Auburn’s Chuma Okeke tore the ACL in his left knee during the NCAA tournament in March, an injury that was one of a few reasons his selection by the Orlando Magic at No. 16 in last week’s NBA draft drew mixed reviews from online pundits.
But the Magic picked Okeke armed with potentially more data and information than any franchise has ever had. In February, Orlando signed an agreement with Chicago-based analytics provider STATS for exclusive use of its new AI-powered AutoSTATS tool for the next year and a half. The platform uses computer vision to mine through television broadcasts of college basketball games and compile player tracking data.
“Now they can go back and basically look at three, four years of the college career of the player with the player tracking, and using all the advanced analytics to determine exactly [whether] this guy is the best guy for us,” Arjan Paulussen, STATS VP of global accounts, said in an interview at the Hashtag Sports conference in New York City on Thursday.
STATS announced the deal at this year’s MIT Sloan Sports Analytics conference, which Paulussen said prompted about 10 other NBA teams to make immediate inquiries about the tool. Often, pro teams can only scout a player a handful of times in person, but AutoSTATS generates data and video clips from every televised game of a player’s career—72 in the case of Okeke.
The soccer version of this AI analytics tool, STATS Edge, already has an early success story. The only national team to use STATS Edge at the 2018 men’s World Cup was Croatia, which advanced to its first final, before losing to France. Among the features of that program is the use of machine learning to identify complex playing patterns and distill those findings into a scouting report.
Croatia analyst Marc Rochon said in a video prepared by STATS that the Edge tool noted how France had a more counter-attacking style of play and made more one-touch passes in its World Cup games than before the tournament. He was able to quickly include that information and accompanying video in his presentation to coaches and players, without having to pore through every second of action himself.
“We had to be able to provide a lot of information very quickly in concise packages, so the playing styles [data] was very useful to me,” Rochon said.
Scouting set pieces was another important part of Croatia’s information gathering. An analyst can select parameters, see a grid showing where kicks go, and then choose an area to see all the relevant video clips.
“You say ‘I want to have an in-swinging free kick from this side of the field and I want this guy to play it—I want to see all those. Where do they go?’” Paulussen said.
Paulussen said an enhanced version of the soccer product will be released later this year and that a number of other major clients across Europe are looking at adding STATS Edge. The update will include penalty kick analysis that can project where a shot will go based on a player’s approach.
“We like to say it gives our clients ‘an unfair advantage,”’ Paulussen said.