Pop quiz. What happens 50 percent of the time following a player suspension? The answer: that player doesn’t appear in another game.
When it comes to football, we’re good at predicting many things. We’re good at projecting performance on both weekly and seasonal levels. We’re even good at estimating injury risk, identifying which positions are most susceptible, how many games a player will miss and how they’ll perform when they return. But there’s one place where our insight lags: Suspensions. We don’t fully understand what a suspension signals about the rest of a player’s career.
Salary cap site Spotrac records over 300 player suspensions since 2011. That’s about 50 per season. Do different types of suspensions have different career impacts? How should fantasy owners value players who get suspended? When I first studied this topic at RotoViz, I focused only on substance abuse-related suspensions. This time around I’ll include other types of suspensions, and develop a fuller answer to those questions.
Some notes about methodology. I’ve measured the impact of suspensions by simply looking at the number of games a player appeared in before and after his suspension. I’m only looking at players who appeared in at least one game before suspension. I did this to weed out the least relevant players. I also eliminated anyone who played in 2016. By excluding those who are still accruing games, the study is effectively limited to players whose careers are over. I’ve also limited the study to the current collective bargaining agreement, which began in 2011. Some of the substance abuse penalties weren’t fully hammered out at that time, but the framework was there. Finally, undrafted free agents were assigned a draft position of 300. Got it? Let’s get to it.
Types of Suspensions Over Time
| TYPE | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | TOTAL |
|---|---|---|---|---|---|---|---|
| PENALTY | 0 | 2 | 1 | 2 | 0 | 0 | 5 |
| OFF FIELD | 5 | 5 | 0 | 4 | 11 | 4 | 30 |
| PEDs | 10 | 17 | 6 | 8 | 5 | 1 | 47 |
| SUBSTANCE | 9 | 9 | 16 | 22 | 21 | 13 | 90 |
| TOTAL | 24 | 33 | 23 | 36 | 37 | 19 | 172 |
“Penalty” refers to suspensions for in-game actions, like fighting or illegal hits. Only five such penalties fit the study’s criteria, so we’ll set those aside for lack of sample size and focus on the remaining three categories. Off field penalties are primarily violations of the personal conduct rule, most frequently arrest-related. Suspensions for performance enhancing drugs (PEDs) and substance abuse are treated separately. The NFL also treats them differently, and actually has two drug policies. One addresses PEDs while the other covers recreational substances of abuse like alcohol and marijuana.
Off Field Suspensions
I’ve lumped a few different types of suspensions together here. They range from domestic violence (Ray Rice), to vehicular assault (Derrick Coleman), but most are classified as either “conduct detrimental to the team” or “violation of the personal conduct policy.”
“Drafted ADP” shows the average and median draft position for players taken in the NFL draft, while “Total ADP” includes undrafted players (assigned a value of 300). Recidivism refers to the number of players in the cohort that were subsequently suspended again.
25 of 29 players were drafted, but 19 (65 percent) never played again. The 10 players that did return to action, however, averaged a healthy 13.3 games post-suspension. The main risk here is that the player won’t return to action. If they do, their career outlook is decent.
Examples
In 2012, Kenny Britt was suspended following a DUI. In his words, he “took it on the chin” and didn’t appeal his suspension. Since then he’s played in 73 games and had no further disciplinary issues.
Contrast that to Ray Rice and Prince Shembo. Once the video of Rice punching his fiancée emerged, he was suspended “indefinitely” which effectively ended his career. There’s no video of Shembo abusing a dog, but in a post-Mike Vick world, his two-game suspension also marked the end of his career.
PEDs Suspensions
First time suspensions for PEDs break down accordingly.
Eighty percent of this cohort are drafted players, averaging a third-round valuation, so it’s very hard to say that the career impact of a PEDs suspension is due to a lack of talent or team investment.
The typical first-time PEDs offender played about one season (15 games) after his suspension. That’s quite good, better than any other suspension category. However, about 23 percent of these players earned a second suspension. A second suspension, for any cause, is associated with much worse post-suspension career length.
Notably, 12 of 40 players never played again following their suspension. In other words, a first-time PEDs suspension has been a career-ending event 30 percent of the time.
Only six players earned a second suspension for PEDs, so it’s difficult to draw conclusions, but let’s take a look.
