« What Is Bill Barr Going to Do? by Larry C Johnson | Main | Fox News–A 21st Century MEDIA Titanic? by Larry C Johnson »

12 November 2020


Feed You can follow this conversation by subscribing to the comment feed for this post.

Barbara Ann

OK, I've watched the video for the 3rd or 4th time and the 21:00 mark is critical. Just before this Dr Shiva's slide say X-axis is % Republican straight party votes (RSPV) and Y-axis is %Trump individual votes - %RSPV. The axes on the very next slide are different. They show the Y-axis as (Trump votes - RSPV) as a %. The difference is critical. In the former case one should expect a flat relationship regardless of how Republican a precinct is. In the latter case the negative slope relationship should be expected. The refutation I linked to assumes the latter analysis was done.

If Dr Shiva has in fact plotted what he initially describes I agree with his conclusions that no credible model of voter preference can explain the result. I hope he soon publishes this work in a more rigorous format so we can all see which it is, it could scarcely be more important.



I think in Wayne County they weren't using those machines, but they sure had a lot of surprise ballots show up. On another surprise note A judge in Pennsylvania has ruled in favor of the Trump campaign after concluding that ballots received after 8 p.m. on Election Day that were segregated should not be counted.

“[The] Court concludes that Respondent Kathy Boockvar, in her official capacity as Secretary of the Commonwealth, lacked statutory authority to issue the November 1, 2020, guidance to Respondents County Board of Elections insofar as that guidance purported to change the deadline in Section 1308(h) of the Pennsylvania Election Code […] for certain electors..."



I don’t know what system they’re using to store/access voting results but can confirm randomizing towards a specific outcome in a specific ballot question is possible in Access/SSMS/Visual Studio etc., after a threshold is met. This little bit of query is so basic I wouldn't even call it an algorithm. It’s so easy to do in fact that I have to wonder why, if fraud was intended, was it made so obvious for others to see? We’re not talking about a small % difference as some are alleging here. And I wonder what kind of testing state/federal governments write into their contracts to opt out if UAT fails — assuming there is any UAT before the production deadline.

It could be a rogue coder(s). I have written business requirements/technical requirements for our clients because sometimes they can’t express, in plain English, what they want in a way that Audit/Controls can read so it makes sense. That kind of trust can be abused.

I’m not convinced there is fraud but I have no problem with the recounts/legal recourse. I wish we had done away with voting machines back when SCOTUS decided it was time to stop the recount.

The Twisted Genius

Shiva rightly observed that down ballot Republicans did better that Trump. That’s been noticed and reported by a lot of people. However, Shiva’s conclusion was that this could only happen if votes were taken from Trump and given to Biden since it is inconceivable that so many voters would split their ballot in that manner. He also fails to understand that a polling precinct or county with a larger percentage of “Republican” ballots would have a higher percentage of those split ballots. That is the point made by Naim Kabir’s rebuttal of Shiva’s claim that Barbara Ann linked to in an earlier comment. Kabir actually provided the algorithm/code to prove his point. Here’s the link again.


Michigan has all paper ballots. Shiva made a point of the ballot images in the counting tabulation not being kept, yet he ignored the actual paper ballot records. His research can be tested against the physical reality of those paper ballots. No algorithms, just simple addition. Georgia’s state-wide hand recount will be very telling as to whether there were widespread shenanigans or not.

A Portuguese Man

I'm probably missing something.

Could someone please clarify this for me: when they talk about straight-Republican and candidate votes, are these exclusive?

If so, then doesn't it make sense that the more people there are that vote straight-Republican (thereby voting for Trump) the less people there would be that vote for Trump alone (because they already did so by voting straight-Republican)?


Just released joint statement from DHS and CISA: “There is no evidence that any voting system deleted or lost votes, changed votes, or was in any way compromised.”

DHS is headed by Chad Wolf, a Trump appointee. CISA is headed by Chris Krebs, also a Trump appointee.

Regarding the so-called software glitch in Antrim county Michigan: There was no glitch. It was human error by the County Clerk, Sheryl Guy, who is a long time registered Republican.


