2019 Adjusted Games Lost: Part II
Earlier this week, we looked at each team's total adjusted games lost to see which clubs had the best (Minnesota) and worst (the Jets) injury luck in 2019. Today we'll break things down by position and find which teams suffered the worst (or most) injuries at each specific spot on the roster.
Offense
Quarterbacks
The average team suffered 2.8 adjusted games lost at the quarterback position in 2019 -- up from the 1.7 average we saw in 2018, but more or less in line with what we have seen over the last decade or so.
Washington suffered the most AGL among quarterbacks last season, but as we mentioned in Part I, that includes 16.0 AGL for Alex Smith and his broken leg, and Washington never really expected that Smith would be playing football in 2019. For the record, Washington's numbers also include 2.4 AGL for Case Keenum (1.4 AGL, foot/shoulder) and Dwayne Haskins (1.0, ankle).
Aside from Washington, the team with the most AGL at quarterback was the Pittsburgh Steelers, led by Ben Roethlisberger (14.0, elbow) and Mason Rudolph (1.0, concussion). Truly, it's hard to argue that any team had their season wasted by quarterback injuries more than Pittsburgh. The Steelers finished last in offensive DVOA, but still won eight games with Rudolph and Devlin Hodges at quarterback. Give that team 16 games of a healthy ben Roethlisberger, and it's easy to see them passing 9-7 Tennessee and getting into the playoffs.
Other teams with major quarterback injuries last year include Carolina with Cam Newton (14.0, foot); Detroit with Matthew Stafford (8.6, hip/back) and Jeff Driskel (5.0, hamstring); and Jacksonville with Nick Foles (8.0, clavicle).
Running Backs
These numbers include AGL for both halfbacks and fullbacks, and that's why New England finishes with the most AGL at the position. Between James Develin (14.0, neck) and Jakob Johnson (10.0, shoulder), the Patriots had the second-most AGL among fullbacks since 2001. (Normally we don't count fullbacks in AGL, but given how often the Patriots use them in their offense, they do count as important situational players.) The 2008 Bengals are the only other team with more than 20.0 AGL among fullbacks; that team lost Jeremi Johnson for 16 games with a knee injury while backup Reagan Mauia also missed significant time with a bad hamstring. Given the way fullbacks are disappearing from the NFL, their record is unlikely to be threatened again any time soon.
The Denver Broncos were another team that had significant fullback injuries, as Andy Janovich suffered 9.3 AGL with pectoral/elbow concerns. Receiving specialist Theo Riddick was also lost for the season with a shoulder injury.
If you ignore fullbacks, no team lost more AGL among running backs than San Francisco. Jerick McKinnon suffered a significant setback while rehabbing the ACL he tore in 2018 and was lost for the entire season for the second year in a row. McKinnon recently restructured his contract in hopes of returning to the field in 2020. Other 49ers running backs with injury issues last year include Matt Breida (2.6, ankle) and Tevin Coleman (2.3, ankle). And San Francisco is another team that relies heavily on fullbacks, and Kyle Juszczyk missed four games with a knee injury. The good news for San Francisco is that they had approximately 93 running backs on the roster, so they were able to weather the storm and reach the Super Bowl.
Other teams with major running back injuries last year, and some of the most notable players:
- Oakland Raiders: Isaiah Crowell (16.0, Achilles), Josh Jacobs (3.0, shoulder).
- Washington: Derrius Guice (11.0, meniscus/MCL), Chris Thompson (4.6, toe).
- Houston Texans: Lamar Miller (16.0, ACL).
Wide Receivers
It was a terrible year for wide receiver injuries -- the average NFL team suffered 10.0 AGL to wideouts, most of any year on record. And those injuries were spread far and wide throughout the league. There have been 36 teams since 2001 with at least 19.0 AGL at wide receiver; nine of them played in 2019. Here are those nine teams, from most injured to least, with some of their most notable injured players:
- Indianapolis Colts: Devin Funchess (15.0, clavicle), T.Y. Hilton (7.1, quadriceps/calf), Parris Campbell (6.3, abdomen/hand/foot).
