The Canucks goaltending as it stands just isn’t good enough.
Since the start of the 2017-18 season, Vancouver has allowed 3.23 goals against per hour in all situations — ranking 29th in the NHL during that span. A significant amount of that blame can be levied to the skaters playing in front of Jacob Markstrom and Anders Nilsson, but there’s evidence to suggest that the Canucks’ Swedish goaltenders haven’t been good enough either.
Corsica’s Goals saved above average (GSAA) metric analyzes the volume and quality of shots that each team gives up and calculates what an average NHL goaltender would concede in terms of goals.
Markstrom and Nilsson lag far behind — combining for a GSAA of minus-11.18 in all situations. In other words, this model believes that the Canucks would have surrendered 11 fewer goals from the start of last season to now if they had a league average goaltender facing the same shots that the Canucks allowed.
Advertisement
When mapping out the team’s rebuild, the blueline is rightly brought up as a major concern, but questions about the team’s goaltending future shouldn’t be overlooked, either. It’s true that Thatcher Demko is one of the top goaltending prospects in hockey, but it’s also naive to have full confidence in him as the team’s proverbial goalie of the future given the volatility of the position. Remember, we’re only a few years removed from Jacob Markstrom being considered a top-five goalie prospect in the world.
There are whispers that the management group isn’t particularly high on Demko, too, which has led to speculation that the team could perhaps take a run at pending UFA Sergei Bobrovsky in the summer — a possibility that my The Athletic colleague J.D. Burke analyzed in great detail.
It may seem premature for the Canucks to worry about their goaltending, but it matters right now because of just how much of a difference it can make for a team’s results.
The man in question, Bobrovsky, led all NHL goaltenders last year in surrendering 32.8 goals fewer than an average netminder. In the Canucks’ case, if you substituted Bobrovsky for Markstrom (and kept Nilsson as the backup), the Canucks would hypothetically improve their all situations goal differential from minus-40 to almost break even at minus-8. Given that three goals are roughly equivalent to one standings point, that would amount to an 11-point swing in a season.
The problem is that finding a consistent, long-term solution in net is akin to searching for a needle in a haystack. Nothing exemplifies this assertion as well as comparing the save percentages for individual goaltenders from 2016-17 to their 2017-18 performances:
There’s no correlation between 2016-17 and 2017-18 save percentage for goalies that played at least 1500 minutes in each season. In other words, 2016-17 save percentage results had essentially zero predictive value for next year’s save percentage. The standard error tells us that each of the goaltenders varied about 10 clicks (eg: .920 to .910) year over year on average.
The unpredictability is captured best when looking at the bottom-five goalies according to save percentage for 2017-18.
You can see that each of the goalies above were at least competent between the pipes in the year prior to their disastrous 2017-18 seasons.
This data is interesting, but many will rightly argue that save percentage is very team dependent. Let’s try and account for that by conducting the same experiment — this time with goals saved above average analyzed instead of save percentage.
There’s some level of correlation here, but it’s really weak. Moreover, the standard error tells us that each goalie deviated by almost 15 GSAA year over year on average. At the same time, what applies generally doesn’t necessarily apply to all. After all, goaltending is a skill and is something that can obviously be controlled to a degree.
As you might expect, there are some upper echelon goaltenders that are able to maintain their performance year over year.
In summation, goaltending is incredibly important, yet there are very few in the NHL who can provide quality play consistently. This spells out bad news for the Canucks, who will need consistently good play in net to bail out a back end that still looks relatively bare moving forward.
Given how valuable such a goalie would be, I believe that Demko is the Canucks’ most important prospect after Quinn Hughes. But how likely is it that Demko is the long-term answer in net? Let’s look at statistical comparables to see just how often goaltending prospects like Demko pan out.
Advertisement
NCAA comparables
To search for comparables for Demko, I went through every NCAA season since 2000 and found the goalies with the closest cumulative U19, U20 and U21 statistical profile. I also set a minimum in terms of games played for each season. Here are the results I found.
It’s a positive sign that all six comparables have graduated to the NHL. Cory Schneider, Jonathan Quick and Jimmy Howard have all had successful careers as starting goaltenders, while Connor Hellebuyck is solidifying himself on a similar path in Winnipeg.
Jon Gillies and Jeff Zatkoff have had little-to-no NHL success, but all in all, it’s a good sign that 4-of-6 comparables for Demko made a name for themselves in the big leagues.
If there’s a concern in all this, it’s that Demko didn’t have a smooth transition to professional hockey in his first year after leaving Boston College — posting a .907 save percentage with the Utica Comets before figuring things out last season. In light of this, we can also look at Demko’s AHL performance to see how often young goaltenders succeed in the American League as well as how predictive it is for eventual NHL performance.
