>
> P.P.S. - can someone out there show a comparison of ranking systems vs. the
> actual results of the tourney over the last few years?  If it can be proven
> that they are more accurate in predicting winners than conference standings,
> then I would withdraw many of my objections.
>
Finally someone gets to the real meat of the matter....do these
techniques for ranking teams really have any predictive power.  If not,
then using them to justify placing teams in the field has no more
validity than any other method (unless you select one that correlates
negatively with results in the tournament).
 
Frankly, I've heard the "ECAC is weak" argument ever since HE split off
7 years ago.  So I pulled out my records and found the following:
 
# Championships:	ECAC	2
			HE	0
			CCHA	2
			WCHA	3
 
# Teams in Final	ECAC	5
			HE	2
			CCHA	3
			WCHA	4
 
Does this mean anything for this year's teams? - No.  But I will bet that
if we went back over the years, we would find that many of the
techniques (polls, computer models, etc.) which penalize Clarkson
because they only beat Dartmouth instead of beating the crap out of them
would not have predicted such strong results for the ECAC.  Sure there
are some weak teams in the league, but the ECAC's best have shown over
the years they can play with anyone.
 
Until strong justification for using any other technique is developed and
until each league is required to play a greater percentage of its games
outside the conference (4 non-league games / team in the CCHA....very
weak predictor), giving a minimum of two entries to each is not
illogical.  Remember, using the most definitive quality of performance
metric available (results within a league which plays a round-robin
schedule) would place only one team from each conference in the
playoffs.  All others could be shown to be "non-deserving" of playing
for even higher honors.
 
Jon Greene
RPI '82   Cornell '84
[log in to unmask]