@Barbara Ann

The graphs use %percent Republican as X axis and % relative to straight R as Y axis because that is the basis of the algorithm used to manipulate the ballot. That means someone has spent the time going through the data looking for any tell tale sign/relationships that indicate an algorithm. They probably plotted R votes straight versus Trump. There were already reports of ballots changing from Trump to Biden so that makes sense. A plot like that would show a variation across counties just the same as the % republican (X axis does), but you wouldn't see any trend in it other than some counties are lower and some are high. It's only when you add in the % Republican that a trend becomes clear.

It also makes sense from a fraud perspective. If the Russians were to influence the vote for Trump you would expect Trump to have higher results across the board or for extra "turnout" in Trump leaning districts. By adjusting the R areas more than D they make it less likely to be spotted. Trump still won those counties so if you didn't do an analysis like this you wouldn't know it. People tend to focus on the districts where their candidate lost as the problem. It's quite clever. It allows you to adjust the count in the districts less likely to be audited. You still lose the district, but you influence the state total. It would have worked if there wasn't a blow out aka a "landslide" for Trump. Notice in the graphs that < 20$ R districts actually saw more votes for Trump than party line tickets. That means Independents and D were voting for Trump even when they weren't voting for Republicans in heavy D areas. Probably because they are tired of the lockdowns and fear of lawless rioting.
Your comment about plotting X versus X-Y is incorrect. It doesn't automatically mean there should be a downward slope and in fact just about every graph has a straight line at 0 to show that party line votes should correspond to individual votes.

The solution is to hand count the ballots then verify them against the previous results.


"Whatever happens. half the country will be convinced the election was stolen. Then what?"

Is there a historical precedent in this country? I don't know.

It seems all of our institutions are breaking down. And we don't have good leadership anymore, just thieves looking to enrich themselves. If it comes to a crisis, these people won't make good choices.

- Eliot


Yes, I agree 100%.
We need handcounts/audits/canvassing in 2 more states.
That would show that the scope of computer fraud is extensive enough to change the outcome.
But we need something to trigger that.
Getting that is my worry.

Martin Oline

I think you nailed it about the hand re-count in Georgia. It should show the difference but it will probably be ignored by the MSM. I watched the whole video and was very impressed with how well it was done. He made it easier for regular people to understand. There are so many people with Master's degrees that couldn't have earned a Bachelor's degree without a "Fun With Numbers" class to get the credit.
I realized that meny elections have been kept closer than they would be if they weren't manipulated. This algorithm probably evens out any big difference either way, but it is available to create the desired outcome when it is needed.
One thing just occured to me that this shows why the Republican party gained seats in the House, won't suffer any significant damage in the Senate, and seems to have lost the presidency by a larger margin. It didn't make sense and now I understand why: it was only programmed to stop Trump despite his army of supporters. Not damaging the rest of the ticket was a big mistake. Romney claimed the drop in support was because "true" Republicans don't like Trump, but as Dr. Shiva patiently explains, if that was what actually happened the result would be a flat horizontal line offset from the straight Republican ticket and not a thirty degree angle down that transfered more and more votes to Biden as the number of Republican votes proportionally increase.

Eric Newhill

Barbara Ann,
Dr. Shiva lays out the crux of his analysis -and his definitions - at the 27:30 mark.

Allow me to put it a slightly different way.

Dr. Shiva has defined "republicanism" as the % straight Republican party vote ballots in a district.

His dependent variable is the % of non-straight party ballots that had a vote for Biden.

Note - It is a simple mathematical reality, per definitions used by Dr Shiva, that as "republicanism" increases the % of individual candidate ballots decreases. A 70% republican district would have 30% individual candidate ballots. A 90% republican district would have only 10% individual candidate ballots. Shiva is taking the % of individual candidate ballots that went for Trump less the % of straight party republican ballots in the same district.

He then goes on to state that he finds in the data that as "republicanism" increases, the % of the individual candidate votes for Biden increases. He further states that such a tight linear inverse correlation is not natural.

