Tampa, Florida – Islanders don’t want to hear about your math equations.
Although they acknowledged that analytics have a role in hockey, they still valued the test of eye and gut feelings over numbers.
“There is will and structure in our game,” coach Lynn Lambert said Thursday. “Those are not really being analyzed by analytics. We are analyzing it.”
There has been a recent growth in the use of numbers in hockey – every team, islanders included, has analytics staff as part of their portfolio – but compared to the other three major sports, the NHL is still far behind. This is partly due to the nature of the game. Hockey is continuous, random, and difficult to break down into the kind of individual events that are easy to analyze mathematically, such as the pitch in baseball, snap soccer, or possession of basketball.
Most publicly available numbers in hockey (individual teams track their own data) use shots as property reserves, a reasonable but rudimentary estimation method. Expected goal models, which are widely cited, can be useful, but because there is no general disc tracking data, they fail to account for factors such as whether a shot was preceded by a pass; Such shots are difficult for the goalkeeper to stop.
It’s everywhere these days, so I think you definitely see it,” said Matt Martin. “I think it can definitely be flawed as well. I think sometimes you can have a really good game defensively against someone higher streak where you lose a Corsi fight or whatever, but you keep them off the plate and you don’t give up a lot of high-risk opportunities. Analytically speaking, it sometimes turns out to be a really bad game where you get a click on the back of your coaching staff about how good a job you guys did.”
The island’s fourth line, of which Martin is a part, is a good test case. Since they spend a lot of icy time in the defensive zone or in the forward check, their analyzes reflect negatively on their play. But the goal of the fourth line is not necessarily to cause an attack, but to have a physical effect on the game, exhausting opponents and keeping them off the board.
“I think you can walk around the locker room, anybody can tell you if they’ve had a bad game or a good game,” Martin said. “We know as the pros. … You’re going to take on Conor McDavid, you’ll probably lose the Corsi fight. But if you limit the amount of high-risk opportunities he has and take him off the board, we take that every night.”
Martin said there is not much interaction between the Islanders Analytics team and the players. Taran Singleton, a longtime video coach with the Devils when he was the current general manager of Islanders, Lou Lamoreello of NJ, is one of five employees dedicated to analytics.
“I’m not a big analytics guy,” said Zach Barris. “I just think there are a lot of factors in this game that are not taken into account. I would go beyond shots on goal, because you give yourself a chance to score. Anything after that, it’s a little bit hard to scale.”
Like Martin, Barris cited factors such as the opponent as the reason behind the inadequacy of the analytics.
“Are you playing against [Patrice] Bergeron and some silk [Trophy] The winners? Or are you playing against a fourth line all night? ‘ said Paris. ‘Where do you begin to face? Does every shift begin a controlled hack? Do you lose a tie? If you lose the pull, the puck disk may not touch the entire shift.
“You don’t reset all the time and start over like a playground [in baseball], you know what i mean? This is just my theory. I may be wrong. I’m sure there’s value in that, but I don’t stress too much about it.”