Can We Predict NFL Injuries?

By Greg Alan (4for4 Founder) on Jul 6, 2010

Greg Alan's picture

Greg Alan is the original founder of 4for4.com Fantasy Football and comes from a deeply statistical and analytical business background.

First, let's get this out of the way... NFL injuries can happen to almost any player at almost any time. That said, there are patterns to be observed and by recognizing them, you can gain a competitive advantage, but not all the time, as few things are 100% absolute.

Next, a time saver test for you...
 

SAVE TIME - READ THIS FIRST

 
If you know everything you need to know about NFL injuries and fantasy football, stop reading.
 
If you're not open-minded to some hardcore, fact-based analysis, save some time and skip this article. This one is not for you.
 
If you hate numerical analysis and/or think math and statistics play no role in managing a fantasy team, stop reading.
 
If you think you must be able to predict injuries with spot on accuracy and you must be able to predict what week a player is likely to get hurt to put injury analysis to good use, stop reading. You won't get this material.
 
Still reading? 
 
Great. I'd really like to share some historically-proven principles I've obtained by leveraging fact-based research via some heavy lifting CHAID statistical analysis.
 

PUT THE ODDS IN YOUR FAVOR

 
Learning and applying the strategies outlined herein will enable you to avoid unnecessary injury risks and in the long run, help you win more and improve your fantasy football record. 
 
Most people play fantasy football to win. But, regardless of how serious you are about winning a title, and despite how successful you’ve been in the past, on any given Sunday, showing no respect for your knowledge, injuries can send you for a loss! Or worse, if not properly prepared, injuries can cost you an entire season. 
 
So what can be done? Aren’t injuries just part of the game? Mostly all luck? The dreaded random factor? 
 
Certainly no one is safe and in the short-run, luck plays a role. However, by applying historically-proven principles and leveraging fact-based research, you can minimize the impact. 
 
You’ll hear other less informed owners whining about how their 1st round stud RB went down to injury. Or, perhaps they’ll gripe because they can’t find a serviceable quarterback now that their QB is seeing stars. Leave the moaning and complaining to less informed fantasy owners. To win, first learn the odds and then put them in your favor as often as possible.
 
Goal: Minimize the luck factor, apply information, and go with the historical odds, not against them. 
 
I’ll provide you fact-based research from 1,400 NFL games. In addition, I’ll show you a few (of many) ways to apply the information and provide some specifics for the upcoming season. 
 
Remember, in the long run, true odds do win out. Through complex computer and statistical analysis of the game of football, probabilities have been derived.
 
Most of the analysis here was obtained via CHAID modeling and validated on large holdout data sets. If you don’t believe an optimal way of playing exists and long-run odds provide insights, then no matter how much I go on about computers, statistics and probability, this information will be of lesser value to you. If you think hunches and random streaks are what winning is all about, all I can really say is “Good luck!” 
 
Just one more thing before we dig in. I realize not every owner wants to spend time studying the intense portions of this research. You’re not interested in all the mathematical research and strategy, but you like the idea of using a few simple rules of thumb to improve your fantasy football record. For that owner, I’d suggest you focus on specific player names bolded herein. 
 
For the rest of you, and you know who you are, since you really want to win consistently at fantasy football, I’d suggest you study everything presented. Periodically reference the concepts and look for new ways to apply the research to increase your chance of a winning season. Think about how you can further exploit the research when working out trades, making start/bench decisions and picking up free agents. 
 
If you don’t want to agonize over player health from week to week, everything else being equal, avoid the “Bumps and Bruises Club.” 
 
While some of these players are studs, they also lead the NFL in being listed as ‘Questionable’ or ‘Doubtful’ due to injury status. Remember, having a player being ruled ‘Out’ is often better for some fantasy owners, especially if you have the injured player’s competent backup at your disposal. 
 
