October 14, 2025
The GIST NFL uses AI to predict injuries, aiming to keep players healthier 
Andrew Zinin
lead editor
Editors' notes
This article has been reviewed according to Science X's editorial process and policies. Editors have highlighted the following attributes while ensuring the content's credibility:
fact-checked
reputable news agency
proofread
Injuries are an inevitable part of the NFL with all the high-speed collisions, crushing hits and high exertion necessary on every play.
Success each season often comes down to which teams can be the healthiest at the end and a string of injuries has already hampered preseason contenders like Baltimore, San Francisco and Cincinnati.
With teams investing hundreds of millions of dollars every season into their rosters, keeping those players available to play is crucial and any small edge has the potential to lead to better results on the field. To help achieve that, the NFL has turned to technology in recent years, partnering with Amazon Web Services on an injury prediction tool that uses data and artificial intelligence to help teams manage the health of their players.
"Fans want their favorite players on the field. The team owners certainly want those players on the field. The athletes themselves want to be on the field," said Julie Souza, the global head of sports at AWS. "Anything we can do to improve that and keep players healthy, that's sort of a noble endeavor."
A 'one-stop shop' for injury data
The Digital Athlete tool takes video and data from players on all 32 teams from training, practice and games, giving every team information on how hard its players have worked, whether they are at risk for more injuries, as well as helping them track leaguewide trends and benchmarks.
This is the third season all teams have had access to the Digital Athlete portal and medical staff say it has been extremely beneficial, calling it a "one-stop shop" for information that previously was never available at one source.
"Basically, it's giving you more information to ask yourself better questions to then make better interventions to make your process more efficient," said Tyler Williams, the vice president of health and performance for the Minnesota Vikings. "At the end of the day, if you sum sports science into one sentence: How can we measure and assess to make ourselves more effective and efficient."
Digital Athlete uses sensors in the shoulder pads, cameras and optical tracking to gather information from practice and games for every player on all 32 teams, similar to what NextGen stats does to determine who's the fastest ball carrier or how much separation a receiver generates on his pass routes.
But the amount of data is far different.
While NetGen Stats generates about 500 million data points in an entire season, Digital Athlete does that on a weekly basis, meaning the only way to parse through all that to glean anything meaningful is through the use of machine learning and AI technology.
"The sheer volume of data means somebody can't be sitting there with a clipboard or Excel figuring that out," Souza said. "This is absolutely a job for high-performance computers, machine learning, artificial intelligence, all of those things."
One of the strengths of Digital Athlete is its ability to aggregate the data from all 32 teams and more than 1,500 players to give training staffs and coaches better insights into which players might be more susceptible to getting hurt at a given time and what steps have helped reduce the impact of injuries.
How Digital Athlete is used
Teams have used it to help determine practice schedules for training camp, how hard they work the players in a given week of a season and what players or position groups have been pushed so hard that dialing back their work might prevent nagging soft-tissue injuries.
"You want to find a sweet spot that's not overworked or underprepared for football," Williams said. "The more football you play, the better at football you are, but the more fatigued you are. It's this seesaw balance of tactical and performance. How do we put the players out there to be the best versions of themselves, in the safest manner to have the longevity in the game?"
Williams said much of the data reinforces his prior beliefs but there are times it helps him catch something he might have missed. Having empirical evidence also can help him persuade a player who might need a day off or a coach that he has to lighten the load at practice or sometimes can push his team harder.
The model can tell the training staff how much a player has worked by tracking decelerations, accelerations, total workload on field, change of direction.
While the NFL said overall injuries have been reduced since the introduction of Digital Athlete, Williams said there are many factors involved and he can't be sure there's a direct causation.
"Everybody is always going to want the smoking gun that if we do A and pair it with B, we will get C," Williams said. "It doesn't ever work like that. Everyone wants to talk about well, this team's really good at preventing injuries. Nobody's preventing injuries. It's what type of recipe can you put together that mitigates the risk the best. With the more we measure, the more risk we'll be able to mitigate."
How it impacted rule changes
The NFL has also used the data gathered from these systems to model the impact of rule changes like the new kickoff that was put in place last season or the crackdown on hip-drop tackles. Digital Athlete was able to simulate 10,000 seasons to help model how the new kickoff rule would impact injuries.
It also has helped inform the league on things like which helmets are best at protecting players from concussions. NFL executive Dawn Aponte said the data helped the league make changes to the helmets for quarterbacks by putting more padding in the back of the helmet.
"Last year we saw the lowest number of concussions in the NFL since we started tracking them," Aponte said. "That really is something that we attribute to the fact of being able to look at all of this data and come up with better equipment, better-performing helmets, ways in which we make the helmets and manufacture them based on the types of hits and impacts these players are having."
Aponte said the initial reaction from some old-timers was to wonder if this was "junk science." But now she sees much more acceptance from everyone involved, from coaches to medical staff to players.
"I think this has now been presented as an additive tool," she said. "It doesn't take away from decisions of coaches that are going to do what they feel is best for that specific athlete, or the team at large. But when you're actually able to point them to certain things and say, hey, this is what we're seeing. This is causing X times more likely for this player to sustain an injury. When they start losing players, particularly in training camp, they pay more attention."
© 2025 The Associated Press. All rights reserved. This material may not be published, broadcast, rewritten or redistributed without permission.
Citation: NFL uses AI to predict injuries, aiming to keep players healthier (2025, October 14) retrieved 15 October 2025 from https://techxplore.com/news/2025-10-nfl-ai-injuries-aiming-players.html This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.
Explore further
Athlete performance model quantifies risk of future injury after returning to game play
Feedback to editors
