𝓤𝓷𝓲𝓽𝓮𝓭 𝓝𝓮𝔀𝓼

Two years ago, the Vegas Golden Knights believed they had found their franchise goalie.

Adin Hill backstopped the team to a Stanley Cup in 2023 and followed it up with a strong 2023-24 season. Logan Thompson, once in the mix, was traded away that summer.

Now Vegas probably wishes it could hit the undo button.

Hill has been one of the worst goalies in the NHL (.869 save percentage), leaving Vegas searching for answers in net — just one year into his mega six-year contract. Meanwhile, Thompson has emerged as one of the league’s best starters since joining the Washington Capitals.

Stories like this are why some people believe that goaltending is becoming increasingly unpredictable. Every season, some highly paid starters like Hill, Sergei Bobrovsky or Linus Ullmark mysteriously struggle, while unknowns or journeymen like Alex Lyon, Brandon Bussi and Scott Wedgewood emerge to carry their teams. Only a few netminders, like Connor Hellebuyck, Andrei Vasilevskiy and Igor Shesterkin, are consistently elite.

NHL goaltending is quickly becoming one of the most volatile positions in all major sports. Here’s some evidence to support that belief, along with possible explanations for why.

---

If an NHL team spends $6 million on a forward, you’d generally expect that player to be more productive and valuable than a forward making $2 million. That isn’t always the case, but on average, in an efficient market, more money should buy better outcomes.

However, in goaltending, that link between price and performance is broken.

Below is a table showing how many cap dollars each NHL team is spending on its goaltending tandem and the results they’re posting. Traditionally, save percentage has been used to measure goalie performance. The problem with save percentage is it treats a long-distance point shot the same as a breakaway. It doesn’t account for shot quality and defensive environment, which can heavily skew a goalie’s save percentage for better or worse. To account for shot quality, we’ll use a metric called goals saved above expected (GSAx).

GSAx analyzes the quality of shots and scoring chances each team has surrendered. It then calculates how many goals a league-average netminder would have allowed based on that team’s shots and chances allowed, and compares that to how many goals a team’s goaltender has actually given up.

For example, Ilya Sorokin, the front-runner for the Vezina Trophy this year, has a plus-29.9 GSAx, meaning Sorokin has allowed roughly 30 goals fewer compared to a league-average goalie plugged into the same defensive environment. On the other hand, Jordan Binnington has a league-worst minus-25.4 GSAx, meaning he’s cost his team about 25 goals compared to an NHL-average goalie.

Interestingly, there is zero statistical correlation between how much each team spends on its goaltending tandem this season and the club’s overall goaltending results, whether you look at GSAx or even a simpler metric like save percentage.

Data from March 16

We have cases where budget goalie teams like the Pittsburgh Penguins, Buffalo Sabres and Carolina Hurricanes are getting significantly better netminding results than big spenders like the St. Louis Blues, Ottawa Senators and Vancouver Canucks. In fact, the 10 highest-spending teams are getting a combined plus-15.1 GSAx from their netminders, whereas the bottom-10 spending teams are receiving a combined plus-57.2 GSAx.

If you look at the NHL’s nine worst goalies in terms of GSAx, seven of them have a cap hit of at least $4.5 million. This includes some established names and Stanley Cup winners, including Ullmark, Bobrovsky, Binnington, Jacob Markstrom, Hill, Tristan Jarry and Kevin Lankinen.

If you zoom out and look at the big picture, there are a staggering number of bad or risky goalie contracts teams have signed over the past few seasons.

Top goalies aren’t stealing playoff series as often as they used to

Some top goalies are also struggling to live up to expectations in the playoffs more often than they used to. Last year, for example, the three Vezina Trophy finalists — Hellebuyck (.866 save percentage), Vasilevskiy (.872), and Darcy Kuemper (.889) all posted very disappointing postseason numbers.

To see if this was part of a bigger trend, I looked at every playoff series since 2020. I identified which team had the superior goaltender on paper before the series started (based on their regular-season GSAx rating) and then tracked whether that team won the series.

For example, when the Florida Panthers beat the Edmonton Oilers in the Stanley Cup Final, I tracked that as a series where the team with the superior goaltender on paper won because Bobrovsky entered the series with better regular-season numbers than Stuart Skinner. Conversely, when the Dallas Stars lost to the Oilers in the Western Conference, I counted it as a series where the team with the better goaltender on paper (Jake Oettinger) lost to a weaker one (Skinner).

