of this research is to evaluate the discipline of a batter compared to the entire league when the pitch is a ball. The probability of the entire league swinging on a given ball will be predicted, and then the Discipline Score (DS) of the batter on this ball will be determined based on the actual result of this pitch. Instead of representing how many walks a batter can generate in a season, this method provides a deeper insight when a batter faces a ball. And we can also know the discipline of batters when facing different types of pitches, such as fastball, breaking ball and offspeed, by adjusting the input pitch data. 2 Related Works Traditional metrics, such as O-Swing%, which evaluates how often a batter swings at pitches outside the strike zone and Z-Swing%, which measures how often a batter swings at pitches inside the strike zone have their limitations [ 3 ] . O-Swing% does not consider the differences between each pitch. If two batters face 10 pitches outside strike zone, and they both swing 3 times, they will get same O-Swing%, 0.3. However, the 10 pitches that they face can be quite different – one may swing at an outside pitch which is very close to the strike zone and another one may swing at an outside pitch which is far from the strike zone. But this difference is not reflected by O-Swing%. In Figure 1 and 2 , although both pitches are balls and the batters have swung, it is obvious that those two pitches are different. Javier Baez obviously swung on a ball which should be identified as such by a big league player. However, Juan Soto swung at a ball that was not too far from the edge of the strike zone. Figure 1: Juan Soto swung on a ball. Figure 2: Javier Baez swung on a ball. In 2023, Ryan Yee and Sameer K. Deshpande introduced a three-step framework to evaluate plate discipline. In their first step, they fitted a flexible Bayesian nonparametric model to predict (i) the probability that a batter makes contact; (ii) the probability that the umpire calls a strike; and (iii) the average number of runs that the batting team is expected to score [ 2 ] . There are two more steps in their research, but only the first step is directly related to this study. When they predicted the contact probability, the model only considers pitches at which batters swung. And they evaluated plate discipline based on the result of contact and expected runs. They also predicted the probability that the umpire calls a strike, but in this research we used the exact definition of a ball based on pitch location data. MLB’s EAGLE (Expected Aggressiveness, Location, and Expectancy) model estimates the run-value impact of a hitter’s swing decision by comparing actual swing choices to league-average expected values given pitch location and count [ 4 ] . 3 Define Discipline Score & Adjusted Discipline Score 3.1 Discipline Score Discipline Score (DS) evaluates a batter’s discipline based on the overall tendency of players in the league to swing at a given ball. For a given ball, DS is defined as: DS=(-1)^{R}\cdot(R(1-P_{S})+(1-R)P_{S}), where R is the actual decision of the batter ( R=1 for swing or R=0 for take), and P_{S} is the probability that the entire league will swing at the given ball. DS ranges from -1 to 1 with larger values indicating better discipline. (-1)^{R} determines the sign of DS, where a negative value suggests that the batter swings at the ball. and a positive value means that the batter takes the ball. Table 2 shows several examples of possible DS values. P_{S} R DS Interpretation 0.7 0 0.7 Take a ball that most batters swing at 0.7 1 -0.3 Swing at a ball that most batters swing at 0.3 0 0.3 Take a ball that most batters take 0.3 1 -0.7 Swing at a ball that most batters take Table 2: Example values of DS. 3.2 Adjusted Discipline Score DS only considers the action of swing or take itself, but ignores the result if the batter swings at the ball. Sometimes, a batter can generate a home run even swinging at a ball. Therefore, we introduced a new metric Adjusted Discipline Score (ADS), which for a given pitch is defined as: ADS=DS+CQ, where CQ is the contact quality which is defined as: CQ=\begin{cases}EV_{score}\times LA_{score},\text{ if batter makes contact and not foul}\\ 0,\text{ otherwise (including swing and miss)}\end{cases} where EV_{score}=min(1,max(0,\frac{EV-70}{28})) and LA_{score}=max(0,1-\frac{|LA-20|}{20}) . EV is the exit velocity and LA is the launch angle of the given pitch. By definition, 98mph is the smallest EV for a barrel [ 9 ] , so when EV is larger than 98mph, EV_{score} will be 1. And when EV is between 70mph and 98mpb, EV_{score} will be between 0 and 1. When EV is smaller than 70, which will be considered

Read the complete text