Over at FiveThirtyEight, Josh Hermsmeyer recently wrote about wide receivers and 40-yard dash times. Using yards per route run as his measure of productivity, Josh concluded “that higher speed isn’t associated with higher on-field production.” Today I want to take a deep dive into the question of how much 40-yard dash times are correlated with wide receiver success. For a very long time, people have argued that 40-yard dash times are overrated (actually, for a very long time, people have argued that just about everything is overrated). But such a comment is paper thin, because it’s unclear exactly how “rated” 40-yard dash times are, anyway. So let’s skip the overrated/underrated analysis and dive into the data.

My sample comprises the 853 wide receivers who ran the 40-yard dash at the NFL combine from 2000 to 2017. ^[1]Why those years? PFR’s data only goes back to 2000, and players who participated at the combine more recently than 2017 have not yet accrued four NFL seasons. All data is publicly available from PFR, via Stathead Football. I then looked at how many receiving yards those players gained in their first four seasons in the NFL. ^[2]Chosen because this represents the average length of a rookie contract. The question of what metric to use to measure production is a complicated one: receiving yards is not perfect (and I will revisit this decision at the end of the article), but it should work well enough for these purposes.

On average, these 853 players ran the 40-yard dash in 4.51 seconds and gained a total of 678 receiving yards in their first four seasons; this includes the 360 of them who never gained a receiving yard in the NFL. The top three wide receivers by receiving yards over this period ^[3]This analysis, of course, excludes players who were not invited to the combine like Tyreek Hill and Josh Gordon, undrafted players like Victor Cruz, Robby Anderson, and Doug Baldwin, and players who … Continue reading were Michael Thomas, A.J. Green, and Anquan Boldin, who ran the 40-yard dash in 4.57, 4.48, and 4.72 seconds, respectively. Strike one for 40-yard dash times mattering. The fastest two players were John Ross and Donte’ Stallworth, who both ran the 40 in 4.22 seconds. ^[4]That might sound like strike two, but Stallworth was tied for the 61st most receiving yards out of this group if 853 receivers. It’s a strike for 40-yard dash time being the only thing that … Continue reading

But anecdotes can only take us so far when we have 853 players, from Ross and Stallworth on the far left, to Thomas up at the top, all the way to Mississippi State’s De’Runnya Wilson, who never played in the NFL and ran the 40 in 4.85 seconds.

A quick glance at this chart seems to support the idea that the 40-yard dash is not correlated with receiving production. But this way of presenting the data can be a bit tricky for the eye to see because the players are not evenly distributed across the spectrum of 40-yard dash times. It’s also easy to kind of lump the players in the bottom half of the graph all together, which is probably not the right way to think about this. ^[5]Although if you want to argue we should only focus on the best players, well, I’d be open to listening. So let’s break down the 40 times into small groups: from 4.22 to 4.24, 4.25 to 4.27, 4.28 to 4.30, and so on. Once we do that, the results are pretty striking.

Let’s once again graph wide receiver production by measuring 40-yard dash time on the X-Axis and receiving yards through 4 seasons on the Y-Axis.  But this time, let’s group them in these ranges and use a bubble chart where larger bubbles represent larger sample sizes.  For ease of reference, I am using the mid-point of each range, so a player who runs in the 4.52 to 4.54 range will get plotted at 4.53 seconds on the X-Axis.  As you can see, there does appear to be a pretty strong correlation between 40-yard dash times and wide receiving production, particularly once you focus on the 94% of receivers in the 4.32 to 4.68 range. ^[6]Which really means the 4.31 to 4.69 range, of course.

While there is lots of noise in the data, the correlation here is probably stronger than you would expect.  Does the 40-yard dash time matter much for an individual player? Probably not.  But if you had 20 wide receivers who ran the 40 in 4.40 seconds, if you knew nothing else, you’d probably expect them to fare better than 20 wide receivers who ran the 40 in 4.50 seconds.

Let’s get back to a question from the beginning of this article: is receiving yards a useful way to measure production?  We are ignoring receptions, first downs, receiving touchdowns, blocking ability, and everything else a wide receiver can do, although given the large sample size, you would hope that any major differences between those statistics and receiving yards would balance out.  Wide receivers who play with great quarterbacks will be at an advantage, but again, we can rely on the large sample size to aid us here (and continue to measure this in the future as our sample size grows).  The two questions that I think worth asking are:

• Is this biased in favor of players who aren’t very good but still see the field because they were higher draft picks or better prospects (in part because of their higher 40 time)?
• Would a measure of efficiency (like yards per route run) be better?

For question 1, I think the answer is undoubtedly yes.  There probably isn’t a meaningful difference in talent between a 2nd round receiver who gains 900 yards through four years or an undrafted free agent who catches 300 yards through four years; we know that higher draft picks (and/or faster players) are more likely to get on the field earlier, which lets them have an easier time getting more yards.   Is that enough to make the data meaningless? I don’t think so.

For question 2, this is also tricky to analyze.  What is the proper minimum number of routes run?  And even at a higher threshold, quantity still matters; a player who averages 2.0 yards per route run on 1500 routes is a more meaningful contributor that one who averages 2.20 yards per route run on 800 routes.   We could try to come up with the best single way to grade each of these wide receivers, and then compare 40-yard dash time to that metric.  But that feels like its own separate project.  What do you think?

And if you are curious…. in the next post, I will look at draft position in addition to 40-yard dash time.

Please leave your thoughts in the comments. I welcome your feedback!