A 10-year Evaluation of the NFL Combine. Do Combine Results Correlate with Career Longevity for NFL Offensive players?

Background: The National Football League (NFL) combine has been found to play a significant role in NFL draft decisions. Previous studies have demonstrated correlations between a player’s combine metrics and draft position and performance, but no study has investigated the purposes of our study, the relationship between these combine tests and NFL career longevity. Methods: A retrospective review of combine test results from all offensive players participating in the NFL combine from 2005-2015 was completed. Tests included height, weight, BMI, 10-, 20-, 40-yard dash, bench press, vertical jump, broad jump, short shuttle, and 3-cone. Career length was determined by online review of NFL rosters from 2005-2017. Statistical analysis using logistic regression was completed to determine the association of each player’s original (O) and normalized (N) combine test results with career longevity. Results: Quarterbacks on an NFL roster at year five were 0.75 inches taller than quarterbacks not on a roster (p<0.05). There was an association between 10-yard (RB: p<0.05; WR: p<0.05), 20-yard (RB: p<0.05; WR: p<0.05) and 40-yard (RB: p<0.05; WR: p<0.05) dash times, as well as broad jump (RB: p<0.05; WR: p <0.05), for running backs (RB) and wide receivers (WR) on a roster at 5-years. Tight ends on a roster at 5-years were found to have performed significantly better in the vertical jump (p<0.05) and broad jump (p<0.05), while also having greater body weight (p<0.05) than those not on a roster. Vertical jump (p<0.05), broad jump (p<0.05) and 40-yard dash (p<0.05) were found to be associated with offensive lineman on an NFL roster at 5-years. Conclusion: The combine has been used by NFL teams to gather objective data about player’s athleticism, and this study shows the specific drills and corresponding traits that are related to longer careers for each position. This study found associations between traits that are reflected by the tests performed at the NFL combine and career longevity. The findings of this study may serve as an additional resource for teams to assist in the assessment of a player’s potential career longevity based on their individual combine results.


Introduction
The annual National Football League (NFL) combine allows teams to assess former NCAA players' medical status, athleticism, and personality traits. The combine has been found to play a significant role in NFL draft decisions [1,2]. While other facets of a team's decision-making process remain subjective and generally not openly discussed by NFL teams, the standard-ized athletic tests from the combine offer an opportunity for objective study. The drills players partake in during the combine are timed and measured, thus serving to provide insight into a player's athletic traits. Due toperformance in combine drills being reflective of athletic and injury-reducing traits, such as ankle flexibility in the 3-conedrill and jump strength in the vertical jump, it is possible that studying a player's results at CrossMark ← Click for updates doi: 10.7243/2055-2386-7-8 the combine may offer insight into their performance and health throughout their career [3,4].
The athletic tests at the combine offer insight into the athletic traits that contribute to the productivity, health and therefore longevity of an NFL player's career. Prior research has studied the relationship between measurements from the NFL combine and their relationship to draft order and NFL success [1,2]. The positive results from prior research offers insight into what traits may be important for productive careers. The existing research studies have varied in terms of positions studied, success metrics used and whether or not they normalizedthe combine drill results by player size [3,4]. Meanwhile, the medical information obtained at the combine has been studied in relation to career length. According to the NFL Players Asssociation the average career of an NFL player is 3.3 years. Quaterbacks average a career of 4.44 years, running backs average 2.57 years, wide receivers average 2.81 years, with tight ends and offensive lineman averaging 3.5 years.To our knowledge, no study has investigated the relationship between both original and normalized combine drill performance and career longevity. Therefore, the purpose of our study is toinvestigate the relationship between combine drills and career length in the NFL for offensive players. We hypothesize that better performance in combine drills will be associated with increased career length across the offensive positions studied.

Methods
Analysis was completed of all players who competed in the NFL combine from 2005 through 2015.Combine results were retrospectively collected from NFLdraftscout.com, a public website that provides historical combine results. Results from each drill that a player had performed at the combine were recorded. These included height, weight, BMI(body mass index), 10-, 20-, and 40-yard dash times, bench press repetitions, vertical jump height, broad jump distance, short shuttle and 3-cone times. The significance of a player's performance on a given combine test can be evaluated by understanding what skills and traits each drill tests (Table 1) [1,5]. Career length data was collected using ProFootballReference.com and cross-referenced using NFL.com team websites from all NFL seasons between 2005-2017.
The analysis included each player'scombine results and a ratio-scaled normalizationof each player's combine results by position. Normalizing refers to adjusting a player's performance measures by his body size relative tothose that he is being compared against, such as a position group. A ratio scaled normalization, which divides the player performance numbers by their body mass index, was used for this study. First, an average BMI was taken for each position. Next, each player's BMI was divided by the average for his position in their respective combine year, so that heavier players would have a higher relative density score. Lastly, combine scores were normalizedfor each player by dividing by their scaled BMI, which gave a normalizedresult for each test. The influence of each performance metric on whether or not a player was on a roster after the combine was evaluated individually using a univariate logistic regression model. Analyses were conducted in R3.4.2. All hypotheses tested were two-sided with P<0.05 considered statistically significant. There was no adjustment for multiple testing.
This study received IRB Approval and deemed exempt from our institution's approval board.

