When it comes to advanced analysis in major-junior hockey, the Greyhounds seem to be leading the pack, Ottawa is anything but sixes and sevens on the subject and the Blazers, like most Canadian Hockey League franchises, aren’t yet burning up with fancy-stat fever.
The growing trend of using advanced statistics — such as Corsi and Fenwick numbers, which aim to measure offensive-zone time and possession — to break down performance and project the future has ignited a fiery debate across the National Hockey League landscape, much like the one that flared up on the Major League Baseball scene in 2003, when Moneyball was released.
While the argument — which has become particularly heated online and on Twitter — has not yet reached its boiling point in the junior-hockey world, advanced statistics are already being used and there are differing opinions on the matter.
Sault Ste. Marie Greyhounds general manager Kyle Dubas is convinced his team has gained a competitive advantage by using fancy stats — in fact, the 27-year-old GM (he turns 28 on Friday, Nov. 29) lobbied team owners to hire a full-time advanced statistician.
Tyson Enfield was named the Greyhounds’ manager of ticket sales and advanced analytics prior to the 2012-2013 Ontario Hockey League campaign.
“There’s Corsi and Fenwick, but there’s much more beneath the surface than those that can help out and I don’t want to totally expose all those because I think it’s advantageous for us,” said Dubas, who took the Greyhounds’ reins at 25, becoming the second-youngest GM in league history.
“We’re just trying to learn, trying to give ourselves more information about how the game actually works and how the outcomes of the games are influenced by various events — individually and in a cumulative manner.”
Studying analytics is a tall task in the CHL, where there aren’t nearly as many stats kept as there are in the NHL, but the Greyhounds have made it work.
“We’ve had to build different software and mechanisms to start tracking stats and we’re trying to build our own models and not just take what’s accepted as being predictors and indicators of team success,” said Dubas, whose squad sits second in the OHL’s Western Conference after 24 games.
The Greyhounds use an intern on game nights to help Enfield crunch the numbers.
Dubas noted the Ontario league doesn’t keep track of who takes shots, nor does it record individual ice time — both stats are essential to feeding the club’s metrics.
James Mirtle of the Globe and Mail began reading about advanced statistics about a decade ago and the Kamloops product is now using them in his journalism work.
“I’ve always been an avid reader of everything that’s out there written about hockey online and in all the blogs and I was part of that community originally, and that’s really where the advanced stats first started,” said Mirtle, who covers the Toronto Maple Leafs.
“They were hobbyists and people running their own websites that started this kind of thing and it wasn’t until 2007-2008, in that NHL season, that Gabriel Desjardins started putting some of the numbers on his site, behindthenet.ca, and that really started things in earnest.
“That’s when things like the possession stats and some of these numbers really entered the lexicon for hockey fans that were really paying attention to this kind of thing.”
That fancy stats came to fruition online and on blogs is a problem for many in the hockey community — and it’s ammunition for the anti-analytics camp.
“That’s part of the reason why some people discount it,” Dubas said.
“You constantly hear it when people are slandering analytics — ‘It’s people in their mother’s basement’ — and that’s an interesting way to look at it. I think that’s a silly way to operate.
“To be honest, if you go through those blogs every day, you can learn something fascinating and it’s almost like getting a degree in that sort of stuff for free, which is incredible. But, I get why others are skeptical. That’s their choice.”
Not every self-published hockey writer is producing worthwhile fancy-stat fodder.
“Approaching any of this stuff with skepticism is probably the right way to go until you actually see it working because there’s a lot of numbers and a lot of statistics that don’t work properly,” Mirtle said.
“Some of it’s quality and a lot of it is not. A lot of it is garbage. Over the years, the stuff that is quality has gained an audience.”
For Dubas and the Greyhounds, it’s not about moulding the club’s philosophy around these new numbers — it’s about capitalizing on any edge they might provide.
The Ottawa 67’s are the latest major-junior squad to invest in analytics, hiring Trent University economics student Matt Pfeffer prior to the 2013-2014 campaign.
He is the club’s resident sabermetrician — a term coined originally for someone who specializes in the computerized measurement of baseball stats (saber: Society for American Baseball Research).
“It gives the 67’s a slight edge, a little more information that teams don’t have,” Pfeffer told Neate Sager, who writes for Yahoo Canada Sports. “That’s just what I try to do, have information that other teams don’t.”
Ottawa is eighth in the OHL’s Western Conference.