So Biden is - in very and suspiciously linear fashion - getting an increasing percent of a shrinking percent of available ballots districts as republicanism increases.

That is the whole argument in a nutshell.

As I said previously, a possible natural explanation is that as republicanism increases the increasingly fewer (as a percent) non-straight party voters in the district voted not straight party, at least in part, because they don't like Trump. Republicanism could be a proxy for conservatism and as conservatism increases, tolerance for Trump's new money brashness decreases. There are many theories one could formulate over a cocktail or two. It remains that the intensely linear relationship is highly unusual. It merits investigation in Dr Shiva's opinion (and mine).

The critic Naim Kabir in your link (today he updated what he wrote) says - "One quantity we care about is the % of Trump votes that split-ticket voters give. If they have a fixed probability of voting Trump, this will be flat." The first sentence is exactly what Dr Shiva says. The second sentence is pretty much Dr. Shiva's null hypothesis; which Shiva thinks has been disproven. Again, all Shiva is saying is that as the % of straight R ballots in district increase, the % Trump votes on the individual ballots decreases and it does so in a suspiciously tight linear way. That needs t be explained.

Kabir goes on to say that "As Ayyadurai assumes: all Republicans who vote straight-ticket will vote for Trump [my edit: duh - no brainer]. Non-Republicans split-ticket voters, on the other hand, only have some small chance of voting for Trump." That second part is not an assumption that Dr. Shiva makes. It's from far left field. Who says that split ticket voters are not Republican or Democrat Trump supporters or Green Party or people like to smother their gay lovers in tapioca pudding while wearing swim fins and snorkels? No one other than Kabir is making any assumptions about who the individual ballot voters are. That's important because he then creates some lame code to simulate something that Shiva never said.

He then goes on a silly tangent about the "line of unity" that adds no value other than to demonstrate that he likes to go on silly tangents.

Then he boldly concludes that Shiva is conning us and he's stupid. IMO, Kabir only proves that any ideologically driven doofus can post anything on the internet as long as his politics are the same as the "fact checkers'". If you can find the hidden mathematical truth to Kabob's argument, I'd love to see it.

Eric Newhill

Kabob = Kabir. Autocorrect is trying to make me into a racist. I like Kabobs. I grew up eating them. I have a real name that is ethnic. I'm 50% middle eastern. I'm not a racist. I swear. Please god let me have a job tomorrow.


Barbara Ann,
I just turned the speed down to .75 and replayed the crucial section. I think that they assume an above-average aptitude/background for this kind of thing.

I think that these fellows are just independent thinkers. (Yes the main one is a bit self-promoting (another inventor of the email?—or *is* he The One.

IMO they’ve given very convincing evidence for not trusting a “voting system” which was an in-needed “improvement” which introduced severe liabilities into the election system just as the boomers reached mature years.

You could have a better election with notched sticks (or poles), and the US is *maybe* just figuring this out. Well, surprise surprise, there’s a lot of stuff like this going on.

It didn’t really take twenty years to realize that the opiate distribution companies knew what was going on perfectly well.

There are plenty of things that don’t add up in this country but that dogleg angle on that scatter graph takes the cake. Obviously they “taxed” the more republican districts—were forced to!—in order to flip enough votes to counter the red wave and the strong showing of trump in Dem districts!

Also it was good insight to compare straight ticket to individual voting. Bennie over there stole the show for me—“my calculator works the same as yours”—& refuting Romney: “Republicans May hate Trump as you say, but republican voters don’t hate Trump from precinct to precinct in a straight line on a scatter-graph.”

The same uniparty that winked at illegal immigration and sent manufacturing overseas winked at this. Deep swamp.

These guys get the uniparty concept.
The best thing elections-wise we can do for any good camdidate (like Tulsi Gabbard), is ensure fair elections—without “software enhancements”.

This is do-able. So I think we should just share and explain the video.

That’s my two cents.



