Stat Analysis

Advanced analytics on player and team performance

Play-Action Offense 2018

Sean McVay
Photo: USA Today Sports Images

by Scott Spratt

The NFL's play-action rate spiked in 2017 to 22 percent -- the highest that it had been in recent years, but not dramatically more than the range from 17 to 21 percent we had seen in the previous decade. However, the 2018 jump to 24 percent makes it clear that play-action increases are a trend, and one not isolated to a handful of newer, play-action-friendly coaches like Sean McVay and Doug Pederson. All but five teams reached the 20-percent benchmark of play-action passes in 2018. And while the +11.3% DVOA differential between play-action and traditional passes was down from the exceptional marks of +17.7% and +14.3% seen in 2016 and 2017, it was typical for most of the last decade and suggests play-action rates could continue to rise.

Play-Action Usage, Offenses, 2018
      With PA
(Pass/Scram)
With PA
(Pass Only)
No PA Difference
Off PA Pct Rank DVOA Rank DVOA Rank DVOA Rank DVOA Rank
LAR 36% 1 56.0% 3 57.4% 3 20.3% 9 +35.7% 6
SEA 33% 2 54.4% 5 50.6% 5 19.6% 11 +34.8% 7
PHI 32% 3 29.1% 12 29.5% 11 14.9% 19 +14.2% 14
NE 31% 4 76.4% 1 76.4% 1 16.4% 16 +59.9% 1
CAR 29% 5 6.0% 24 5.0% 23 15.2% 18 -9.2% 24
TEN 29% 6 33.4% 9 27.2% 12 -12.1% 28 +45.6% 4
KC 29% 7 43.4% 6 43.0% 6 74.8% 1 -31.4% 32
BAL 28% 8 13.3% 18 12.0% 17 17.7% 13 -4.5% 20
ATL 26% 9 38.5% 8 39.1% 8 27.3% 7 +11.2% 16
NYJ 26% 10 7.0% 23 6.3% 22 -6.0% 26 +13.0% 15
SF 26% 11 18.0% 16 20.9% 16 -0.6% 23 +18.5% 13
CIN 25% 12 -5.2% 29 -5.6% 29 19.5% 12 -24.7% 30
DAL 25% 13 29.9% 11 31.5% 10 -7.5% 27 +37.4% 5
DEN 25% 14 23.7% 15 21.3% 15 -4.7% 25 +28.4% 9
BUF 24% 15 12.6% 19 -2.4% 28 -33.7% 31 +46.3% 3
HOU 24% 16 26.1% 13 25.5% 13 20.1% 10 +6.0% 18
NYG 24% 17 8.7% 22 8.7% 21 17.1% 15 -8.4% 23
CLE 23% 18 5.8% 25 3.3% 24 11.2% 20 -5.5% 21
IND 22% 19 32.2% 10 32.1% 9 21.7% 8 +10.4% 17
LAC 22% 20 63.1% 2 66.0% 2 34.9% 2 +28.2% 10
MIA 22% 21 -19.4% 31 -23.0% 31 7.4% 21 -26.8% 31
CHI 22% 22 10.9% 20 11.6% 18 17.3% 14 -6.4% 22
WAS 21% 23 -9.1% 30 -10.5% 30 -14.1% 30 +5.0% 19
NO 21% 24 55.4% 4 55.3% 4 28.5% 5 +26.9% 11
MIN 21% 25 40.3% 7 40.4% 7 6.4% 22 +33.8% 8
ARI 20% 26 -3.3% 28 -2.3% 27 -54.3% 32 +51.0% 2
GB 20% 27 5.0% 26 2.7% 25 27.6% 6 -22.6% 29
JAX 19% 28 -31.0% 32 -36.3% 32 -13.5% 29 -17.5% 27
DET 18% 29 24.5% 14 24.8% 14 -1.7% 24 +26.1% 12
OAK 18% 30 0.4% 27 0.4% 26 16.1% 17 -15.7% 26
TB 17% 31 15.4% 17 11.1% 19 30.0% 3 -14.6% 25
PIT 12% 32 9.0% 21 9.0% 20 30.0% 4 -21.0% 28

Play-action is McVay's favorite item in the toolbox to make life easier for quarterback Jared Goff. In Goff's rookie season under head coach Jeff Fisher, the Rams used play-action on just 16 percent of their pass attempts, which ranked 26th in football. And while they may not have been outstanding on those play-action passes, the Rams were markedly better with play-action (15.6% DVOA, 21st) than without it (-48.9%, 32nd). McVay identified that split and ratcheted up the team's play-action percentage to 29 percent in 2017 (second-most) and 36 percent last year (first). They finished top-10 in play-action efficiency both seasons and enjoyed the sixth-best DVOA difference between play-action and traditional passes in 2018.

