Last Modified: 2011 October 9
Joe Schlobotnik / joe@amurgsval.org
URL for this frameset: http://www.slack.net/~whelan/tbrw/tbrw.cgi?2002/pairwise.020317.shtml
Game results taken from US College Hockey Online's Division I composite schedule
Now that the CCHA title game has been played, all the results are in for selecting and seeding the NCAA tournament field. Central to this process are the pairwise comparisons among the 27 teams who finished the season at or above .500 (in games against tournament-eligible teams, which doesn't include Findlay). Those are as follows:
Five teams received automatic berths for winning their conference tournaments: Denver in the WCHA, Michigan in the CCHA, Harvard in the ECAC, Quinnipiac in the MAAC, and New Hampshire in Hockey East. This leaves seven at-large bids, which will go to Minnesota, Boston University, Michigan State, Maine, St. Cloud, Colorado College, and Cornell, since these seven win all their pairwise comparisons with all other teams in contention for at-large bids. This gives a tournament field of:
| Team | lPWR | RPI | Comparisons Won | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| New Hampshire (H) | 11 | .6239 | DU | Mn | BU | MS | Mi | Me | SC | CC | Cr | Qn | Ha |
| Denver U (W) | 10 | .6259 | Mn | BU | MS | Mi | Me | SC | CC | Cr | Qn | Ha | |
| Minnesota (W) | 9 | .6241 | BU | MS | Mi | Me | SC | CC | Cr | Qn | Ha | ||
| Boston Univ (H) | 7 | .5991 | MS | Me | SC | CC | Cr | Qn | Ha | ||||
| Mich State (C) | 7 | .5940 | Mi | Me | SC | CC | Cr | Qn | Ha | ||||
| Michigan (C) | 6 | .5826 | BU | Me | CC | Cr | Qn | Ha | |||||
| Maine (H) | 5 | .5882 | SC | CC | Cr | Qn | Ha | ||||||
| St Cloud (W) | 4 | .5947 | Mi | CC | Qn | Ha | |||||||
| CO College (W) | 3 | .5824 | Cr | Qn | Ha | ||||||||
| Cornell (E) | 3 | .5793 | SC | Qn | Ha | ||||||||
| Quinnipiac (M) | 1 | .5032 | Ha | ||||||||||
| Harvard (E) | 0 | .4999 | |||||||||||
The top four teams in the nation receive first-round byes. UNH, DU, and Minnesota all win all comparisons against the rest of the field, so they get three of them. The final bye comes down to a choice among Boston University, Michigan State, and Michigan, with each team winning one comparison and losing one with the other two. In this case the ratings percentage index should be applied as a tie-breaker, giving the last bye to BU.
The assignment of bye seeds is actually a tricky thing. Boston University, as host of the East Regional, must play there. But according to the NCAA's Championship Handbook for Division I Men's Ice Hockey, #4 BU must be placed in the bracket so as to play #1 UNH in the national semifinal if they both advance. With Boston University as the second seed in the East, that means New Hampshire must be the first seed in the West, and either Denver or Minnesota must be sent East as a number one seed. (The documents we've seen (see a recent hockey-l post for an explanation) describing a policy change to reduce travel seem to imply that the preceding reasoning is unaffected, as they say that the "seeded" teams (the top four in the hockey tournament) should be placed into the bracket first, and travel-reducing considerations then applied to the remaining teams.) This would likely make New Hampshire fans unhappy, and the "reward" of being bracketed to play BU is likely offset in the Wildcats' minds by the penalty of flying out to the West Regional. Nonetheless, it appears to be the only course of action consistent with the NCAA's published guidelines.
Given that both Minnesota and Denver are more than 400 miles from both regionals, either team would be flying to either regional site, so there's not much to favor sending one or the other East in UNH's place, except that Minnesota might bring more fans to the West Regional in Ann Arbor Michigan. Denver, as the national #2, should get the more "favorable" seed, although it's not clear whether this would be the second seed in their own regional or the top seed in the other. So as a pure guess, let's suppose Denver goes East.
There are four teams from each region without byes, and since the NCAA is trying to minimize travel in accordance with the aforementioned policy, it seems safe to assume that the eight teams will all stay in their own regions, which gives us regionals of:
Western Regional | Eastern Regional | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Both regionals can be seeded directly according to the pairwise comparisons without any first-round intraconference matchups. In principle, the committee might overrule the Quinnipiac-Harvard comparison because the MAAC is a substantially weaker conference than the ECAC and thus the selection criteria are unreliable in making the comparison. However, moving Harvard up to the fifth seed would set up a rematch of the ECAC championship game with Cornell, so the ordering is best left alone. That makes the expected brackets
5W Michigan (C) 6E Harvard (E)
4W St Cloud (W) 3E Maine (H)
1W New Hampshire (H)--+--2E Boston Univ (H)
|
2W Minnesota (W) --+--1E Denver U (W)
3W Mich State (C) 4E Cornell (E)
6W CO College (W) 5E Quinnipiac (M)
The selection committee's actual choices will be known in the next hour or two, but with the reduced travel, the unpredictable decisions about regional placement are avoided. The only question is the identity of the #1 seed in the East, with the bracket above being about as likely as one in which Minnesota and Denver are reversed. One other possiblity might be to interchange Denver and New Hampshire, and keep all twelve teams in their own regions. However, as noted above, this would seem to require that the committee disregard the details of the process described in the championships manual.