I think you're on to something with this. As a way of comparing two
teams that haven't played, you could compare records against common
opponents. That might be prove inadequate--after all, what happens if
they haven't played any of the same teams? So you could add maybe some
other methods of comparison, and then weight each of those and thus
rank every team in college hockey. Neat!

There's some German guy working on something similar, I think he
published it recently in Mathematische Zeitschrift. Volume 29, pages
436-460; you should check it out.

Pat Carr


On Mar 25, 2005, at 12:08 PM, Nathan Hampton wrote:

> Cornell in the WCHA. Who are they most like?
>
> Try this comparison:
> Cornell vs
> Sacred Heart win by 7-1
> Yale win by 6-2 and 5-2
> Princeton win by 5-3 and 5-0
> St. Lawrence win by 1-0 and 3-2
> Maine in Estero win by 4-3
> BC in Estero lost by 2-4
> Overall scoring an average of 4.1 goals
> And allowing 1.88 goals against
>
> St. Cloud State vs
> Sacred Heart win by 9-1 and 3-1
> Yale win by 10-0
> Princeton win by 7-2
> St. Lawrence tie 1-1
> Maine in Estero tie 1-1 win in shootout
> BC in Estero lose 1-2 in double overtime
> Overall scoring an average of 4.57 goals
> And allowing 1.14 goals against.
>
> So, in having six identical opponents, St. Cloud State did better.
> St. Cloud State was 9th place in the WCHA.
>
> Nathan Hampton