The application of sabermetrics to the teaching and coaching of collegiate baseball.

Abstract

This study was designed to improve the teaching and coaching techniques presently used by many in leadership positions in the realm of collegiate baseball. By applying the principles of sabermetrics, the in-depth analysis of baseball statistics, coaches and teachers of the sport would benefit in many key situations involving the sport: choosing personnel for the best potential lineup, selecting appropriate game strategies, gaining statistical support for time-honored theories while refuting others, and boosting bench morale. The author collected data contained in individual scoresheets used in collegiate games played by Union University, a National Collegiate Athletic Association (NAIA) member school, in the years from 1983 to 1985. Sabermetric principles were applied to the statistical information gathered from these games to determine which numerical indicators are most useful and reliable when dissecting the college game. Statistics in a multitude of areas were collected, as indicated in a seasonal sample illustrated in Appendix B of this paper.

The writer applied theories and formulas pioneered by founding sabermetricians Bill James and Pete Palmer, among others, including "Runs Created," "Total Average," on-base percentages, run-scoring probabilities, and several other categories of measurement. Among the major discoveries of the study were: (1) more times than not, using a bunt in an inning reduces the chances of scoring, despite the belief of most coaches that it helps the chances; (2) many player personnel selections would be made more easily if a coach used more pertinent information than is currently employed, such as substituting on-base percentages for batting averages; (3) the stolen base attempt usually will hinder an offense as much as it helps it; (4) scoring the first run in a game results in a victory well over 65 percent of the time; and (5) the concentration of pitchers can be greatly enhanced by keeping statistics of their performances after two men are out in an inning. The study also includes run-scoring probabilities in all possible ball/strike combinations and baserunning situations. Many of the statistics used were compiled from major league baseball games, but the author has found them relevant at the collegiate level. (Abstract shortened with permission of author.).