Any prediction about the changes Todd Bowles might make to the Eagles' schemes needs to start with an honest assessment of the team's performance so far and some educated guesses about what Castillo had been trying to do.
As any number of people have pointed out, the Eagles' defense has actually been pretty good this year -- most of the time and in most situations. Looking back over the entirety of Juan's tenure, a non-exhaustive list of specific concerns would include:
1. Missed assignments. We haven't talked much about it this year, but this was a problem in 2011, both up front (gaps) and in the secondary (blown and freelanced coverages). I'm including it here because it pertains to the later discussion.
2. Nnamdi's vanishing wheels. This town blows awfully hot and cold on Nnamdi these days. He gives up a couple big completions and he's a bum; then he does a great job on Megatron (with lots and lots of help) and hey look, "he's figured it out again." He is, in both cases, the same guy -- an exceptionally skilled cornerback without much catch-up speed left.
3. Two rookies in prominent roles. Boykin and Kendricks have boatloads of talent, but they're both inexperienced. This has shown up most often when they're challenged in zone.
4. Sub-par play from the WIL spot. Not to pick on just one guy, but on an individual level there aren't too many weak links.
5. Castillo's inexperience as a playcaller. It's been reported a couple different places this week that Bowles has already had some involvement in the playcalling this year. My standing theory since the first game has been that he's called the coverages while Castillo was making the calls up front. Such a system comes with inherent compromises.
Watching the Lions game one last time, I was struck by the contrast between their wide nine and our wide nine. The base defenses are awfully similar up front, but Detroit showed far more variety and aggressiveness over the course of the game.
In fact, my initial plan for this post was to highlight some similar formations/plays from the two offenses, then to show how much more creative Detroit was in attacking them up front. But as I was working on that, I got a pretty healthy reminder of the power of recency bias. Looking back at games earlier in the season, Castillo actually had been more creative up front. Unfortunately, some of that stuff didn't work and I think we can best understand the current (pre-Bowles) state of the defense as a reaction to that.
After the jump, less jabbering, more pictures.
Contrast that with Detroit's alignment against the same look (1TE 1RB):
Not much difference in the front fours, but you can see that their linebackers are "stacked" behind hte DTs. Teams stack their linebackers for a range of reasons (often to protect them from blockers), but for our purposes it's interesting to speculate on why we don't -- and I think it at least partially goes back to point #1 at the top.
After all, it's pretty had to forget what gap you're supposed to be in when you begin the play by standing in it.
Moving to the second point, we shouldn't forget how much trouble Nnamdi had earlier this year playing outside without help. Baltimore got him when we brought the house:
Against Pittsburgh, Antonio Brown is going to run right past him, but drop the ball in the end zone (with the safety in no position to help on the other side):
And we all remember the Giants game.
This is likely an area where Bowles and Castillo disagreed. My guess is that Bowles would tell you he can work around that by mixing up the coverages and giving different looks. Castillo's answer was more along the lines of: "If in doubt, keep two safeties back."
This does a few things. First, it's harder to run against this:
Than against this:
More importantly, in the modern NFL, you're also making big plays possible against you in the passing game. And that's where trying to keep two guys always deep can really hurt you.
If you tell a safety "you've got a gap AND deep coverage responsibilities," you're going to see play action into that gap and the tight end running up the seam:
Or play action into that gap to attack your quarters coverage:
For as much crap as Kurt Coleman gets, I sure wouldn't want his job. And what they're asking him to do is just fundamentally unsound, especially as a regular feature of the defense and not just something we mix in, but don't let people count on it being there.
Buttressing the interior of the defense at the second level also allows the ends to stay in the wide nine, because you're less worried about getting gashed:
The Eagles didn't show the same look every time, but one of the issues they had in terms of pass rush was that the Lions could get an auto-double on the outsides whenever they went with two tight ends:
Quick aside, while we're on the subject of two tight ends. Jordan has contain responsibility on this play. He doesn't keep it, but DeMeco and DRC shut things down for him, in large part because the WR ignores DRC and heads to the safety:
What I find fascinating though is to compare these two frames just before and just after the snap:
A lot of linebacker play happens at an extraordinarily reptilian level.
Three more points before I wrap this up.
First, I think we'll see more variety in the defense now at least partly because we're back to a single playcaller. For the old plan of "you vary the coverages and I'll get everything else in," Bowles needed to know before he made his call how many guys he had to work with and who would be on the field. Blitzes had to be called as packages, because there wouldn't time for Castillo to say, "I'm bringing five, what do you want to do." With one guy calling everything, that should all be easier.
Second, Bowles' new focus on the entire defense should help fix some of the issues we've seen in coverage over the middle. The Cruz touchdown against the Giants is probably the best example. It looks at first like he's running a slant/post right into the safety:
But then Allen walks up and shows blitz:
Which puts DeMeco on the tight end and leaves man outside and a gaping void in the middle:
At the snap, however, Allen locks up the tight end and DeMeco races to undercut Cruz's route:
He just doesn't do it all that well, sort of bumbling out there rather than really committing to it:
For which Boykin didn't particularly care:
I'd like to think some of that Bowles cut-up and coverage magic can rub off on the other groups. We'll see.
Lastly, it sounds from the interviews since he's taken over that Bowles has identified the pass rush issue as being related to two things:
- Teams are getting the ball out quickly.
- Teams are max protecting on the handful of plays where they want to take shots.
The pass rush itself isn't the problem. So what he needs to figure out how to do is beat max protect with scheme up front and take away the easy, quick throws in the back. I'm looking forward to seeing what he has in store on both counts.