Quote:
Originally Posted by deelo
If he's showing a lot of "no bias" days, it actually gives his days listed as biased more credibility in my eyes.
|
I agree with this. My thinking is still evolving after all these years.
There is no way to prove there was a bias. Even on days with 9 races to analyze it's possible that a series of seemingly biased results were just random fluctuations of form that happened to line up the same way. That can even be true when the horses come back.
On some level it's how you want to think about these things.
There's evidence of a bias.
There's evidence for how strong it was.
There's conflicting evidence (horses that ran well against the grain.
So where's the cutoff point for when you make a note and when you don't?
I don't know.
It depends on what you want your degree of confidence to be.
The tighter the standards the more often you will be right, but those are the ones everyone else sees and you might miss some.
The looser the standards, the less likely you miss any, but you might make some bad ones.
I usually put the sample size in my note and sometimes add in a degree of confidence. ex. (GR -
sample size only 5 dirt races, but clear cut good rail). Still, I'm sure I miss some and make some mistakes.
I don't think there are any perfectly correct answers to this. There are probabilities.