What is an injury-prone player but a story we believe?
NFL evaluation departments are taught not only to evaluate players, but also to assess risk of future injury. That risk is one of the biggest driving forces in the difference between player value on tape and player value come draft day.
Jay Ajayi's knee issues drove him down into the fifth round of this year's draft, and as a skill position player with considerable upside, he's the most memorable present example of this. He's hardly the only one.
Offensive lineman Brandon Thomas hurt himself at a predraft workout and fell all the way to the 49ers in the third round last year. Wideout Ryan Swope went in the sixth round in 2013, despite several Wes Welker-esque qualities, because of concussion concerns. Swope retired from the Arizona Cardinals before he even played a preseason down. And these are just a few recent examples.
This is, in many ways, the biggest black box of the NFL right now. It's compounded by the fact that most teams don't have the same evaluation practices for injuries. For instance, a torn ACL didn't keep Georgia's Todd Gurley from going 10th overall in the draft. That also means it's a big opportunity for general managers to leverage some additional skill to their roster for a price that is lower than it should be—if they believe in the player's talent and fit.
What does it actually mean to be "injury-prone"?
Prior to last season, many hung the injury-prone label on running back DeMarco Murray. Murray, of course, then slogged out 392 carries for the Dallas Cowboys, the most since Larry Johnson set the NFL record in 2006 and the seventh-highest single-season total ever. Then, teams were so turned off by his injury-prone nature that (a) Murray received a five-year, $42 million contract with $21 million in guarantees from the Philadelphia Eagles, and (b) at least two other teams leaked to the media that they'd offered comparable money.
What does that mean for Murray's status going forward? I interviewed Jene Bramel of Football Guys, one of the doctors I respect most in the football prognostication business, for this piece.
Bramel said, on Murray: "He had been prone to injury. But deciding how much weight to give a recent injury pattern is a more complicated question. Some players are absolutely prone to future injury. Scar tissue, cartilage loss, poor conditioning, normal aging, a genetic predisposition to certain types of soft tissue failure and lots of other factors apply. Maybe Murray was unlucky in his first three years. Maybe he was lucky to avoid injury last year. As observers, it's hard to know which is closer to the truth."
Bramel has been writing about the NFL for a long time—here is a passage from Bramel's Second Opinion about Murray's injury-prone nature back in 2012:
Demarco Murray suffered a dislocated kneecap as a freshman at Oklahoma, then had hamstring injuries and a knee problem in future seasons at OU, before missing time in camp as a rookie with a hamstring strain. Is that list of injuries sufficient to call him injury prone? Is it reasonable to think his conditioning and strength have improved from what it was between 2007 (age 19 season) and 2010 (age 22 season)? How many other running backs would’ve broken their leg if tackled in the same way Murray was last season? ...
... Are certain players injury-prone? Almost certainly. Can you predict which players may be more injury-prone than others? Probably. Can you reliably predict whether a player will suffer an injury in any given week, month or season? Good luck – especially if you’re unable to make a precise argument based on one of the above factors. [emphasis added]
So, even knowing what we know about the likelihood of players getting injured, one of the doctors who studies this most has basically concluded that predicting future injury is like throwing darts at a board.
What we can do, according to Bramel, is look at factors that make a player an injury risk. To him, those factors are:
- A player with a smaller frame, greater than average laxity in his joints, or less than average tendon flexibility
- A player with poorer strength, poorer conditioning, or less endurance than his competition
- A player with below-average bone density or differences in the microscopic makeup of his connective tissue
- A player with poor biomechanics or technique, especially when performing repetitive motions
- A player with poorer reaction time or slower neuromuscular processing speed
- A player more willing to take chances or put himself in a position to be injured
- A player who has had repeated injuries to the same area (e.g. scar tissue, cartilage loss, overuse syndromes)
- A player more willing to play through pain that limits his conditioning, flexibility, reaction time, etc.
This is all useful information when it comes to identifying injury risks. However, as we've already noted, none of these things inherently means a player is predictably injured. They simply give him a higher chance to be injured. Playing the game of probabilities means also acknowledging a player has a chance to stay healthy in spite of any of the above habits.
