How Incomplete Passes to Sammy Watkins Are Holding Back the Buffalo Bills' Offense

There's no such thing as a positive incomplete pass, but some teams throw their incomplete passes better than others.

Back in May, during the most boring part of the NFL calendar, I wrote an article discussing the value of an incomplete pass, and considered the damage teams were doing to their expected point outcomes by force-feeding their top receiver the football.

You can check out that article here, if you'd like, but the takeaway was as follows: Teams with bad quarterbacks shouldn't force one receiver the ball, and teams with strong deep-threat receivers generally shouldn't force-feed their best receiver the ball, either.

Calvin Johnson, Andre Johnson and A.J. Green were the primary examples of receivers who were "hurting" their teams by being targeted so heavily, and negative outcomes seemed to happen more often when those receivers were being targeted.

And when you think about it, that makes perfect sense. A team will almost always put their best cover corner or provide safety help on those receivers, meaning that a mistake from the quarterback is more likely to result in a turnover. Those players are also more likely to be targeted in a key situation, like a 3rd-and-4, a play that would sharply swing the expected points for a team if it doesn't result in a first down.

So, with three months worth of the 2014 season under our belts, have we learned anything from last season? Yes and no. Let's take a look at some of the players whose 2014 incompletion data stands out through 12 games.

Here is a chart of the 20 receivers who have seen between 90 and 110 targets this season, along with their Incompletion Net Expected Points (NEP lost on incomplete passes targeted in their direction) and the per-incompletion ratio of NEP loss.

NameReceptionsTargetsCatch RateInc NEPPer Inc NEP
Randall Cobb659072.22%-20.05-0.80
Anquan Boldin6810068.00%-29.75-0.93
T.Y. Hilton6810167.33%-29.81-0.90
James Jones609066.67%-30.02-1.00
Mike Wallace549855.10%-32.05-0.73
Steve Smith549457.45%-32.18-0.80
Mohamed Sanu529057.78%-33.93-0.89
Mike Evans539158.24%-34.63-0.91
Alshon Jeffery6710762.62%-35.42-0.89
Andrew Hawkins529057.78%-35.90-0.94
Calvin Johnson499054.44%-37.42-0.91
DeAndre Hopkins609066.67%-41.31-1.38
Reggie Wayne549556.84%-44.27-1.08
Kelvin Benjamin5711051.82%-46.56-0.88
Brandon Marshall5810256.86%-46.70-1.06
Dez Bryant6711060.91%-48.18-1.12
Rueben Randle5610155.45%-48.29-1.07
Sammy Watkins519852.04%-52.27-1.11
Keenan Allen7211065.45%-52.38-1.38
Vincent Jackson5010746.73%-59.63-1.05

The Pressure of Two First-Round Picks

Sammy Watkins has been targeted 98 times this season, and those targets have resulted in 51 receptions. Those 51 receptions have provided the Bills with plenty of big plays and key first downs, as Watkins' impressive 92.16% Success Rate would imply. Success Rate is a measure of how often a player's actions (in this case, receptions) result in a positive Net Expected Points outcome for his team.

But what about those 47 incompletions? They've been as damaging to the Buffalo offense as his receptions have been positive. Among the 20 receivers in the chart above, Watkins ranks in the bottom third in "Incompletion NEP" and NEP lost per incomplete pass targeted at him.

His per-incompletion NEP loss is similar to that of Jimmy Graham a year ago, who was one of the players near the top of this dubious metric in 2013. But like Graham last year, there's good reason for the Bills to want to get the ball to Watkins.

There are also bad reasons to try to get the Clemson product the ball as often as possible, including the two first-round picks surrendered for the right to draft him. So should Buffalo be targeting Watkins as heavily as they have so far this season?

That's a tough call to make. Buffalo's top receiver suffers from the "bad quarterback" issue and the "deep threat" issue, but there are smarter ways to get him involved than the Bills are currently using. Chris Trapasso, an NFL contributor to various websites and a Bills fan and writer, was either being facetious or alarmingly honest when he exclaimed on Twitter this past Sunday that the Bills threw only their second screen to Watkins this season.

Whether Chris is literally reporting the second such pass to Watkins, or jokingly jabbing the team for not throwing enough screens to Watkins, the sentiment remains the same. Sammy is seeing the top cover men in the league on a regular basis, but could still be used effectively in the screen game if the Bills drew up those plays. He ran screen plays at Clemson so often that it negatively impacted his perception as a well-rounded NFL receiver coming out of college, because NFL receivers aren't used as a screen option nearly as often as Watkins was with the Tigers.

Which is why it's startling to see him used downfield so often, especially in an offense with such poor quarterback play. Neither Kyle Orton nor E.J. Manuel rank among the top-20 quarterbacks in Passing NEP per attempt (among those with over 100 drop backs), and Buffalo is in the middle of the pack in our opponent-adjusted team passing metrics. Receivers who typically see shorter targets also have less penalizing per-incompletion metrics, and Watkins is seemingly custom-made for those short, quick plays.

Watkins clearly does big things once he has the football, which is why the Bills continue to try to get him involved. But the way he's used could use some tweaking, and some extra targets for his less exciting counterparts Robert Woods and Chris Hogan could help, as well. Neither has as harsh of a per-incompletion NEP loss, but neither is as productive of a receiver on a per-reception basis, either.

Be Like Mike

So how do the Bills get Watkins involved in a less risky fashion? Be more like the Dolphins.

Miami's top receiver Mike Wallace is once again among the "best" in terms of NEP lost on incompletions thrown in his direction. His 2013 data stood out as one of the league's most efficient recipients of incomplete throws, and this season he's on a similar pace. Among the group listed above, his per-incompletion loss of NEP is the best by a decent margin.

Wallace's overall production this season isn't as impressive as Watkins', as his Reception NEP and Success Rate are both lower than the rookie's numbers. But with both receivers sitting at 98 targets, incompletions thrown at Wallace have resulted in over 20 Net Expected Points worth of improved results versus Watkins. Why is that?

Both players see a similar share of their team's passing offense, and neither team is particularly bad about throwing interceptions. Both rank in the mid-teens in our passing offense metrics, and have quarterbacks who hover near the middle of our player rankings at the position. So with everything else nearly identical, we can safely assume that the quality of the passes thrown to Wallace are better than those thrown to Watkins.

Throwing the football in the NFL is a risk/reward situation, and the difference between Mike Wallace and Sammy Watkins reveals how teams differ in their approach to getting their best receivers involved. The Bills are taking more risks by throwing the ball to Watkins in more dangerous situations, and he's produced better positive metrics while also being among the worst when considering how incomplete targets to him impact the offense.

Wallace, on the other hand, has slightly worse production but with a much better average non-catch "result." Is the risk worth the reward? Not in every case, and especially not in this one.

There are better ways to use Sammy Watkins than the way the Bills are currently deploying their first-round pick, which is why his low catch rate and extreme NEP loss per incomplete target numbers are frustrating. Watkins is an accomplished catcher of the football and has run more screens in his life than most 15-year NFL veteran receivers. Getting him involved more in the short game, on easier throws, would possibly help vault the Buffalo offense into the upper echelon in the NFL, despite their lackluster quarterback options.