Toughness Predicts Performance In College Football

A premium college football player is estimated to generate over $1 million for his program, so optimal player assessment and selection are paramount. Lean body mass, back squat, and vertical jump have been the most predictive physical test scores, but such metrics typically account for less than 10% of variance in football performance. NFL scouts have tended to rely on vertical jump, 40-yard dash, and 20-yard shuttle scores, but interestingly, reliance on none of the physical tests conducted at the NFL Combine was predictive of team success; in fact, teams that relied on fewer total physical test scores tended to win more games. Dr. Tom Osborne suggested the Performance Index for interpreting physical performance metrics, but also insisted that psychological attributes such as toughness were equally important to his evaluation process since players deal with immense adversity during the course of a season. Toughness can be characterized as the ability to cope with stressors more effectively, and is measured in this study by examining cortisol reactivity. When physical test scores (i.e., lean body mass, hang clean, back squat, bench press, 10-yard dash, 20-yard shuttle, and vertical jump) were converted into Performance Index scores among 47 Division 1 freshman football players, players who contributed on the field differed from players who did not contribute only in vertical jump and 10-yard dash scores. A multiple regression model that included only physical test scores accounted for 28% of variance in football performance, whereas a model that included cortisol reactivity measured during a physical testing session, in addition to traditional physical performance prediction metrics, accounted for 39% of variance. Results from this investigation suggest that cortisol reactivity may capture a distinctive attribute of players and predict on-field performance better than many individual physical performance variables. If cortisol reactivity, as a measure of toughness, can be effectively used to predict performance on the football field, then the metric should be included in performance prediction models that have traditionally included only physical attributes of players. Advisor: Daniel W. Lege