Results
Retrospective review identified 131 quarterbacks, 277 running backs, 428 wide receivers, 150 tight ends, and 263 offensive linemen that participated in the NFL combine from 2005-2015. Of these players, 37 quarterbacks, 197 running backs, 233 wide receivers, 121 tight ends, and 192 offensive linemen completed both ten NFL seasons and the entire list of combine drills. For the 37 quarterbacks, the vertical jump and bench press were not completed. Quarterbacks ( Table 2).

5-Year Roster
Quarterbacks that were on a NFL roster at year five were found doi: 10.7243/2055-2386-7-8 to be 0.75 inches taller than QBs that did not make a roster at year five(mean=75.44 inches (sd=1.57) vs mean=74.676 inches (sd=1.45), p=0.032). For every 0.25" taller, the odds that a QB made a roster at year five increased by approximately 9% (95% confidence interval (CI): 1.008-1.186, p=0.04). For QBs, no other combine results were significant. Table 3)

5-Year Roster
Logistic regression analysis of running backs showed a significant association between making a NFL roster for five years and original and normalized 40-yard dash (  For every unit (0.02 seconds) increase in normalized 10yard dash, the odds of a wide receiver making a roster for one season decreased by 5.6% (95% CI: 0.901-0.988, p=0.013). The odds of being on a roster for 1 season decreased by 13.9% for every 0.05 seconds added to a wide receiver's normalized 20-yard dash (95% CI: 0.797 -0.928, p<0.001). For wide receivers, the odds of being on an NFL roster for one season decreased 16.5% for every 0.1 seconds added to their 40-yard dash times (95% CI: 0.764-0.909, p<0.001). For the normalized shuttle drill, every 0.1 second increase in time decreased the odds of a wide receiver making an NFL roster for one season by 14.9% (95% CI: 0.775-0.933, p<0.001). For every 0.1 second increase in the 3-cone drill, the odds of a wide receiver making the roster decreased by 8.2% (95% CI: 0.867-0.971, p=0.003). The odds of a wide receiver being on a roster for 1 season increase by 3.4% for every 0.5" added to their vertical jump (95% CI: 1.002-1.067, p=0.037).

5-Year Roster
Receivers that were on a roster for five years had a significant association with normalized 40-yard dash   The odds of a tight end making a roster for one season decreased by 25.6% for every 0.1 seconds added to a player's 40-yard dash time (95% CI: 0.579-0.942, p=0.016). For every 0.05 second increase in 20-yard dash time, the odds of a tight end making a roster decreased by 20.3% (95% CI: 0.65-0.969, p=0.025). The odds of a tight end making an NFL roster for 1 season increased by 6.3% for every 0.5" increase to their vertical jump (95% CI: 1.006-1.129, p=0.035). For tight ends, the odds of making a roster for 5 seasons increased by 3.6% for every 0.5" increase in broad jump distance (95% CI: 1.006-1.071, p=0.024).

5-Year Roster
In the logistic regression analysis of tight ends, there was a significant association betweenmaking a NFL roster for five years and normalizedbroad jump(116.993 inches ( Offensive Line ( Table 6)

1-Year Roster
For offensive linemen that played at least one year in the NFL, there was a significant association between roster status and normalized 40-yard dash ( The odds of an offensive lineman making an NFL roster for 1 season decreased by 9.5% for every 0.02 seconds added to their 10-yard dash time (95% CI: 0.839-0.974, p=0.009). For the 20-yard dash, every 0.05 seconds added to an offensive lineman's time decreased the odds of making a roster for one season by 14.8% (95% CI: 0.755-0.954, p=0.007). Offensive lineman odds of making a roster for 1 season decreased by 20.9% for each 0.1 seconds added to their 40-yard dash time (95% CI: 0.687-0.902, p<0.001). For every 0.5" added to an offensive lineman's broad jump, the odds of making a roster for 1 season increase by 2. 1% (1.003-1.04, p=0.025). The odds of an offensive lineman making an NFL roster for 1 season decreased by 1.9% for every 0.01 seconds added to their shuttle time (95% CI: 0.965-0.996, p=0.014). For every 0.5" added to an offensive lineman's vertical jump, their odds of making a roster for 1 season increased by 4.3% (95% CI: 1.002-1.087, p=0.044).

