Instead of doing a Power Rankings post, each week I am going to go over all the rankings. (Note: All the rankings includes the USCHO Poll, SB Nation Poll, PairWise Rankings, my Power Rankings, and my ballot for the SB Nation Poll.)
Some thoughts on each of the ranking mechanisms:
- USCHO Poll – Garbage. Voters are lazy and (I suspect) don’t actually watch the games.
- SB Nation Poll – We generally get it as right as biased humans can.
- PairWise Rankings – Limited, but functional.
- My Power Rankings – My rankings are built on ranking teams across 10 categories and then summing those categorical rankings. Those categories are Winning Percentage, Goal Differential, Shooting Percentage, Save Percentage, Power Play, Penalty Kill (both *.333 as teams spend on average 1/3 of a game on special teams), Percentage of Shots Taken in Games, Percentage of Shots taken Relative to Opponent Quality, and the last two against teams that take more than 50% of shots in a game. This is also limited, but is more well-rounded than the PairWise, which is based solely on wins and losses. If Team A gets outshot 100-1 by Team B, but wins 1-0, I am more apt to reward Team B and think of them as a better team than Team A. My rankings attempt to account for this with the limited data available.
- My Ballot – The best, obviously.
The SB Nation Poll:
This week the final results were almost identical to my ballot. Good job everyone. My thoughts on the slight differences:
Clarkson / Boston College
I’ve watched every Clarkson game this season save for one I believe, and have seen all three of their losses. The theme was pretty constant, the Knights were pretty dominant, they just couldn’t win. They had solid possession and good chances, but they couldn’t score goals when they needed and took inopportune penalties.
Watching Boston College, none of their victories give me warm feelings about them. We know Minnesota-Duluth is not good now; all we have are solid victories over decent Providence and St. Lawrence teams, and even then the defense has been suspect. BC has allowed one or zero goals just twice this season, and three or more six times. Those numbers for Clarkson are 11 (seven shutouts) and four respectively.
Cornell / Ohio State
While I won’t come right out and say I think Ohio State are frauds, I think they fit into the top five of most polls because there’s no other obvious choice to put there. They’re riding their back to back weekends against Minnesota and Minnesota-Duluth whom we now know aren’t great. Since getting blown out by Wisconsin (whom Cornell played close), Ohio State has looked pretty mediocre against Minnesota State and Bemidji State.
Meanwhile, Cornell has played the toughest schedule (including Clarkson and St. Lawrence away, Quinnipiac, Princeton, and Wisconsin at home along with Brown and Yale) and is 5-3-0. It doesn’t get any better either. After Penn State at home next, the Big Red play at Providence twice, a home and home versus rival Colgate, and then on the road against Princeton and Quinnipiac. If they come out of that with their heads above water, they shouldn’t be below Ohio State on any poll.
Minnesota / St. Lawrence
This one is the closest of the bunch and I think largely a wash. Both teams played questionable games against Merrimack with Minnesota losing a 4-3 contest and St. Lawrence playing them to a 0-0 draw. St. Lawrence has more losses, but all of them came to ranked teams; two to Clarkson, one to Cornell, and one to Boston College.
The difference is largely that St. Lawrence has quality wins over Colgate and Boston College and Minnesota doesn’t. The most impressive Gopher result is…hmm. Losing to Wisconsin in overtime? Know what, maybe this one isn’t so close.
I know this is silly, but if the season ended today…
- Wisconsin – 10. Robert Morris
- Colgate – 7. Minnesota
- Boston College – 6. Cornell
- Clarkson – 5. Ohio State
Moving beyond the travel-aiding shuffle that would occur, those eight teams feel about right. Colgate would probably luck out as Wisconsin and Minnesota would surely play to minimize the number of flights needed, giving them a first-round matchup against Robert Morris, which is kind of disappointing because I’d like to see a tournament newcomer like Colgate play a team that has practically lived there in Minnesota.
Sometimes weird stuff happens, and here’s why: Because my strength of schedule tracks performance against teams that take more than half of a game’s shots on average, and because teams can move in and out of that pantheon week to week, it can dramatically impact another team’s ranking in that field. This season’s prime example is Robert Morris, who has either played zero quality opponents, or put forth one of the best shot totals against them depending on how Maine does in a given week.
Fortunately Robert Morris will play Colgate and Ohio State (really, four times? huh) in addition to Mercyhurst, so ending the season with zero quality opponents is unlikely.
This is also true of Minnesota-Duluth, whose shots percentage versus quality opponents metric looks a lot better when Bemidji State dips off that list.
Nothing like games against Yale and Brown to help you recover from getting statistically bulldozed by all the ranked opponents you’ve played. Once their schedule evens out, they should be a lock to sit above Northeastern and Harvard, and perhaps Robert Morris who plays Ohio State four times (yes, really) in addition to Colgate, and will probably not go undefeated in-conference.
Ups and Downs
It’s always frustrating when a team has a good week, and still moves down, or a bad week and moves up. That by and large hasn’t happened this season with the metric even capturing the nuance of Minnesota looking unimpressive in a sweep of St. Cloud and RPI netting two draws against a New Hampshire team that is probably better than them.
Things I think About Teams
- I don’t trust anyone, but I mistrust Clarkson and Wisconsin the least.
- I think Colgate may legitimately be better than Clarkson though.
- I think Providence is playing short of their potential.
- I think Harvard is bad. They got shut out by Holy Cross (who is probably on par with like Lindenwood).
- I think Minnesota-Duluth and Bemidji State are better than their records show.
- Maine and Northeastern confuse me. Are you good, are you not good, will you make up your minds please?
- I’m keeping on eye on Vermont. They’re going to be a pain for someone I think.
- If I could pick a team from the bottom five that might do stuff, I’m picking Union.
- I think St. Cloud would beat most of the bottom teams in the other conferences.
- I think I would like to see Minnesota upset Wisconsin in the first round of the NCAA tournament.