- Cincinnati Bengals: A.J. Green (16.0, ankle), John Ross (8.0, shoulder), Auden Tate (3.3, knee).
- New York Giants: Corey Coleman (16.0, ACL), Sterling Shepard (6.0, concussion).
- New York Jets: Quincy Enunwa (15.0, neck), Demaryius Thomas (6.0, hamstring).
- Los Angeles Chargers: Travis Benjamin (11.3, quadriceps), Dontrelle Inman (7.0, quadriceps).
- New England Patriots: N'Keal Harry (9.0, knee/ankle), Cameron Meredith (4.0, unknown), Julian Edelman (3.3, chest/shoulder/knee), Josh Gordon (2.3, knee).
- Green Bay Packers: Equanimeous St. Brown (16.0, ankle), Davante Adams (4.4, turf toe).
- Philadelphia Eagles: DeSean Jackson (13.3, abdomen), Alshon Jeffery (4.6, calf/ankle/foot), Nelson Agholor (2.1, knee).
- Detroit Lions: Jermaine Kearse (16.0, broken leg), Marvin Jones (3.0, ankle).
Yes, that says "unknown" by Cameron Meredith's name. You can always count on the Patriots to make injury reports interesting. Meredith tore his ACL and suffered other knee damage in the preseason with the Bears in 2017. He played six games for the Saints in 2018, but went on injured reserve that November with a knee injury. The Patriots signed him last August but put him on the PUP list before the season started, then released him in October. He never played a game for the team. All of Meredith's New England-related news items on Rotoworld say that he suffered a knee injury, and there's no real reason to doubt them.
Tight Ends
Congratulations, Jacksonville Jaguars! You set a new record with 31.3 AGL by tight ends! The Jaguars broke the old mark of 30.4 held by the 2013 Tampa Bay Buccaneers, a team that put Nate Byham, Tom Crabtree, and Luke Stocker on injured reserve with knee, biceps, and hip injuries, respectively.
The Jaguars' problems started in the preseason, when third-round rookie Josh Oliver hurt his hamstring. He didn't return to the field until Week 7, then suffered "small fractures" in his back against Indianapolis in Week 11 and was out for the year. Veteran James O'Shaughnessy tore his ACL against Carolina in Week 5; Geoff Swaim went on IR with an ankle injury in Week 8. Seth DeValve was signed off waivers in September, then missed four games in the second half of the year with an oblique injury. Jacksonville's active tight ends in Week 17 against the Colts: DeValve; seventh-round draftee Ben Koyack; and Charles Jones, an undrafted rookie pulled off the practice squad in November.
Elsewhere, Washington lost both Jordan Reed and Vernon Davis to concussions for most of the year. (Reed missed the entire year after taking a hit to the head against Atlanta in the preseason. Davis didn't play after Week 4, though he was still on the active roster and not placed on IR until Week 13.) Buffalo's Jason Croom missed the entire season with a hamstring injury, while Tyler Kroft racked up 4.3 AGL with a broken foot and ankle injuries. The Chargers' Sean Culkin went on IR with a torn Achilles in Week 5, while Hunter Henry missed several games early in the year with a knee injury.
Offensive Linemen
The Arizona Cardinals suffered the most AGL to offensive linemen in 2019. Newly signed right tackle Marcus Gilbert missed the entire season with a torn ACL; his backup Jordan Mills went on IR with his own knee injury in Week 5. Max Garcia, another free-agent addition, missed the first half of the year recovering from his own torn ACL, suffered in November of 2018 while with Denver. When he finally returned to the active roster, it was as a backup to J.R. Sweezy and Justin Pugh.
The Jets had seven offensive linemen rack up at least 1.0 AGL -- a significant number, considering only five linemen start in any game. Those seven, in order of AGL: Ryan Kalil (9.3, knee), Brian Winters (8.2, shoulder), Chuma Edoga (5.3, hip/knee/ankle), Kelvin Beachum (5.0, ankle), Kelechi Osemele (3.3, pec/shoulder/knee), Alex Lewis (2.2, shoulder/neck/elbow/ankle), Tom Compton (1.0, back/calf/ankle).
