Writers of Pro Football Prospectus 2008

Most Recent FO Features

PrescottDak15.jpg

» OFI: Blowouts, Upsets, and Narrow Escapes

The College Football Playoff field narrowed on Saturday. Some teams got upset, some barely escaped an upset, and a few had big record-setting blowouts.

10 Mar 2009

2008 Adjusted Games Lost

by Bill Barnwell

The Dallas Cowboys weren't the only team who had their season dramatically impacted by injuries in 2008. In fact, many of the season's biggest surprises have a fair amount to do with their health -- or lack thereof -- relative to 2007.

Take the Miami Dolphins. The addition of Bill Parcells, Chad Pennington, and about half a full roster of new players had a lot to do with their ten-win improvement upon the 2007 campaign; they were also helped by a shift of over 36 Adjusted Games Lost (henceforth AGL), the largest positive difference in the NFL last year. It wasn't just the shenanigans related to keeping the injury report clean, either: The 2007 Dolphins had 56 instances of a starter missing a game due to injury; the 2008 Dolphins had just 25, and 15 of them were rookie right guard Donald Thomas, who managed all of two quarters before going down for the rest of the season.

Even more impressive were the Atlanta Falcons. Helping the Falcons' resurgence was the fact that they had only one starter -- tackle Sam Baker -- miss time before they clinched a playoff spot. The division-leading Carolina Panthers had the second-fewest AGL in the league, with only 6.7 AGL outside of their offensive line.

The Baltimore Ravens, though, were the trickiest team of all. Unlike the other teams that surged to success, the Ravens were riddled by injury, finishing with the third-most AGL in football with an even 95.0.

For those of you unfamiliar with AGL, it's introduced in an essay called "The Injury Effect" in the back of Pro Football Prospectus 2008. While I won't go into a detailed explanation here, AGL uses the injury report, the status and relevance of a player to his team, and historical data regarding how likely a given player is to play to produce a measure of how dramatically a team is affected by injury. A team's shift in AGL from year-to-year bears a significant correlation with changes in their win-loss record as well as their team DVOA.

Team AGL Rank Difference Rank
NYJ 17.8 1 -20.3 5
CAR 19.9 2 -6.0 10
TEN 23.4 3 8.3 19
MIA 25.7 4 -36.4 1
ATL 27.9 5 -18.8 6
SF 30.3 6 -21.0 4
CHI 33.3 7 -35.5 2
MIN 33.5 8 14.1 22
ARI 33.5 9 -29.6 3
SD 34.7 10 -2.6 11
NYG 35.9 11 7.5 18
PHI 39.0 12 0.4 13
TB 42.8 13 -11.7 8
HOU 44.0 14 -6.6 9
KC 47.3 15 20.3 23
BUF 48.2 16 -18.0 7
GB 48.2 17 23.4 25
DAL 53.3 18 7.5 17
OAK 54.3 19 10.9 20
WAS 55.0 20 1.6 15
PIT 55.2 21 26.4 26
NO 56.4 22 21.3 24
JAC 64.2 23 13.1 21
IND 70.2 24 -0.1 12
CLE 73.7 25 30.6 27
DEN 74.8 26 7.0 16
NE 79.1 27 48.2 30
SEA 79.9 28 39.2 29
STL 80.3 29 0.4 14
BAL 95.0 30 36.4 28
DET 102.3 31 73.0 32
CIN 108.1 32 62.8 31

These numbers are strictly the figures for starters, those players that replaced them, and players who get significant playing time without starting (think someone like Jonathan Stewart, Adam Terry, or Joselio Hanson). The numbers are not weighted by player quality, although we'll be experimenting with adjusting the numbers for players with Pro Bowl and All-Pro appearances in the offseason.

Going position-by-position, the most- and least-injured teams are mostly who'd you expect:

At quarterback, Kansas City (21.4 AGL) finished just ahead of Detroit, thanks to injuries to both starting quarterback Brodie Croyle and backup Damon Huard, who assumed the starting job for a short period of time. Nine different teams recorded zero AGL on the season.

It's also worth noting that Bill Belichick did not put Matt Cassel on the injury report once all season -- maybe that Tom Brady shoulder injury was serious after all.

As expected, Denver's traumatic set of running back injuries led the league. As players came onto and fell out of the starting role, we ended up with the same figure of 21.4 AGL. If we weighted starters and reserves equally, the Broncos would have ended up with 57 AGL -- no team in the eight years we have data for have had more than 38.4 AGL. The Jets, Dolphins, and Cardinals all made it through an entire year without an injury to their starting halfbacks.

