After watching Dallas lay an egg in Detroit, I thought about my favorite scene in all of filmmaking: the blood test from John Carpenter’s, The Thing. When it happens, we’ve already been put through the ringer. Our likeable crew of American researchers stuck on ice — led by the fantastically-bearded Kurt Russell — have already witnessed a series of Lovecraftian nightmares: from dog aliens, blood flowers, and leviathan eyes to a human chest that transforms into Venom’s mouth.
But the alien antagonist isn’t Victor E. Green or plants from the Northeast — it’s just a parasite that can clone other organisms, including humans. To figure out who the potential imposter is, Russell’s character ties up all the surviving members, and sticks a heated piece of copper wire into their respective blood samples. It’s a tense scene involving the literal blood and guts of true identity. And it’s this grand guignol of sci-fi spectacle that reminds me of the Dallas Stars.
I come back to this analogy not because it’s an excuse to think about the film (but that too), but because there’s something apt about it all. Two playoff appearances in 10 years is the first-world nightmare Stars fans have been tortured by. But who goes there?
What kind of team is Dallas, at the level of the drafting, coaching, and management? A fast, north-south team? A methodical, physical team? A creative, east-west team?
After 10 years of dysfunction — in which three of Dallas’ first round picks have been traded, two haven’t earned the trust of the team’s coaches at one point or another, and it’s taken three years to give Jason Dickinson meaningful minutes after a 53-point rookie campaign in the AHL — I’m not sure Dallas knows either. It was hard to think of anything else after watching them lose to the worst team in the NHL. With that said, onto to the numbers...
1. Teams four points (or more) out of the playoffs at that point rarely overcome the deficit. Since 2005-06, 46 of 52 teams (88 per cent) that far out on Nov. 1 ended up missing the playoffs.
This quote from Jonathan Willis isn’t a fancy stat. It’s a pattern. Teams don’t win the right to fight for Lord Stanley’s cup 10 games into the season. But they can certainly lose the right to play 10 games into the season. Stats aren’t gospel, which is part of the joy of watching sports to begin with; teams defying expectations. Nonetheless, it emphasizes the importance of the next two games to make or break that statistical benchmark — on the road against Montreal (3rd in the Atlantic) on Tuesday, and Toronto (currently 1st in the Atlantic) on Thursday. Yeesh.
2. At even strength, Dallas ranks 16th in the NHL in expected goals-for, and 9th in expected goals-against.
So far, Jim Montgomery’s system seems exactly as advertised: not as icing-happy, and dump-and-chasy as Ken Hitchcock’s, but not as To Valhalla! as Lindy Ruff’s. As such, they’re getting higher shot quality than Hitchcock’s team (last season, despite the gaudy percentages, Dallas ranked only 18th in scoring chances for per hour), but aren’t bleeding opportunities going the other way like they did under Ruff.
3. At even strength, but on the road: Dallas ranks 30th in expected goals-for, and 2nd in expected goals-against.
You already know this in your bones, but it’s still an incredibly odd pair of stats that we can reasonably say will change with time. Basically, Dallas’ types of shots — wrist shot, slap shot, deflection, distance, rebounds, rush shots, different angles, and strength states — become a lot less varied on the road, but they’re able to hunker down on defense quite well. As for why that is. I’ll get to that in a minute...
4. Justin Dowling leads the team in time-on-ice per game at even strength with 14:34 per game...on the road.
I’m never funny. Which means this is not a joke. Natural Stat Trick has only tallied one game from Dowling on the road, so it’s silly to read much into it. However, even as a small sample size, this is kind of bizarre. First off, I think Dowling is a very solid player. He brings quiet playmaking and speed to a lineup that needs it. But I have two problems with this. One is minor. The other is major. 1) if Dowling was brought in to be a top six forward, why wasn’t he on the roster sooner? And 2) Dowling’s icetime represents what I consider a problematic theme. Which brings us to number five...
5. At even strength and on the road — Tyler Seguin, Jamie Benn, Alexander Radulov, and Jason Spezza are ranked 2nd, 5th, 6th, and 10th (respectively) in time-on-ice per game among all forwards.
Two weeks ago I got bored and looked into how teams divide their icetime to their top five producers at even strength. In ranking teams from most productive (raw goals-for) to least, there was a clear (albeit fairly self-evident) pattern: goal scoring teams either play their top producers more, or around league average, while less productive teams “balance” their lineup. You’d expect Seguin, Benn, and Radulov in particular to be some mix of 1st/2nd/3rd but instead only one of them breaks into the top three. Why? Radek Faksa and Forwards Next to Him During Morning Skate are getting big minutes. Checking lines are the ultimate risk policy for NHL coaches, after all.
6. At even strength and at home — Seguin, Benn, Radulov, and Spezza are ranked 1st, 3rd, 2nd, and 10th (respectively) in time-on-ice per game among all forwards.
