Will Fuller Drops the Ball and I Don’t Care

The most common complaint about Houston Texans wide receiver Will Fuller is that he has a problem with drops. Here’s why I think that’s a non-issue. But before I explain why drops are a non-issue for Fuller, let me explain why I think they’re a non-issue in general.

What’s in a Drop?

Almost is worse than not-even-close

Obviously it’s better to catch a pass than drop one. But only receivers that are good enough to get their hands on the ball can drop a pass. If the receiver can’t get to the ball, then it’s not a drop. So drops have the potential to penalize players that are good enough to (a) be targeted and (b) make a play on the ball, even if that play ultimately fails.

No stickum

According to Sporting Charts, the WRs with the most drops in 2015 were: Mike Evans, Ted Ginn, Amari Cooper, Brandon Marshall, Martavis Bryant, Demaryius Thomas, and Julian Edelman. The only one of those that is considered to be a bad WR is Ginn – and he just had a career year. I haven’t seen any articles floating around suggesting you fade Cooper because of his bad hands, and if you had any of those guys on your team last year you were probably pleased with the production you got. Yes, Mike Evans was a bit of a disappointment, but still finished as a WR2 in PPR. Maybe you don’t like Marshall because he’s old, or Bryant because he’s suspended, or Edelman because he’s injured. But I’m guessing that drops don’t factor much into your evaluation of these players, even though they have a “drops problem.”

We rightly don’t consider drops very much because drops don’t appear to be a very sticky statistic. Comparing the drop rates of WRs in year N to year N+1 yields a correlation coefficient of just 0.1. So drops in one year don’t have much correlation at all with drops in the following year. Here’s a quick illustration of that using two of the players just mentioned.

Capture

The graphs show Thomas and Marshall’s drop rates by season, along with the percentage change from the previous year.

Where’s the beef?

But even more importantly than a lack of year-to-year stability, drops don’t seem to have an appreciable impact on a player’s utilization. I analyzed 382 WRs from 2011 through 2015, using data from Pro Football Focus. I took each player’s drop percentage in year N,1 and compared it to the change in targets they received in year N+1.2 If drops mattered, we’d expect to see a negative relationship: as drop percentage increases, targets in the following year should decrease. But that’s not at all what actually happens.

Capture

Drop percentage is on the x-axis, and the following season’s target change is on the y-axis. The trendline is essentially a flat line. If you’re scoring at home, the correlation between drop rate and target change is -0.023.

 

Are drops just a measure of usage?

So drop rate isn’t consistent from year to year, and doesn’t impact a player’s utilization. So what are drops measuring? Raw usage. Here’s the relationship between drops and targets.3

Capture

It looks like drops are just a reflection of usage, not a predictor of usage or production.

Will Fuller

Full disclosure: I don’t have drops data for college players. So I’m going to assume that college football works like the NFL, and that drops are primarily a reflection of usage (targets and location thereof).

Fuller was a very heavily targeted player – 28th in the nation over the past two seasons.4 So right off the bat we can assume that he’s likely to have more drops than usual. Therefore, we shouldn’t be concerned about those drops.

Instead, let’s focus on how Fuller was utilized, and see how he compares to similar prospects. We can roughly approximate how a player was utilized by using yards per target and yards per reception. Let’s look at players within a half-yard either way of Fuller’s numbers (11.2 YPT and 17 YPR) and 90 or more targets over the past two years. The table is sorted by targets; Fuller is fourth on the list.

PLAYERTARRECYDSCATCH RTYPTYPR
Roger Lewis269158263758.7%9.816.7
Corey Davis255169285466.3%11.216.9
Josh Doctson224145234964.7%10.516.2
Will Fuller210138235265.7%11.217.0
Thomas Sperbeck210140231466.7%11.016.5
Sterling Shepard197137225869.5%11.516.5
KD Cannon186109191158.6%10.317.5
Taywan Taylor172131223476.2%13.017.1
Josh Reynolds163105179264.4%11.017.1
JuJu Smith-Schuster13589145465.9%10.816.3
Jalen Robinette13169144752.7%11.021.0
Mekale McKay13071123254.6%9.517.4
Simmie Cobbs, Jr.12967114951.9%8.917.1
Thomas Duarte12681141264.3%11.217.4
Nicholas Norris12685134267.5%10.715.8
Chaz Anderson12463103750.8%8.416.5
Jay Lee12280141165.6%11.617.6
Mack Hollins11865135855.1%11.520.9
Rodney Adams11367113959.3%10.117.0
Sam Martin1125186345.5%7.716.9
Dijon Paschal11166120659.5%10.918.3
Devin Lauderdale11174122266.7%11.016.5
Malachi Jones10961104856.0%9.617.2
Tony Lippett10565119861.9%11.418.4
Brandon Sheperd10562107559.0%10.217.3
Penny Hart10571109967.6%10.515.5
Ricky Jones1035796455.3%9.416.9
Drew Morgan9673102476.0%10.714.0
Cam Worthy9555101657.9%10.718.5
Mike Williams9559105062.1%11.117.8
Breshad Perriman9450104453.2%11.120.9
Trevor Davis9464107168.1%11.416.7
Antonio Vaughan9463101967.0%10.816.2
Garrett Brown9063109270.0%12.117.3

As you can see, Fuller’s numbers compare very favorably to Josh Doctson, Roger Lewis, and Sterling Shepard among current NFL rookies. Other favorable comps include Corey Davis, JuJu Smith-Schuster, Mike Williams, and Breshad Perriman. 

If we included 2013, we’d find this remarkably similar and highly encouraging comp:

Player Targets Catches Yards CatchRate YdsPerTarget YdsPerCatch
Will Fuller 95 62 1258 65.3% 13.2 20.3
Odell Beckham Jr. 90 59 1152 65.6% 12.8 19.5

Conclusion

Just to be clear, I’m not saying Fuller will be as good as Odell Beckham, or any of the receivers in the previous table. I’m saying that focusing on drops, without contextualizing based on usage and location, adds more noise than value to our analysis. What matters isn’t what a player doesn’t do (e.g. doesn’t catch some passes). What matters is what a player does do. And in Fuller’s case, he catches the ball very well relative to other receivers with similar yards per target and yards per reception. And that, in turn, means he’s a very productive, but shockingly undervalued, prospect.

  1. Percent of targets scored as a dropped pass.  (back)
  2. So if the WR got 100 targets in year N, and 110 in year N+1, the change was +10.  (back)
  3. Total targets and drops per player from 2011 – 2015. Comparing drops to targets on a year-to-year basis still yields an R-squared of 0.6  (back)
  4. That doesn’t sound that high, until you realize he ranks 28th out of 1,532 WRs in the data set over that time frame.  (back)