Mr. Neal writes:
 
> Quick statistical note: Taking the top 15 vote getters in a top 10 poll
is not
> the same thing as taking a top 15 poll.  (demonstration snipped)
 
This is certainly, 100% true (in the case of a "top 1" poll it's even more
obvious ;-)  There are other statistical effects as well which vary
depending upon what number of teams is included, though I'd still argue (of
course ;-) that my basic point is retained: a top 10 poll is "elastic"
enough to stand for whatever reasonable number of teams you wish ranked.
 
Using top 12 might be the best of all possibly numbers, since there is then
a measure of "significance" attached to the dividing line on any one ballot
(essentially, "which 12 teams do you think have earned a trip to the
tourny").
 
-- Greg
 
HOCKEY-L is for discussion of college ice hockey;  send information to
[log in to unmask], The College Hockey Information List.