There will come a time in these Stanley Cup Playoffs—and heck, it’s likely happened several times already—where the vagaries of puck luck and hot goaltenders and bad penalties and bounces will inspire you (if you’re a certain type of fan) or someone you know (if you’re not) to point gloatingly at a badly-outshot-yet-victorious team and declare, “So, how’s that analytics crap working out for you now?!”
Don’t do this.
For one thing, it’s snide and rude, and lord knows the analytics-in-hockey debate already has more than enough of that on both sides. But most of all, it’s a stupid thing to say—though not for the reasons you might think.
Mostly it’s because the analytics-supporting friend or colleague or lover you are being snide and rude to likely agrees with you. When it comes to playoff hockey, anything can and does happen, and everyone knows this. So why can’t we stop arguing about the role possession metrics play in determining who wins and loses? When it comes to playoff hockey, this debate only has one side: Luck.
STANLEY CUP PLAYOFFS: | Broadcast Schedule
Stanley Cup Playoffs Fantasy Hockey
New Sportsnet app: iTunes | Google Play
If you don’t have at least a good chunk of it, and it can come in many forms, you’re done. Put it this way: In the regular season, having a goalie with an unsustainable save percentage and a third-line forward with a shooting percentage several notches above his career norms is a sign that you’re due for a correction and a drop in the standings before year’s end. In the playoffs, those two things are basically prerequisites for a deep run.
Hell, even Billy Beane–you know, the A’s GM? The guy who started using #fancystats in baseball, which eventually led to #fancystats in hockey and eventually got you so riled up because someone dismissed your favourite grinder’s grit and toughness? Yeah, Billy Beane, that guy agrees with you about the playoffs.
If there was one quote from Moneyball, Michael Lewis’s book about Beane’s use of Sabermetrics to turn his Oakland A’s into an on-base-percentage juggernaut (a simple-yet-undervalued baseball statistic that drives everything else but wasn’t widely cited as critical … kind of like Corsi, for instance, in hockey), that applies across all sports and sums up what advanced statistics and their applications can and cannot do, it is this:
“My s–t doesn’t work in the playoffs. My job is to get us to the playoffs. What happens after that is f—ing luck.”
So please, I urge you, as a hockey fan, to take this quote to heart over the next two months whenever you feel like making a declarative statement one way or another about analytics.
The phrase “small sample size” has become something of a buzzword, used to explain away hot starts and bad stretches and careers in their infancy. But it’s a buzzword because it’s real and it exists, and it applies to every single seven-game playoff series ever, in any sport.
If you hate what the rise of analytics has done to hockey discussion–and I do sometimes, as much as I love possession metrics, if only because it often turns what should be an enjoyable sports conversation into a battleground that requires everyone to pick a side–then you should know that there is nothing to argue about in the playoffs. We’re all on the same side here. Chaos reigns and all of our favourite teams, possession monsters or weaklings, are subject to it.
So listen, have the argument if you must, but keep it to items that are actually ripe for debate. You hate those Corsi freaks and you want to tweak them a little bit about the Kings missing the playoffs over an 82-game season? Well, go right ahead. There’s a few reasons you’re wrong, and I’m sure they’ll happily clue you in, but at least it’s a fair basis for an argument about the predictive power of possession metrics.
But if you want to point to a first-round series and do the same thing? Don’t bother–the person you’re yelling at agrees with you. Analytics is largely useless in any one series. At best, a great possession team facing a weak one has, what, a 62-38 shot attempt advantage? Enough to swing an even strength goal or two over the course of a few games? No doubt, and that adds up. But one bad game from your goalie, one good game from theirs, one missed call and one bad bounce? That’ll swing a lot further in a series that–and this is important to remember–is a seven-game series in name only. Most series do not go the full seven games, of course, which means every bad bounce is that much more critical.
I’m not here to tell you not to argue about hockey during these playoffs. I’m here to tell you to argue about the right things. Teams with good possession numbers have a slightly better chance of winning a seven-game series. Just like Beane’s A’s, a team that makes the playoffs almost every year, but has to scratch and claw and catch a whole lot of luck to go any further–just like every single other team in the playoffs.
If you think analytics don’t work in hockey, you’re wrong. But if you think they’re much less likely to work in the playoffs, you’re spot on–but so are the people you’re yelling at.