Projecting the Eagles' injuries on 'D'

This afternoon, we continue our Q&A with Bill Barnwell of Football Outsiders.

Barnwell wrote the Eagles chapter in the Football Outsiders Almanac, which projected them to win 9.3 games this season.

One of the major reasons Barnwell wasn't as optimistic as some others was that he projected the Birds to suffer more injuries, specifically on defense, in 2009.

But is there really a formula to project injuries? Isn't it mostly luck?

I asked Barnwell those questions yesterday, ironically the same day we found out the Eagles would be without middle linebacker Stewart Bradley for the entire season.

Q: One of the major reasons you projected the Eagles to have less than 10 wins this season was more expected injuries on defense. A skeptic would say that trying to predict injuries is ridiculous, and it’s all luck from season to season. How do you go about projecting something like that?  Have you had success in the past doing so? And lastly, you have the Eagles’ medical staff ranked 14th. How did you arrive at that number?

A: That skeptic is right, and we agree with them!

What we've found is that teams aren't consistently healthy or injured from year-to-year. So when a team or a given unit stays particularly injury-free or is dramatically afflicted by injuries, you can safely forecast that they'll suffer more injuries in the subsequent season.

We track injuries with a metric known as Adjusted Games Lost (AGL), which takes into account the various lists and injury reports put out by the NFL and assigns each injury a value based upon the player's importance to the team (namely, whether he's a starter or a reserve) and the historical likelihood that a similar player would miss a game. We know that quarterbacks listed as Probable almost always play, for example, so we assign them a small sliver of a fraction of a Game Lost if they appear as such on the Injury Report.

Last year, the Eagles' defensive starters combined for an AGL of 3.5. (That was the second-lowest total in the league, behind only Miami, who had a microscopic 1.6.) Those 11 players combined to miss a total of exactly one game, thanks to Asante Samuel's absence in Week 13 against the Cardinals. That's worth 1 point of AGL; the other 2.5 points comes from players being listed as Probable and Questionable on the injury report.

Now, how can we forecast that the Eagles would suffer more injuries defensively? Let's take a look at all the teams that have put up an AGL between 0 and 5 since 1996, the first year we have injury data for. That's 46 teams, not including the Dolphins and the Eagles from last year (since we don't know how well they'll do next year). In the subsequent season, 38 of the 46 teams (83 percent) had more injuries than they had in the previous season. The average team over those seasons had 13.2 AGL in the following season; the league-wide average for defensive AGL is 16.4, so they were very close to average.

We're not at the point where we can say "This team's going to have two guys go down with season-ending injuries, and a third guy's going to get hurt in Week 8.", nor will we ever be. It's possible that the Eagles could have everyone else stay healthy for the rest of the season, absolutely. Even without Bradley's injury, though, expect the Eagles to see more defensive starters miss time in 2009.


That fourth paragraph really stuck out to me. Basically 83 percent of the teams that had the same injury luck the Eagles had last year were far less lucky the following season.

I'm in the camp that says you can't really predict injuries, given the way the game is played, but it's interesting to look at the likelihood of the Eagles once again being one of the league's healthiest teams.

Click here for my earlier post about the Eagles' morning practice, including observations on Joe Mays and LeSean McCoy.