I wanted to sort out how I feel about the Leafs this season vs how I think about them, and I decided to just do that in this post so you can see the process.

I've found it hard to talk about the team this year because I often find the games boring. I hate the poor quality of offence, and I don't like seeing the team with a five-on-five Corsi and Expected Goals below 50%. I don't like that the Expected Goals isn't better than the Corsi. I don't like that they're 20th at Corsi For per 60 minutes and a whopping two places higher by Expected Goals For.

See: this is not a fun conversation for anyone but the perpetually aggrieved.

I also don't like this state of affairs because the Leafs have talented forwards and this feels like they're playing like a bad team that wants to use their goalie and their power play to win close games while pretending luck isn't really the decider.

Here's why I don't like this:

  • Corsi or Expected Goals percentage correlates with future success
  • The closer a game, and the later a game stays close, the greater the chance that random events not tied in any way to skill, or lack thereof, will decide the outcome

Here's the other thing, though:

The Leafs have a Points % of .618, they have a positive goal differential, and they are in the top half of the league in total goals scored. They are better if judged by goals against, but this is a formula that gets you into the playoffs. The Atlantic is tight, with several teams good enough to pick points off the top teams, so that percentage is good for second place in the division right now.

The Leafs didn't get all those points by some fluke. It's not some guy with a wild shooting percentage or that run Martin Jones went on last year. You can point right at the tangible things that are creating the points:

  • goaltending – it's been incredibly important in getting wins, but it's hardly been unsustainable
  • power play – it was horrible, now it's not
  • goals from talented players (Oh no, runs in circles, the guys scoring goals make a lot of money!!! Who could have ever foreseen this? Surely the 3RD needs to score more!!!)

Notably absent in that list is defending. The Leafs are 20th in Expected Goals Against. The Leafs are 20th in just about everything, and they're 20th in Corsi Against too. Also absent is quality offence. The Leafs shoot from dumb places and it's annoying.

Corsi Correlates

Back to the top and this business where Corsi correlates with future success. The usual way that's stated is "correlates better" and you can look at various and sundry Expected Goals models and find some slight differences to the phenomenon, but it's largely the same effect: it correlates better.

Correlates does not mean creates. So the way you play in the first bunch of games is not creating wins later. It's actually a key component of your wins in the first part of the season. But wins, and therefore standings points, take a long time to settle down into something like a realistic picture of a team's strength. Sometimes they never do. Wins come from goals, and near the end of the season team goal differential usually sorts very closely to standings order.

That's the thing Corsi is "better than". It's been so long now since this was first analyzed and charted and shown to be true that I think a lot of people have no idea what the point of it all was. In hockey, in the NHL, goals correlate to future goals fairly poorly. Corsi is better. But poorly is not the same as unlinked. And it pays to remember that it is goals you need, and you need them in a proportion that favours you.

The Leafs are sixth in the NHL in Goals For % at five-on-five. They rank high in Goals Against per 60 at 12th, and their Goals For rate is 15th. Finally they aren't 20th. But make no mistake, most of that ranking is the goalies and a few forwards.

Lucky bounces and odd-man rushes

The closer a game, the more luck plays a role. This is one of those things that I honestly believe anyone should be able to understand by just casually watching games. No numbers. But whether it's the constant drone of narratives about effort and desire and the explanations offered for why every goal against is someone's fault, this seems to be a controversial fact. It goes against the idea that what happens in a game is by intent.

Take some luck, add in the loser point, and you can roll on up the standings by winning close games. You can just let the game come to you, take what they give, and play conservatively in the offensive zone. I swear the Leafs looked like they were trying to hold the lead in their last game when it was 1-0 Vancouver.

We get told on TV that a key thing about the Leafs this year is that rush chances against – a big problem last season – are way down. This is unverifiable because it's all proprietary information that can get doled out in tidbits that tell a story that just sits there untethered to any understandable context. I assume it's true.

Here's another thing that's true: shots that come off the rush are more likely to become goals than non-rush chances when all else is equal. So what that means is, that with the actual numbers to compute this, you could say exactly how much the Leafs benefit from low rush chances against.

You can also just look at Expected Goals which has this baked right in in various ways, though not always in the same way in each model. Every Expected Goals model holds to two core realities, and that's the fact that number of shots (Fenwick is the input to these models, not Corsi) and the distance of the shot from the net have the most impact on the probability of a shot being a goal.

I think this rush chance against improvement is helpful for sure, but to claim it mitigates the shot volume allowed ... I'm sorry, but no. It's goaltending.

Risk

The way the Leafs play looks conservative. It looks like it's mitigating risk. And you get comments from the coach when mistakes are made about those skilled players wanting to make plays for the home crowd. And that sounds like if only three of the five players tether themselves to the blueline in the offensive zone, and no one shoots from close in, then there'll never be anything for the Goalies to do.

The conservative defence-first approach might help. And it lowers the risk of goals against. I think its effect has been dramatically overstated, however, because the Leafs allow a lot of shots.