The other set of injuries widely recognized as absurd was that of the Seattle wide receivers (22.9 AGL), which includes injuries to an impressive six different wideouts at one point or another during the season. Our system considers Deion Branch and Bobby Engram to be the expected starters at the beginning of the season, with Ben Obomanu and Nate Burleson being the immediate replacements. When Obomanu went down in the preseason, he was replaced by Logan Payne, who himself only lasted through Week 2. Burleson couldn't make it out of the first week. Once Koren Robinson took over as a starter, we stopped counting Burleson, and the return of Engram got us to stop counting Payne and Obomanu. That's some of the more ridiculous (and subjective) accounting we've ever had to do. Randy Moss and Wes Welker didn't make a single appearance on the injury report, giving New England's the league's healthiest wideouts.

St. Louis led the league in tight end AGL with 12.0, which matches the 12 games Randy McMichael missed with a broken leg.

The most injured offensive line belongs to the Jacksonville Jaguars, who saw long-term injuries to Brad Meester, Maurice Williams, and Vince Manuwai en route to 36.2 AGL. Seattle was right behind with 34 AGL. Naming the healthiest offensive line in football makes me shudder, because it's the Giants and their 0.7 AGL. The starting five of Diehl-Seubert-O'Hara-Snee-McKenzie hasn't missed a single regular season game since the beginning of the 2007 season. That's extremely unlikely to continue in 2008, and as Giants fans saw when Kevin Boothe came into the lineup at the end of the year, it's a scary thought.

Wondering how the Ravens could've been so beat up without being mentioned once? Welcome to the biggest miracle of the 2008 season: The Baltimore defense. They led the league with 32 AGL on the defensive line, with season-long injuries to starters Dwan Edwards and Kelly Gregg. The Cowboys only had two instances all year of a starting defensive lineman even appearing on the injury report, and they were both Probable and played.

18.3 AGL were enough to make San Diego's linebackers the most injured bunch in football, thanks almost totally to Shawne Merriman's knee and the 15 games it kept him out of. Arizona didn't name a single starting linebacker on the injury report all season.

My vote for Ed Reed as defensive player of the year came thanks to what surrounded him in the Baltimore secondary. To review quickly:

  • Reed starts at one safety spot. He's questionable for half the year, but plays all 16 games at an elite level. This will be more impressive in a minute.
  • Dawan Landry starts beside him. He suffers a scary spinal injury in Week 3 and misses the rest of the season.
  • Chris McAlister starts at one corner. He gets benched after a poor performance against the Colts, and has consecutive weeks where he goes from Questionable to Doubtful to Out to Injured Reserve with a knee injury. He ends up playing seven games.
  • Samari Rolle starts at the opposite corner. He misses five games with a neck injury before returning for good in Week 10.
  • Rolle's replaced by Fabian Washington. Washington goes down in the Titans game with a shoulder injury and misses Week 6.
  • Washington's replaced by Frank Walker. Frank Walker does not get hurt, which forces the Ravens to continue playing him.

Think it was fun being Ed Reed last year? The Ravens ended up with 36.5 AGL in the secondary last year, just barely beating out the Lions' 36.3. There was only a slight difference in performance. The Jets' starters were perfectly healthy, the only team in the league to be as such.

Although we don't count special teams totals in AGL, if we did, the Steelers and Vikings would be tied with 16 AGL each, thanks to Daniel Sepulveda and Heath Farwell, respectively.

The good news is that thanks to the work of Chris Povirk, Devon Teeple, and our newest batch of interns, we'll have even more data to work with in the near future. We're expanding our injury database all the way back to 1996, and we've gathered more information about specific injuries in order to provide more accurate details on how long it takes to heal and the effects they have on players and teams. We're excited to present that information, and we'll be doing so in the coming months.

Posted by: Bill Barnwell on 10 Mar 2009

60 comments, Last at 13 Mar 2009, 3:15pm by Chris

Comments

1
by bgrimm420 :: Tue, 03/10/2009 - 1:41am

First - Are these values adjusted in any way for players that sustain major injuries in the last game of the season? e.g. Carnel Williams for the Bucs in the last game of 2008.

2
by Dave from Oz (not verified) :: Tue, 03/10/2009 - 2:10am

Anyone else think that the trainers in the photo for this article look like they have had a terrible fake tan accident?

Oh, and good article and all that...

4
by Keith (not verified) :: Tue, 03/10/2009 - 5:05am

I thought that was Shanahan. Seriously.

