By Kevin Meers
(Huge thanks to TeamRankings for providing the data used in this post)
Last November, I wrote a post on how the lockout did not significantly affect injuries in the NFL this season. Now that the regular season is over, I wanted to take another look. The much discussed lock out could have exacerbated or limited the mounting “wear and tear” and “buildup of cuts and bruises” that develop throughout the season. In reality it did neither. There was almost no difference in injuries this season when compared to the past two years. For various reasons, there has been a lot more attention paid to the injuries that did happen. That increased attention has made it appear that there were a lot more injuries this season, despite efforts by the NFL to limit concussions and hits on defenseless players. If the league is serious about reducing injuries, that attention needs to focus on prevention going into next season.
As the graph below clearly shows, injuries this season are almost exactly the same as the past two years. This constant level of injuries suggests that neither the lockout nor the recent rule changes have had any significant effect on the number of injuries in the NFL.
While it appears that the past two seasons have had a lot fewer injuries than the 2009-2010 season, that difference stems from an extremely high injury rate in the first two weeks of the preseason that year. Removing the preseason, it is clear how little the number of injuries has changed over the past three years.
To accompany this consistent injury rate, injury severity has not changed either – despite the NFL’s recent efforts to penalize and fine players for dangerous hits. Those policies are good for the health of the players, but more should be done if the NFL wants to protect the men on the playing field. To calculate severity, I created a system where a player listed as “probable” had a severity of 0, “questionable” equaled 1, “doubtful” was 2, “out” equaled 3, and “injured reserve” was 4. A metric like “weeks missed” would be better than this system, but those data, unfortunately, are hard to come by. Below is a graph of average severity in each week of the season. As you can see, severity is well within the normal range of the past couple of seasons.
Furthermore, through the end of the regular season, 308 players have been placed on I-R, compared to 259 in the 2009-2010 season and 338 in 2010-2011. Even the most serious injuries have not fallen significantly.
With so many big name players seriously injured this season and recent studies on the effects of concussions on players’ long term health, talk about injuries has become commonplace this season. It may simply be that football as it is played requires this level of injuries. It is up to people who know much more about football than I do how much, if at all,  the NFL should change the game to ensure the safety of its players. What I am putting forward here is simple: neither the lockout nor recent rule changes has significantly changed the injury rate or the severity of those injuries. For that to change, the NFL must continue to change its rules.
Seems to me that one obvious possibility has been missed. What if the two variable are offsetting. The rule changes reduced injuries, however the effect of the lockout increased them by a similar amount. You would achieve the same results.
Chris – the variables could be offsetting, but it would be quite a coincidence for the lockout and rule changes to have exactly equal and opposite effects. It is more likely that neither have had a significant effect, especially given how identical this season was to the past two seasons.
Kevin,
I’m not sure if I agree with the usage of the statement “it would be quite a coincidence for the lockout and rule changes to have exactly equal and opposite effects.”
Statistically, of course that’s true. The chance that any two continuous variables are exactly equal is always 0%. However, the two influences, lockout and rule changes, could have similar (in absolute value) affects on injuries, and be deemed “statistically equal.” In other words, close enough so that the injury index fits nicely in with 2009 and 2010. Since previous years are your point of reference, there is a margin of error there that allows for the influences of the lockout and rule changes to be “close enough” in my opinion.
Thanks uoduckfan, that’s exactly what I’m getting at. Also, I’m not saying that I think this is true. I’m just saying to not even mention the possibility seems a bit sloppy, and like a very easy part of the study to attack for anyone adamant that either there were less injuries b/c of rule changes or there were more because of the lockout. The more conclusive update will be after next season with no lockout variable in play.
If we assume that the rule about keeping concussed players out of games has really taken hold this year, and not previous years, then injuries in previous years are actually under-reported while the numbers are more accurate this year. I don’t know how many players that would represent, but it would mean that the number of injuries actually has dropped some amount. On the other hand, risk compensation says that we should see exactly what you’ve found.
Suggestion for future article – Is it possible to perform any quatitative analysis on a team by team basis compared to the rest of the NFL that shows if the amount and severity of injuries that any particular team incurs during a season actually has an impact on wins and losses. I am tired of hearing each season that team XYZ suffered all sorts of injuries and that is the reason for their poor record as opposed to mediocre players and/or coaching.
It’s hard to come by experienced people for this subject, but you sound like you know what you’re talking about! Thanks