Are the Cowboys really a passing team?
Their run-heavy days are behind them, but the offense's identity depends on situation.
It’s no secret the Dallas Cowboys have moved away from their run-heavy reputation as they’ve transitioned to a retooled offense under offensive coordinator Kellen Moore.
In recent years, the Cowboys have been among the teams most committed to establishing the run early—starting in the 2014 season and reaffirmed in the first few years of the Dak-Zeke era.
When the team moved on from Moore’s predecessor Scott Linehan, we knew they were likely to take the offense in a new direction.
Some have argued that the team’s struggles to put together a consistent season have come because they abandoned the run while trailing and embraced a pass-happy approach. There are several problems with this framing, but I will save my thoughts on this for another time.
The best evidence we have suggests the Cowboys’ shift toward the pass in most situations has been real, but far from radical. One of Moore’s defining traits in his first season as a playcaller has been a commitment to a particular brand of balance—not to the pass, per se.
What makes a team pass-heavy?
Too often, narratives of team identity are driven by raw counting stats like total passes, or simplified rate versions like passes per game.
Unfortunately, the raw totals don’t give us enough information to work with. In fact, without accounting for why teams pass in some situations more than others, we can’t even trust raw per-play tendencies as a true indicator of a team’s commitment to the pass.
Here are three of the main factors that complicate a team’s run-pass ratio.
Down and distance—Team A runs the ball on 3rd and 1. Team B passes the ball on 3rd and 14. Which team is more committed to the run? We couldn’t possibly know by looking only at whether each play was a run or a pass. Accounting for down and distance is essential if our goal is to quantify a playcaller’s true tendencies.
Field position—How close you are to the goal line matters. Passing is better than rushing in most parts of the field, but this advantage starts to erode in the red zone.
Game state—Teams that face steep deficits are more likely to pass as they scramble to catch up. Teams with a big lead, on the other hand, tend to run more in an effort to burn clock. The end result is that raw team pass rates are skewed heavily by the game score and time remaining.
So how should we account for these factors?
I usually begin by removing 3rd and 4th down, when coaches’ decisions are largely driven by the down and distance. Then, using nflscrapR’s win-probability model, I filter out all situations where the team on offense has less than a 25% chance to win, and is therefore inclined to pass. Finally, I take out plays where the team with the ball is more than 75% likely to win.
This type of filtering has become a standard way of analyzing teams’ true passing tendencies in the public NFL analytics community. It’s simple, intuitive and appears to get us most of the way to a clear picture of team tendencies.
But I’ve been intrigued for some time by the even more precise approach taken by Sean Slavin, who runs the excellent @CardinalsViz account on Twitter.
Instead of throwing out all the situations that most skew team pass rates, Slavin built a model to estimate how often a team should be expected to pass based on down, distance, field position and game state.
Then, by comparing a team’s actual pass rate to their expected rate, Slavin can get an even more comprehensive view of passing tendencies across all situations—including 3rd down and in games with lopsided scores.
(Numbers below were current as of Dec. 13, which is good enough for our purposes.)
Cowboys’ true identity
This is where the Cowboys’ general tendencies become clear. Under Kellen Moore, the offense has been more willing to pass than they have been in previous seasons. However, it still remains committed to running the ball about as often as the average team would under the same circumstances.
Perhaps the most encouraging improvement under Moore has been the Cowboys’ newfound willingness on 2nd down to attack through the air—and beyond the sticks.
Through his 2nd-down playcalling, Moore seems to have a healthy respect for the importance of avoiding third down altogether, rather than playing for the suspect concept of third-and-manageable. (I’ll go into more detail on this in a future piece.)
Here’s the same chart as above, but for 2nd down only:
You can see that the Cowboys are among the more pass-friendly teams on 2nd down, a development that has played a significant role in their offensive successes this year.
But confusingly, Moore has yet to treat 1st down with the same sense of urgency—even when trailing by a score or two.
The Cowboys lean moderately toward the run in these situations by NFL standards. It’s not until the team drops below a 10% chance to win that we really see the Cowboys pass offense let loose on 1st and 10. By then, it has generally been too late to make much of a difference.
Some final thoughts
This offense, to my eyes at least, appears to have fundamentally different priorities in mind for first and second down. This raises a few questions for me.
For one thing, being consistently more run-prone in one situation than another makes the offense more predictable, not less. Defenses know they’re more likely to face a handoff on 1st down than they are on 2nd down, when they could theoretically key in a bit more on the pass.
The other issue is that the run-heavy approach on 1st down isn’t consistent with Moore’s apparent goal of avoiding 3rd downs with his aggressive 2nd-down play selection. It’s almost like the offense is trying to use the run on 1st down to “set up” an aggressive pass beyond the sticks on second down.
But NFL data researchers have found time and time again that 1st down rushes result in worse looks on 2nd down than passes do, even after accounting for the risk of sacks, interceptions and incompletions.
It’s tough to say how much of these tendencies are a direct result of Moore’s priorities as a coordinator, as opposed to Jason Garrett’s preferences as head coach and ultimate authority over the offense and defense. In the end, from a purely offensive perspective, I think Cowboys fans should be mostly encouraged by the progress the team has made in the past year.
But there’s still an opportunity for the Cowboys to employ a more prolific pass offense in the future. And that would be a scary prospect for NFL defenses.