My response to the 'debunk' of Dr. Ayyadurai's Michigan precinct analysis
(media.patriots.win)
Comments (16)
sorted by:
DEFAULT COMMUNITIES
•
All
General
AskWin
Funny
Technology
Animals
Sports
Gaming
DIY
Health
Positive
Privacy
Who did this "debunking" analysis you replied to?
This is the video
https://www.youtube.com/watch?v=aokNwKx7gM8&t=486s
All he did was flip the chart 180 degrees and he thinks that proves something...
Had me fooled for a bit too. The easy mistake to make is to think that the precincts all are the same, left to right. They are not.
All he does is really make the inverse case for Shiva's work.
Frankly, unless I am overlooking something badly, his refutation is more of a confirmation.
I did change one thing in my comment on the video. Just woke up with this in my head and was still a bit groggy.
I've seen this video yesterday. Now I'm far from being a professional mathematician myself, but I did notice that the question he is asking is not the same as Dr. Ayyadurai. It could be that it was done in good faith and just a misunderstanding, or it could have been done in bad faith to obfuscate and show something that would result in a similar graph in Biden's favor. But he's not answering the same question as in the other video in my opinion.
I don't know anything about his education, so I wouldn't judge.
What I would like is to bring the video to the attention of our weaponized autists with background in statistics and data analysis to make a thorough comparison.
Dr Shiva did make a critical mistake in his first video analysis (mathematically proven). People are making similar mistakes in debunking it and indeed one of those is that the inverse result is typically going to simply mirror the other result when you're looking for flipping.
It's a very painful mistake because it means that any subsequent finding that might not be flawed is going to be mistaken for it.
Shiva's mistake shows the left aren't any better. Hardly any of them noticed that the graph is just Y = -X plus random fuzz.
I disagree. In the beginning he is describing the constraints for his data range, which may or may not be justified, and this is the part that should probably be examined closer.
The way I understand it, his analysis focuses on the discrepancy between people voting full republican ticket + Trump, compared to people voting for the presidential candidate freely. He is showing that the more republican voters there are in a precinct, it appears that Trump is getting less votes for himself, compared to the other Republican candidates on the ballot. Now if Trump was just unpopular, you'd see this depicted as a straight line in the negative percentage. Not a sloped line. Because if he is unpopular, he would be unpopular uniformly among the population.
His depiction of the data shows that for some reason either Trump becomes more and more unpopular the more Republican voters there are in a precinct, or that there's some kind of counting bullshitery going on.
The problem is I've proved their error and you can see it in the video (pause where they show their working out)...
https://thedonald.win/p/11Q8c5qCff/x/c/1BjYf4thsC
They are involved in other analysis that do look more interesting but generally I'd take Shiva with a grain of salt.
Not that I don't suspect fraud, in fact I'm certain they gamed the election but their first analysis was the result of a collective brainfart.
The fraud is almost certainly within a few percent so anything that stands out so strongly I'd triple check for an error in the plot.
So how come it's not Y=-X in the 0 - 25% range? There shouldn't be more "gap" between voters the more voters there are if the data is shown in percentages (for the rest of the area, I mean. The 25 - 100% range)
I'm no great expert, but I don't understand your proof. Please explain it to me like I am 5.
They did:
Consider you have...
Then you have...
In effect, Y is X and Y.
It should have been...
If you have Y = -X then you get a 45 degree angle.
For example if X is 10 then Y is -10. If X is 20 then Y is -20. You get a linear line going down.
But with the way you describe it, then the function would be the amount of independent voters as a function of the amount of republican voters? Which may not necessarily have such a relation (unless you assume everyone voted for someone and didn't leave the space blank but voted for everything else on the ballot).
Still, I don't understand why this isn't showing at the beginning of the plot, in the 0-25% area (timestamp 33:28 of the video).
In the graph shown on 41:12, there aren't that many precincts in the 0 -25% area, so you could argue that Y is indeed -X there, but it doesn't look like that in the previous graphs.
His scales for POTUS and Biden we're also different. He altered the scale for Biden to have the slope of his line appear the same as the plot for POTUS. That's an immediate red flag.
They're were also people on here claiming the same thing. They don't understand fucking math. You're 100% right, they both of the same graph because of the interference. Look at random counties across the US and you will see a relative flat line.
can you show us the "relative flat lines" in random counties across the US?