Stars Stuff: Revisiting a Wyatt Johnston prediction, Bourque hates plus minus and why you should too, and one final remark on PDO

Performance or not, things are getting interesting.

An intrepid reader in our Discord server¹ prodded me after the Seattle game about how I would probably call the Stars PDO Merchants after the victory.

Publicly I took it in stride, but internally I was in Scanners mode (kidding). After all, whoever said being a PDO merchant was bad?

If this all gibberish and inside references too soon, then I recommend reading my explainer for the so-called “luck stat.” The concept of PDO is much broader than simply whether teams are getting lucky. It’s about respecting the nature of randomness and trying to give it context. It’s about understanding that talent alone (as opposed to a constantly-evolving combination of talent and fortune) does not nor has it ever driven results by itself. But just because variation shrinks doesn’t mean the drama does too. On the contrary.

The Stanley-Cup winning Vegas Golden Knights were famously an example of this — of the ‘PDO merchant’ being crowned king. As Dom Luszczyszyn wrote at The Athletic in the middle of their 2023 run:

At five-on-five the average Stanley Cup champion scores on 8.4 percent of their shots and saves 93 percent. A PDO of 101.4. Only three Cup winners — Pittsburgh in 2008-09 and 2015-16 and Colorado in 2021-22 — did so with a PDO under 100. And they were all above 99.75 percent…At five-on-five the Golden Knights are scoring on 12.5 percent of their shots and are saving 94.4 percent of shots the other way. That’s a PDO of 106.9!

This is the positive side of the PDO spectrum, and a takeaway just as important as profiling the merchants preparing for a fall like Detroit and Vancouver last year. Teams that execute at a high level to begin with are more likely to be more dramatically fortunate.

Speaking of, here are the top ten PDO teams since Pete DeBoer took over for the Dallas Stars.

There are some weirdos there, like Boston and New York, but again — this is over the last three seasons. Both teams had two very good/fortunate seasons prior to this year. Besides, there are some common denominators here, and I suspect it goes without saying (either teams with strong shooters, or strong netminding). Regardless, these are mostly very good teams.

I also think that coaching, despite being understudied, wields its own power. And that was always my issue with Dallas under Rick Bowness. They didn’t play trying to create their own bounces. They played as if they could prevent them. And that’s why they were a lot less fortunate under his leadership.

DeBoer, conversely, has a system that actually maximizes the strengths of the roster. He favors offensive depth, structures the attack around, and the execution has followed. They’re playing the odds in situations where the odds favor them, like on the rush. Under DeBoer, Dallas is 5th in goal-scoring.

So again, PDO is not a bad…word. It’s something that chart huggers and spreadsheet watchers like me are interested in because it’s a quick and dirty way to read the tea leaves at the team level. But it’s not the mark of the beast or anything. It’s a way for us to frame variation in hockey. Nothing more. Nothing less.

The Seattle win was fitting since it was the dramatic example of this ‘positive’ side. Yes, the Stars were once again outshot; in this case 21 to 36. But they won the battle between high danger chances 14 to 10. If I had the choice between which shot battle I’d rather lose to a team of talented shooters like Dallas, I know which one I’d choose every day of the week and twice on domingo.

Mavrik Bourque: based

I don’t say this lightly. If you’re someone that uses it, please know that this is not an attack on you. Gavin Spittle, if you’re reading this, I still love you very much. But here it goes: plus-minus is dumb. Not only is it dumb, but worse — it’s misleading.

Razor mentioned on the telecast that he had a brief discussion with Bourque. Bourque told Razor that it doesn’t make sense for players to receive a plus/minus in empty net situations. Razor agreed. Until Razor’s comment, it hadn’t occurred to me that many fans probably don’t even know how the NHL calculates plus-minus. So let’s get back to basics. Plus/minus is calculated by looking at who’s on the ice when a goal is scored in the following situations:

• Even strength

• When a goaltender is pulled

• When the penalty-killing team scores

If you’re new to hockey, or just never really thought about it because you assumed a stat that’s commonly-referenced in broadcasts would actively make sense, you might be scratching your head. “Wait, what?” Exactly. Not only does plus/minus make no distinctions between different gamestates, but it doesn’t even treat goal-scoring itself equally. To summarize, your plus/minus is:

• Not affected by power play goals

• Adversely affected by empty-netters

• Does not include penalty shot goals

In other words, if you are good on the power play, plus/minus won’t reflect that. If you defend late leads, plus/minus won’t give you credit. If you’re on a bad team, and you’re always used to score for the late-game tie, plus/minus won’t respect your status. Because plus/minus isn’t broken down by rates, there’s no distinction between role players and superstars.

