Sport analytics leverage AI and ML to improve the game

These three globally recognized organizations are using AI and machine learning to transform how players and coaches can not only extract but effectively wield valuable data to help achieve superior results.

Nearly 10 years ago, Bill James, a pioneer in sports analytics methodology, said if there’s one thing he wished more people understood about sabermetrics, pertaining to baseball, it’s that the data is not the point. The point is to use the data like a razor to cut through false convictions to find the truth.

“The reason that understanding is so difficult to build in baseball is because there’s an entire industry of people selling nonsensical ideas about the data all the time,” James said at the time.

In the years since author Michael Lewis popularized sabermetrics in his 2003 book, Moneyball: The Art of Winning an Unfair Game, sports analytics has evolved considerably beyond baseball. Computer vision, AI, and machine learning (ML) all now play a role.

Here are three examples of how international sports organizations are using AI and ML to change the way players and coaches approach their sports.

Improving player safety in the NFL

The NFL is leveraging AI and predictive analytics to improve player safety. Working with partner Amazon Web Services (AWS), the NFL has developed Digital Athlete, a platform that uses computer vision and ML to predict which players are at the highest risk of injury based on plays and their body positions.

Digital Athlete draws data from players’ radio frequency identification (RFID) tags, 38 5K optical tracking cameras placed around the field capturing 60 frames per second, and other data such as weather, equipment, and play type. During each week of games, the platform captures and processes 6.8 million video frames and documents about 100 million locations and positions of players on the field. It also pulls data from practices, for a total of more than 500 million data points.

“We’re running millions of simulations on in-game scenarios to tell teams which players are at the highest risk of injury, and they use that information to develop individualized injury prevention courses,” says Julie Souza, global head of sports at AWS.

Risk Mitigation Modeling can then be used to analyze training data and determine a player’s ideal training volume while minimizing injury risk.

Souza’s advice: Cultivate curiosity.

Like other data-driven initiatives, Souza says Digital Athlete uses data rather than hunches and instinct to understand what’s happening on the field during games and practices.

“It’s really about having a mindset where you’re curious,” she says. “The first thing is having a data strategy, having a foundation of data, and then asking questions of it.”

Computer vision transforms tennis coaching

For 2023’s Billie Jean King (BJK) Cup, the International Tennis Federation (ITF) partnered with Microsoft to develop an AI-based platform that provides in-match insights to help coaches fine-tune player performance.

The BJK Cup is the largest annual international team competition in women’s sports, with 16 national teams qualifying to compete for the prestigious title each year. Like the Davis Cup for men, it’s one of the few tennis competitions that allow the team captain to coach players during matches as they change ends.

The ITF partnered with Microsoft in 2021 to power its match insights platform for the BJK Cup. The platform uses ball-tracking cameras and 3D radar systems to generate live on-court match data, which is fed into Azure and combined with live score data to reveal insights into serving patterns, returns, and player movements around the court. The insights are then provided to the team captain via a dashboard on Microsoft Surface devices.

“We’re really starting to focus on how that data can be used to support the players, coaches, teams, and everyone involved behind the scenes on the performance side,” says Mat Pemble, executive director of IT for the ITF.

Capel-Davies’ advice: Focus on communication.

Jamie Capel-Davies, head of science and technical for ITF, says metrics don’t mean much if you can’t communicate them effectively in time to make use of them.

“One of the key things we were looking at was what were the most important metrics and how can they be communicated effectively,” he says. “The great thing about the app is it’s very visual and it also has a reasonable amount of customization.”

LaLiga adopts AI and ML for peak performance

LaLiga, Spain’s premier football division, is leveraging AI and ML to deliver new insights to players and coaches.

With the help of Microsoft, LaLiga has created a data analysis platform called Mediacoach, which uses Azure infrastructure to collect, interpret, and showcase insights from approximately 3.5 million data points captured in near real-time per match via 16 optical tracking cameras. These cameras are installed in each of the league’s stadiums to capture data on player and referee positioning, and the ball’s movements.

“With this huge amount of data per month, we’re able to offer stats and reports,” says Ana Rosa Victoria Bruno, innovation manager at LaLiga. “With 112,000 reports in the system and 8 million bits of information, it’s a huge amount of information for 42 clubs.”

One of the tools that’s also provided to broadcasters for fan engagement is a Goal Probability model, which leverages a range of variables, including the player’s line of sight (taking into account the positions of opposing players); distances between the ball and the goalkeeper, and the ball and the goal; and the distance and angle to the nearest defender to measure the probability of finishing a given scoring chance. The calculation also takes into account a player’s efficiency indicator based on variables such as the player’s ratio of goals per match and per shot.

Bruno’s advice: Create a multidisciplinary team.

Bruno says it required a multidisciplinary team of football analysts, business intelligence analysts, and the LaLiga analytics team to find success. “One of the challenges is, in order to turn this raw data into knowledge, we need not just data scientists, but also football analysts, UX experts, and coaches,” she says.