Look, I’m not going to shame anyone for doing whatever it takes to get a job. Especially in 2020. Being honest … in this economy? Telling a potential employer whatever they want to hear is good practice and if you need proof of that concept, look no further than Cowboys coach Mike McCarthy.
Before he was hired by the Cowboys, McCarthy was the subject of numerous redemption pieces before the NFL’s annual coaching carousel started up. He did an interview with ESPN, recorded a segment with NFL Network and let Peter King follow him around for a day. The theme was the same for each story: After hearing criticism for letting the Packers offense get stale, McCarthy was using his sabbatical from coaching to modernize his approach.
What did that entail? A trip to Pro Football Focus’ offices in Cincinnati, where he got a crash course in analytics. McCarthy claimed he studied every offensive snap of the 2019 season – but did walk that back during his introductory press conference in Dallas. While McCarthy didn’t totally agree with the critics saying he let his offense in Green Bay get stale, he told King that a self-scout of his offense did reveal some issues:
“We got away from motion and shifts and multiple personnel groups that we used in the past, so you look at the why . . . and quite frankly you apply it to the next opportunity.”
And his study of the NFL’s top offenses revealed there were some tactics he wasn’t tapping into nearly enough in Green Bay.
“As a play-caller, you’ve got to stress the defense,” McCarthy said, “and one of the things watching all these teams has shown us is how good some teams are at challenging the eye discipline of the defense. Makes ‘em think at the snap of the ball, which is huge. This bullet-motion sweep, this jet motion, at different tempos, different speeds. I just really like what it does to a defense. We call those things ‘nuisances’ for the defense.”
Before we look at how McCarthy applied what he had learned, let’s recap what we should have expected to see on Sunday night in Dallas’ season opener:
More varied personnel usage
More pre-snap motion
So what actually happened during the Cowboys’ 20-17 loss to the Rams? Let’s take a look at their usage of each of those tactics…
More varied personnel usage
The Cowboys used 11 personnel (1 RB, 1 TE, 3 WRs) on 83.3% of their snaps. They used 12 personnel (1 RB, 2 TEs, 2 WRs) on all but two of the remaining snaps. In 2018, McCarthy’s last year in Green Bay, the Packers used 11 personnel on 78% of their snaps, and 12 personnel on 17% of snaps.
There was actually less variation in personnel usage! We’re not off to a good start.
More pre-snap motion
According to ESPN’s Seth Walder, the Cowboys used pre-snap motion on only 4.3% of their plays. That ranked 26th in Week 1.
Maybe next week!
The Cowboys did have a decent amount of snaps without a huddle, but … almost all of them came in end-of-half situations. According to the NFL’s play-by-play logs, there were only three snaps of no-huddle in more neutral situations.
So that’s a 0% success rate. According to the analytics, that’s not good!
Of course, it’s possible to design a good offense without leaning heavily on those tactics. Play design is probably more important than all that, for instance. But based on what we saw on the broadcast of the game — the NFL has yet to release the coaches film from Week 1, so that’s all we have to go off — McCarthy’s offense did not make Dak Prescott’s job any easier. It felt like Prescott had to scratch and claw just to find an open receiver anywhere on the field, and, more often than not, that receiver was not very open or very far downfield.
According to Next Gen Stats, Prescott’s average attempt traveled 7.1 yards downfield, and he had an Expected Completion Percentage of 63.2%. Neither figure is very high, which is not a great sign considering that Expected Completion Percentage tends to go UP when a quarterback’s average depth of throw goes DOWN. Based on last year’s numbers, a quarterback’s aDOT (average depth of target) explained about 50% of his Expected Completion Percentage. Using the 2019 data, we can figure out what we should expect a quarterback’s xComp% to be based on the average depth of his throws. Here’s what the relationship looks like, in scatterplot form:
According to those results, a quarterback throwing the ball 7.1 yards downfield should have an Expected Completion Percentage of 65.6%. The difference between that number and his actual Expected Completion Percentage was -2.4%, which was the sixth-biggest difference of Week 1.
So that gap could be explained in a few different ways, right?
Maybe the receivers aren’t getting separation, maybe the quarterback isn’t seeing the field well and missing open receivers downfield or maybe the coaching isn’t doing a good job of scheming his receivers open (which was a big criticism of McCarthy in Green Bay).
The quarterbacks who fell below Dak in this metric are either notoriously bad at reading the field (Mitch Trubisky and Tyrod Taylor), were playing against suffocating secondaries (Ryan Fitzpatrick), have underwhelming talent at receiver (Aaron Rodgers), or all of the above (Daniel Jones).
But none of those things apply to the Cowboys offense: The receiving corps is one of the best in the league and Prescott doesn’t routinely miss open receivers downfield. So that leaves one obvious explanation: McCarthy still isn’t doing a good job of getting his very good receivers open.
There were some things McCarthy did well on Sunday. The Cowboys run-pass split on early downs was heavy on the pass (that’s good!). And he was aggressive on fourth down when it would have been easy to settle for a field goal (despite the result, also good!). But McCarthy had been doing those things during his time in Green Bay, so we can’t really point to that as a sign of growth. And I should point out that Dallas used play-action at a below-average rate, which the nerds surely didn’t like.
It’s still extremely early but it’s looking like McCarthy didn’t learn much during his year away from the NFL. Outside of using the media to get a job, that is.