Last week in Trending Topics, we had a nice long chat about how the New Jersey Devils’ success is unsustainable given the quality and depth of that roster and the general high percentages they’ve gotten in all situations.
There was a little bit of pushback on the idea, which is to be expected, but the general consensus seemed to be, “Yeah, we all understand the team isn’t this good.”
Ditto, then, for the Golden Knights. But you knew that, too.
Just like the Devils, everyone knew a month ago that this wasn’t a team built to compete long-term in the NHL, and just like the Devils, they’re doing it short-term through a crazy percentage run.
Currently, Vegas is sitting on a 103.7 PDO at full strength, third-highest in the league ahead of Thursday’s games (though they didn’t actually play on Thursday themselves) and propped up mostly by an unsustainable .941 5-on-5 save percentage. In all situations, the number is 103.4, despite a poor power play, which shows how far the goaltending so far has gotten them.
That’s what makes the Golden Knights different from the Devils: Yeah, New Jersey has been getting good goaltending, but that’s probably something you’d expect given that it had mostly been Cory Schneider between the pipes (until he dropped off more recently). Schneider had a bad season in 2016-17, no doubt, but his track record shows he’s closer to a .920 goaltender than the guy who’s stopping something like 91 percent of the shots he faces. Obviously Keith Kinkaid, meanwhile, isn’t going to be .930 forever, but he too has a pretty good track record overall.
This is something you absolutely cannot say for the Golden Knights. Marc-Andre Fleury? That’s an NHL goaltender, for sure. He’s a bit overrated and that’s not his fault, but you have to say he’s at least roughly average when things go well. Everyone else? Not so much. They’ve already rostered four other goalies in less than three weeks, from Calvin Pickard (since traded to Toronto) to Malcolm Subban to Oscar Dansk to Maxime Lagace. Once Fleury got hurt — after a hot-but-not-unbelievable start of .925 — you had to be seriously concerned about how that goaltending would hold up.
Pickard is the one other guy for whom it’s fair to say he’s an NHLer, at least as a backup. He’s played 86 games over the last three seasons (and actually got into 50 last year when the Avs ran into injury trouble) and he’s .914 in those appearances. That’s perfectly fine for just about any goaltender, but also: He’s not on the team anymore.
So that leaves Vegas with a trio of Subban (now also hurt), Dansk, and as an absolute last-resort fallback position Lagace. We have very little evidence that any of those three guys are even close to being NHLers, despite their pedigrees as such. Subban was selected No. 24 overall in 2012 and Dansk went seven picks later. In theory these are two guys who can play. In actual practice, the evidence might not be there.
Subban has been a professional hockey player for the previous four seasons and was a little above average at the AHL level (career .918 on about 4,000 shots between the regular season and playoffs). Dansk spent the last few years bouncing around between the ECHL, AHL, and Swedish league. He’s a career .877 in the American League and .907 in Sweden’s first division. Lagace hasn’t gotten into a game, and for good reason: He’s a career .895 AHL goalie whose ECHL stats are also bad. The fact that he’s even on an NHL roster right now speaks to the Knights’ paucity of options.
Suffice it to say, the numbers for Subban and Dansk do not translate to NHL save percentages of .936 and .929, respectively, at the NHL level. But here we are.
One interesting thing, though, is that while Vegas gets outshot (badly) they also give up more shots against per hour of 5-on-5 time than almost every team in the league. What that basically means is that they’re keeping their goalies very busy and also relying on them to make a ton of saves. The Vegas defense is porous, and not only that, but they’re bottom-10 in the league in both medium- and high-danger scoring chances against. The idea that you’d expect any goalies — and we’re even talking Henrik-Lundqvist-in-his-prime — to keep stopping 94 percent of the shots they face, given that quality, over any reasonably long period of time is, flatly, bonkers. It does help, however, that Vegas takes relatively few penalties per game, so the goalies have less PK time to deal with.
On the whole you’ve got one goalie who’s clearly league-average in Fleury, two guys who might be replacement-level in certain instances, and one guy who’s not AHL-replacement-level. And yet Vegas is outperforming its all-situations xGA/60 by half a goal. It’s not, honestly, as bad as you might think, but that’s a number that has to come back to earth nonetheless. Especially because xGA assumes league-average goaltending, and taken as a group, this is a team that should reasonably expect sub-average goaltending, especially behind that particular defensive group.
The question, then, is whether the offense can keep this up. After all, 27 goals in eight games isn’t exactly a scoring assault. It is, in point of fact, right around league average. But that’s roughly how many they “should have” scored in all situations. And this is a team that’s not very good at 5-on-5. You therefore can’t expect them to keep things up on the power play for one simple reason: They don’t have the weapons.
Obviously if you look at the on-paper talent the team carries with it, this isn’t something that’s going to last, because All-Star teams would have trouble shooting 11 percent for a full season against modern goaltending. They have eight guys with at least two goals right now, and James Neal has six already. Good on them for generating decent looks so far, but it’s a long season.
We can reasonably expect the goaltending to take a step back for Vegas, and the reasons why are obvious. The thing to watch, then, will be how much longer the offense is getting the kind of looks that lead to a roughly average offense, just given what they have on hand. Let’s put it this way: They might have one bona fide first-line forward (Neal, and that’s if you’re being a little generous), a handful of second-line guys, and then a bunch of fourth-liners of various age and potential. Plus, in the modern NHL, you need defenders to get involved with the offense, and right now their top-two in terms of TOI is Nate Schmidt and Luca Sbisa.
There’s no questioning that Vegas is off to a weird start. The Golden Knights’ share of expected goals comes in at almost 55 percent in all situations. That’s significantly out of whack with their other underlying numbers (corsi, fenwick, SOG), all of which are below 50. You basically never see teams’ xGF% get that far separated from the underlyings. That’s true even in small samples. In big ones? It just doesn’t happen unless you have elite rosters and spend a lot of time leading in your games. So far, Vegas has certainly checked off that second box, but it’ll be a few years at least before it can check off the first.
Until then, this is a nice little storybook start, but let’s not make it out to be bigger than it is.
All statistics via Corsica unless otherwise noted.