Please forgive me if I seem a bit slow to catch on to what has been
said on this subject lately.   I get the Digest form of the list,  and
sometimes it's just too long to get through.
 
     Anyway,  I guess I'm just AMAZED and APPALLED that the weighting
factors were applied at values of  0.25,  0.50,  0.25;  and,  that this
was done to mimick B-ball.   Let me get this straight:  the factors are
win-loss,  strength of schedule (opponents' win-loss),  and the opponents'
stength of schedule.   Have I missed something?
 
     The situations in B-ball and hockey are completely different.   B-ball
has a million conferences,  a large number of automatic bids,  and HUGE
differences between the conferences.   The difference between the Big Ten
and the Big Sky,  for instance is an order of magnitude larger than the
difference between HE and the ECAC.   In the B-ball situation,  schedule is
the all-important factor.   The primary challenge is to determine if
West Texas' 30-0 record MEANS anything compared to Providence's 19-11
in allocating the few remaining at-large bids.
 
      There is a lot more parity in hockey despite the fact that there
isn't a tremendous amount of interlocking play.  There is enough to draw
some basic conclusions.   The ECAC,  year in and year out is slightly
weaker than the other three conferences  --  but the difference is pretty
small.   It is thus ENTIRELY appropriate to choose the first order,  the second
order,  and the third order effects in a manner unique to the HOCKEY
situation.   And,  I would submit to you that it is intuitively OBVIOUS
that those should be (in decreasing order of importance):  own record,
strength of schedule,  opponents' strength of schedule.   (Remember the
Taylor series,  folks?)   We can argue about the relative magnitudes of
the factors,  but for me  --  0.50,  0.25,  0.25 makes a lot of sense.
It comes as no surprise that the rankings using the modified factors
seem to have more validity from a subjective standpoint.
 
     --  Dick Tuthill