5-Year Roster
For offensive linemen that played five seasons, there was a significant association between normalized vertical jump (p=0.02)and roster status. The odds of an offensive lineman being on a roster for 5 seasons increased by 4.3% for every 0.5" added to their vertical jump (95% CI: 1.002-1.087, p=0.026). (Table 7) Significant results for all positions at each time are listed in Table 7.

Discussion
The NFL combine is regarded as a measure of a player's athleticism and is a tool used by teams when making draft decisions [5]. This study found that specific combine drills correlated with longer NFL careers for each position for offensive players. While other studies have investigated the use of combine data to predict success or injury, this is the first strictly focused on career longevity. In agreement with prior evidence that increased height is a favorable trait for quarterbacks, associated with higher draft position and career performance, we show that taller quarterbacks are more likely to have longer careers [6,7]. Additionally, a study investigating the relationship between individual performance awards, Pro-Bowl and All-Pro selection, and combine performance found that award winning quarterbacks tended to be taller and weigh more [7]. Hedlund's study also showed that award-winning tight ends tended to be taller, however, our study found no association between height and career longevity for tight ends. There was an association with decreased body weight and at least a five-year career for players at the tight-end position, suggesting that lower weight is an important factor for a tight end's longevity and performance.
Greater first-step quickness and short area speed, as described by the 10-yard dash, and sprint speed, as described by the 20-and 40-yard dashes, were significantly correlated to greater career length for all offensive positions except quarterback. Similarly, prior studies have shown an association between these athletic qualities and several measures of success, including draft order, Pro-Bowl, and All-Pro selectionfor running backs, wide receivers, tight ends, and offensive linemen [5,7]. One study of the relationship between sprint speed of running backs and wide receivers showed a positive correlation with a running back's number of yards per carry, number of games played, and player salary, with the 10-yard dash being the greatest predictive measure [1,8]. The priority placed on speed across all non-quarterback offensive positions in the draft seems to be reinforced by its association with success and how this may translate to increased career longevity.
Agility is an important quality for players to execute offensive plays and routes effectively and themost accepted measures of athletic agility are the 3-cone drill and shuttle. NFL teams have emphasized agility for quarterbacks, running backs, wide receivers, and offensive linemen, evidenced by the relationship between a player's 3-cone time and the increased probability of higher draft position for these positions [2,5]. Career performance studies have shown the 3-cone drill to be associated with greater likelihood of award recognition for wide receivers, tight ends, and offensive linemen [7]. The shuttle was shown to be associated with greater salary for running backs and higher likelihood of being selected for the Pro-Bowl and All-Pro awards for running backs and offensive linemen [7]. We show that faster shuttle and 3-cone drill times are significantly associated with greater career length for running backs, wide receivers and offensive linemen.It is reasonable to conclude that players who are more agile are likely to be more successful and would also be more likely to make an NFL roster for multiple years.
Explosiveness and power, as demonstrated by the broad and vertical jumps, has been shown to correlate with higher draft position, and elite individual performance awards for running back, wide receivers, tight ends, and offensive linemen [2,5,7]. In regards to performance, wide receivers with greater vertical jumps were found to have significantly higher yards per reception throughout their career [8]. Our data shows a strong association between the jumps and career longevity with wide receivers, tight ends, and offensive linemen having a significantly higher chance of being on a roster for at least five years.
Of note, our study also initially included 10-year roster status in our statistical analysis. This was ultimately not included given the paucity of NFL players that play ten seasons, which resulted in only one significant association, increased bench press reps for tight ends. In order to better understand the athletic traits that may contribute to careers of ten years or more, future studies that include 10-year roster status would be well served to have a greater range of career length and combine performance data beyond the ten year windows used for each in our study in order to have a greater sample size.
A limitation to our normalized data is that we did not use allometric adjustments.While we do realize that BMI can be less useful in individuals at the the extreme of body mass, we did try to account for this by normalizing the data and comparing like sized individuals by position group. This did help us to create some validity when studying individuals that may be outliers with a high muscular component. By using this scaled BMI approach in interpreting the data we feel it allowed for the most accurate comparison in the absenceof a consistent exponent for allometric adjustments in football players.Allometricadjustments are intended to provide a more accurate normalization of athletic drill performance by body shape and size rather than simply BMI and therefore could be useful if an accurate exponent for American football athletes is developed [1]. This study was limited to only the select offensive positions and therefore cannot be generalized to other player positions. Future studies will evaluate the association of combine metrics with career longevity for other positions, and potential contributing factors to career length, such as performance and games missed due to injury.

Conclusions
The combine has been used by NFL teams to gather objective data about player's athleticism, and this study shows the specific drills and corresponding traits that are related to longer careers for each position. This study found associations between traits that are reflected by the tests performed at the NFL combine and career longevity. The findings of this study may serve as an additional resource for teams to assist in the assessment of a player's potential career longevity based on their individual combine results.