Beauty and Truth: Advanced Stats & the Average Fan
A thought-piece about the things we believe and the things we learn
There is a war going on in the hockey world. A war waged in the back pages of local papers, on internet chatrooms, and in the corner booths of your local sports bar. It’s being fought between the hockey establishment (a group of fairly uniform-looking greying white men that want hockey to be "the way it always has been") and the "advanced stats guys" (a group of rebellious young upstarts using the soapbox of social media to spread the idea that new statistics can be used to better evaluate and predict hockey games). The field – and prize – in this conflict is us, the great unwashed, whose hearts, minds, and ultimately, pageviews will determine the winner of this ghastly war of attrition.
Sometimes I don’t really know which side I'm cheering for.
The easy answer, for me – the one that aligns best with my pseudo-intellectual self-image – is the advanced stats camp. I am, amongst other things, a supporter of (but not really a contributor to) math and science, and in this fight I see the same anti-intellectual, anti-progress, luddite nonsense that has opposed every other advance in science since the discovery of fire. When you stand against statistics, you stand shoulder-to-shoulder with the idiots who opposed vaccination, industrial machinery, medicine, personal hygiene, agriculture, and the wheel. I mean seriously: you’re fighting for the superiority of bronze tools. Every time someone clicks on the link to a Steve Simmons article, some prehistoric forebear, sitting on his haunches outside the circle of the campfire light and stubbornly gnawing at his uncooked meat, nods his head in approval.
More importantly, I also know the advanced stats people are right about a lot of things. There can be no doubt that if possession correlates strongly with success, and Fenwick correlates strongly with possession, a team that demonstrates a high Fenwick over a sufficient period of time is going to be successful.
But while Corsi, Fenwick, PDO, and their confusing legion of variants are indisputably useful, their emergence has also proven to be a bit of a downer. The problem is that whereas vaccinations killed polio and smallpox (or should have, anyways), Corsi kills narratives and romance.
Anyone who has ever found out that one of their favourite players is actually an anchor dragging down a team’s possession numbers knows what I’m talking about. When you’ve spent years embracing some of hockey’s favourite archetypes (like the big immovable defenceman, the gritty third line forward, or the affable brawler) it’s hard to come to terms with the fact that most of them make your team worse.
I’ll be honest with you, I liked thinking that every team needed an energy line to "bang some bodies", "make the other team afraid to go into the corners", etc. etc. I used to love when Chris Neil got sent out on a powerplay to show the skill guys how to take some punishment and set a screen. I enjoyed the blissful ignorance of thinking that every team needed a guy like Douglas Murray to clear the front of the net. It was unpleasant to have the veil of these illusions ripped away, and to find out that so many of hockey’s old adages were myths. It was like growing up and finding out that most of your family was just gritting their teeth and barely tolerating each other during your idyllic childhood Christmases.
The predictive uses of advanced stats can really burst your bubble too. It used to be fun to cheer for a team of plucky underdogs that got great goaltending and "found a way to win". These days by December you can probably predict 14 of the 16 playoff teams based on Fenwick Close 5v5, PDO, and/or a few other measurables. You know that teams like the Leafs are likely to fall off, and teams like the Rangers will likely climb. Sure, a few teams like this year’s Avalanche will slip through the cracks, but even then you know that it’s extremely unlikely that they can actually win more than a round or two in the playoffs. And even if by some miracle a team with low possession numbers and a high PDO somehow gets lucky enough to win it all, the sweetness of that victory will always be tempered – at least a tiny bit – by the knowledge that they never really deserved it; they’d be the hockey equivalent of Greece’s Euro 2004-winning side.
So that’s where advanced stats have brought us: we know which teams deserve the Stanley Cup, and we can be pretty certain those teams will end up competing for it. Despite the illusion of parity in the regular season standings, the rest of the NHL may as well be the 26 NBA teams that enter each season with no hope of winning a title. Thanks, math!
And none of that is even the worst part. The most terrible thing about advanced stats is that I can’t even figure these things out for myself. After a lifetime of playing and watching hockey, I never could have told you that Erik Condra is a secret weapon that makes everyone around him better, that the 2013-2014 Leafs were a far worse team than their mid-season record showed, or being perfectly honest with myself, that little Erik Karlsson is one of the league’s best defencemen (good, of course, but the superlatives really show themselves in the numbers). I’ve been trained to watch for all the wrong things. For all my experience, it turns out I need to wait until after the game for someone with a spreadsheet to tell me what actually happened.
But despite the romance-killing nature of advanced stats and the often-infuriating smugness of the people who relay them, I’m ultimately glad that advanced stats have finally emerged in hockey, and here’s why: because my TV is lying to me, and my team is lying to itself, and I’m sick of both these things.
Anyone who has paid any heed to advanced stats has experienced the former. You’ve seen Paul MacLean send out the Greening/Smith/Neil line and heard the commentators describe them as the "shut-down" line, but you know full well that the Smith’s line’s underlying numbers belie that title. Or you’ve heard the bonhommes at the CBC spend an intermission lavishing effusive praise on Douglas Murray, and you’ve sat and wondered how they can possibly believe the things they’re saying. For all that I miss living in that warm bubble of familiar lies, I’m glad it’s gone. Advanced stats are going to end the absurd hegemony of lazy commentators and other members of the hockey establishment, and those gentlemen know it as well as we do (why else would they be fighting it so hard?). I, for one, won’t miss them.
As for my team, there’s nothing I want more than for them to pay attention to the statistics. At some point, every NHL organisation is going to be using them; in a salary cap system, the market advantage is too valuable to ignore. Advanced stats can be useful in almost every facet of hockey operations, from scouting (both at the minor and NHL levels) to game management (teaching the value of zone entries/exits, the importance of not playing a goalie on consecutive nights, or the ideal distribution of ice times, etc.). These stats are obviously not the only thing that a team should be looking at – that would be absurd – but an organisation will ignore them at their own peril. You can be certain that the advanced stats camp will be keeping a close eye on these issues, and I’m glad to have a resource that will help me evaluate the decisions my team is making.
In reality, it really doesn’t matter whether I want support the old ways or ride the zeitgeist into the modern era of sports. The advanced stats revolution is happening, and the only thing hockey’s old guard can do is slow it down: TV broadcasts are already starting to display Corsi numbers; NHL owners and GMs are starting to understand that the market advantage will be biggest for the first teams to seize it; Matt Kassian is finally off the Senators roster. Times are changing.
It’s a brave new world, and maybe a harsher one, but at least it’s real. And hey, that "fire" thing turned out alright, didn’t it?