Part 2: The NFL Combine: Do Big Investments Yield Big Returns?

Recall that we are interested in seeing whether combine performance, draft position,
opportunities as a rookie, and offensive performance as a rookie are related to each other. To do this, we used combine results, draft position, and first year NFL statistics for running backs and wide receivers from 2008 to 2012. More specifically, we explored four relationships within the running back and wide receiver data:

(1) Combine performance should predict where rookie running backs and wide receivers are taken in the draft.
(2) Draft position should predict the number of playing opportunities for rookie running backs and wide receivers.
(3) More opportunity to contribute should predict the offensive performance of rookie running backs and wide receivers.
(4) Combine performance should predict performance for rookie running backs and wide receivers.

Relationship 1: Combine Performance and Draft Position We set out to build a statistical model3 to show which of the six combine events (40-yard dash, bench press, cone drill, shuttle drill, vertical jump, and broad jump) best predict players’ draft position. We found that three of the six combine events together predict where a RB was drafted: 40-yard dash, cone drill, and shuttle drill. In fact, these three combine drills together mathematically account for 20% of the reason why a player is drafted with a certain pick.

Getting back to Jahvid Best, it seems like he was fortunate to improve his 40-yard dash and cone drill times, as it resulted in his being drafted higher than expected.

For wide receivers, only one combine metric – broad jump – accurately predicts draft position (broad jump accounts for 8% of the reason why a player is drafted a certain pick). In other words, the farther a wide receiver broad jumps, the higher he will be drafted. This relationship seems to hold true for Jon Baldwin, who posted one of the longest broad jumps at the 2011 combine.

Figure 1 illustrates the relationship between what our statistical model would project as a player’s draft position and where that player was actually taken in the draft. For many players, this model is spot on. In 2008, the Bengals drafted wide receiver Jerome Simpson 46th overall. Based solely on his combine broad jump, our statistical model predicted that he would be drafted 49th. While not every prediction is this closely aligned with their draft position, the statistical analysis does provide some statistical insight into the logic of coaches and scouts observing the combine.
Relationship 2: Draft Position and Opportunities as a Rookie Once a player is drafted, the number of opportunities he gets as a rookie is largely contingent on the team to which he is drafted. Does the player fit a need? Is there an established NFL starter already at his position? Is he being drafted so that the team can “develop” him? How quickly can he pick up the team’s system? All of these questions clearly play a role in determining playing time in a player’s first NFL season. It seems logical, though, that in general, players who are taken higher in the draft would have more opportunities to contribute in their first year. Often, teams draft for need — they use their first handful of picks to select players they think will have the most immediate impact; guys who can step in and contribute right away.

Is a player’s draft position related to his rushing and receiving opportunities as a rookie? We measured opportunity by combining two metrics – the total number of times a player was thrown to plus the total number of rushing attempts.We used targets as opposed to receptions to better reflect total opportunities. To statistically examine these relationships, we looked at the relationship between draft position and opportunities for running backs and wide receivers drafted between 2008 to 2012.

Overall, it seems that draft position does have a lot to do with how many opportunities a player gets as a rookie. For running backs, draft position is clearly related to the number of opportunities that a rookie received, as running backs who are drafted higher get more opportunities. Jahvid Best fit this mold – taken in the late first round, the Lions expected him to contribute right away. Best earned 245 combined rushing and receiving attempts in his first full season as the starting running back. From 2008 to 2012, the average running back drafted in the first round got an average of 171 rushing and receiving attempts their rookie year. Clearly, the Lions were excited about Best’s potential, and awarded him with plenty of opportunities.  

The relationship between draft position and opportunities was nearly identical for wide receivers – players drafted higher tended to see more opportunities. Jon Baldwin is another archetypal case. He saw 51 targets (he had no rushing attempts) in the 11 NFL games he played in his first year. Baldwin was averaging about 4.7 targets per game, whereas the average rookie wide receiver drafted in the first round from 2008 to 2012 saw about 4.38 targets per game. Had Baldwin not suffered a wrist injury (which he incurred from a locker room fight with a teammate), he was on pace to finish the 2011 season with 75 targets – right in line with what we would expect for a first round WR.
Figure 2 displays this relationship between draft position and total opportunities for both running backs and wide receivers, indicating that the higher the draft position, the more opportunities players will see. First round running backs should expect to see about 170 total opportunities, while first round wide receivers should expect to see about 60 total opportunities. For every three draft positions a running back or receiver falls, that rookie should expect to 1-3 fewer opportunities within their first year.

