Stat Analysis
Advanced analytics on player and team performance

2019 Adjusted Games Lost: Part I

New York Jets TE Chris Herndon
Photo: USA Today Sports Images

The Football Outsiders injury report database now goes back nearly two full decades, and for most of that time injuries have steadily been rising. That may be changing, however, as it appears that the NFL has -- to use contemporary terminology -- flattened the curve when it comes to injuries. The news is not good for everyone, however; it was a very rough year on the injury front for the New York Jets, and for wide receivers across the league.

We have collected the data from the NFL's weekly injury reports for every season since 2001. This allows us to measure not just who played and who didn't, but who was able to play with injuries, even if those injuries meant the player was at something less than 100%. That's why we call this metric adjusted games lost -- in addition to players who missed games entirely, we also count those who hit the field after appearing on the injury report at an adjusted rate. Further, we track whether the injured player was a starter, a situational reserve, or simply bottom-of-the-roster fodder. Obviously, an injury to a starting tackle is more important than one to a guy who only plays on special teams.


Changes in the Process

As a reminder, players are listed as either questionable (theoretically meaning they had a 50% chance of missing that week's game), doubtful (75%), or out (100%) on weekly injury reports. In prior years, there was also a probable (25%) designation, but the NFL did away with that in 2016; since then, teams have left those players with a blank designation on the injury report, and they have ended up playing far more often than not.

This year, we have reevaluated our AGL calculations based on how a player is listed in the injury report, especially situational/reserve players. That includes going back over prior seasons and making slight adjustments to the earlier numbers before the probable designation was removed in 2016. We have also changed the numbers since then so they are based on the odds of playing in all four years instead of just in 2016. Those odds are:

Player Participation Rates, 2016-2019
Designation Missed Game?
Starters Reserves
Probable/Blank 2% 17%
Questionable 27% 47%
Doubtful 99% 99%
Out 99.9% 99.9%

The average team suffered 75.8 AGL in 2019, which is up from 74.1 in 2018, which was up from 72.2 in 2017. It's also nearly triple the average of 25.8 in 2001, the first year for which we have injury report data. In 18 seasons since, the average AGL has gone up from one year to the next 13 times, while only dropping five times. However, long-term trends show the NFL's injury numbers have plateaued. Over the five seasons from 2015 to 2019, the average NFL team has suffered 75.2 AGL; that's just a smidge lower than the average of 75.6 from 2014 to 2018. That tiny difference may not seem significant, but in the big picture it's enormous: it's the first time on record that the five-year average for AGL has ever gone down.

Average AGL by Team, 2001-2019
Year Avg. AGL Dif. 5-Yr Avg.
2001 25.8 -- --
2002 32.6 +6.8 --
2003 33.4 +0.8 --
2004 46.8 +13.4 --
2005 34.4 -12.4 34.7
2006 36.8 +2.4 36.8
2007 44.4 +7.7 39.2
2008 53.0 +8.6 43.1
2009 52.1 -1.0 44.1
2010 49.0 -3.1 47.1
2011 59.1 +10.1 51.5
2012 61.4 +2.3 54.9
2013 65.9 +4.5 57.5
2014 77.8 +11.8 62.6
2015 71.8 -6.0 67.2
2016 82.2 +10.4 71.8
2017 72.2 -9.9 74.0
2018 74.1 +1.8 75.6
2019 75.8 +1.8 75.2

We also went back and checked the numbers looking at starters only since 2002. This has nothing to do with any changes we've made subjectively in how we mark situational players, so it should be more consistent and accurate over the long term. Here it appears injuries are still on the rise -- the average team suffered 64.3 AGL to starters in 2019, the most of any year in our database:

Average AGL, 2002-2019, Starters Only
Year AGL
2002 31.5
2003 32.2
2004 42.2
2005 32.7
2006 34.4
2007 42.0
2008 45.1
2009 46.2
2010 42.2
2011 50.0
2012 51.1
2013 49.7
2014 54.4
2015 56.4
2016 62.1
2017 58.6
2018 57.3
2019 64.3

Team Results: J! E! T! S! Jets! Jets! Ow! My Knee!

