Prediction Models of Speed and Agility in NFL Combine Attendees

PURPOSE: To develop prediction models of sprinting speed and agility in NFL football players using vertical jump (VJ), broad jump (BJ), body mass (BM), and height (HT). METHODS: Secondary data analyses were performed on data from the 2003 and 2008 NFL combines. Data were extracted from a publically accessible website used previously for publication. A total of 446 athletes completed VJ, BJ, BM, and HT measurements in the 2003 (n = 203) and 2008 (n = 243) combines and were used in this analyses. Scores for VJ, BJ, BM, and HT were correlated to 10-yard sprint time (10sp), 20-yard sprint time (20sp), 40-yard sprint time (40sp), 20 yard shuttle time [i.e. pro-agility time (20ag)], and the 3-cone agility drill. Briefly, 10sp, 20sp, and 40sp are measured in a single test using laser timing equipment. 20ag and 3-cone drill are also measured using laser timing equipment. Vertical jump was measured using a Vertec, broad jump with a tape measure, HT using a stadiometer, and BM using a calibrated scale. Pearson's correlations (r) were calculated to determine univaritate relationships between each performance test (10sp, 20sp, 40sp, 20ag, and 60ag) and predictor variables (HT, BM, VJ, and BJ). Combined regression models were derived using backwards elimination. Alpha was set at p < 0.05 for all comparisons. RESULTS: Combined models are listed in Table 1. 10sp was significantly correlated with HT (r = 0.60), BM (r = 0.75), VJ (r = −0.56), and BJ (r = −0.73); however, HT and VJ did not contribute significantly to the final model (p > 0.05). 20sp was also correlated with HT (r = 0.60), BM (r = 0.75), VJ (r = −0.56), and BJ (r = −0.73). VJ did not contribute significantly to the combined equation (p > 0.05). BM (r = 0.65), VJ (r = −0.65), and BJ (r = −0.80) were significantly correlated to 40sp. BM (r = 0.75), VJ (r = −0.62), and BJ (r = −0.69) were correlated with 20ag. BM (r = 0.65), VJ (r = −0.56), and BJ (r = −0.63) were also correlated with 3-cone time. HT was not significantly correlated with 40sp, 20ag, or 3-cone times and did not contribute to the final model equation for these performance tests (p > 0.05). Additionally, VJ was eliminated from the final model for 3-cone drill time (p > 0.05). CONCLUSIONS: BM and BJ appear to be the best predictors of speed and agility in NFL football players. Interestingly, when BJ was included in the model, VJ was eliminated as a significant predictor of three of the five tests. PRACTICAL APPLICATIONS: Exercise regimens that decrease body mass and improve horizontal and vertical ground reaction forces may be beneficial for improving speed in NFL football players. Research is needed to determine level of transfer of increases in BJ scores and decreases in BM to improvements in interval speed and agility scores.