Click the Pic
Support Free Republic.
Become a monthly donor today.
Posted on 11/12/2020 2:30:20 AM PST by gattaca
MIT Scientists: Data Analysis Of Election Fraud In MI Shows 69,000 Votes Were TRANSFERRED From Trump To Biden! by Michael Suede • November 10, 2020
Odysee backup link BitChute backup link
Dr. Shiva Ayyadurai, who is a MIT trained data scientist and Fulbright Scholar, along with Bennie Smith a Democrat, who is a software engineer, data analyst and an election commissioner and Phil Evans, who is also a trained engineer and data analyst, discuss some very troubling findings after reviewing the election data coming out of Michigan. Dr. Ayyadurai is a US Senate candidate in Massachusetts, which is a contributing factor as to why he began looking into the integrity of our voting systems.
Dr. Ayyadurai’s discovered through a series of FOIA requests related to his own election that ballot counting machines don’t directly count ballots. Instead, they take an image of the ballot that is submitted to them, and then they tabulate the votes based on the ballot images collected.
Federal election statues dictate that all election records must be held for 22 months. However, Dr. Ayyadurai discovered that many states do not retain the ballot images produced by the machines. The default settings of the machines automatically store the ballot images, so this means election officials are manually disabling the image storage feature.
Because ballot images are not being retained in states like MA and MI, the election results cannot be meaningfully audited, leaving ambiguity as to the actual election outcome. The only way to audit the results of an election without ballot images would be to manually count the ballots, which is very difficult to do.
Dr. Ayyadurai learned that in 2001, a weighted race feature was added to Global Election Systems (GEMS 1.18.1) voting systems. This feature allows elections to be skewed to one candidate or another based on the settings of the software. This is not a hack, but a built-in FEATURE of the software. Why this feature exists in US ballot tabulation machines has never been explained.
Not only does the software allow for the weighting of election outcomes, but the way votes are stored and counted is also problematic. The systems store the votes as fractional decimals instead of whole numbers. This means the machines can be set to tabulate half a vote, which makes no sense at all, yet that’s how they were designed.
After voting “glitches” were reported in Michigan, Dr. Ayyadurai’s team decided to analyze the top four counties in MI for irregularities. They looked at data from Oakland, Macomb, Kent and Wayne counties. What they found was absolutely shocking. Data analytics show, at a minimum, 69,000 votes were TRANSFERRED from Trump to Biden in these four counties, in a pattern that is consistent with algorithmic weighting of a race.
In Michigan, there are two different kinds of ballots a person can chose to vote on. A person can chose to vote a straight party ticket or they can chose to vote for each individual candidate separately. They cannot do both. This means straight party ballots and individual selection ballots are tabulated in separate piles which produces two sets of data.
The fact that party ballots and individual ballots are separate in MI allowed the data scientists to compare the differences between how often people in a given precinct voted for Trump vs how often people voted for a straight Republican ticket.
For example, in this image, the bottom X axis percentage indicates how often people in a given precinct voted a straight Republican ticket. The Y axis indicates how often people voted specifically for Trump on an individual ballot minus the percentage who voted a straight Republican ticket.
So if a given precinct had 50% of people vote a straight Republican ticket and 50% of them vote for Trump on an individual ballot, we would see a dot at 50% on X axis and at 0% Y axis. Intuitively, we would expect most dots to fall along the 0% Y axis line.
In other words, if a given precinct tends to vote 60% straight Republican tickets, we would most likely expect to see about 60% of people in that precinct who vote on the individual ballots to chose Trump as well.
Instead, what the data shows is a clear linear relationship that starts at about the 20% X axis mark, where the more Republican a precinct is, based on straight party tickets, the less likely people who vote on individual ballots in that precinct are to vote for Trump.
Not only is this relationship inverted from what we would expect to see, the manner in which it displays itself is what really gives the fraud away. Notice the linearity of the slope. This linearity can only be the product of an algorithm. The perfection of the slope and distribution is what gives the game away.
This same slope is observed in three of the counties that were analysed. It shows up in early voting data as well as election day data. The pattern is consistent spatially as well as temporally. There is absolutely no way this is the product of anything other than an algorithm.
Consider what this data is telling us. This data is telling us that the more likely a precinct is to be Republican, the more likely it is the Republicans in that district voted for Biden over Trump, and they did so at a mathematically perfect rate.
