Advanced stats predict second round of Stanley Cup playoffs
By Ian Cooper, Phil Curry, IJay Palansky and Mikal Skuterud
Yeah, ok, our 4–4 record in the first round wasn’t good. By way of explanation (or, if you prefer, by way of excuse) we lost both coin-flip series: the Blues lost to the Wild and the Canucks lost to the Flames. Those are the breaks.
But to borrow a few tired clichés from old-school analysts, we at the Department of Hockey Analytics have grit, determination and, well ... a whole bunch of spreadsheets. We’re not going to let a few bad games scare us off, especially when both our picks to meet in the Stanley Cup finals are still playing. (Now seems like a pretty good time for the Department of Hockey Analytics to say thanks to the NHL’s Department of Player Safety for suspending Red Wings defenseman Niklas Kronwall for Tuesday night’s Game 7 against the Lightning in Tampa Bay.)
Which is why we’re doubling down on the Bolts ... sort of. We’re still predicting Tampa Bay to meet the Blackhawks in the finals, and if that happens we’re still giving the Lightning a 52.3% chance of winning that series, but let’s point out an obvious fact: For that to happen both teams have to get to the finals in the first place.
Taking the relative difficulty of each team’s road to the finals into account, our updated predictions give Tampa Bay a 22.2% chance to win it all. But with two of the tougher Western Conference teams having fallen in the first round (St. Louis and the Jets), Chicago’s path to the championship has gotten a fair bit easier. Our model now makes the Blackhawks the slightest of favorites, with a 22.4% chance that they will be the ones drinking champagne out of the Cup in some random bar in Muskoka in July.
So what does the second round have in store?
Blackhawks vs. Wild
Prediction: Blackhawks in 6
In round 1 Chicago held serve against the Predators while Minnesota won a toss-up series against the Blues. But that’s history, and while the Blackhawks’ goaltending has been inconsistent, the Wild can’t count on a repeat of the two gifts that St. Louis goalie Jake Allen served up in the deciding game of that series.
Chicago has been the better team all year long. And while Minnesota had a great run over the last two-plus months of the regular season, the Wild’s underlying numbers aren’t particularly flattering. In fact their shot-attempt differential (aka, Corsi For Percentage) has been on a slow and steady decline all year long, and their scoring-chance differential has fallen off a cliff in the last six weeks. During the regular season the Blackhawks ranked second overall in both categories; Minnesota was 16th. (Stats courtesy of war-on-ice.com). Devan Dubnyk has been good in net, and anything can happen in a best-of-seven series, but the favorites on paper are also the favorites according to our model, which pegs Chicago as a 58.4% favorite.
Ducks vs Flames
Prediction: Flames in 7
All year long we’ve been waiting for Calgary to crash and burn in the inglorious tradition of this year’s Avalanche. As impossible as it seems, the Flames’ shot and scoring-chance metrics this season are actually worse than those of the god-awful Maple Leafs. But Calgary rode the second-highest team shooting percentage in the league to a surprising playoff berth and a first round win over fellow playoff-lightweight Vancouver. So how can our model predict the Flames to defeat top-seeded Anaheim?
The numbers don’t lie—we-triple checked this particular prediction just to be sure. Our model severely penalizes the Ducks for their ridiculous, unsustainable 33-1-7 regular-season record in one-goal games, which grossly inflated their points in the standings. And Anaheim’s 51.1% score-adjusted shot-attempt differential isn’t exactly setting the league on fire either. On the other hand, Calgary has so far overcome its glaring shortcomings. The Flames actually earned one more standings point in the second half of the season than the Ducks, and also had a better power play, and better shooting and save percentages.
It’s a close call, but the model says that Calgary is a 51.7% favorite to take the series.
Lightning vs. Canadiens
Prediction: Lightning in 6
Montreal is one of those teams that give people like us heartburn. On the one hand, the Habs tend to have fairly hideous possession numbers for a Cup contender, with a regular season Corsi For Percentage of 48.5%, which was just a shade worse than that of the Coyotes (48.6%). Outside of Max Pacioretty, P.K. Subban and maybe Tomas Plekanec and 36-year-old Andrei Markov, there really isn’t much to get excited about among the team’s skaters. Which explains why Montreal ranked 18th in goals-scored during the regular season.
The Canadiens pretty much rely on goalie Carey Price to stand on his head and win them games. More often than not, he does, which is why they can be a scary opponent.
It’s not like our model doesn’t give good goaltending its due, but if Price really is a singular talent, it probably underestimates Montreal’s chances, particularly in a short series. That said, if Price is anything other than exceptional, his teammates aren’t well equipped to handle Tampa Bay’s skill and depth the way Detroit’s top shutdown players did. Price can probably expect to continue to face a lot of shots.
We’re probably going to get hate mail from Canadiens fans over this (laden with reminders of Montreal’s playoff sweep of the Lightning in 2014), but our model gives Tampa Bay a 65.1% chance of victory.
Prediction: Rangers in 7
We’ve seen this movie before. Alex Ovechkin vs. Henrik Lundqvist. Offense vs. Defense. High-end skill vs. Depth. If the past is any indication, this series will be a boring, low-scoring war of attrition.
But there are good reasons to think that won’t be the case this time around. Washington, which completely revamped its blue line last summer after hiring defensive-minded coach Barry Trotz, is a very different team. And New York may have allowed fewer goals-against this season, but they also had 10 more goals-for than the Capitals.
Washington was below .500 in one-goal games, which suggest that its 101 regular-season points may understate the Capitals’ true abilities. But we’re also talking about a team whose regular-season power play (25.3%) plummeted (15.4%) against the Islanders, who had a pretty horrid penalty kill (78.0%) during the regular season.
Washington, particularly Ovechkin, needs its power play to click in order to dominate, but they’re facing a Rangers team whose penalty kill ranked sixth in the NHL, at 84.3%. We’re betting that the Caps’ power play will continue to sputter.
Despite all the changes in Washington, our model is telling us that while the script may have a few different twists and turns, the movie will end the same way. There’s a 54.0% chance that New York will take this one, likely after a long, hard-fought battle.
For updated picks from Ian’s four-year-old (who is gunning for us this time, by the way), click here.
The Department of Hockey Analytics employs advanced statistical methods and innovative approaches to better understand the game of hockey. Its three founders are Ian Cooper (@ian_doha), a lawyer, former player agent and Wharton Business School graduate; Dr. Phil Curry (@phil_doha), a professor of economics at the University of Waterloo; and IJay Palansky, a litigator at the law firm of Armstrong Teasdale, a former high-stakes professional poker player and Harvard Law School graduate. Please visit us online at www.depthockeyanalytics.com
Dr. Mikal Skuterud (@mikalskuterud) is a professor of economics at the University of Waterloo.