As opposed to previous predictions of what the NCAA *will* do
this year, this post speculates on what the NCAA *would* do if they
used a *different* selection process. In this scheme, teams within a
conference are ranked by their finish in the regular season and the
conference tournament. Looking at the final standings (only the top
four are neccessary), we find:
Tournaments:
1 NoDak 1 Michigan 1 Cornell 1 BU
2 Minnesota 2 MSU 2 Clarkson 2 UNH
3 CC 3/4 Miami/BGSU 3 RPI 3 [BC]/[Lowell]
4 St. Cloud 4 Princeton
Regular season:
1 NoDak 1 Michigan 1 Clarkson 1 BU
2 Minnesota 2 Miami 2 Cornell 2 UNH
3 St. Cloud 3 MSU 3 Vermont 3 [Providence]
4 CC 4 BGSU 4 RPI 4 [Merrimack]
[Teams in brackets finished the season with overall losing records are
are not eligible.]
The various conference champions (North Dakota, Michigan,
Cornell, Clarkson and BU) are all automatic bids, so we give them the
top seeds and rank them by individual pairwise comparisons. Michigan
beats everyone, Clarkson all the rest, North Dakota the other two, and
BU beats Cornell to get the last bye.
Okay, now we need to fill the seven at-large bids; we take the
second place teams from the WCHA (Minnesota), CCHA (Michigan State and
Miami), and Hockey East (New Hampshire), along with the third place
teams from the ECAC (RPI and Vermont). Minnesota beats all of the
others to claim the #6 position, and are replaced by the WCHA's
third-place finishers (Colorado College and St. Cloud). Vermont now
beats the remaining teams (CC, SCSU, MSU, Miami, RPI and New
Hampshire) to grab #7. They would be replaced by fourth-place
(regular season) RPI, but the Engineers are already being considered
as the third-place team in the tournament. Of the six teams now being
compared, UNH beats all the others for #8. There is no Hockey East
team to replace them, since all of the others have losing reconds.
(This is practical on my part, since the comparisons have already been
worked out for all teams above .500, but it's reasonable to expect
that none of those teams would win enough pairwise comparisons to
qualify anyway.) Of the five teams left, Miami beats the others to
claim #9. They would be replaced by third place (regular season) MSU,
but the Spartans are already in the pool as the second-place finisher
in the tournament. Of the remaining four, Michigan State wins all the
pairwise comparisons, taking #10 and being replaced by Bowling Green,
who finished just behind them in both the CCHA tournament and regular
season. Colorado College beats the other three teams left in the pool
(SCSU, BGSU and RPI) to grab #11. They would be replaced by
fourth-place (RS) St. Cloud, who are already in the pool as
third-place tourney finishers. Finally, St. Cloud beats the other two
teams left in the pool to get the 12th and last spot.
So the overall rankings by this scheme would be:
1 Michigan
2 Clarkson
3 North Dakota
4 BU
5 Cornell
6 New Hampshire
7 Minnesota
8 Vermont
9 Miami
10 Michigan State
11 Colorado College
12 St. Cloud
The byes are the same, and the main differences are that Vermont is
the odd man out in the East, and St. Cloud replaces Denver becuase
they finished ahead of them in the WCHA. Cornell placing ahead of UNH
means that the UNH-BU matchup is removed. The raw brackets, if this
alternate selection process were used, would be
W1 Michigan E1 Clarkson
W2 NoDak E2 BU
W3 Minnesota E3 Cornell
W4 Vermont E4 New Hampshire
W5 Miami E5 Colorado College
W6 Michigan State E6 St. Cloud
It just so happens that under this ranking scheme, there's only one
second-round matchup, between Minnesota and North Dakota. Removing it
would mean switching Minnesota and Vermont, who were seeded as they were
because Minnesota beats UVM in a comparison
Minnesota vs Vermont
RPI 0.5809 1 0.5786 0
L20 13- 6- 1 0 12- 5- 3 0
TUC 12-10- 1 1 10- 9- 2 0
H2H 0 0
COP 6- 2- 1 1 2- 1- 0 0
============================================
PTS 3 0
since it's clear-cut, it would be rough to make Minnesota play
Michigan. It would depend on the NCAAs priorities in this imaginary
world.
To consider an even more imaginary world, suppose we use the
alternate scheme I proposed a while back for the tournament, where the
NCAAs are a three-week double-elimination tournament, where everyone
who didn't win their conference tournament starts with one loss
(representing the loss that eliminated them from their conference
playoffs). In this scheme, the first round is hosted by the four
tournament champions, and the regions are kept separate until the
final round. Here's how such a tournament would go, if every game
were won by the team with the higher Ratings Percentage Index:
At North Dakota: NoDak plays MSU Friday and Miami Saturday (the lower
of the two visiting seeds has to play Friday when the host is fresh).
According to RPI, North Dakota would sweep and be the only team to
advance to the regionals.
At Michigan: Michigan hosts Colorado College friday and Minnesota
Saturday. They sweep to advance to the West Regional.
At Cornell: Cornell beats SCSU (who replace the missing third HE team)
Friday to ensure a trip to the regional, but New Hampshire defeats
Cornell Saturday to join them there.
At Boston University: UVM and Clarkson beat BU on consecutive nights
to advance to the regionals and eliminate the Terriers.
Then we go to the Regionals:
In the East, we have a field of four; we would reseed by individual
comparisons to give Clarkson, Vermont, New Hampshire and Cornell,
except that to prevent a rematch of the Cornell-Clarkson game from the
ECACs, Clarkson plays UNH and Vermont plays Cornell. Clarkson and Vermont
advance to the Eastern championship game, which Clarkson wins
In the West, Michigan and North Dakota play a three-game series to
decide the West champion; Michigan wins.
Finally, Clarkson and Michigan face off in a best of three in
Milwaukee to determine the national champion.
Okay, back to reality.
John Whelan, Cornell '91
<[log in to unmask]>
<http://www.cc.utah.edu/~jtw16960/jshock.html>
Cornell Men's Ice Hockey: Back-to-back ECAC and Ivy League Champions
HOCKEY-L is for discussion of college ice hockey; send information to
[log in to unmask], The College Hockey Information List.
|
