The NHL regular season is over as of last night, and the playoffs begin on Wednesday. That means it’s time for predictions! I’ll be using my very simple NHL model. The short description is that it picks a winner from the two teams’ goal differentials during the season (technically, the difference of their differentials) and who has home ice. Or, it would use home ice except that there’s no statistical advantage to having home ice over the past 5 years in the NHL. In fact, the number is slightly negative, meaning a (non-statistically significant) home ice disadvantage.
When the NBA is set, I’m going to try to predict the whole thing as well as go round-by-round as is done for the TrueHoop smackdown picks. However, the NHL handles things differently. Instead of being a set bracket, match-ups are picked by seed every round. So if number 7 upsets number 2, and the number 1 team wins, they will play number 7, not number 4. In other words, there’s no set bracket like March Madness. That means that predicting the entire NHL playoffs would be a mess, especially given how noisy they are (which I mentioned in my linked post above). So I’m just going to pick winners round by round. Also note that the multinomial model I use to pick how many games the series will take in the NBA is virtually useless in the NHL; there’s really no telling, so I’m not going to try. We’ll start in the East.
Washington-Rangers: Caps are +27 on the season, Rangers +35. So the Rangers may actually be the better team. I have Rangers with a 54% chance of victory and the 8 seed moves on.
Philly- Buffalo: +36 versus +16. The Flyers have a 59% chance of winning and the 2 seed moves on.
Boston- Montreal: +51 versus +7. The biggest mismatch in the first round so far; by goal differential the Bruins are the best team in the East and the Habs are tied for worst in the playoffs. The Bruins have a 70% chance of winning, and the 3 seed moves on.
Pittsburgh – Tampa Bay: +39 versus +7. The Penguins win 65% of the time.
If everything turns out as predicted, the second round will be Philly hosting the Rangers and Boston hosting Pittsburgh. Since I don’t consider home ice to matter, the Bruins are the team to beat in the East. How about the West?
Vancouver-Chicago: +77 versus +33. Vancouver is the best team in the league this year, by record and goal differential. But Chicago is the third best team in the West despite being in 8th. The Canucks should have a 70% chance of moving on, and taking out a very tough team.
San Jose – Los Angeles: +35 versus +21. The Sharks win 57% of the time. The Sharks are notorious playoff choke-artists, but that could be in part just noise in the NHL playoffs (I should look at that sometime). But they are favorites here, for what it’s worth.
Detroit – Phoenix: +20 versus +5. As a Wings fan, this is the match-up I wanted. The Wings aren’t that strong according to differential, but they got about the best match possible in the Coyotes. The Wings have a 57% chance of moving on.
Anaheim- Nashville: +4 versus +25. Not as dramatic as the Caps/Rangers in the East, but the home team here is not nearly as good as the visitors. The Predators should win 60% of the time.
This sets up a second round of Vancouver-Nashville and San Jose-Detroit. The Canucks are the team to beat and the predicted Stanley Cup champs. But the NHL playoffs are a cruel mistress; you see that not many of these match-ups are much different from coin flips. I’ll be adding results to the model as we go along, so the numbers might tweak a bit when we get to the second round. We’ll see how many predictions turn out correct.