Dan Olsen provides the following starting point (pet peeve: I
know we're both Eastern folk, but I like to put the West on the left
because that's where it is on a map):
 
>East     West
>1. Clarkson (2)(E)   1.  Michigan (1)(C)
>2. New Hampshire (3)(H)  2.  North Dakota (4)(W)
>3.  Vermont (5)(E)   3.  Minnesota (6)(W)
>4.  Cornell (9)(E)   4.  Miami (8)(C)
>5.  Boston (11)(H)   5.  Denver (10)(W)
>6.  Michigan St. (13)(C)  6.  St. Cloud St. (12)(W)
 
        As an aside, let me point out that all the teams still in the
running for RS conference titles are alread on the list: Clarkson and
Cornell in the ECAC, UNH and BU in Hockey East, Michigan in the CCHA
and North Dakota in the WCHA.  Also, the elimination of RPI from the
ECAC RS race takes away the slim chance that existed for five ECAC
teams to make the tournament (if the RS and tournament championships
had gone to teams not in the top twelve).
        If we simply sent the two lowest West teams East and the
lowest non-host East team West, without regard to conference, we'd get
 
W1 Michigan (1)(C)         E1 Clarkson (2)(E)
W2 North Dakota (4)(W)     E2 New Hampshire (3)(H)
W3 Minnesota (6)(W)        E3 Vermont (5)(E)
W4 Miami (8)(C)            E4 BU (11)(H)
W5 Cornell (9)(E)          E5 St. Cloud St. (12)(W)
W6 Denver (10)(W)          E6 Michigan St. (13)(C)
 
        This has an intraconference matchup in the first round and
one, maybe two, in the second.  Now, as we know, the NC$$ likes to
avoid first-round interconference matchups and, where possible,
second-round ones as well.  There are two ways to do this: shuffle the
seeds within a region, and swap teams between regions.  I've been
using the former method (which seems fairer to me), Dan has tended
towards the latter.  If we can shuffle the seeds, switching Minnesota
and Miami in the above configuration removes all the intraconference
matchups but one second round game, which would require Denver to
upset Miami.  However, looking what the NC$$ did last year, it seems
that they are loath to have teams within a region be seeded in an
order other than that in which they're ranked.
 
        So instead if we want to model the selection committee
behavior, we should swap teams between the regionals, keeping the
seeding in order of PWR.  (The top two seeds in each conference seem
to be cast in stone, so it's only the 3-6 teams we're messing with.)
As Dan points out, swapping Miami and Michigan State will eliminate
all but one second-round matchup.  Swapping Minnesota and St. Cloud
would also do this.  (BTW, swapping Cornell and Vermont would take
away the first round matchup, but still leave at least two in the
second round, even if another pair of teams were swapped.)
        So the options for "correctly seeded" brackets with a minimum
of intraconference matchups would be:
 
W6 MSU/SCSU               E6 SCSU/MSU
W3 Minnesota/Miami        E3 Vermont
      W2 NoDak                  E2 UNH
W5 Denver                 E5 BU
W4 Cornell                E4 Miami/Minnesota
      W1 Michigan               E1 Clarkson
 
        Either plan involves one potential second-round WCHA game, but
Plan A (advocated by Dan) has that come about if the higher-seeded
Minnesota beats Michigan State, while Plan B requires St. Cloud to
upset Miami.  However, I think in this scenario, the NC$$ would go
with plan A, since a Minnesota/Michigan State game would be a bigger
draw in Grand Rapids than Miami and St. Cloud.
        Three more weeks and we get to do this for real.
 
                                        John Whelan, Cornell '91
                                        <[log in to unmask]>
        <http://www.cc.utah.edu/~jtw16960/jshock.html>
 
Cornell Men's Ice Hockey: 1996-7 Ivy League Champions
WE WANT MORE!  WE WANT THE ECAC!
 
HOCKEY-L is for discussion of college ice hockey;  send information to
[log in to unmask], The College Hockey Information List.