For every Dubas, there is likely a whole fleet of GMs in the CHL who are not buying into advanced statistics. It’s the same in the pro ranks.
Said Mirtle: “I was at a sports conference and Dave Nonis, the Leafs’ GM, spoke there and basically came out and said, ‘We don’t see any value in analytics. We have a large budget that we’re able to dedicate toward analytics, but we often don’t even use it at all.’
“I don’t think that they’ve looked into this close enough.” Mirtle added.
“I’ve been writing about this in the paper and I can tell the Leafs have read some of the stuff and it just seems to me they’re dismissing it out of hand, in part because I think they don’t like some of the conclusions that it has about their hockey team.”
Kamloops GM Craig Bonner does not care to comment on advanced statistics — not because he is against them, but because he is unfamiliar with Corsi, Fenwick or any other analytics system.
Kamloops head coach Dave Hunchak, meanwhile, does not see their value in the WHL.
“We’re asking scouts to identify 14-year-old kids for the draft and a kid might be 5-foot-6 and put up 160 points in bantam hockey, yet when he gets to us, he could be 6-foot-2 and still developing,” Hunchak said.
He explained the revolving-door nature of a junior-hockey roster renders advanced statistics ineffective.
“Most teams in the NHL have their core signed and their base players signed and their elite guys that they’ve drafted and developed — and now you take a look at the numbers and you try to plug holes,” Hunchak said.
“That’s where free agency comes in . . . in hope that they produce certain numbers to support the core group.
“When it comes to Moneyball-type stats, I don’t think it can apply to junior hockey.”
The Blazers do keep more than just traditional stats, making note of individual battles won and lost and scoring chances — real-time stats, as they’re referred to by Hunchak.
There are no plans in the works on Mark Recchi Way to hire an advanced-stats intern or a full-time analytics specialist.
“Paralysis by analysis is a common phrase that we use,” Hunchak said.
“[Blazer associate coach] Mark Ferner and I will watch hours and hours of video and, in reality, you already know what the issues are.
“At the end of the day, when you’re dealing with human beings, you’re still dealing with the personalities and the players and that’s the most important thing.”
Neither Mirtle nor Dubas envision the use of analytics by NHL or CHL clubs becoming commonplace any time soon.
“They’re in their infancy in hockey,” Mirtle said. “This is the very, very beginning.
“I think it’s probably going to be 10 or 15 years before we can say that every team in the NHL is using something like this or something even more advanced, but it’s happening.
“Two years ago, hardly anyone was talking about this kind of thing. This season, I get asked about it all the time.
“There’s a huge appetite from hockey fans and there’s a growing appetite in the NHL.
“As younger people get in to be GMs, as more people become hockey executives that didn’t play in the NHL and have more of a new-school kind of thinking, it will become much, much more popular.”
Dubas represents that new school.
“If you’re stubborn and you believe one thing works and one thing only and that’s the way it’s been your whole life and it’s never going to change, then you’re always that way, you could be 20 years old or 70,” he said.
“And I think there’s other people, and I’d hopefully be classified in this, we’re always trying to learn and never discount anything regardless of how foreign the concept might be, to see whether or not there might be something to it that can help us be better managers and a better hockey program and better at scouting and developing players.
“People will always doubt it.
“I look at it as every piece of information is valuable, as long as you have it in the proper context and you’re willing to invest yourself in it.”
Before you dismiss them . . .
KTW asked Mirtle to provide one advanced-statistic measurement that might turn the doubters into believers.
“Fenwick Close, which is basically attempted shots for at even strength minus the ones against, and only when the games are close, so you throw out the games when it’s 3-0 or 4-0 because, of course, if you’re up 3-0, you’re not as worried about directing pucks on net,” Mirtle said.
“The last six NHL seasons, to give you an example, teams that average 52 per cent or better over a season in Fenwick Close, so they’ve basically been in the offensive zone 52 per cent of the time, their average finish in the standings is 101 points.
“Conversely, teams that are below 48 per cent in that number are 80-point teams. That’s a huge swing.
“Amazingly, the statistics are very, very accurate and are good at predicting future outcomes for teams and teams that are good at that, at directing more shots on net than are directed on their own net, they do very, very well.”
Hear all about it
To hear from Hunchak, Mirtle and Dubas, tune into CBC Radio Kamloops, 94.1 fm, at about 7:45 a.m. on Friday, Nov. 29.