Just to illustrate how sensitive this outcome is, and at the risk of
offending either ECAC or WCHA folk or both, I checked Bill's "math" but
misread it.  I had UMD beating Wisconsin and Denver beating North Dakota -
the higher PWR teams, but lower seeds.  With changing just those TWO games
from one reasonable outcome to another (higher seed to "better" team per
PWR), Cornell falls from 14 to 17 and misses the tourney, while UMD rises
from 16 to 13 and makes the dance.

It gets better.  If UMD and Wisconsin were to TIE the WCHA consolation,
and DU beat NoDak, then UMD and Cornell would BOTH make it and the Gophers
would be out.  So the conspiracy theorists among us would suggest that
somehow UMD should blow a late one-goal lead in the consolation game, JUST
TO SCREW THE GOPHERS.

Obviously this is only one slice of a gillion different scenarios.  But
it's fun to ponder.

Mike
Bill Fenwick wrote:
> ---- Mike Abegg <[log in to unmask]> wrote:
>> Probably the bottom line is that there's no way for 4 ECAC teams to play
>> their way into the big dance.  SOMEONE will have to go 0-2 (or two teams
>> will go 0-1-1) against TUCs and that will be costly enough to RPI and
>> TUC
>> record.
>
> I got this from the ELynah forum (elf.elynah.com) and tried it out with
> the You Are the Committee script at
> http://www.collegehockeynews.com/ratings/yatc.php .  There is a way to get
> four ECAC teams to the NCAA's, although it will probably no longer be
> possible by the time you read this.  Every higher-seed has to win every
> remaining league tournament game, including the consolations, except for
> Cornell losing to Princeton in the ECAC semis and to St. Lawrence in the
> ECAC consolation.
>
> In this scenario, among other things, auto-qualifier Air Force flips seven
> pairwise comparisons and climbs to #13 in the PWR, and Cornell barely
> edges out Minnesota for the #14 spot and fourth ECAC spot.  Minnesota
> still makes it at #15, as only Bemidji State remains as an auto-qualifier
> to take the #16 spot.
>
> --
> Bill Fenwick
> [log in to unmask]
>