Montreal’s loss in the first game of the Eastern Conference final was made all the more awful by Carey Price’s injury. It’s one thing to get blown out—1-0 or 15-0, you’re still only down a game in the series—but losing your No. 1 goaltender, who posted a phenomenal .927 save percentage in the regular season and a very good .919 in the playoffs, that’s a death blow, right?
Well… sort of. It hurts, but it wasn’t quite the death blow to Montreal’s playoff run that history will likely make it out to be. As it so happens, math (and reasonable assumptions) can help us estimate of the difference between the Canadiens with Dustin Tokarski, the goalie charged with saving Montreal’s season, and the Canadiens with Carey Price.
The first challenge is coming up with an estimated save percentage for Tokarski, who is 24 years old and hasn’t really had much of an NHL career. He’s played 451 minutes in which he’s faced 225 shots and has allowed 22 goals, a .902 save percentage. That’s not much to draw any conclusions from—seven and a half games tells us virtually nothing about a goaltender. Brian Boucher, who had a career save percentage of .904, played 332 minutes of shutout hockey once. Crazy things happen in small samples.
What if we look at Tokarski’s time in the American Hockey League? He’s played a lot more there, 13,685 minutes, and has allowed 578 goals on 6,524 shots, a save percentage of .912. That tells us a little bit more. It’s reasonable to think that it’s harder to stop shots in the NHL than it is in the AHL. Stephan Cooper found that the average goalie’s save percentage was about eight points worse in the NHL. We can apply that here and estimate Tokarski would put up a .904 if he was a full-time NHLer.
If the Habs can expect to get a .904 from Tokarski, what could they expect to have received from Carey Price? As mentioned above, his save percentage in the regular season was .927 this year. We know that goalie save percentages fluctuate, which makes using a bigger sample more sensible. Price has posted a .918 save percentage over the past three seasons, so for the sake of discussion, we’ll use that.
The question then becomes: How does the Canadiens going from an expected .918 to an expected .904 change their outlook of winning the series? We can reason this through. The Habs scored 2.53 goals per 60 minutes played this season. Neither Price nor Tokarski provide any offence, so substituting one for the other presumably makes no difference offensively. If you take into account Montreal’s shots allowed per 60 minutes this season (30.6), we see the difference with Tokarski and Price. With the .904 goaltending, the Habs would allow 2.94 goals per 60 minutes. With .918 goaltending, we’d expect Montreal to give up just 2.51 goals per 60 minutes.
Obviously, that’s a big difference. But what can it tell us about the probability of the Habs winning? One of the things that hockey analytics types have found is that if we know how many goals a team scores and allows on average, we can estimate how often a given score will come up. If you know that, you can estimate how often a team will win given its ability to score and prevent goals.
Applying that to this situation, you come up with a 50.3 percent chance of the Canadiens winning a game with Price in net versus a 43.2 percent chance with Tokarski. It’s important to keep in mind that this is a pretty simple model—it doesn’t take into account that the Rangers are stronger than the average team or the location of the games. Down 1-0 in the series, the Habs had to win four of their next six games. Montreal’s chances of doing that fell from about 38 percent with Price to 24 percent with Tokarski. It’s a big drop, although nowhere near impossible to overcome.
Losing the second game doesn’t help things. By my math, we’d expect Montreal with Tokarski to win four out of five games about 9 percent of the time. I note that the betting markets are much less pessimistic about Montreal’s chances—my math says that it’s an 11/1 or 12/1 shot while the markets seem to have settled at about 6/1. It’s always worth paying attention to the people backing their opinions with money. My estimate is based on certain assumptions—everyone who is betting is going through some process to determine whether they think a bet is worth a certain price. I’ve got a hard time seeing how Montreal’s odds could be that good but I’m always leery of going against the market.
If you’re a Habs fan looking for hope, the fact that the betting markets give Montreal a 14 percent chance of pulling it off beats anything I have to offer. Just try not to think about the possibility that diehard Habs fans are still dumping money on Montreal regardless of logic.