Russell Wilson, Brian Schottenheimer, and Seahawks Completion Percentage Above Expectation

Over the past two years, many fans have pointed to analytics to argue the Seahawks offensive attack forces Wilson into unnecessary difficult throws. But digging deeper into the data, is that really the case? Analyst Matty Brown investigates the analytical conundrum.
Author:
Publish date:

When the analytics on the Seahawks’ 2019 offense were published, many were quick to pile on offensive coordinator Brian Schottenheimer. Maligned by much of the team's fanbase, the general perception surrounding Schotty’s attack is that it lacks creativity and regularly forces the star quarterback, Russell Wilson, into difficult throws.

The discussion surrounding the Pete Carroll-influenced run-pass ratio has been exhausted and can be had another day - including 2019's increased passing rate on second down compared to 2018. Where fresh Schottenheimer criticism arrived was via a combination of two metrics: Completion Percentage Above Expectation and Expected Completion Percentage.

Completion Percentage Above Expectation, or CPAE for short, takes the difference between a quarterback’s ‘Expected Completion Percentage’ and ‘Actual Completion Percentage.' It produces a '+ or -', percentage figure for each passer.

Actual Completion Percentage is what you see in your back-of-the newspaper, first take, vanilla box scores. For instance: Russell Wilson completed 61.3 percent of his passes in 2017. Or Alex McGough had a 75 percent Actual Completion Percentage in Week 2 of the 2018 Preseason.

Expected Completion Percentage, meanwhile, is defined by NFL Next Gen Stats as: “an indication of the level of difficulty of a quarterback's throw.” NGS explains that “the level of difficulty of a specific throw is determined by in-play features specific to each play.” Such “features” include the distance of the throw, the pressure the passer is under, and receiver separation. Expected Completion Percentage is abbreviated as xCOMP%.

Ben Baldwin, Ethan Douglas, and Sheil Kapadia praised Completion Percentage Above Expectation in The Athletic’s excellent analytics glossary:

“This is a better measure of accuracy than traditional completion percentage because it takes into account where passes are being thrown.” This enables us to contextualize a quarterback’s box score. We are also able to gain some idea of the offensive play calling before grinding the tape.

In Next Gen Stats’ tracking of Completion Percentage Above Expectation, Russell Wilson finished 2019 fourth out of 39 total quarterbacks at +4.8 percent. Meanwhile, NGS had Wilson’s xCOMP% on throws attempted at 61.3%. This was 37th out of the 39 quarterbacks.

As Sheil Kapadia explained in his ‘A football nerd’s guide to the Seattle Seahawks’ 2020 season,’ the low Wilson xCOMP% “suggests that the scheme wasn’t producing a lot of high-percentage throws.” This was the damning evidence for Schottenheimer critics.

I don’t think Schotty is trying to produce conventionally high-percentage throws either. But why would he? Wilson’s main strength as a passer is that he throws an extremely sexy deep ball. Since 2016, the 31-year old leads the league in a multitude of deep passing categories (attempts of 20+ yards):

We can get into a long debate about the merits of efficiency versus volume. And we can question what the ideal balance or trade-off is in that relationship. The heralded 2015 stretch of offense where Wilson enjoyed quick game elements from a shotgun attack will always be the reference point for fans wanting easier throws for Wilson.

A number of complicating factors exist, especially with the make-up of the offensive line and Carroll’s guiding philosophy. A lot of the same quick-hitting concepts still exist in the Schotty attack though. They've even been refined to a less erratic, more concentrated group of reliability. Furthermore, while we don't have Wilson's xCOMP% for 2015, it’s interesting that his 2015 CPAE was higher than 2019 with a league-leading +8.1 percent:

Ultimately, the design of Schotty’s passing attack so far has reflected Wilson’s main strengths. Wilson’s Intended Air Yards (IAY) tied for fourth-highest in the NFL at 9.4 in 2019, per NGS. (Detroit’s Matthew Stafford led the league with a gaudy 10.7 IAY).

