Introduction
Physical test batteries are routinely used at both the amateur and professional levels of sport to evaluate performance characteristics for the purpose of monitoring training adaptation and player selection. Commonly, test batteries are made up of measures evaluating various attributes. As part of its evaluation of future prospects, the National Football League (NFL) implements such a test battery at its annual combine. Specifically, the combine battery is made up of 9.1-, 18.3-, and 36.6-m sprint times; vertical and horizontal jump measures; the 18.3-m shuttle and 3-cone drill, designed to measure change-of-direction ability, and the 102.1-kg bench press as a measure of upper body strength. The participants invited to the combine are predominantly seniors from National Collegiate Athletic Association (NCAA) colleges and make up approximately 3% of all football players representing the NCAA Division I teams (6).
A considerable amount of research has been conducted to examine the relationships between physical attributes such as jumping, sprinting, and change-of-direction ability (3,8,9). Evidence exists indicating significant correlational relationships between sprint ability at distances ranging from 5 to 36.6 m (3,11) and also between sprint and jump abilities (3,8,11). Evidence also exists suggesting a significant correlational relationship between sprint speed and change-of-direction ability (9,11). The relationship between jump ability and change-of-direction performance is less clear (9,11). Although it has been suggested that upper body strength plays a role in activities such as sprinting (13), relationships between upper and lower body measures have received less attention.
Similar to the test battery used annually at the NFL combine, batteries implemented across sports and levels commonly include sprint, jump, and change-of-direction ability measures. Given that research exists suggesting that many of the tests commonly incorporated into evaluation test batteries are associated, and thus may be measuring similar skills, it is confusing that such high profile batteries as that performed at the NFL combine continue to incorporate multiple tests that may be measuring similar attributes. It would seem inefficient and unnecessary to include tests measuring similar skills or to include tests able to predict the outcomes in other tests incorporated into the same battery. It may be that research to date is viewed as inconclusive or statistically insufficient.
Because of the nature of much exercise science research, sample sizes in studies investigating the relationships between various performance measures have tended to be quite small and can therefore lack statistical power. With one exception in which the research used a sample incorporating multiple sports and multiple age categories, the research previously described employed sample sizes ranging from 10 to 38 participants (3,8,9,13). The NFL combine tests a relatively large (i.e., <300) sample of athletes annually. Thus, the combine is an excellent source of sprint, jump, change of direction, and upper body strength data. The purpose of this study was to investigate the relationships between the physical attributes measured at the NFL combine, specifically, to provide support using a large sample, for the hypothesis that sprint, jump, and change-of-direction abilities are strongly correlated and to examine the relationships between these attributes and upper body strength.
Methods
Experimental Approach to the Problem
National Football League combine data were examined to determine the relationship between sprint, jump, change-of-direction ability, and upper body strength. Correlations between combine performance measures are presented.
Subjects
Data included were for players who attended the combine from 2005 to 2009 and were drafted in the same year. Those players invited to the combine but not selected in the subsequent draft were not included. It is also possible that data for certain players invited to the combine and subsequently drafted the same year have not been obtained. Data were collected from NFLdraftscout.com and are deemed accurate. Place kickers and punters were not included because players in these positions are required to perform the 36.6-m sprint only. Although a total of 1,136 players were included in the study, all combine draftees did not necessarily complete all physical tests making up the combine. Furthermore, the quarterback and wide receiver positions are exempt from performing the bench press test. Therefore, in many instances, correlations have been calculated from samples that were somewhat <1,136. Sample sizes are presented for each correlation. Institutional review was considered a nonissue because of the retrospective nature of this study and the fact that no names are revealed. Further, all data were retrieved from public access domains.
Procedures
Data from the following combine tests were analyzed.
The 36.6-m Sprint
From a 3-point stance, players run 36.6 m as fast as possible. Split times are also recorded at 9.1 and 18.3 m. Thus, the 36.6-m sprint test provides 3 separate outcome measures.
Vertical Jump
Jump height is measured using a device (e.g., Vertec) whereby players jump for maximal height from a standing 2-footed position in a countermovement manner using arm swing. At the peak of the jump, the player reaches as high as possible with a single hand to move horizontal vanes of the Vertec. Vertical jump height is calculated by subtracting the player's standing-reach height from the height of the highest vane moved.
Horizontal Jump
Horizontal jump distance is measured. From a standing 2-footed position, players jump forward for maximal distance using a countermovement and arm swing. Jump distance is measured as the distance from the start line to the nearest body part upon landing (this is typically the point of heel contact).