What does that mean for player acquisition? It means that even though players a front office evaluates as injury-prone effectively are a zero-sum game ("Can this guy stay on the field?" is a yes/no question), the injury-prone question should actually hit multiple levels and go beyond that.
The case of Marcus Lattimore
Let us go to the most extreme example in recent history.
The 49ers drafted South Carolina running back Marcus Lattimore in the fourth round in 2013. He'd had two horrific knee injuries in college, including the most recent one below, which still makes for cringeworthy viewing and an opportunity to criticize the fact that college football players are not compensated for their work:
However, while drafting Lattimore at 131st overall was a failed gamble, the 49ers actually wouldn't have needed much from Lattimore to make the pick a worthwhile endeavor—even in selecting someone they knew was going to miss a quarter of his rookie contract. I'll use the two most well-known publicly available multiposition comparative systems to demonstrate: Pro-Football-Reference's Approximate Value (AV) and Pro Football Focus' grades.
Per Chase Stuart's work at Football Perspective, we know the average five-year return of AV on the 131st overall pick is 3.6 total AV.
I actually believe Football Outsiders' DVOA statistics are the best true gauge of how a player performed empirically, so let me bring up the worst possible qualifying running back from last season: Houston's Alfred Blue.
No matter what empirical evidence you believe in, Blue didn't have a very good season, whether judging by raw rushing stats (3.1 yards per carry), DVOA (dead last) or Pro Football Focus grades (-6.1 in just 341 snaps).
|Alfred Blue, Rookie Season|
|Basic Stats||Advanced Stats|
|Yards Per Carry||3.1||DVOA||-21.3%|
|Rushing Yards||528||PFF Grade||-6.1|
|Sources: Pro Football Focus, Football Outsiders, Pro Football Reference|
And yet...Blue created 4 AV. Just under 350 snaps of bad running back play is worth that many AV. If Blue were a fourth-round pick, we'd deem him a successful one in the guise of AV history. As a sixth-rounder, Blue is already a wildly successful pick by that measure.
So when you really approach historical standards of what to expect of the 131st overall pick, there's not a whole lot of risk. People remember Lattimore's retirement because he tried to overcome so much to make it in the league. People will remember the players the 49ers could have selected over him who succeeded. But 10 years from now, will you also remember Devin Taylor or Sanders Commings? I barely remember them now.
This is why talent has to come first. A badly broken, 350-snap season of Lattimore was one of the worst-case scenarios. He didn't fulfill that, but if he had, he'd have been an average pick for his draft slot by the (admittedly) somewhat flawed total value measurements we have to work with.
I asked Bramel to speak to the reasoning that perhaps teams fall into the cognitive trap of finding a long-term fixture in a league that seems to turn over rosters faster than ever. Bramel said: "Coaches and coordinators and general managers and players seem to have a shorter 'long-term window' than ever. Some teams and coaches may see their effective window as 1-2 seasons. Those teams may be perfectly comfortable with a player with an injury file suggesting a shorter career.
"Teams with a longer effective window would probably take that player off their board altogether. Or those teams with the shorter window may not want to take a risk on any injury risk derailing their depth chart."
It's also worth noting these are historical averages, and if you believe the totality of scouting is becoming easier due to increased video availability and better practices in a more information-driven world, you should bump up these numbers. (I believe in these things.) Still, even if you double the expected AV of Lattimore's pick, the 49ers could recoup their investment rather easily, even if they only have three years to do it.
Fungible positions, meet fungible players
Nobody would ever argue you should draft an injury-prone quarterback just because he was the best available talent. While a quarterback-desperate team might do it anyway, there's a certain stability a quarterback needs to have to be a true franchise player. Chad Pennington and Matt Schaub are good examples of modern quarterbacks who frequently missed time, and while each had stellar statistical seasons, you'll never hear them mentioned in the same breath as the star quarterbacks of their day.
