And, most importantly, people on the internet are arguing about statistics. We all were foolish to think the Toronto Maple Leafs' crash-and-burn would bring that to an end — but we were younger then. It's been a long five weeks.
The latest log on the very hot, very obnoxious fire: A story by SportsGrid's Matt Rudnitsky headlined "MONEYPUCK: Why Most People Need To Shut Up About ‘Advanced Stats’ In The NHL." Hockey Twitter, like regular Twitter, is ephemeral by nature, but a good portion of Tuesday morning was spent dissecting it.
And really, it was OK enough. There were some good points; it's always worth noting that some teams have a higher baseline shooting percentage, and counting even-strength shot attempts, particularly when the game is close, is a useful tool for anyone trying to understand the game a little better.
There were issues, too; Rudnitsky's claim that using shot attempts (Corsi) won't help you beat Vegas lines doesn't seem to hold up:
There was also an undercurrent of anger to the whole thing. That's understandable to a pretty large extent, particularly for anyone who had to put up with, essentially, the same argument in baseball a few years ago. This is "Groundhog Day," and we're somewhere in the middle of the movie, when Phil starts jumping off bell towers.
So sure, it can be aggravating — but you either want people to start using this stuff, or you don't. And personally, I'm cool with seeing Sean McIndoe and Pierre LeBrun and Dave Lozo incorporate all this into work. Rudnitsky, for whatever reason, seems … not to be, I think?
Let's leave that for another time…or never talk about it again. Either way works.
The most useful thing to take out of the piece, though — and this may be projecting, because whether Rudnitsky feels this way himself isn't totally clear — is that we are fast approaching a time where simply posting stuff you found on Extra Skater won't be enough to pass as actual insight. It shouldn't be subversive or all that interesting to pick series winners by checking out the Fenwick Close rankings, tweeting a screencap, then washing your hands of it.
Whether you're a media professional or a hobbyist or a fan, doing that may be effective. For the most part, it is — though it's just as easy (and possibly just as effective) to say "hey, my conference finalists are the last four Cup winners." Neither of those, though, are inherently interesting, insightful or unique. If you're in that rat race, you will need to do something else. Dartboard incorporation, perhaps. It's tough to argue that we're there already, considering, uh, the Maple Leafs, but sooner or later, it will happen. And that should make everyone feel pretty good.
And I'm guilty of this — hopefully less now than in the past, because the point of all this is to get better. Here's an example: I'm based in Pittsburgh, so I see the Penguins more than any other team. Over the winter, they relied way, way too much on their top two lines and power play. Every available stat — and my eyeballs, dangerous as that can be — backed that up. They didn't have a third line to speak of, let alone a fourth, and an injury-demolished group of defensemen was struggling to make up the difference.
Over the last couple months, though, that has changed. Kris Letang and Paul Martin, two All-Star-caliber puck-possession demons, are back in the lineup and, as of Tuesday, playing together. Olli Maatta and Matt Niskanen are a great second pairing. At forward, Beau Bennett and Brian Gibbons are healthy, Lee Stempniak and Marcel Goc were positive deadline additions, and Dan Bylsma is using them effectively — now Crosby and Evgeni Malkin can play together if need be, and center Brandon Sutter has better players on his wings.
That stuff is big. It was also easy enough to see coming, and it's been reflected on the ice and, as a direct result, their possession numbers. They have an infinitely better shot at beating Boston (and New York) today than they did in February. People need to change how they think about them, because they're not the middle-of-the-pack Corsi team of a few months back — those guys probably wouldn't have beaten the Columbus Blue Jackets, let alone put up numbers like this, even over a few games.
This is a graph depicting their percentage of unblocked even-strength shot attempts during tied or one-goal games in the first/second periods (Fenwick Close). Reread that if you have to, because it's as good as tool as any for assessing the way they've played.
Now, Pittsburgh is at least capable of winning games if Crosby goes a few without bounces going his way, or if the power play coughs up a few shorthanded goals. They beat Columbus largely because of secondary scoring and solid possession numbers throughout their lineup. Given what they'd worked with for much of the regular season, that borders on amazing.
It's not earth-shattering stuff — but it is, if nothing else, some proof that checking a simple possession number (like Pittsburgh's Corsi For during the regular season) and using it as the be-all, end-all for a prediction isn't smart or special. This is not news for anyone with a long-standing interest in the topic, but as more people are exposed to "advanced stats," or whatever you want to call them, it's worth remembering.
