Using a computer vision technology developed at McGill, Sportlogiq is ushering professional hockey into the 21st century. The Montreal startup tracks hockey players’ on-ice movement to generate detailed statistical analysis. Already, ten NHL teams and two broadcasters have signed on to access the company’s strategic data.
By Kathryn Jezer-Morton
Mehrsan Javan, co-founder and Chief Technology Officer at the Montreal startup Sportlogiq, used to be a sports fan. As a PhD student in Computer Engineering at McGill, he loved watching basketball, and he was starting to catch hockey fever. Not anymore. “When I’m watching a game now, I’m thinking, ‘What would we have trouble tracking? How would we improve? I’m always thinking about the technology.”
Watching the game – closely – is what Sportlogiq does. Javan co-founded the company with CEO Craig Buntin, also a McGill grad (he earned his MBA in 2013) and former Olympic figure skater. The company – partially backed by American billionaire and owner of the Dallas Mavericks Mark Cuban – uses proprietary software based on Javan’s PhD research to track hockey players’ on-ice movement to generate detailed statistical analysis.
While game data is an established part of the cultures of some professional sports, professional hockey is by and large a late adopter. “Hockey guys are old-school guys,” says Javan. “However, the paradigm has recently shifted and some teams are realizing the value of the data and statistical analytics.” Sportlogiq currently provides data to ten NHL teams and two broadcasters, RDS and SportsNet. “We have to dominate the whole NHL,” he says, before moving on to other sports.
Javan didn’t set out to revolutionize statistical analysis in sports. As a PhD student, he worked on computer vision – a field of study in which computers analyze the content of images and videos. Javan developed surveillance software for identifying unusual activity in crowded areas, like train stations or sports arenas. “The technology was fully transferable from one domain to another,” he said. Today, that software is giving Sportlogiq a competitive advantage in the growing field of sports analytics.
Sportlogiq can take footage from any camera or video stream – even a smartphone – and use it to generate data about a hockey game. Other stats generators rely on costly static cameras installed around the game or radio-frequency identification devices (small electronic devices consisting of a chip and an antenna). “We are relaxing the constraint of using hardware,” explains Javan. Sportlogiq’s software has two modes of analyzing gameplay: by tracking player location, and activity recognition. That information is then used by the game analytics engine, which builds a statistical game model in order to generate the high-level descriptive and predictive statistics of a game. This enables Sportlogiq to address some questions such as: How do certain players respond in certain game situations? Are they making the best decision? What are their playing styles? How does the scoring chance change with a certain line? This puts Sportlogiq ahead of the competition, which rely solely on player location as a way of analyzing play.
Because teams can access data about opponents’ gameplay, Sportlogiq can provide strategic information for coaches. “We can see patterns. We can see what the save percentage of a goalie is when a shot comes from different regions. Let’s say I have a player who shoots from the same location all the time, and we know that the save percentage from a certain goalie from that area is 100 per cent. So we know to never shoot from that angle against that goalie. Or, if they want to trade players, or are looking for a player with specific characteristics and playing style, [Sportlogiq] can find that player.”
Montreal’s startup community has been a boon to Sportlogiq. The founders were supported by local technology incubator Tandem Launch and angel investors Anges Capital Québec.
There’s one Montreal institution, however, that has yet to embrace Sportlogiq’s innovations: The Canadiens. “When Mark Cuban came to visit here, he was kind of surprised that the Habs aren’t using it. In an interview he did with RDS, he mentioned that. They don’t see any value in the data. One day, they’ll realize.” After this past season, paying attention to the data might deserve some serious consideration. If the call comes in, Javan and his team will be ready.