Bracket ABCs: Late Regular Season Edition
The Latest Look Into the Machinations of the Pairwise
by Adam Wodon/Managing Editor
This is our last look at the Pairwise before the conference tournament playoffs begin. Since we last looked at it, Wisconsin has been a big mover up, and Michigan has dropped the most. But there has otherwise been much less volatility than years past, as we suspected.
As mentioned in the last article, because there are only three components left, and because ties are broken by the RPI, which is one of the components already, the Record vs. Common Opponents criterion becomes meaningless in most instances. That means head-to-head records can mean quite a bit, but at this time of year, you're only playing teams in your own conference. So the most you're able to do is flip a comparison here or there with a conference opponent that is close to you in the Pairwise. It can be important when you get down to the nitty gritty, but the likelihood of big swings in the Pairwise is a thing of the past.
Meanwhile, just to clear up any lingering confusion, it appears as though all the other major sites that publish a Pairwise, have moved their calculations to be in line with ours. As we said from the beginning, we believe ours is the correct representation of the formulation that the Committee will use on selection day, and with everyone in line, there should be no more confusion over it.
(See bottom of article for potential bracket matchups and issues)
1-2. Boston College / Minnesota
Nothing has really changed for these two. They are going to be the top two seeds in the tournament, No. 1 seeds in the Worcester and St. Paul Regional, respectively. There's a month to go, but this is all but locked in.
The only issue is which gets No. 1 overall and which gets No. 2. Boston College's RPI is significantly higher than Minnesota's, but Minnesota has the edge in head-to-head (win/tie). That means it comes down to Record vs. Common Opponents, and right now, that's dead even. There are many common opponents which both teams can still play — in Minnesota's case, Michigan, Wisconsin and Penn State; in Boston College's case, Notre Dame and New Hampshire. If Boston College loses to one of those teams in the Hockey East tournament, even one game of a best-of-3, it would drop its Record vs. COP and cause Minnesota to be No. 1, even with a higher RPI. Likewise, if Minnesota improves its .500 record against Wisconsin along the way, it would improve its Record vs. COP. There is no way for Boston College to improve its Record vs. COP because it already has a 1.000 winning percentage against New Hampshire and Notre Dame.
3-6. Union / Ferris State / St. Cloud State / Wisconsin
Union and Ferris State have limited ability to move up, but could certainly slip down and out of a top seed. Ferris has a low Quality Win Bonus, and because of its schedule the rest of the way, really needs to win out, or close, just to maintain its spot as a top seed in Cincinnati's Regional.
For Union, going 0-2 against Lake Superior State is holding it down in comparison to nearby teams, though the Dutchmen do get a boost from other strong wins and a good RPI overall, not to mention its head-to-head win over St. Cloud.
Without any head-to-head games against Ferris State by any nearby teams, it's all going to come down to RPI. But Ferris' RPI can be more volatile because it's playing teams that are low in RPI. A couple losses and it could provide enough swing, so long as St. Cloud and Wisconsin are playing well.
The big disparity in Wisconsin's home/road record is holding down its RPI, because of the way things are now weighted. It would have a much better chance of a top seed with some more road wins, but it won't get much chance at that.
Some recent losses in league, such as to St. Lawrence and Colgate, is hurting QU in Record vs. COP with nearby Pairwise teams, as is an earlier loss to Alaska-Anchorage. Then again, as we now know, Record vs. COP is basically meaningless unless teams have also played head to head, which Quinnipiac hasn't against any of the teams above it except Union. But that was a split, so it also makes it moot, unless they meet again in the ECAC tournament.
So, the ability for Quinnipiac to get back to a No. 1 seed spot is somewhat limited without running the table. If Quinnipiac holds its No. 2 seed slot, you could easily see it winding up in Union's regional for a second straight year, assuming the committee slots teams more geographically again. Prior to last year, I wouldn't have made that assumption, but it seems to be the tendency now, moreso at least. But the committee is made of different people every year, so who knows. Otherwise, the committee would try to keep the 1-8, 2-7, 3-6, 4-5 pecking order, so Quinnipiac would go where whatever team it matches up with is.
