Reply to thread

Message

[QUOTE="jncuse, post: 847973, member: 1969"]
That is the issue I sometimes have with these "as of now" approaches.    I have no problem with them from a theoretical standpoint, but there is always two problems:
 
1) Too little data to separate early on.
2) Bias given to those who they think will do well vs who has done better.   Sort of a half assed way of doing it.   They don't really want to punish teams who are likely to do quite well, and alternatively are conservative on teams that have surprised.
 
Not that Joe Lunardi is doing the bracket badly, but I find looking at an RPIforecast projected resume gives a different perspective of where a school may be tracking.
 
 
I just use RPIforecast to project where a team may be tracking- it takes year to date performance and puts in a 30 game picture that is similar to resumes on Selection sunday.   The big difference with this approach is 1) your conference weighs more heavily since it projects a conference record, and the strength of your conference matters (RPI, top 50 record), and 2) One or two losses right not a big deal if your Sagarin is good.   
 
[url]http://www.rpiforecast.com/index2.html[/url]
 
Of course the limitation is that it uses Sagarin going forward based on what you have done so far.   Sagarin has no bias, but if you disagree with the number it spits out, you will disagree with the projected resume.   I am not sure why we are #19 in Sagarin's predictor.   There will be some strange things but its a quick reference and like any model need to consider limitations.
[/QUOTE]

Verification

What is a Syracuse fan's favorite color?