This is why so many modern analytics are focused on even-strength shot metrics, adjusted for minutes, and adjusted for score². By including more events (in this case, shot attempts) with a clearer distinction between gamestates, we drown out more noise. That doesn’t mean we drown out the noise completely, but at least it puts us on a clearer path.

So good on you, Bourque.

More Wyatt Johnston praise

I know this is broken record stuff. Wyatt Johnston: elite. But broken records can still carry beautiful tunes.

Johnston still continues to impress. The player I was watching closely the other night wasn’t Johnston, but Matty Beniers. I’ve always like Beniers. I thought Seattle was lucky to avoid Owen Power not because I didn’t believe in Power but because I think a top line center is slightly favorable to a top pairing defender. I thought of Beniers as a poor man’s Barkov: a solid two-way pivot. I still think Beniers will be a very good player for many years. But Beniers was much more physically advanced in college. While this tends to be touted as an asset at the prospect level, it’s also a double-edged sword: having the physical advantage in juniors or lower levels is not an advantage a player can take with them to the NHL. It was very clear versus Dallas that Beniers is still struggling with that transition at times.

Ultimately this is why Johnston has surpassed him. There’s a really easy way to avoid a prizefight with a bigger man: don’t agree to his terms. Using his mind, Johnston is able to win battles that he never has to engage in. You simply won’t see cards like this for players this young.

Click here for a casual fan’s guide to Jack Fraser’s hockey cards.

Admittedly, there’s a sense of mother’s pride here. “Haha! What did I tell you?!?”

I put a lot of work into the film room in July two years ago, which makes me happy less because I got to say I was right, and more because it wasn’t a waste of time. (This is a paid post, granted.)

Certain hills you have to be willing to die on, and my hill was Johnston scoring 60 points in his sophomore year before he scored 60 points.

A full year with Benn and Dadonov will naturally buffer his point totals. Combine that with him refining his skills — both seen, and unseen — and there’s no reason why he can’t break 60 points.

Maybe this is just poor form. But can’t I gloat? Just a little. He’s on pace for 75 points this year, and a big part of why Dallas can beat anybody when fully healthy. Johnston stands alone on his own hill, and still he outnumbers the opposition.

Could Dallas catch Winnipeg?

The Stars have certainly made things interesting. Where before it looked like Winnipeg finishing first in the West would be mathematical certitude, now it’s merely mathematical likelihood. With a win over Vancouver yesterday, the Jets are now six points ahead of Dallas, but Dallas has a game in hand. Five of their next eight games are against playoff opponents (VGK, LA, DAL, STL, EDM) while only two of Dallas’ next nine are playoff bound (WPG, and MIN). Needless to say, the Stars are very much in this as the potential top team in the Central.

However, there’s a new wrinkle: the wild card spots. Where before the second wild card spot had potential to be the Rick Tocchet-led³ Vancouver Canucks or the offensively anemic Calgary Flames, now the top teams in the West will be up against a strong Minnesota team when healthy, or a scorching hot St. Louis team that Jim Montgomery appears to be getting the best out of.

St. Louis and Minnesota are not legit threats IMO (I think the jury’s still out on the Blues, especially with the loss of Colton Parayko) but they do have the elements of an upset team: hard to play against. The shot quality they allow is a testament to this.

No, I don’t think of any of the top teams have to actively worry about either squad. But unlike in prior trends, there won’t be the gimmie matchups that previous top teams could count on.

Just saying.

Programming note

Tomorrow I’ll have a paywalled piece on a few of Dallas’ prospects, including NCAA free agent signings Ayrton Martino and Trey Taylor making their debut with the Texas Stars, and Aram Minnetian and Boston College up against the unstoppable David Carle.

1. I know I keep plugging this place, but a lot of new people recently joined, and it’s gotten better as it’s gotten bigger.

2. Modern analytics adjust for score because trailing teams shoot more while leading teams shoot less. This became a feature of modern fancy stats in 2011 when it was found that a better predictor for future success wasn’t points in the standings but shot differential in tied game states.

3. Are we sure he does?