Player Missing Time for an Injury and the Impact on Fantasy Points When They Return

In honor of my two high draft picks, Fournette and Cook

Introduction to Gridiron AI:

We are a technology company leveraging Artificial Intelligence to make fantasy football projections each week. Our goal is to provide data-driven predictions that are unmodified by personal opinions.

Each week we provide the top 10 players at each position within the app on our site https://gridironai.com/. For free. We have a subscription for $5 a month which grants access to the full predictions and other features (saving lineups, customized scoring, etc.).

We are available on Slack to answer specific questions you have about your team or players. We also tend to release new data in our Slack channel as it becomes available and before it gets integrated into the app (e.g., the AI we created for game outcome and box scores), so there are reasons to visit us there!

Onward to the article:

Let’s try to answer the questions: If someone is injured or coming back from an injury, should we start them? Should we trade someone who is injured if we can get close to full value for them?

Let’s start by looking at the probability of playing based on injury designation:

A few things I noticed about these graphs:

• Removing Probable was the right call by the NFL, over 90% of the time, the person played.
• Probable was removed (starting in 2016), the probability of playing when marked as Questionable increased ~10%. It is likely some people who would have been Probable were labeled questionable.
• If someone is Doubtful, they aren’t likely to play. Going by the same logic the NFL used for removing Probable, they could get rid of Doubtful and just let people be Out. All-though, Doubtful is a proxy for: probably not playing, and the data shows this.
• When broken out by skill position, you don’t see significant differences in the likelihood of playing except for QB, even when a QB is probable, their probability of playing is lower than all other offensive positions.

Now for the good part, let’s look at in-season fantasy trends when someone is injured or playing while listed as Q.

The data:

• I took an average of the points from the three games before missing a game due to an injury compared to average three-game points after injury.
• The player had to have played three games in a season before the injury and three games in that season after the injury. I am just looking for the impact of in-season injuries and using examples that support that.
• The error bars are everything within one standard deviation from the mean for the given position and injury designation that caused them to miss a game.

General Trends:

• If a player misses a game for an injury during the season, there is a downside risk in the points they are going to score when they come back. The graph below is a more natural way for us to visualize this. There is skewing to the downside.

• It is rare for a player to come back and play when tagged with IR, in this dataset, only RB and WR had cases where a player had three games before and after they went on IR in a season.
• I get the Probable designation is gone because most players ended up playing but, when someone did miss a game, their performance decreased. I’ll make another analysis analyzing point differences when playing injured.

QB:

• QBs perform the worst when coming back from a game after missing an injury, given the positions of the error bars. I feel the conventional wisdom in the NFL for QBs is: an injured starting QB is more effective than a backup QB. This wisdom needs to be tested, after looking into this data, I don’t know if that is generally true from a fantasy perspective.

RB:

• RBs have minimal downside risk when missing a game when the designation is Questionable. The distribution of points looks pretty uniform, the downside risk is still present, but I am not second guessing myself with these players.

WR:

• In general, WRs have a volatile week to week scores and I am not controlling for said volatility. When a WR misses a game, the data is still skewed toward worse performance.

TE:

• The performance loss is the least pronounced. If you don’t have the one or two top TE’s so is their scoring. Given the prior two sentences, we can move on.

Let’s revisit our questions:

If someone is injured or coming back from an injury, should we start them?

There is a threshold/caliber of player that should be benched when coming off an injury if a game was missed(depending on who you have as their replacement). For example, when Cook comes back from his hamstring injury, he will be in my lineup but I shouldn’t expect the same performance as if he was healthy. I am going to use the following heuristics for the first game back from players coming back from injury:

• Start a near comparable replacement player. For example, when Devonta Freeman comes back if you have an RB 2 that is a certified starter, that player is the safer bet. Assuming the point below isn’t valid.
• Start the injured player if I need his non-injury adjusted points to win. If you have a guy like Cook or Fournette, you probably need their best self to win most weeks. I can live with them underperforming or not playing a full game for the upside.

Should we trade someone who is injured if we can get close to full value for them?

If you don’t need a lottery ticket from the player each week, finding a trade partner that values your injured player close to OG market value of the injured player could help minimize the overall risk to your projected score each week. If you have two injured players, it’s even easier to make this case to yourself.

General Musing on Injuries and FFB

In a short-season sport like football, it is a difficult situation when a starter goes down with an injury. This analysis shows the performance you are going to receive when someone comes back needs a discount rate. If you need points, making a trade and dumping the injured player could net you more points in the aggregate. I understand this is not always going to be the case but it’s something I am going to consider in the future.

Other Analysis That Needs to Be Completed

I don’t have time during the season to look into everything about injuries. Below is some additional analysis I’ll try to get to during the offseason.

• Points scored for a player over the rest of the season when coming back from an injury.
• Probability of re-injury
• Snap count changes

If you liked this article, give us a follow and a clap. Our goal is to keep publishing articles and analysis on topics that help you win your league.