Perhaps one of those who crunch the numbers on PWR could answer this one
for me.
 
It seems to me that there is an ugly bias lurking in the "Common
Opponents" criterion of the comparisons, namely the number of times each
team has played the common opponents.  I'll cite the most extreme
example I can hypothetically construct off the top of my head: a WCHA
team and a HE team who's only common opponents are Boston University and
Alaska-Anchorage.  Say that each team has played the non-conference
opponent once (in, say, a holiday tournament).  With regular in
conferencing schedules, our HE team has played much stiffer "common
opponents" than our WCHA team, since they played a top notch team (BU)
three times while the WCHA team played UAA four times.  If each team has
won all of its games against Anchorage and lost all of its games to BU,
the WCHA team would ge a point in the comparison for its 4-1 record vs.
a 1-3 record.
 
It hardly seems like a fair comparison.  Is there any sort of correction
run in a case like this?  While my extreme example might not happen, I
suspect that this weighting problem is very common on a smaller scale.
 
J. Michael Neal
 
HOCKEY-L is for discussion of college ice hockey;  send information to
[log in to unmask], The College Hockey Information List.