What’s wrong with the Eagles’ pass rush?

“Well, it starts with our guys up front.”

Stop me if you’ve heard it before. I’m sure you won’t because I’m sure you have. One of the most common narrative crutches for a coach when describing his defensive philosophy, or a general manager detailing his team-building approach, is to put the first onus in the trenches.

Such has always been the case for the Philadelphia Eagles under defensive coordinator Jim Schwartz, the king of “pressure with four,” curator of the hockey-style line shift, zealot of third and long. When Schwartz first arrived in Philadelphia with HC Doug Pederson in 2016, the Eagles ranked 11th in Football Outsider’s Adjusted Sack Rate at 6.6% — that is to say, a sack on 6.6% of their opponent’s dropbacks.

Hey — that’s pretty good! But what’s even more impressive is the pressure rate: in 2016, the Eagles were 3rd in total pressure rate, despite bringing only 4 man rushes on over 75% of their snaps defended. The value of a pressure has been extolled in recent years, to the point that football viewing canon has integrated pressure into the viewing experience. This is especially for Eagles fans: defensive end Brandon Graham is the poster boy for the argument that low-sack rushers who generate a ton of pressure still provide value. Pressure equals hurried throws, lower QB ratings, increased likelihood of interceptions; pressure makes quarterbacks worse.

So Schwartz’s arrival in Philadelphia turned up the heat, to that 3rd-highest pressure ranking in the league: 31.6% of snaps defended saw a pressure. Interestingly, that number became 33.5% in 2017 (8th in the league), and 30.8% in 2018 (17th in the league). The rates stayed mostly consistent, but when subjected to the context of a league that’s got better at rushing the quarterback, they dropped relatively.

The same is true of the Adjusted Sack Rate I mentioned above. 6.6% was good for 11th in 2016; in 2017, 6.3% put them at 19th; in 2018, 6.5% put them at 26th. And now, in 2019, the Eagles sit with 2 sacks through 3 games, staring down the barrel of a 3-0 Green Bay Packers team on Thursday night. Their Adjusted Sack Rate is 1.6%; only the sackless, winless Denver Broncos are worse.

This inertia would be troubling for any team — in large part because every team says that it starts with their guys up front. But for Philadelphia, it steps beyond worrisome and into perplexing, in large part because their defensive line hasn’t just been good — it’s been great.

Fletcher Cox is ranked by film analysts as perhaps Aaron Donald’s only peer, and was positioned as such by AP on the 2018 All-Pro team. PFF put Cox 5th overall to Aaron Donald’s 1st overall on their 2018 player rankings; Cox is also second to Donald in terms of average contract value. Brandon Graham actually had a higher grade than Fletcher Cox did coming into the 2018 season for PFF, and finished the 2018 season 21st among EDGE rushers in PFF’s Pass Rush Productivity metric.

Did the Eagles’ top rushers suddenly get worse in 2019? It doesn’t seem so. Per ESPN’s Pass Rush Win Rate, Graham is sixth among NFL EDGEs with a 28% win rate, and opposite DE Derek Barnett is eighth at 26%. Overall, the Eagles have a 59% Pass Rush Win Rate as a team, good for third in the league. On Pass Rush Productivity from PFF, Graham is 22nd among NFL EDGEs, and Fletcher Cox is 12th among defensive tackles — not his best, but still strong.

Philadelphia’s top players are still rushing with success, even if it isn’t the world-beating levels they hit in previous seasons — and as a team, the Eagles have one the top rushes in the league. Yet the sack numbers aren’t there, and the defense is bleeding long drives, passing yards, and points. The rush is as potent, but less effective; as successful, but less impactful. It wins in a vacuum, but not longer in context.

Offenses aren’t losing to the Eagles’ pass rush anymore.

If it starts with the guys up front, then it’s worth taking those guys out of the picture as an offense, and seeing if it ends with the guys up front, too: in Philadelphia’s case, it certainly does. Through the first three weeks of the season, the Eagles’ secondary has been shredded beyond recognition, despite the fact that its staffed by many of the same players from the last couple seasons, and deployed in much the same technique as before.