In Cincinnati, Jonah Williams, the 11th overall pick in the draft, missed the entire season with a torn labrum. Veteran Cordy Glenn racked up 7.0 AGL with a concussion, while guard John Miller had 2.3 AGL with groin and concussion issues.
In Oakland, journeyman tackle Jordan Devey started only four games before tearing his pec. Other Raiders linemen with significant injuries include Gabe Jackson (5.4, knee), Trent Brown (5.9, pec/knee/calf/ankle), and Richie Incognito (2.3, back/ankle/knee).
Defense
Defensive Line
Three teams led the way in defensive line AGL in 2019. The first two battled each other down to the wire for the NFC East championship. The third won the Super Bowl.
In the table at the end of this column, you'll see Dallas and Philadelphia tied with 31.3 defensive line AGL. Take it to one more decimal point and the Eagles win this sad race, 31.32 to 31.26. Malik Jackson was carted off the field in the season opener against Washington and missed the rest of the season. Hassan Ridgeway missed the second half of the year, going on IR with an ankle injury. Timmy Jernigan missed most of the first half of the season with a broken foot.
For the Cowboys, Daniel Ross missed the entire season with a shoulder injury, while Tyrone Crawford only played four games due to a hip injury. Antwaun Woods also missed six games due to knee, shoulder, and hip injuries.
Finally, in Kansas City, Chiefs defensive tackle Xavier Williams went on IR with a high ankle sprain in Week 6. Emmanuel Ogbah (6.0 AGL, torn pec), Alex Okafor (4.4, ankle/torn pec), Chris Jones (2.3, groin), and Frank Clark (1.9, neck) also had more than their share of aches and pains.
Linebackers
The most injured team of 2019 suffered the most injuries to linebackers of any team we have ever measured. The New York Jets were struck by 48.5 AGL at linebacker, surpassing the 42.2 mark of the 2017 Chicago Bears.
The struggles for New York began in their second preseason game, when Avery Williamson collided with teammate Tevaughn Campbell in Atlanta, tearing his ACL and going down for the year. C.J. Mosley didn't last much longer -- he hurt his groin in Week 1 against Buffalo, missed the next four games, returned for the Monday nighter against New England, then missed the rest of the season. Blake Cashman went on IR in Week 9 after tearing his labrum. Neville Hewitt missed four games in October and November with a stinger. Jordan Jenkins missed two games with calf and ankle issues earlier in the year. With all those injuries, Albert McClellan might have started down the stretch, but he missed the second half of the year with a concussion.
Finishing in a distant second place in this category were the Jacksonville Jaguars. They signed Jake Ryan in free agency after he missed all of 2018 with a torn ACL, but he struggled to return and didn't see the field until Week 13. He played two games, then went on IR with a hamstring injury. Najee Goode (6.0 AGL, toe/knee), Myles Jack (4.3, knee), and Quincy Williams (4.1, meniscus/hamstring/illness) also struggled to stay healthy. Note this doesn't count Telvin Smith, who took 2019 off to "get his world in order."
In third place is Washington, a team we have already discussed quite a bit today. They lost Reuben Foster for all 16 games due to an ACL tear. Ryan Kerrigan also missed four games in December with a calf injury.
Our final team to discuss in this space is the New Orleans Saints. Alex Anzalone only played the first two games of the year; in Week 3, he went on IR with an "undisclosed" injury. Anzalone has a history of shoulder injuries, and indeed, it was shoulder surgery that ended his 2019 season. Kiko Alonso also missed three games with a quadriceps injury.
Defensive Backs
Somehow, the Baltimore Ravens had the most injuries in the secondary of any team in the NFL last year, and still finished with the best record in the league. Slot corner Tavon Young missed the entire year with a neck injury. Tony Jefferson joined him on the sideline after tearing his ACL in Week 5 against Pittsburgh. Deshon Elliott probably would have started in Jefferson's place, but he missed the last 10 games of the year with his own knee injury. Finally, cornerback Jimmy Smith missed seven games early in the year with a sprained MCL.
