(Bloomberg) — After years of making inroads in sports ranging from baseball to soccer, the computers are taking over hockey now too.

The National Hockey League signed a deal with Amazon.com Inc. for advanced data, analytics and video capabilities to give fans and teams high-technology features that have already filtered into other sports. That could be a selling point for legions of hockey-lovers whose only choice during Covid has been to watch games from home.

Soon the speed of a puck may appear on viewers’ screens or they may see analytics predicting the outcome of a face-off or the probability of a team scoring on a power play, said Matt Garman, vice president of sales and marketing at Amazon Web Services.

“We’re going to pretty quickly look for some of those analytics that, in real time, can provide interesting data,” Garman said in an interview. “Fans like having that extra insight, and it adds another dimension to watching the game.”

The analytics will be introduced throughout the upcoming season, Seattle-based Amazon said Wednesday. Financial terms of the deal weren’t disclosed. The NHL is playing an 868-game regular season, with 56 games per team. Because of border restrictions related to the pandemic, the league realigned its divisions with a North Division that consists of the seven Canadian teams.

Amazon will build on the NHL’s existing Puck and Player Tracking system and use machine learning to develop advanced analytics and metrics. The technology giant also will help the league automate video processing and store footage in a central repository that’s easier to search and retrieve. The NHL also will get footage from a series of new camera angles that typical broadcast cameras don’t capture.


It’s not just fans who will benefit, according to Amazon. Partnerships with other leagues — including the National Football League, Formula 1 Racing and German soccer league Bundesliga — have shown that much of the value from the tie-up will be reaped by players and coaches, Garman said.

“There is a lot of value to the teams in doing and applying machine learning to improving athlete performance,” Garman said. “Over time, that’s an area that the NHL will encourage.”

(Updates with season plans in fifth paragraph)

For more articles like this, please visit us at bloomberg.com

©2021 Bloomberg L.P.