But there are also examples in today's NFL of positions that are essentially plug-and-play. The obvious example is running back, a position that has become so devalued that it was a big deal when two backs went in the top 15 of this year's draft.
And if you believe what Bramel says about a team not wanting to derail its depth chart, what better spot than a place where it's already unsteady? Even beyond running back, certain teams have begun "punting" different positions, a conciliation to the fact that in the NFL's salary-cap era, you can't fill every hole. So if you can't fill every hole, where are the places you can go cheap?
|Teams Who Have "Punted" a Non-RB Starting Spot|
|Atlanta||LG||Harland Gunn, Mike Person, Adam Replogle|
|Carolina||T||Michael Oher, Jonathan Martin, Mike Remmers, Daryl Williams (?)|
|Carolina||SS||Roman Harper, Robert Lester, Kurt Coleman, Brian Blechen|
|Detroit||RT||La'Adrian Waddle, Michael Williams|
|Indianapolis||FS||Dwight Lowery, Dewey McDonald, Robert Smith|
|Jacksonville||FS||Sergio Brown, Josh Evans, Desmond Cooper|
|Kansas City||ILB||James-Michael Johnson, Josh Mauga, DJ Alexander|
|New Orleans||TE||Ben Watson, Josh Hill, Orson Charles, Harold Spears, Jack Tabb|
|Oakland||SS||Charles Woodson, Tevin McDonald, Jimmy Hall|
This above table lists depth-chart battles for teams where there is no free agent or highly drafted player in place. Note that right tackle is a place that a few different teams have conceded in recent years, especially the Lions, who have been doing this for years now.
Bramel believes this is "cyclical, in the same way some draft classes are stronger than others and some schemes become more popular than others." Perhaps those teams have just not found the circumstances to help them fix the hole more than they lack an actual desire to fix it.
Different coaches bring different ideas of fungibility to the table, as do different circumstances. You'll never see a Jeff Fisher offense without a lot of focus on the ground game. At the same time, it's hard to ignore the impact of a mauling right tackle when your passing game is led by an untested quarterback. Much easier if you have Matthew Stafford throwing to Calvin Johnson.
I'd just argue that, if a team has a position it treats as a concession, it should be taking risks on players with injuries that lose stock on the market because of them. If the Lions are happy to play right tackle patty-cake, they should be interested when T.J. Clemmings falls to the fourth round over a stress fracture in his foot. When a team doesn't invest much in tight ends and finds a talent like Travis Kelce falling into its lap because of sports hernia concerns, it should consider taking him.
After the first round of the draft, floor and ceiling matter much more than current status. If a highly graded player keeps falling, NFL teams shouldn't let pack mentality continue pushing him down—they should be the aggressors, especially if it's a late pick at a position where they're already not going to invest much anyway.
What I'm arguing here isn't that NFL teams are necessarily wrong. NFL teams have a lot more riding on each pick than the statistical measurables we have. They also embrace plenty of risk when, say, selecting a player from a lower division of football. An NFL coach can't go to an owner and argue that picking Lattimore was smart simply because of the expected AV of the pick. There's a lot more to it than that.
As Bramel says, "It's probably as simple as recognizing that every team will have its own philosophy of risk tolerance. In the same way some teams may be more aggressive on fourth down, some teams may be more willing to accept certain types of injury risk."
But given the relative drop in the bucket that perceptually injury-prone players are, it seems that at relatively fungible positions, stockpiling injury-prone players with talent and rolling the dice isn't a bad strategy.
Jay Ajayi was a terrific pick for the Dolphins. I don't know that his knee will hold up in the NFL—the reports we heard very well may be true. But odds are the 149th player in the draft will live up to the projected 2.7 AV he's signed up for within 250 snaps of play. And if he's as good as his original talent indicated, team executives may look back and shake their head at the idea he was ever available.
Injury risks matter. But in most cases, it's smart to not overthink them.
Rivers McCown is an NFL Analyst for Bleacher Report and the co-host of the Three-Cone Drill podcast. His work has also appeared on Football Outsiders and ESPN.com. Follow him on Twitter at @riversmccown.