Data story · September 2, 2026

Injury Risk by Position, 2015 to 2025

I wanted to know which fantasy positions actually miss the most time, and whether any of the things people say at the draft table (workload, size, running style, mobile quarterbacks) show up in the record. So I pulled every player drafted in the first 150 picks from 2015 through 2025 and classified every game each of them missed. This is what came out. The full per-player table and a CSV are further down, so you can check any row.

Some definitions, because the numbers depend on them. A player-season is one drafted player in one year. Drafted means inside the top 150 of FantasyPros consensus PPR ADP that August, taken from Internet Archive snapshots of their ADP page. A missed game is a game the player's team played in which he did not record a single stat. That last definition is blunt on purpose. It cannot tell a healthy scratch from a benching, so I use the NFL's weekly injury reports and the roster reserve lists to sort injury from suspension, holdout, illness and everything else.

The sample is 1,509 player-seasons: 534 running backs, 601 receivers, 171 tight ends, 203 quarterbacks. Every chart has its table behind a "Show the data" toggle. The method, with its weak points, is written out at the bottom.

Running backs miss the most games. Tight ends get hurt slightly more often.

Running backs played about 79 percent of their possible games, the lowest of the four positions. Part of that gap is not injury. Backs lose jobs to committees and benchings more than anyone else, and the occasional holdout (Le'Veon Bell in 2018, all 16 games) lands here too. On injury alone the positions are closer than I expected. Tight ends and running backs each lose about 2.6 games a season, receivers 2.3, quarterbacks 2.0.

About 61 percent of drafted tight ends missed at least one game to injury in a season, 56 percent of running backs, 53 percent of receivers, 38 percent of quarterbacks. Roughly one in four at every position except quarterback missed four games or more. I would not read much into the tight end lead. The sample is 171 seasons, and if you swap a few of the worst ones out it shrinks.

Quarterbacks are the outlier in a different way. They get hurt least, but 42 percent of their injury stints ended the season, against 27 to 34 percent at the other positions. Some of that is how teams handle a hurt quarterback. Once he is out for a month, there is often no reason to bring him back.

Look up any player

Every player-season in the sample, with the games he missed and what the pipeline thinks happened. If a row looks wrong, it probably is the pipeline and not the player, and I would like to hear about it.

Ten cases you can check yourself

These are ten seasons most readers will remember, next to what the table says about them. My first pass got two of these wrong. It labeled McCaffrey's 2020 ankle as a shoulder because the code looked at the wrong report, and Anthony Richardson's 2023 shoulder as a concussion for the same reason. Both are fixed, and the rule that caused it is described at the bottom. There are almost certainly others I have not found.

The type of injury tells you more about the missed time than the position does

A stint is one continuous run of missed games from a single injury. When I grouped every stint by body part, the spread between injuries turned out to be much larger than the spread between positions.

Hamstrings, groins, quads and concussions cost two to three games and rarely end a season. Ankles, shoulders, knees and feet cost three to four and a half games on average, and a third or more of them end the season. Back and neck injuries ended the season 59 percent of the time. Of the nine Achilles stints in the sample, eight were season-ending, and the ninth was McCaffrey's 2024 tendinitis, which cost him eight games.

Stints of eight or more games are a minority of injuries but the majority of the damage. They account for 41 percent of quarterback games lost, 37 percent for running backs, 32 for receivers, 31 for tight ends. Receivers have the most one-game absences of any position, 180 of them in eleven years.

Each position has its own signature injury

Running backs lose 30 percent of their injury games to knees and another 19 percent to ankles. Nothing else is close. For receivers it is more even: ankles 19 percent, knees 18, hamstrings 16. The hamstring is the most common receiver injury by count (87 stints), it just does not cost many games each time.

Quarterbacks lose games to the throwing arm. Shoulders lead at 20 percent, hands and arms 13, then feet, knees and ankles at 11 to 12 each. Concussions are the second most frequent quarterback stint by count, 15 of them, but short. Tight ends look like a mix of the other two: knees 20 percent, ankles 18, hamstrings 14, backs and necks 9.

