The NHL playoffs start on Wednesday, which means it’s time to make some predictions. If you aren’t familiar with how I make the predictions, you can check out my original post here. The short version is that I take the season-long goal differentials for each team (for example, the Red Wings were +45 this year), compare them, and use a binomial regression to pick who will win, or a multinomial regression to pick how many games the series will last. An interesting feature of the NHL playoffs is that, at least in the span I have data for, home ice is meaningless. I’m taking that as good news for the Wings, who had a chance to get into 4th place at the end of the season but ended up in 5th.
Western Conference: We’ll start in the West because the West is the Best (I’m not even going to try to hide my blatant Wings homerism).
#1 Vancouver Canucks vs. #8 LA Kings: The Canucks finished with the top record in the league and the third-best goal differential (+51) while the Kings locked up their playoff berth a few games before the end of the season but have the 11th best differential (+15). The Canucks have about a 69% chance of winning the series according to the binomial model. Given that the Kings are likely to lose, the series will probably go 6 games.
#2 St. Louis Blues vs. #7 San Jose Sharks: The Blues, I think, surprised a fair number of people with a very impressive season, ending up ahead of both Detroit and Chicago in a crowded Central division. The Sharks ended up only a point behind Phoenix, which would have put them in the 3 seed as a division winner. Instead they get the Blues and a 35% chance of advancing. 2 wins for the loser tends to be the usual favored outcome, so this one should go 6.
#3 Phoenix Coyotes vs. #6 Chicago Blackhawks: I think people will tend to pick the Blackhawks given their pedigree and larger number of points, but going by goal differential the Coyotes actually have a tiny advantage (+12 to +10). The Coyotes get the slight edge on a coin flip, winning 51% of the time. The Blackhawks are pretty equally likely to win 1, 2, or 3 games, so I’ll call this one a 7 game series just to up the excitement.
#4 Nashville Predators vs. #5 Detroit Red Wings: The Wings were actually number one in the conference earlier in the season, but got derailed by some injuries and kind of limped their way to the end of the season. Thus they sport the conference’s second-best goal differential (tied with the Blues) despite being seeded 5th. That, and the lack of any kind of home ice advantage, makes the Wings a 60% favorite to advance. The Preds will probably win 2 games.
Thus we more or less are calling chalk here: the Canucks, Blues, Coyotes, and Wings advance, although of course nothing is guaranteed. The Coyotes in particular are vulnerable, being the best team in a poor division. On to the East!
#1 NY Rangers vs. #8 Ottawa Senators: The Rangers got the top seed despite having the third-best differential in their conference. On the plus side, they still did better than the Senators. The Rangers should move on 66% of the time while the Senators take two.
#2 Boston Bruins vs. #7 Washington Capitals: The Capitals, as usual, are all promise and no delivery. They actually had a negative goal differential, getting outscored on the season despite winning more games than they lost. The Bruins have the best differential in the league, so this may not be pretty. Give the Bruins a 84% chance to move on, which is about as good as you’re going to get in the NHL, with the Caps taking two.
#3 Florida Panthers vs. #6 New Jersey Devils: If you thought the Coyotes were taking advantage of the division winner thing, gaze upon the -24 goal differential Panthers. Only four teams in their conference had a worse differential, yet they managed to win the division. The Devils weren’t great, but they managed to outscore their opponents at least. Make the Devils the favorite, winning 72% of the time while the Panthers take two.
#4 Pittsburgh Penguins vs. #5 Philadelphia Flyers: The Penguins are always a strong choice given their talent, but they’ll get even more attention since Crosby came back at the end of the year. Hopefully the Flyers don’t hire Gregg Williams to coach the defense during the playoffs! (football joke). Even with Crosby missing so much time the Penguins still had the second-best differential in the league, although the Rangers weren’t slouches either. Give the Penguins a 66% chance to win with the Rangers taking… you got it, two.
So if that all turns out as is most likely (which it probably won’t), you’ll have the Rangers, Bruins, Devils, and Penguins moving on. The NHL playoffs re-seed for later rounds, so there isn’t too much of a point in projecting ahead, but the Rangers would be favored over the Devils and the Bruins would have a war with the Penguins, with the winner being favored over the Rangers. In the West the Canucks would be slight favorites over the Wings and the Blues would be stronger favorites over either the Coyotes or Blackhawks; the Canucks would be slight favorites over the Blues as well. But we’ll see how it all turns out!