6683
Comments (326)
sorted by:
You're viewing a single comment thread. View all comments, or full comment thread.
6
grassshrimp 6 points ago +6 / -0

This assumes that the Country had only the one voting machine. I doubt that is the case.

4
alabasterCrown 4 points ago +4 / -0

there are a lot of assumptions, and this is a highly conservative estimate. I'm being upfront about that in my original post.

we could talk all day about other ways in which this could be more accurate, but I have more pressing things to attend to than generating PCA plots of what accounts for the greatest variation in voting deltas, like county urbanification, income, etc. add more voting machines, sure! it only helps our case!

2
grassshrimp 2 points ago +2 / -0

Thanks, I didn't intend my post as a criticism, hope you didn't take it that way. Trying to point out exactly what you just stated - this is a highly conservative estimate. Appreciate your analysis.

2
alabasterCrown 2 points ago +2 / -0

no offense taken at all! thanks!

2
2scoops2genders2term 2 points ago +2 / -0

Also, the math included 36,000 (denominator) people in the county. Is that total population or registered voters? And did everyone vote? Frequency & Delta goes up as the denominator goes down.

3
alabasterCrown 3 points ago +3 / -0

i'm playing fast & loose with the math. its total population ("the county itself has 36000 people" i thought i was being clear), which is a proxy for # voters (i know full well they arent the same thing, but they should map proportionally)

not everyone votes, of course. that however is unrelated to frequency and delta. re-read what I wrote. The 1/1000 becomes 1/500 when you turn 37 votes into a delta calculation, making it 74, thus, doubling the occurrence from 1/1000 to 1/500.

Perhaps the switching algorithm dynamically does it and it isn't based off the # registered voters (or county population) a priori, though. Who knows. that is just my hunch, that they just use a simple but reasonable metric to base their fraud on. like, if they had expert algorithm writers, they wouldn't be working at dominion, would they? wouldn't they be at google or something? or making video game engines? or brain-machine interface decoders?

they probably hire shitty programmers at dominion that probably wouldn't be able to code a decent dynamic heuristic for undetectable voter fraud. chinese people are CHEAP and wouldn't want to spend much money (source, work with cheap chinese)

2
sirtra 2 points ago +2 / -0

It gets better, i'm very familiar with the Georgia data. I'll let you do your magic but this will help you provide an even better estimate.

In Ware there was...14218 votes in total cast as per the hand-count audit.

Trump / Biden / Jorg / invalid writein / valid writein / overvote 9902 / 4174 / 117 / 16 / 6 / 3

It could be even better depending which piece of equipment he's talking about and how many they have. But purely on the 14,218 instead of the approx 36,000 we'd be looking at least double your guesstimate

1
alabasterCrown 1 point ago +1 / -0

ah thanks! yeah i don't know where or care to learn how to get those exact #s. So yeah. i'll just have to reiterate my mathematical guess is conservative, and clearly other corrections can simply and reasonably double it!

"and then we doubled it"