JohnD wrote, in part..
>When did they start using the formula?
>I always remember them having a selection
>committee just like the NCAA basketball tournament does.
 
It's been several years since they went from using polls and selection
committee judgments.
 
The current system has been tweaked over the past few years, but was
essentially unchanged from last year.
 
>Maybe it's time to change
>that formula so the best teams are competing come tourney time.
 
What would you change?  Just to bring ME-Hockey readers up to speed
(apologies to those that follow Hockey-L or other descriptions of the
selection process or those that fall asleep while reading) ...
 
o The regular season conference champions are awarded an automatic bid.
 
  If you were the best team, start-to-finish, in your conference,
  you get in.
 
o The conference tournament champions are awarded an automatic bid.
 
  If you were the best team in your conference at the end of the year,
  you get in.
 
o The remaining (4-to-8) at-large bids are awarded to teams that best
  measure-up by a method we have come to know as PWR or the Pairwise
  Rankings.  PWR compares each team to others under consideration.
  These comparisons include RPI, head-to-head, last-20-game record,
  record against teams under consideration (TUC) and record against
  common opponents.  I've simplified this portion a bit, but this shows
  the criteria, if not the mechanics involved.
 
This year, the selection of 12 teams was very easy, meaning that the
13th, 14th and 15th teams weren't close to the #12 and higher teams.  In
fact, there was so much difference in ratings that *MOST* rating systems
agreed with the set of teams selected.  Our Black Bears were not among
the #13-15 teams referenced above (#16, as I recall)
 
I like the current system (maybe you can tell!). :-)
 
It is not political, it is not one-sided wrt late-season (conference
tourney champion, last 20 games) and entire-season (regular season
champion, RPI, head-to-head and common opponents).
 
It takes direct (head-to-head) and indirect (common opponents) effects
into account, as well as overall play (RPI, regular season).
 
It discounts the effects of games against weaker teams (TUC).
 
It does not put too much stead in any one game (for example, BUs
quarter-final losses to MC were just 2 losses to a relatively poor team,
not the roof falling in).
 
One of the major advantages of all this is that the at-large bids are
determined by looking at teams "on the bubble" and how their performance
compares to the others "on the bubble".  Even though this doesn't
normally change the ranking from other methods such as RPI or HEAL that
simply come out with a single rating number, it makes the final
selections according to actual comparisons of teams.  This is *good*
IMHO.
 
There may be some room for improvement, but not necessarily in the what
I have described here.  The NCAA Men's Ice Hockey Selection Committee
(recent and present) should be highly commended for a world-class job
done.
 
Likewise, our Black Bears should be highly commended for the effort they
put forth, right down to the last minute of the season.  I'll miss Brian
White and the others that have played their last game, but look forward
to next season.  With 1 or 2 or 3 coaching positions opening up, and
Shawn always willing to listen, break out the Pepto-Bismol this
off-season.
 
cheers,
 
Wayne T. Smith
[log in to unmask]                          Old Town Landing
Co-owner of the College Hockey lists - Hockey-L/Info-Hockey-L/Hockey3