The 2019 New York Jets were largely considered a joke, best known as fodder for memes about their coach's unusual appearance and mannerisms:

… or the feud between that coach and his big-money running back:

… or their quarterback's illness (and ESPN's ensuing graphic on Monday Night Football):

… but beyond those yuk-yuks was a nigh-unprecedented injury list. The Jets weren't just the most injury-struck team of 2019; they were the second-most injured team we've ever found.

2019 AGL Results
Team 2019 AGL Rk 2018 AGL Rk Dif DVOA
MIN 25.6 1 70.2 12 -44.6 15.3%
TB 39.5 2 105.6 30 -66.1 1.4%
TEN 43.6 3 64.7 11 -21.2 8.1%
DAL 51.8 4 75.2 17 -23.5 17.2%
ATL 53.2 5 75.9 18 -22.7 -5.3%
SEA 56.5 6 62.7 10 -6.2 13.5%
CLE 56.8 7 47.9 6 +8.9 -9.7%
NO 58.3 8 50.0 7 +8.3 29.3%
PIT 60.5 9 38.5 5 +22.0 -5.4%
LAR 60.6 10 37.9 4 +22.7 5.2%
BUF 64.6 11 31.0 2 +33.6 3.0%
IND 65.3 12 108.6 31 -43.3 -5.0%
CHI 65.4 13 34.5 3 +30.9 -2.1%
GB 65.6 14 89.7 22 -24.1 7.7%
CAR 66.2 15 102.6 29 -36.4 -26.7%
BAL 68.7 16 28.5 1 +40.2 41.8%
NYG 72.6 17 51.2 8 +21.5 -17.4%
KC 75.8 18 60.1 9 +15.6 30.2%
HOU 78.2 19 85.3 21 -7.2 -5.8%
CIN 83.2 20 97.9 27 -14.7 -25.3%
PHI 84.4 21 117.0 32 -32.6 6.6%
ARI 85.0 22 92.3 23 -7.3 -5.8%
NE 85.0 23 75.2 16 +9.9 30.8%
DET 87.8 24 74.7 15 +13.1 -12.0%
DEN 89.2 25 73.6 14 +15.7 -9.0%
OAK 93.4 26 77.3 19 +16.1 -11.6%
SF 95.8 27 97.2 26 -1.4 27.9%
MIA 100.0 28 93.1 24 +6.9 -37.4%
JAX 100.2 29 102.5 28 -2.3 -17.8%
LAC 103.0 30 83.0 20 +20.0 -6.8%
WAS 131.0 31 95.6 25 +35.3 -25.6%
NYJ 160.1 32 70.4 13 +89.6 -15.6%

The Jets' total of 160.1 AGL is the second-most for any team we have ever measured, behind only the 171.6 AGL of the 2016 Chicago Bears. (We listed those Bears with 155.1 AGL in our original article, but our calculations and methodology have shifted slightly since then.) New York's year-to-year jump of +89.6 AGL is by far the largest we have ever measured, shattering the record of +74.5 set by the 2008 Detroit Lions that went 0-16. Their 76.4 AGL on offense is the eighth-highest we have ever measured; their 83.7 AGL on defense is the fourth-highest. The injury bug hit the Jets everywhere in 2019 -- they were among the top ten teams in AGL by quarterbacks, wide receivers, tight ends, offensive linemen, tight ends, and defensive backs. But they were hurt most of all at linebacker, where they suffered an all-time record of 48.5 AGL, surpassing the 42.2 mark of the 2017 Chicago Bears.

And yet, for all of that comedy, and all their injury struggles, the Jets finished with a respectable 7-9 record. That includes -- and I never would have guessed this if I hadn't looked it up -- a winning record in games started by Sam Darnold. Now, there was a lot of good fortune in those seven wins. The Jets went 5-1 in one-score games, and their only victory against a playoff team came against a Buffalo Bills squad that was resting its starters in Week 17. The Jets were only 25th in Pythagorean wins, so there is certainly room for improvement. But if they can just get average health in 2020, in an AFC East where no current quarterback has ever won a playoff game? The window for the Jets to sneak into the playoffs may be open wider than it has been in a long time.


