In baseball, advanced statistical analysis first brought to the masses by Bill James more than 30 years ago are now standard procedure for most MLB teams and broadcasts. In the NFL, teams such as the Atlanta Falcons and Chicago Bears use #fancystats to make personnel moves and evaluate draft choices. SportVU cameras, adopted from missile defense technology, track the movements of every NBA player on the court, plus the ball in every arena.
The statistical revolution is happening everywhere but in the NHL.
“It seems like hockey as a sport is a bit more old school and not as stats-focused as basketball,” explains Brian Kopp, senior vice president of Sports Solutions at STATS LLC. “Basketball was ready for more data.”
It is true that the NHL doesn’t have the best track record when it comes to embracing statistical analysis. Games played did not become officially tracked until the 1960s. It wasn’t until 1967 that the league started keeping track of goals from special teams, game-winning goals and shot total data, though shots against at the goaltender level wouldn’t be officially available until 1982. In 1997 the NHL would roll out their computer-based Real-Time Scoring System (RTSS), and while it was a step in the right direction, it has its fair share of issues. For example, there is evidence that not all NHL arenas record shots equally and some operators are unsure of where the faceoff dot is.
But that could all change soon.
“We have done some initial data capture for hockey, especially in arenas where both NBA and NHL teams play,” Kopp said. “We did some testing about a year ago and the data capture, camera setup and player tracking is pretty straightforward. Puck tracking is going to be a bit of a challenge because of the speed. The other challenge is the line shifts and identifying players. The way we do it in basketball is read the color and number of their jersey so we have the cameras doing the identification. We can leverage some of that and put cameras behind the bench to read them easier. The biggest challenge is the line shifts, which we don’t have in basketball, and [tracking] the puck.”
Until then, the statistical void will continue being filled by sites like Behind the Net, Extra Skater and Hockey Analysis, as well as bloggers like Eric Tulsky, whose paper on zone entries was presented at last year’s Sloan Sports Analytics conference. But even at an event dubbed “Dorkapalooza,” skepticism among the NHL ranks in attendance remains high.
“Statistics are like a lamp post to a drunk: Useful for support but not for illumination,” former Maple Leafs general manager Brian Burke said during last year’s Sloan conference. But that type of resistance wasn’t any different than Kopp experienced when he was first tasked with rolling out SportsVu to the NBA.
“For some reason, people like to be put in two camps: You are either old-school coaching or numbers-geek analytics and it should be a combination of both,” Kopp said. “One is not going to replace the other. Numbers are never going to replace the coach and their experience.”
There are a few NHL teams who have dipped their toe in the water when it comes to the next level of data collection. The Pittsburgh Penguins have used shot location analysis and goal-scoring probabilities from the Sports Analytics Institute, while Tampa Bay and Chicago, among others, have hockey analytics people on staff. Kopp is optimistic that more of the NHL will emerge from the dark ages sooner rather than later.
“We are planning some more development this season, and I am hoping next season,” said Kopp. “Whether that is with a handful of teams or through the league, I don’t know yet but I hope that by next year we are doing more significant work to figure out how to roll this out more broadly.”