The Vezina Trophy, the accolade awarded to “the most outstanding goaltender” of the season, has been voted on by the NHL’s fraternity of general managers since 1981. At first glance, the men responsible for building teams and assessing hockey talent voting on the performances of players whose position is notoriously hard to evaluate doesn’t seem like a bad idea.
However, when you begin to look at the voting results in recent years, it becomes apparent that one of two things is happening:
1. GMs clearly believe wins are the most important statistic in evaluating an individual goaltender’s performance.
2. GMs just don’t care enough or have the time to do the research required to accurately vote for this award.
Let’s get this out of the way: goaltender wins are a terrible stat to use to describe individual performance. While goalies have a significant effect on stopping the other team from scoring, they have almost no effect on their team’s actual goal production. They’re not the ones scoring goals. This is a very basic statistical principle: a variable that an individual has little control over is not valuable at describing said individual’s performance.
If a goalie stops 40 out of 41 shots in a game, but his team is shut out and loses 1-0, should he be punished for that? Any reasonable person would tell you: absolutely not.
There are far better numbers out there that much more accurately describe an individual’s goaltending performance, and they are listed right on NHL.com’s own stats page.
Now, I want to reiterate my earlier point that goaltending is arguably the most difficult position to evaluate in hockey for a number of reasons that people much smarter than I can express more clearly. Any of these numbers still only tells part of the story, and each one has its flaws in terms of individual effect.
However, it’s a fact that a goaltender has much more control over other factors than he does wins, and these statistics are inherently more valuable when deciding who deserves the Vezina.
The most basic numbers used by literally every front office, media outlet, and casual fan are save percentage (SV%) and goals allowed average (GAA). While GAA has some problems as goals allowed can be significantly influenced by the defense in front of the net, and because there is a smaller sample of data to work with, SV% is generally a better and more solid number to examine. This is primarily because there are many more shots on goal recorded in a game than goals, often by factors of around 10 or more. The more data points you have, the more accurate the number gets over a season-long sample at showing a goaltender’s true skill at the thing he is there to do: stop pucks.
Let’s look at this year’s Vezina winner and finalists. Andrei Vasilevskiy won the award this year, and apparently it wasn’t even a close race.
Let’s look at how the top vote-getters performed according to wins, GAA, and SV% and their ranks in each category (minimum 45 GP):
- Andrei Vasilevskiy: 39 wins (1st), 2.40 GAA (4th), .925 SV% (3rd)
- Ben Bishop: 27 wins (16th), 1.98 GAA (1st), .934 SV% (1st)
- Robin Lehner: 25 wins (18th), 2.13 GAA (2nd), .930 SV% (2nd)
It’s clear to the observer that the main reason Vasilevskiy won the Vezina was that he was the starting goaltender on one of the greatest regular season teams in NHL history. That’s not to say the young Russian wasn’t one of the top goaltenders in the league this year; looking at these numbers, you can make a good case that he was. But looking at the numbers that a goaltender has far greater control over like SV% and GAA, it becomes apparent that he wasn’t the best in the league.
For the interested evaluator, the GAA and SV% fall short at evaluating goaltenders, because as useful as they are, these stats lack context. Consider the following example.
Goalie A: .930 SV% (930 saves on 1,000 shots on goal)
Goalie B: .930 SV% (651 saves on 700 shots)
Who is the better goalie? The logical choice here is Goalie A, because he had a 30% greater workload than Goalie B and saved the same percentage of shots. The more shots a goalie faces, the more chances the puck has to go into the net. This is an extremely simplified example, but you get my drift.
I believe that the context of a goaltender’s performance should matter in the choosing of Vezina candidates. For that, we have a an even more advanced evaluation tool, called Goals Saved Above Average (GSAA). Also sometimes called Goals Saved Above Replacement, this stat attempts to find out how many goals a goaltender has saved compared to a league-average netminder when factoring in workload, shot location, and type of shot. These are all numbers published by the NHL. Stats sites corsica.hockey, naturalstattrick.com, and evolving-hockey.com have their own versions with slightly different criteria, but they’re all close to one another and all attempt to give context to a goaltender’s performance.
For example, John Gibson had a GSAA of 28.63 this season in all situations (5 on 5, penalty kill, power play, etc.) according to corsica. This means that if Gibson were replaced with a league average goaltender, he would theoretically have given up roughly 28.63 more goals in a season given the workload, shot location, and type of shot faced. As you can see, the numbers are pretty easy to understand. And yes, Gibson’s 28.63 GSAA was the best in the NHL.
Let’s take a look at the Vezina finalists’ GSAA numbers this season:
- Vasilevskiy: 10.64 GSAA (10th)
- Bishop: 24.66 GSAA (2nd)
- Lehner: 20.38 GSAA (4th)
Looking at this metric (and factoring in Gibson’s eye-popping GSAA), the general managers got this one wrong.
These aren’t the only ways to evaluate goaltending performance, but they are some of the best numbers available to try and fairly determine the best goaltender in the NHL in a given season. Given these numbers and how readily available they are, why exactly are general managers using wins to determine who wins this award?
Is it because they genuinely believe wins are an accurate way to evaluate individual performance? If so, then the credibility of every single hockey executive should immediately come under question when it comes to determining who the best goalies in the league are. As talent evaluators, they have a duty to their teams and to their fans to utilize every tool available to them to make their best judgement. Clearly, they are not doing that with the Vezina Trophy.
Or is the answer simpler than that? The nature of a general manager’s job is to watch his team on a nightly basis and use what he sees as well as the tools available to him to assess both the short and long term performance of his players in order to construct the best hockey team possible. Due to this reality, these executives simply don’t have the bandwidth to pay attention to every other team’s goaltender all season long. It’s unfair to ask them to make these judgements in this context.
Yet, they are asked to evaluate the performance of a goaltender that they likely haven’t had the ability to pay close attention to all season. Also, let’s be real: general managers have much more important things to do with their time than to research and vote on awards looking back on a season already played. They need to plan for the playoffs and the draft on the horizon.
The NHL and the fans believe these awards are important. They winners are etched in the history books and on the trophies themselves. They enter the lore of a sport that carries a tremendous amount of reverence for the most successful players in the game’s history. But year after year, the general managers’ voting record for the Vezina Trophy stirs up new controversy, even if they occasionally get it right.
This isn’t specifically about John Gibson being continually disrespected by the Vezina voters (though I’ll admit that the idea for this article came from that feeling); every year deserving goaltenders are left out of the conversation either by archaic approaches to evaluating goaltending talent or by cop-out decisions by men who aren’t in a position to vote on the award in the first place. This year, very good cases could be made for goaltenders like the Bruins Jaroslav Halak or our old friend Frederik Andersen of the Maple Leafs ahead of Vasilevskiy.
The results speak for themselves, but not in a good way. It’s time to hand the Vezina voting to a group other than the league’s general managers. Will handing it over to a group like the PHWA make a difference? That’s impossible to say at the current moment, but there is no doubt that the writers are in a better position to vote for the league’s most outstanding goaltender.
At the very least, the league owes it to both the goaltenders and the legacy of the sport to at least try to fix what is very clearly a broken system.