When NHL Seattle's newest quantitative analyst, Dani Chu, was younger, he played all sorts of "ball hockey, road hockey, but not organized hockey." Chu's brothers both played ice hockey and his dad is a lifelong NHL fan.
"We grew up watching NHL games as a family at home," recalls Dani, a Vancouver native. "During the 2010 Stanley Cup Final [the hometown Canucks losing in seven games to Boston], we went downtown to watch Game 7 on the big screens outside the arena."
Chu's dad's favorite player? Ron Francis, then Hall of Fame player and now general manager here in Seattle. Let's just say Chu knows his new boss has got game.
What perhaps impresses Chu even more is Francis' reputation as an early innovator in hockey data analytics, hiring one of the first specialists in the field during Francis' time as an executive with the Carolina Hurricanes. While interviewing with Francis, fellow hockey operations colleague Alexandra Mandrycky and NHL Seattle CEO Tod Lieweke, Chu realized a "common buy-in and alignment" about the value of analytics and "belief in hearing opinions."
Mandrycky is a fellow quantitative analyst herself and among the field's early pioneers at the NHL level. She signed a new NHL Seattle contract Nov. 1 expanding her responsibilities with the new title of Director of Hockey Strategy & Research.
Yes, that sounds like a cool job and it is.
"When I started working in hockey analytics back in 2015, only a handful of teams were even admitting they were using analytics [as part of player evaluation]," says Mandrycky. "Now the expectation is a team will have at least one quantitative analyst. Analytics teams are growing."
Mandrycky and Chu met at an analytics conference three years ago. Mandrycky was impressed and made sure to keep in touch with Chu, who was still a student but already conducting research across several sports, including basketball, football, soccer and rowing. In fact, Chu remains a student. He will graduate from Simon Fraser University is December with a master's degree in statistics under the supervision of esteemed PhD researcher Dr. Tim Swartz. He will begin his NHL Seattle position in early January.
Last summer, Chu interned at the National Basketball Association as part of the league's basketball strategy and analytics groups. He was part of a team that took first place in the Student Division of the 2018 NFL Big Data Bowl for their project using all available Next Gen data (powered by AWS) to identify pass routes, then analyze which routes on strong in both success rates and big-play capability.
As a former high school and club basketball competitor, Chu made it a point to give back to his alma mater by volunteer coaching and individual player development sessions. As a grad student, he and colleagues organized recurring "sports hackathons" as part of founding the Vancouver Sports Analytics Symposium and Hackathon group.
Chu's hiring is an impressive result for NHL Seattle. Not surprisingly, Chu turned down several other serious offers in sports to move a couple hours south.
"It comes down to the organization, how Alex, Ron and Tod carries themselves and live their values," says Chu. "They treated me like family as soon as I stepped in the door."
PRESS RELEASE: NHL Seattle Hires Dani Chu as Quantitative Analyst
"We are thrilled that Dani has chosen to join NHL Seattle among his other opportunities," says Francis. "His experience in sports and proven statistical expertise will be an integral part of how we build our team."
When asked for three words to describe Chu, Mandrycky squeezes in four: "Innovative. Communicator. Really excited!"
"I admire the way Danny approaches a problem and comes up with a way to share it visually," says Mandrycky, noting a skill that encourages front office executives, scouts, coaches and players to understand and more readily embrace the data presented.
Chu says he is attracted to the prospect of the coming mega-supply of NHL puck and player tracking data that will start filling team databases with the 2020 Stanley Cup Playoffs. Data points about to go galactic.
Mandrycky explains the league has been supplying about 300 "data events" per game. Teams employing a sports software data service (the vendors are typically startups and vary by team) can produce about 3,000 data points per game. The league's puck and player tracking will jump the number to 3,000 data outcomes per second of game action. The math: 3.6 million data events per 20-minute period or almost 11 million each game night.
"It's an exciting opportunity to be a part of the puck and player tracking from the beginning," says Chu. "Hockey is still the Wild West [especially compared to analytics in baseball and basketball]. We will now have enough data to prove out claims about a player's value."
Chu offers an example. A hockey fan sees a spectacular spin move for a goal by a forward. It that a repeatable play or unique act? Without the puck and player tracking, that answer is hard to provide. By the time NHL Seattle takes to the regular-season ice in October 2021, Chu and colleagues will have a data-packed answer to any number of repeatable plays from NHL scorers.
Another example: Early video tracking work has uncovered "zone entries" are a relevant statistics as players skate into the opponent's zone to create scoring chances. Chu says there are all sorts of variables that are not currently known or easily identifiable about a zone entry, such as strength/weakness of the other team's defenders, whether the opponent's formation is "crowding the blue line" (entry point) and who is joining the player entering the zone from his own team. Player tracking will account for every player on the ice, giving the quantitative analysts more data and, in turn, more answers when Francis, Olczyk and others pose questions of proof.
"One of the most exciting things about Danny is he will be the first of many for us," says Mandrycky. "We will build an infrastructure with the expectation of working together as a culture. Ron has been really vocal about that. And Tod right from the start said he wants our analytics people to be involved and listened to. As a research and development team, we will have open conversations [across hockey operations]."
Mandrycky says Chu will be helping out on the pro scouting side from Day 1, working with the five new scouts hired this fall, along with Francis, Mandrycky and assistant general manager Ricky Olczyk. The goal is identifying players who might be available for the 2021 NHL Expansion Draft or free agency beginning July 1, 2021. She says there will be times when the quantitative analysts and scouts agree on player evaluations and other instances when there is a difference of opinions.
"That's when we have the fun conversations," Mandrycky says.
As an indicator of how NHL Seattle will build a team with a cutting-edge organization and inclusive hockey operations culture, Mandrycky and the team has posted current job listings for another quantitative analyst ("we have received 412 applications as of Nov. 1," she says), plus a data engineer and a developer.
The data engineer will make it easier for quantitative analysts to focus on providing deeper and the most relevant insights to Francis and hockey operations colleagues. This role involves collecting all sources of data about players, then "optimize it" so it is a reliable and statistically sound set of data. Analytics people refer to it as "clean data."
Much fewer teams employ data engineers (Carolina is one) and it is arguably even more rare that NHL teams are hiring full-time analytics developers (Toronto has four). The developer turns the quantitative analysis being done into an easy-to-digest form, which will ultimately include NHL Seattle players.
"I welcome the collaborative aspects that are possible across analytics and hockey operations," says Chu. "I know I am going to rely on people to give me feedback and ideas. I am excited for those conversations."