With centuries of tradition behind it, tennis as a sport has been highly resistant to change. Other sports have been quick to embrace the use of data and analytics to transform how athletes are recruited, trained, and prepped for competitions, how they adjust to changing circumstances during play, and how they break down successes and failures after competition.
“It’s fair to say that tennis has lived up to its roots as a traditional sport,” says Mat Pemble, executive director of IT for the International Tennis Federation (ITF), the sport’s governing body. “We’ve not been one of the fastest sports when it comes to embracing new technology and data analytics, particularly on court.”
Still, the appetite for innovation is there. Electronic line calling is a prime example that Jamie Capel-Davies, head of science and technical for ITF, points to. So too is its embrace of smart racquets and wrist-worn devices, the first wave of which proved limiting in that they could provide data on how fast a player was swinging a racquet, for example, but couldn’t collect data on the outcome.
“You didn’t know whether that was a good shot or a bad shot or any other context around what was going on when you played it,” Capel-Davies says.
Serving performance data courtside
In the early 2000s, the ITF started working with Sony’s Hawk-Eye Innovations, whose computer vision system uses timing data from multiple high-speed video cameras to triangulate the position of the ball in relation to the court. The technology made its debut at the Australian Open in 2003 and Wimbledon in 2007, and it provides the foundation for electronic line calling for the sport.
“One of the byproducts of that, because you’re tracking the ball to see whether it’s going to land in or out, you actually get a lot of data through the process: how fast people are hitting the ball, where they hit it from, and where it lands,” Capel-Davies says.
The ITF was quick to make use of that wealth of data on the presentation side of the sport but struggled to unlock its value for competitors. So, in 2021, the ITF partnered with Microsoft to power its match insights platform for the Billie Jean King Cup (BJK Cup) with the idea of transforming performance. The BJK Cup is the largest annual 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 is one of the few tennis competitions that allow the team captain to coach players during matches as they change sides between games.
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 provide insights into serving patterns, returns, and player movement around the court. Those insights are provided to team captains during the BJK Cup finals via a dashboard on Microsoft Surface devices.
“We’re really starting to focus on how that data can be used to support the players, the coaches, the teams, everyone involved behind the scenes on the performance side,” Pemble says.
By allowing court-side coaching, the BJK Cup presents a unique opportunity for the ITF to showcase the platform’s match-insight capabilities.
“They’re getting live data coming through as the match progresses, and that gives teams an opportunity to look at how they are performing against the game plan,” says Capel-Davies. “Does it need updating or adapting to the situation?”
Fine-tuning performance with in-match data
The ITF worked closely with Microsoft and representatives from BJK Cup teams to develop the analytics and their presentation to ensure the dashboard would provide meaningful insights. The captain is only allowed a few moments to coach their players when they change sides between games.
“One of the key things that we were looking at was what were the most important metrics and how can they be communicated effectively,” Capel-Davies says. “The great thing about the app is it’s very visual and it also has a reasonable amount of customization.”
For instance, it can be used to display serve placement and returns to help captains and players identify patterns. It can display how players return on breakpoints. After last year’s finals, the ITF worked with Microsoft to add new features based on feedback, including a visualization called “serve plus one,” which visualizes the third shot of a rally.
Beyond court-side insights, Capel-Davies says the platform also provides teams with value before and after matches. Prior to matches, the platform can provide insight into opponents to help develop a game plan by understanding their strengths, weaknesses, and tendencies. After the match, the platform can be used to perform a post-mortem, providing insight into what worked, what didn’t, and what players can improve in the next match.
For now, the system is limited to the finals because it requires deployment of four to 12 cameras that are calibrated to a specific court. It’s also not used on clay courts because those typically don’t have electronic line calling.
“Tennis courts are the same size, but in practice the lines don’t always go in exactly the same place, so there’s a process by whereby the system is calibrated to the actual position of the lines and you also have to know something about the topography as well,” Capel-Davies says. “You don’t assume that the court is perfectly flat; you have to map that.”
For now, in-match insights are available only at the BJK Cup, but Pemble and Capel-Davies believe change is coming.
“There’s been a sort of slow, slow burn in tennis in terms of innovation and coaching, and I think this is demonstrating how good, quality coaching and good information to base that coaching on can really enhance the game,” Capel-Davies says.
“The appetite is there, and I think that extends from the largest tournaments, the grand slams and the world championships, the Davis Cup and the Billie Jean King Cup. But now I think that’s moving down to some of the lower levels of the game,” Pemble adds. “The accessibility of the technology and the data now is so much greater than it was even two, three years ago. It’s filtering down to the club level and a lot of the systems require much less in terms of technology support and teams to put in place and run them. We’re seeing a big amount of development across AI-based camera tracking systems. It’s automating a lot of that data processing and analytics generation.”
Digital Transformation, Machine Learning, Machine Vision