Throughout Schwartz’s tenure in Philadelphia, teams have devoted growing resources to extinguishing the Eagles’ four-man rush, even if just staying the swelling tide for that moment more. It matters in general, and against Philadelphia, it matters a lot: In 2017, the Eagles were tied for the second-fastest time-to-pressure rate in the league, at 2.33 seconds. That was their average figure — they were also the only team to register a pressure under 2.5 seconds on more than 30% of their pressures.

In 2019, teams don’t want to give Philadelphia that time. Matt Ryan and the Atlanta Falcons had the second-fastest time to throw in Week 2, at 2.18 seconds. When we step into the film, we see how this affected Philadelphia’s rush: Ryan and OC Dirk Koetter fed the flats and worked the quick screen/RPO game to keep Philadelphia’s hungry dogs on short, unforgiving leashes.

Look at the variety of offensive line angles and sets given to Philadelphia’s front four here. The jump/pop passes from Ryan on RPO looks; the ability to audible into WR screens against off-man coverage (I didn’t include the game-winning Julio Jones touchdown because it was a Cover 0 blitz, but the same logic applies). These concepts are low depth-of-target throws that typically have only one pre-snap read assigned to them — they are, actually, the easiest plays in the book.

So why have a talented QB like Matt Ryan running them? Because they get the ball out of the quarterback’s hands. Typically we assign these plays the character of a bad quarterback: when you don’t have talent at QB, run a spread offense that gives him pre-snap reads and requests of your pass-catchers an extra effort to generate YAC and chunk plays.That spread and shred system has the adjacent benefit of entirely nullifying a pass-rush; forcing a defense to rally and tackle.

We must remember now the shifting landscape of the league’s offense, and the correspondingly fluid priorities of a modern defense. As Establish the Run’s Evan Silva put it:

It’s important to remember that plays like RPOs (which quite literally could be a running play) or WR screens (which, like runs, involve giving the ball to one of your best athletes behind the line of scrimmage with blockers in front of him) are effectually running plays for offenses. A flat route for Matt Ryan is pitch-and-catch automatic; the transferring of the ball from quarterback to skill player is as predictable as a handoff.

Yet the Eagles’ secondary is not well-equipped to handle the newfangled passing attacks in the NFL, while the Eagles’ run defense remains perennially one of the best in the league. Watch again and count how many defenders the Eagles have in the box; watch how quickly the linebackers flow against play-action. If Jim Schwartz insists on stopping the run as his first priority, he must expand his definition of a run.

Again, offense in the NFL is evolving: every year, passes become a higher-percentage play. We’re getting better at throwing the football, but the Eagles’ defense continues to leave the disruption of those plays as a second priority, and the cost of that hierarchy falls on the defensive backfield. Corners play far off the ball and are unsupported by safeties; they are exposed in space to great route-runners and tackle-breakers alike.

Accordingly, it is all too easy for teams like the Falcons to throw in rhythm to their first read — such as the last play in the cut-up above, which you’ll see again as the first play in the cut-up below. Several of these plays have a fast release time, that could come in under the 2.5 second threshold — but what matters more in the eyes of a play designer is the quarterback footwork and rhythm. Even when the Eagles have rushers that win their reps, passers are throwing in time to their first read, unabated and unafraid.

And first reads aren’t always shallow, quick game routes.

Media and opposing coaches laud Bill Belichick for “choosing what will beat him” as a head coach. He’ll scheme to take away this, or get Player X against Player Y and win that matchup until you adjust. He addresses that at which you’re best and says “Hey, if you’re gonna beat me, you’re gonna have to do it with something else.”

The model for offenses against the Eagles follows this framework. There is no reason to anticipate the Eagles’ secondary taking away your first read — so why would you throw it anywhere else? It’s regularly open, in rhythm, and frequently beyond the sticks — and if you’re really worried about the pass rush getting to you before then, they’re struggling to tackle before the sticks as well. Until the Eagles’ secondary can take away the opponent’s first read — with better route recognition in zone coverage; or an in-season acquisition to improve their man coverage ability; or with a press alignment to disrupt route stems early — make them beat you with their corners. Don’t let them beat you with their defensive line.

We see a similar idea in the Lions’ approach to the game plan against the Eagles. Unlike Ryan, Matt Stafford had one of the longest times to throw in the league when he faced the Eagles, at 2.91 seconds. Why wasn’t the pass rush getting there? Because the Lions were expanding the surface of their offensive line, regularly asking their backs and tight ends to chip the Eagles’ EDGEs.