Can you really use injury risk information to improve your team? The answer is yes, but not with 100% accuracy and nothing is absolute. But, you don't need to be 100% accurate here. In fact, if you can shift the odds from 50-50 to 51-49 in your favor, you have an edge. By employing several techniques, I’ll show you how to increase your odds of avoiding injury problems. 
 
Short term, a rather inept fantasy football owner could easily prevail over the best owners in the world. Why? Due to the fact that football results can vary greatly from expectation. And in the short run, they often do. The following illustration brings this all together. 
 
Visualize two jars. Each jar holds 100 marbles, some red and some white. If you select a red marble your fantasy player will score big that week; pick a white marble and your player stiffs. Many fantasy football owners, when faced with very hard decisions, use “gut feel.” In essence, the uninformed owner is randomly picking from a jar holding 50 red (winning) and 50 white (losing) marbles. Without appropriate historical facts, they simply “win some and lose some.” In the short run, this type owner could certainly pick 3 or 4 red (winning) marbles. When that happens, they walk away happy. Unfortunately for them, it just won’t last. 
 
However, by applying the strategies outlined herein, you can avoid the 50-50 situation. In fact, armed with appropriate and historically proven information, the wise fantasy owner can gain an advantage. In this example, faced with that same hard “gut feel” decision, the wise owner will not be forced into selecting from the 50-50 jar. Rather, the wise owner selects from the container holding 51 or more red (winning) marbles. The actual number of red marbles depends on the expected value derived from historical computer analysis. But, as long as it’s greater than 50, the wise owner has an advantage over his uninformed competitor. 
 
The key takes: First, the uniformed “gut feel” player will occasionally experience good fortune and select an excessively large number of red (winning) marbles. Second, the wise player, despite using historically-proven computer analysis to make all the right moves will at times experience down turns. However, if the wise player sticks with the historically-proven execution and never deviates, in the long run, they will prevail. Finally, the wise player’s advantage need not be based on 100% accuracy. In fact, when faced with what looks to be a true flip of the coin decision, the accuracy need only be 51% to gain an advantage. 
 
All that being said, it’s important to mention, what follows is not only my advice on how to manage injuries. Rather, it’s the way to manage based on looking at thousands of NFL players - exhaustive review of years of NFL action. 
 

DRAFT - Injuries and Preparing for the Draft

 
In the context of preparing for the draft, I’ll use a ‘Questionable’ or ‘Doubtful’ listing on the official NFL weekly injury report as a way to define “Bumps and Bruises.” For the prepared fantasy owner, having a player ruled ‘Out’ in advance, can often be more desirable than dealing with the uncertainty of a ‘Questionable’ or ‘Doubtful’ situation. 
 
With this in mind, computer analysis was required to develop a historically valid predictor of the number of games an NFL player will appear as ‘Questionable’ or ‘Doubtful’ in the upcoming season. In addition, I developed the analysis so it would be concise, exact and yet offer a very powerful strategy. 
 
After structuring the challenge, it became clear. Predicting the number of times an NFL player will appear ‘Questionable’ or ‘Doubtful’ in the upcoming season involves many factors. In some cases, factors become circular. For example, everything else being equal, better players tend to get more playing time. The more time an NFL player spends in a game, the more likely he is to get hurt. The more likely he is to get hurt, the more likely he is to show up as ‘Questionable’ or ‘Doubtful’ or miss a game. The more games one sits out, the less time they spend on the field. The less time they spend on the field, the less likely they are to get injured and/or re-injured. As such, they have less of a chance to show up as ‘Questionable’ or ‘Doubtful’. And I said I wanted to keep this concise? 
 