The 18.3-m Shuttle
From the starting position, players run 4.6 m in 1 direction, quickly change direction, and run 9.1 m in the opposite direction and then change direction again and run a final 4.6 m in the opposite direction (i.e., the direction initially run). The test is run in both directions (i.e., left and right) for maximal speed, and the average of the 2 tests is recorded as the score. The 18.3-m shuttle run measures the change-of-direction ability (1).
Three-Cone Drill
Players run around 3 cones placed in the shape of an “L,” with 4.6 m between each cone. From a 3 point stance, players run a predetermined route as quickly as possible. The 3-cone drill also measures the change-of-direction ability (4).
Bench Press
Players bench press 102.1 kg for maximum repetitions to measure upper body strength. The bar must touch and briefly pause on the chest before being returned to the start position, of full arm extension, for a repetition to be deemed countable. Because of the achievement of high repetitions (Table 1), this test is considered a measure of upper body strength endurance.
;)
Table 1: National Football League combine descriptive and performance characteristics for the years 2005–2009 (mean [SD]).
Statistical Analyses
To determine the strength and direction of relationships between the performance measures, Pearson's r correlations were generated using SPSS 18. Magnitude of effect of the correlations was determined as follows: trivial, <0.10; small, ≤0.10–0.29; moderate, 0.30–0.49; large, 0.50–0.69; very large, 0.70–0.89; and nearly perfect, 0.90–0.99 (5). The coefficient of determination (r2) was calculated to investigate the explained variance among tests. In cases where r2 < 0.50, the variables are considered to be independent of each other (10). A 5% level of significance (p ≤ 0.05) was used to determine statistically significant correlations.
Results
Descriptive and performance characteristics are presented in Table 1. Correlations are presented in Table 2. Depending on the relationship between performance measures, the interpretation of the sign of the correlation may not be obvious. Lower scores are preferable for the 9.1-, 18.3-, and 36.6-m sprints; 18.3-m shuttle; and 3-cone drill. Conversely, higher scores are preferable for the vertical and horizontal jumps and for the bench press. All correlations are significant at p < 0.05. The correlations between the 9.1-, 18.3-, and 36.6-m sprint times were nearly perfect (r ranged between 0.900 and 0.967) and those between the linear sprint times and jump performance were large to very large (r ranged between 0.593 and −0.777). The correlation between the jump tasks was very large (r = 0.742) and that between the change-of-direction tests was nearly perfect (r = 0.948). The common variance (r2) was >50% among all linear sprints and between horizontal jump performance and the linear sprints. The associations between the jump tasks and linear sprint times become progressively stronger with longer split distances. The bench press is positively correlated with all running drills and inversely correlated with the jump tests.
Table 2: Intercorrelation coefficient matrix between National Football League combine performance measures for the years 2005–2009.*
Discussion
Arguably, because of the large sample size, the magnitude of the correlation coefficient is of greater importance than statistical significance. For example, although the correlation (r = 0.250) describing the relationship between performance in the 18.3-m shuttle and 9.1-m sprint time is statistically significant (p < 0.01), the shared common variance (r2) between the 2 is only 6.25%. Thus, in the following discussion, attention will be paid to the magnitude of the correlation coefficient rather than to statistical significance.
The present findings provide evidence supporting the hypothesis that stationary-start sprints of 9.1, 18.3, and 36.6 m are very strongly associated. The relationship between sprinting and jumping is less strong. The shared common variance (r2) between sprint and jump abilities ranges between 35 and 60%. The hypothesis that a strong relationship exists between change-of-direction ability and sprint and jump abilities is refuted. It would appear that change-of-direction ability is less strongly associated with linear sprinting and vertical jump performance than was previously considered (9,11). Importantly, evidence is provided using a relatively large sample size.
The 3 linear sprint times exhibited nearly perfect correlational magnitudes of effect and shared common variances implying that the tasks involve similar motor skills. Presumably 3 separate linear sprint times are collected because they are deemed to measure different athletic attributes. The first phases (9.1- and 18.3-m splits) arguably represent acceleration, whereas the 36.6-m time is likely influenced by maximum speed (2). If indeed the split times do measure different athletic skills (e.g., acceleration and maximum speed), it would seem that maximum speed is dependent on acceleration, and vice versa, or that success in one test predicts success in the other. In either circumstance, a single time measure is likely sufficient. Alternatively, it is possible that acceleration and maximum speed are independent motor skills and should be tested independently. Previous research has suggested that “flying” (i.e. running start) sprints are a better measure of maximum speed as compared with stationary-start sprints (12). Flying times are calculated by allowing the athlete a running start (e.g., 20 m) to achieve maximal, or near-maximal, velocity at the initiation of timing. It is possible that a “purer” measure of maximum speed would be less strongly associated with acceleration and should be considered in the construction of test batteries designed to measure independent motor skills.