5
by Keith (not verified) :: Tue, 03/10/2009 - 5:07am

Except now I see a Seahawks uniform. Go me.

3
by Aaron (not verified) :: Tue, 03/10/2009 - 2:34am

I wonder if there is any correlation between individual teams and injuries over a longer period of time. It would be interesting to see if there are some teams that are consistently healthy/unhealthy. Injuries might not be completely random and may be a result of the franchise's environment(scheme, facilities, practice techniques, surface, trainers, etc.) Looking at the Eagles and Rams, they both suffered about the same amount of injuries over the last two years but the Rams had over twice as many adjusted games missed.

6
by Theo :: Tue, 03/10/2009 - 5:46am

How do injured backups count when the starter is down? In the case of the Steelers running backs; Willie Parker went down, then Mendenhal.

8
by JasonK :: Tue, 03/10/2009 - 8:41am

From the article, right below the chart:

"These numbers are strictly the figures for starters, those players that replaced them, and players who get significant playing time without starting (think someone like Jonathan Stewart, Adam Terry, or Joselio Hanson)."

(Grammar rodeo note: the "that" should be a "who.")

7
by Temo :: Tue, 03/10/2009 - 8:25am

At some point, you have to wonder if Bill Parcells has some kind of system that reduces injuries. The Cowboys were relatively injury-free as well in the years he was here as well as the year after he left. The years before and last year they had relatively normal amounts of injuries.

And specifically, it seems like it's the recurrent, nagging injuries that are absent, as plenty of players have had season-ending catastrophic injuries that require surgery. It just seems that no one ever comes up with a pulled groin or a strained hamstring with Parcells around.

It may just be that he has a preference for drafting/signing players with good injury history, as many of the players who were injured last year for the Cowboys were not "Parcells Guys" (Romo excepted, obviously).

"Then again, I'm a Bobby Carpenter believer." -- Barnwell

19
by MJK :: Tue, 03/10/2009 - 11:44am

Or it may be that Parcells browbeats and bullies players into playing through injuries and pain more than some coaches do.

28
by Purds :: Tue, 03/10/2009 - 4:14pm

Or that he quickly cuts those with nagging injuries.

9
by Russ (not verified) :: Tue, 03/10/2009 - 8:57am

How are the Browns not up there for QB AGL? We had to pick up an play Bruce Gradkowski, our 4th QB of the year. If your system doesn't show that as a problem, it needs some tweaking.

11
by JasonK :: Tue, 03/10/2009 - 9:22am

The stat is additive over the entire season. I'm sure that Cleveland is high in the QB injury rankings, as the Browns finished the year with the most decimated QB roster, but they also had their #1 guy healthy for more than half of the season. KC and DET lost their starters within the first few weeks (and their backups shortly thereafter).

10
by Karl Cuba :: Tue, 03/10/2009 - 9:21am

How is Dwan Edwards considered a starter? Price, Ngata, Gregg, Suggs, so where does Edwards start?

14
by dcaslin (not verified) :: Tue, 03/10/2009 - 9:56am

The Ravens play a 3/4. The original 3 (I think) were Pryce, Gregg and Edwards. Gregg goes down and Banaan replaces him. Edwards goes down and the Ravens pick up Marques Douglas to replace him.

18
by Hurt Bones :: Tue, 03/10/2009 - 11:34am

Haloti Ngata, Batman! I think you're forgetting someone.

22
by Disco Stu (not verified) :: Tue, 03/10/2009 - 12:29pm

+1 to you Hurt Bones... that one def got a laugh.

26
by math_geek (not verified) :: Tue, 03/10/2009 - 2:45pm

It's a flaw in the AGL system for sure... But it's a very logical one. Dwan Edwards is listed as a starter because he started 13 games in 2007. Now, the only reason he started 13 games in 2007 is because starter Trevor Pryce got hurt, however I believe from the quote "AGL uses the injury report, the status and relevance of a player to his team, and historical data regarding how likely a given player is to play to produce a measure of how dramatically a team is affected by injury" that they are trying to use simply data, so something like 13 games started in 2007 might be one of those things that tricks a data-based algorithm into thinking you're a starter.

Alternatively, the fact that Edwards is a heavily used rotational defensive lineman might qualify him in the metric as a "starter" but I doubt it.

12
by JasonK :: Tue, 03/10/2009 - 9:25am

The breakdown by position is interesting, but didn't PFP08 have AGL broken into Offense and Defense? I think that would be more useful-- any chance that Bill or some other Outsider could post it in the Comments?

