The NHL is at the beginning of an analytics awakening.
Advanced statistics, such as the tracking of total shots attempted for and against, have been around for decades, giving teams a glimpse of information behind the final score and players’ goals and assists. As stats and the game have evolved, more hockey people are seeing their worth to wins and losses.
Over the summer, several teams, including the Toronto Maple Leafs and Edmonton Oilers, hired analytics specialists to get the most out of the data available. Recent Stanley Cup champions like the Los Angeles Kings and Chicago Blackhawks have had success with advanced stats, so it stands to reason that others will follow.
"You have to eliminate some of the noise and present the data or the information that’s going to best help the team in whatever regard that is: It’s scouting, it’s strategy, lineups," said new Leafs assistant GM Kyle Dubas. "It’s taking what I know works and then incorporating that into the whole team structure and knowing that you’re dealing with a number of subjective pieces to the puzzle, as well."
The most basic of the advanced stats are called Corsi, Fenwick, zone starts and quality of competition.
Corsi and Fenwick imply puck possession because taking more shots than the other team generally means you have the puck more. Corsi measures all shots attempted (on goal, misses and blocks) for and against, while Fenwick does the same but takes out ones that are blocked.
"It’s just probability and it’s just getting the odds on your side," Oilers winger Taylor Hall said. "Teams that have the puck a lot are going to win games and that’s really what that advanced stat boils down to."
Zone starts show where on the ice players began their shifts (ie, more often in the offensive or defensive zone.) And quality of competition essentially shows the difficulty level players face based on who they match up against.
Many teams have tracked that kind of information for years. Now it’s faster to obtain it.
"It’s been going on, but it’s been a tedious work that’s being done on a long-term basis, and now with the process that analytics is included in, the information that took days, weeks, months, now is done on a game-by-game, period-by-period, shift-by-shift basis," St. Louis Blues coach Ken Hitchcock said. "Because of the technology, the information that you slaved over for hours now can be brought forward instantly."
Hockey analytics are still developing 20 years after Bill James ushered in the sabermetrics revolution in baseball. Sabermetrics deliver information like how many wins an individual player is worth compared to a replacement and tries to explain more about the process that leads to the results how batters get on base and how runs are scored.
Old-school and new-school thinkers can agree on one thing: Hockey is not baseball, and therefore it’s more difficult to quantify.
"The game is more impulsive, more spontaneous, more dynamic, more inventive, more creative, more imaginative, faster," Hockey Canada president and CEO Tom Renney said. "At the end of the day, there’s 10 players on the ice and a couple of goaltenders that are playing this high-paced game intuitively."
Renney, who worked under Mike Babcock on the Detroit Red Wings’ staff last season, believes from his time as an NHL head coach that analytics can "help set the compass" for where teams need to go.
For executives, general managers and scouts, analytics are a way of giving evidence beyond the so-called "eye test." Ottawa 67’s coach Jeff Brown, an ex-teammate of Brendan Shanahan’s with the Blues and Hartford Whalers, said the new Leafs president is "extremely into the analytics side of things."
"He opened my eyes to it," Brown said. "He’s a big believer in it. I have all the respect in the world for Brendan Shanahan. If he’s buying into it, I certainly am going to have to, I guess, get up to speed on it."
In one of his first acts on the job, Shanahan sought out and hired Dubas, a 28-year-old prodigy who used analytics along with a lifetime in hockey to help turn around the Ontario Hockey League’s Sault Ste. Marie Greyhounds. Then the Leafs brought in "hockey research and development" analysts to get up to speed with other teams around the league on advanced stats.
The Oilers hired blogger Tyler Dellow, a Toronto lawyer, to join their front office, and the Philadelphia Flyers added Ian Anderson, who had been a team services manager for the Washington Capitals, as manager of hockey analytics. None of the hires were publicized, as there remains a cloak of secrecy over using advanced statistics to gain an edge.
Behind the scenes, teams do more than track Corsi and Fenwick to make judgments. Winnipeg Jets winger Blake Wheeler correctly called those stats "primitive."
"How do you really judge a player’s value based on his overall shots for or against or whether his shots are getting blocked or whatever? I think there’s a little bit of a ways to go in those areas," Wheeler said. "You can watch and you can say this guy, he’s head and shoulders above the rest of the guys on the ice or this guy’s not. I think it’s more of a visual game."
Hockey is getting there.
Following the lead of the NBA, the NHL is exploring the possibility of micro-chip technology that could provide real-time information on the speed of shots, time in the attacking or defensive zone and how long the puck is actually in a player or a team’s possession. That’s the next step.
"There’s going to be a lot of changes, I think, in terms of technology and how they’re rating us," Florida Panthers centre Nick Bjugstad said. "I think it’s good, it’s going to keep you honest. It’s going to make sure everyone’s doing their job."
For coaches, the cutting edge right now has to do with figuring out how to put forward lines and defensive pairings together and how to create the best matchups.
"Where it’s helped a lot of us is surprisingly on chemistry issues," Hitchcock said. "The data that has been brought forward is very helpful on things that you’re thinking of. So for instance, you might think you’ve got a good line going and then you make a change and you think the line’s going to be better and all of a sudden you find out the data doesn’t back it up at all."
The biggest challenge, Hitchcock said, is delivering the information to players in a way that doesn’t burden them with too much information.
"We all use analytics to help us, but you’ve got to be careful that the analytical information gets in the way of the feel and it gets in the way of being information overload," he said. "I find that when you coach hockey players and you get them into information overload, now you get a lot of confusion and hesitation in their games."
In a sport that values instinct and quick thinking, players aren’t worried about their Corsi stats, and they shouldn’t be. But understanding what goes into advanced statistics is something that can help, as New York Islanders captain John Tavares has learned from working with independent performance coach Darryl Belfry.
"Sometimes that can kind of fuel when things are going well, what you’re doing well, and maybe when things aren’t going as well what you can do to kind of make some changes and obviously improve," Tavares said.
Because hockey is such a fast, fluid sport, there’s a natural resistance from players that it can be measured so much statistically.
"It just isn’t relevant all the time," Vancouver Canucks defenceman Kevin Bieksa said. "Puck possession and all that stuff, it’s not relevant. It doesn’t matter how much you have the puck, it’s what you do with it. And there’s so many things happening away from the puck, too, that it’s tough for me."
Many star players don’t understand analytics, but it doesn’t seem to hurt their games. Like in auto racing, the best drivers don’t need to know how every gear turns to be successful.
"That’s just the nature of the game," Red Wings defenceman Niklas Kronwall said. "But at the same time I think it’s amazing that guys have come up with a system that actually can measure things, and I think it’s great and I think it’ll be something that will be around for many years to come."
Analytical thinking in hockey isn’t going anywhere, though it could look much different in five or 10 years. The stats that fans are just getting accustomed to now are already old hat for those working in the game, and new puck- and player-tracking technology should open the door to more useful information down the road.
As those stats evolve, so are the people who are using them.
"I think you’re starting to get to the point with advanced stats where they’re incorporating video and they’re explaining why a stat is a certain way and what’s producing that outcome," Hall said. "I think it’s a useful tool for coaches and GMs, and I think it’s starting to become a useful tool for players, as well."
Analytics are already a way of life for GMs and coaches in a sport where the difference between winning and losing is razor-thin.
"It’s at our disposal, why wouldn’t we use it? It’s not going to hurt, right?" Brown said. "It’s just another tool to make you a championship team. I don’t think anybody’s saying that solely analytics is going to win you a Stanley Cup. But certainly if it’s there and if it makes sense, why wouldn’t you use it to your advantage?"