Greenie:
 
>>Since this thread began, I've heard four things:
 
>>1. The committee uses the PWR to seed the tournament. "Case closed."
 
Adam:
 
>more or less true -- I overstated the case at first .... but not really
>... since, as further explanation shows, the committee method and PWR
>are essentially identical.
 
        But not always, as I've pointed out in a separate message.
 
>>2. Now when there's a tie, the committee looks to the RPI.
 
>I don't think anyone ever said that.  Only when there's a tie within the
>comparison does the RPI get used to break the tie, and determine who
>wins that one particular comparison.
 
        They also use the RPI if a three-or-more team tie cannot be
broken on the basis of comparisons.  Say Rock University wins the
comparison with Scissors College, which wins its comparison with Paper
Tech, which in turn wins the comparison with Rock U, and the three are
all competing for a particular seed.  Then they rank the three teams
on the basis of their RPI.  (The committee did this to send Cornell
out West last year instead of Vermont or New Hampshire.)
 
>None of the points above are in conflict with each other .... they are
>all correct.
 
        Exccept for the "case closed" part.
 
>>process also showed that it's NOT purely based on the PWR.
 
>Well it is really .... unless you throw in the stuff about automatic
>byes and automatic bids ... which I didn't think was an issue ... I'm
>sorry.
 
        I think the point is that it's based on Pairwise
*comparisons*; "purely based on the PWR" to me means you always rank
teams in order of the total number of comparisons won, which you
don't.
 
>You also asked whether Clarkson could drop ... and I emphatically stated
>no!  I wasn't upset that you asked, only trying to re-state that it was
>more or less fact, not opinion.  Just like with Vermont last year.
 
        Given the numbers I crunched in the other post, Clarkson looks
pretty bulletproof, although there may be an outside shot given a
specific set of results--including four upset tournament
champions--and a jump in the RPI by Ohio State.  If the Knights do go
down, we should be able to answer the question with more authority
next week by looking at the possible outcomes of the conference
tournaments.  (Too many variables to play a convincing game of "what
if" by hand right now.)
                                         John Whelan, Cornell '91
                                     Official Scorer/PA Announcer
                                        U of Utah Ice Hockey Club
                                               <[log in to unmask]>
                      <http://www.cc.utah.edu/~jtw16960/joe.html>
 
HOCKEY-L is for discussion of college ice hockey;  send information to
[log in to unmask], The College Hockey Information List.