The primary purpose of this study was to determine whether tests performed at the National Hockey League (NHL) Combine could distinguish draft status (ie, the round selected). A secondary aim was to provide performance ranges and percentiles for each of the dependent variables.
A retrospective, cross-sectional study design was used with performance data and draft order from 2001, 2002, and 2003 Combine participants. Draft round was divided into 5 classifications (rounds 1, 2, 3, 4, and 5 through 9), and performances on 12 physical tests served as dependent variables. Three multiple analyses of covariance (MANCOVAs) were used to determine the significance of performance scores at the NHL Combine on draft selection. Age (years), body mass (kg), height (cm), and percentage body fat were treated as covariates.
Overall, MANCOVA results indicated no significant effect of performance on draft selection for 2001, 2002, or 2003. Subsequent univariate tests revealed that no single dependent variable was able to distinguish between draft rounds for any of the 3 years sampled.
Using draft status as an indicator of ice hockey performance, it appears that off-ice tests cannot accurately predict ice hockey playing ability in an elite group of athletes. This might stem from homogeneity of the Combine participants, a lack of validity of the tests, or other factors (eg, on-ice hockey skills, psychological variables, etc) that play a role in draft selection.
Vescovi and VanHeest are with the Human Performance Laboratory, Dept of Kinesiology, University of Connecticut, Storrs, CT 06269. Murray is with the Strength and Conditioning Dept, University of Louisville, Louisville, KY 40292. Fiala is with the Dept of Health, Physical Education and Human Performance, Salisbury State University, Salisbury, MD 21801.