Here’s the deal. After going over numerous structuring exercises like the one above and employing advanced segmentation software, I was able to develop three simple yet generally encompassing decision rules to capture most complex situations. The idea was to keep this as simple as possible, but no simpler than that. The upshot to structuring the computer analysis revolves around the basic truth that the more an NFL player competes, the more likely they are to get hurt and end up ‘Questionable’ or ‘Doubtful.’ Exactly how I derived these rules is not in the scope of this offering. What really counts is how good these exact strategies are at helping you avoid players that will end up with “Bumps and Bruises” in 2009. That’s the bottom line. Based on how drastically these guidelines improve odds, I think its safe to say, we’ve done it. These findings apply to QBs, RBs, WRs and TEs. 
 
1) Long live the Iron-Man!
Players that started 14 or more regular season games in the prior year and never appeared as ‘Questionable’ or ‘Doubtful’ will show up ‘Questionable’ or ‘Doubtful’ 61% less often than all other players starting 14 or more games the prior season. Everything else being equal, you want to aggressively seek out and draft players that started 14 or more games last season and were never listed as ‘Questionable’ or ‘Doubtful.’ 
 
Editor's Note: To see if a player was an Iron-Man last season, check our Cumulative Injury Report to see if he started 14+ games and was never listed 'Questionable' or 'Doubtful.' If the player doesn't show up on the report, he wasn't listed on the IR at all last season and is an Iron-Man.
 
2) Bumps & Bruises Club
If a player started between 9 and 13 games last year and appeared ‘Questionable’ or ‘Doubtful’ two or more times, you should expect trouble the following season. In fact, players with multiple ‘Questionable’ or ‘Doubtful’ listings in a season in which they only started between 9 and 13 games are 232% more likely to appear as ‘Questionable’ or ‘Doubtful’ in the upcoming season than the average NFL player. If a player falls into this category that often spells problems and despite how good their prior accomplishments, the torch may be getting passed. Devalue accordingly and draft with caution. 
 
Editor's Note: To see if a player is a member of the Bumps & Bruises Club, check our Cumulative Injury Report to see if the player played between 9 and 13 games and appeared 'Questionable' or 'Doubtful' two ore more times. If so, draft with caution.
 
As a side note, on draft day, if faced with a virtual toss-up between drafting an individual player or a team defense, I’d suggest you minimize the injury risk and select a defense. Certainly the entire defense won’t end up on the IR list. In addition, this can be especially effective late in the draft. Historically, defenses are harder to predict. The chance of obtaining a top tier defense late in the draft is much better than predicting which journeyman wide receiver might suddenly have a breakout year. 
 

USING INJURY DATA IN-SEASON

 
Fantasy football owners often focus on their player’s health and playing status for the upcoming week. And with good reason! You certainly want to know if your player is hurt this week and if he’s going to play this Sunday. However, as you are about to see, this is only part of the equation. 
 
On average, does just one appearance on the official NFL injury report (due to injury) signal a possible decrease in fantasy football output for the remainder of the season? In other words, if a starting NFL player has been ‘healthy’ this season (never been listed as Probable, Questionable, Doubtful or Out) and then suddenly shows up Probable, Questionable or Doubtful for the upcoming week, what should you expect from this player during the rest of the season? 
 
To address this issue, over 1,400 regular season NFL games were examined. We focused only on NFL starters. To facilitate comparisons, each player was scored using the following scoring system: 1 point per 10 yards rushing or receiving, 1 point per 25 yards passing and 6 points for all TDs. 
 
Below, ‘Before Injury’ simply refers to all regular season games before the player appeared on the injury report. Likewise, ‘After Injury’ reflects all games after the injury was officially announced. 
 
Next, we carefully identified, examined and compared ‘Before Injury’ and ‘After Injury’ fantasy production for almost 500 NFL starters. 
 
Some amazing and useful things started to crystallize. Clear as day, player production decreased 60% of the time once a player appeared on the injury report. Folks, that means you can use the official NFL injury report as an accurate leading indicator for the rest of the season. 
 
If a starting NFL player is listed as ‘Probable’ (football related injury), ‘Questionable’ or ‘Doubtful’ just once, its not going to be the end of your fantasy season, but its not a great sign based on the true odds. Based on hundreds upon hundreds of occurrences, you can expect a decrease in fantasy football production relative to what that player had achieved earlier in the season. 
 
