On Tue, 15 Nov 1994, Chuck Sackett wrote:
 
> As I'm sure Charlie will point out, his rankings of teams by points per game
> shows how many points per game a team is averaging as far as wins and
> losses. . . . .not goals for or goals against.
 
ah yes, Charlie pointed this out... I read his original post too quickly...
 
>  For instance, a team that has
> played four games and is 2-1-1 will have 5 points in the league standings.
> Averaging that out over 4 games gives the team a per game average of 1.25
> points per game.  I think that it's a very effective way of comparing teams
> that have not played an equal amount of games.  Minnesota is at the top of
> the WCHA with 13 points and Colorado College is in third with 10 points.
> But, CC has played two fewer games.  This way, we can compare their progress
> more easily.
>
 
well, ok - but how valid is that stat given games against uncomparable
competition? I mean, if you have one team that is 3-0 (against 'easy'
opponents) and another who is 3-3 (but against much stiffer competition),
then who is stronger?
 
I grant you that it gives a number to point at, but I question the
validity of the conclusions you can draw from it...
 
-art