Terry Long writes
 
> It is unlikely that those "3 or 4 other teams" John mentions could have
> beaten New Hampshire and also made a respectable showing against North
> Dakota.  So, it looks like there must not have been any "flaws in the
> selection criteria" as John asserts with the selection of NIagara.
 
Since J. Micheal Neal has already admirably pointed out the weakness
in this argument, I won't bother to recount the respectable
performances of the teams in question against UNH and NoDak over the
course of the season (or Providence's win over UNH in the Hockey East
regular season), but just point out that if this sort of anecdotal
reasoning were valid, one could just as easily state that Wisconsin
didn't deserve their #1 seed as a result of their loss to BC despite
an extra day's rest.  Does this mean that a system which would seed
Wisconsin below BC based on this season's results "must not" have any
flaws?
 
There is a valid point hiding under all that rhetoric, which is a
reason why I was not surprised Niagara pulled off the upset.  The
Purple Eagles have overachieved this season, pulling off wins like BU,
CC and RPI that are at odds with other data such as their losses to
Canisius and Western Michigan and their tie to Brown.  Perhaps this
makes them a better candidate for the tournament than a rating based
on their overall performance would indicate.  However, the NCAA's
criteria are also intended to favor "big game teams" by considering
performance against teams with non-losing records as a separate
criterion.  Niagara performs well in those games (and would do even
better if Canisius and Quinnipiac were not considered TUCs); their
KRACH rating based only on those games would be 67% higher than that
based on their entire schedule.  This means for instance that they
would also win the comparison over Rensselaer if the effects of
strength of schedule were factored in, despite having a worse overal
KRACH rating, on the strength of their performance against TUCs (big
game) and in the last 16 games (the NCAA also favors hot teams).  But
Mankato's performance against winning teams was still better when the
strengths of those teams are taken in to consideration, and Minnesota,
Providence and Colorado College would all win enough other criteria to
take their respective comparisons despite being weaker against TUCs.
Again, this is all assuming a particular method of accounting for
strength of schedule which appears at
http://www.slack.net/~whelan/cgi-bin/tbrw.cgi?rankings
 
> Also, it should be noted that Mankato was unable to beat UAH this year
> (1-0-1 at Mankato).
 
It should be further noted that games against Alambama-Huntsville DID
NOT COUNT towards selection this season.
 
                                          John Whelan, Cornell '91
                                                 [log in to unmask]
                                     http://www.amurgsval.org/joe/
 
HOCKEY-L is for discussion of college ice hockey;  send information to
[log in to unmask], The College Hockey Information List.