Within a given season, on average, you should expect a 15% decrease in fantasy football production among NFL starters after their first appearance on the official NFL injury report. By position, expected decrease is as follows: 
 
QB -14.5%
RB -11.0%
WR & TE -17.7%
 
There are several points to be made here. 
 
First, recalibrate your expectations after a player appears on the injury report. Don’t fall into the trap of thinking this player will be part of the 40% that keeps trucking along despite the announced injury. If you do, you’re going against the odds! In the long run, you won’t prevail. Also, don’t assume that after your player has supposedly  recovered, his production will return to ‘Before Injury’ levels. Perhaps it might, however the facts show, on average, he’s not going to deliver as much as he did before the injury. 
 
Next, don’t overreact. If your stud receiver shows up ‘Questionable’ one week, expect a decrease in his production the rest of the regular season. However, 85% of Calvin Johnson's output is still top-rate. 
 
Finally, be sure to take full advantage of the fact that you have a proven leading indicator, explore a trade. As soon as your player appears on the official injury report, flag him and look for another owner’s player. The other owner’s player should be equal in value to your player before he was reported as injured. After several weeks, assuming your guy is still playing and the other owner’s player has avoided the injury report, pull the trade! Most owners will almost completely disregard the fact that your player was ‘Questionable’ several weeks ago. However, you know, on average, his production will drop by 15%. In addition, if the two players in the trade were truly of equal value prior to the injury announcement, then everything else being equal, you have a 60% chance of winning the trade. 
 
Some might ask why trade if you “only” have a 60% chance of winning. Certainly, we’d all like more. However that’s the point. Unless you victimize a clueless owner, you’re not typically going to be able to get away with highway robbery; the other owner wants a good deal as well. To properly consider 60%, consider the long run! 
 
In fantasy football, the more you go with the odds, the more likely you’ll walk away a winner. In the case of doing a trade based on a historically-proven leading indicator, the more trades you make, the closer your results will be to the expected return. Several of my statistician friends call this the ”Law of Large Numbers.” The long run power of having a 60% likelihood of winning a single trade is illustrated below. 
 
Long Run Probability of Improvement (given a 60% chance of winning each trade)
Number of Trades = 1 Overall Chance of Winning = 60%
Number of Trades = 10 Overall Chance of Winning = 78%
Number of Trades = 100 Overall Chance of Winning = 93%
 
The table points out that if you make 100 different fantasy football trades, and have a 60% chance of winning each trade, the likelihood that you will have improved, as a result of all 100 trades, is 93%. Finding anything in fantasy football with a 93% chance of success is a winning habit you need to take advantage of. 
 
Now I realize that unless you’re a very busy trader and participate in several dozen leagues, you won’t be making 100 trades this season. The idea here is that over a period of time, and perhaps a number of leagues, you’ll work your way into the long run. In the end, you’ll have been a better owner 93% of the time as a result of ‘grinding out’ a 60% tactic. Applying this principle over and over to numerous techniques has allowed me to consistently win at fantasy football. In addition, I’ve seen this principle consistently get results for many successful fantasy football owners. In the long run, it simply works! 
 
One final thought for those of you thinking just one appearance on the official injury report can’t possibly signal a decrease in a player’s rest-of-season output. First, realize the burden of proof is on you. Exhaustive computer analysis of every starting NFL player shows it’s a fact. It’s not my opinion. Saying you can expect a decrease in fantasy football production in a given season, after a player appears on the injury report, is like saying you expect a fair coin to yield an equal number of heads and tails after you flip it 500 times. 
 
Think about these techniques and your situation. Then put them into action. Don’t deviate, and use them over and over and over again. If you do, in time, you’ll make it to the long run, and you'll be standing inside the winner’s circle!
Filed Under: Preseason, 2010