Where does Matt Ryan rank among playoff quarterbacks now? Was 2016 even a top-five postseason in Tom Brady's career? Scott Kacsmar's annual look at playoff drive stats also includes the first look at 1986-88 postseason DVOA.
02 Feb 2017
by Scott Kacsmar
The star players going into Super Bowl LI are obvious: Tom Brady, Matt Ryan, and Julio Jones. However, the players we may be talking the most about on Monday could be the running backs: Devonta Freeman, Tevin Coleman, Dion Lewis, and James White. New England's LeGarrette Blount is one of the favorites to win Super Bowl MVP, but I left him out since he had just seven catches for 38 yards this season, and today we're going to look at running backs in the passing game. The four other backs are all skilled receivers, and receiving backs have been feasting on the Falcons and Patriots this season. In what should be one of the most offense-driven Super Bowls in history, the team that gets more mileage out of its backs in the passing game might come out on top.
When looking at 2016 defensive DVOA by receiver type, both the Falcons (No. 26, 16.7%) and Patriots (No. 20, 5.7%) have their worst DVOA against running backs. No defense allowed more targets (141), catches (109), receiving yards (870) and receiving touchdowns (six) to backs than Atlanta. New England was close, allowing 101 catches (tied for second-most) and 801 yards (third-most).
Should this come as a surprise? The young Atlanta defense has struggled this season, and has started four rookies, including linebackers Deion Jones and De'Vondre Campbell. Meanwhile, the Patriots traded Jamie Collins and have relied on players such as Elandon Roberts, Shea McClellin, and Kyle Van Noy at linebacker at times.
Then there are the game scripts. No defenses took the field with a greater average lead than the Patriots (9.76) and Falcons (8.23) this season. Opponents had to throw to catch up from large deficits, and garbage time is literally littered with garbage completions to backs. The 655 passes thrown against the Falcons this year are the second-most against a defense in a season in NFL history (the 2013 Eagles faced 670 attempts).
Still, even when the scoring margin was within eight points either way, the Falcons (20th) and Patriots (26th) did not fare well in DVOA against passes to running backs. When only looking at the first three quarters, both were still average at best. Average defensive play against players such as Coleman and Lewis would be a big problem on Sunday.
So I looked at the 517 passes to running backs in games involving the Patriots and Falcons this season to see what trends popped up, and how these offenses might be able to exploit the defenses for huge gains.
Please note that the research below only includes the regular season. Lewis and White each caught a touchdown in the playoffs against Houston, but were held to a combined five catches for 16 yards against Pittsburgh in the AFC Championship Game. Ben Roethlisberger effectively checked down seven times to DeAngelo Williams for 51 yards in that game. As for Atlanta, Freeman and Coleman combined for 102 receiving yards and a touchdown against the Seahawks. They combined for 77 more yards and another touchdown in the NFC Championship Game. Neither Seattle nor Green Bay had much of a receiving back with which to test the Falcons.
The first thing I should mention from my viewing of 517 plays was the dullness that was so often on display. We have known completion percentages to be very high for running backs -- typically above 70 percent. Former Atlanta back Jacquizz Rodgers was once the NFL's 21st-century leader at catch rate at over 85.0 percent. Even bad quarterbacks can complete a high rate of passes to running backs. A huge reason for this is the short distance on these throws. In 2016, the average pass to a running back traveled 1.05 yards beyond the line of scrimmage. Backs are also often left uncovered on purpose by the defense to protect from plays being made down the field.
This pass from Mike Glennon to Peyton Barber with the Falcons up 43-20 in the final minutes of the game is a good example of how easy it can be to complete a pass to a back in garbage time. The middle of the field was left wide open for Barber.
A quarterback could complete an absurd rate of passes if he made decisions as poor as this one by Landry Jones against the Patriots. Le'Veon Bell still turned the play into a 1-yard gain, because he's Le'Veon Bell, but that took so long to happen that I did not even capture that part.
