An introduction to hockey analytics

By Case Keefer (contact)

Saturday, June 10, 2017 | 2 a.m.

When the Florida Panthers fired Vegas Golden Knights coach Gerard Gallant last year, rumors persisted that his stubbornness to embrace the organization’s aggressive use of advanced statistics precipitated the change.

Gallant allowed the whispers to propagate for several weeks before he broke his silence and denied analytics playing a major role.

“No, it wasn’t huge about analytics,” Gallant told ESPN.com. “For me, analytics is certainly a part of coaching, but it’s not the whole thing. In my mind, if I take a job, analytics is part of it for sure, 25 to 30 percent, whatever percentage you want to put on it. It’s definitely a tool. If you get the right information, you’re happy with that. Every coach uses analytics. We all go over the same stuff.”

“I wasn’t fired because of analytics.”

The true extent to which a philosophical divide on analytics impacted Gallant’s ouster is irrelevant now, but it helped illustrate a larger point on the state of the NHL. The analytics revolution has fully arrived in hockey.

It’s no longer a question of whether a franchise is using newer-age statistics; it’s a question to what extent they’re employing them.

A preponderance of numbers is available to teams’ coaching staffs and front offices, more information than ever before. We could move from our Hockey 101 series to a semester-long graduate-level course and still not comprehensively cover all the data.

But since this was meant to be introductory, let’s stick to the basics in the final Hockey 101. And, yes, analytics are basic at their core.

There’s a misconception that a secondary math degree and a graphing calculator are required to understand the numbers. That’s far from the truth.

The basic premise of analytics in hockey is that goals are a deeply misleading statistic and offer little predictive value. Anyone who’s ever watched a few hockey games can surely understand the random and wacky ways that the puck can find the net.

A team’s true strategy, therefore, should be maximizing their scoring opportunities while limiting the chances of their opponents. But it’s extremely difficult to track time of possession in hockey, and not something the NHL has implemented.

The best way to measure opportunities, therefore, comes from shots. The NHL’s traditional method for determining a shot, however, is also flawed because it only counts when a puck goes into the net or is stopped by the goalie for a save.

It excludes both blocked shots and ones that miss the net. Enter Corsi, which adds all shot attempts together regardless of how they ended.

Named after current St. Louis Blues goalie coach Jim Corsi, it swept the NHL more than any other advanced statistic. Every team has a Corsi for, which is all of their shot attempts, and a Corsi against, all of their opponents' shot attempts.

It’s often manipulated into a percentage or a per-60 minute statistic. For example, the Western Conference champion Nashville Predators had a Corsi of 3,790 this season with their opponents posting a 3,581 Corsi.

Add the two together and divide to calculate the Predators’ Corsi percentage of 51.42, or multiply by 60 and divide by total ice minutes (3,935.37) to get a Corsi for of 57.78 per-60 minutes and a Corsi against of 55.29 per-60 minutes.

The Predators barely snuck into the playoffs, as the final entrant in the 16-team field by points, but their Corsi percentage indicated they were much stronger as it ranked sixth in the league.

Corsi creates a much larger sample size than goals or shots, the idea being that it can more accurately rate a team’s success. It’s also calculated for players, but more flawed in that regard because it doesn't account for how much more valuable a shot attempt by someone like the Pittsburgh Penguins' Sidney Crosby is compared with a random role player.

Another gripe with Corsi is including blocked shot attempts, because some argue that’s more a defensive skill. Fenwick, named after creator Matt Fenwick, was created to address that problem by taking out blocked shots but otherwise including everything Corsi incorporated.

The NHL has made Corsi and Fenwick readily available, tracking the numbers itself and logging them on its advanced stats site. The numbers can also be found at Puckalytics and Hockey Analysis.

Corsi and Fenwick are a good start, even though there are countless other metrics available to evaluate the NHL. It’s a guarantee that the Golden Knights, including Gallant, will be making use of them.

Case Keefer can be reached at 702-948-2790 or [email protected]. Follow Case on Twitter at twitter.com/casekeefer.