AHL comparables
To create a historical precedent, I searched for goalies who succeeded as AHL starters in a similar age range as Demko. The former second-round pick was 21.8 years old at the start of last year and so I filtered by age to only include goalies that were plus or minus half a year during their AHL season of success (20.8-22.8 years-old). In terms of performance, I looked at goalies who posted a .918 or higher save percentage in that season with a minimum of 30 starts.
There’s some degree of assumption involved here. Tristan Jarry, for example, looks like he’s going to hit 100 games played so we’ll give him the nod. We’ll also count Markstrom as a starting goaltender in spite of the fact that he’s better suited to being a backup.
In all, more than half of Demko’s AHL comparables played or are on track to hit 100 games played, while one in three became starters at the NHL level.
Advertisement
I tend to believe that the actual odds of Demko panning out are a bit higher than this sample indicates when you consider the context. As we outlined earlier, save percentage is influenced by the team that plays in front of the goalie and in the case of Demko, he was on a mediocre Comets squad that was dragged into the playoffs in large part due to the 22-year-old netminder’s play. If he played behind a better team, I’d like to think that he’d have better numbers for both the year we’re looking at in this sample, as well as for his rookie season in the minors.
On the other hand, when we examine the busts in the group we see a clear pattern emerge — one that we can look for this season to project Demko.
All five of the outright busts in the group saw a significant decline in their second AHL season and generally weren’t able to sustain their U23 level of play for the rest of their careers. On average, each of these goalies dropped by more than 10 save percentage points. In these cases, it would appear that their initial success was simply a fluke.
In other words, this season will be a major test to see if Demko’s performance last season is something he can sustain this year and if so it bodes really well for his odds of flourishing in the NHL. Some might think the answer to that question is an automatic yes, but goalies really are voodoo and it’s especially important not to make assumptions considering Demko is coming off a concussion that delayed the start of his season by almost two months.
But if we assume Demko can perform well again this year now that he’s back in action, these are the comparables that we’re left with once we subtract the five players from our aforementioned group.
- Tristan Jarry
- Matt Murray
- Connor Hellebuyck
- Petr Mrazek
- Jacob Markstrom
- Martin Jones
- Josh Harding
- Laurent Brossoit
Jarry’s a year older than Demko and is currently splitting time between the NHL and American League. His long-term future is up in the air like Demko’s, but aside from that, everyone else on the list is already established one way or another.
Murray, Hellebuyck and Jones are league-average or better starters, while Markstrom looks like a 1B type goaltender at best. The rest either are or were backups.
Advertisement
When might Demko be ready?
It seems that goalies are jumping into the NHL at younger ages nowadays. John Gibson, Andrei Vasilevskiy, Juuse Saros and the aforementioned Murray are just some of the modern-day netminders that have graduated to the NHL at a young age. The difference with these players is that they had almost immediate success in the minors.
Every player’s development path is different, though, and there’s a potential template for the Canucks and Demko to follow with Cory Schneider.
Schneider spent three full seasons in the AHL before graduating to the pros. I’d like to see the Canucks exercise similar caution and ensure they give Demko another full season to marinate in Utica — especially with the current state of the Canucks blueline.
Conclusion
Good goaltending is extremely important, yet finding a goalie who can consistently hold down the fort is really difficult. As our analysis on 2017-18 goaltending showed, there was virtually no repeatability league-wide when it comes to performance between the pipes.
All goalies are volatile, but the skill comes in the fact that the good ones will generally be on top of their game for longer stretches than their dips — something that will bear itself out over a larger sample of multiple seasons.
Going back to the Canucks, the success of their rebuild is heavily dependent on the goaltending situation moving forward. Markstrom and Nilsson are not feasible No. 1 options in the long run and Demko is the best shot the organization has in producing one.
If we look at Demko’s development strictly through a statistical lens, we can mark this season as a critical point in evaluating his chances of blossoming into an NHL goalie. At this point, more than half of Demko’s AHL comparables have gone on to play in the NHL at some capacity, while one-in-three have become No. 1s.
Advertisement
Those sound like disappointing odds, but when you consider that only two of the 17 goalies drafted in the first two rounds between 2009 and 2013 became legitimate starters (John Gibson and Andrei Vasilevskiy), Demko appears to be in a good place.
The Canucks’ unsettled view on him, however, suggests they’re aware of both the volatility and importance of goaltending. For the rebuild’s sake, you’d hope the 22-year-old American can buck the trend and become one of the few goalies selected in the first two rounds of the draft that actually pans out.
(Top photo: Jeff Vinnick/NHLI via Getty Images)