But what about the hidden risk of this system? If you take the puck away from the most talented players for much of the game and rely on the goalie and the hands of fate to get you wins in low-scoring games, you've intentionally left unused your best and most reliable weapons. Why do this? Because the risk of a butt goal, or a bout of puck-watching by the Leafs that lets a goal go in even Stolarz/Woll can't save is sitting right there, obscured by the feeling of safety of a dull and boring system.

That's how it feels watching. But the Leafs are not the Devils of old or the Islanders. They don't actually nerf their top players all the time. The Leafs are two teams – a bottom six that can't score, but can actually limit shots against and a top six that's flipped the other way. And when holding the other team's 1-0 lead stops working, you will see some lines of skilled players go over the boards.

But why play this system designed for a team with no good shooters as much as they do? Is the illusion of safety that comforting? Safety can also come from dominating at five-on-five and having a solid foundation that withstands goalie injuries, scoring slumps, power-play troubles or having a weak third line. What is the allure of this trying to win by not trying too hard to score?

But the playoffs, though... Ah, yes, the idea that the playoffs need some completely different game.

When is this future success happening?

Before I started writing this I downloaded all seasons of Evolving Hockey's team summaries for the regular season and the playoffs. I glued them together and had a look at how Corsi or Expected Goals correlates to playoff success.

First, it's been understood for years the in-season Corsi % correlates very heavily to which teams make the playoffs and which don't. This isn't in question and all arguments against the idea are false. Sitting at 20th in the NHL in that measure is a concern for a very good reason. Sure there are exceptions, but there aren't many.

Move into the playoffs itself and consider this: do you expect in-season Corsi % to correlate with how deep a team goes?

It does not. Neither does in-season Expected Goals. And this is another of those things that is so mythologized that a thing you should be able to just know is impossible to convince nearly anyone of. The playoffs are short stretches of games. There's so much more random chance in the outcome of the playoffs than in the regular season, it's hard to overstate it.

The most difficult subjects can be explained to the most slow-witted man if he has not formed any idea of them already; but the simplest thing cannot be made clear to the most intelligent man if he is firmly persuaded that he knows already, without a shadow of doubt, what is laid before him.
Leo Tolstoy quote that opens The Big Short by Michael Lewis

Playoff series are also not enough games to see a correlation between in-playoffs Corsi or Expected Goals and wins (or games played). If you look at the history, you will see you have to be at least competent, so no laying down 45% on the regular, but you don't have to be excellent. You're more likely to go deep into the first-round series if you're well over 50%, and more champions were very good than just okay, but it's a weak indicator. And I bet you can guess why.

Goalies

They aren't like other people, and their results in hockey are not like shot statistics. Looking at Goals Saved over Expected is what I do to see who has good results. But it's not really much more useful than Save % – a nearly useless stat. It cuts out some of the team effects, but it won't tell you what that goalie will do tomorrow or next week or in the playoffs.

There are so very few goalies that have the consistency to be relied upon for sure, that they are extremely expensive. A team with one of them has to cut the skater budget by a lot, and you can't just nip down to the store and buy one. There's also a lot of high-priced knock-offs out there that can fool almost anyone for a year or two. Jack! Sammy!

The smart move is therefore to have many inexpensive goalies you can rotate through your lineup and the proponents of this idea despair of the stupid GMs who just won't do it. These proponents also don't seem to know what the salary cap, waivers or roster limit is.

Coming as close as you can to a Club Pack of minimum salary netminders is what that dummy Treliving has done this season. And it's working – luck plays a big role here because Stolarz could as easily been a one-season wonder as not. But it's worked, and while I don't believe for one second in Dennis Hildeby, having Matt Murray available was a zero-risk move that's paid off.

Acceptance

After all of that, the Leafs are the Leafs are the Leafs. And they are winning at five-on-five. It's actually a higher-risk way to do things than a nice solid puck possession style that produces quality offence and good enough defence. And it's not like the Leafs don't do that. They just don't do both on the same shift.

Go here to make your own: https://moneypuck.com/t.htm?t=TOR

Two teams. Upper Right team is all offence, and, eh, kinda horrible defending. Which they've always been. These leopards did not change spots. Notice the few that are at least okay defensively - the top line.

Lower Left team is all defence, and for all of that, if you watch the game, those guys are in the offensive zone a lot – holding onto the puck to give it to the leopards.

Simon Benoit plays for some other team.

The result is poor offence on average because Lower Left don't shoot much (and therefore don't give the puck away to the wrong people) and poor defence on average because of the lack of spot-changing by the leopards. Auston Matthews, William Nylander and Mitch Marner are all shooting at their normal recent-season pace. The all have totally normal individual Expected Goals numbers. They aren't low event. John Tavares is shooting less, but – that's a longer story about how his role wasn't working, so it's not a worry.

No Worries

This is fine, totally fine. Or maybe I'm rationalizing a bit. But there's obvious ways to make the team better. The third line could stand to be something other than neither bad nor good at anything. If a defenceman who can genuinely add value can be found for limited investment that would be good. But he need to be someone who functions well when the forwards for half the game are kinda clueless in their own end. Just big isn't enough.

The truth is, you can't count on any system to get you to the cup in the playoffs. You can't be bad, you have to be good at more than one thing, but there is no perfect formula that will guarantee you the right outcome in a highly randomized contest. That's how it is.

Go Upper Right Go!