Quote:
Originally Posted by Jeff P
But the data over the past seven years for 7f on the dirt at SAR tells a different story.
I'm using a logisitic regression model that I developed myself.
Early speed is one of the factors in the model and the coefficients for the factors in the model are unique for each track-surface-dist --
Meaning that if the same race had been run at say 7f on the Tapeta at Woodbine:
I very well may have come up with 15-1 for the same horse.
But here, at 7f on the dirt at SAR, the model came up with 9-2.
I Hope I got most of that out in a way that makes sense,
-jp
.
|
I totally love this.
I'm certain our methods are different, but I have an algorithm that creates profiles like this. I can run the algorithm against "all races" of a specific type or do it by sub categories like track, track condition, field size, etc... I think the public has general perceptions, but sometimes specific categories are different than those perceptions.