database/numbers guys...tool to identify potential cheats?
I imagine that all the data that's collected by dedicated database guys could be used to suggest certain trainers for add'l scrutiny.
W% and WPS% over time. Then parsed out by a few criteria. I can think of a few random ones off the top of my head.
start with any notable delta in year-to-year W% and WPS%
then stuff like
1st and 2nd start off claim.
frequency +10pts BSF since last start
all controlled for conditions, pars, etc.
I wish I had a database. And some smarts.
|