Usually, Canada is the team to beat at the Olympics. Usually. This year, Canada enters the tournament as the co-favorite with the United States despite a markedly better forward group led by the two best players in hockey, and a comparable stable of defensemen. Team Canada also has one of the best goalies in the world as a potential starter in Logan Thompson.

The problem? There is a high likelihood that Canada makes the wrong choice in net by starting Jordan Binnington, one of the league’s worst goalies this season. 

It’s that decision alone that has massive ramifications for Canada’s quest for gold and puts the Canadians neck-and-neck with the Americans going into the tournament. What was once a somewhat marginal choice among arguably average options amid a nationwide goaltending crisis has now become a major storyline heading into the tournament. 

Since the 4 Nations Face-Off, the chasm between Binnington and Thompson has grown to an extreme point, so much so that starting the former has turned into a genuinely risky decision. Going into the first game against Group A’s top opponent, it’s a risk Canada can’t take.

The Thompson-Binnington divide

Goaltending is extremely difficult to predict; everyone in hockey knows that. When two goalies are close, the error is large enough that it truly can come down to a matter of preference. There isn’t a wrong answer between Jake Oettinger and Jeremy Swayman, for example. When the gap is wider, though, the certainty grows. No one is picking Connor Ingram to start over Connor Hellebuyck.

That brings us to Thompson and Binnington and it’s vital to view the difference between the two within a timeline of events. 

When rosters were announced for 4 Nations, Thompson did not have much body of work as a starter, while Binnington was coming off one of his strongest seasons in 2023-24. Even with Binnington’s slow start to the season, the difference in goals saved above expected was close enough at the time that opting for the veteran was completely defensible. 

By the time the gold medal game was played, the gap between the two continued to grow. There wasn’t anything to do about that with Thompson off the roster, but at the time, Binnington still looked like a decent starter. He wasn’t trending in the right direction, but he was at least average over his prior 41 games and his resume made up for it. Even if Thompson was on the roster, the small sample of his elite play would’ve still put a Binnington start on the side of defensibility. I wouldn’t have done it, but it would at least be understandable.

But a lot of hockey has been played since and the gap between the two continues to grow at a rate where it becomes harder and harder to understand how the decision could even be debatable. Thompson has maintained his level of elite play for over a year, while Binnington has continued to consistently trend in the wrong direction.

When the 4 Nations rosters were announced, Binnington had saved 0.25 goals above expected per game over his past 41 games compared to Thompson’s 0.03. By the time the final was played, that flipped to 0.57 for Thompson and minus-0.06 for Binnington — a 0.63-goal gap. Today, Thompson has stayed steady at a high-end level (0.44). During that stretch, the difference between the two is now 1.01 goals. Over the past 41 games, the difference between Thompson and Binnington is equivalent to a team starting Binnington already being down 1-0 before the game even begins.

Obviously, Canada should not make decisions based solely on a goalie’s most recent output and I genuinely don’t think Binnington is this bad. 

But even if all the numbers are regressed to reduce error — with Binnington’s past and Thompson’s small body of work prior to 2024-25 considered — it’s hard not to look at the two goalies’ peers and think this is anything but an obvious decision. At the very top of the goalie hierarchy, according to our projected output, is Ilya Sorokin (0.33 GSAx per game), Hellebuyck (0.33), Andrei Vasilevskiy (0.30) and then Thompson (0.3). To think that Canada has its own Hellebuyckian option and that it’s still up for debate is difficult to comprehend. Binnington, on the other hand, is at minus-0.11. His peers: Alex Nedeljkovic (minus-0.12) and Charlie Lindgren (minus-0.10), Thompson’s backup.

Between the two, it’s a 0.41 gap in goals allowed per game. The average leaguewide difference between a team’s starter and backup goaltender is 0.23 goals. The difference between Thompson and Binnington has simply become far too large to ignore.

The winning pedigree fallacy

There is one argument that the Binnington truthers love to use and it’s that Binnington is a clutch winner. That was proven by his Stanley Cup win in 2019 and his clutch performance in last year’s 4 Nations final. I’m not here to take either of those things away from him, but the overarching logic of the sentiment warrants scrutiny. 

Let’s suppose Binnington’s winning pedigree is innate, as is his ability to rise to the occasion. That it’s something he can summon whenever he wants and when it comes time to backstop Canada at the Olympics, he will lock in and play above the level he’s played at for the last calendar year. It could happen! 

If that were true, however, two instances from the last year immediately call that logic into question. 

The first: Game 7 against the Jets in last year’s playoffs. The Blues were up 3-1 with two minutes left in the third. They ended up losing 4-3 in overtime, which included a back-breaking game-tying goal in the dying seconds.

The second: A quarterfinal matchup for Canada against Denmark in the World Championships, where Canada had a 1-0 lead with two minutes left in the third and lost in regulation. Let me repeat this one. Binnington, in a Team Canada sweater behind one of the most stacked World Championship rosters that Canada has ever had, lost a game in regulation to Denmark in the final two minutes.

There is obviously plenty of blame to go around for both losses. But it seems like there’s a lot of cognitive dissonance when it comes to talking about Binnington’s 4 Nations final compared to his Game 7 loss and his loss to Denmark. On one side, Binnington gets a lion’s share of credit for winning behind a loaded roster of superstars. On the other side, mostly crickets. With the Denmark loss in particular, it’s hard not to blame both goals on Binnington, especially with the winner coming off incredibly poor rebound control.

If Binnington’s ability to win is so innate that no statistical argument against his actual playing ability over the last year is apparently warranted, why is it that two devastating last-minute losses in big games where Binnington was the opposite of clutch are ignored? Why does one clutch game matter, but two choke games don’t?

