John Deetjen writes:
>I think it's interesting how Maine ranked as low as sixth on one of the polls,
>while BU, whom Maine is 2-0-1 against, is ranked second on the same poll.
>Or is it the computer ranking? I'm not sure, but Maine obviously has proven
>themselves against the Terriers this season.
 
Maine is rated sixth not in the polls, but in CHODR.  It helps to
understand the ratings to see why teams are ranked where they are.
CHODR, unlike RPICH, takes into account game scores.  The fact that
Maine has won many close games hurts them in CHODR in comparison with
other highly ranked teams.  This is because a team's CHODR rating is
the difference of its offensive and defensive ratings (as I understand
it).  Maine's defensive rating is the best, but because the difference
of its offensive and defensive ratings is less than five other teams,
including BU, Maine is ranked sixth by CHODR.
 
Perhaps this could be considered a limitation of CHODR, that it gives
more credit to teams who win fewer games but do win by large margins,
as opposed to teams that win more consistently but do not tend to blow
out opponents.  But that depends on what you expect to get from CHODR.
I don't believe CHODR is intended to specifically rate the teams from
1 to 44, at least I have not gotten that impression.  Rather, I look
at it as an indication of how teams have fared overall in the
offensive and defensive aspects of the game.  CHODR's predictions tend
to be correct more than they are incorrect, but the fact that they are
incorrect a reasonable percentage of the time shows that like any
other system, it is not infallible.  The CHODR ratings may often
translate into wins, but not always.
 
>This is why I personally don't put much faith in polls, which are based mainly
>on opinions, and not always the type of schedule a team plays. It's my feeling
>that if Maine and BU each had to play a steady diet of teams like Michigan,
>Colorado College, Minnesota, etc., they wouldn't be ranked as high as they are,
>mainly because they probably wouldn't do quite as well against these teams.
 
I don't know what basis you have for making this claim.  By RPICH,
both Maine and BU have strength of schedules that compare to the other
teams you mentioned.  That is a big part of the reason that they are
ranked as high as they are not only in the polls, but also in "the
only poll/rating that matters", RPICH.  There is no evidence to
dispute the counterclaim that Maine and BU *would* fare just as well
if they had played the competition you mentioned.
 
I agree with your lack of faith in the polls, but I also feel that
many people's criticism of polls is as unfounded as the polls they
criticize.  For example, some have lambasted Chaz Scoggins' poll that
was posted earlier by SteveW.  But their criticism is based on one or
two minor factors - BG should not be three spots behind MSU because BG
beat MSU last week, Clarkson should not be behind UNH because UNH beat
Clarkson twice earlier in the year, etc.
 
Meanwhile, in games played more recently than the last Clarkson-UNH
game, UNH has gone 3-3-0 to Clarkson's 5-1-0 - and in that stretch,
Clarkson has wins over two teams that people regard well, Brown and
Vermont (despite Vermont's apparent underachievement).  I can see why
Scoggins might rank Clarkson one spot higher, although I don't
necessarily agree with it.  (then again, maybe I do.)
 
It is very easy to take only the evidence that supports one's claim
and ignore all of the evidence against it.  That doesn't erase the
fact that that counter-evidence still exists.
 
>When you only get to play certain teams once or twice a year, you don't tend to
>get a huge idea of how well you can do against them, or the teams they have to
>play. This is probably one reason why Maine was undefeated for so long this
>year...
 
On the other hand, as you said earlier, Maine played BU four times and
went 2-0-2.
 
>...and why a seemingly harmless Princeton team, who I believe lost to Colby
>(a Div. III Maine school) earlier this year, could defeat Maine on Maine's
>home ice.
 
Actually, Colby tied Princeton.  But still, I believe you are making
the same selective choice of evidence here.  No team is as good as
their greatest win, or as bad as their worst defeat.  Otherwise, the
same claim could be made for BU's loss to Merrimack.  Both Merrimack
and Princeton had to play near-perfect games to post those wins.  It
doesn't mean that Maine and BU are undeserving of their status as two
of the highest ranked teams in the nation.  It just means that like
everyone else, they are not invincible.
 
>Essentially, the polls don't mean much, and I think should almost be thrown
>out. All it is based on is opinion, which opens the whole system up for bias.
 
This is true and is why it is a good thing that the polls do not
determine the seeds.  But similarly, all too often our own bias creeps
into our criticism of the polls.
 
>I don't think Maine is #1, and I'm one of the biggest fans of the Black Bears.
>I think Michigan is #1, especially since they play a much tougher schedule.
 
Again, the strength of schedule of both of these teams are remarkably
similar.
 
>Hockey East is getting stronger, so teams like Maine and BU may not be getting
>wins as easily as in the past. However, unlike the western conferences,
>especially the CCHA, many tend to forget the teams in the lower half of
>Hockey East and the ECAC.
 
I don't think this is the case.  As well, out West, both of the
conferences there tend to have teams that are struggling on the same
scale as the lower half of the Eastern conferences.  HE has UMA; the
CCHA has OSU.
 
And, I would submit there aren't any more consistent powers in the
West than in the East.  Several of the teams you mentioned -
Minnesota, Michigan, MSU - have their equals in consistent Eastern
powers (over the last few years), such as Maine, BU, Clarkson, and
Harvard.  Several of the other teams, like Denver and CC, are schools
that have just begun to shine again after years of struggling, like
Brown, Mass Lowell, and Northeastern in the East.  (well, until NU
began struggling again themselves.)
 
Plus, whatever you have done in the past doesn't really matter this
year.  Sure, tradition can help programs who have become accustomed to
winning to a certain degree, but if you don't win this year, you
aren't going to the show.  It is as simple as that.  LSSU is a good
example.  They could still make it, but it is going to take a
fantastic finish.  Right now, they are a middle of the pack team at
best.
 
I know that in the East, there is perceived to be a bias in the polls
from the West against Eastern teams.  But from talking to people, some
who vote in these polls, I think there is as much of an Eastern bias
regarding Western teams.  And it takes two forms.  Some voters will rank
teams from their own region more highly than voters from the other
region because of a bias towards their region's teams.  On the other
hand, sometimes the same voters will rank teams from the other region
higher because of the tradition you mentioned - for example, if I see
TeamA a lot and think they're not as good as some of the established
programs from the other region that I know about.  Meanwhile, the
people from the other region only see TeamA's scores and record
and think hey, this team has to be pretty good, even better than the
teams I see everyday (who likewise don't seem as good when you see
them as often).
 
The fact that all of these things take place is really why I don't put
any stock in the polls.  Voters are ranking teams they rarely if ever
see.  But again, fortunately it has no bearing on the NC$$ seedings.
So because of that, I just don't think it's worth getting worked up
over anything you see in the polls.  It is just the collective opinion
of a group of people, an opinion that may or may not represent the way
things really are.
---                                                                   ---
Mike Machnik                                            [log in to unmask]
Cabletron Systems, Inc.                                    *HMM* 11/13/93