13
by Kaveman :: Tue, 03/10/2009 - 9:48am

If we weighted starters and reserves equally, the Broncos would have ended up with 57 AGL -- no team in the eight years we have data for have had more than 38.4 AGL.

This does nothing to take away my still continuing stunned response to the Broncos coaching change. And Denver was still 5th in offense DVOA (4th in rushing DVOA!) even with 7 different RBs.

I wonder where Shanahan might end up coaching next year. There should be no shortage of suitors.

15
by JimZipCode (not verified) :: Tue, 03/10/2009 - 10:21am

The Ravens took some hits on offense too: Marshall Yanda, Demetrius Williams, Quinn Sypniewski, Adam Terry, Willis McGahee. Somehow Derrick Mason played thru the separated shoulder, so I guess it doesn't count; and Jared Gaither played too, thru various ailments; but those guys were nicked.

Really a very impressive season for them.

36
by jonnyblazin :: Tue, 03/10/2009 - 7:11pm

They were also playing with their 3rd string QB for the entire season. The QB competition in up until the 3rd preseason games was basically between Boller and Troy Smith, Joe Flacco was well behind them on the depth chart. He only started because the Ravens had no choice.

16
by Purds :: Tue, 03/10/2009 - 11:30am

Did the numbers start with opening day, or did you account for losses in preseason? Just curious.

17
by are-tee :: Tue, 03/10/2009 - 11:33am

"The Ravens ended up with 36.5 AGL in the secondary last year, just barely beating out the Lions' 36.3. There was only a slight difference in performance. The Jets' starters were perfectly healthy, the only team in the league to be as such."

I think Eric Smith was technically the Jets' starting strong safety went he went out with a concussion (and suspension for his hit on Boldin on the same play.)

20
by MJK :: Tue, 03/10/2009 - 11:48am

Nothing that people don't already know, but I think you really need to adjust for quality of player to make this really meaningful. For instance, I bet most folks would agree that losing Tom Brady for the entire year is probably worse than losing both Brodie Croyle and Damon Huard.

Actually, QB might be a position where it makes sense to only consider injuries to the starter... Losing two CB's is much much worse than losing one, because it reduces your depth and weakens your entire secondary. But the dropoff from a backup QB to another backup QB is not that large, so losing you starter and your backup isn't really that much worse than just losing your starter.

25
by NCE (not verified) :: Tue, 03/10/2009 - 2:38pm

Thats not always true. The drop off from Seneca Wallace to Charlie Frye was probably bigger than the drop off from Hasselbeck to Seneca. The drop off from Brady Quinn to Ken Dorsey was substantial as well and Kansas City likely had their best QB third on the depth chart.

56
by RickD :: Thu, 03/12/2009 - 1:13pm

Not "always true", but almost always true, and that's what we want from a statistic.

Yes, some teams have Steve Young on their bench, but actually, only one team in NFL history has had a backup QB as good as Steve Young. Discard the outliers!

21
by MJK :: Tue, 03/10/2009 - 11:55am

Sorry for consecutive posts, but I had another thought.

Does it make much sense to include rookies lost at the start of the year in this evaluation? After all, no one knows how good the rookie will be. For example, we know now that if Jerod Mayo or Joe Flacco had gone down in week 1, it would have been a major loss for the Pats or Ravens, but if Vernon Gholston had, well, it really wouldn't have mattered all that much.

With a rookie, even if they're penciled in as a "starter" on opening day, you never know if they're going to turn into a quality NFL player or not. So if they miss significant time their rookie year, we have no way of knowing what they cost their team by being injured.

23
by bubqr :: Tue, 03/10/2009 - 2:21pm

I think that involving "value" of players is way too tricky.

24
by Bill Barnwell :: Tue, 03/10/2009 - 2:37pm

Some thoughts/comments:

- No, we don't factor in injuries that happen in Week 17, because really, there's so few of them that really matter. How many starters go down with an injury in Week 17? I can only really think of two -- Cadillac and Shaun O'Hara a year ago. We don't track injuries that happen in the middle of games in any other week, so I'm comfortable with missing a couple of Week 17 injuries here and there.

- Absolutely, I think there are teams that are better at preventing and rehabilitating injuries than others. I also think there are teams that do a better job of identifying, acquiring, and employing sturdier players, and Bill Parcells' teams are one of them.

