Better late than never. I finally finished my revised "Inside The College
Hockey Computer Rating" paper. It is 3 pages of figures and examples that
explain the theory behind TCHCR. It tries to answer questions like:
 
  --What does a connected schedule graph look like?
  --Where do the game outcome measures go?
  --What happens when there is a loop in the graph?
  --What effect do repeat-performances have?
  --How do ties work?
  --Where does the basis team fit in?
 
In fact, there are 6 figures, one for each question.
 
I can't merely email it to you, however, because the text part is pretty
worthless without the pictures. I do have, tho, a POSTSCRIPT version of the
paper available. If you have a Postscript printer near you, you may be able
to print it out yourself (it comes out fine on an Apple LaserWriter).
 
Another problem is that I can't email the Postscript file--out mailer eats
some of the long lines.
 
There is a way to get it, if you are able to ftp. Do an anonymous ftp to
andy.bgsu.edu and go to the pub/Hockey directory. There sits the file, called
"InsideTCHCR.ps".
 
I suspect that not too many people are able to ftp, know how to ftp, and
even want to read the paper, so let me know if anyone actually successfully
gets it printed out.
 
If someone wants to read the paper *real bad*, I can work out some other way
to get them a copy. (Uuencode, snail mail, etc.)
 
Back to our regularly-scheduled NCFF tournament bantering.... (^;
 
Keith [log in to unmask]