That’s more like it. However, it’s still weird to see Spezza where he’s at. The only forwards who get less time on ice per game are Dickinson, Smith, Hintz, and Ritchie. And this despite the fact that Spezza is Dallas’ 4th best even strength producer. On one level it makes sense. At Spezza’s age, giving him tough minutes might look like a bad risk policy. Or maybe he can’t physically handle it. On the other hand, he’s getting it done, and probably wants more. I won’t pretend to know the answer, but I do know that — to quote Mark Twain — ‘twenty years from now you will be more disappointed by the things that you didn’t do than by the ones you did do. So throw off the bowlines. Sail away from safe harbor.’ Pairing Spezza with Valeri Nichushkin and Denis Gurianov might not be safe harbor, but Dallas might want to loosen their grip on those bowlines. IMO.
7. Julius Honka and Miro Heiskanen draw penalties at a rate of .7 and .68 (respectively) per hour. They’re pretty good at creating rebounds too.
To bloviate for a second, I suspect hockey men hate the term “luck.” It implies a lack of control, and free will. Despite the fact that luck plays a significant role in how teams climb in the standings. Or that when you look at scoring events across multiple sports, including professional basketball, and both college and American football, scoring is more erratic in hockey than anywhere else. The NHL is messy, chaotic — especially at even strength — and counterintuitive; if you predicted that rebounds from the point are as likely to lead to goals-for as goals-against, you’re much more intuitive than I am.
I say embrace the chaos. Embrace the risk. Risk in hockey is like a tide. You can either try to stop each rise and fall with a broom—overplaying your checking line and balancing your blueline with stay-at-home defensemen—or you can ride the wave and strengthen each rise and soften each fall i.e. trusting players who project to drive offense long term over judging them for That Big Mistake on that rainy day.
Last season, Dallas’ blueline didn’t have anyone drawing penalties at that rate. They also only had one defenseman creating more than half a rebound per hour. Thus far, they have four this season: Carrick, Honka, Heiskanen, and Lindell. This is what I consider ‘riding the wave’, strengthening each rise (putting a deadly power play on the ice to potentially pad the lead), and softening each fall (creating more rebounds for forwards to clean up to overcome a deficit).
8. At even strength, the Nichushkin-Spezza-Radulov trio lead all forward combos in shot attempt differential at 66.67 percent.
Technically, Dallas has a potentially dangerous second line. Even though it’s clear that Benn and Seguin miss Master Yoshi. But Dowling, Benn and Seguin are doing quite well. The goals just aren’t coming for the top producers. Benn’s production is more concerning. He’s 9th on the team in even strength goal rates. His shot rates are fine, though, which tells me that this is just your yearly Is Benn Injured or Is It That Goals are Distributed in Arbitrary Intervals situation. Well, that and a teaspoon of ‘peak production.’
9. Gemel Smith has produced .93 goals per hour at even strength since last season. That is second only to Benn, and above Seguin for the team lead.
This stat requires no comment except that I think when a small sample size is a (really) good sample, you don’t healthy scratch a good sample.
10. Some words about what “stats” are really about.
Ben Bishop had some harsh words about analytics. Instead of arguing with a professional athlete, or crunching a bunch of numbers that I myself am not certified to talk about, I’ll just pontificate with a capital ‘p.’ To me, when I think about fancy stats, it’s not the numbers that interest me. It’s the psychology. So many of these debates are broken into Watch the Game, Nerd! versus Read My Histogram, Grandpa!
That’s not a coincidence. What we’re really talking about is how information allows us to frame certain decisions or problems. In hockey, teams frame risk. But do they frame their decisions narrowly or broadly? Let’s consider roster decisions. I suspect most NHL coaches are like most humans. When given the opportunity, they will sequence their roster options into their simplest scale: the risk of preventing offense versus the reward of creating it.
Scorers are risky. Checking lines are not. Let’s take risks sometimes, and avoid risks most of the time so we can score more than our opponents do.
Under this narrow frame, a roster that is more effective at preventing goals is more attractive than a roster’s raw ability to score. But what if we frame it more broadly?
Am I constantly creating a situation where the roster can better manage the game when trailing, is within a goal through two periods, tied, or down a goal? Am I adjusting to how we manage the game when leading, or up by a goal?
Suddenly we generate more choices. This isn’t always a good thing. At the level of the power play, narrow framing is good. Good shooters play with good passers and voila! At the level of even strength, broad framing works best, and “fancy stats” help clarify how teams frame their ability or inability to be effective. Suddenly we’re asking questions like: What defense pairings are more effective at preventing offense when we’re leading? Which ones are more effective at building offense when we’re trailing? What forward lines are more effective at preventing offense when we’re within a goal? What forward lines are more effective at building offense when we’re tied? What combination of forward lines and defense pairs better manage multiple gamestates instead of just one? It doesn’t mean numbers alone should decide who plays, and how. It just means you’re more open to possibilities, and maybe enough self-awareness to consider whether the safe harbor is really the best retreat.
As for what this means for Dallas’ highly average start? Well, what kind of team Dallas is trying to be? If they wanted to be younger, why sign Comeau? If they wanted to be faster, why sign Polák? If they want to win now, why not trade their best prospects for big name veterans? I guess it all depends on how Dallas answers that essential question. Who goes there?