A couple more losses and Quinnipiac could drop to a 3 seed, at which point, all bets are off
Two losses to Quinnipiac earlier in the year could be offset if Lowell flips the Record vs. COP. There is a chance, of course, Lowell gains enough momentum to grab a No. 1 seed, but if it's 4th overall, it would mean going to Cincinnati.
9-12. Northeastern / North Dakota / Cornell / Vermont
Currently, what would be considered the No. 3 seeds in each region. Of course, this is all going to change, even though the Pairwise is less volatile than in the past. Keep winning, move up; lose, move down. It's that simple for these guys.
Colgate is in this mix almost solely thanks to its tremendous record against top teams, known now as the Quality Win Bonus. Even a mediocre QWB would drop Colgate to around 18th, that's how close the RPI is otherwise with teams in this part of the Pairwise.
14-15. Michigan / Notre Dame
These teams are hanging on the bubble, and, of course, susceptible to conference tournament "upsets," i.e. tournament wins by teams that wouldn't automatically make it but would get automatic bids. Michigan has a comparison loss to Western Michigan thanks to a loss to the Broncos at the GLI. Michigan could flip the Record vs. COP component by beating Michigan State in the playoffs, or WMU losing to Nebraska-Omaha in its playoffs. That would be an enormous help to the Wolverines, so keep an eye on that.
Notre Dame has to be careful of losing games to Maine and/or Providence the rest of the way, because those are teams right there on the bubble. Notre Dame also has to play Boston College in the final regular-season Hockey East game this weekend, which probably won't help. An upset there, however, will be a nice boost.
If Michigan gets in as a 4 seed, it would avoid Minnesota because of the intra-conference matchup thing. So then it could luck out and get a game in Cincinnati, relatively close to its base.
16-17. Providence / Maine
See above about Notre Dame. Wins there would be a big boost for these teams. Providence also has a chance to flip a comparison with Northeastern on the same grounds.
18-20. Yale / Minnesota State / New Hampshire
Any team below 20 is going to have a pretty hard time making the NCAAs at this point. Even these three have hurdles to overcome. Obviously, they just need to win games and hope teams above lose games.
* * *
We've always said that showing what the bracket would look like if the season ended today, is somewhat pointless, since the season doesn't end today. But it can be a fun exercise to see how things might look, and to identify any possible trouble spots. So here's a bracket going straight 1-16 based on today's Pairwise.
1. Boston College vs. 16. Atlantic Hockey tournament champ
8. Massachusetts-Lowell vs. 9. Northeastern
2. Minnesota vs. 15. Notre Dame
7. Quinnipiac vs. 10. North Dakota
3. Union vs. 14. Michigan
6. Wisconsin vs. 11. Cornell
4. Ferris State vs. 13. Colgate
5. St. Cloud State vs. 12. Vermont
The 8-9 matchup couldn't take place because of the edict to avoid intra-conference first-round matchups. In some prior years, you'd see 9 and 10 simply flip places, and that would be that. But the last two years, the committee placed a lot higher importance on geography than it had tended to do since the tournament went to 16 teams in 2003.
Consequently, it's conceivable you would see a lot of shuffling within bands. Maybe Wisconsin goes to St. Paul with Quinnipiac going to Bridgeport against Northeastern, and Cornell facing Lowell in Worcester. Something like that. We'll see.
The split of the conferences makes things a lot easier for the committee to tilt things geographically. You could have a Wisconsin-North Dakota first-round game and not worry about it, because those teams aren't in the same conference anymore.
You also have to start wondering what the committee will do to help Cincinnati, if anything. Michigan, Notre Dame? Those teams won't necessarily help attendance there. So maybe the committee throws its hands up with Cincinnati and just doesn't worry about it.