Keith Instone writes:
 
>  If you want to read the paper where the original algorithm (for
TCHR) was written up, I'll gladly give you the reference.
 
I would like the reference.
 
I'm skeptical that any system, no matter how well thought-out, can
overcome the simple lack of statistical significance consequent of
there being so few interconference games. There is no good way to
compare teams between conferences. If I may advance a modest proposal -
just pick three teams from each conference and let them thrash it out
in the NCAAs -the truth will out.
 
As for the latter idea, I'd like to pose a heretical question - why
insist on choosing the best twelve teams to go to the tournament,
anyway? (Gasp!) The mindless selection of three teams from each
conference (seeded 1, 2, 3 in a conference-specific bracket) would
eliminate everybody's claims of regional or conference bias and ensure
a field with equal regional interest, and a little creative scheduling
could guarantee that conference rivalries would not be replayed until
the NCAA semi-final. And hey, in the *very* worst case, if the best
twelve teams in the country were in a twelve-team conference, the best
three would still make the tourny :-)
 
Yeah, I know, there would still be some controversy over picking (and
ranking) the three teams from a conference (look at Cornell-SLU last
year). But there would be a *much* better chance of doing the right
thing when comparing teams, with at least two (and probably more)
head-to-head games and many games vs. common opponents.
 
Anyway, who's thinking about Albany? I'm thinking about Princeton and Yale.
 
 
Greg
Boston
Let's Go Red!