Every year, some veteran receivers decline and some young receivers break out. Every offseason, receivers change teams, impacting the passing offense they just left as well as the one they just joined. And every August, we try to make sense of what these changes mean from a fantasy standpoint.
One prominent theory is that receivers who ascend in the passing game will garner more targets and therefore produce more points. On the other side of the aisle, some believe that these receivers will face tougher coverage, which will hurt their production.
This debate is especially relevant today in the wake of news that Arian Foster is likely to miss at least half of the 2015 season, if not more. While the loss will certainly impact Houston’s running game, some have questioned what this will mean for its passing game, too.
DeAndre Hopkins was already looking to climb into a more prominent role with the departure of Texans’ mainstay Andre Johnson. Now, with the running game primed to struggle, he could see his role in the offense skyrocket over last season’s. Which of the two theories will best help us predict what will happen to Hopkins this year?
Making Sense of Competing Theories
The contrasting narratives are problematic because they provides us a ready-made explanation for any possible outcome. When Peerless Price flopped in 2002, it was because of tougher coverage. When Muhsin Muhammad exploded in 2003, it was because of more targets. In short, we’re left with a model that explains everything, and a model that explains everything actually explains nothing.
This isn’t to say that tougher coverage can’t lead to reduced efficiency. I’m sure differences in coverage certainly did play a role in Peerless Price’s struggles in 2002. (The change in quarterback and offense likely didn’t help, either.) Likewise, undoubtedly Muhammad wouldn’t have been as productive in 2003 had a week 1 injury to Steve Smith not opened the door for a larger workload.
The problem is how we stitch these two narratives together. Anecdote alone is an awfully thin thread and easily broken. For best results, one should use something a bit more substantial.
To that end, I wanted to examine the impact tougher coverages had on a receiver’s per-target efficiency in a more rigorous manner. Consider this an attempt to upgrade the seams of our theory.
How Do We Measure Tougher Coverage?
This is the crux of our problem. Short of analyzing decades worth of film, there’s no way for me to measure the quality of coverage a receiver faced. I can either abandon all attempts… or I can take a few shortcuts. I opted for the shortcuts.
For the rest of this article, I’m going to let a receiver’s offensive role stand in as a proxy for the coverages he faced. The idea is that opposing defenses devote more resources to covering a heavily-used first option than they do to covering a lightly-used second option. Therefore, I will assume that anyone who saw a dramatically increased role from one year to the next will have also seen a commensurate increase in the quality of coverages.
If we want to examine receivers with an increased role, we first must identify what qualifies as an increased role. An increased role is not particularly obscene, but it's still one of those things where we just "know it when [we] see it". Crafting a set of parameters that create a good list of candidates while minimizing false positives proves fairly tricky.
In the end, I settled on the following: I wanted the complete sample of receivers who finished 25 or more targets behind the team leader in year N, and 25 or more targets ahead of the second-place option in year N+1. This gives us guys who were clear “second fiddles”, then went on to become clear “top bananas”.
Using this criteria helps weed out long transitions or “passing of the baton” situations. For example, in 2007 Reggie Wayne led all Colts in targets by more than 50. This transition came, however, after a long apprenticeship; in 2006, Wayne trailed Marvin Harrison by just 11 targets. Does Reggie Wayne really fit the spirit of the “increased role”? In cases where I’m not sure, I’d rather err on the side of being too exclusive than too inclusive.
Now, with my list compiled, I went through and removed everyone who only appeared to be a “second fiddle” in year N because of injury. Julio Jones finished 5th on the Falcons in 2013, then led them by nearly 40 targets in 2014. Of course, the only reason he had so few targets in 2013 is because he was injured in his 5th game. On a per-target basis, Julio was the clear “top banana” both years.
Finally, I removed all instances where the receiver played on a different team in year N and year N+1. This, sadly, cost us examples like Peerless Price. At the end of the day, though, if we’re trying to isolate the impact of the coverage alone, we need to control as much as possible for other variables. Changing teams changes everything, to the point where it’s hard to isolate what is causing any effect we observe.
(This is not to suggest that remaining on the same team keeps everything else constant. Sidney Rice might have played on the Vikings in both 2008 and 2009, but there’s a slight difference between playing with Gus Frerotte and Tarvaris Jackson, and playing with Brett Favre. Obviously, this is something we cannot really help unless we want to reduce our sample to unusable levels. We can only hope that any changes in supporting cast will be relatively randomly distributed and will wash out in the analysis.)
Footballguys’ target data goes back to the 2002 season. After culling my data set, I was left with 31 pairs of seasons. Let’s dig into the numbers and see where they lead us.
What Happens When a Receiver’s Role Increases?
The average year N production over the 31-receiver sample was 65.5 targets and 506.5 yards. The average year N+1 production was 128.5 targets and 1028.0 yards. At first blush, these averages look promising for the cutoffs I used. That certainly looks like the profile of a receiver who saw a dramatic increase in his role from year one to year two.
Here is the complete data set, including the player, years, targets, yards, and yards per target (or YPT):
|Name||Year N||Year N Targets||Year N Yards||Year N YPT||N+1 Targets||N+1 Yards||N+1 YPT||Target Change||YPT Change|
Yes, you’re reading that average right at the end. On average, the receivers saw their yards per target increase by 0.27 when they became a featured part of their respective offenses. And that’s the weighted average; the unweighted average increase was 0.37 yards per target.
Of the 31 receivers, eight saw their YPT increase by at least one full yard, four saw it decrease by one full yard, and 19 remained relatively unchanged from one year to the next.
So… that’s weird, right? What the heck is going on here? Why are receivers getting better as their targets increase?
The first answer is that any sample like this is going to be very prone to selection bias. More specifically, the results reflect the outcome of survivorship bias. There have undoubtedly been more than 31 times in the past 13 seasons where a wide receiver was projected to receive an increased role, but failed to achieve it because they were unable to handle the increased defensive attention.
For an easy illustration of this, look at Julian Edelman. Edelman was not actually projected to lead the Patriots in targets in 2013. Instead, newly-signed Danny Amendola was the favorite to be the top receiver in New England, and both were projected to take a back seat to Rob Gronkowski. But Amendola struggled with the transition, so when Gronkowski was injured, it was Edelman who stepped into the increased role.
Because that chart is filled with players who “survived” the process of seeing an increased role, we can’t necessarily generalize its conclusions to every receiver who is projected to see his role expand going forward. We can’t say “Player X is projected to see a bigger role, and the data says he’ll probably do just fine with it”. Again, Danny Amendola truly did struggle in adjusting to an increased role. What we can say, however, is that receivers who successfully see their role expand are likely going to be no more or less efficient than they were before the change.
Or, to put it another way, I think the way we conceptualize the impact of tough coverage is all wrong. If the impact of coverage is going to show up in a receiver’s stats, it will show up in their absence. If a receiver can’t beat the coverage he faces, the result isn’t bad targets, it’s no targets at all.
Since DeAndre Hopkins is already relatively proven at the NFL level, I would expect he is more likely to wind up in the “survivors” bucket when all is said and done. As a result, I view any increase in his usage due to Foster’s injury as a net positive. For receivers who are still able to get quarterbacks to look their way in the face of extra defensive attention, there’s no such thing as a bad target.
More articles from Adam HarstadSee all
More articles on: Player ArticlesSee all
3 Lessons Learned After Week 6 - Allen
3 Lessons Learned After Week 5 - Allen
3 Lessons Learned After Week 3 - Allen