In the East, we have two teams that we know will be in: Maine and SLU.  There
is no argument (nor any need for one) as to the placement of Maine, so we'll
put them #1E.  The rest of the picture is not as clear.  This is probably
not going to be as comprehensive as I'd like, but I hope it will still be
interesting.  I haven't touched the West since I still think all seeds are
up for grabs, especially in the event of an upset.  My goal is not so much
to predict the seeds yet as it is to show some information I've put together
that will likely be considered by the committee.
 
Re: my earlier posting on Wis-NMU-PC, it is still possible that PC will get in,
so we'll keep that on the back burner.
 
There are a lot of outside factors that may influence the committee's
decision on seeding.  These include:
* BU losing in the first round
* UNH going to the HE Championship (beating PC in semifinal)
* SLU winning the ECAC Championship
* Harvard finishing first in the regular season
* Harvard getting knocked out in the first round
* Jack Parker being on the selection committee
 
I call these "outside" factors because there is nothing in stone that says
these must play a part in the seedings - but undoubtedly, they will.  And there
is no way to know just how much of an effect they will have.  For example,
as Mike mentioned, Clarkson both won the regular season title and the playoff
title last year yet was seeded fourth - that is a clear indication that the
committee placed a heavier emphasis on the numbers.  They may do it again,
they may not.
 
I see five factors that are known to definitely have some sort of an effect
on the decision, from the past.  These are:
* RPICH rating.  Supposedly to have a bigger impact this year.  It did in the
  basketball seedings.  But that doesn't mean the hockey folks will go with it
  in the end.
* Division I winning percentage.
* Record against common opponents.
* Head-to-head record.
* Record against "teams under consideration" (we can only take an educated
  guess as to exactly which teams these are).
 
Prior to last year, the committee announced the seeds by telephone press
conference and took questions from people as to why certain teams did/did not
get in.  They were often very explicit as to the criteria used.  That's how
I came up with the above list.  I'm leaving out things like "available players",
etc. since I don't think they will play a part this time around.
 
In looking at how the teams could be seeded, I'm trying to look at all teams
that appear to be under consideration and comparing them head-to-head.  This
may shed some light on how certain teams stack up.  The RPICH rating is
supposed to be the key criteria, but other criteria also come into play for
teams that are "closely rated".  That's subjective.  I went with last week's
top 15.  They have changed slightly but won't affect the "Top15" records in
most cases, and then only very slightly.
 
The teams I'm looking at are: SLU (they're in, but where will they be seeded),
Clarkson, Harvard; BU, UNH, PC.
 
First, since two teams must be chosen from each conference, I think it is safe
to say that we need to choose from Harvard or Clarkson as the 2nd ECAC team.
          RPICH  DivI%  CommOpp%  H-to-H  Top15%
Harvard   .5220  .630   .750      .250    .214
Clarkson  .5160  .683   .696      .750    .500
 
Despite the difference of three places in the standings of RPICH, the actual
rating difference between these two is small enough that other factors will
likely come into play.  The key may be Clarkson's 1-0-1 record against Harvard.
Harvard has the slight edge in CommOpp, but I don't think it is enough, so
I'd pick Clarkson as the 2nd ECAC team.  (Might as well start off with an easy
one.)
 
Next, the second HE team from UNH, BU, PC.
     RPICH  DivI%  CommOpp   H-to-H   Top15%
UNH  .5401  .639   .667      .500     .300
BU   .5454  .697   .652      .500     .545
 
UNH  .5401  .639   .696      .250     .300
PC   .5352  .611   .520      .750     .500
 
BU   .5454  .697   .646      .500     .545
PC   .5352  .611   .580      .500     .500
 
Here's where it gets tough.  BU appears to have the edge over UNH - first a
slight edge in RPICH, but head-to-head is equal.  UNH has the slight edge in
CommOpp, but BU has a bigger edge in both win% and record against top teams.
 
BU gets the clear nod over PC, in my mind, so I put them as the 2nd best HE
team.
 
The other comparison is between UNH and PC.  This is particularly difficult.
Rating is close, and PC has the huge 3-1-0 head-to-head advantage as well as
the better record against top teams.  But UNH has a heavy CommOpp lead and
a slight lead in win%.  We have no idea which factors will get the heaviest
emphasis or if the UNH win over PC in the quarterfinals will sway them.
Many people seem to feel it will, but that's not necessarily true.
 
Here's why SLU may get the #2E seed.
     RPICH  DivI%  CommOpp  H-to-H  Top15%
BU   .5454  .697   .714     -----   .545
SLU  .5321  .726   .722     -----   .600
 
UNH  .5401  .639   .767     1.000   .300
SLU  .5321  .726   .700      .000   .600
 
No comparison was done with PC because I can't see PC beating out BU, but
UNH could beat out both BU and PC.
 
SLU has a lower rating than BU & UNH - but not by much.  SLU seems to have the
clear edge over BU.  It's closer with UNH because of UNH's head-to-head win
and better CommOpp%, but 1) they may not place as much emphasis on the win
since it came after SLU had played two tough games on the road and was SLU's
third game in four nights (speculation ONLY), and 2) SLU has a much better
win% and record vs top teams.  So, while I think that just winning the ECAC
championship and finishing 2nd in the regular season is NOT enough in my mind,
I do think that they stack up well enough against whoever the 2nd-best HE team
is.  So seeding SLU 2nd would be okay in my book and I think it will happen.
 
The rest of the seeds are so close it's tough to try to get inside the
committee's heads and determine what they will look at closely.  Suffice it
to say there is enough evidence to do almost anything else.  BU should be 3rd
or 4th.  Clarkson's rating is low enough that I cannot see them going to
other factors in possible comparisons with BU, PC or UNH.  PC could have
wrapped up a bid by beating UNH, but they still aren't out of the running.
And poor performances by NMU or Wisconsin, especially NMU, could open the
door for all three of BU, PC and UNH to get in.
 
And there are still the "outside" factors.  Quite a mess if you ask me.  I
invite anyone else to put forth a different interpretation...
---
Mike Machnik    [log in to unmask]   mikem@{beanpot,bubba}.ma30.bull.com