| Previous studies with college football players have determined a difference in starters versus non-starters when comparing the results of selected combine tests. Data from 636 National Football League players were analyzed. The data were compared with the playing performance status of the player whether a Pro Bowl, starter, or non-starter. Two ANOVAs were performed on these three separate player groups, first for the total data set using means based on the total sample to calculate the total combine score and the second using means calculated among the specific position groups to calculate the total combine score. The combine tests used in this study were the 40-, 20-, 10-yard dashes, 225 bench repetition test, Pro Agility, vertical jump, broad jump, Wonderlic Personnel Test, and height and weight. Additionally, correlation and regression analysis were performed when comparing Pro Bowl playing status with the combine tests. Results showed neither ANOVA was significant for the total combine score analysis. Correlation only produced four significant relationships among the position groups and five significant relationships for the total data set. Regression showed only two significant models for the position sets (offensive backs, linebackers); results for the total data set were significant as well.;Although these results are compelling, problems with multicollinearity and Pro Bowl player sample size should temper any conclusions from this study. However, despite these problems, this study's results raise validity questions about the combine and care should be taken by player evaluators in placing too much emphasis on the test battery. Much is still to be determined about the combine's true effect in predicting player success; this should be the topic of future research. |