Here we see a much smaller percentage of games after suspension. Only three of the six players ever played again. It’s a small sample, but it fits the overall trend for second suspensions.
Examples
Ravens RB Kenneth Dixon was just hit with a four game PEDs suspension. I think it’s fair to wonder if he has any fantasy value at this point. Between the suspension and bye week, the most he can provide is 11 games in 2017. After missing the first four weeks, he’ll have to contend with Terrance West (or maybe a draft pick) on the ground, and Danny Woodhead through the air.
Jets wide receiver Jalin Marshall was recently suspended. He showed some ability as a rookie and on a wide-open Jets depth chart, he had a chance to make a bigger impact in 2017. He still might. But he’s gone from competing for significant playing time to being in danger of getting cut.
The other notable player here is Alshon Jeffery. He was suspended four games for PEDs in 2016, but returned to action thereafter. Arguably the most talented free agent WR, he wound up settling for an incentive-laden one-year contract. His salary ranks a respectable 18th among WRs, but his total guarantees come in just 41st. In other words, the NFL has already risk-adjusted their valuation of Jeffery. There’s a one-in-five chance he’s suspended again, and he’s only managed 21 games the past two seasons. I think he’s a very risky fantasy asset, and I’d try to sell high based on his recent signing.
Substance Abuse Suspensions
First-time substance abuse suspensions break down like this.
This cohort has similar draft value as the first-time PEDs cohort. They also played a similar number of games pre-suspension. It’s after the suspension where we see a big difference.
Players hit with a first-time substance abuse suspension average just eight games thereafter. That’s much less than the 15 post-suspension games averaged by the first-time PEDs offenders.
We also see a major difference in recidivism, with substance abuse offenders earning a subsequent suspension at a robust 50 percent rate. 21 of the 52 players in this group never played again, so a first-time substance abuse suspension has a 40 percent chance of marking the end of a career.
Things get even worse for second-time offenders.
This sample includes 21 players, of which Josh Gordon is actually the success story. He managed to play five games following his second substance abuse-related suspension. Nobody else managed more than one. Of these 21 players,17 (81 percent) never played again.
The set of players earning a third (or more) suspension is very small.
Much like the group of players earning a second PEDs suspension, we should be cautious about the small sample size. It’s worth noting however that none of these players have played after being suspended for a third time. And given the history of those receiving either a substance abuse suspension or multiple suspensions for any reason, we shouldn’t really expect anyone from this group to play again.
Examples
I talked specifically about Josh Gordon and Martavis Bryant last season, and really nothing has changed. Having been suspended multiple times, they face very long odds of being relevant again. If you’re still holding Bryant, take advantage of the news that his reinstatement is imminent and sell now. Remember that Gordon nearly made it back last year before ultimately going back to rehab and missing another season.
Le’Veon Bell is a premiere player and fantasy asset, and will be the highest paid back in the league. He’s also been suspended twice for substance abuse infractions. Add in a troublesome injury history and the odds are very good Bell misses time or has a shorter than hoped for career. You might consider selling for less than you would have previously accepted.
Conclusions
I think the median information included in the tables above is informative, but to keep this simple we’ll stick to averages for first time suspensions.
No type of suspension can be considered a good thing, but I believe substance abuse suspensions to be the most significant. They have the longest average length, and happen earliest in a player’s career. They also affect the highest number of players and have the highest recidivism rate. Players suspended for substance abuse infractions also have the shortest career length.
A PEDs suspension is also significant, but the lower recidivism rate and longer post-suspension career length mitigate the concern. Off field suspensions are in some ways the worst: 65 percent of affected players never see game action again. However, these types of suspensions are often easy to diagnose.
Where we get into trouble as fantasy football players is not properly accounting for the risks presented by suspension. I think we view them as similar to injuries: this player is out four weeks with a hamstring; that one is out four weeks due to suspension. Instead, we ought to use what we now know about the relationship of suspensions to career length to apply a more stringent discount to affected players.
If we realize that a player suspended for substance abuse will play on average just eight more games, and has a fifty percent chance of being suspended again, we might be less willing to hold onto them.
The analysis of how suspensions affect player career length is far from complete. Having established correlation, the question of causation is still open. It would also be interesting to look for positional or draft-related trends. In the meantime, I’ll post a table of suspensions at RotoViz, where you can engage me in follow up discussion.
