OK, My table is at best a first-order approximation which I can do in
a few minutes, and it does not account for the kind of analysis which
John has done.  That is, I only compare points already earned against
potential points, disregarding opponents for future games.  All that
means is that a team will finish no worse than the table says (VT
no worse than 11) should not be construed to say it may not finish
better (10th by John's analysis).  I should put that statement in a header
to that table so as no to be misleading.  The combinatorics for
a complete analysis would completely wipe me out.  It gets worse before
it gets better since there are so many "split seasons" (especially in
the ECAC).
 
  So what you missed John was the laziness on my part and no qualifying
statement to show how lazy I am.  What you say looks good to me.
 
  I was called to task a couple weeks ago for allowing Princeton to still
finish first, when by a similar analysis they couldn't.  X %
 
  So my analysis (such as it is) is conservative (that is, a team is
not precluded from finishing in a given position until long after it
could no longer accomplish that.  The rationale I have is that what I
do is quick and dirty (and incorrect) but can be done in a few minutes
AND by the next week will be correct at which time I will be wrong
with some other teams.  On top of all that, it will take someone
a week after then end of the season to figure out the ECAC pairings
anyway (look at the algorithm for that).
 
  Anyway, cheers to John, maybe he'll do a complete analysis for the
current standings.
 
-glen-
 
HOCKEY-L is for discussion of college ice hockey;  send information to
[log in to unmask], The College Hockey Information List.