Jacob Markstrom and Anders Nilsson were given a nearly impossible task heading into the season. One of them had to suddenly go from being a long-time backup to a legitimate number one goaltender at the age of 27.
Very few goaltenders have managed such a feat in NHL history. Most goaltenders that break into a starting job in the NHL that late in their careers played in Europe or were exceptional in the AHL but weren’t given an NHL opportunity. One of the few examples that Markstrom and Nilsson could look to was Olaf Kolzig, who went from being an underwhelming backup to carrying a team to the Stanley Cup Finals and winning a Vezina trophy in the span of a few years.
To up the difficulty, Markstrom and Nilsson had to do it behind one of the lowest scoring teams in the NHL, giving them little margin for error. Unfortunately, they’ve made many errors. They’ve given up too many goals early in games, and allowed soft goals at inopportune times.
55 goaltenders have appeared in at least 20 games this season. Among those 55 goaltenders, Markstrom is 34th with a .909 save percentage and Nilsson is 49th with a .901 save percentage. Both goaltenders are well below the league average save percentage of .913.
Would the Canucks have done better this season with a different goaltender? It’s hard to say definitively — if this hypothetical goaltender made a save that Markstrom or Nilsson didn’t, that changes everything about how the rest of the game is played, like the proverbial butterfly that causes hurricanes.
We can provide a theoretical answer to this question, however, by looking at one particular goaltending statistic: Goals Saved Above Average (GSAA).
GSAA, as calculated by Emmanuel Perry’s Corsica site, compares a goaltender’s performance to how an average NHL goaltender would perform facing the same shots. This includes components of shot quality, such as the distance, angle, and type of shot, as well as if the shot is a rebound or rush chance.
A positive number indicates that the goaltender has saved his team that amount of goals compared to this imagined average goaltender.
For instance, the defending Vezina winner Sergei Bobrovsky has a GSAA of 31.67 in all situations, providing a massive contribution to the team’s goal differential. Considering the Blue Jackets’ goal differential is just plus-2, it’s fair to say the Blue Jackets wouldn’t be anywhere near as good without Bobrovsky in net. It’s entirely likely that they wouldn’t even be in the playoffs.
As for the Canucks, it’s clear that Markstrom and Nilsson have been far below average.
Both Markstrom and Nilsson are well into the negatives. They’re not the worst goaltenders in the NHL by this metric — that dishonour goes to Scott Darling of the Carolina Hurricanes — but they’re definitely in the bottom tier.
The biggest struggles for Markstrom have come on the penalty kill, where only Mike Condon of the Ottawa Senators has a worse GSAA. As an interesting note, both Markstrom and Nilsson have saved the Canucks a little over a goal while on the power play, making some good saves on shorthanded chances.
One note before continuing: there are some elements of shot quality that are not included in the GSAA metric, simply because the data is not available without manual tracking. Two of the main elements would be puck movement before the shot (was there a pass across the slot or from below the goal line?) and whether there was a screen in front of the goaltender preventing him from seeing the shot.
With that caveat in mind, let’s see what difference an “average” goaltender would make for the Canucks.
The Canucks currently have the second worst goal differential in the NHL at minus-50, ahead of only the Buffalo Sabres, who have been truly brutal at minus-66. It shouldn’t come as a surprise that the teams with the worst goal differential are at the bottom of the standings.
In fact, the only team with a negative goal differential that is still in a playoff position is the New Jersey Devils at minus-3. At this point in the standings, goal differential is a pretty solid indication of how good a team is.
If the Canucks replaced Markstrom and Nilsson with two theoretical “average” goaltenders, what would be the result? That would remove 16.53 goals against; let’s round it up to 17 to give us a nice round number.
That would make the Canucks total goals against this season 220 instead of 237 and bring up their goal differential to minus-33.
That’s still pretty bad.
A goal differential of minus-33 would place them in 25th in the NHL between the Edmonton Oilers (minus-29) and Carolina Hurricanes (minus-35). In essence, a pair of league-average goaltenders would still leave the Canucks well out of the playoff picture and would likely still mean picking inside of the top-10 at the NHL draft.
What if they didn’t have a league-average goaltender? Let’s say they instead had one of the best starters from this year — Sergei Bobrovsky — and, with him, a fantastic backup, like Ryan Miller who has a GSAA of 18.11 in 24 games played this season.
That’s a combined GSAA of 49.78, so let’s give the Canucks a 50-goal swing in goal differential. That brings their goal differential to an even zero.
Even that might not be enough to make the playoffs in the Western Conference. The team with the worst goal differential in playoff position in the west is the Anaheim Ducks, and they’re plus-16. Two teams with positive goal differentials, the St. Louis Blues (plus-6) and Dallas Stars (plus-10), currently sit outside the playoff picture.
Essentially, even if this season’s Canucks had two of the best goaltenders from this season, it would still require some exceptional luck for them to make the playoffs.
Really, this article is a long way of saying that if the Canucks or their fans are hoping that Thatcher Demko will fly in to save the day next season, they should temper their expectations.
*All statistics via NHL.com and Corsica.hockey.
**A note on goal differential: I’m using true goal differential instead of the numbers on the NHL standings, which include shootout goals.