Giant NC$$ Mess

[Greg Berge]

NC %s:

.716 Hockey East
.675 WCHA
.595 CCHA
.420 ECAC
.394 CHA
.162 MAAC

[jtwcornell91]

DeltaOne81 '03 wrote:
[Q]
Here's the quote I was talking about:
[Q]On other matters, the committee recently published its
guidelines for seeding the tournament, which include creating
four "bands" of four teams. Those bands will be the No. 1-4
seeds, respectively. There can be flip-flopping from region to
region within a team's seed, but there will be no more
flip-flopping of seeds within regions.[/Q]

from http://uscho.com/news/2003/02/26_006290.php
[/Q]
Keep in mind, though, that that's a secondary source.  What the committee actually said was
[Q]
The committee favors a "banding" model, whereby four groupings of four teams are created and assigned to the respective regional sites.  The Division I Men's Ice Hockey Committee favors the selection of teams on a national basis.  As a revenue-generating championship, it also favors the pairing of teams on a national basis.

The committee feels that the bracket was impacted negatively due to the implementation of the new policies for 2001-02.  It felt that it was not able to provide competitive equity in bracket, as it had been able to do in the past, with the top teams being from the West region and having to compete in the West region.  With the previous policy, the committee would have had the ability to move teams across regions.
[/Q]
in http://www1.ncaa.org/membership/governance/division_I/champ_cabinet/2003/Supp_13_MIH.htm
and
[Q]
Once the six automatic qualifiers and ten at-large teams are selected,the next step is to develop four groups from the com- mittee  s rankings of 1-16.The top four teams are No.1 seeds. The next four are targeted as No.2 seeds.The next four are No. 3 and the last four are No.4 seeds.
[/Q]
in http://www.ncaa.org/library/handbooks/iceHockey/2003/d1_m_ice_hockey.pdf

The official statements are a lot less explicit than the USCHO interpretation.  In fact, since they explicitly mention the placing the #1 seeds as a higher priority than avoiding conference matchups, and say no such thing about 2-4 seeds, you could read it to say they will swap seeds to avoid first-round conference matchups if they absolutely have to.

[Greg Berge]

Key is the phrase: "four groups from the committee's rankings of 1-16."  That's the committee's rankings, not the PWR (which is a model thereof).  At the end of the day, the committee could always set the brackets, then work backwards to set its "rankings."  That would be disingenuous, which makes it even more attractive.

[dadeo]

exactly

[Adam Wodon]

jy3,

There is no need to speculate. It's clearly spelled out what a good win is. Win vs. a team that finishes the season in the Top 15 of RPI.  The ambiguous situation you describe also cannot happen, since the Top 15 is determined via a snapshot of RPI after all pre-NCAA tournament games are complete.

[CrazyLarry]

People are getting very hung up on the possibility of WCHA intraconference matchups.  In the past, such machups were occassionally unavoidable.  Usually, the committee then explained their choice to allow such a matchup by noting that it was not a rematch of a conference tournament game.  If lots of WCHA teams get in, some of them won't have played in the WCHA tourney, and I'd expect the NCAA to allow such a game based on previous history.

Harvard getting in as a #4 seed would certainly increase the chance of seeing a CHA/MAAC 1st round opponent as a 1 seed, which seems nice.

Also, looking at the comparisons, I've never seen this much non-transitivity, so how the committee actually seeds the teams based on a PAIR-wise appraoch is probably very unsure.  PWR may not indicate the seeding bands very well.  There are a lot of comparisons won by teams much lower in the overall PWR, which are likely to be irrelevant because those teams never get compared directly.

Of course, this is what makes it fun!

[jy3]

Adam,
that is what i thought, thanks for clearing it up. :-)
I find it interesting that it is top 15 and not 16 since 16 teams make the tourny.