Offense vs. Defense

The following table shows each team's AGL on offense and defense, and the difference between those two numbers:

Offensive and Defensive AGL, 2019
Team Offensive AGL Rk Defensive AGL Rk Dif
BAL 9.7 1 59.1 31 +49.4
DAL 11.3 2 40.5 21 +29.2
MIN 11.7 3 13.9 2 +2.1
CLE 15.4 4 41.4 23 +26.0
ATL 15.5 5 37.7 18 +22.2
TB 17.0 6 22.5 4 +5.6
TEN 19.7 7 23.9 6 +4.2
NO 23.2 8 35.1 17 +11.8
LAR 29.8 9 30.8 15 +1.0
PHI 30.3 10 54.1 29 +23.8
PIT 30.6 11 30.0 14 -0.6
SEA 31.5 12 25.0 7 -6.5
KC 32.2 13 43.6 24 +11.4
CAR 36.5 14 29.7 13 -6.8
CHI 36.9 15 28.5 12 -8.4
IND 38.6 16 26.7 9 -11.9
GB 39.5 17 26.2 8 -13.3
DEN 41.0 18 48.3 27 +7.3
MIA 43.0 19 57.0 30 +14.0
ARI 45.6 20 39.4 19 -6.3
BUF 46.5 21 18.1 3 -28.5
NYG 49.0 22 23.6 5 -25.4
SF 49.5 23 46.4 26 -3.1
HOU 50.5 24 27.7 11 -22.8
OAK 52.4 25 41.1 22 -11.3
JAX 55.4 26 44.8 25 -10.6
CIN 55.9 27 27.3 10 -28.6
DET 56.5 28 31.3 16 -25.2
LAC 62.7 29 40.3 20 -22.5
NE 73.7 30 11.4 1 -62.3
NYJ 76.4 31 83.7 32 +7.2
WAS 81.9 32 49.1 28 -32.9

The two teams with the most disparate results on the two sides of the ball just happened to be two of the top three seeds in the AFC. The Baltimore Ravens had the fewest injuries on offense but the second-most on defense, while the New England Patriots had the third-most injuries on offense but the fewest on defense. Each team placed in the top five since 2001 in difference between offensive and defensive AGL, but their results are hardly unprecedented -- just last year, the Buccaneers had a bigger swing in favor of healthy offense, while Washington and Jacksonville had bigger splits in favor of healthy defense.


Other Notable Results

The team with the most injuries on offense (and second-most overall) was Washington, although that comes with an asterisk. Though he was on injured reserve for the entire season, Alex Smith was still on Washington's roster, and if he had been healthy you can be sure he would have been starting. So he counts as 16 AGL by himself … but there was no point in the calendar year 2019 when Washington seriously thought Smith was going to be on the field. Smith's horrific broken leg suffered in Week 11 of 2018, followed by a severe infection and 17 surgeries (yes, 17), effectively removed him from Washington's plans. That's why they signed veteran Case Keenum in free agency and drafted Dwayne Haskins. They had many other injuries -- notably at running back and tight end -- but if you remove Smith from the equation, then Washington becomes the No. 2 most injured offense of the year behind the Jets. Regardless, none of this is anything new for Washington -- they have now ranked 22nd or worse in AGL six years in a row, finishing last in 2017 and next to last in 2015 and 2019.

As you'd expect, the teams with the fewest injuries generally had good seasons. The leaders in AGL on offense (Baltimore, 9.7), defense (New England, 11.4) and overall (Minnesota, 25.6) all made the playoffs. Both Super Bowl teams, however, finished below average. Kansas City was in the top six for most AGL at both defensive line and linebacker. San Francisco had the third-most AGL at running back (first if you ignore fullbacks) and was also beat up at defensive line, linebacker, and defensive back. Overall, the correlation coefficient between AGL and DVOA was -0.393 -- generally, teams that suffered more injuries played worse on the field. That's not always true, of course; the Ravens and Panthers were nearly tied in total AGL, but one of them had the best record in the league while the other ended the year on an eight-game losing streak.

We also need to discuss the Tampa Bay Buccaneers. In 2018, the Bucs suffered 91.4 AGL on defense, the most we have ever measured. Last year, Tampa Bay's defensive AGL dropped to 22.5. That decline of 66.1 defensive AGL is biggest year-to-year improvement on record, surpassing the 61.0-AGL decline the New York Giants enjoyed from 2015 to 2016. We started to put together a table of the Jets, the Bucs, and other teams that had radical increases or declines in defensive AGL, but surprisingly there wasn't a substantial effect on DVOA one way or another. Yes, more injuries usually means worse performance, but there are enough exceptions that the total figures don't paint a compelling picture.


