This is a review of the baseball analytics research conducted by David Michael Vock and Laura Frances Boehm Vock, applying a potential outcomes framework and G-computation algorithm.
Baseball is an intricate sport, intertwining offensive and defensive strategies. Baseball offense depends on the pitches the batter faces, the batter’s choice to swing, and the batter’s hitting skill. Deciding whether or not to swing is one key component of plate discipline, which is the probability that a player will swing at pitches with different characteristics in different scenarios. Players with poor plate discipline swing at pitches regardless of their location or characteristics while a player with good plate discipline is more selective.
A potential outcomes framework is used to estimate the effect of various plate disciplines as this framework allows for precise assumptions and facilitates the models needed to estimate the causal effect of deciding to swing or not to swing at a pitch. A parametric G-computation algorithm is used to assess the causal effects.
For every pitch, two decisions must be made. First, the pitcher must decide what type of pitch to throw and second, the batter must decide whether or not to swing the bat. Analyzing potential outcomes is done by looking at what the result would have been if the pitcher and batter had made different choices. This is accomplished with a random dynamic treatment regime , which is a sequence of conditional probability distributions which provides the likelihood of making a certain decision by examining past covariate history and prior actions taken.
With the G-computation algorithm, the distribution of a long-term outcome can be modeled using different intervention strategies. This is accomplished by modeling the effect of the intervention and prior outcomes on the intermediate outcomes. A model is developed regarding the effect of swinging or not swinging, pitch characteristics and prior history of the at-bat on the outcomes of each pitch of a plate appearance.
The framework facilitates direct comparison of players who face different types of pitches and separates the impact of pitch selection, plate discipline, and hitting ability have on a player’s batting performance.
Analysts gain a clearer understanding of what effect changing the plate discipline of a batter will have regarding their batting performance. Notably, coaches can look at the ability of players when it comes to hitting pitches outside of the strike zone. Batters who demonstrate strong skills in this area should not be criticized for poor plate discipline. However, players without this skill will need to learn better plate discipline. Being able to differentiate on this point will provide coaches with better information on which to create coaching strategies for individual players. Analysts and coaches could look at the effect changing a batter’s plate discipline could have in specific scenarios. As batters alter their plate discipline scouts will be able to pass this information on to their team, allowing pitchers to alter their strategies in response.
Analysts could create standardized performance statistics for players across the league or team using a common pitch selection distribution, which would facilitate direct comparison between players. Analysis would also help identify how players could improve their outcome in plate appearances by improving the various aspects of their batting.
The circular effect between batter and pitcher adjusting as the other changes would bring about continual growing opportunities for the players, resulting in strong competition across teams.
Analytics Used: Potential Outcome Framework, G-Computation Algorithm, Random Dynamic Treatment Regime
————————
Find out how Sports Analytics Expert Victor Holman can give your team the competitive advantage.
How mature is your team’s analytics program? Take the Sports Analytics Maturity Assessment.
Discover the Groundbreaking Sports Analytics Application and Framework coaches and sports analysts are talking about!
Learn all about sports analytics in Victor Holman’s Sports Analytics Blog.