No, You Build *Around* Brandon Graham

As we move into free agency and the draft, a lot of assumptions are being made about the kind of scheme the Eagles are going to run on defense next year.  Assumptions that aren't really supportable if we go back and look at Chip Kelly's introductory press conference:

That's one of the things about Billy's background is his versatility because he's coached in both. What direction we end up ultimately heading in, I like the 3-4 better...

One of the things that really attracted me to Billy was his versatility and being able to coach in both systems...  And that's what I wanted in a coordinator, is a coordinator with versatility. Then it's our job as coaches to figure out what is the best scheme for the guys we have in place.

Everybody has a wish list of how they want to do things and what they want to do. But everything we do offensively, defensively, and special teams wise will be driven by personnel.

...

We won't have a really good understanding of our guys until we get to that first camp in April ... So we've got time, and I think we've got versatility in our coaching staff, and I think we have some versatility in our roster.

...

What we're going to do is put our guys in the best position for them to make plays. I don't know if that's being a 3-4 team, a 4-3 team, a 5-2, a 6-1 team...

We're also not caught up in that. It's about making sure we play sound defense on first, second and third down. We could look drastically different on first and second downs than third downs. And that's going to be entirely personnel driven for us. Could it be a 4-3 under defense? Yeah, or it could be a 3-4 under defense.

I'm not caught up with labels...

The message here seems pretty clear:

• In the long run, Chip Kelly would prefer to run a 3-4 (for a number of reasons, including an interesting point about improved special teams from carrying more linebackers and fewer linemen).
• In the short run, he's going to structure the defense to fit the personnel, not the other way around.
• One of the reasons he hired Billy Davis is because he's comfortable coaching a mix of 3-4 / 4-3 / hybrid-whatever, so he can help carry everyone through this transition period.

I could make an argument that Chip Kelly's background as a college coach is driving his thought process here.  When you're in an environment where you're losing six to eight starters every year and aren't recruiting at an Alabama level, you're going to have to be flexible with adapting what you do to the players you have.

On the other hand, Bill Belichick never coached at the college level and this is also what he does.  There's more than one path.

So let's talk about Brandon Graham.

There's been a fair amount of Graham talk lately.  Does he fit as a 3-4 OLB? Would it make sense to move him for someone who might fit better?  Why do we even care about Graham anyway since he's a bust and we should have taken JPP?

All of these things are wrong.*  Let's work backwards.

On a per-snap basis, Brandon Graham was strikingly productive in 2012.  Pro Football Focus offers a "signature stat" it calls "pass-rushing productivity" that mashes together sacks, hits and hurries -- with a weighting factor for each piece -- and then controls for the number of attempts a player gets.^  I don't actually like the stat much, but it's hard to argue with it at least in directional terms.

Brandon Graham led the league in PFR in 2012.

Obviously there are ways to argue with that stat.  He was a part-time player who didn't become a starter until late in the year.  Maybe he was only effective because he got the bulk of his chances when he was fresh and everyone else was tired. 

Nope:

You can slice those stats by use periods (little early, some in the middle, lots late) or pre/post-departure of Babin.  The numbers are, allowing for small sample sizes, consistent.

You get similar results looking at his rush defense stats.  He had tackles on 12.7 percent of run plays as a backup and 12.5 percent as a starter (again, PFF stats).  Overall, Graham was in the top three in tackles, tackles+sacks, and stops per snap among all defensive ends.  

Let's flip sites as a sanity check and head to AdvancedNFLStats.  Trent Cole got 75 percent more snaps than Graham, but the two came out almost even in terms of EPA+, WPA+ and EPA+/game.  

Sort that table by EPA/game and look at the top of that table.  Then consider this per-pass rush comparison:

No, I'm not saying that.  But there's something going on here.

There are plenty of coaches -- even some good ones -- who would look at Brandon Graham and see a square peg that won't fit a round hole.

Chip Kelly's gonna re-cut that damn hole. 

-------------------

*Maybe not the part about how we should have taken JPP.  

^Actual formula is sacks + hits * 0.75 + pressures * 0.75.