- We have the Baltimore starting defensive linemen as Edwards, Gregg, and Ngata, as Edwards started all of last year when Pryce was out. Therefore, we have Edwards and Gregg going down for the year with injuries as starters, and Pryce coming in with the "NEW STARTER" designation. If this is grossly inaccurate, please let me know.

- Guys like Gaither and Mason absolutely are included in the totals, but because they were listed as Questionable, their totals aren't as strong as someone who's Out or on IR.

- I was under the impression that Abram Elam was the starting safety in New York, not Eric Smith.

- Browns were fifth in the league in QB AGL, behind KC, DET, NE, and CIN.

- It's absolutely meaningful without adjusting for quality of player, since we already adjust for quality of player between starter and reserve. I expect it'll be even more meaningful once we are able to adjust for quality of starter. Not sure what your point about quarterbacks are -- absolutely, there are teams that suffer mightily when they're forced to go to their third-string quarterback than their second-string quarterback. Look at Cleveland.

- If a rookie's good enough to win a starting job and then he gets hurt, he's good enough to be included.

27
by math_geek (not verified) :: Tue, 03/10/2009 - 2:56pm

Only reason Trevor Pryce wasn't starting in 2007 is because he was injured in Week 2. He tried to come back for 1 game in the middle, but was lost for the season with a torn Tricep.

Trevor Pryce is definitely the starter. Dwan Edwards is his backup. He'd still probably count for some points as a rotational guy, he would have seen all of Marques Douglas' time, as the Ravens traded for Douglas because they lost Edwards.

29
by bballer2294 :: Tue, 03/10/2009 - 4:55pm

Pryce is the started on the Ravens when healthy, but Edwards get a significant amount of time being the number 1 player coming off of the bench in D-Line. (He sometimes starts even, being he is better against the run than Pryce).

30
by Jacob Stevens (not verified) :: Tue, 03/10/2009 - 5:21pm

When you said you stopped counting Seattle wideouts after Koren Robinson came in, does that mean that their injury situation was actually worse than the metric indicates?

Shawne Merriman's year was counted, but it sounds like Burleson's year was only counted until a replacement came along that stuck. I don't understand the difference?

31
by Jacob Stevens (not verified) :: Tue, 03/10/2009 - 5:33pm

I think I understand, now. You don't count the replacement's replacement as a starter. That's understandable.

You guys have a few Seahawks fans on staff, so that might have made it possible -- or even dictated -- to slice that one open a little deeper. But if it wasn't too difficult to distinguish between a non-starter getting injured, and a non-starter who assumes a starting role due to injury subsequently getting injured, I think it would be great to include those numbers.

I see that the momentum already sounds to be headed towards making a greater distinction between quality of players, and I thoroughly understand why. On the other hand, particularly when evaluating positions, or sides of the ball, as opposed to overall, I think it would be more valuable to look at time lost by designated starters, as in the injury replacement getting injured ala Seattle wideout.

I'm not going to split hairs between dropoffs from starter to replacement and replacement to 2nd replacement, but obviously multiple iterations of replacements at one position is roughly multiple standard deviations away from league average quality at the position, whereas a taxonomy according to quality that is objectively applicable league-wide is something I imagine would be difficult not to just pull your hair out trying to identify, and it would STILL be guaranteed to be criticized ruthlessly once published.

33
by Jimmy Oz (not verified) :: Tue, 03/10/2009 - 5:57pm

Jacob, you understand nothing becuase Burleson was a starter or no.3. Thats a solid contributor. Robinson was signed 16 September. Thats a street free agent.

Unless "We don't count games missed by a starter if they sign a guy off the street" was in the preamble and i missed it.

46
by Jacob Stevens (not verified) :: Wed, 03/11/2009 - 2:19pm

I suppose you have a point about #3 wideout in a WCO being a solid contributor. Generally speaking. Not all #3 WCO wideouts contribute, of course.

My initial confusion was about whether the number of games counted that initial starters and Burleson & Obomanu lost were stopped after permanent replacements came in, OR

whether the total number of games lost by the initial starters were completely counted, and the designated starters for the season who went down on injury (Obomanu, and then Burleson & Payne, who actually started the season) were only counted as starters until the starters Branch & Engram returned, and then would have been counted as #3 contributor games lost, except that Koren Robinson finally replaced them, and it wouldn't be fair to count Seahawk wideout replacement games lost and not other team's replacement players' games lost, solely because the Seattle replacement replacement came in off the street, compared to other teams having the guy on the active roster, so they decided to compromise and count the games los by replacements, until starters/permanent replacements came in.

