Hockey analytics have arrived in the mainstream – and they aren’t leaving any time soon
Posted by Dean Holden at May 12th, 2013
by Jonathan Willis, 8 May 2013
Critical mass refers to a minimum: the minimum number of customers to maintain a business, the minimum amount of fissile material for a nuclear reaction, the minimum number of adopters for a technology to reach mainstream acceptance. Once critical mass is reached, the business or nuclear reaction or technology is self-sustaining; the core is big enough to ensure further growth.
It feels like the moment of critical mass for analytics in hockey has been reached.
It wasn’t all that long ago that public discussion of hockey analytics was largely limited to quiet corners of the internet where a few adherents discussed theory and snapped up whatever small amount of information leaked from NHL organizations. Certainly, at least some major league teams were well ahead of the public discussion – things like shot attempts (often commonly referred to as Corsi numbers) and shot location were being tracked by NHL teams at least as early as 1972 – but until recently they have been all but ignored by the mainstream discussion.
That’s changing, and the signs of it are everywhere. Mainstream reporters with a national platform – people like CBC’s Elliotte Friedman, The Globe and Mail’s James Mirtle and many others reference statistics tracked at places like behindthenet.ca frequently – and that site’s owner, Gabriel Desjardins, does analytical consultation for NHL teams. Too, events like the Sloan Sports Analytics Conference have received more frequent coverage in hockey media.
In Edmonton there’s no doubt that analytics have received more attention of late, particularly from the team. Much was made of new general manager Craig MacTavish’s use of the term “hocus pocus” in reference to statistics at his year-end press conference, but the full context of the exchange paints a different picture than those two words do. After MacTavish referenced a statistic about faceoffs (“every 40 faceoffs that you lose in the defensive zone, one will end up in your net in the next 30 seconds”) Oilers Now host Bob Stauffer asked him about analytics.
Stauffer: “You’ve referenced the analytics people three times. Is this something that maybe you see differently now today than five years ago? Was it more hocus pocus five years ago or…”
MacTavish: “It’s still a little hocus pocus. It’s amazing, the inferences they can draw from scratching the data off. Our guy, as I said the other night, Dan Haight he – self-professed – knows little about the game of hockey but he can tell you a whole lot about every player in the league. I think it’s just part of the tool package and you have to be participating, but it’s just one of the tools available to try and make a decision on a player. ”
Stauffer was a logical candidate to ask the question, not only because of his close ties to the team but because he’s one of the mainstream media personalities covering the Oilers with a real familiarity with statistical analysis, often referencing that work on his show. MacTavish’s answer reflected what seems to be emerging as the consensus in the mainstream – that while the statistical work being done both in public and behind closed doors is far from perfect, it is a valuable tool and that NHL teams “have to be participating” in the process.
With NHL teams more frequently willing to publicly admit the value of such analysis, the fact that it’s getting a higher profile in the mainstream discussion isn’t a surprise. Can a reporter ask an intelligent follow-up question when MacTavish references shot differential if he doesn’t understand what the Oilers’ trend on the season has been? MacTavish’s comment on faceoffs – long a sore point in Edmonton – basically indicated that concern over them was overblown; if he’s right, why have the Oilers’ struggles in the dot been so frequently cited, and if he’s wrong, shouldn’t there be a follow-up question on that point? Increasingly, those covering NHL teams need to be able not just to follow up that sort of comment, but ask those sorts of questions – and to their credit, many are.
This is the direction the conversation is going, and it’s a good thing. Statistics aren’t meant to replace watching the game, but rather to enhance understanding of it. As former NHL head (and current assistant) coach Craig Ramsay put it in an interview with the Journal’s David Staples, “your perception of what you thought you saw [watching live] is often quite off.” There, Ramsay was talking about the importance of video review to understand what actually happens in a game; statistics serve much the same function but in a different way. And as they’re incorporated further into mainstream hockey coverage, that coverage will more accurately reflect what’s actually happening on the ice.