Data envelopment analysis, a linear programming technique, is used to estimate an objective measure of coaching efficiency. This method is applied using the performance records of 147 Division I college basketball teams from 1984 through 1991. The results indicate a wide dispersion in coaching efficiency, with the typical coach being very inefficient relative to the best coaches in the sample. It is also shown that additional experience apparently has no effect on coaching ability. Finally, coaching efficiency estimates provide a much different assessment of coaching performance than does a coach's winning percentage. These results imply that if coaching retention is based on winning percentage rather than coaching efficiency, excellent coaches will be ignored in hiring decisions and inappropriately dismissed in firing decisions if coaching retention is based on winning percentages rather than coaching efficiency.
John Fizel is with the School of Business at the Pennsylvania State University-Erie, Erie, PA 16563. Michael D'itri also is with the Pennsylvania State University-Erie.