And to fortify that approach, consider that once Seattle pulled the trigger and brought Koren in, even as a spot contributor who wasn't rightly conditioned, when everyone was healthy he would likely be higher on the depth chart than Obomanu, and definitely higher than Payne.

Once I conceived of their approach, I noted that I think I figured it out, and my guess is you're still understanding it as the first described approach. I guess Bill could clarify, if he cares to.

I guess this is all relevant only because as Seahawk fans we want to believe that our 4-12 season can be accounted for by the WR devastation. For overall analysis of trends, and injury impact, it's not necessary, but for that badge of pride that says we were better than what happened last year and to found expectations of bounceback, I think we're more interested in the true impact of the actual carnage, and how that fits in with the rest of the league. At least, that's my vested interest.

51
by Jimmy Oz (not verified) :: Wed, 03/11/2009 - 6:53pm

Robinson wasn't active until the week 6 Packers game, when he wins his place by beating out Colbert, Bumpus, and McMullen.

It doesn't fit with the guidelines FO gave at the start of the piece. Its games lost by starters and their replacements, and Burleson was either a starter or the replacement of a starter as he's listed as a starter with Branch and Engram out in the official game book.

But by looking at who was active for the Green Bay game, the reason FO have given to stop counting Burleson's missed games (with Branch also out) is because a street free agent beat out Colbert, Bumpus & Billy F'ing McMullen.

Payne and Obomanu have nothing to do with this.

38
by Travis :: Wed, 03/11/2009 - 1:48am

How many starters go down with an injury in Week 17?

Ben Roethlisberger?

I was under the impression that Abram Elam was the starting safety in New York, not Eric Smith.

Smith started 5 of the Jets' first 7 games. (Exceptions: the Week 3 game against the Chargers, when he was benched for David Barrett, and the Week 6 game against the Bengals, which he missed due to suspension for the hit on Boldin.) Elam started that game against the Bengals, and took over as the full-time starter after Smith suffered a second concussion in the Week 8 game against the Chiefs.

47
by are-tee :: Wed, 03/11/2009 - 3:27pm

"I was under the impression that Abram Elam was the starting safety in New York, not Eric Smith."

Elam became the starter when Smith went out with his concussion/suspension. But he sucked even when healthy.

32
by Jimmy Oz (not verified) :: Tue, 03/10/2009 - 5:45pm

"Burleson couldn't make it out of the first week. Once Koren Robinson took over as a starter, we stopped counting Burleson"

Why did you stop counting Burleson when a chronically injured wash-out at three teams with bad hands was signed as a stop gap measure and started over injured players? Is it because Chewbacca lives on Endor?

34
by mike (not verified) :: Tue, 03/10/2009 - 6:11pm

Does adjusted games lost account for starters lost in the pre-season (e.g. Antwan Peek and Daven Holly of the Browns)?

35
by Will Allen (not verified) :: Tue, 03/10/2009 - 7:06pm

Maybe I missed it, but are suspensions counted? The Vikings didn't miss Mckinnie any less for the first four games because he was suspended instead of injured, and the same will be true of the Williams Wall if they lose their lawsuit.

If Mckinnie's four missed games aren't already county, I'm even more impressed with the Vikings finishing 10-6, with a high number of injuries, against a harder than average schedule, and Gus Frerotte taking snaps for most of the season. They really were extremely good on the line of scrimmage.

37
by Benn (not verified) :: Tue, 03/10/2009 - 11:55pm

As always, this is fascinating work. It does occur to me that the teams on the very bottom of the list (DET, STL, SEA, CIN) are disproportionately dreadful teams, and that their dreadfulness may be not only a consequence of their injuries, but also a cause. I wonder if there is a way to look at the injury records of teams that are out of the playoff race and see if they seem to rack up injuries at higher rates than do those teams that are still competitive. Anecdotally, it has seemed to me that when the teams I follow are firmly out of the playoff race, the number of players who suddenly need season-ending treatment seems to escalate quickly, but I'd be curious to know if that actually happens.

39
by ammek :: Wed, 03/11/2009 - 8:21am

I wondered the same thing. Perhaps the data running back to 1996 will help build a correlation between bad breaks and bad teams. You have to wonder if, say, Peyton Manning had been on the 2008 Bengals and Carson Palmer on the 2008 Colts if their games-missed totals would not have been reversed.

I've seen the same argument applied to fumbles: RBs, in particular, seem to fumble more often in 'down' years than in 'up' years. And while their fumbles may contribute to losses, they may also be a consequence of playing from behind: over-thinking, trying to do too much, selling out for that extra yard.

