For years, I’ve written about how the big shift that would change sports would be to figure out not just the best ways to treat sports injuries, but to prevent them in the first place. I even discussed it in the conclusion of my book That’s Gotta Hurt: The Injuries That Changed Sports Forever. And I’ve watched how professional sports franchises have been collecting all kinds of data, but we didn’t know yet what that data meant. We are now getting to the point where artificial intelligence (AI) might be able to predict injuries, and teams could use that information to prevent them. I discussed the idea in my latest newspaper column.
The benefits of predicting and preventing sports injuries
What would it mean to a professional sports team to be able to accurately predict when its star player would suffer a season-ending injury and make adjustments in training or games to prevent it from happening? It could mean saving millions of dollars in lost salaries, fan attendance and merchandise sales. It could mean the difference between winning a championship and failing to even make the playoffs. Artificial intelligence could be making it happen right now.
I talked to Billy Beane, the Oakland A’s executive vice president of baseball operations, about this idea back in 2015. As described in Moneyball, Beane ushered in the use of data to predict player performance and maximize a team’s wins for every dollar spent. I asked him about similarly using data to predict and prevent injuries. He told me baseball teams were quietly using data for that very purpose but were keeping it secret to maintain a competitive advantage.
Using artificial intelligence to prevent soccer injuries
Apart from being an orthopedic surgeon who treats athletes with these injuries, I care about this topic from a fan’s perspective. I’m a lifelong fan of Liverpool, the English soccer giant. Last season, Liverpool won the English Premier League for the first time in 30 years. This season, they have been ravaged by injuries, losing many of their key players for the entire season or long stretches of games. Liverpool currently sits in sixth place. I blame injuries as the biggest factor in their slump.
One of Liverpool’s greatest players was Steven Gerrard, who now manages Rangers of the Scottish Premiership. Gerrard has his team poised to claim its first league title since 2011. The team credits much of that success to a partnership with Zone7, a company that specializes in injury risk forecasting for professional teams around the world, resulting in a large drop in injuries.
Zone7 has generated attention in the soccer world recently. Getafe, a Spanish team perennially toward the bottom of La Liga, started working with Zone7 and saw quick success. In the 2018-19 season, when Real Madrid, Barcelona, and Atletico Madrid suffered between 27 and 47 separate player injuries, Getafe only had eight, the lowest of any team in the league.
Artificial intelligence enters the sports world
To be fair, I’m not promoting Zone7 specifically, as I expect there are several companies working on this idea. Last year, for example, Sparta Science, an artificial intelligence startup, announced a partnership with the NFL to collect data and help predict where a player is likely to be injured during a game.
The use of artificial intelligence to predict injuries could be a potentially game-changing opportunity for sports franchises. Just look at the cost of these injuries.
Recovery treatments: Fad or the future?
The cost of injuries in the NFL and NBA
According to an Associated Press study, NFL teams lost over $500 million to injuries in 2019. Jeff Stotts, an athletic trainer who tracks NBA injuries on his site In Street Clothes, found that in the 2018-19 NBA season, teams lost anywhere from just over $5 million in salary due to injury (the Sacramento Kings) to almost $40 million (Cleveland Cavaliers).
A theoretical example of how artificial intelligence could predict injuries
This is just an idea for how artificial intelligence could predict an injury. Let’s take an Achilles tendon rupture, which for pro athletes almost always means surgery and 9 to 12 months out of the sport.
Professional sports teams now have their players wear all sorts of devices to track movement, sleep, recovery and more. Let’s say that based on GPS and accelerometer data from hundreds of Achilles tendon injuries, the data shows that players who tear their Achilles tendons drop their left shoulder more than normal when cutting to change direction. Now just knowing that a player drops his shoulder to a certain level might not mean anything.
Based on following his personal data at every practice and during every game, though, AI could detect a change, showing that along with his recovery numbers dropping, his shoulder is dropping more than normal. AI might warn the performance staff that he is approaching high risk for an injury. The athletic training staff could then adjust training for that player, reduce his minutes in the next game, or sit him out entirely.
Can AI predict all sports injuries?
If AI can effectively predict and prevent injuries, and I do expect we are reaching the point where it will, I’d bet it will be more effective for predicting non-contact injuries, like ACL tears, Achilles ruptures, and hamstring strains than fractures and dislocations from hard tackles or collisions. But big data might figure out those injuries too.
We’re just at the beginning of what could prove to be a data-revolution in sports injuries. As both a fan and a sports medicine specialist who works with teams and athletes, I’m excited by the possibilities.
Note: A modified version of this article appears as my sports medicine column in the March 11, 2021 issue of The Post and Courier.