Bend a Little More, Break a Little Less

Back in June, I took a look at the way the Eagles have traditionally outperformed their yardage statistics -- both offensively and defensively -- in terms of points gained and allowed.  At the time, I wondered whether special teams' performance could account for the gap, but I didn't push much beyond that speculation.

I decided to run a quick test on the numbers to see what I could find.  I want to stress that this wasn't a very in-depth analysis, so don't take these numbers as gospel, but I think it's at least broadly correct.

First I pulled yards gained/allowed and points scored/allowed for every team from last year.  I then graphed the yards/points figures for both units in a scatterplot and had Excel add a trendline to find the predicted ratio between yards and points.  Finally, I took the total difference between predicted and actual points scored and allowed, and then ran a correlation analysis with the special teams' DVOA figures provided by Football Outsiders.

In English, I figured out what each team's point differential should have been, based on their yardage stats, and then looked to see how much of the variation from the actual stats could be explained by special teams performance.

Like I said, it wasn't the world's most intricate study, but when I ran the numbers as described above I ended up with a correlation coefficient of .804 -- which is really, really high.   (And by the way, if you want to see the excel file I used, focus on sheets 4 and 5 of this document.)

This is interesting for a couple of reasons.  First of all, it suggests that the FO guys are really on to something with their special teams DVOA numbers.  They're explaining an awful lot of variation with just one stat (at least for 2006).  Secondly, it makes me wonder about one particular essay in the 2007 Pro Football Prospectus book I've plugged so many times.  The two authors take a lengthy look at "bend but don't break defenses" and argue that such a creature exists, although it's rare for a team to be able to sustain such a performance for more than a single year.  The funny thing is, though, that their measure is most highly correlated to "opponent's average initial line of scrimmage," which basically means it's a hell of a lot easier to bend without breaking when the other guys have a long way to go to score.  I'm not sure that's a conclusion we needed a lot of math to reach.

With that said, there does seem to be something to their whole theory.  New England under Bill Belichick and the Eagles under Jim Johnson have nine of the top 44 seasons in terms of points prevented per drive (aka bending without breaking) during the last nine years.  That makes sense and squares with our intuitive understanding of the two teams.  And one of JJ's seasons came last year when, as we've already seen, the Eagles' special teams weren't that good.

So what does that tell us about '07?  My theory about the Eagles defense this year is that we're going to see a fair amount of bending, but a lot less breaking.  The bending is going to be a direct result of the inevitable mistakes made by some young players.  I expect the Eagles to be a lot tougher to run on this year, but it could be a little while before Gocong and Gaither (especially Gocong) have enough experience to avoid being taken advantage of by opposing offensive coordinators.  There could be some ugly plays out there.

As much as I'm worried about Chris Gocong -- and boy am I worried about him -- I am absolutely in love with this year's nickel defense.  We saw a fair amount of it in the preseason, and it was easily the best look we could throw at people.

Up front, we get Cole and Juqua flying off the edges, with Howard or a healthy Reagor -- or both -- as our pass-rushing DTs.  At the next level it's Spikes and Gaither, both of whom are better than the guys we used in the nickel backer package last year.  (They can even go dime and bring in Mikell to play one of the linebacker spots, where he also makes plays -- think of the Cowboys.)  Finally, in the back William James comes in to match up on the outside and Sheldon Brown moves into the slot, where he gets to mix it up and do his thing.  His shoulder looked completely healthy during the preseason, so I expect a return to the physical play and excellent tackling of his early years. 

Now that's one mean-looking unit.  And it will be even better as long as JJ decides that the guys who are paid big bucks to rush the passer should probably rush the passer, rather than dropping into coverage and looking lost out there.

This unit is likely to be the one Johnson goes to in the distant red zone (say between the eight and 20 yard lines), both because of its versatility and the way it allows him to get creative with his blitz schemes.  So long as guys aren't giving up those big scoring plays from other areas of the field, this could be one very tough defense to score touchdowns against. 

There's also a connection here to Green Bay, but I'll save that one for later...   

Comments

[pawnking71]

It's obvious there is a real connection between special teams and defenses based on your analysis, but why not mention the best reason for hope in '07? Sav, if he can pin opponents inside their 10 or 20 every time the Eagles advance as far as their own 45, might become the MVP of the defense this year.

This will allow for a whole lot of room to bend without breaking by JJ's boys.

[Me]

Good point.