Yes, lots of people have done it, but met with hostility every time. I saw one pede with a master's degree in data analysis post about 40 comments, with graphs, math, formulas, everything to explain it, but he just got completely downvoted and then deleted everything he had posted, because he didn't have the energy to receive constant abuse for posting the truth.
I had independently done the same work, and shared my comments. I will not bow down and delete the truth. If this site overwhelmingly chooses ignorance over truth then that's just a sad reflection on us. I even consider the possibility that the people pushing Shiva are our ENEMIES and are doing it to WASTE our time.
Here are some explanation comments which go over the math:
And the post that all of those comments are posted under made an attempt to explain it all clearly.
Sadly there are many Common Core victims here too, that are incapable of grasping the math. :'-( I can understand the feeling. It definitely isn't easy to grasp at first if the viewer has no background in math/statistics. But I have made countless efforts to explain it clearly.
I have encountered a dozen other pedes who have done work to attempt to explain these concepts to the masses. The truth is slowly getting out there. But the misinformation is still going to hurt us tremendously since it's being spread more than the truth, and making us look retarded. I am really despairing about the situation. It feels like fighting a tide of misinformation to try to get us to focus on the ACTUAL fraud such as the Dominion machines flipping votes all over the fucking place, and Trump most likely WINNING the states that would have a hand-recount. THAT is where to focus.
Absolutely. Those machines were caught flipping votes in like a dozen locations. Think about all the locations where nobody noticed that the machine totals don't match the paper stacks, due to not having hand-counted their ballots. The ones we know about are just the people who said "wait a minute, this is weird!" and did a manual check.
There has also been massive "multi-scanning fraud" where Democrat shitheads took boxes of Biden votes and let the machine tally them multiple times, to boost his numbers. A hand recount is necessary to find those situations.
As for Shiva, he's hurt us badly with his clownish math. :-(
I had a little bit of a look at them during the early stages, and I read the code that was being used. The code itself was an absolute mess. And it took the truncated (only ONE decimal, rounded) "vote share percentage" numbers from the Edison election data and then compared the shift in the "vote share percentage" to the "actual percentage" shift in the total votes for each candidate. But because the number published in the Edison data is truncated, I warned that the analysis method is most likely just misusing the rounded number. Whenever a number is rounded, you lose precision. His "analysis script" (again, very poorly written) was interpreting those rounded percentages as "stolen votes".
It's as if I say "You have 100003 votes, which is 63.6% of the total share" and then I say "You have 100007 votes, which is 63.6% of the total share" (while in reality it may be 63.64% (meaning not yet enough to round up to 63.7%)). His script interprets that lack of decimals as a "shift in votes", even though it really doesn't tell us anything. The vote percentage number is rounded in the Edison data because it saves filesize that way. Instead of saying "Trump has 69.47473747347364613558% of the votes", the Edison data simply rounds it and says "Trump has 69.5% of the votes". It is extremely common to do that in programming. The percentages are calculated in realtime in the code that spits out the Edison data files. It is just included in the JSON files for convenience. But the real percentages can be recalculated at any time, in the same way that the pede script author did. His problem however, was that he interpreted the difference in the "Real, multi-decimal percentages vs the rounded percentage" as fraud, while it's really nothing more than a way to save filesize for an insignificant number. The only number that matters in the vote data files is the total vote counts for each candidate (ie 3000 Trump, 2000 Biden, 5000 Total). The percentages assigned to it for display purposes are "metadata" that don't matter at all. With the vote numbers, we can always calculate the actual percentage with all trailing decimal numbers if we truly want to.
So, no, I haven't gone in depth into that analysis. I decided not to do anything when I saw that the author was misinterpreting the intentionally missing decimal numbers from a rounded percentage as "election fraud", while his code was an absolute mess that was almost unreadable.
His code is quoted in my comment. As you can see, it's an absolute mess.
What I will say, however, is that his code will definitely catch any sudden massive "glitch" swings such as "-7000 trump, +3000 biden". But such detection code would have been written much better by just writing code that looks at each successive vote total number (ignoring the "rounded percentage" red herring) and just ensures that it's always greater than or equal to the previous number (no regressions/subtractions). However, even such fixed code would not detect the votes that were NEVER published as Trump votes. In other words, if the machine originally received 5000 ballots and 4000 were for Trump, but they published online "3000 biden, 2000 trump", then no amount of analysis of the Edison data will ever find that fraud. The only way to find it is to look at and hand-count the actual paper ballots that the machines lied about the tallies for.
And by the way, we must pray that paper ballots are always kept (not destroyed), since there are plenty of reports of people multi-scanning boxes of ballots over and over (since ballots don't have any uniqueness identifiers) into the machine, to count Biden votes multiple times. The only way to catch THAT will be to have the original papers, or to have the image files (scans) of the paper ballots and look for identical multi-scanned ballot images (same style and position of checkmarks etc).
We have a lot of work ahead of us but I am sure that recounts will continue to find massive fraud. :-)
Wait we disproved Shiva? The MIT guy?
Yes, lots of people have done it, but met with hostility every time. I saw one pede with a master's degree in data analysis post about 40 comments, with graphs, math, formulas, everything to explain it, but he just got completely downvoted and then deleted everything he had posted, because he didn't have the energy to receive constant abuse for posting the truth.
