Quote:
Originally Posted by sjk
I think the best use of pars is to anchor the numbers to keep them from creeping. Periodically I update the par times to align them with the pars. I also update the pars to reflect the last three years actuals and normalize that to prevent number creep.
I have a par table with 15k+ entries and I will admit that I am now less than 100% at adding new conditions if they arise.
If anyone needs pars for Spt or GS I have them.
|
One of the stumbling blocks I run into is having a field in the data with a consistent and unique value for each class. Without that I can't automate a lot of the system. I wind up with different categories that are actually the same class and others with the same value that are actually different. I was doing some of it manually for awhile (in NY and when horses moved into stake company) but I threw in the towel. It was too much work for too little gain since I don't generate automatic plays or odds lines from my data. I still wish I had it though because there are many tests I'd like to run and I'd get better results with higher quality class info in the database. In some cases I substitute purses just for testing purposes, but that's fraught with peril.
__________________
"Unlearning is the highest form of learning"
Last edited by classhandicapper; 01-23-2019 at 11:19 AM.
|