It’s entirely within the realm of possibility that some goalies win more than others and that they are, in fact, clutch. Binnington may even be one of them and he rises to the occasion in Milan after taking a year off from competent goaltending. 

But if we take that concept at face value, we still have one instance where that’s actually true for Binnington over the last year and two glaring incidents where it’s the exact opposite. Binnington folded at the worst possible time in high-leverage situations two times in games more recent than the 4 Nations final. If his “clutch” play in one is a point in his favor, his non-clutch play in Game 7 and against Denmark easily cancels it out.

Not only has Binnington struggled to make saves since the 4 Nations win, but he’s struggled just as much at being clutch and it’s difficult to take any argument about his “winning pedigree” seriously.

The effect on Canada’s chances

The question now is, how much does all of this actually matter? This is Canada we’re talking about; they should win regardless, right?

Yes and no. Canada should still win because they have the best group of skaters, but the margin between Canada and the USA has shrunk considerably over the last decade. That was on full display at the 4 Nations in two tightly contested games, one that the USA won and the final that Canada won in overtime. The gap between the two countries is now so tight that who starts becomes a genuinely big deal. With Thompson, Canada is the favorite to win gold. With Binnington, that honor goes to the United States.

The effect size ends up being quite substantial, as you’d expect given the current difference between Thompson and Binnington is nearly twice as large as the usual difference between an NHL starter and backup. 

From getting a bye to advancing through the knockout round and winning gold, here’s how Canada’s odds are affected.

A 15 percentage point loss in gold medal odds and a 12 percentage point loss in medal odds going from Thompson to Binnington is a massive deal. It’s the type of win probability difference that should be an absolute no-brainer.

The reason the discrepancy is so wide is that every game in a short tournament carries a lot of leverage and has implications on seeding and path. Canada is still likely to win all three of its group games, but its chances are slight enough to create more possibilities that it doesn’t get a bye and ends up with an unfavorable path in the knockout round.

That’s why Canada must make the right call between the pipes and that starts with the very first game, the highest leverage game for Canada in the group stage. Canada has a stronger team than the Czech Republic, but the Czechs are still its biggest threat to winning Group A. The Czech Republic also has the exact formula that can cause Canada headaches: a dynamic offense that doesn’t need possession to do damage and game-stealing goaltending. It’s not difficult to imagine a game where Canada loses 3-1 despite outshooting the Czech Republic 45-18 because of a masterclass from David Pastrnak and Lukas Dostal.

It’s a game that Canada wins a majority of the time, but not all of the time. With Thompson, the Canadians are heavy favorites at 80.8 percent. With Binnington, those odds drop down to 72.8 percent — the kind of odds that hockey fans have seen many underdogs pull off in the NHL every year.

Those heightened odds of Canada losing, even if the team is still heavily favored, are worrisome. Some might say that it’s no sweat, that Canada will figure it out and that it may even be a good thing if it shifts Thompson into the starter’s role. But it’s imperative that Canada gets the decision right before the game, not after. 

The big issue is what a loss to the Czech Republic does to Canada’s path. A regulation loss would mean a deep cut to Canada’s bye chances and while they would have a high likelihood of winning that game, their post-group stage path would change drastically. At the moment, Canada likely plays Switzerland in the quarterfinal and Sweden in the semifinal. A loss to the Czech Republic would shift that to a date with Finland in the quarterfinal and the USA in the semifinal. Path odds are far from set in stone, but Canada’s chances of facing Team USA before the gold medal game, jumping from 20 percent to 35 percent, would be far from ideal.

In the event that a loss to the Czech Republic happens, here’s how Canada’s tournament odds change (relative to starting the tournament with Thompson, assuming Canada likely turns to him after a Binnington Game 1 loss).

It may just be the first game and a little adversity never hurt, but a loss creates a huge ripple effect toward Canada’s path. A collision-course with the USA in the semifinal rather than the final is not a risk worth taking — not with how close the two rosters are now.

Look, I know this has a decent chance of being a Chicken Little forecast. The sky is not falling if Canada goes with Binnington and/or loses Game 1. Canada still has a high chance of winning gold, no matter who’s in net. And I do recognize that analytics aren’t perfect, even less so for goalies. It’s possible that Binnington, powered by the spite of many questioning his ability, indeed puts up the greatest goaltending performance of all time en route to gold. It’s possible that with St. Louis’ season in the tank, he’s saving himself for this exact moment. As a Canadian, I could not be happier to be more wrong if that is the case.

But as a realist, it’s impossible to ignore the mountain of evidence that puts a strong lean toward Thompson. The difference between the two has grown to a point that it’s nowhere near close enough to come down to a matter of preference. The difference is large enough that running counter to the obvious logic of who the better goalie is would be a massive unforced error. 

We’re not talking about gut feel or intuition within a marginal difference in talent. We’re talking about a difference so wide that it should be an incredibly easy choice. We’re talking about the best statistical Canadian goalie since Carey Price vs. a goalie who lost his starting job to a different Canadian goalie who was nowhere close to making this team.

That it’s even up for debate feels like an entire country being gaslit about the last year of hockey Binnington has played. That the debate is likely already closed in Binnington’s favor is even harder to fathom.

Binnington may have earned the benefit of the doubt during the 4 Nations final. But he’s lost it repeatedly for a year, ever since, with his play trending downward and multiple moments where his play melted down at the worst possible time. 

There should be no question who Canada’s starter is at the Olympics. If Canada wants its best chance to win gold, Binnington can’t be their Game 1 starter.

— Data via Evolving Hockey