Most of what we argue about at the draft table does not show up in the data

For this section I measured each player's usage from the season before the one being tested, since that is what you can see on draft day, and compared it with games missed the following season. I split each metric into thirds (low, middle, high) and looked at whether the thirds differ. One limit up front: there is no public data on where a receiver lines up, slot or outside. Depth of target and share of targets over the middle are the closest stand-ins I have, and they are not the same thing.

Almost nothing moved. Carries per game, the share of a back's runs that went between the tackles, how much of his work came as a receiver, his weight, a receiver's depth of target, yards after the catch, height, targets per game. In every one of those the low, middle and high thirds land within half a game of each other. The 350-touch idea is the one I expected to see and did not, quite. Eighteen backs had 350 or more touches the season before, and they averaged 3.9 games missed against 2.4 to 2.8 for everyone else. Eighteen is not enough to call that anything. I am reporting it because leaving it out would look like hiding it.

Draft price does not help either. Players in the starter tier and the bench tier at the same position missed about the same number of games. Paying up for a player buys you his production, not his health.

Three things did move. Quarterbacks who ran fewer than three times a game the year before missed about 1.6 games and 30 percent of them missed any game at all. Quarterbacks who ran six or more times a game missed 3.2 and 54 percent missed a game. That second group is only 24 seasons, so treat the size of the gap loosely, but the direction held in the middle bucket too.

Age matters for backs and receivers, though less than I assumed. Backs 29 and over missed 3.1 games against 2.5 for backs 24 and under. Receivers 29 and over missed 2.9 games and 32 percent of them missed a month, against 2.1 games and 20 percent for the youngest group. The tight end result is the one I trust least: tight ends under 250 pounds missed 3.2 games and tight ends over 255 missed 1.8, but that is 57 seasons a side.

Where the games go also depends on the quarterback's job. The running quarterbacks lost their games to ankles (29 percent), feet (18) and backs (16). The pocket group lost theirs to hands and arms (21 percent), Achilles tears (18), knees and concussions (15 each). Two Achilles tears, Rodgers in 2023 and Cousins in 2023, make up most of that 18 percent, so it is a small-number artifact more than a pattern.

A prior injury raises the odds, but not by as much as "injury-prone" implies

Players who missed no games to injury the year before still missed a game 47 percent of the time the next year, averaging 2.2 games. Players who had missed eight or more games the year before missed a game 65 percent of the time, averaging 3.6. So the label is not nothing. It moves the chance of a missed game by about 18 points and the expected games missed by about one and a half. It is not the difference between a safe pick and a landmine.

Recurrence of the same body part is more specific. Knees and ankles came back the next season 18 to 19 percent of the time, feet and hamstrings 13 percent, shoulders 7, concussions 4. If you are worried about a repeat, the lower body is where it happens.

Week 3 and weeks 16 to 17 are when it happens

Ninety-two injury stints started in week 3, against 46 to 64 in a typical week from October on. I do not have a clean explanation. The usual one is that bodies that were held back in the preseason are getting their first full-contact reps. Stints starting in weeks 16 and 17 (68 and 81) are inflated some by teams shutting hurt players down once they are out of the playoff race, so the late bump is real but smaller than the chart makes it look.

Is it getting worse?

Running backs lost between 10 and 24 percent of their games to injury in any given year, and the last five years sit at the higher end of that. Some of it is the seventeenth game. The 2023 quarterback spike is Rodgers, Burrow, Cousins and Richardson all going down the same year. With eleven seasons and 40 to 60 players per position per year, I would not call any of this a trend yet.

How I am using this

The practical version is short. I assume my running backs and tight ends will miss two to three games and draft depth I would actually start, rather than counting on a handcuff. When a player gets hurt, I pay attention to the body part before the reporter's timeline. A hamstring is usually two or three weeks. A knee, foot or ankle is usually a month and sometimes the year. Back, neck and Achilles injuries are the ones where I move on. I do not fade a back for his workload or his running style, since neither showed up here. I do discount backs and receivers from 29 on, and I keep a real backup behind a quarterback who runs.

If you find a row in the table that is wrong, or a case the method mishandles, I want to know. The whole point of publishing the data is that it can be checked.

How I built this