41
by tuluse :: Wed, 03/11/2009 - 10:04am

I doubt that would be the case with Palmer and Manning. 1) The Colt's line was terrible until Saturday got back, and still wasn't good afterwards Palmer would have got killed back there. 2) Manning is a robot.

49
by Jacob Stevens (not verified) :: Wed, 03/11/2009 - 4:40pm

Dr. Z once said something along those lines. One of those articles from the POV of an anonymous, journeyman lineman, who doesn't like the self-aggrandizing "motivation" of guys like Ray Lewis, or the flak from the QB when there's a little bit of pressure.

His character said it's obvious. Bad teams, less talented teams, you know you have to pick up their slack. You have to try that much harder, lay it all out there to try and get just a little success, and you open yourself up for more injuries. So more guys go down, the team is less talented, and if you're going to do anything at all, you have to dig in and find that much more effort.

It makes sense. I think there's a bit of a feedback correlation between the two, but would expect that injuries are more of a cause for teams being bad than poor talent is a cause of more injuries. Who knows the ratio. 65-35?

In defense of my Seahawks, I like to believe it was more about injury than initial lack of talent, but in general I could see that lack of talent can be a contributing factor.

40
by Justin Zeth :: Wed, 03/11/2009 - 9:36am

It would seem like it would be more helpful if there were SOME way to adjust for which players are actually injured (though doing it this way is certainly a good start, and more useful than nothing.) Lions starters are not the same quality player as Giants starters. The Steelers ranked as one of the most injured teams in the league and won the world championship; how do you explain that? Well, it's because all of their important players stayed healthy all year. Almost all of the Steelers' AGL are players that suck anyway (Marvel Smith, Kendall Simmons) and missed almost the entire season or players that aren't that important (Willie Parker, Rashard Mendenall) missing big chunks of time. Contrast that with the Cowboys, for instance, who weren't terribly injured (middle of the pack in ACL) but lost their best and most important player for four games, which is season-altering (ask the Patriots).

Then again, I don't know that there's any real utility in trying to figure out a way to properly adjust for quality of players lost; it's something the reader can glance over the list and decide for himself pretty easily...

42
by Danish Denver-Fan :: Wed, 03/11/2009 - 10:26am

Good stuff.

As a Broncos fan I can't help but think that the Denver O-line was the healthiest. I don't recall any injuries at all... Of course "probables" don't stand out to you as much as "outs" do. Still i can't remember any, and Clady-Hamilton-Wiegmann-Kuper-Harris are all 16/16. Unless you're counting Tom Nalen? If you are, then that's definately worthy of a subjective tweak, Casey Wiegmann's pro bowl performance considered.

48
by An Ominous (not verified) :: Wed, 03/11/2009 - 3:40pm

Why does that deserve a subjective tweak? Starting center, himself a pro bowl player and potential HoFer, lost for season. Just because Denver had the forethought to secure an exceptional backup doesn't change the fact that the starting center was lost for the season. The reason Nalen was the starter is because the team believed he was better than Wiegmann, so even if Wiegmann played well, it still counts as a loss. If Champ Bailey gets lost for the season and some rookie we just drafted winds up coming in and playing at a pro-bowl level, should FO not count the fact that Denver lost Champ Bailey? When Bledsoe went down in 2001, Tom Brady played very well... but the Patriots still lost their starting QB for 14 games.

50
by Danish Denver-Fan :: Wed, 03/11/2009 - 6:42pm

Sure, I agree.

But it was definately my impression, that Nalen was never likely to play a down. If, say, Brady doesn't get ready for the start of this season, should he count for an injury suffered a year earlier? The days where the report from camp was a 100% healthy Nalen, could be counted on one hand.

But you're right, probably no tweaking required. I just don't get how Denvers 'line isn't the healthiest, then. Oh well, I'll have to wait for PFP '09.

43
by E :: Wed, 03/11/2009 - 11:01am

Great work guys. There are 2 big questions that commenters raised that I don't think Bill answered, which would be really helpful in understanding these numbers:

1. How are pre-season injuries factored in? For example, did the Giants lose 16 games on the D-line when Osi went down for the season?

2. How are "starters" determined? Bill's comments about the Baltimore D-line seem to indicate that this is based solely (or at least mostly) on 2007 starts, regardless of 2007 injuries, though to me this makes no sense. (e.g., is Anthony Gonzalez listed as the 2008 starter over Marvin Harrison?)

