Kirk Cousins’ Best Route

Any time a new offensive coordinator takes over, one of the first things they say is that they’ll “continue to do the things that worked well last year,” while promising to add their own variations to spice up the offense. The nebulousness of that phrasing aside, it forces us to ask a difficult question: what worked?

For the Vikings, the answer is invariably “play-action,” a topic we’ve covered in detail and something that Kirk Cousins has been uniquely good at over the years. All that does, however, is give us a framework for plays — not the plays themselves. For an accounting of which throws work best for Cousins — play-action or not — we’ll have to dig a little deeper. Fortunately, NFL’s Next Gen Stats and Sportradar have given us route-by-route breakdowns for each quarterback and receiver in the NFL.

Before we take a look at that, we should quickly look at the basic route tree for a receiver — some of the most common routes found in NFL playbooks:



Even a basic route tree looks complicated, and the 18 routes up there do not fully account for what receivers end up getting asked to do over the course of the year. Compounding that, there are multiple names for each of the routes above, adding a layer of confusion to a lot of football analysis. Luckily, there is a tendency to run a set group of routes more than any other, and the most common routes are the go, post, corner, dig, out, slant, curl, comeback, cross, drag, quick out and smoke.



Next Gen Stats has been able to track those routes, and that can give us insight into what receivers and quarterbacks work with best. This kind of analysis parses pass attempts into 12 different categories, diluting their statistical power, but it can give us a starting point for analysis.

There are a lot of limits to this type of breakdown — good quarterbacks know their limits and the structure of the defense, so they will frequently choose the best routes for themselves based on their strengths and weaknesses. For example, if a quarterback decides to throw a crossing route instead of a corner route on a consistent basis, that might mean he’s better at throwing the crossing route even if the success rate for the corner route is higher; he might prefer the crossing route by default and only throw the corner route when it’s relatively easy.

While quarterbacks will generally throw what is “open,” they could have different standards for what they perceive to be open based on any number of idiosyncrasies in the route concept.

Different receiver talent distributions and offensive designs will have a big impact as well. Some teams will have receivers run go routes as “clear-out” routes to create space underneath, while other offenses may choose to engage in play-action more often to further distort the coverage geometry. Each offensive tendency changes the progression and defensive spacing in different ways. For example, if a passer ends up doing a better job throwing underneath when the offense enables those routes in unique ways, that doesn’t necessarily tell us that he’ll be better at that particular route in another system.

For how much route success is on the quarterback and how much is a product of the offense or randomness, the truth is likely somewhere in the middle. The throwing motion, arm strength, familiarity with the route concept and ability to process the progression all also impact route success — and those are more innate to the quarterback than the offensive design is.

We can goose things a little bit in our favor, however. Next Gen Stats data goes back to 2016, and Cousins has been with three different offensive coordinators and has had eight different starting receivers in that time. That should tell us with some degree of authority if this route data represents his “true” ability with these routes, especially because it functionally triples the sample data while accounting for the potential pitfalls.

In order to reduce the impact of individual big plays skewing the data, we’ll look at relative first-down rate instead of pure yards-per-attempt. The reasoning is simple — if one quarterback was successful on a route only three times on 13 attempts for 90 yards each, he would have a 20.8 yards per attempt on that route, the best in the league. But, on the other hand, he would only have three first downs in 13, a paltry 23.1-percent success rate — one of the lowest in the league.

More likely than not, this example quarterback was blessed with a long field on his successful attempts but is generally fairly poor at those deep routes. First-down rate tells us a bit more about how generally good a player is for those individual plays. Still, we’ll also look at yards per attempt as well just to see if there’s something there we may be missing.



When ignoring routes that seem to have little to do with quarterback talent — flat routes and wide receiver screens — the clearest example of a positive outlier for Cousins is the corner route, also called a flag or seven route. For the past three years, his first down rate on that route has been higher than the NFL average, and his yards per attempt on those routes consistently beats most NFL quarterbacks.

Abstractly, being the best at throwing a particular type of route doesn’t mean that the quarterback should throw it at a higher rate. Cousins happens to crush the NFL average on first downs for comeback routes, but if he threw it more often, he would lower his overall yards per attempt. Luckily, the corner route is one of his most successful routes compared to other routes on the tree — not just compared to other quarterbacks on the same route. He creates first downs on 45 percent of those throws and averages 11.8 yards per attempt on them.

The obscenely high yards-per-attempt mark is likely a product of small samples and the high amounts of variance that yardage stats often have to deal with, but it’s notable that he averaged over 10.2 yards per attempt in each of the last three seasons despite the NFL average of 8.7 yards per attempt.