February 15, 2015 - Tony Villiotti
Which combine drills are most important?
No one seriously believes that the NFL Combine allows scouts or their geeky counterparts to sit in an office and evaluate talent based solely on Combine results. Where there is disagreement, though, is in how to interpret those results. Some believe the Combine data is relatively useless and that the true value of the Combine lies in its medical exams and interviews. I tend to believe that the data shows that the results of some drills, dependent on the playing position, do seem to be an indicator of success and should be an element in the decision-making process.
This article presents the results of an analysis that seems to support that case. It should be said up-front that no drill that provides an absolute guarantee of success. Rather, it is a matter of improving a team’s probability of success on draft day. Think of it like basic strategy in blackjack.
The analysis was performed using Combine results for the past 10 years (2005-2014) and reviewed each drill by playing position for top Combine performers and all Combine participants. The metric used in the analysis to measure success is whether a player has started one NFL season, an admittedly low hurdle but one that fits the time frame used in the analysis. A starting season is defined as any season where a player starts at least eight games. Top Performers are defined as the top five performers (plus ties) at each drill in each year for each playing position.
The analysis was intended to serve three purposes:
A large difference between Top Performers and All Participants indicates the importance of focusing on Top Performers at the Combine. Cornerbacks and defensive ends (both small and large) have the largest difference between Top Performers and All Participants.
Centers, Quarterbacks and Large Running Backs have only minor differences, indicating that Combine drills may not be all that important for those positions.
Importance of the Drills
The importance of the drills is measured by calculating the percentage of one-year starters for each Combine drill and playing position. The premise is that the higher the percentage, the more important the drill. Table 2 reflects the percentage of Top Performers that started at least one season by Combine drill for each playing position.
The information in Table 2 is then used to place Combine drills for each position in the order of importance. Please note that the Bench Press is not relevant for quarterbacks and wide receivers (very few even do the drill) and are omitted for those positions. There are also a few ties in order of importance that cannot be reflected in this table but will show up in the degree of difference.
The degree of diversity in the top drills for each position is noteworthy. Of the 15 drills ranked first (one for each playing position), six are some variation of the 40-yard dash, three are the 3-Cone drill and there are two each for the Bench Press, Broad Jump and 20-yard shuttle. Only the vertical jump is not represented.
Table 3 shows the ranking of the drills for each playing position:
A kinesiology expert might look at this table and say it does not make sense. But the results speak for themselves. Whether it makes perfect sense or is a statistical oddity, I will leave to others to debate.
Degree of Difference
The degree of difference calculation is intended to answer the question of how much more significant one drill is than another. The data from Table 2 was used to develop this index. The index value for the drill having the highest percentage for each playing position was set at 100. The index value for all other drills by playing position was calculated by dividing (1) the percentage of starters from each drill excluding the top-rated drill by (2) the percentage of starters for the top drill. For example, assume that (1) 50% of the top performing offensive tackles in the Bench Press started at least one season and (2) the Combine drill with the best outcome for top performing offensive tackles is the 10-yard split at 67.2%. The index for the 10-yard split would be set at 100.0. The index for the Bench Press would be 74.4, calculated by dividing 50% by 67.4%.
Table 4 presents the index values that were calculated. For centers there is very little difference among all the drills and the least predictive drill has an index value of about 87% of the most predictive drill. For small running backs, on the other hand, there is a significant difference between the 10-yard split and the flying 20, indicating that the 10-yard split is clearly the most important drill.
Please note that the indices should be used within each playing position. In other words the Center index of 87.1 for the 10-yard split cannot be compared to the Guard index of 74.9 for the same drill.
One Last Question
A logical follow-up question is whether the probability of success for a player increases if he ranks as a Top Performer for multiple drills. The short answer is yes. An analysis was performed that identified players who were Top Performers in each of the top three drills for each playing position. Table 5 shows the outcome.
