The PFF QB Annual is a yearly publication that breaks down each quarterback’s season in excruciating detail. It’s worth the PFF subscription on its own, even if you don’t like PFF’s grades. In it, they break down stats by down, time to throw, pressure, and much, much more. Their conclusion on Kirk Cousins is a familiar one:
[Cousins] has clean footwork, is accurate and generally understands opposing defenses but can’t operate with anything less than immaculate surroundings. One of these years it’ll happen, we promise.
It’s an agreeable conclusion on its face and a mostly foregone conclusion with Cousins by now. But there’s a lot to unpack about a relatively simple statement. The insinuation is that the difference between a good supporting cast and a bad one is uniquely large with Cousins. But we can test that, and further explore it, using only the data in the QB annual.
My personal favorite part of the QB Annual is the accuracy charting. PFF uses different buckets to define accurate or inaccurate passes. They also make plenty of exceptions depending on defender positioning, the route and play design, and of course, drops and receiver mistakes. The middle bubble is the most important one, exhibiting the frequency of accurate passes.
2019 was a better year for Cousins, and his 2018 or 2020 performances are closer to his career average. That average is nothing to sneeze at, either. Generally, Cousins’ accuracy and raw throw-to-throw ability have never been at issue. While his 2020 wasn’t quite as accurate as his 2019, that dark green circle is still plenty to be happy about.
Receivers
Compared to previous years, his receiving help has been about the same too. He threw about 55% of his passes to open receivers in 2020, defined as players with more than a step of separation. That almost exactly matches 2018 and 2019. You could say the effect Justin Jefferson has replaces the effect Stefon Diggs had, at least in terms of how much separation they give Cousins to work with. Another huge facet is the Kubiak scheme, which hasn’t changed from 2019 and likely won’t in 2021.
Some quarterbacks need wide-open players to function. Some can turn a little separation into an easy catch. Cousins was accurate about 2.8% more often than average on those “open” throws. He was open about 8% above average on throws where the receiver had just a step on his defender. He threw accurately 6.8% more often than average when a defender was “closing,” like on a comeback or snag route. He was only slightly below average on “tight” passes, but those only comprised 19% of his attempts.
All that suggests that Cousins won’t waste the separation his receivers create and maximizes middling separation. Were Cousins to have lesser passing targets, it’s hard to see him falling apart any more than anyone else would. But that’s not the heart of the surroundings argument, is it?
Scheme
One common critique of Cousins is how unwilling he is to color outside the lines. Per PFF, he only resorted to a scramble drill on 3% of his passes. Those turned out pretty well but were much less frequent than the 5% league average. The QB Annual has this heat map of routes vs targets. Essentially, if you see blue in the left map and red in that same spot in the right map, that means Cousins prefers those throws more than the Vikings planned for.
As you can see, there isn’t much deviation. That tracks with our preconceived notions. Perhaps the biggest difference is how often Cousins would check down. Maybe he favors the left a little bit, but not so much that it’s particularly instructive. That should get you thinking about the real issue at hand here.
Offensive Line
Cousins has had absolutely putrid offensive lines in the last four years. Even dating back to Washington, he has always dealt with a lot of pressure. The Vikings haven’t helped this, and that’s the main argument for PFF’s assessment. No offensive line? Kirk Cousins can’t be good enough. So let’s truly test that.
When kept clean, Cousins’ dropbacks generated 0.38 EPA/play, which ranked sixth leaguewide. Under pressure, they fell to -0.32 EPA/play, ranking 11th. So Kirk Cousins loses 0.70 EPA/play on average due to pressure. That’s the 13th most among 35 qualifying quarterbacks. Put another way, Cousins is the 22nd best pressure eraser in the league. It’s certainly not a strength. But what if the Vikings had an average offensive line for the first time since the Ponder era?
Cousins had 221 of his dropbacks plagued by pressure. If you wanted to simulate a median offensive line performance, you’d overturn about 40 of those. Let’s say instead, only 181 of Kirk’s dropbacks came under pressure, around 32%. That would increase his EPA/play from 0.11 to 0.16. The Vikings would go from 11th in EPA per dropback to seventh. You could say the difference between the Vikings’ offensive line and an average offensive line amounts to about a point or two per game. That would make a pretty big impact, considering four of their nine losses came by a score or less, and two came by one point.
Now let’s say Cousins was average against pressure but suffered the same amount of it. The median pressure eraser in the league was Aaron Rodgers, who suffered a 0.66 EPA/play loss on pressured dropbacks. That’s not too far off from Cousins, but the impact would still be measurable. We can do the same hypothetical there, shown in the table below.
“Average” is an uninspiring goal. There’s room between that and “immaculate,” so let’s just do the top quarter of the league. If we gave the Vikings a 75th percentile offensive line and made Cousins a 75th percentile pressure eraser, where would the offense end up?
Fix the pressure problem and you get a top-10 passing offense. Whether you target Kirk’s half of the problem or the line’s half of the problem is the real question. Keep in mind also that Cousins is responsible for 27 of his own pressures, per PFF. Still, the Vikings’ pressure rate is further from average than Cousins’ ability to handle that pressure. It shouldn’t be a surprise to anyone that it’s more efficient to eliminate pressure than it is to improve at dealing with it. It’s probably cheaper, too.
