Alexandra Mandrycky gets that some local hockey fans, especially complete newbies, might view using advanced analytics as akin to flying a spacecraft.

Mandrycky knows about lacking familiarity, having had zero hockey knowledge until her Buffalo Sabres fan future husband introduced her to the sport in 2008. But she’d studied enough data analytics and programming for an industrial engineering degree at Georgia Tech to start building predictive models around hockey numbers as a side gig, boosting her interest level to where she landed an NHL job in Minnesota four years ago.

Now, as NHL Seattle’s chief numbers-cruncher, Mandrycky wants to show advanced stats are approachable and can similarly increase local fans’ hockey interest ahead of her team’s October 2021 debut. Asked for three easily grasped advanced hockey stats, she provided some to help fans follow the current season with a sharper focus.

NHL Seattle


“There are so many moving parts to a hockey game, it can be overwhelming to try to learn and pay attention to everything,” Mandrycky said. “Having numbers to refer back to, or particular players you want to watch more closely, can allow fans to understand and digest the game a little easier.”

Her chosen stats — Corsi percentage, Expected Goals and 5-v-5 Points/60 — aren’t nearly as advanced as what NHL teams use as part of their proprietary data. But they’re found on public websites, comprising basic premises that spawned many of today’s more secretive analytics.

The so-called “Corsi’’ stat is likely the most commonly known, named after onetime NHL netminder and current Columbus Blue Jackets development coach Jim Corsi. Mandrycky said Corsi is a fancy name for counting all shots “either on-net, missing the net or blocked by opponents.”


The “Corsi percentage’’ — also known as “shot attempt percentage’’ — counts total shots a team takes with a specific player on the ice. “This stat is useful because it predicts future team success — goal scoring — better than only shots on goal, and goals themselves,’’ Mandrycky said.

The premise is simple: Using players whose teams shoot more when they are on the ice should generate more chances and more goals.

Not exactly a perfect theory. But even Corsi, who helped groom Hall of Famer Dominik Hasek, is fond of saying: “Statistics are like a lamppost for a drunkard. You can either use it to lean on, or illuminate.’’

Those seeking illumination on the best Corsi percentage players can find them on under the “SAT %’’ field, which tracks it in five-on-five situations.

As of Wednesday, Los Angeles Kings defenseman Sean Walker led everyone with at least nine games played at 65.58% — meaning nearly two-thirds of the shot attempts are n his team’s favor when he plays. The Carolina Hurricanes had the most players in the top-10, with Teuvo Teravainen second at 65.41%, Sebastian Aho fourth at 62.16% and Nino Neiderreiter seventh at 61.31%.

Not surprisingly, the Hurricanes were No. 2 overall with their players averaging 55.47%. The Philadelphia Flyers led all teams at 57.98% while the New York Rangers were dead last at 42.45%.


But how dangerous were all those shots? Carolina sat just 10th at 3.22 goals scored per game, the Flyers 13th at 3.00, while the Rangers were 19th at 2.71, meaning some shots were clearly more lethal than others. Corsi percentage won’t tell us much, but Mandrycky’s second stat — expected goals — will delve further.

It registers things like a shot’s location and whether it was taken on a rebound or a rush opportunity — helping gauge the “likelihood” a shot will result in a goal. Typically, the closer a player gets to the net before shooting, the more likely he’ll score.

“This is a way to describe if teams or players are getting to the dangerous areas of the ice,’’ Mandrycky said.

Expected goals are tallied at the website, where Washington Capitals star Alex Ovechkin led the league midweek with 5.8 in all situations. Thing is, Ovechkin actually had seven goals — meaning he was 1.2 over his “expected’’ total and might have gotten “lucky” a shot or two.

On the opposite end, Flyers winger James Van Riemsdyk hadn’t scored as of Wednesday but his expected goals tally was four. So, things weren’t going his way — he finally scored Thursday — partly explaining why the Flyers were 13th in goals per game despite leading everyone in Corsi percentage.

Mandrycky’s stat list concludes with “points per 60 minutes’’ but in five-on-five situations only. She feels isolating “5-v-5’’ play is useful because that’s when the bulk of play occurs, as opposed to power plays and penalty-killing situations.


She said players average from 8 minutes, 7 seconds to 17 minutes, 28 seconds of 5-v-5 time per game, so the stat standardizes things by estimating how many combined goals plus assists they’d tally if playing an entire 60-minute contest. A player’s point total is divided by 5-v-5 time spent on the ice and multiplied by 60.

The “Natural Stat Trick’’ and “Evolving Hockey’’ websites showed Filip Forsberg of the Nashville Predators leading by midweek at 5.25 points per 60 minutes of 5-v-5 play. Nashville had five players in the top-15, helping explain their NHL-best 4.22 goals per game.

But somewhat less expected was Montreal Canadiens forward Brendan Gallagher at No. 2 with 4.67 and linemate Phillip Danault No. 7 with 3.72. Both are largely why Montreal midweek was a surprising fifth in goals per game at 3.67.

But they had only one other top-50 player — Jonathan Drouin at No. 48 — perhaps explaining why the Canadiens still have trouble scoring outside their top-line and an improved power play.

“These aren’t the most advanced in the world,’’ Mandrycky said of her stat offerings. “But I think they are good starting points for things a general fan could pay attention to.’’

And perhaps delve in enough to be floating out NHL résumés of their own by the time Seattle’s team takes the ice.