45
by woefully inappropriate (not verified) :: Wed, 03/11/2009 - 11:35am

3. How are pre-season or season injuries counted that don't make the injury report, because the individual is not on the team? It (hopefully) shouldn't happen, but it did with the Jaguars last year, and could have happened if Winslow's and Roethlesburger's non-football injuries from a couple of years ago led to their team dismissing them outright.

44
by bengt (not verified) :: Wed, 03/11/2009 - 11:33am

Methinks the Steelers' special teams AGL does not include the torn ACL of long snapper Greg Warren after about half the season. Reserve punter Mitch Berger missed a couple of games, too. (I assume that an AGL of 16 can be fully attributed to Sepulveda missing the complete season, as you insinuate in the text.)

52
by BD (not verified) :: Wed, 03/11/2009 - 7:34pm

As a Bengals fan, seeing Cincinnati at the very bottom of the AGL list is at once encouraging and demoralizing. It means our ineptitude this past season might have been due to injuries more than anything else (I know it's unlikely, but a fan's got to hope, right?), but why on earth did we have so many?

53
by greybeard (not verified) :: Thu, 03/12/2009 - 1:31am

How did Kansas City loose 21.4 games in the quarterback position in a 16 games season? Do they normally play more than one quarterback? Did they secretly make playoffs, won Superbowl and then played another 2.4 games in some secret location with some secret teams?

54
by Bobman :: Thu, 03/12/2009 - 4:08am

What? You mean to say you missed that? Those extra games, and the post-game press conferences were Herm-a-licious.

I'll assume QB1 = 16 games before the season starts. He goes down in week 2 (guess) and that means 14 GL. The next guy counts only if he is close in talent level maybe... or a former starter... and when goes down 5 weeks later, that's 8 GL. It's a guess. If it was the dropoff from Manning to Sorgi and Sorgi gets hurt, it would not register. But in KC, they were rated as close. As I say, a guess. About as likely as yours;-0

59
by greybeard (not verified) :: Thu, 03/12/2009 - 9:11pm

I guess you are right about how they calculate it. But that smells bullshit to me.

If the second QB is close to first one in talent than they should not count 14GL if the first QB gets injured second week. They should count it as 2GL or whatever the talent difference between the two QB amounts to in 14 games. The (regular) season is 16 weeks long, there is one QB on the field at a time, so the most you would loose as game lost from a starter is 16 weeks. And it would be only 16 GL if somehow that team forget to sign a second QB and are playing a kicker in QB position.

57
by RickD :: Thu, 03/12/2009 - 1:18pm

"At quarterback, Kansas City (21.4 AGL) finished just ahead of Detroit, thanks to injuries to both starting quarterback Brodie Croyle and backup Damon Huard, who assumed the starting job for a short period of time."

That explanation is copied verbatim from the article.

58
by greybeard (not verified) :: Thu, 03/12/2009 - 9:02pm

I *can* read. Did you think that I had read the first 5 words in that sentence and immediately started typing?
Tell me how that explanation convinced you.

55
by Rich Conley (not verified) :: Thu, 03/12/2009 - 9:51am

"It's also worth noting that Bill Belichick did not put Matt Cassel on the injury report once all season -- maybe that Tom Brady shoulder injury was serious after all. "

i'm really getting sick of this sort of stuff. Belichick is required by the NFL to put Brady on the injury report every week because he doesn't participate in certain drills. If a player is "withheld from practice" they must be put on the injury report. I expect better from you guys.

(now, preventing more shoulder damage is what he holds Brady out of the drills for)

60
by Chris (not verified) :: Fri, 03/13/2009 - 3:15pm

When Bill Parcells went to Miami he would go in the locker room before the season started and talk about reducing injuries. He slammed that mentality into players heads and pushed for more conditioning. He made a commitment to it and it worked. He let players know that it was not acceptable to use injuries as an excuse. Now if your leg snaps your leg snaps, but players are often known for trying to sit out practice and ride that injury out.

There are also different programs you can use in the gym ( The redskins used HIT training) which requires less weight in the weight room and will keep you healthier than traditional weight lifting.

I do think that teams out of the playoff hunt do see more injuries because borderline injuried players have an excuse to throw in the towel, and also teams don't want to risk further damage to star players so they more or less allow it. Teams throw in the towel.

If you don't believe me, then why do players in all sports perform better during contract years, and worse AFTER they sign the big contract? Check out the injury history for guys in their final year of thier contract, vs guys int the first year of new contracts and see injuries go up dramatically.