Here are the general stats for passes to running backs for the offenses and defenses of New England and Atlanta this season. ACC% is an adjusted completion percentage that accounts for drops and throwaways.
|2016 Passing to Running Backs: General Results|
|ACC%: Completion percentage adjusted for drops and throwaways|
The Atlanta offense was very effective with Freeman and Coleman doing a lot of the work for Matt Ryan this season. Coleman led all 2016 running backs with three receptions of 40-plus yards, helping to boost his offense's YPA up to 8.03, which is considerably higher than the numbers of these other three units. Unfortunately, the Atlanta defense was almost as good as the offense at seeing on-target passes completed to backs at 88.5 percent. It was also the unit to get the biggest boost from dropped passes.
Drops were also a problem for the New England offense, though the sub-70.0 percent completion rate is not just due to Tom Brady's suspension to start the season. Brady was 71-of-103 passing to backs this year, or 68.9 percent. He did have all seven of the intentional throwaways for the Patriots. Yes, even a clear throwaway can still end up as a receiver's target, which has long been a pet peeve of mine.
I was most interested to see where the running back was aligned at the snap when he was targeted for these four units. I broke down each play into three categories: backfield, slot, and wide. The Browns actually had a play where fullback Malcolm Johnson was an in-line tight end, and he tripped on his route, which almost caused an interception for the Patriots. The other 516 plays all fit into one of these three categories though.
|2016 Passing to Running Backs: Target Alignment (Pre-snap)|
Not surprisingly, the two well-coached offenses were more varied in their approach than the opponents. We have talked a lot about New England's defense having the easiest schedule this season, and that showed up a little in these numbers with 88.3 percent of opponents' running back targets coming out of the backfield. Meanwhile, the Falcons played more versatile offensive teams like the Eagles (Darren Sproles) and Saints (twice). Arizona was also using David Johnson as much more of a receiver in Week 12 in Atlanta than it did in Week 1 against the Patriots.
New England's games rarely saw a running back get targeted out of the slot. The offense did like to send a back out wide though, and rarely did a defender bother to cover him when it was fullback James Develin. When the Patriots went with a back out wide, more than half the time he was targeted on a curl (also a hook or hitch, depending on your preference) or slant. These plays were not always a late shift either, and typically the defense stuck with the outmatched linebacker out wide. The same can be said for all of these units when a back went out wide. Defenses simply do not show enough respect to these players' receiving ability. Generally speaking, running backs receive less defensive coverage in the passing game than any other position, and this is simply not exploited enough by the league's offenses, as you are about to see.
The last thing I tracked was the play type, though I largely grouped together passes thrown to the flats, checkdowns and those little swing passes into one category (FCS). As a charter, I have never been a fan of pretending to know the play call or what the quarterback's first read was, so if it was a pass thrown within 10 yards of the line of scrimmage that wasn't an obvious design (screen, smoke, shovel, wheel, etc.), then I likely marked it FCS. It's true that there are some flat passes that are designed to be the first option, and there are little angle routes and such, but a lot of this stuff looks similar. I marked the actual route on the plays where the running back was in the slot or out wide.
|2016 Passing to Running Backs: Play Types|
A few routes were left out for space reasons, and since the total sample size among the four units was really tiny. We're talking about something like Atlanta's two shovel passes in Week 1 against Tampa Bay. After the second was stopped cold in the backfield for a 5-yard loss, the Falcons never called it again this season.
Again, not a surprise to see the New England defense get attacked with the blandest approach by the opponent, with 78.1 percent of the passes marked as FCS. The offense was only at 50.0 percent. That is surprising that the defense was challenged by just seven screens all season (and zero slants). The New England offense loved traditional running back screens with 28 of them, compared to just two bubble screens with the back out wide or in the slot. The Falcons had 27 total screens to backs on offense, but nine of those were bubble screens. There were nine quick outs against Atlanta's defense, but Darren Sproles accounted for five of them in one game.
The really intriguing play was the wheel route down the field. Out of the 14 attempted, six were completed, three more were dropped, and one saw a linebacker (Corey Nelson on Freeman) beat so badly so quickly that it led to the only defensive pass interference penalty in the study. Some of the drops were pretty bad, but the overthrow by Alex Smith in Atlanta to Spencer Ware with a chance to put the game away was especially heinous.