Change-of-direction ability exhibited small-to-moderate correlational magnitudes of effect with the linear sprint times and vertical jump performance. These findings are in contradiction with previous observations (9,11). Regardless of the arbitrary scale assigned to describe the magnitudes of the correlation coefficients and statistical significance using p values, close examination of the correlation coefficients themselves reveal differences among this study and those described in the previous sentence. Specifically, the correlations in question in this study range between 0.250 and 0.464. Those in the studies referred to ranged between 0.53 and 0.99 in 1 study (9) and between 0.590 and 0.717 and 0.533 and 0.831 in the other study (11). The correlation ranges referred to in the second study were determined for 2 independent samples of collegiate athletes. The varying results between this study and those referred to in previous research could be because of a number of factors, including change of direction test implemented, athletic profile of sampled population, gender, and population size. The sample sizes of the previous research were 10, 51, and 79, respectively, and the populations sampled were female softball, soccer, and lacrosse players, respectively. Further, the relationships reported are based on somewhat different techniques of performance measurement. It is possible that the above-described factors, or others, play a role in the varying results. Nonetheless, the sample size of this study adds considerable weight to the argument that change-of-direction ability is less strongly associated with linear sprint and vertical jump abilities than was previously considered. The inclusion of a change-of-direction ability measure into a test battery also including sprint and jump measures is likely warranted.
Interestingly, performance in the horizontal jump correlated much more strongly with change-of-direction ability than did vertical jump. Although speculative, it may be that motor skills necessary for the horizontal jump are common to those necessary for change of direction. It is possible that the forward propulsion associated with the horizontal jump, as compared with vertical lift associated with the vertical jump, aids in acceleration-influenced sprinting tasks. If this is the case, it would also help to explain the stronger correlations exhibited between the horizontal jump and the acceleration-influenced linear sprint times examined in this study, as compared with the observed correlations between the vertical jump and same linear sprint times. Previous research (11) has indicated that vertical jump ability is more strongly associated with maximum speed, as compared with acceleration, and suggests characteristics associated with vertical jumping (e.g., vertical propulsion) may be more important for maximum speed, as compared with acceleration.
The correlations generated between the various sprint and jump tasks are of a magnitude implying that similar motor skills are involved. Nonetheless, the size of the shared common variance (i.e., ±50%) is such that the inclusion of both sprint and jump tasks is likely warranted in the construction of test batteries. The magnitudes of effect between the jump tasks and linear sprint times become progressively stronger with longer split distances, suggesting that jump ability is more strongly associated with sprint ability at longer (e.g., >9.1 m) distances. With respect to the countermovement vertical jump, this is supported by previous research (11). The current findings support the notion that the stretch-shortening cycle is particularly important at maximum speed (7,13).
Bench press performance is positively correlated with all running drills and inversely correlated with the jumping tests, indicating that upper body strength endurance may not be of benefit to these tasks. Because upper body strength has been suggested to play a role in sprint ability (13), the direction of the correlations presented here may be somewhat counterintuitive and deserve further attention. It is possible that the body type and skill requirements of a rather large number of the athletes making up the current sample were such that they may help to explain the direction of the correlations involving the bench press. Specifically, over one-third of the athletes included in this study were linemen. Linemen are the heaviest of the players making up an American football roster and commonly have relatively large percentages of body fat (1,2). They are very strong and rarely required to run distances of >10 m during intermittent play often lasting <10 seconds. Given that these athletes are relatively heavy and have great upper body strength, and are not required to be good sprinters or jumpers, one might expect this group to exhibit correlations in the direction observed. It may be that, in general, the anthropometric measures of the entire sample are quite different from those of a group of sprinters in which upper body strength may be positively correlated with sprint time.
There are limitations to this study. It has been assumed that the collection of the test data at the combine was done appropriately. Because the data were mined and not directly collected by the author, it is impossible to comment on collection technique rigor. Although all participants were NCAA American football players drafted to the professional ranks, the diversity of skills necessary to succeed at the various team positions means that a variety of athletic profiles were included. Care should be taken while extrapolating the current findings to a less diverse group.
Practical Applications
If various linear sprint times are collected with the intention of independently measuring acceleration and maximum speed, flying sprint times, as a measure of maximum speed, are likely less contaminated by acceleration than are stationary-start sprints. Within a single test battery designed to measure various athletic traits, an “acceleration” sprint (e.g., 5–10 m) and a “maximal speed sprint” (e.g., flying 20 m) may be preferred to varying length stationary-start sprints.