There are some scatter plots out there that may or may not indicate something amiss. There is a discussion in Bret Weinstein’s Twitter feed that goes into details but I have not been able to fully review either side. Bret Weinstein is the professor from Evergreen State in Washington who was run off campus by a woke mob. He is scrupulously fair to those he disagrees with. The scatter plot debate is worth reviewing.

Benford’s Law is not really valid here. At the district or precinct level there are not the orders of magnitude difference in populations to make it valid. Different counties do have those population differences but only Texas has enough counties to make anything resembling a correct graph simply due to sample size. I ran analyses on the county level for various states, red and blue, contested and not, and other than Texas all violated Benford’s Law.

I would need to see Dr. Shiva’s source data to comment, as well as to see further control counties as comparison but I do believe he is on to something. Really, someone with the data sets and good R or Python skills could quickly confirm or debunk this. I suspect it would be confirmed. His stated method makes a lot of sense.

I would be careful of just taking credentials. There is a forensic accountant misapplying Benford’s Law online.

There is a young man named Matt Braynard who is doing a record by record comparison and following up with affidavits. This would detect voter level fraud. He is worth looking up. The downside to his work is that his report outs are frustratingly vague but he is making progress. He reports out what he has found but not at a level that permits an understanding of magnitude. Part of this appears to be an overworked small team working fast.

Thank you for all of your insightful analysis. I always enjoy reading your writings.


Jopseph Baio

I don't post comments but in this case I can't resist because I don't understand why no one has asked why voting machine software would have a weighted election feature. What is a weighted election? Why the feature?



That's not quite correct. Shiva called for MI to count the paper ballots so as to determine the actual results. They need to be validated as well as there is plenty of evidence of thousands of fraudulent ballots being injected into the system, especially in Wayne County.


Joseph Baio

Neither I or my wife (a former election judge) have heard of such a think before watching Ayyadulai's lecture.


Ghoti "Fish?"


Posted by: Patrick Armstrong | 12 November 2020 at 04:10 PM

personally I am a fan of anecdotes and yes, why not Benford.

thus I did like this piece of evidence, brigades challenging brigades. What would be my chances as (innocent?) work bee no 20 on desk 228 that allegedly disappeared Republican votes? Hopefully there are surveillance cameras and the Challenging Rightful Supervisor noted the exact time at which I supposedly disappeared ballots from a predominant Republican precinct.

Anyway here we go::
Here is a new Simone Gao interview of a Republican poll challenger/watcher on duty at the Detroit (Cobo Hall) vote counting center during the Nov3-4 ballot counting period. Very informative; 22 min long.


Barbara Ann


"Really, someone with the data sets and good R or Python skills could quickly confirm or debunk this."

Yes, I would like to do just that and am trying to find the source for the data. I can only find county-level results with no distinction between single candidate votes and SPV so far. If anyone has identified the source please post a link so we can get to the bottom of this. AFAIK Dr Shiva has yet to publish his work in a form where it may be subjected to peer review. Given the gravity of his allegations, the longer he delays in doing this the more suspicious we should be of his motivations IMO.

Eric Newhill

Barbara Ann,
I said, "Note - It is a simple mathematical reality, per definitions used by Dr Shiva, that as "republicanism" increases the % of individual candidate ballots decreases. A 70% republican district would have 30% individual candidate ballots. A 90% republican district would have only 10% individual candidate ballots. Shiva is taking the % of individual candidate ballots that went for Trump less the % of straight party republican ballots in the same district."

Then I got distracted. I should have concluded that paragraph by noting that if the objective is to systematically code the counting software to throw the election to Biden, you'd have to switch a greater % of the shrinking volume available individual candidate ballots in the district to Biden as republicanism increases. That is exactly what Dr. Shiva's results show.

It's really a simple concept.


Barbara Ann,

"Given the gravity of his allegations, the longer he delays in doing this the more suspicious we should be of his motivations IMO."

You mean like settle science climate change analysis? Just kidding. Nothing is preventing anyone else from getting the same data from the MI SOS or those county/city clerks who possess the data. Nothing other than their resistance to providing it, which is itself a red flag that should trigger further scrutiny.


