Turf vs Grass
Players have argued about artificial turf for decades. We went to the injury reports. Since 2009, 36,873 new injuries showed up on a report the week after a game, and we matched each one to the surface of that game. Knee, ankle, foot and toe injuries were reported +8% more often after turf games than after grass games (19.1 against 17.7 every 10 team-games), and the serious ones +12% more. Turf came out worse in 15 of 18 seasons. That is a link, not proof that turf caused them: read the three limits below first.
Where the reports pile up
↳Bottom lineLower-body injuries are reported +8% more often after turf games (19.1 against 17.7 every 10 team-games), with a clear gap for knee, ankle, foot and heel.
?How to read this chart
- The figure
- colors each body part by how its injury rate on artificial turf compares with natural grass. Green-yellow (lime) = more injuries on turf, blue = more on grass, grey = about even. The stronger the color, the bigger the gap.
- The percentages
- read like this: +20% means that body part gets hurt 20% more often per game on turf than on grass. A ● means the gap is too big to be chance (it stays outside the 95% noise band).
- The bars on the right
- show the raw rates, new injuries every 10 team-games: green striped bar = grass, lime dotted bar = turf. The word "clear" = the same as ●; "within noise" = could be chance.
Muscles, arms and heads
↳Bottom lineClear gaps outside the lower body: head and concussion (+11%), neck (+15%). The rest comes out close to even.
?How to read this chart
- One row = one body part
- outside the legs and feet. Green striped bar = injuries every 10 team-games on grass, lime dotted bar = on turf.
- The percentage
- on the right is turf against grass (+ = worse on turf). "Clear" means the gap is too big to be chance; "within noise" means it could be.
Turf worse in 15 of 18
↳Bottom lineTurf came out worse in 15 of 18 seasons: one season can swing, the pattern across years is what counts.
?How to read this chart
- One tile = one season.
- The percentage is lower-body injuries on turf against grass that year: +10% means 10% more on turf.
- The fill
- rises with the size of the gap: lime = turf worse, blue = grass worse, grey = about even. A glowing border means that season's gap is clear on its own, not chance.
The positions with the biggest gap
↳Bottom lineDL show the biggest gap (+15%), QB least (−8%).
?How to read this chart
- One jersey = one position group,
- sorted from the group turf hurts most to the one it hurts least.
- The percentage
- is lower-body injuries on turf against grass. Underneath: injuries every 10 games on grass → on turf. "Clear" = too big to be chance.
Every home field, on its own surface
↳Bottom lineMost lower-body injuries: Allegiant Stadium (23.1, grass). Fewest: GEHA Field at Arrowhead Stadium (15.2, grass). Single stadiums are small samples.
?How to read this chart
- One tile = one team's current home field,
- with its surface: striped green background = grass, dotted background = turf.
- The big number
- is lower-body injuries every 10 team-games there, counting both teams. Ranked from most (#1) to fewest. Faded tiles marked "new" have too few games to rank.
Not all turf is the same turf
↳Bottom lineHighest: AstroTurf (20.9). Lowest: AstroPlay (17.0).
?How to read this chart
- One row = one playing surface
- (a grass type or a turf brand) with at least 60 team-games.
- The number
- is lower-body injuries every 10 team-games on that surface.
Same team, new surface
?How to read this chart
- Each card
- is a stadium that changed surface for good: before on the left, after on the right, with the years.
- The numbers
- are lower-body injuries every 10 team-games, both teams, in each period. Same team, same building: the cleanest comparison we have, but still a small sample.
MethodHow this is built
- What counts as an injury
- A player who appears on a team's official injury report with a body part that was not on the report before the previous game. That injury is credited to the previous game and to its surface. Illness, rest and personal reasons are left out. Source: nflverse injury reports, regular season, 2009 to today.
- What it cannot tell
- The report is a proxy, not a medical record: an injury picked up in practice lands on the game before it, contact and non-contact injuries are mixed, and minor knocks that never reach the report are missed. Practice injuries and missed knocks should hit both surfaces alike, but turf games also differ in roofs, cities, rosters and eras, so the result is an association, not proof that the surface caused the injuries.
- Rates and noise
- Every number is injuries per team-game, so the 4,855 grass team-games and 3,695 turf team-games are compared fairly. The ● mark and the word "clear" mean the 95% interval of the turf to grass ratio stays on one side of even.
- Surfaces
- Surface of each game from nflverse; hybrid grass counts as grass, every synthetic brand as turf. A stadium that changed surface is split into two periods.