The World Cup of Hockey returns in just over a year and could very well represent the only opportunities to see best-on-best competitions between nations until at least the 2022 Winter Games in Beijing. What would Team Canada’s 2016 roster look like if it were to assembled right now? We put together a hypothetical roster based on statistical analysis; here’s what it looks like.
Up front
I’ve started with a list of names over the age of 23 as of September 2016 considered obvious possibilities.
The key numbers here are relatively straight-forward. Five-on-five skill is critical; Canada ideally wants players who can put up big numbers at even-strength, can play tough minutes and whose teams out-chance the opposition while they are on the ice. Being able to take defensive-zone draws and handle power-play minutes matters, but is less important given the amount of talent on the team.
Some players are relatively easy eliminations. Bryan Little, Kyle Turris and Patrick Marleau haven’t scored enough over the past three years and are coming off poor offensive seasons at even-strength. Mark Stone plays sheltered minutes. Brad Marchand, Patrick Sharp and James Neal all had tough offensive campaigns in 2014-15. We can also nix Milan Lucic, Jarome Iginla, Logan Couture and Ryan Johansen at this point; all have poor combined point production and on-ice scoring-chance numbers.
Others are easy locks. Sidney Crosby, Ryan Getzlaf and Jamie Benn score too much to be ignored. Jonathan Toews plays the toughest minutes of any player on this list. Patrice Bergeron has ridiculous on-ice scoring-chance numbers while taking a crazy number of defensive zone shifts. All of Corey Perry, Tyler Seguin, Taylor Hall, Matt Duchene and Rick Nash fare well both by point production and by on-ice scoring chances. Finally, Steven Stamkos and John Tavares score a ton at both even-strength and on the power play, so they make the cut.
That leaves two roster spots and an alternate slot for the nine remaining players. This is when the cuts get increasingly difficult. Projecting ahead, Joe Thornton will be 37 by the time this team plays; it’s not too much of an assumption to put him on the outside. Claude Giroux is a power-play wizard, but that doesn’t carry as much weight with this club as it would with an NHL roster. Jordan Eberle doesn’t score enough for the minutes he plays relative to this competition.
The final choices are the hardest of all, but there’s obvious value in eliminating risk. Eric Staal, Jeff Carter and Andrew Ladd all have significant international and NHL playoff experience; with little to choose from by the numbers that gives them a slight edge over younger options Jaden Schwartz and Ryan O’Reilly.
The blueline
Defence is a tough position to assess, particularly because assignments can differ so much. A top defenceman can be cast in an offensive or defensive role, with scorers or with checkers; as a result it can be tough to do a fair statistical analysis of the position.
One way to solve part of those problems is to look at how each blueliner played with the top scoring forwards on his team. To that end, the chart above includes Corsi differentials over the past three years for each club’s top offensive option, with and without the defenceman on the list.
It’s helpful first to eliminate the obvious outsiders. Jake Muzzin and Mike Green played softer competition than the rest of the list; they’re gone. Marc Staal, Alex Pietrangelo and Brent Seabrook failed to distinguish themselves when out with top lines and there don’t appear to be obvious mitigating circumstances in those cases; they’re out, too.
Four players are obvious locks, based on their performance with top forwards and their own impressive offensive totals: Mark Giordano, Kris Letang, P.K. Subban and Duncan Keith. That leaves four other spots, and here it’s vital to dig into context.
• Marc-Edouard Vlasic has mostly played with middling partners and been brilliant when posted with stronger options; to me he’s a hard guy to cut.
• T.J. Brodie has played lower-end competition and mostly been paired with Giordano; his numbers away from Giordano prior to last season simply aren’t good enough. He’ll be worth a long look in 2020, though.
• Shea Weber’s usage has been extreme, both in terms of quality of competition and zone starts, and that shows here with Filip Forsberg’s minutes getting considerably easier away from the Nashville captain. He makes the team.
• Jay Bouwmeester has regularly played with either Pietrangelo or Giordano, and looks relatively ordinary away from those guys. His totals during his time with a weak Calgary team weren’t extraordinary either.
• Dan Hamhuis has played well with a range of partners on a pretty good team and posted good scoring totals all down the line despite his reputation as a defensive defenceman.
• Drew Doughty’s numbers don’t look great overall, but they get fantastic if we separate him from Robyn Regehr, with whom he spent fully one-quarter of his time over the past three years. Regehr has typically brought down the totals of all his regular partners, and the Doughty/Regehr tandem looks miles better than the Voynov/Regehr pairing. He’s an easy lock if we ignore those minutes.
Digging into the numbers, I’d add Vlasic, Doughty and Weber to the quartet outlined at the start. That leaves one spot open on left defence for Hamhuis and Bouwmeester, with Hamhuis likely the better option by the numbers.
In net
There really isn’t any question as to the identity of Canada’s starting goalie. By either regular or shot quality-adjusted save percentage, Carey Price has the best numbers of anyone in the field. He’s also close to the top of the list in terms of minutes played over the past three years (giving us a high degree of certainty as to his quality) and in terms of save percentage on high-danger shots.
After that, there’s a tier of four goalies who are pretty much interchangeable by adjusted save percentage and five others who could conceivably play their way into the mix with a strong season. As it stands, Corey Crawford and Braden Holtby would be my picks to round out the goalie trio; both have a large number of minutes played and strong high-danger save percentage numbers.
The team
There are various ways to parse the numbers, but the roster above is how I’d slice them right now. Naturally, Team Canada will be assembled based on far more than simply looking at the analytics, and regardless of the exact composition it will be the gold medal favourite at next fall’s World Cup.