The Discourse: The Dallas Stars Doomer edition (and a measured response to it)

Losses to Pittsburgh and Minnesota. It wasn't the best weekend.

There are many different ways to interpret a loss. Especially in hockey, where a goalie can let in a few softballs. The schedule might factor into the outcome too, as teams tend to be worse on the second half of a back-to-back. Maybe a team gets “goalied.” Sometimes the hockey gods aren’t on your side. Sometimes the officials play an outsized role. Broken plays, lucky shots, bad decisions, etc. The list goes on.

Why not the same for victories? Why don’t we pick them apart the same way we do losses?

I ask this because it feels like the heart of why the discourse feels so heated following Dallas’ losses to Pittsburgh and Minnesota. For some fans, the outcomes are manifest destiny. If Dallas won, whether they deserved to win or not, the stars were aligned. “Two points is two points.” If they lost, well, hey — they’re missing two key players. There are good excuses for why Dallas lost this weekend. But there’s no such thing as a good reason for why victory shouldn’t be earned. In this worldview, wins always tell the truth. Losses, however, can lie. But isn’t that just a double standard?

That’s the case for doomerism: the wins are deceiving while the losses have been telling the truth. Over the last 10 games, Dallas has been outshot 251-358. Only four of those teams were playoff caliber, and of those four, only one had a fully healthy lineup¹.

Defense wins championships, right? Because Dallas has aspirations of winning a Cup, I figured it might behoove us to compare Dallas’ shot quality — or expected goals against per hour of even-strength play — allowed (without Heiskanen) versus the last 10 Cup winners. Here’s the quick and dirty breakdown.

I have to admit I was a little surprised. Turns out, four of the last six Cup winners were outside of the top 10 in shot quality allowed. On the surface, this doesn’t seem like such bad news for the Stars. Perhaps it even raises questions about how much defense actually wins championships. It’s not a good look, but it doesn’t seem fatal, right?

Well, two things: 1) it would be the lowest mark of shot quality allowed by a serious team in the last ten years and 2) none of the last six Cup winners were outside the top ten in EV expected goals against. Given that this coincides with the recent scoring boom, this feels a little extra relevant.

So this is the crux of doomerism. It’s not that Stars fans are rooting against the Stars². It’s not that Dallas’ fate is sealed if they have to play Colorado in round one. And it’s not that they’re a bad team. It’s that, without Heiskanen, they’re bringing with them the kind of defense that hasn’t won championships, at least lately. That doesn’t guarantee they can’t, but it seems like kind of a big deal.

But doomerism doesn’t tell the whole story.

What doomerism gets wrong

As much as I love watching the spreadsheets, it’s easy for fans who aren’t used to what the numbers mean to look at things like the Deserve to Win-O-Meter and stats with words like “expected” in the title, and confuse predictive with prophetic.

But that’s not what fancy stats are. They’re not prophecy. They’re language. In the case of the numbers above, they’re a language to separate team defense from goaltending; and within that they try to separate outcomes from performance in some ways. The other piece is this: the negative connotation to the idea that outscoring your problems is bad.

Why? Who cares how many problems you’re outscoring if you’re outscoring your opponent? Isn’t that kind of the point? Looking at goals above replacement — explained here — Dallas has the second-most potent offense of any Cup winner since the Avalanche.

However pessimistic you’re feeling about Dallas, they have the most important ingredient in the fight against “regression”: chaos. Talented shooters and talented netminding make for a chaotic mix. Like what we’ve seen in Dallas’ recent stretch, even in games where they’ve allowed way too much zone time, and been outshot, their opponents have had to go above and beyond. The amount of Dallas’ multigoal losses, you can count on one hand all year: really just Colorado and Winnipeg. Columbus and Pittsburgh required empty netters. That’s kind of insane, and a testament to Dallas’ bend-not-break ability.

