Here's an issue I am running into running a regression.
I have 3 factors and the finish position. I'm running a regression trying to get the correct weighting of each factor to produce the most winners.
What I think the regression is trying to do is weight the 3 factors to get as close to matching the entire order of finish instead of just trying to maximize the winners.
I don't really care about being more accurate about who is going to finish 6th, 7th, 8th etc.... I just want to maximize winners.
Is there anything that can be done to trick it in some way. I was thinking of keeping all the winners as a 1 and everyone else as a 0 or 99.
(Also, since each field has a different number of horses that could also be confusing it)
__________________
"Unlearning is the highest form of learning"
Last edited by classhandicapper; 12-01-2022 at 04:37 PM.
|