Chayka, who co-founded the sports data company Stathletes in 2010, chose to invest time in Finland, even during a busy time in her schedule with the Stanley Cup playoffs. She forged new connections during the thrilling tournament in nearby Espoo. The 35-year-old Canadian entrepreneur was rated #95 in The Hockey News’ People of Power and Influence this year for her groundbreaking work with Stathletes, which is based in St. Catharines and Toronto, Ontario.
In addition to working extensively in the NHL, Stathletes has contracts that span more than 20 leagues worldwide to provide data tracking and analytics. That data supplies the basis for everything from in-game coaching decisions to selections in the annual NHL Draft.
For women’s hockey fans, nothing tops Olympic competition. Conversations with Trevor Pilling, the long-time head of programming for CBC Sports, led to Stathletes supplying publicly available data for the landmark 2018 Olympic women’s hockey tournament in PyeongChang, Korea.
When did you get interested in applying analytics to women’s hockey?
We actually pitched the CWHL four or five years ago, because it was always my mandate that if we did men’s hockey at a certain level, I wanted to do as much as I could for the women’s game once we were profitable and had the resources within our office. College hockey is great. Historically, both the CWHL and the NWHL fed into the national teams. And there’s a hotbed for amazing women’s talent around the Toronto area. There’s no real excuse not to do it. It was just a matter of timing. So I saw the NHL not going to PyeongChang as sort of a breakthrough where maybe women’s hockey would have a bigger spotlight.
How would you describe your feelings about women’s sports?
For me, I really do enjoy them. I watch tennis and golf as well. As a young adult, I played all sorts of sports and I found it just as interesting – and in some cases even more interesting – to watch the women battle it out. Like in tennis, they have more rallies than the men. Volleyball is sometimes more interesting too. So it totally is sport-specific or based on the type of athletes you get.
That said, I always felt like there needed to be more attention paid to women’s hockey and that the audience would grow. Coming from an analytics background, having a company that works in the NHL, it just seemed like a natural fit to really do some extra projects in the women’s game.
What were your preparations like heading into PyeongChang?
We’d already done some women’s projects. We did some women’s college hockey as well. So it wasn’t very foreign to us in terms of how we could roll that out or what that would look like. It was basically a process of pitching the right people.
I think we found some good people who realized that women’s commentary was underserved. Instead of just telling their stories, you could actually talk about their performance, talk about how they played in quantitative terms, not just someone’s opinion. They saw it as strengthening the game and the understanding of the women’s game in particular, not just their usual storylines around it.
It was about a year out that I pitched. Then within three or four months, we had a deal to go forward. It was short enough that it was just a nice runway to do a tournament. We’ve done the World Juniors. We do a lot of tournaments on the men’s side. So it’s just sort of taking the same approach.
I think everyone will agree that it’s tough to see a championship decided on a shootout. Even when Canada wins! [laughs] We like seeing the 3-on-3 hockey now in the NHL. I think both the USA and Canada are very close teams. They’re very competitive teams that are a lot closer than people perceive, and the games are as well.
In the gold medal game, Canada did a great job of competing despite most metrics being against them. Fatigue from the U.S. attack could have helped the Americans secure the win in the shootout.
The Americans had 30 more shot attempts, and eight more on net at 5-on-5. There were high-quality rebound shots for both teams. Team USA had substantially more power play time, which led to a goal on special teams. They also had six odd-man rushes to Canada’s four.
It comes down to a lot of the very detailed aspects of the game, which analytics can actually help with a lot, whether it’s coaching, or player usage. I foresee as more countries compete and jostle to be number one that this could be a difference-maker.
Interestingly, we found that when the U.S. and Canada meet in gold medal games, the Americans have consistently outshot their opponents over the last decade by about five shots per game on average.
That could be just a matter of training. They just have this mantra of getting more pucks on net. But I think when you look more deeply into analytics, it’s about the type of chances that you’re creating, too.
That really came out in the Olympics, too. In the gold medal game, there were two or three really high-quality chances that weren’t scored on for both teams. That’s sort of the determinant of who won or lost. There was even one for the U.S. that they missed where they would have won in regulation, which I think was like a tap-in with the goalie down. So there are those momentum-changers that you look at from an analytics perspective and it really changes your gaze on the game.
I think as a coach or GM, too, it’s hard when you go into OT. You forget those things to a certain extent. You just file it away and you keep going. So it’s nice to see that overall picture of what that game actually looked like.
I tried to attend as many games as I could. We were doing the men’s tournament too, so it was a bit of a grind! And then the NHL was still playing, too. So we had our normal workload back in the office. I think most of the games were at an OK time, but there was one women’s game that started at 2:40 am or something. So we didn’t like that call to start working on the game at that time! [laughs] But otherwise, it was pretty straightforward.