The Miami Dolphins didn't have as many big injuries as the Ravens, but they had seven different defensive backs with at least 1.0 AGL. Xavien Howard first suffered a knee injury after Week 4. He didn't return to the starting lineup until Week 8 against Pittsburgh on Monday night, then went on injured reserve the next day. Aqib Talib went on IR in Week 10 with broken ribs. Ken Webster suffered an ankle injury halfway through the year, tried to play through it in December, and finally went on IR himself. Reshad Jones only played in four games scattered throughout the year, missing the others with ankle and chest injuries. Ryan Lewis (3.3 AGL, hamstring), Bobby McCain (1.5, shoulder/hamstring), and Jomal Wiltz (1.3, groin/shoulder) round out Miami's walking wounded.
The last team with a plethora of injuries in the secondary is the Atlanta Falcons. The list of those injuries, at least, is short: Keanu Neal missed 13 games with a torn Achilles, John Cyprien missed 11 with a knee injury, and Desmond Trufant missed seven with a toe injury and a broken arm.
AGL by Position for all Teams
The following table shows AGL at all positions for each team in the NFL in 2019. For those wondering, specialist (K/P/LS) injuries are not included in AGL.
| 2019 AGL by Position | |||||||||
| Team | Total | QB | RB | WR | TE | OL | DL | LB | DB |
| MIN | 25.6 | 0.0 | 2.1 | 5.6 | 0.1 | 4.0 | 3.1 | 4.0 | 6.8 |
| TB | 39.5 | 0.9 | 0.0 | 6.6 | 2.3 | 7.1 | 7.2 | 13.8 | 1.6 |
| TEN | 43.6 | 0.0 | 0.4 | 5.4 | 9.3 | 4.5 | 9.0 | 2.5 | 12.4 |
| DAL | 51.8 | 0.1 | 0.0 | 2.3 | 0.0 | 8.9 | 31.3 | 6.0 | 3.2 |
| ATL | 53.2 | 0.4 | 2.1 | 3.5 | 3.3 | 6.3 | 3.5 | 0.1 | 34.1 |
| SEA | 56.5 | 0.1 | 4.0 | 2.3 | 12.6 | 12.5 | 10.2 | 2.7 | 12.1 |
| CLE | 56.8 | 0.0 | 0.0 | 0.8 | 10.6 | 3.9 | 5.8 | 15.6 | 20.0 |
| NO | 58.3 | 5.3 | 4.9 | 3.1 | 2.1 | 7.8 | 10.3 | 18.1 | 6.6 |
| PIT | 60.5 | 15.0 | 6.4 | 5.4 | 2.4 | 1.4 | 10.0 | 3.6 | 16.3 |
| LAR | 60.6 | 0.0 | 3.0 | 2.1 | 3.3 | 21.4 | 0.3 | 13.1 | 17.3 |
| BUF | 64.6 | 0.3 | 1.9 | 1.6 | 20.3 | 22.4 | 14.4 | 0.5 | 3.2 |
| IND | 65.3 | 1.0 | 2.6 | 28.4 | 6.2 | 0.4 | 6.5 | 3.3 | 16.9 |
| CHI | 65.4 | 1.3 | 0.3 | 7.3 | 10.1 | 17.9 | 16.6 | 10.0 | 1.9 |
| GB | 65.6 | 0.1 | 2.2 | 21.1 | 0.7 | 15.3 | 2.8 | 5.3 | 18.1 |
| CAR | 66.2 | 14.0 | 0.1 | 1.6 | 2.9 | 17.9 | 18.7 | 6.4 | 4.7 |
| BAL | 68.7 | 0.5 | 1.8 | 3.0 | 1.9 | 2.4 | 3.5 | 10.6 | 44.9 |
| NYG | 72.6 | 2.0 | 4.1 | 24.6 | 14.0 | 4.3 | 0.0 | 17.0 | 6.6 |
| KC | 75.8 | 1.5 | 4.9 | 6.7 | 0.0 | 19.0 | 24.8 | 16.0 | 2.8 |
| HOU | 78.2 | 0.3 | 16.1 | 5.4 | 10.0 | 18.6 | 8.3 | 2.2 | 17.2 |
| CIN | 83.2 | 0.0 | 0.3 | 27.3 | 0.0 | 28.3 | 6.4 | 0.3 | 20.5 |
| PHI | 84.4 | 0.0 | 2.4 | 20.1 | 1.3 | 6.5 | 31.3 | 10.1 | 12.7 |
| ARI | 85.0 | 0.3 | 2.1 | 3.7 | 0.3 | 39.2 | 18.6 | 2.6 | 18.2 |
| NE | 85.0 | 0.4 | 27.4 | 21.8 | 4.1 | 20.0 | 1.4 | 1.9 | 8.1 |
| DET | 87.8 | 13.6 | 8.8 | 19.6 | 4.3 | 10.3 | 14.3 | 8.4 | 8.7 |
| DEN | 89.2 | 0.0 | 25.4 | 0.8 | 0.6 | 14.2 | 6.8 | 15.4 | 26.1 |
| OAK | 93.4 | 0.0 | 19.0 | 6.2 | 0.0 | 27.1 | 15.5 | 0.7 | 24.9 |
| SF | 95.8 | 0.0 | 25.1 | 4.8 | 2.9 | 16.6 | 12.6 | 8.0 | 25.7 |