The results of this study support those of previous research and the hypothesis that sprint and jump ability are associated. Although significantly associated, it appears that vertical jump ability can be regarded as an independent motor skill as compared with acceleration-influenced sprint tasks. Horizontal, as compared with vertical, jump ability is more strongly associated with linear sprinting and change-of-direction ability. Practitioners should be aware of the stronger association horizontal jump ability has with these motor skills. Previous research has reported a lack of significant association between vertical jump and measures of sprinting and change of direction (9) and thus supports the argument that the vertical jump is a reasonable alternative for a test battery designed to measure specific independent skills. In the construction of test batteries, if acceleration-influenced linear sprints and change-of-direction measures are to be included along with a single jump measure, the vertical jump is recommended over the horizontal jump.
The nearly perfect correlation exhibited between the 2 tests of change-of-direction ability suggests that a single measure is likely sufficient in any test battery. Similar to the argument recommending the vertical jump over the horizontal jump (i.e., less shared common variance with other measures), the 18.3-m shuttle may be preferred to the 3-cone drill. Although a detailed discussion as to the appropriateness of the combine battery is beyond the scope of this study, and not the purpose of the research, the NFL may be well advised to revise the current combine physical test battery.
In the current sample, upper body strength would appear to be independent of lower body tasks. As upper body strength is an important attribute to successfully perform in many positions in American football, the inclusion of a measure of upper body strength is warranted. In any sport in which upper body strength is deemed of importance, incorporation of an upper body strength measure into test batteries would seem prudent. When devising test batteries for the purpose of monitoring training adaptation and player selection, coaches and practitioners should attempt to measure independent motor skills. Inclusion of multiple tasks measuring similar skills is inefficient and unnecessary. Careful consideration of the skills necessary to achieve success should be followed by careful determination of the tests to measure those skills.
References
1. Barker, M, Wyatt, TJ, Johnson, RL, Stone, MH, O'Bryant, HS, Poe, C, and Kent, M. Performance factors, psychological assessment, physical characteristics and football playing ability.
J Strength Cond Res 7: 224–233, 1993.
2. Brechue, WF, Mayhew, JL, and Piper, FC. Characteristics of sprint performance in college football players.
J Strength Cond Res 24: 1169–1178, 2010.
3. Cronin, JB and Hansen, KT. Strength and power predictors of sports speed.
J Strength Cond Res 19: 349–357, 2005.
4. Hoffman, JR, Ratamess, NA, Klatt, M, Faigenbaum, AD, and Kang, J. Do bilateral power deficits influence direction-specific movement patterns?
Res Sport Med 15: 125–132, 2007.
5. Hopkins, WG. A scale of magnitudes for effect statistics A new view of statistics [online]. Available at:
http://sportsci.org/resource/stats/effectmag.html. Accessed May 25, 2010.
6. McGee, KJ and Burkett, LN. The National Football League combine: A reliable predictor of draft status?
J Strength Cond Res 17: 6–11, 2003.
7. Mero, A, Luhtanen, P, Viitasalo, JT, and Komi, PV. Relationships between the maximal running velocity, muscle fibre characteristics, force production and force relaxation of sprinters.
Scand J Sports Sci 3:16–22, 1981.
8. Mikola, M and Mehis, V. The relationships between jumping tests and speed abilities among Estonian sprinters.
Acta Axademiae Olympiquae Estoniae 15: 9–16, 2007.
9. Nimphius, S, McGuigan, MR, and Newton, RU. Relationship between strength, power, speed, and change of direction performance of female softball players.
J Strength Cond Res 24: 885–895, 2010.
10. Thomas, JR, and Nelson, JK.
Research Methods in Physical Activity. Human Kinetics, Champaign, IL, 2001.
11. Vescovi, JD and McGuigan, MR. Relationships between sprinting, agility, and jump ability in female athletes.
J Sport Sci 26: 97–107, 2008.
12. Young, W, Russell, A, Burge, P, Clarke, A, Cormack, S, and Stewart, G. The use of sprint tests for assessment of speed qualities of elite Australian Rules Footballers.
Int J Sports Physiol Perform 3: 199–206, 2008.
13. Young, WB, Newton, RU, Doyle, TLA, Chapman, D, Cormack, S, Stewart, G, and Dawson, B. Physiological and anthropometric characteristics of starters and non-starters and playing positions in elite Australian Rules football: A case study.
J Sci Med Sport 8: 333–345, 2005.
