Which Goalies Can You Trust These Stanley Cup Playoffs?
A Closer Look at What Makes a Netminder Reliable
Even in my younger years as a junior hockey play-by-play announcer, one debate gave players and outsiders alike a lot of consternation:
Is quality goalkeeping a skill or just luck?
On the one hand, a goalie can position himself properly at the right spot on the ice, know shooter tendencies and make sound decisions with the puck to make sure no easy goals are allowed. On the other hand, a goalie does not have any control over the quality of shots he faces and the puck can bounce and deflect in funky ways.
Deciding these debates cannot be generalized so easily, but with more refined data, maybe we can inch closer to a correct answer. Those portions we can isolate as based on skill may be predictive, which means we can then figure out which netminders will continue to excel as long as they’re in the Stanley Cup Playoffs.
“Goals Saved Above Average” (GSAA) is a statistic from Natural Stat Trick that consists of the difference between the goalie's “Goals Against” and a “Goals Against with the same Shots Against and the average SV%”. In other words, this stat looks at how much better this goalie is saving shots than what is considered average. Ultimately what we want to know is how predictive or sticky these stats are from one season to the next, because something based on skill should naturally tell us more about future performance than anything with luck or variance. Looking only at the last two regular seasons and goalies who played at least 25 games in those two campaigns, here is what their GSAA’s looked like:
Goalies on the upper-right part of the graph have been excelling both last season and this one. Those on the upper-left struggled in 2023-24 but improved drastically to where they have been massive assets1. Also important is the red line representing a goodness-of-fit regression, or how well the 2023-24 performance explains this regular season. It turns out, not well at all, with an R^2 of 0.092 and a Cohen’s D of 0.08. However, one thing that is clear is just how dominant Winnipeg’s Connor HelleBuyck and Toronto’s Anthony Stolarz have been recently.
Now, is it possible to isolate something from GSAA that can be more predictive? One consideration is looking at shot selection. Natural Stat Trick has it broken down into three zones largely based upon proximity to the net: low-danger shots, medium-danger shots and high-danger shots. The closer a shot is to the goal, the more dangerous it is. First, let’s look at high-danger save percentage of these goalies over the last two regular seasons:
First, by looking at the red regression line, we see a slight improvement when isolating just high-danger shooting percentage, albeit still a small effect (R^2 = 0.106, Cohen’s D of 0.03). Second, many of the same goalies we saw previously are also excellent when needed most. Stolarz stands out with a 0.875 high-danger save percentage this season, but Tampa Bay’s Andrei Vasilevskiy has also come on strongly with his 0.873 high-danger save percentage.
Finally, let’s look at low-danger save percentages:
Interestingly, low-danger save percentage is less predictive in this small sample size (R^2 = 0.080, Cohen’s D of 0.02). These rates should naturally be higher because the shots are easier to stop, but perhaps it’s more difficult to predict exactly when the puck will somehow sneak through. If those freak occurrences happen enough, a save rate will plummet and predictability becomes less likely. Still, seeing names like St. Louis’ Joel Hofer, Anaheim’s John Gibson and Stolarz lead the way over the last two seasons highlights the incredible impacts they have had.
So, when it comes to isolating a goalie’s skill, it may be slightly more advantageous to look at high-danger situations for consistency and predictability, though I must emphasize “slightly”. Still, in terms of takeaways for these Stanley Cup Playoffs, uncomfortable as it might be, the situation at goalie for Toronto should remain outstanding. Couple that with the hockey team’s forward depth and you have enough talent and examples to suggest betting on them to win the East and finding a skater or two to have the support necessary to win the Conn Smythe (e.g. Auston Matthews at 38/1 on FanDuel).
NHL logos represent goalies who switched teams this season such as Mackenzie Blackwood or Spencer Knight.