Goddamnit this analysis is wrong!! I wish it were this easy, but this is going to make us look dumb when it gets "debunked" by MSM. This really is an information war, and we need our credibility.
If you look at the python script, line 12, you can see that the script ONLY LOOKS AT DATA THAT SWITCHED FROM TRUMP TO BIDEN.
I am a software engineer. I don't use python, but you don't have to understand python to see what I mean here. If there was a check both directions, you would see another line like line 12, except with Biden and Trump switched places.
Please see the comment made by user fsvantesson for a more detailed explanation. He posted on the same thread, but it isn't getting the traction it needs.
Also kind of a side-note, I can tell the Pede (who meant well!) is a novice programmer by the poorly named variables. Makes the code difficult to read and no seasoned developer would write this way.
For the love of god please get the word out about this. It's getting quoted by journalists at this point.
Can some more Pedes go over the Dr. Shiva video showing his X-Y plot of mixed and straight ticket Republican voters too. Myself and another engineer went through Shiva's analysis and it does not add up at all to us. But maybe we missed something or misunderstood something. If it's incorrect we need to debunk it too so we don't look like fools. https://www.youtube.com/watch?v=Ztu5Y5obWPk
He's plotting down-ballot Republican share of vote on x axis, by precinct and Trump's share of the vote relative to the down-ballot Republican share on the y-axis. Normally, this should be a more or less straight line on the graph since you would expect Trump to either underperform, outperform, or be in line with other Republicans.
Indeed, this is the case for heavily blue Wayne County where he slightly over performed other Repubs in most districts. But in red counties it shows that as Republicanism increases, Trump underperforms more and more in a linear fashion.
This could be a natural phenomenon, but it defies an easy explanation other than algorithmic vote siphoning.
They could demonstrate that it's NOT natural by simply analyzing Biden's data too. Why aren't they doing this? Show Biden's data as a straight line and Trump's data with the downward linear slope. That would convince me. They used Trump's Detroit data as the comparative data, that's good too but not enough.
If they are correct then the swamp is waaaaay deeper than I ever thought. They are essentially saying that America does not have a free election at all. That's some scary shit.
SPEZ: Ok I learned more, any chart with that shape of line screams unnatural. I think this is real
The data is represented as a share of the votes, so you have total votes and a number from 0-1 representing how many go to each candidate Your idea would just flip the graph, but the names on the graph would also be reversed
Yes thanks I get it now. Been looking into it. I think they really found something
It wouldn’t be a straight line, it’d be the inverse of the too perfect drop off line.