As can be seen in Table 5, those players who rank as a Top Performer in all three of the top drills at each position do outperform all Top Performers. The overall difference, though, is modest. Several positions had relatively few data points, though, and it is difficult to draw any conclusions from those.
Follow Tony on Twitter @draftmetrics
- Identify the playing positions for which the Combine is most important
- Place the drills in order of importance by playing position
- Determine the degree of difference among the drills for each playing position
- This recognizes that simply putting the drills in order of importance does not provide enough information to evaluate importance
- More about the degree of difference later in the article
- 40 yard dash (“40)”
- Further broken down into the first 10 yards (“10”), the first 20 yards (“20) and the final 20 yards “F20”
- The final 20 yards are often referred to as the “flying 20” because the player has a running start (the first 20 yards) to begin the timing
- Bench Press (“BP”)
- Vertical Jump (“VJ”)
- Broad Jump (“BJ”)
- 20 yard shuttle (“20S’)
- 3-cone drill (“3C”)
A large difference between Top Performers and All Participants indicates the importance of focusing on Top Performers at the Combine. Cornerbacks and defensive ends (both small and large) have the largest difference between Top Performers and All Participants.
Centers, Quarterbacks and Large Running Backs have only minor differences, indicating that Combine drills may not be all that important for those positions.
Importance of the Drills
The importance of the drills is measured by calculating the percentage of one-year starters for each Combine drill and playing position. The premise is that the higher the percentage, the more important the drill. Table 2 reflects the percentage of Top Performers that started at least one season by Combine drill for each playing position.
The information in Table 2 is then used to place Combine drills for each position in the order of importance. Please note that the Bench Press is not relevant for quarterbacks and wide receivers (very few even do the drill) and are omitted for those positions. There are also a few ties in order of importance that cannot be reflected in this table but will show up in the degree of difference.
The degree of diversity in the top drills for each position is noteworthy. Of the 15 drills ranked first (one for each playing position), six are some variation of the 40-yard dash, three are the 3-Cone drill and there are two each for the Bench Press, Broad Jump and 20-yard shuttle. Only the vertical jump is not represented.
Table 3 shows the ranking of the drills for each playing position:
A kinesiology expert might look at this table and say it does not make sense. But the results speak for themselves. Whether it makes perfect sense or is a statistical oddity, I will leave to others to debate.
Degree of Difference
The degree of difference calculation is intended to answer the question of how much more significant one drill is than another. The data from Table 2 was used to develop this index. The index value for the drill having the highest percentage for each playing position was set at 100. The index value for all other drills by playing position was calculated by dividing (1) the percentage of starters from each drill excluding the top-rated drill by (2) the percentage of starters for the top drill. For example, assume that (1) 50% of the top performing offensive tackles in the Bench Press started at least one season and (2) the Combine drill with the best outcome for top performing offensive tackles is the 10-yard split at 67.2%. The index for the 10-yard split would be set at 100.0. The index for the Bench Press would be 74.4, calculated by dividing 50% by 67.4%.
Table 4 presents the index values that were calculated. For centers there is very little difference among all the drills and the least predictive drill has an index value of about 87% of the most predictive drill. For small running backs, on the other hand, there is a significant difference between the 10-yard split and the flying 20, indicating that the 10-yard split is clearly the most important drill.
Please note that the indices should be used within each playing position. In other words the Center index of 87.1 for the 10-yard split cannot be compared to the Guard index of 74.9 for the same drill.
One Last Question
A logical follow-up question is whether the probability of success for a player increases if he ranks as a Top Performer for multiple drills. The short answer is yes. An analysis was performed that identified players who were Top Performers in each of the top three drills for each playing position. Table 5 shows the outcome.
As can be seen in Table 5, those players who rank as a Top Performer in all three of the top drills at each position do outperform all Top Performers. The overall difference, though, is modest. Several positions had relatively few data points, though, and it is difficult to draw any conclusions from those.
Follow Tony on Twitter @draftmetrics 