The end of the regular season has long since passed, and for Leafs fans that means there's not an awful lot going on that's worth writing about. To fill in some of the time between now and draft/free agent speculation really ramping up, 67sound and I are going to try something a little bit different than what you'd normally see on a hockey blog. Over the course of the next several weeks we're going to exchange a series of letters (or their equally exciting modern equivalent: emails!) discussing our views about advanced statistics, the Toronto Maple Leafs, and whatever else seems to fit into the flow of the conversation. Once a week (if possible), I'll post the results here for everyone to read and, if you're into that kind of thing, discuss. These posts aren't likely to be filled with fancy charts (sorry nhlcheapshot) or complicated math, just what I hope will be an interesting discussion about where we're at as hockey fans and where we think the Leafs are at as a team. And so, without further ado, here's the first installment:
I proposed that we start our discussion of advanced hockey statistics by explaining how we first came to learn about them, but the truth is that I don't entirely know. I remember when I first signed up to comment at Pension Plan Puppets, which was when the Leafs were doing poorly early on in the 2010-11 season and Brian Burke gave an interview on TSN about how the team's season ticket holders were happy with the direction of the team. And I also know that I first came across advanced stats from people in the PPP community; you were probably one of them. So I can trace the beginning of my desire to understand things like "Corsi" and "Qualcomp" to some time back around the beginning of 2011. That's not really all that long ago, which is strange to think about because it feels like I've been going on about those kinds of things for a while now.
I have to admit that at first I was pretty skeptical. I knew that +/- wasn't a very reliable indicator of ability for skaters, but beyond that I think I accepted a lot of fairly traditional beliefs about hockey stats, like that shooting percentage indicated players who shot the puck better (the mythical "shot quality") or that goals against average was a good way to measure goalie talent.
I like to think that I'm open to changing my mind as long as I'm presented with a compelling set of evidence that contradicts what I believe. Accepting that Corsi or Fenwick had some value was pretty easy. It's simple to understand why Corsi might be valuable – even people who have no knowledge of advanced statistics know how hard a good cycle down low in the offensive zone is to deal with, or how important it is to have "puck-moving" defencemen who can quickly and efficiently get the puck out of the defensive zone. Ultimately that's what Corsi is attempting to measure, and it's easy to see why that would be useful. As I read more articles by smart people like Gabriel Desjardins or Vic Ferrari I saw lots of evidence that Corsi was actually useful as a measurement of, if not individual ability, certainly team success.
What took me longer to come around on is the issue of shot quality. It seems pretty obvious that shot quality (that is, that some players are better at shooting the puck than others) is a real thing. We've all seen Ovechkin or Kovalchuk blast laser beams into the top corner often enough that it certainly seems like they must have a talent for it that other players don't. And to a degree, I do actually think that shot quality exists, but maybe not quite in the way that it's often conceptualised. I think that being able to evade defenders and get off shots is a hugely important part of what makes good goal scorers good at scoring goals (with exceptions for guys like Holmstrom who mostly score garbage goals). So maybe Mike Brown really can shoot the puck as well as Phil Kessel under the right circumstances, but the difference is that Kessel, whether through skating, positioning, hockey sense, or whatever else is an awful lot better at creating "the right circumstances". That's where I currently sit on that issue, but it's also the case that the more data I see on shot quality, the less important I think it is. I definitely do think that shot quantity is generally a more important skill, and that being able to generate a high volume of shots is a real skill that we ought to value.
It's kind of funny that I've gotten into this line of inquiry (fancystats, that is). Even though I'm employed as a computer programmer now, I studied history rather than computer science at university because I wasn't really interested in doing academic math. I do kind of wish sometimes that I had a better grasp of statistics and other kinds of math that I never bothered learning, because I think I'd have a better understanding of a lot of the more complicated hockey stats articles that go over my head a bit. And I'd definitely like to be able to do more of the hard work myself, rather than relying on other people to generate numbers and then just doing interpretation on them.
My interest in advanced hockey stats actually does line up with my educational background somewhat, though. I've always relied heavily on statistics in the history writing that I've done (my thesis has an entire chapter that's predominantly numbers) and I'm definitely interested in the idea of trying to piece together a puzzle in a way that provides an objective answer, or even a hint as to what an objective answer might look like and where we might want to look if we were to continue trying to find one. So from that standpoint, learning and writing about hockey stats fits my love of logical puzzle solving.
How about you? How did you first come across advanced hockey stats, and did you accept their usefulness right off the bat or struggle to come to grips with the idea that some of the ways you'd thought about hockey growing up were misinformed? What drew you to them in the first place, and what makes you continue to be interested in them now? Is it just that they're more reliable than other statistics or do you find that, like me, they satisfy some more fundamental aspect of your personality?
Looking forward to your response,
* * *
I actually got into hockey analytics because of Joe Morgan. Or to be more specific, Fire Joe Morgan, the epic blog with the well-earned slogan "Where Bad Sports Journalism Goes to Die". Ken Tremendous (aka Michael Schur) and the boys combined a working knowledge of sabermetrics with biting sarcasm to eviscerate tired cliche after tired cliche. I had already read Moneyball and had a passing familiarity with baseball analytics, but Fire Joe Morgan made it fun.
Some time after the blog closed its doors in 2008 I started to poke around the Internet for a hockey equivalent. I think the first one I came across while looking for a similar hockey "fisking" site was Cox Bloc (@GoddTill and Kim Jorn), which led to places like Down Goes Brown, Pension Plan Puppets (@mlse and @felixpotvin) and Bitter Leaf Fan (@mforbes37). I also learned from a rec hockey teammate had one of our other teammates had this weird hockey stats blog--none other than Tyler Dellow, aka mc79hockey.
I'll always be grateful to PPP for creating a community where people like me discussed, debated and educated each other about hockey. I don't think any of us at PPP in those days were stats experts but there was a solid group of us who were open-minded, analytical and curious. Moreover, people would link to guys like Gabe Desjardins and his incredible Behind the Net database. I was also still reading mc79hockey, and he would link to a lot of other statistically-minded Oilers writers like Vic Ferrari.
The next step for me was Twitter. I had seen people at PPP linking to Twitter feeds around trade deadline or July 1, so I signed up to get the latest. I then learned that as good as it is for breaking news, it's even better for sharing opinions and analysis. I found out who the best hockey analytics people were around the league and started reading their stuff: @kent_wilson, @jonathanwillis, @camcharron, @BSH_EricT, @geoffdetweiler, @andrewberkshire, and many others I'm shamefully forgetting to mention here.
As a result of the education I've received, pretty much everything I used to believe about hockey has been overturned. Most fundamentally, five years ago I would have sworn that shot counts didn't tell you much because some teams just float around the perimeter and shoot 50 footers, while others go to the dirty areas and consistently get higher quality chances. While that may be true for individual periods or even games, it just doesn't hold up over a season. The most important thing in hockey, I now believe, is keeping the puck in the offensive zone. If you can do that consistently you'll get shots, some of them will be good shots, and some of them will go in.
What really clinched it for me is that every year around November, "surprise" teams will emerge. Despite getting outshot consistently, they'll defy the odds and the talking heads will credit it to "shot quality", "grit", "determination" or some other intangible characteristic. The hockey analytics crowd will predict their demise--and inevitably be proven correct. This year it was Minnesota. It's not even difficult to do. Everyone who pays attention to possession metrics could see the Wild's downfall a mile away.
So that's my back story. Let me ask you: what about our beloved Leafs? What do you think is their biggest need right now? Is it goaltending, forward or defence?
I wish I was funnier,