Schottenheimer finds the mismatch for Wilson and Wilson hits it, often deep. The play caller understands the available talent, then aligns and assigns it suitably. Schotty gets coverage indicators for Wilson via pre-snap work and Wilson understands which one-on-one to attack on each concept. And the quarterback has been pretty damn good at it throughout his career.

Every coach in the nation preaches the mantra of “don’t get beat deep” to their defensive backs. Defenses are focused on stopping this play. Meanwhile, in the greater time it takes for a pass to arrive the deeper it is thrown, a defender is naturally able to better recover to the catch point. @friscojosh’s sorting of 2019 NFL play action passes into four buckets of “openness” supports this. It shows that receiver separation on play action became less and less common the further downfield the ball was thrown. 

Deep passing is supposed to be tough. Wilson defies convention. He thrives in hitting passing concepts that spawn these difficult downfield throws and is incredibly accurate when throwing deep. He's one of the best we have seen. Even with less receiver separation, partially a result of downfield concepts and low xCOMP%, Wilson’s 2019 numbers were highly efficient. Just look at his fourth-highest CPAE.

This isn’t a new argument - it was made well by @cmikesspinmove, writer at Beastpode and podcaster at SeattleOverload on Twitter back in December 2019 - however, the line of thinking certainly challenges the conventional narrative. That of Wilson being put in terrible situations and succeeding despite the circumstances; the easy, lazy, group-think conclusion of those not wanting to consider or see the full picture.

“It's wildly impressive Wilson is as efficient as he is throwing the passes that he does, and it's not scheme that is making those "easier" for him. He just happens to be very good at them, better than the anticipatory throws that often lead to "wide open throws" so the scheme will always follow from that. He's a paradox of a quarterback. To make it work though, you have to have the targets at receiver. Give him an elite or near elite receiving option and their connection will have all-time efficiency numbers. Whereas an elite receiver with the next quarterback might not.” Griffin (@cmikesspinmove)

It’s why Tyler Lockett has managed to establish an integral role in the offense. He isn’t the burner he once was, but he has become a master at catching the difficult deep receptions. In 2019, Lockett had a catch rate of 80.5 percent, yet posted an expected catch rate of 67.8 percent according to NGS. This was a difference of +12.7, which NGS had as second-highest in the league behind New Orleans’ Michael Thomas.

Lockett is a low-percentage baller - as I explained, with video, in a July article. The 27-year old has become a downfield possession target despite being just 5-foot-10. He can catch the inefficient routes that are weirdly high percentage for Wilson. The stuff Wilson prefers to hit, and actually can hit, as opposed to routes that are more difficult due to his shortness. (I’ve seen the heat maps; Wilson is very hesitant to throw short-to-intermediate over the middle against anything other than man coverage - scrambles and improvisation excluded)

Video 3 Lockett

Remember when the Seahawks’ offense started to struggle last season following their bye in Week 12? Was it because teams caught up to the passing design? Or because Wilson was playing at a lower level. Or because Lockett was hurt? Really, it was a combination of all three. Seattle, though, did not stop trying to throw the ball deep. Their overall offensive approach didn’t change. Wilson just stopped completing said deep balls. See the post-Week 14 work of @DeryckG_:

In the past, we have grown willing to accept that Wilson’s play style is at least partially responsible for higher pressure and sack rates.

“From time to time, Russ will get back there and he’ll be hunting, and he’ll end up holding it a little bit too long,” was part of Schottenheimer’s August 25 answer on pressure rate being the worst-allowed in 2019. “Those numbers can be very misleading at times.”

Now, ahead of the 2020 season, we should also open our minds to Wilson’s play style being a major factor behind the Completion Percentage Above Expectation and Expected Completion Percentage figures. Rather than using these metrics as automatic ammunition to attack Schottenheimer, consider that the offensive design is catered to Wilson’s strengths, comforts, and preferences. And it always has been.