In an example of Sportradar’s AI-powered analysis, the data predicts the likelihood of the route run by Arizona Cardinals receiver Larry Fitzgerald when he is on the outside of a “trips” formation.
NEW YORK — While analytics have already permeated the basketball discourse enough that anyone can identify that the most efficient shots are close-range 2-pointers and 3-pointers, those broad parameters lack pertinent details in evaluating the quality of a shot: the defender’s proximity and wingspan, his matchup history with the shooter and even his passing options. Those are among the essential, yet seemingly incalculable details, that will soon be fan accessible in the wake of Sportradar’s acquisition of Palo Alto-based Mocap Analytics.
“There’s so much richness that the data is able to tell us,” said Ashok Balakrishnan, Sportradar’s senior vice president of technology and product. “It’s just been untapped in the past.”
Sportradar, a global leader in sports data and player tracking, announced the deal Thursday, lauding Mocap’s sophisticated machine learning algorithms and six years of experience in harnessing artificial intelligence to analyze this information. The entire team at Mocap Analytics — best known for its five-year relationship with the NBA’s Golden State Warriors — will remain in place and serve as “a starting point to accelerate our own product development in the space,” Balakrishnan said.
“Mocap’s pedigree with the Golden State Warriors and in the performance analysis domain is really what underpins the credibility of what they’re bringing to us,” Sportradar communications director Alex Inglot said. “When you’re good enough for the Golden State Warriors’ analytics team, you’re good enough to provide fans what they’re looking for, that eureka insight that cuts through all the noise and cuts through all the speculation and gets you to the root of why what happened just happened.”
As the official real-time data provider for the NBA, NFL, NHL and NASCAR, Sportradar sits on a wealth of information that the company wants to package into “snackable” content for fans, particularly in the U.S. market. Already the company has launched the NFL Game Stream — which creates interactive visualizations of key plays, counting Sports Illustrated as its first partner to date — and will soon release an NBA companion, Game Stories, that will be a fully-automated, AI-generated feed targeting social media platforms and a “who’s who of digital publishers,” Balakrishnan said.
“We have a very strong infrastructure built up around getting accurate, fast data,” he added. “Now what we’re trying to do is, How do we take advantage of that? For the fans out there, we believe there is so much storytelling that can be done on top of this data.”
In the video example, Sportradar is able to discern the Arizona Cardinals’ formation and tendencies when star receiver Larry Fitzgerald is aligned in a particular place on a particular down.
Balakrishnan, a longtime Oregon resident and former Nike executive, mused about potential basketball insights relating to the fruitful interplay between guards Damian Lillard and CJ McCollum on his beloved Portland Trail Blazers. With AI poring through millions of data points accumulated over several seasons, the algorithm could discern a key component of their backcourt dynamics, whether it be number of passes or floor spacing or any other of a myriad options.
Sportradar has access to the entirety of NBA tracking stats, dating back to the first year of the league-wide use of SportVU in 2013. A new tracking system, Second Spectrum, is being installed for the coming season. Balakrishnan said one advantage of having this historical data is its use in training the AI algorithms. His team can load two years of data, have the machine predict the third season and then compare that projection with what actually took place. This can be repeated over and over to hone the AI’s accuracy.
“It’s really about offering a solution to anyone from broadcasters, digital publishers, media companies, fantasy sports but also things like brands and marketing agencies who want to really tap into how they activate around micro-events in competitions and contests rather than around broad [sponsorships of player or event],” Inglot said.
Sportradar has concrete plans for how to apply this technology as a tool to aid fantasy sports players either through analysis or as a way to track performance in real time. Balakrishnan said he believes “the fan ecosystem is ripe for disruption” with public analysis “plateauing in mediocrity” for years.
“Our initial focus is the fans,” he said. “How do we go after the masses and truly go after our vision of up-leveling the fan experience?”