It's possible to read McVay's play-action transformation as a simple embracing of modern football, no different than what any Football Outsiders reader could have done. After all, Goff is far from the only quarterback who maintains that sort of split -- the league-wide DVOA advantage on play-action passes is over +10.0%. But I suspect McVay is more deliberate with his strategic decisions, and my best evidence for that opinion might be the Rams' complete avoidance of run-pass options (RPOs).

Sports Info Solutions charts which play-action passes are and are not RPOs, and their data hub offers a tool to discover team tendencies and effectiveness with those and other strategies. As the "it" football strategy of recent seasons, the RPO is an armchair indicator of which coaches and teams "get it." It's hard to argue with the offensive success of RPO proponents like the Chiefs, Bears, and Eagles. But the Rams have reached their perch of play-action frequency and success while ignoring the RPO fad almost entirely. Just two of their 206 play-action passes in 2018 were run-pass options.

The term "fad" carries a negative connotation that I do not intend here. Traditional play-action passes and RPO passes are similar in that they both fake a run, but the typical NFL version of those plays have two different objectives. Traditional play-action passes buy time for longer routes to develop and can help a team gain chunk yardage. Teams tend to deploy traditional play-action passes on first downs (36 percent of pass attempts) and second downs (27 percent) rather than third downs (4 percent) because the resulting downfield throws have lower chances of completion. Run-pass options create quick separation, stemming typically from a linebacker read, and are reliable plays to move the chains. As such, traditional play-action passes average 8.6 yards per attempt while RPO passes average just 6.8 yards per attempt.

McVay might reasonably have leaned on RPOs to increase the Rams' play-action percentage, and not just to shape his reputation. Goff frequently relied on them for big plays at Cal. But as The Ringer's Robert Mays distilled from his conversation with Princeton head coach Bob Surace, Goff's Cal RPOs often required a read of a safety or cornerback and led him to take 81 sacks in his three-year tenure. Sacks were a major problem for Goff in his freshman NFL season -- his 11.3 percent sack rate was the worst of qualified passers in 2016. McVay has been smart to rely on traditional play-action passes for Goff's attempts at big plays. They have helped cut the Rams' sack rates to 5.0 and 5.6 percent the last two seasons, both lower than average.

There were 11 teams that owed at least 10 percent of their play-action passes to RPO passes in 2018. Andy Reid and his coaching tree -- which includes Matt Nagy in Chicago, Frank Reich in Indianapolis, Ron Rivera in Carolina, John Harbaugh in Baltimore, and Doug Pederson in Philadelphia -- make up about half of that list. This table is courtesy of John Shirley of Sports Info Solutions.

Run-Pass Option Pct of Play-Action Pass Leaders, 2018
    Yards Per Play
Off RPO% RPO Other PA Diff
CHI 24.1% 6.1 6.6 -0.5
PIT 23.3% 3.9 8.2 -4.3
KC 20.0% 7.4 9.6 -2.2
IND 14.9% 5.0 9.9 -4.9
CAR 14.2% 8.3 7.7 +0.6
TB 13.6% 10.4 9.0 +1.4
BAL 11.8% 9.6 8.0 +1.6
GB 11.5% 5.0 7.1 -2.1
WAS 11.2% 7.9 7.4 +0.5
OAK 10.9% 5.5 7.0 -1.5
PHI 10.8% 5.0 9.5 -4.5

Neither Rivera nor Harbaugh were offensive coaches under Reid, and so they may have implemented RPOs because of their specific quarterbacks rather than a general offensive philosophy. Cam Newton was already used to making reads in the Panthers' zone read running game. RPOs are a natural progression, and the Panthers are well-equipped to succeed with them, having speed and quickness at receiver and a quarterback who can run the ball himself. Meanwhile, the Ravens nearly doubled their percentage of RPO passes from 8.9 percent in the first half of the season with Joe Flacco under center to 16.1 percent in the second half with rookie Lamar Jackson at quarterback, and they added more than a yard per RPO pass to threaten the Buccaneers' top spot.

The Panthers and Ravens were two of the four teams on the leaderboard that produced more yards per play with their RPO passes than on their traditional play-action passes. But while their personnel may have identified them as likely candidates to do so, sample size is a major factor here. The Rams were the only team to use play-action on more than a third of their pass attempts. These teams are throwing RPO passes on 10 to 25 percent of an already small percentage of their total pass attempts. The Steelers, Buccaneers, and Raiders in particular have deceptive RPO pass percentages since they were the bottom three teams in play-action pass percentage. The Chiefs led all teams in raw total of RPO passes with 35, which is just more than two per game.

