Just figured out how to add weight to the follower data!

Comment #3

hypersoniq — Sat, 23 Sep 2023 03:22:06 GMT

The first live test of the weighted follower system will be Monday with the PB and Tuesday with the MM. Paper play for pick 3 this weekend. I would clean up the old print lines that were commented out, but you never know when you want to break down the stats, so for now this version will be copied and archived, and a newer, leaner version crafted when time permits. Summary of changes... 1. The weighted pick was added, achieved by simply combining the full and most recent lis... [ More ]

Comment #2

hypersoniq — Fri, 22 Sep 2023 14:38:10 GMT

The first setup uses a hard coded 10% That means if there were 2,000 observations of the full list, there would be 200 for the recent list, and when weight is applied, it rings the list to 2,200 followers. Because it uses a percent, it is still auto ranging to the number of full observations, which varies per draw position and drawn number. Before undertaking the next step (converting the entire script) I will pass the percentage when calling each function. <br... [ More ]

Comment #1

hypersoniq — Fri, 22 Sep 2023 08:44:45 GMT

The goal is to generate both the full and recent lists, then combine them to add weight to the recents while maintaining the scope of all follower counts. Although the auto ranging feature was cool to figure out and implement, there needs to be a different mechanism that can select the recent list lengths based on the full list length. The lists vary, such as the number of direct followers is usually longer than the skip28 followers, so it is here that the ratio should be calculated a... [ More ]

Original Blog Entry: Just figured out how to add weight to the follower data!

hypersoniq — Thu, 21 Sep 2023 02:52:25 GMT

So, we have a list that captures ALL follower data for the last drawn number in each position from 7 different vantage points. Then I added the most recent followers that auto range based on their last number drawn. So here is how the weighting will work...

1. Keep the full combined list of follower data

2. ADD to it the combined recent data lists, this adding weight to the recent followers.

Step 2 acts like a theta weight in linear algebra, as it makes recent followers a feature , and then the theta is what gets changed to fine tune the overall data.

Maybe it works, maybe it does not. It is adding 36 lines of code to a 2,200+ line script and editing another 36 lines of code, so not that bad of a project. The bottom line is getting back to just one pick per game for the games it supports.

I have no idea how many recent picks are the right amount to add weight, perhaps a percentage of the follower count or the draw history length?

It must be narrowed down to just one before back testing can be coded, so hopefully there is a glimmer of hope in this new modification.... [ More ]