I had independently done the same work, and shared my comments. I will not bow down and delete the truth. If this site overwhelmingly chooses ignorance over truth then that's just a sad reflection on us. I even consider the possibility that the people pushing Shiva are our ENEMIES and are doing it to WASTE our time.
Here are some explanation comments which go over the math:
https://thedonald.win/p/11Q8XO2YxF/x/c/1BjFNXpNSu
https://thedonald.win/p/11Q8XO2YxF/x/c/1BjFgKmo6q
https://thedonald.win/p/11Q8XO2YxF/x/c/1BjFgLvCMh
https://thedonald.win/p/11Q8XO2YxF/x/c/1BjFkvm4Ir
https://thedonald.win/p/11Q8XO2YxF/x/c/1BjFNYzQXq
And the post that all of those comments are posted under made an attempt to explain it all clearly.
Sadly there are many Common Core victims here too, that are incapable of grasping the math. :'-( I can understand the feeling. It definitely isn't easy to grasp at first if the viewer has no background in math/statistics. But I have made countless efforts to explain it clearly.
I have encountered a dozen other pedes who have done work to attempt to explain these concepts to the masses. The truth is slowly getting out there. But the misinformation is still going to hurt us tremendously since it's being spread more than the truth, and making us look retarded. I am really despairing about the situation. It feels like fighting a tide of misinformation to try to get us to focus on the ACTUAL fraud such as the Dominion machines flipping votes all over the fucking place, and Trump most likely WINNING the states that would have a hand-recount. THAT is where to focus.
So you still think the Dominion software flipped votes to Biden but that Shiva’s analysis of the “algorithm” was incorrect?
Absolutely. Those machines were caught flipping votes in like a dozen locations. Think about all the locations where nobody noticed that the machine totals don't match the paper stacks, due to not having hand-counted their ballots. The ones we know about are just the people who said "wait a minute, this is weird!" and did a manual check.
There has also been massive "multi-scanning fraud" where Democrat shitheads took boxes of Biden votes and let the machine tally them multiple times, to boost his numbers. A hand recount is necessary to find those situations.
As for Shiva, he's hurt us badly with his clownish math. :-(
Well shit, what about the other data analyses the people on this site have done? I think there was a PA data dump, and michigan too.
I had a little bit of a look at them during the early stages, and I read the code that was being used. The code itself was an absolute mess. And it took the truncated (only ONE decimal, rounded) "vote share percentage" numbers from the Edison election data and then compared the shift in the "vote share percentage" to the "actual percentage" shift in the total votes for each candidate. But because the number published in the Edison data is truncated, I warned that the analysis method is most likely just misusing the rounded number. Whenever a number is rounded, you lose precision. His "analysis script" (again, very poorly written) was interpreting those rounded percentages as "stolen votes".
It's as if I say "You have 100003 votes, which is 63.6% of the total share" and then I say "You have 100007 votes, which is 63.6% of the total share" (while in reality it may be 63.64% (meaning not yet enough to round up to 63.7%)). His script interprets that lack of decimals as a "shift in votes", even though it really doesn't tell us anything. The vote percentage number is rounded in the Edison data because it saves filesize that way. Instead of saying "Trump has 69.47473747347364613558% of the votes", the Edison data simply rounds it and says "Trump has 69.5% of the votes". It is extremely common to do that in programming. The percentages are calculated in realtime in the code that spits out the Edison data files. It is just included in the JSON files for convenience. But the real percentages can be recalculated at any time, in the same way that the pede script author did. His problem however, was that he interpreted the difference in the "Real, multi-decimal percentages vs the rounded percentage" as fraud, while it's really nothing more than a way to save filesize for an insignificant number. The only number that matters in the vote data files is the total vote counts for each candidate (ie 3000 Trump, 2000 Biden, 5000 Total). The percentages assigned to it for display purposes are "metadata" that don't matter at all. With the vote numbers, we can always calculate the actual percentage with all trailing decimal numbers if we truly want to.
So, no, I haven't gone in depth into that analysis. I decided not to do anything when I saw that the author was misinterpreting the intentionally missing decimal numbers from a rounded percentage as "election fraud", while his code was an absolute mess that was almost unreadable.
Here's my comment from back then:
https://thedonald.win/p/11Q8O2wesk/x/c/1ATp1CLTV3
His code is quoted in my comment. As you can see, it's an absolute mess.
What I will say, however, is that his code will definitely catch any sudden massive "glitch" swings such as "-7000 trump, +3000 biden". But such detection code would have been written much better by just writing code that looks at each successive vote total number (ignoring the "rounded percentage" red herring) and just ensures that it's always greater than or equal to the previous number (no regressions/subtractions). However, even such fixed code would not detect the votes that were NEVER published as Trump votes. In other words, if the machine originally received 5000 ballots and 4000 were for Trump, but they published online "3000 biden, 2000 trump", then no amount of analysis of the Edison data will ever find that fraud. The only way to find it is to look at and hand-count the actual paper ballots that the machines lied about the tallies for.
And by the way, we must pray that paper ballots are always kept (not destroyed), since there are plenty of reports of people multi-scanning boxes of ballots over and over (since ballots don't have any uniqueness identifiers) into the machine, to count Biden votes multiple times. The only way to catch THAT will be to have the original papers, or to have the image files (scans) of the paper ballots and look for identical multi-scanned ballot images (same style and position of checkmarks etc).
We have a lot of work ahead of us but I am sure that recounts will continue to find massive fraud. :-)