PDA

View Full Version : Hockey Analytics Conference Set to Expand Conversation



Phil in Absentia
04-07-2016, 08:54 PM
“If you think Corsi is analytics, you think the crossbow is a modern weapon,” said Tingling, a management information systems professor at Simon Fraser University. “The fact of the matter is hockey’s in the dark ages. Nice to see them finally step up a little bit.”

Corsi — plainly referred to as “shot attempts percentage” by the NHL — is a fairly basic formula that helps figure out how much a team possesses the puck relative to its opponent.

But as more data has become available in recent years thanks to better tracking of players on the ice, advanced statisticians have rapidly expanded with new, more detailed ways to both measure, and attempt to predict, performance.

"Analytics are about how do we get the most juice out of the lemon?" said Tingling. "How do we get the right piece of the jigsaw puzzle in place?"

With questions like those in mind, Tingling is part of a group at SFU hosting a hockey analytics conference in Vancouver on Saturday that's expected to attract executives, academics, students and fans.

And while advanced stats are being viewed more favourably in the NHL — many teams have analytics departments and the league's website has been updated with enhanced statistical categories ——vent organizers say it's still an uphill climb.


The small-market Oakland Athletics overcame budget constraints to topple big-spending opponents in the early 2000s using the "Moneyball" strategy. Swartz doesn't expect there to be as big an impact in the NHL, but the proper implementation of analytics could still be significant.

"Teams will get an edge by doing things a little bit different," he said. "Sometimes small gains are important. If you move a couple points up in the standings, that could be a playoff spot."

http://www.sportsnet.ca/hockey/nhl/hockey-analytics-conference-set-expand-conversation/

fletch
04-07-2016, 08:59 PM
Be interested to see what they come up with. Hockey is such a team sport it's hard to generate individual statistics. In baseball, you've got a large sample size for at-bats for batters, or pitches for pitchers, or fielding plays for fielders. Quaterbacks have completion percentages in football. Perhaps they can steal some of analyzes done in soccer (possession time, pass completion percentage, etc) but it's still a tough nut to crack

CreaseCrusader91
04-07-2016, 10:11 PM
There have been a number of conferences this year with similar participants. One held in Rochester, another in Florida and one in Ottawa. Some good stuff has come out of them and I think a lot of the presentation materials are online. I know a few people presenting on Saturday.

Pete
04-07-2016, 10:51 PM
As long as they move away from shot attempts=possession, then I'm eager.

CreaseCrusader91
04-07-2016, 11:02 PM
As long as they move away from shot attempts=possession, then I'm eager.

Yup. Some of the more interesting studies going on now are looking at zone exits, entries and passing.

IE: what teams are effective and now. Is it a team with a number of players who can recover puck and skate it out? Is it a team with good transition play? Is it a team that dumps it in and so on.

Ultimately there will be incomplete, for lack of a better word, results until SportVu type tracking is implemented.

I agree with the shot attempt premise you brought up.

Shots attempts for me was piggy backing off the soccer approach of possession. The team that shoots more and holds it chases less and therefore drives play.

Puck Head
04-08-2016, 12:59 PM
As long as they move away from shot attempts=possession, then I'm eager.

This, well said.
The Soviets of the 70's and 80's owned the puck the entire game, but was not necessarily reflective in shots for or against.

They didn't waste stupid shots on net. At the same time, opposition had the puck so little they'd throw it on net whenever they got a chance.

Puck Head
04-08-2016, 01:02 PM
One "stat" we'd somewhat track during video review, was the amount of times a player set up a scoring chance.

Scoring chance = any shot from within the house, or any odd man break.

I wanted to know who was creating these chances vs killing them