1180
posted ago by whythis ago by whythis +1181 / -1

Guys..

I think found a key element to how dominion's voting algorithm works.. Specifically the "weight setting" they used and the most likely method they used to transfer votes. I would like some talented pedes to check my work and maybe check if other data sets conform to the same pattern.

Please note: I have only been able to check the data set listed here which is a small sample from Virginia: https://www.thegatewaypundit.com/wp-content/uploads/Virginia-Election-from-11-4-5am-to-11-7.jpg

Other areas may have used differing weight settings..

Here is a quick summary:

  • On the data I analyzed, they all used the same "weight".
  • The weight was applied to the "Total Votes".
  • The weight is a specific number: "0.984" for this data set
  • You can divide this number into your votes and get an accurate vote total.
  • The "accurate vote total" is a whole number (or close to it because of rounding) indicating this is extremely close to (or is) the weight setting.

From there I do not know how the algorithm works but I speculate like this:

  • Each 1 vote is actually counted as ".984".
  • This creates "unassigned" votes over time. For instance after 1000 votes passing through there will be 16 "unassigned" votes.
  • 1000 * .984 = 984 || 1000 - 984 = 16
  • From here you can assign the "unassigned" votes to your chosen candidate.
  • Using this method is much harder to detect as votes are not added "linearly" to a candidate which would happen if you weighted a candidate instead.
  • You can also add these votes "randomly" to the tally which also helps evade detection.

How to recreate this or look into it further:

1 - Make sure you are using lower and upper bounds. NOT THE ROUNDED NUMBERS.

What I mean by this is: the data numbers we have are rounded decimals to (2) decimals. That means the numbers most people work off of are not "precise". For instance "4,315,854.4950", "4,315,854.5049" and every number in between that will display as "4,315,854.50" when rounded to two digits. "4,315,854.4950" would be the lower bound of "4,315,854.50" and "4,315,854.5049" would be the upper bound. Because the weight is more precise than 2 decimals you need to make sure you are testing on the range not the rounded number.

2 - Look at the total votes and divide by the weight.

Lets use an example.. Lets say I had 100 votes and I weighted it .95. It would show our "total votes" as 95. In order to "undo" that calculation I would divide by .95. We would get 100 even.

What is extremely telling is that when we divide by our weight ".984" we get extremely close to a "real number" (no decimals) ---> This means this could be or is the correct weight setting.

I am probably being overly complicated for a lot of people but take a look at my table below:

Every number I tested with the ratio ".984" fell in between the upper and lower bounds.

https://i.maga.host/d8u5fV0.png

TLDR; 0.984 is the weight for total votes. Siphon's votes and then can add them back.

Comments (46)
sorted by:
You're viewing a single comment thread. View all comments, or full comment thread.
4
Super_degenerate 4 points ago +4 / -0

one digit away from 1984

1
RUFishing 1 point ago +1 / -0

they probably wrote the code as (1)984.