From 2020 to 2022, the team with the superior goaltender on paper won 33 of 45 playoff series, which is a 73 percent win rate. Having the more established netminder made a team more likely to win the series during those years.

That has flipped lately, though.

From 2023 to 2025, the team with the more established goaltender won only 17 of 45 playoff series, a 38 percent win rate. For every situation where a star goalie like Bobrovsky is winning over a weaker one in Skinner, there are actually more instances recently where the Hellebuyck types underperform and lose, or where a Skinner gets hot for a series and outplays an Oettinger.

Maybe this is just a blip on the radar that will normalize in future years. But it’s also possible we’re entering a new golden age of offense and scoring, where the quality of a team’s skaters and their two-way play, both defensively and offensively, is a more significant driver of wins and losses than the quality of a team’s goaltender. This isn’t to say “goaltending doesn’t matter” because that would be an extremely foolish take, but it is fair to wonder if the position has less of an influence on winning compared to 10-15 years ago because of how advanced offenses are becoming.

NHL offenses have found a new level, and goalies haven’t caught up yet

In 2015-16, the average save percentage for a goaltender was .915. Fast-forward a decade later, and the average save percentage has collapsed down to .896, which is the lowest mark since 1993-94, when the butterfly goaltending style was still in its infancy stage. The NHL has recently changed how it tracks shots on goal — goalies no longer get credited with a save for stopping borderline shots if the puck’s trajectory was estimated to miss the net — but that’s only part of the drop in save percentage.

There are two possible explanations for this dramatic decline: Offenses have evolved to become significantly more dangerous, and/or goalies have become worse. The first part of that equation is quite obvious to the eye. Players are faster, more skilled and better at shooting than ever before. There’s also been a massive tactical shift — with analytics growing in popularity, teams have a far better understanding of what types of scoring chances goalies struggle with, and have optimized their offensive strategy accordingly.

“It’s as challenging as it’s ever been in my career,” said Canucks goaltender Kevin Lankinen. “I’ve only been in the league for six years, but I see the (offensive) development and where it’s going… The game has changed where there’s not many teams who just send (low-danger shots) toward the net. Teams are doing more with the advanced stats and (targeting) what are the high-quality scoring chances. There’s a lot more east-west (movement) than ever before.”

The stats back this up. According to former NHL goalie Steve Valiquette, who founded Clear Sight Analytics, the number of unscreened shots from long range — which are very easy shots for goalies to save — declined nearly 28 percent between 2018 and 2024. Meanwhile, the number of shots right after plays or passes that crossed the middle of the ice — which is the challenging “east-west” movement Lankinen was referring to — jumped by 41 percent between 2018 and 2023.

“One of the reasons that I’ve always said that if I was a general manager I would try my best to avoid term when it came to goaltending is because we’ve literally seen guys sign contracts and the strength of their game when they signed a big deal no longer works, because the NHL attack has changed so much by the time the contract expires,” said Kevin Woodley, a goaltending expert for InGoal Magazine, which has relationships with goalies and teams all around the league.

The rapid growth in offensive chances originating from east-west passes across the slot is a meaningful development because it’s arguably the most dangerous type of scoring chance in hockey, with shooting percentage on those plays estimated at a sky-high 30 percent. These east-west plays are so difficult to stop because goalies have to move from side to side and often don’t have a chance to get set and squared before the shot is taken.

The publicly available GSAx models, which are designed to account for the defensive environment in front of each goalie, lack the data needed to measure east-west passes. In other words, even the goalie statistics we referenced earlier in the story aren’t perfect measures for netminder performance.

Overall, NHL goalie performance is becoming increasingly tied to how effectively the team in front can limit deadly scoring chances that netminders struggle with.

“I had a goalie coach tell me early this year that the position and performance have never been more closely tied to that (defensive) environment,” said Woodley. “I’m not saying goaltending doesn’t matter, but the amount of goalies that can consistently outplay bad environments (is low) … because when guys create the types of scoring chances they’re creating more regularly now, a lot of them are no-chancers.”