| MIA | 100.0 | 0.2 | 6.1 | 12.6 | 0.0 | 24.1 | 18.6 | 0.8 | 37.7 |
| JAX | 100.2 | 8.4 | 0.3 | 11.8 | 31.3 | 3.6 | 10.7 | 29.5 | 4.6 |
| LAC | 103.0 | 0.0 | 0.0 | 22.2 | 17.5 | 22.9 | 9.1 | 1.6 | 29.6 |
| WAS | 131.0 | 18.4 | 17.3 | 9.6 | 23.0 | 13.6 | 2.9 | 20.3 | 25.8 |
| NYJ | 160.1 | 5.2 | 1.1 | 23.0 | 12.9 | 34.3 | 11.7 | 48.5 | 23.4 |
Comments
22 comments, Last at 20 Apr 2020, 10:06am
#2 by Aaron Schatz // Apr 17, 2020 - 2:06pm
Of course, there isn't DYAR for any players besides "skill players." Quantifying the effect of specific players (non-QB, particularly) on DVOA is very difficult. But we've played around with different ways to try to score AGL by the quality of players being lost to get something that's more of a predictor of how much a team will decline due to injury. We've never quite found anything we like. But we're still playing with it every so often.
#21 by ssereb // Apr 19, 2020 - 4:19pm
Well, it certainly helped that Marlon Humphrey and Earl Thomas each played more than 90% of defensive snaps, as did Marcus Peters once he came over from the Rams. Brandon Carr and Chuck Clark also had 16-game seasons. Just a ridiculous amount of depth in the secondary.
#4 by reddwarf // Apr 17, 2020 - 2:28pm
How can Denver have 0 AGL at the QB position? Brandon Allen started 2 or 3 games because Flacco was injured. I get not counting Lock coming in (though arguably when Flacco went down Allen only started because Lock was on IR and hadn't come back yet) as you could say that was (lack of Flacco) performance related, but you might want to check your injury records for Flacco.
#6 by Fishbol // Apr 17, 2020 - 3:29pm
I was shocked to see that Corey Coleman earned the Giants a whopping 16 AGL, as he hasn't done much of anything, production-wise, in well, ever. Just because he was technically starting doesn't mean he had more impact on a season, than say, the Eagles losing Desean Jackson
#7 by Aaron Schatz // Apr 17, 2020 - 3:41pm
That's one of the issues with AGL, that we do treat players of different quality the same if they count as starters, but Coleman was pencilled in as a starting outside wideout for the Giants last year (one of three WR with Shepard and Tate) so he counts.
#9 by MarkV // Apr 17, 2020 - 4:38pm
I like the idea behind AGL, but the more I look at the data, the less sense it makes.
I am a Broncos fan, so I know them best, and looking at their splits:
OL - JaWuan James played 6% of snaps, functionally missed the entire first week, then played half of two different games. A well calibrated system should probably assign a team missing a starting OL for 15/16 of the year at 15. Ronald Leary also missed 4 games entirely, presumably a straight forward 4 games lost (technically he lost another 9 snaps preceeding that,. Elijah Wilkerson lost a full game and half another (he was replacing James after he went down, but it was week 17 - I get that calculating the value of starters who became starters because of injuries is hard, but this has to count for something). Dalton Risner lost half a game as well.