What can really make this play so effective is the pick that delays Campbell in getting to Ware. Smith just missed the throw, but this play could work for a lot of offenses out there if it was just called more.
The Falcons have some fine receiving options in Julio Jones, Mohamed Sanu, Taylor Gabriel, and whichever tight end is open, but Freeman and Coleman are both very good receiving backs. Atlanta would be wise to exploit New England's linebackers by using Coleman at wide receiver with Freeman still in the backfield. There were only nine plays (five passes) all season where the Falcons had Freeman and Coleman on the field together.
Freeman runs with good power, but Coleman's speed is a big weapon. In Week 5 at Denver, Coleman lined up in the slot and linebacker Brandon Marshall never had a fair fight in keeping up with him. The result: a 31-yard touchdown right down the seam.
If Coleman doesn't go deep at least once against a New England linebacker in this game, then the Falcons are missing a golden opportunity.
Have you heard that the Patriots are 16-0 since 2015 when Dion Lewis plays? No, he's not that valuable, but he is a great fit in this offense, and is a nice dark-horse bet for Super Bowl MVP. Since Lewis returned in Week 11, the Patriots have been able to use him and James White as great complements to the power back in LeGarrette Blount. One of my favorite plays from this study was another one in Denver. While the result was not a long touchdown this time, the advantage of unpredictability was on display again.
Lewis lined up out wide, but Denver had plenty of time to adjust. Cornerback Aqib Talib stood his ground, giving a huge cushion to Lewis on a second-and-5. Talib stood flat-footed after the snap, watching Brady and likely never expecting Lewis to test him deep. Brady stepped up for the deep throw, and Lewis had a step on the Pro Bowl corner, but the pass was well overthrown.
Again, the separation on these deep shots -- whether it's a wheel route, a double-move, or a straight-up go/fly route -- is usually there with these running back targets. The quarterback just has to make the throw.
In Super Bowl XLIX, Brady used Shane Vereen against Dan Quinn's Seattle defense to the tune of 11 catches for 64 yards, substituting as the running game for New England that night. That's the kind of impact Lewis could have on Sunday night, and he is a more elusive runner with the ball in his hands than Vereen ever has been.
The Patriots shouldn't have to test Atlanta deep too much with their backs in this matchup. The "death by 1,000 papercuts" might work just fine against Atlanta's young front seven, which has had some issues this year. Quarterbacks were 38-of-40 when throwing to a running back in Weeks 1 to 4 against Atlanta, and that included one drop. Rookie Paxton Lynch missed four of his first five throws against Atlanta in Week 5, but the defense still had some tackling issues. Just look at this touchdown allowed in Week 1 to Tampa Bay's Charles Sims, who looked like a few Falcons had him dead to rights before he was able to break tackles.
You would like to think those were Week 1 growing pains, but come Week 12, the Falcons were still looking lost at times on defense. Arizona's David Johnson should command some respect from a defense, so it is pretty comical to see Campbell turn his back to chase Larry Fitzgerald down the middle of the field, leaving no one in sight near Johnson for a 25-yard gain.
Expect Brady to try exploiting the two rookie linebackers in this game. When a player like Lewis or White does get the ball in their hands, the Falcons better wrap up or the YAC is going to pile up.
It goes without saying that the Falcons are the best and most complete offense the Patriots will play this year. That includes the potential impact Coleman and Freeman should have. One area where New England has looked vulnerable is when the linebackers have to defend deep, which is exactly why I said Coleman in particular has to have a big hand in this game plan for the Falcons. The Patriots were targeted by six wheel routes this season, and they allowed four completions for 75 yards. The linebackers were also beaten on two other deep shots, including a 38-yard catch by Seattle's C.J. Prosise. The Seahawks were one of the few offenses to really challenge the Patriots with a receiving back in a variety of ways. They were also the only offense to outscore the Patriots with Brady this season.
There is nothing revolutionary about putting Coleman at wide receiver and having him run slants and go routes against players not suited to keep up with him. However, it would be unconventional, but sometimes going against the grain is necessary if you are going to break through and win a title.
7 comments, Last at 07 Feb 2017, 1:41pm by Guest789