The NHL is actively testing Sportlogiq to provide the optical half of its planned puck-and-player tracking system. The league began installing dedicated cameras in four arenas two months ago to collect trial data. That collection process will continue during the Stanley Cup Playoffs.
At a demonstration in Las Vegas in January, the NHL announced that JogMo will provide the location-tracking ultra-wideband sensors in the puck and on the players’ pads. The company is a spinoff based on technologies developed at the Fraunhofer Institute in Germany.
David Lehanski, the NHL’s SVP of business development, has said the combination of optical and sensor systems is the “optimal solution” and described Sportlogiq as the “leading contender” to be that optical vendor.
“We’re hopeful we can get a good sample set that we can then look at the results from to determine whether or not we need to make changes, tweaks, or move on, or move forward,” he said.
Sportlogiq CEO Craig Buntin confirmed his participation in the NHL trial and offered a statement: “We are proud to be working with the NHL on its optical tracking system. We are really excited about the innovative approach they are taking — a multi-pronged strategy combining optical player, chip, and puck tracking. The NHL are ahead of the curve with their vision for how this information will transform the fan experience.”
Lehanski said the NHL has been working with Sportlogiq for about a year and a half and that ensuring accuracy and reducing latency have been the two primary objectives. Sportlogiq already provides its analysis to 25 NHL organizations on a team-by-team basis but does that work using video from television broadcasts. Having its own camera systems in a venue will provide a broader view of the rink and keep every player in sight at all times. Sportlogiq is also the official data provider for the Swedish Hockey League.
The original impetus for implementing a league-wide tracking system was primarily for broadcast and coaching analysis. Last year’s Supreme Court decision on PASPA, however, enabled the legalization of sports betting and created a new use case for the data. The NHL announced a partnership with MGM Resorts last year. The demands of betting require near real-time stats to fuel in-play wagering.
“There are some great optical tracking solutions out there, but we don’t want to get the data three minutes after the fact,” Lehanski said. “We want it like 200 milliseconds after the fact.”
While the NHL has trailed other North American sports leagues in implementing a tracking system—MLB created Statcast, the NBA has Second Spectrum, and the NFL uses Zebra Technologies—Lehanski said his league has taken a deliberate approach in order that its partners will be able to maximize its use. NHL commissioner Gary Bettman has said the league will own any technology that it uses.
This is part of a broader technology initiative from the NHL. In collaboration with Catapult’s XOS Digital, the league began outfitting teams with iPads on the bench for in-game review of video during the 2017 Stanley Cup Playoffs. In January, an additional program with Apple and SAP led to a Coaching Insights app on a fourth iPad for data analysis. Lehanski said a more comprehensive program geared toward general managers is under development.
Lehanski gave two examples as to why an optical system is needed in addition to the sensors. If two players are vying for the puck, the positioning of their sticks is important, he said, and that wouldn’t be captured by the player and puck sensors. Similarly, to better determine the scoring probability of a shot, knowing whether it was a forehand or backhand attempt is important. The optical system is needed for both.
Also, he added, while goalies are often assessed by the number of goals they allow, that’s often a reflection of overall team defense. Optical tracking could provide better context on that.
“If we’re able to track the puck and know who shot it, where was it shot from, how fast was the puck moving, was it deflected, was there someone standing in front of the goalie, add in contextual elements like the score of the game, the time of the game, we’ll open up a new opportunity for everyone to maybe re-evaluate the way that they appreciate goalie performance,” Lehanski said.
Though the system the NHL is building will be expensive, Lehanski said he hopes a lighter version could be offered to minor and amateur leagues. If many of the same base metrics can be captured, that could provide a standardized system for evaluating players.
For now, the focus is on implementing both halves of the tracking system at the NHL level. Sportlogiq’s computer vision-based AI system will learn and improve as it collects data. Lehanski said the league was strategic in choosing the home arenas of four teams expected to make deep playoff runs in order to increase the number of analyzed games. He added that he hadn’t bothered looking at the data yet, knowing its precision will get better. All data will be shared with the NHLPA as well.
JogMo’s sensor-based system is expected to be installed in all 31 home arenas next season. The rollout of the optical component may take longer, depending on the results of the ongoing trial and whether additional testing is needed. When both halves are running and synced together, the NHL hopes the venture will revolutionize the understanding of the sport with immediate, data-driven commentary.
“We have 60 minutes of actual play, and the stoppages are very short,” Lehanski said. “If we want a casual sports fan to really understand what’s going on the ice—how fast the player’s skating, how fast the puck is moving—it’s really hard to tell those stories in the context of a game.”