Connor Hellebuyck’s results this year exemplify that. Statistically, Hellebuyck, the NHL’s reigning MVP, is having a middling season relative to his elite standards. His .897 save percentage is a sharp drop-off compared to last year’s .925, and his public GSAx is only slightly above average.

In reality, though, this “volatility” in Hellebuyck’s numbers this year has far more to do with the Jets’ sinking defense than with his individual performance.

“Do you know that before he went for knee surgery (in late November), he was actually playing better by Clear Sight’s numbers than he was last year when he won the Vezina and Hart?” said Woodley. “But nobody could see it because the defensive environment had totally cratered.”

A generational loss of workhorse starting goalies

While some of the NHL’s goaltending instability can be attributed to evolving offensive play, the pool of elite starting goaltenders also isn’t as deep as it once was. Woodley pointed out that the league “lost an entire generation of workhorse No. 1s in a pretty short window” and that the next wave hasn’t been strong enough.

“I definitely think it’s (a lack of) talent,” agreed one NHL executive. “You can even just isolate Canada as the whole cause of the problem. Look at this list of goalies you named: (Marc-André) Fleury, (Roberto) Luongo, (Carey) Price, (Martin) Brodeur, all Canadians. Imagine if you added four new, elite Canadian goalies to the NHL right now. You’d be talking about how amazing the league’s goaltending is.”

Why fear of the unknown causes teams to sign bad contracts

Earlier in this piece, we showed several teams have recently signed regrettable contracts. Part of that relates to the factors we’ve already discussed: offenses have evolved to a new level, a team’s defensive volatility can dramatically affect a goalie’s performance, and after most of the top goalies of the 2010s retired in a short time frame, the next wave of starters hasn’t been strong enough.

However, there’s also a huge psychological component. Because goaltending can be such an unstable position, once teams find a goalie who is performing well and appears to fit their system and coach, they want to keep what’s already working, because the alternative is often having to bet on an unproven goalie from the minors or a second-tier goalie from the increasingly weaker free-agent market. The latter feels like a really risky, unknown decision, even though it’s sometimes the correct one.

“If you look at how goalies are signed, how contracts happen, almost every team is like, ‘I don’t want to sign my goalie to this contract but he put up a great year, here are the comps,’” explained the NHL executive. “My options are either this guy or to not have a (reliable) goalie. I’m not surprised that everybody’s making bad decisions.”

Vancouver’s decision to extend Lankinen last year is a prime example.

The Canucks originally signed Lankinen to a cheap, one-year deal during 2024 training camp because of Thatcher Demko’s extended injury absence. Lankinen was lights-out as the club’s starter, and arguably their second most important player behind Quinn Hughes in the first half of the 2024-25 season. He was on an expiring contract, and the Canucks had to decide whether to sign him to a lucrative extension or let him go.

The Canucks knew they needed a goalie capable of handling a significant workload, and perhaps even a No. 1 role again given the uncertainty around Demko’s health. They also knew the 2025-26 season was one of the most important in franchise history because they needed to have a strong year to keep their chances of re-signing Hughes alive. Vancouver’s alternatives to re-signing Lankinen were massive question marks: Arturs Silovs’ NHL performance that year was so brutal (.861 save percentage in 10 games) that it seemed reckless to believe he was ready for NHL starting goalie duties — especially since this was before Silovs went on his stellar AHL playoff run — and the upcoming free-agent goalie market didn’t look great either. With all that in mind, the Canucks signed Lankinen to a splashy but seemingly justifiable five-year extension at a $4.5 million cap hit.

This year, Lankinen’s numbers have collapsed amid a significantly worse defensive environment, and the contract is aging poorly.

---

It’s easy for outsiders who don’t have skin in the game to say that teams should be bolder, more unconventional and more willing to take risks in goaltending decisions. In reality, it’s far harder to follow through on that as a GM when you know that a gamble gone wrong in net could jeopardize your team’s season and even your job.

This is why, as much as the league has evolved in its understanding of goaltending technique, workload, biomechanics and development, we’re likely to see teams continue to make mistakes when they’re forced to bet on this increasingly unpredictable position.

“That’s the biggest problem with goalies: What are you supposed to do?” said the NHL executive. “Everybody is a critic of goaltending, but very seldom are people (like) ‘Here’s what you should have done right.’”

— All GSAx data courtesy of Hockeystats.com