“What is the cause of management’s fantastic faith in the machinery?”
That was Richard Feyman’s humbling question to NASA after being appointed by the Secretary of State to figure out how the Space Shuttle Challenger exploded in 1986. The shuttle itself was unique—built “at the edge of previous, or outside of, previous engineering experience.” A lot went into understanding the unique interaction of solid fuel rockets, liquid fuel engines, and avionics. But in the end, despite analysis from engineers, astronauts, and military men, it was a Cal Tech physicist who found that the gaskets (or O-rings) couldn’t survive in the cold.
As complex as systems can be, sometimes it’s just the one thing.
The Dallas Stars have failed in a season where even cynical fans felt shielded beforehand from Facebook’s version of Nelson’s mocking finger. GM Jim Nill signed Martin Hanzal, Marc Methot, Alexander Radulov, and Ben Bishop in the offseason to put a shot in the arm of the roster. Despite all the bells, whistles, and offseason gold belts, Dallas currently has a 1.1 percent chance of making the playoffs, thanks to franchise-setting failures.
How does this team fail to beat Lindy Ruff’s 2013-2014 squad — a squad that didn’t have John Klingberg (no longer leading defenseman in scoring, but still just two points shy at 62), Radulov, Jason Spezza, or Radek Faksa—that did make the playoffs? Ken Hitchcock’s Stars would have to score 3.4 goals per game in the next five to get even with Ruff’s offense from four years ago. Using a sci-fi DeLorean, Ruff’s Stars would have to give up 3.6 goals per game in the next five to get even with Hitchcock’s defense from this year. In other words, Ruff’s team had better offense, worse defense, and still found a way. That can’t be right, can it? Because systems can be complex, observers naturally look for indirect answers.
It’s the injuries!
Wrong. Dallas is 19th in cap hit of injured players (or “CHIP”). They’ve only had 28 games in which they were above the CHIP average, and pretty much all of that was either a busted knee, or blown hamstring from Hanzal and Methot (two players accounting for 11 points in 69 games between them).
It’s the goaltending!
Wrong. Going back to the 2014 comparison, Ruff’s team had a combined save percentage of .907 thanks to the musical chairs of colorful veteran backups (without including the one-hit wonders like, umm — Jack Campbell?). Hitchcock’s team sits at combined .914 save percentage.
It’s secondary scoring!
You almost got me there, but let’s face it. Ruff’s 2014 playoff team of crusty veterans and pesky wingers was supported by Cody Eakin, Alex Chiasson, and Ryan Garbutt. With Devin Shore, Mattias Janmark, and Tyler Pitlick in their departed ice, this isn’t exactly the difference between Matthew Barzal or Alex DeBrincat and Denis Gurianov (guh).
It’s the coaching!
Bingo. And it’s okay. It’s not reductionist to blame the one thing in a complex system, assuming the evidence permits.
As I argued in December, I noticed a worrying trend with this Hitchcock team. Shot attempt-wise, they were downright awful on the road — whether down a goal, up a goal, tied, leading, or trailing, and their possession numbers were abysmal. With Hitchcock having less control over match-ups, I felt his roster decisions were more emphasized and vulnerable. Hitchcock wants to dominate one or two game-states (when leading or up by one), instead of push to compete in other game-states by displaying balance in every skate, puck, and bench spot of his chosen strategy (get the lead, and then preserve it).
So someone like Julius Honka gets lost in the dump-and-chase shuffle. Despite having more exit success than any other defensemen. Despite the fact that when he’s on the ice against top 10 teams, Dallas is 6-5-2, but are 1-11-1 without him. Despite the fact that Dallas takes 21 percent fewer penalties, and draws 15 percent more. Despite the fact that in the last 10 games against top 10 competition, Dallas has outscored their opponents 16-12 with Honka, but are outscored 20-10 without him. Despite the fact that Dallas has a six percent higher chance of scoring with him, and without him, Dallas’s opponents have a 12 percent higher chance of scoring. Despite the fact that Dallas takes an extra 1.45 penalties per hour in games where Honka is scratched. That last part is kind of a big deal, since their improved power play couldn’t make a difference in their overall special teams performance; a power play that has seen enough lows to keep a unit of Benn, Seguin, Klingberg, Radulov, and Spezza just 18th in the NHL.
But it’s not just Honka (nor am I arguing that Honka would have saved the season singlehandedly). Forwards like Spezza and Gemel Smith have become lost in the dump-and-chase shuffle too. Despite the fact that Smith has a higher goals-per-hour rate than everyone not named Pitlick, Benn, and Seguin, yet remains healthy scratched in favor of underperforming forwards. Despite the fact that Spezza’s first assists rate per hour is well above players like Ritchie, and Shore (who get more time on the ice). Despite the fact that Faksa was never a third line tweener — even before last season — but through the first 20 games that’s how exactly Hitchcock inexplicably deployed him. And so even the big guns have gotten lost in the dump-and-chase shuffle, going from triggering the expected offense of a dynamic playmaker to the expected offense of a static (but still effective) shooter. The end result is that a dynamic player like Jamie Benn makes less zone entries. A dynamic player like Tyler Seguin chambers less shots per hour. A dynamic player (even as far back as last year) like Spezza magically forgets how to enter the zone with possession. And a dynamic player like Radulov has less shot assists (Josh Lile’s analysis contrasts his time in Dallas to Montreal’s own inert system of grit, grind, and grease so it’s weird to see such a drop in certain traits unless they weren’t the product of the team’s system).
Maybe just one change doesn’t lead to an extra win. But all of them at once with the evidence for each suggesting they’d make a difference? I couldn’t say “meh, probably not” with a straight face. But maybe that’s just me.
When Feynman found the one thing that led to the Challenger explosion, he had some prudent questions for the Federal Aviation Administration. The turbine blades of these shuttles were allowed to have cracks. The administration made a system classification; blade cracks were maintenance problems, not flight safety problems. Why? Because cracks in the turbine blades can’t grow to become full-blown fractures in the time of an allotted mission.
How well do we know that cracks always grow slowly enough that no fracture can occur in a mission?
Feynman’s question is the 2017-2018 Dallas team in a nutshell. Hitchcock could accept the cracks of Greg Pateryn constantly icing the puck (who’s not responsible for Dallas’ strong league lead in icings, granted), or Devin Shore playing an invisible game even with visibly talented linemates because he assumed those cracks couldn’t grow into fractures. Except they did in spectacular fashion. And through it all, Hitchcock remained too obstinate to try different things, except spam his favorite defensive pair and a dynamic quartet who have succeeded in spite of Dallas’ system.
Hockey is a sport of unique weather. A lot of different elements — like promising young prospects and aging veterans — interact to create complex systems. Those elements are defined, identified, and controlled by the head coach. A lot of things could have helped Dallas potentially succeed this season, such as better drafting (a good prospect might have helped secondary scoring if Dallas wasn’t so dreadful at it), and the walk-on-water miracle of being 100 percent healthy, for example. But we know the one thing that failed. It doesn’t mean Hitchcock is a bad coach. It just means that for this season, and this system with this roster in this day and age of hockey, Hitchcock was the one thing putting too much faith in his own machinery; a system that crashed when that one thing failed to properly interact with different elements.