The other piece that doomerism gets wrong is that some assign a higher significance to the value of certain stats than they deserve because it’s all “black box” math. The fanciness comes from the models, and the math itself. Figuring out how to assign a shift-to-shift value for a player takes smart people doing smart work. Figuring out a team’s intrinsic value rather than the value we get from outcomes (i.e. wins and losses) also comes from smart people doing smart work.

However, the application of this fancy math comes from human decision-making. Above I haven’t done any real analysis here. All that’s being said here is that Dallas is leaking high quality shots without Heiskanen. That’s it. The doomerism line of thought or implication that “Dallas will be upset in the first round” is a totally different sentence. That would require new answers to new questions.

• Can goaltending bail out the lack of team defense?

• Is the lack of team defense sustainable?

• What’s the threshold for expected goals becoming observed goals?

• Who is Dallas’ first round opponent?

• How does Dallas’ offense line up against their opponent’s defense, and vice versa?

• What’s the special teams matchup?

• What can we expect from the goalie battle?

Do I think Dallas has a serious problem? Yes. There’s no need to sugarcoat what’s going on defensively with the Stars. But we’re getting ahead of ourselves when it comes to matchups. Not only are we still a long way from knowing who that will be, but we’re only a short way from seeing Dallas get better in small (the return of Tyler Seguin) and significant ways (the return of Heiskanen).

The path ahead

Several years ago I put Dallas through the so-called contender checklist for D Magazine. While there’s no formula for a Cup champ, Cup champs obviously have to be good at more than one thing. What those things are tend to be universal, which is being top 10 in at least four of the following categories. Here’s the updated version. (Taken from football, SRS refers to a team’s point and goal differential weighed against their strength of schedule.)

Numbers courtesy of Evolving-Hockey, HockeyViz, and Hockey-Reference.

Does this help clarify some of the doom? I think so. Granted, Dallas’ expected goal differential drops to 17th without Heiskanen, but their total goal differential stays the same, believe it or not. It’s actually slightly higher without him (60.09 percent), but their rank is the same. The main takeaway here is that Dallas can outshoot and outgoalie most teams, and they can win a special teams battle. Without Heiskanen, Dallas’ power play jumps all the way to third. So while a generic glimpse at Dallas without Heiskanen poses a lot of questions, perspective is still the word of the day.

Stray observations

The running theme lately is that individuals on this Dallas team can be the tide that lifts all boats (especially the netminders). Evgenii Dadonov nearly beat Pittsburgh all by his lonesome. Thomas Harley was hard on himself after that game, and I saw a lot of smoke directed at Cody Ceci, but it’s important not to take any broad lessons from one-off mistakes. If Harley had a pattern of high-risk passes in the defensive zone that became turnovers, we could low-stakes harass him for it, but he doesn’t. Harley’s defense is a big reason why he’s earned his keep. Yes, the offense is great and that goes without saying, especially on the power play. But he’s one of Dallas’ most composed defenders in his own zone.

Normally I roll my eyes over gushing that goes on over any team’s fourth line. Call it the Islanders Effect. But man, what Oskar Back, Mavrik Bourque, and Sam Steel are doing is something else. In 60 minutes together they’re at 70 percent in expected goal share. Just complete domination in limited minutes.

Speaking of unsung, Lian Bichsel and Matt Dumba haven’t had much puck luck, but they rate the highest of any defense pair in expected goal share and are winning the shot attempt battle. Their role on the team is small by comparison, as seen with their deployment.

But it’s important because while their role is minimal, it’s not absent as was the case last year when Dallas rolled five defenders versus Edmonton. And within that role, they’re winning possession. Dumba has been the number one pariah and was before he wore victory green. But credit where credit is due. He’s not perfect, and could very well be the odd man out when Heiskanen returns, but he’s very quietly pulled himself up from the bowels of Hated Per 60.

And with that, tomorrow I’ll be taking a look at Wyatt Johnston (paid), and where he ranks against his peers.

1. It’s one thing to beat a team with injuries but Edmonton (McDavid and Draisaitl) and Minnesota (Kaprizov and Ek) were missing the guts of their respective offenses.

2. Although I did wonder about people during the Ken Hitchcock year.