The Maple Leafs have played three games vs Alberta so far and have three more to come before a few days off at the start of February. With some troubles so far this season with offence, defence, and injuries, the timing of a couple of hot goalie performances couldn’t have been better. Even better, the Leafs didn’t waste these gifts, and the goalies brought the team four points this week.

Wednesday Oilers Game

The first Oilers game on January 20 was called a defensive duel by some, but I just called it dull. With five days of sober reflection, I’d say it was really dull.

The story of that game was shot-blocking by the Oilers as they turned a 56% Corsi for the Leafs into a 44% share of unblocked shots which dwindled down to a very, very alarming 40% Expected Goals share. (All numerical information is from Evolving Hockey.)

If that was a defensive duel, the Oilers won it. They won the game too.

Frederik Andersen was in net, and he let in two goals, which is hardly an outrage. His Expected Goals Against was two, so he just was an average goalie. He only saw 22 shots on goal against on 30 unblocked shots, his lowest this season.

When a goalie faces fewer shots, randomness plays a bigger role. This is more true over a whole season, so a team that limits shots against is counting on their goalie to be good enough to not be too hurt by the random bounces that can ruin your game, but low-event hockey favours luck over skill. One unusual event, like an own-goal, and you look average with very few saves to make to reset your numbers.

The Leafs failed to capitalize on their own defence in that game, but their goalie gave them a chance. The result was a 3-1 win for the Oilers.

Friday Oilers Game

In the next game against Edmonton, the situation seems no different at first look. Andersen faced 23 shots on goal on 31 unblocked shots, so déjà vu all over again there. He let in two goals, on an expected of 1.84, so he was actually a tiny bit worse than average. And yet the perception from all over was this:

The game at five-on-five was 57% Corsi for the Leafs turned into 54% Fenwick and 48% Expected Goals. The Leafs offence fired much better, and the Oilers either didn’t or couldn’t block as many shots. By the numbers this was a game won by the offence with average goaltending.

But the numbers are far from perfect in a single game, and shot rates only tell some of the story. Expected Goals weighting helps some, but not enough to really know why of two nearly identical goaltending performances, one felt okay to the viewers, and the other excellent.

In the Wednesday game, Connor McDavid had three shots at even strength and none of them were great ones. He was also as good as it got. All the Oilers offensive oomph came on the power play which has some gigantic flaws to it this year, and it just never looks that hard to defend.

In the Friday game, McDavid got loose from the tight checking that had worked in the first game. He had six shots at even strength, and the other Oilers shooters were good scoring forwards, not their depth. The power play was still bland for them, but overall they looked dangerous offensively.

That’s how you get identical all-situations shots and goalie results where they felt vastly different in quality.

The Leafs won the Friday game 4-2 in part because Andersen gave them a good enough baseline to win from. There were a lot of other good performances in the game from the skaters, as well, but they needed that goaltending to put them over the threshold to a win.

Sunday’s Flames Game

Jack Campbell got the start against the Flames, and he was on fire! That was a genuine stellar performance, and if the Leafs had blown that, even Jack might have frowned briefly.

The Leafs, still struggling to really own the ice at five-on-five, had 43% Corsi, which increased to a 49% share of unblocked shots and a basically static 48% share of Expected Goals.

The big change in Corsi to unblocked shots from Leafs standard (this is the only game so far where there was a meaningful rise) is worth an investigation. There were only two shots blocked by the Flames, and the Leafs blocked 12. Blocking is a combination of willingness and opportunity, and the truth is, the Leafs shot a fair bit from the points and from the circles, and that’s the sort of shot that is usually blocked.

Meanwhile, at the other end of the rink, that Leafs blocking helped, but that’s about all they did to give Campbell a hand in his second start of the season.

The Leafs allowed 33 shots on goal on 48 unblocked shots. If the Flames had missed a little less, this might have ended differently. One thing to consider about those 48 shots,  not all of which Campbell had to save, is that he doesn’t actually know that when the shot is released. He has to be ready to save the ones that whizz on by. He worked hard in that game.

Campbell allowed two goals against on 3.87 Expected Goals for a performance that counts as Excellent. Possibly a game-stealing performance, even. Put that with his other good game, and he’s a top-10 goalie in the NHL in the early going (by expected goals measures, you’re on your own if you want to look at traditional save percentage). So is Brian Elliott and Marc-Andre Fleury, in case you need a reality check.

The Leafs won 3-2 by letting Campbell handle the Flames while they produced almost as much five-on-five offence as they did against Winnipeg, their only dominating-esque performance to date.

This wasn’t a win the Leafs “deserved” by the numbers. It was one Campbell made possible. When a goalie does that for you, you had better step up and win it.

The Leafs play the Flames again tomorrow, and we’ll see if the Flames change their approach or if the Leafs find a little tiny bit of the defending they managed against the Oilers.

Analysis