Outside of the RPO bubble, the Seahawks, Patriots, and Titans are the three other most notable teams, both because of their top-10 play-action frequencies and their top-10 DVOA differentials with play-action. Seattle has been top-10 in play-action percentage for all seven years of Russell Wilson's career. Their success with play-action has fluctuated relative to their success without it, but they always fall in a range from average to above-average compared to other teams with play-action passing. Wilson's mobility is undoubtedly a factor in helping him extend plays to take big shots down the field, but he does not boost the team's play-action efficiency to any great degree with scrambles. The Seahawks have had the same DVOA rank on play-action passes with and without scrambles every year except 2013, when they were one ranking apart.

The Patriots landed in the upper half of teams in play-action percentage for most of the last decade, but their third- (28 percent) and fourth-place (31 percent) finishes the last two seasons are their highest of that timeframe. The Patriots have a reputation as a quick-strike offense, and they certainly have had success without play-action, finishing no worse than eighth in DVOA on traditional passes every year between 2009 and 2017. But they have still maintained a positive DVOA differential on play-action passes for all of those seasons except 2017. When the team slipped to 16th in DVOA on traditional passes in 2018, play-action passes became the cornerstone of their offense. They finished first in both DVOA and DVOA differential on play-action passes.

With Brady just a week away from his 42nd birthday, any decline in any part of the Patriots' offense implies a decline in his ability. It's easy to read the play-action split that way since traditional passes require quick decisions and throws while play-action passes buy Brady more time. But it may simply capture the changes in the Patriots' personnel. The trio of Rob Gronkowski, Chris Hogan, and Josh Gordon combined for nearly 200 targets last season, and each averaged more than 14 yards per catch, but none of them may play a snap for New England in 2019. Instead the Patriots will have a deeper stable of running backs than ever to sell their play fakes. It will be interesting to see how heavily the team relies on play-action with new players like Demaryius Thomas and rookie N'Keal Harry on the outside.

Teams don't actually have to run the ball frequently or well to have success with play-action. Still, the run-dominant Titans are the archetypical play-action team with power rusher Derrick Henry in the backfield and disenchanting passer Marcus Mariota under center. The Titans have finished in the bottom five in DVOA on traditional passes each of the last two seasons, so they have been smart to ramp up their play-action percentage to 23 and 29 percent. But as successful as they have been with play-action, the Titans will likely need Mariota to improve on his other passes. They and the Cowboys were the only two of the the 10 teams with negative DVOA rates on traditional passes that finished with a record above .500.

Comments

5 comments, Last at 25 Jul 2019, 1:11pm

1 RPO Pass YPP

Great stuff here, Scott. What's really interesting to me is that 2nd chart. While the sample size is smaller as you say, for the most part it's the teams with dual-threat QBs who had the most success passing out of the RPO, both compared to the teams there without a QB who runs and compared to their own traditional PA numbers. Could this be because when showing an RPO play fake with a dual threat QB, more edge defenders and LBs and box safeties bite toward the line because they've got two running threats to worry about? (Both the RB and the QB, perhaps thinking it's a read option play)

2 RPO Pass YPP

In reply to by c-note711

Teams with a dual theat QB didn't always fare better with the RPO pass vs the traditional PA pass, I'm realizing. But for the most part, teams with QBs who don't run much in a read option way did far better with the traditional PA pass than the RPO pass (Pit, Indy, Oak). Maybe it's a sign the traditional PA pass is the clear better option of the two if you don't have a QB who'll threaten with his legs out of the RPO. 

3 Re: RPO with dual-threat QBs

In reply to by c-note711

I think your hypothesis makes a lot of sense with dual-threat quarterbacks, but there could be other explanations too.  Cam Newton actually carries the ball himself on a lot of the Panthers' option plays and hits some big runs that way.  Those carries aren't captured by any of the numbers in the article because they aren't passes, but I wonder if that rushing success raises the standard of expectation for Newton to actually want to throw the ball on an RPO?  Where a team like the Eagles might be content to pass on an RPO to gain a few yards that might also net a first down, Newton may tuck and run unless he sees what he thinks is going to be a 10+ yard completion?

4 That's good, I like it…

That's good, I like it. Could definitely be the case that QBs like Cam could be searching for more in terms of a longer pass play when pulling it out of the back's gut and throwing it. And yeah they can also at times have three options on the play, whereas non mobile QBs or QBs looking to run less (Wentz in 2018) would only have two. 

5 Saints w/ different QB's

In reply to by c-note711

Here is an obvious manifestation of your theory: which QB is taking the snap for the Saints.

If Brees is back there, there's no point to even faking a read-option/RPO--just run the play! While Brees is good on those jump-over-the-goal-line-and-extend-the-ball sneaks, nobody would say that he is a threat to run. But put Taysom Hill back there--now you have a credible threat for a read option, and a competent passer for an RPO (backup-QB competent, at least). 

At the end of the day, though, let Brees run a traditional PA pass, and use Hill as a credible option to throw if teams try to stop those read-option plays. Yes, it gives an obvious "tell"--but the Saints have used Hill in other capacities to try to combat that "tell."