Relationship 3: Opportunities and Offensive Performance Does the more opportunities a rookie running back or wide receiver get – both rushing and receiving – equate to better offensive production? To explore this, we created a hybrid metric of offensive performance, composed of the average of all receiving and rushing yards, also known as the average yards from scrimmage. This measures the offensive impact a player makes each time they touch the ball.

Figure 3 suggests that the rookie running backs and wide receivers who get more opportunities get more yards per attempt, but only up to a certain point. Running backs who get around 75 opportunities produce about the same offensive performance as running backs who get 400 opportunities (about 4.1-4.6 yards per opportunity).
Are better players getting more playing time or does more playing time make for better players? Without an experiment, we can’t know for sure. But, as we have seen, Jahvid Best and Jon Baldwin got plenty of opportunities their rookie years based, at least in part, on their outstanding combine performance. But did they deserve the extra opportunities because they were great players, or were they getting opportunities for another reason?
Relationship 4: Combine Performance and Offensive Performance as a Rookie Recall that the 40 yard dash, shuttle drill, and cone drill predicted where a running back was drafted. For wide receivers, broad jump was the most reliable combine event that predicted draft position. As we have demonstrated, that draft position then predicts the number of opportunities, both rushing and receiving; in turn, those opportunities predicted offensive performance.
So it would follow that the same combine events used to predict draft position (40 yard dash, shuttle drill and cone drill for running backs; broad jump for wide receivers) should predict offensive performance.

Surprisingly, the data showed that 40-yard dash time, shuttle drill, and cone drill time were, in fact, not predictive of rookie running back success.
Figures 4a – 4c are the offensive performances for all rookie running backs between 2008 to 2012 matched with the three predictive combine events. There is nearly an equal distribution on each side of the “average” for each of the three graphs. If any of the three combine events were predictive of first year performance, the points on the graphs would be higher on the right side of the average, than on the left.

Best had the fastest 40 yard dash of all running backs at the 2010 NFL draft. He finished with the 11th best offensive performance for all rookie running backs selected in the first round from 2008 to 2012. Best’s offensive performance was 4.27 yards per touch. The average offensive performance for all first round rookie running backs from 2008 to 2012 was 4.75 yards (or 4.4 yards if you remove Felix Jones who posted 8.4 yards per opportunity).

We then looked at the relationship between broad jump and wide receivers’ first year offensive performance. Again, we found no relationship between the two. The length of a wide receiver’s broad jump at the combine had no bearing on his offensive performance as a rookie.

Figure 5 shows the offensive performance for all rookie wide receivers between 2008 to 2012 matched with the predictive combine event, the broad jump4. Again, there is nearly an equal distribution on each side of the “average” for each of the three graphs.

There were 16 wide receivers picked in the first round of the NFL draft from 2008 to 2012, each with varying distances on their combine broad jump. Jon Baldwin’s offensive performance was 4.98 yards, the average offensive performance for all first round wide receivers during this time frame was 8.35 yards.
There’s lipstick on the Princeton pig
Best and Baldwin could be called “good” players in their rookie seasons, but were they first round caliber players? They were first round picks, due in part to their performance at certain combine events.
They both got first round playing time their rookie seasons, but were they worth a first round pick? Rookie players may get better with more experience, but if teams are willing to give the ball more often to higher drafted players, wouldn’t these teams do so in the hopes that these players merit the touches? Wouldn’t teams want to give more opportunity to a player who truly deserves their status as a first round selection and the high price tag that comes along with it?
If a running back draft prospect wants to improve his draft position, he should focus on getting his 40 yard dash, cone drill, and shuttle drill times as low as possible. A wide receiver on the other hand, should buy a pair of shoes with springs in them and broad jump as far as he can. These are the combine events that can push players to higher draft spots and maximize their first contract. According to the data, better performance on these combine events are some of the criteria that GMs, owners, and coaches may be using (whether they realize it or not) when they decide where to draft future NFL running backs and wide receivers.

However, despite the fact that these combine events can increase a player’s draft position, these combine events do not predict his first year offensive performance.If the same combine events that predict where a player is drafted do not predict players’ rookie year performance, might GMs, owners, and coaches picking based on the wrong criteria?

Read Part 3:

Leave a Reply