clock menu more-arrow no yes

Filed under:

Calvin de Haan and Corsi, hockey’s favorite advanced statistic, don’t mix well

Corsi, a fancy term for shot attempts, has become a common statistic in the hockey world. But de Haan isn’t convinced of its value.

Calvin de Haan actually has a positive Corsi rating for his career. But don’t tell him that.
AP Photos

LAS VEGAS — Calvin de Haan hears about Corsi a lot more than he’d like.

He has heard it in free-agent negotiations. He hears about it all the time on Twitter. And he’s hearing it tossed around as a criticism of the Blackhawks, who currently rank 28th in the NHL in Corsi percentage.

“You can’t win a game with a calculator,” he said Tuesday. “I know everyone uses it and it does come up in negotiations and free-agent signings and stuff. I went through it, it’s a part of that now, it’s a big bargaining chip. But no one in here really cares about, I don’t think. No one goes home and reads books or does their own formulas or anything.”

De Haan attracted a sizable dose of vitriol from the outspoken and growing hockey analytics community when he casually called Corsi — a term for shot attempts — an expletive after the Hawks’ loss to the Penguins last Saturday.

But the defensive defenseman doesn’t fit the stereotype of an old-school player rejecting all statistics but goals. And he clearly knows a lot about Corsi and its statistic cohorts, even if he doesn’t believe in it entirely.

So hearing his opinion was interesting.

“There’s so many variables in a hockey game,” he said. “There’s 10 guys on the ice, two goalies, four refs. It’s not like baseball, where it’s pitcher vs. batter. That’s the analytical part of that sport, one-on-one. There’s so many variables on the ice that I think there’s not really an algorithm or formula to show how the game goes sometimes.”

De Haan readily admitted there are situations where Corsi numbers do accurately depict a game, citing the Hawks’ ugly recent loss in San Jose, when they took 38 shots to the Sharks’ 58.

He also sees plenty of examples where Corsi doesn’t align with the true outcome of a game, though. He mentioned the Islanders’ recent 10-game winning streak, which they pulled off despite ranking 30th in the league in Corsi percentage right now (they had a 46.6 percentage over that 10-game span)

“You can get outshot 30-10 but still win the game,” he said. “Doesn’t mean your chances weren’t good. You could’ve had 10 quality chances versus maybe 25 from the perimeter that aren’t going to go in.”

That comment implies that scoring chances are a better indicator of effective play and success than mere shot attempts, and the data does support that assumption. Over the last three seasons, team scoring chance ratio has a correlation of 0.58 with standings points, while team Corsi ratio has a correlation of 0.47.

But 0.47 is still a moderately positive correlation — thus why Corsi is so commonly used throughout the hockey community, and why de Haan’s criticism of it sparked a reaction.

Coach Jeremy Colliton, whose job duties necessitate a closer relationship with analytics than de Haan’s does, said Wednesday he sees both sides of the argument.

“It’s a big conversation,” Colliton said. “There’s reasons why it’s valuable; there’s reasons why it’s noise. I think just like all the other stats, you’ve got to dig deeper into it. There are multiple things that we track and try to create a full picture of how we’re playing.”

Ironically, de Haan actually owns a positive Corsi differential over his seven-year career, though he has struggled (along with most of the rest of the Hawks) in that regard this season.

Yet it would be wise not to mention that to him.