OL should be the easiest to calculate (along with QB), since starters don't sub. So the Broncos "pure" games lost for OL should probably be able to arrive at a number between 20 and 22. 14.2 Doesn't seem to adequately describe this.
Similar concerns with Denvers RBs. Its hard to project what they would have done with Theo Riddick, since he was new and never played a down. But the presumption seems to be as a 3rd down back/recieving back. If he had ended up with 30% snap count player, and Janovich was close to that when healthy, its really weird that their injuries read as costing the Broncos almost twice as much as the OL injuries. I would expect them to total somewhere around 8-12 games lost because of the smaller roles they were presumed to play. Even if you presumed that Riddick would have been a starting RB major player, I still don't see that total rising above the 18-20 games lost threshold.
Looking at LB, Denver lost 15 games from clear starters (Chubb, Davis, Miller). Jewell, Johnson, Reed, Hollins, and Nelson all had injury appearances and lost time from injury (its not obvious that all of those calls are, or should be wrong - Nelson and Jewell were both early season starters (ish) that got benched before getting hurt, while Reed and Hollins were replacing Chubb snaps when they got hurt. I don't know what the right number for the time lost is from those gentlemen, but .4 seems wrong.
I get that 2 down, 1 down, flex, key sub players etc are going to be hard to calculate, but looking at the only team I follow closely, the numbers just look weird. They accurately describe that Denver had injury luck at WR and TE, and I happen to basically agree that the DL number is reasonable, but (excluding the QB error), the rest of the numbers just seem silly.
#18 by Vincent Verhei // Apr 17, 2020 - 10:08pm
Most of those questions are gray-area determinations of which players should and should not be counted as starters. Yes, we count third-down receiving specialists like Theo Riddick as starters, but not second- and third-string linebackers. And "new starters" who are filling in for other injured players count less when they get injured as well.
The big discrepancy is Ja'Wuan James. Yes, he only played three games, but he was listed as questionable eight times (0.27 AGL per game), doubtful once (0.99 AGL), and out six times (1.00 AGL each). The assumption is that questionable players could play if absolutely necessary, but the team makes the decision to go with the healthy backup instead. So James counts for 8.9 AGL, even though he only played three games.
#14 by mrh // Apr 17, 2020 - 6:36pm
At p-f-r, Wilson's player pages show him at 98.6% of the offensive snaps in the 1st Rams game and 98.5% in the 2nd Rams game, with 73 and 66 snaps respectively. But going to those game box scores, Wilson has the same numerical counts but credit for 100% of the offensive snaps in both games. So there is some discrepancy in the data somewhere (not sure of FO, p-f-r, and CBS' sources).
#16 by Vincent Verhei // Apr 17, 2020 - 9:26pm
Wilson was listed with a knee injury in Weeks 8 and 9, and a hamstring injury in Week 12. He was full participant in practice all those weeks, and was listed with a "blank" designation, which means players almost always play (see Part I). But the assumption is that he played at less than 100% those weeks, so he racked up a tiny amount of AGL.
#17 by Vincent Verhei // Apr 17, 2020 - 9:38pm
Gordon wasn't injured, he was holding out. AGL only measures injuries, not holdouts or suspensions.
Justin Jackson's injuries are not counted because he was strictly a reserve player. Even in Gordon's absence, he never played more than 28 offensive snaps in a game. And most of the games he missed were after Gordon's return, so he would have been the third back behind Gordon and Austin Ekeler.
#20 by Granite // Apr 18, 2020 - 4:28pm
Out of curiosity, why use adjusted _games_ lost instead of adjusted _plays_ lost? I know it can be hard to estimate the quantity of something that never existed (plays missed), but using games seems to effectively make other assumptions that seem less accurate than estimated plays.
Using plays would eliminate the need to draw an artificial line between starters and backups, and it would include valuable backups - for example, a team that loses all its backup receivers or DBs is in trouble, and right now (if I understand correctly) the stats don't account for that: The team with all backups injured looks the same as the team with none.