Chipping doesn’t necessarily negate pressure — and at times, it can assist rushers, who use the momentum to careen into interior gaps and generate new angles on the offensive tackles they’re facing. However, no matter how good your defensive end is at playing off chips, chipping delays rushes. It protects the offensive tackle’s initial set by allowing him to play aggressive into the rusher, unconcerned with the potential of a quick speed rush beating him around his outside shoulder. It forces EDGEs to restart their momentum and allows tackles a stunned target to latch onto; it disrupts timing of handwork and inside counters. It increases the OL’s surface by expanding the sheer body count across the width of the pocket, making paths to the quarterback longer in distance and thereby longer in time.

You see Matt Stafford get pressured here; you see him get hit. You see him throw incompletions. But the story that the chipping tells matters more than the results: the Lions are perfectly content releasing only three receivers into the Eagles’ defensive backfield. That’s the cost of the chip: your tight ends and running backs can, at most, release late into the route concept as check down options.

The Lions would rather have spent seven men blocking four defensive linemen for the Eagles than sent five men against seven defensive backs in the route concept. That’s an oversimplification, but it hearkens back to the same idea: if the Eagles corners aren’t going to win one-on-one match-ups, why bother sending a RB swinging into the flat when you’re unlikely to need him? The speed out, the deep comeback, the back-shoulder fade — they’re all open! He’d be more useful in the backfield, ensuring that Brandon Graham and Derek Barnett can’t get to the quarterback.

Same idea, different manifestation all the way back in Week 1 against the Redskins. Here, Washington regularly presented the Eagles with hard play-action concepts, which included boots, half-rolls, and straight dropbacks. Play-action is often credited with suckering linebackers down into the line of scrimmage, which creates intermediate throwing windows behind their flow — and that’s true. But it also initially puts the defensive line in the paradigm of run defense, which involves retaining leverage and responding to offensive line flow. The focus on penetration, on the quarterback’s set point, on a quick time to throw, fades away in the bright immediacy of defending the run.

Accordingly, Washington used play-action to not only manipulate Philadelphia’s zone defenders and score free completions as the struggling Eagles’ coverage shells failed to adjust, but also to buy time in the pocket for downfield shots, on which they would almost certainly have the advantage in one-on-one coverage.

Again, you see pass-rushers win. On that last rep, they may have won just enough to help create that incomplete touchdown bomb. But we already know that the Eagles’ pass-rush is generating pressure; we’ve seen that in the stats and it’s there on the game film as well. What isn’t there is the sacks, and that’s easy to see: the camera sits on the quarterback in the pocket, the secondary safely off screen, and our brain records our heart’s disappointment every time the passer gets the ball out before the rush arrives.

But in that sacks are situational, and that pressure rates are high, Philadelphia’s sack rate is expected to regress upward to the mean. It is highly irregular than an NFL team have this few sacks on these successful of pass rushes, and eventually, statistical normalization conquers us all.

What also isn’t there, however, is the coverage — and if it seems like it’s lacking worse than it ever has before, that’s largely because opposing offenses have decided to stop losing to the Eagles’ pass rush. Offenses will take their best guy against your best cover guy — who may very well be Rasul Douglas on Thursday night against the Packers, God be with us — they’ll take him early in the down, and they’ll take him in contested situations, too.

We don’t feel the deficit in coverage as sorely as the deficit in pass rush because we don’t see it as easily: again, our rampant selection bias is attuned to that which the camera shows us, that which happens first, before the ball is even released. But the issue with the Eagles team, as it was when the front four dominated opponents in 2016 and 2017, continues to be their secondary, both in talent and in scheme. Now that the pass-rush has hit a dry spell, the defensive backfield has lost its cover; we can see the wound beneath the bandage.

The pass rush is worse in 2019 — that’s measurable and clear. But it isn’t nearly as bad as we think, and it will improve in the sacks column (or else become a shocking statistical anomaly). But as it becomes more productive, it will do so in the face of overwhelming, focused opposition: teams will continue to devote the resources of the quick game, the play-action game, and chip blocks to dull its edge. Until they are presented with resistance in the secondary, there is simply no reason to do otherwise.

This time, it starts with the guys deep downfield.