Saving Data to Models
 

When saving race data for models is it recommended to include hidden horse if the hidden horse did not win? I have info by track, distance and surface but always included the hidden horses. Just looking for ways to make the data more telling.
Thanks,
Ken

Yes Ken if you don't then you dilute your models. Also I don't include mdn,2yr old or Graded races in my models.
Tim

Ken
When I develop my models...every horse is included not just my selections.
Unless it is a first time maiden...then I throw it off the board so it does not post to
excel model.
I know most folks seperate there selections and throw the others out our down into
contenders or non contenders
I am pretty sure that when you throw out horses so they do not post to the excel model
it changes the values of the horses left behind same thing if you throw them into
condenters not winners.
I am not saying my way is the only highway...I look at it like all the horses in a race
effect the out come....so why throw them out.
Now let me reiterate this is the way I do it...I am not suggesting that this is the only
way to heaven. Everybody can choose to build there models as they please.
Good luck
Don

Oh by the way Ken
Just so you know Tim is a much better handicapper than me.
He has taught me pretty much all I know.
So if you take advice I would highly suggest you listen to Tim...

Don

Thanks for the feedback,

I was rereading the follow ups and seeing forms to track last 20 races used both Hidden and Unhidden data on winners. Hiding allowed horses to move up in categories just as your suggesting to better evaluate. Then I was wondering if they should be left out of the downloaded data.

I don't recall reading anything about models specifically in the follow ups, at least in 70-88, maybe I missed it? Is there literature recommended that discusses the topic you are aware of ?

Ken

Towards the end of the methodology, Howard was against models so there wouldn't be much about them in the last 3 years of the follow ups. Ted has the Brohamer Model book online, which is very different from the published MPH. I think you can find it here:

https://paceandcap.com/forums/showthread.php?t=6156

The Doc wasn't against models he was against folks modeling too many factors so as to confused themselves.
Tim

I agree with you on this Tim. I find that when I use modeling I find that a Track/Distance/Type of surface and Class. One can come up with various variable being
like a 90% others only being like 20%.
And where a variable like POH f3 is 80% in one model it may be 10% in another model.
So my rule of thumb I take the highest variables like if top is 100% and another is 80%
I will use both...but if you go to low % percentage it is nothing but noise.
Then you still have to do some regular handicapping looking at Running Style ESP etc.
and prayer every now and then or horses color etc.
Don

Very helpful, I have tended to use to many variables in my analysis for sure and that leads to my confusion for sure! Thanks for the article it was an interesting read, the continued improvements to RDSS has changed some of the most dependable variables like FX over FW but the article gave me a better sense of how to use leading factors. Still working to improve my use of Supplementary Factors to fine tune or support my selections after Primary factors. I also never gave enough focus on %Med. Thanks for sharing info!

Ken

simply used the e/l model for each distance, each track...WORKED well for years