It’s a bit belated because I was out of town, but here are the predictions for the NHL conference finals. Having played back-to-back games to start the series for some crazy reason, the Blackhawks are already up 2-0 on the Kings, and the Bruins stole game 1 from the Penguins. The series predictions are below, along with a little summary of the previous round, along with what the games already played mean for the predictions.
Chicago was the best team in the regular season, racking up an adjusted goal differential of 91. The Kings were ok but not great, putting up a 26. That difference leads the model to predict that Chicago will win 78% of the time with the Kings probably winning two games. According to NHL.com, teams trailing 2-0 in a best-of-seven have come back to win 12.7% of the time, which means the Blackhawks should be up to something like an 87% chance of winning the whole thing. Of course the outcomes so far, while suggesting that Chicago is a much stronger team, still allow anything from a sweep to the Kings winning in 6.
Speaking of the Blackhawks, they beat my beloved Red Wings by winning three straight, including game 7 in overtime. That same NHL.com page also has the outcomes for teams that have come back from being down 3-1, which has happened about 9% of the time. Of course, many teams that are down 3-1 are the worse team; if someone is going to be up 3-1, you would think it would be the better team. However, Detroit was probably not the better team in that series. You would think that Chicago had better than a 9% chance to come back, and that they did. I’m not too upset though; the Wings extended their playoff streak and did better than expected once they got there.
The Kings were kind of a playoff favorite since they’re the defending champs despite the fact that they came in as the 5th seed. That seems odd to me since they made their playoff run from the 8 seed last year and they were only two games ahead of the 8 seed this year. Everyone likes to place importance on the playoffs, since it’s at the end of the season and they give someone a trophy at the end, but they forget that the sample is so small. Across the last two regular seasons everyone (excluding Boston and Pittsburgh, who had a game cancelled because of the marathon attack) has played 130 games. The Kings are 11th in the league for wins in those games, tied with the Devils. In last season and this season’s playoffs combined, the Kings have played the most games of any team in the league at 35. That is, in these two playoff runs so far the Kings have played less than this year’s lockout-shortened season. Had we just cut off the season at 35 games this year, the Blackhawks would already have the Stanley Cup.
As mentioned, if you add in playoff games the Kings have played the most of any team, which isn’t surprising given their playoff runs. But they’re still only 4th in total wins behind Pittsburgh, Chicago, and the Rangers. They’re only 7th in total win percentage (games won divided by games played; ties don’t count towards wins) behind those three as well as Boston, St. Louis, and Vancouver. What the Kings do lead the league in is difference between playoff and regular season win percentage. Theirs has jumped about 17%, much better than second place Phoenix’s 7%. Of course, to make a jump like this you have to be mediocre in the first place. Chicago, who has done well in the playoffs and the regular season, has seen a 2% drop in win percentage from the regular season; the Penguins have an 11% drop. Boston is about even. St. Louis, who I was somewhat surprised to see as the 3rd-winningest team of the last two regular seasons, has had a 20% drop to the playoffs.
Some people like to say that the playoffs are where you learn who the best team is. I think that’s a little silly, especially in the NHL. If the Kings are the best team in the league, or nearly there, why don’t they win in the regular season? Do they not want home ice in the playoffs? Are they waiting to ‘turn it on’, which nearly cost them a playoff berth at all last year? It’s a tough argument to make. The Kings are fine but mediocre, and have happened to have good runs when they start the playoffs. If they make things interesting against Chicago and make it against the Eastern conference champ, we’ll have a very unlikely back-to-back champion.
Speaking of the Eastern Conference, Pittsburgh is favored over Boston. The Pens had a 41 goal differential advantage over the Bruins in the regular season, which the model translates into a 69% chance of winning with Boston winning two. Of course, they’ve already won one game. According to whowins.com (which is what I wanted to find earlier but Google didn’t help me out until now), the team that wins first in the NHL is 68.5% to win the series. But that ignores who was at home, which tends to ignore which team is better. If you look at teams that are up 1-0 on the road, the probability goes down to 55%. Either way Boston is obviously in better shape than the 31% the model gave them before the series started. But that drop from 68.5% to 55% is the ‘being the worse team’ effect, even though Boston has a win in hand and has ‘stolen’ home ice from Pittsburgh. This series is still a toss-up from either perspective, although it certainly feels more like a six game series than Chicago-LA, doesn’t it?