View Single Post
  #9  
Old 6th December 2004, 11:56 PM
KennyVictor KennyVictor is offline
Member
 
Join Date: Jan 1970
Location: Mt Tamborine
Posts: 574
Default

Quote:
Originally Posted by shoto
Interesting breakdown of results. What do you use for a 'base' measure for the class of the race in your ratings?


Each race gets a rating based on the horses which run in that race. It's an average of the expected ratings all the horses will return based on their previous runs. A sort of self governing system.
If none of the horses have run before I temporarily assign an average starting rate for the race just to keep the system happy. It doesn't matter in the short term because I only bet on races where I have a previous rating for at least half the horses. Then when those previously untested horses have run a couple of races agains horses for which I have some ratings I run the system backwards to create a more accurate starting rate for these 'unratable' races.

KV
Reply With Quote