In article <[log in to unmask]>, Wayne Smith
<[log in to unmask]> wrote:
>Andrew wrote, in part..
>>I think an ideal system would be something like this:
>>
>>A team gains points for each win. More points for beating a better team.
>>A team loses points for each loss. More points for losing to a worse team.
>>Ties would probably average the two.
>
>Gee, you've just described RPI.
>
>Of course almost any system will do this, since What you've said
>basically says "strength of schedule" should count.
The question is where and how much. I'd have to look at the RPICH formula
again, but it looks like you can get lower your rating by beating a worse
team, and raise your rating by losing to a better team. Anything that
just
factors in "strength of schedule" (like RPI and associates) won't distinguish
between winning or losing to good teams, winning or losing to bad teams. My
'ideal' would factor that in.
--
- Andrew Brecher ([log in to unmask]) <insert disclaimer here>