I think they had a hockey program at noon on our time in North America. So it was nice to see the results of what we did go to air and be used. That’s a bit unique. Usually with analytics, if you’re working on the team side, there could be a lot of competitive advantages. So they don’t want to share. But women’s hockey is a bit more open, I think. That’s sort of a nice aspect of developing and growing the game instead of pure competition.
How did you manage your time during the Games?
I was there just to support. Actually, I had lived in Korea. Kind of a random story, but in fourth-year undergrad, I lived in Korea for three months. So I felt pretty comfortable with the country and how things were. It was pretty easy for me to operate there and just make sure there were no technical issues. That’s always the fear of any sort of IT/analytics type. We definitely have learned to have backups to backups!
I tried to go to as many games as possible, especially the women’s games. That’s what was meaningful to me, because there are times when I go to seven or eight men’s games a week. It was really refreshing with the spirit of the women’s game. I think you can feel the passion. This is what they’ve been working really hard for, for four years. So anyone who works in that environment, you feed off the excitement of that event. It felt really different than the normal work that we do.
Were there any differences in terms of the quantity or quality of raw data that you could access compared to your usual NHL working environment?
It doesn’t differ at all. It was quite good. There wasn’t too much we really had to change. It was just about not only collecting it efficiently, but also accurately. As you know, data is really important to make decisions, but you have to make sure that it’s not only correct, but it’s delivered to people and teams where they can make decisions. Especially in a short tournament, they need that turnaround pretty quickly and in a way that coaches can interpret it. In this case, too, media could look at it.
So bringing that aspect to the general public with the Olympics, I felt we weren’t going to approach it like a niche CHL radio segment or anything. We tried to go pretty broad with the type of metrics and modelling that we did, to make it consumable for the public. But we also wanted to provide that value of, who made differences? Why was the game decided that way? And what could these teams do proactively to change their fate?
There are some things that really stand out, whether it’s like players overperformed compared to usual, or odd-man rushes. If you get more odd-man rushes, that can be a huge predictor of winning. And then shot quality, too.
At that level, too, it’s like, as women on Olympic ice, you’re farther out. It’s a bit of a different game than playing at the Scotiabank Arena, let’s say. So it’s harder for them to get out of the corners. It’s harder for them to get into the high-probability shots, to which they have to add being a bit closer than the men, traditionally. There’s strength, there’s sticks...there are all sorts of physiological reasons. They’re probably not going to shoot as fast as Brent Burns. Just in that sense, their game is a bit different in that they have to get a bit closer to the net to increase their chances of scoring.
Also, the goalies for the U.S. and Canada are both so good. Look at how Shannon Szabados played with men’s teams. I think this was her first season playing with women. You’re not going up against normal women’s hockey goalies, either, in the sense of average. These are elite goalies.
And your work also enables you to get a deeper appreciation of the players who don’t necessarily show up on the scoresheet, like stay-at-home defencemen, right?
Absolutely. A lot of the data points that don’t show up on box scores are highly relevant in analysis or modelling. So for example, even when you’re talking about a first pass, pass completion, was there pressure? What areas of the ice is she passing from? Does the possession that she starts end in a goal? Those are all really important questions that won’t show up on a box score necessarily. But you know that defenceman is doing a good job for you from an offensive side if they start a lot of possessions that end up with good scoring chances.
There’s a lot of ways that higher-level modelling or analytical work can help illuminate players that you know are great and you want in your lineup, but don’t have the “Wow!” factor for the average fan.
In the big picture, what do you think could be done to cover women’s hockey better?
I think my biggest pet peeve about female hockey or women’s sports in general is that we like to pick out a few females that get all the attention. And hey, Marie-Philip Poulin is awesome. Being Canadian, you can’t help but admire her battles for her size, determination and Brian Burke-style grit. That does show up in those metrics. It’s really a joy to watch.
But then I also think that when it comes to the women that are in the development programs or coming up in the NCAA ranks, there needs to be more attention on them as well. So it’s one thing to identify that. But I think there are so many women and groups of women that should be profiled and looked at.
It’s not like you have any shortage of NHL work. How motivated are you to keep on applying your models to women’s hockey and seeking out new opportunities on that side right now?
It’s one of those things where, to me, it’s not a business model based on profit. It’s a business model based on equity. I try to do what I can with the resources I have. You feel obligated to help raise the profile in all ways, not just in spurts. You know, I recall this WNBA quote where they talked about data and how it’s a legacy issue as well. So I really took that to heart. Some of these women will have retired, not really knowing the data they had or what types of games they played. You love the game and you see an opportunity to push it ahead.