Ottawa hockey analytics conference goes beyond the boxscore
'The more information, the better,' say organizers about advanced NHL stats
It used to be that goals, assists, penalty minutes and plus-minus were about all you needed to know about a particular NHL player's skill set.
The people behind the Ottawa Hockey Analytics Conference at Carleton University, however, say those stats are just the tip of the iceberg.
"We're getting into graduate-level sort of mathematics and statistics here," said Michael Schuckers, a statistics professor at St. Lawrence University in Canton, N.Y. and one of the organizers of today's conference.
Co-hosted by St. Lawrence University and Carleton's school of mathematics and statistics, the conference represents a meeting of the minds in the growing field of advanced hockey statistics.
New stats measure value 'beyond basic counting'
The general goal of advanced hockey statistics is to look at the player's impact on all of the events that happen on the ice, said Shuckers, and go beyond the surface-level stats that have traditionally appeared in newspaper box scores.
Take a player like Erik Karlsson, says Schuckers. A traditional statistic like plus-minus — which measures how often a player is on the ice when goals are scored, either by their team or by their opponent — would suggest that the Ottawa Senators' lone all-star, who doesn't even crack the top 200 in that stat, is a tad overrated.
According to the advanced numbers Schuckers prefers, though, Karlsson is actually one of the top three defencemen in the league, he told CBC Ottawa's In Town And Out on Saturday.
"I think we can get at some value there that is beyond just the sort of basic counting and the traditional metrics we've used," he said.
Some advanced statistics like Corsi — which measures a team's shot attempt differential while playing five-on-five, with the higher the number the better — are gradually becoming better known among hockey fans.
Saturday was the conference's second edition and more than 200 people showed up, including representatives from the NHL's Florida Panthers and Tampa Bay Lightning and the NBA's Boston Celtics.
"Carleton has made a big commitment to move in the area of data science, and this is just one of the applications that's certainly possible," said Shirley Mills, a statistics and mathematics professor at Carleton and the conference's co-organizer.
Mills was scheduled to deliver a talk on the emerging field of player tracking, in which wearable technology is used to capture data about things like a player's speed, movement, shift length and total ice time.
It's technology that's already been used in other sports, especially soccer.
- NHL experimenting with tracking technology
- Hamilton-bound company uses GPS, chips to track and train athletes
While not everyone is ready to give up goals and assists and accept advanced analytics as the new hockey gospel, Mills says there's no harm in giving them a try.
"I think, the more information, the better. Keep an open mind, try the different methodologies, and see what you get out of all this," she said. "You've got nothing to lose."