This is an impossibility. People are sent to jail in fraud cases with far less evidence than this.
If this was a legitimate election, we would expect to see a scatter distribution along the 0% Y axis where the red arrow is, with perhaps either a positive or negative bias along axis depending on how much people liked or hated Trump.
Here’s what Wayne county’s results look like where no algorithm was detected. Wayne county is strongly Democratic, so its likely they didn’t bother rigging the votes here. There could still be cheating going on here, but this what we would expect all the graphs to look like. We can see voters here liked Trump a lot more than the Republican party.
When Dr. Ayyadurai began complaining on Twitter about the lack of transparency and the fact that ballot images were being deleted, he was attacked by an organization called The National Association of State Election Directors. This organization is funded by corporate interests with ties to the Rockefellers and other international business conglomerates. The NASED actually contacted Twitter and tried to get him banned. Keep in mind he’s a Senate candidate. This organization has been working with the election software companies involved in this fraud.
This software is in use far beyond the four counties that were analysed by Dr. Ayyadurai’s team. It explains how Joe Biden, a man who couldn’t get more than a handful of people to turn up at one of his “rallies” managed to get more votes than Donald Trump, a man who routinely held rallies that drew over 50,000 screaming fans who were collectively chanting “WE LOVE YOU!”
Biden did not win this election, and now there is ZERO doubt about that fact.
Changing the difference by 138,000 votes (subtract 69K from one and add 69K to the other). Twice as effective vote fraud then mere addition fake ballots with half to zero effort.
Great of “Dr Shiva” and his team to do that, but usually when you speak of whatever university scientists, you are referring to scientists on the faculty of that university. Not just one researcher on a team being a graduate of the school.
> he systems store the votes as fractional decimals instead of whole numbers
You do not use floating point numbers when integers will do the job. There are negative impacts to memory usage and performance, for starters.
Thank you! I hadn't seen it either.
> Consider what this data is telling us. This data is telling us that the more likely a precinct is to be Republican, the more likely it is the Republicans in that district voted for Biden over Trump, and they did so at a mathematically perfect rate.
It’s *these* data *are* telling us. Pet peeve.
“It has been posted under the 69,000 Michigan vote swap and 138,000 (the total effect) before that. Scroll back.
http://freerepublic.com/focus/f-bloggers/3904669/posts
http://freerepublic.com/focus/f-chat/3905002/posts
Maybe other duplicates were deleted.
Who cares for Gods sake. We are at WAR and if it’s posted several times and folks are seeing it for the first time who cares.
Well, this is the first time that I’ve seen it. You might not realize it, but some of us don’t sit on FR 24 hours a day. I also have a job that requires my attention and a life outside of FR. I’m grateful that I found it this time so that I can share it with fraud disbelievers that I know.
The machines that did the questionable ballot counting must still exist. Every machine can be tested for accuracy by simply counting against a set of pre-verified hand tabulated ballots. The software can also be analyzed by coders to verify possible manipulation of votes. Both checks would be hard evidence in any courtroom.
Not that science.
Let’s see why such an algorithm might be constructed in that way. The author of the algorithm wants to inflate Biden’s vote, but do it subtly. He doesn’t know what kind of precinct a given machine will be in, and in any event you can’t actually tailor the algorithm to the precinct. So it has to learn what kind of district it’s in, by detecting the percentage of straight Republican vs. straight Democrat ballots.After a certain point, the logic goes, “OK. Too many Republicans in this precinct, we gotta fudge the number of Trump votes down and the number of Biden votes up.
Or do you just count a certain number of “excess” Republican over Democrat votes, and flip one ballot from Trump to Biden?
Strange how vote flipping only goes in one direction.
Sorry but I appreciate it. I’ve been running two 18 hour days in a row. Today will only be 12
I get to my hotel and crash. I’ve been hearing about this but this is the first synopsis be caught.
It would seem that when such a flagrant failure of the system occurs, the courts should call for a new election.
In other words, if a given precinct tends to vote 60% straight Republican tickets, we would most likely expect to see about 60% of people in that precinct who vote on the individual ballots to chose Trump as well.>>> this is not a valid assumption as the reason to not vote straight party are varied. The fact that the results are linear is the tell as he explains. All software used by government must be open source. period. completely auditable.
bkmk
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.