Quote:
Originally Posted by 46zilzal
NO ONE can predict the 4th horse, long term...NO ONE
|
Really?
You do realize it's possible to model the likelihood of a horse finishing exactly first or exactly second from the data using logistic regression and other statistical modeling techniques right?
I know that it's possible because I've actually done it.
When modeling the likelihood of a horse finishing exactly first one of the requirements when preparing the data for analysis is to create a column for your dependent variable that contains a 1 for the winner of each race and a 0 for all of the non-winners in each race.
I can't see any reason somebody couldn't prepare their data for analysis by creating a column for the dependent variable that contains a 1 for the horse in each race that finished exactly fourth and 0 for all of the other non-fourth place finishers in each race.
And from there run a logistic regression analysis on the data to generate an accurate estimate for the likelihood of a fourth place finish.
That said, if you don't want to go down the logistic regression rabbit hole, you can get a pretty good ballpark estimate just by looking at the data.
Attached at the bottom of this post is a text file that contains two cut and pastes showing what I have in my current database from July 1 2022 through yesterday July 26 2022:
#1. The Fastest Horse in the Race, or horses that are rank=1 for one of my time decayed final time speed fig factors with the data broken out by actual finish position.
#2. The Fourth Fastest Horse in the Race, or horses that are rank=4 for the same time decayed final time speed fig factor with the data broken out by actual finish position.
Note that the rank=1 horses won about 30% of the time and and finished exactly fourth about 12% of the time.
Also note that the rank=4 horses won just 10% of the time but finished exactly fourth almost 19% of the time.
Imo, it's a big enough sample that I feel confident saying that the fourth fastest horse in the race is about 19% likely to finish exactly fourth and about 1.5 times more likely than the fastest horse in the race to finish exactly fourth.
-jp
.