2019: A Bad Year for Pass-Catchers

Finally, we must address the brutal year that pass-catchers had last season. The average NFL team suffered 10.0 AGL to wide receivers and 6.6 AGL to tight ends, both the highest marks on record. No team came close to the wide receiver record of 36.2 set by (here they are again!) the 2017 Bears, but many gave it a run. There have been 36 teams with at least 19.0 AGL at wide receiver; nine of them played in 2019. (The Indianapolis Colts led the league with 28.4.) The Jacksonville Jaguars did set a record with 31.3 AGL at tight end, passing the mark of 30.4 set by the 2013 Tampa Bay Buccaneers. Three other teams last year were at 17.0 or higher; three other teams were at 12.0 or higher.

Later this week, we'll look at the teams with the best and worst injury luck at each position in 2019.

Comments

11 comments, Last at 16 Apr 2020, 5:15pm

3 Oops! That was literally the…

In reply to by Aaron Schatz

Oops! That was literally the last part of this we were working on before this went live. I spent four months assuming the Pats were going to be a top-two seed, and then they went to three-seed and out of the playoffs so quickly I forgot it happened that way.

4 Diving into the causes of injury

This seems like a very fruitful area, though maybe you've looked at some of these questions in prior years. In particular we could learn what factors affect injuries:

 * Is there a correlation between age and AGL? I'm guessing there is, but how strong is it? Maybe older guys take better care of themselves, or maybe the 28 year-olds must be durable to survive that long while the rookies include many who can't make it through an NFL season.

 * Is there a correlation between offensive and defensive AGL? If injuries are random, there shouldn't be.

If there is, maybe both correlate with a third factor:

 * Is there a correlation between a team's prospects and AGL? Maybe players lose motivation when playoff chances dim. You'd need a proxy to represent the team's prospects, of course; maybe current record, projected record, projected playoff chances, or the remaining tenure of coaches and executives (if that data is available!).

 * Is there a correlation between weather and AGL? Do cold weather teams have more injuries? High-humidity teams? (Was Brady smart to move to Tampa or should he have chosen Vegas?)

 * Do some teams, or even some positions on some teams, have consistently higher or lower injury rates every year?

 * Is there a correlation between coaches and AGL? Head coaches? Assistants?

 * Is there a correlation between trainers and AGL? (Does even Football Outsiders track trainer data?)

 * Is there a correlation between schemes - offensive, defensive, or special teams - and AGL?

 * Maybe the most interesting one: Is there a correlation between opposing teams and AGL? That is, is it out of a team's hands - maybe it depends on how their opposition plays. Or maybe it depends on the opposing team's schemes, coaches, playoff prospects, or the opposing players' contract statuses. Imagine the implications of finding that certain teams or coaches cause more injuries !

5 My only question is a simple…

My only question is a simple one: does having Adam Gase as your coach lead players to be injured so they don't have to deal with him?

8 Combined with the post above…

Combined with the post above:

 * Is there a correlation between a team's prospects and AGL? Maybe players lose motivation when playoff chances dim. You'd need a proxy to represent the team's prospects, of course; maybe current record, projected record, projected playoff chances, or the remaining tenure of coaches and executives (if that data is available!).

I think you actually see this with Detroit. A few departed players have indicated they did not hurry back from injuries/missed early games because they were out of condition because they hated the coach and the front office.

6 A bit surprised to see the…

A bit surprised to see the Packers finish close to average rather than among the most fortunate teams. I wonder if G Lane Taylor, who was put on IR after Week 2, pumped up their number a little bit. He technically started the first two games, but was splitting snaps with rookie Elgton Jenkins, and there was a decent consensus that he was going to lose his job even before he was hurt - Jenkins turned out to be pretty good. Not at all a criticism of the methodology, but just thinking that the Packers might have been pretty fortunate relative to what an "average" season normally looks like and might still expect some regression to the mean in injury luck next season.

10 Love this

Last piece of my puzzle, thanks for the help guys! (aggregated, while have NOT weighted, but should, 3rd down rebound, fumble recovery, Pythag., Net close wins, (this one is my own) "net penalty gifts on 3Down" and now AGL: used 2 tables from FB, thanks boys, keep up the great work!