Having lived in Michigan for more than a dozen years I'm familiar with the straight line voting. It's done to ensure winning all those down ballot races because downballot votes in democratic areas drops off a cliff. That's why the democrats have fought so hard to keep it in place.

Further, and rather importantly, there is the 'spoiling' of the ballot that occurs if someone votes straight party by makring that on the ballot, then votes for any partisan race down ballot individually. I would be interested in seeing that data. Also, in parallel to GA where there are reportedly tens of thousands of Biden only ballots it would be interesting to see how many Biden only/Trump only ballots there were in Michigan; then how many straight party votes with no other ballot issues being voted on. A comparison to prior elections would be telling.

Barbara Ann

Fred, if I had discovered algorithmic vote counting fraud I would move Heaven & Earth to prove it beyond reasonable doubt and that means peer review, not a single live-streamed YouTube video. Yes the peer review system has its flaws, especially when it comes to academic grant factories like Climate Change, but this is very simple to prove or disprove. And on that note:

I came across someone who has done just that and invested the effort to replicate Shiva's work for one county - Kent*. The author achieves the same down slope Shiva does for Trump vs. Rep. SPV. He then does a plot for Biden using the same methodology and waddayaknow another down slope. The author says; "By Shiva's logic, the vote tallying software is stealing votes from Trump in high R areas, and stealing votes from Biden in high D areas. Of course, this is only an artifact of the choice of y-axis."

*It appears Kent county uses the services of a Grand Rapids-based election services & software co. https://electionsource.com who publish results down to precinct level for this and other counties using their services (in pdf form) at https://electionreporting.com.



I'm sure this won't end the discussion, but at least now the data for one of the counties Shiva uses to demonstrate the alleged vote-stealing algorithm is out there, so I'd encourage the suitably skilled and concerned to prove it or otherwise for themselves, I intend to do so over the weekend.


Be interested to hear your views on the above and btw I had the most intense déjà vu reading your latest comment, most bizarre.


Barbara Ann,

physically count the lawfully cast ballots. That means people who did not vote in person don't get extra time to vote, otherwise they are being privileged over other citizens. Physically recounting fake votes is no different than George Floyd counting his fake $20s. They are still fake.

Eric Newhill

Hi Barbara Ann,
I just modeled this out myself, again, with graphs, etc, but using dummy numbers (that doesn't matter because we're looking at the concept, not the actual results). My model was based on 20 "districts".

I can create scatter plots that look exactly like Dr. Shiva's. To do so requires the following:
a) the % of individual ballot voters going for Biden (or, conversely, going for Trump) remains constant as a % straight republican vote increases. I arbitrarily used 35%, but you could use any %. The line will do the same thing.

b) the % of individual ballots going to Biden increases as the % straight republican vote increases

The only difference between a and b is the steepness of the slope of the line.

Those (a and b) are the required conditions to get a line that looks like Dr. Shiva's.

I also modeled where the % individual votes going to Biden, as a % straight republican ballots increases, is neither constant, nor constantly increasing. Rather, that % is random across districts. In that case, of course, you don't get a straight descending slope line. You get a zig zagging line, as expected.

To say that the downward sloping line is merely a function of the Y-axis is like saying that getting an answer of 4 when adding 2+2 is merely a function of math and the choice of 2 and 2. It's a meaningless statement of the obvious that doesn't discredit Dr. Shiva in the least bit. If you add 2+3 you don't get 4. The question is, seeing a result of 4 and seeing inputs of 2 and 2, does that make sense given what we know about the natural world.

Dr. Shiva is saying that a constant % (actually I think he says increasing %) of individual ballots going to Biden as straight republican votes increase is highly suspicious.

There is nothing wrong with Dr. Shiva's model/approach.

You can critique his conclusion. As I have suggested, there may be natural explanations for his downward sloping line and the tight correlation between % straight R ballots and % individual ballots going to Biden. I agree with Dr. Shiva in that I think it is very odd that the correlation is so tight, but the critics hand waving away his analytical approach are saying nothing and trying to sound smart or deliberately attempting to obfuscate.

The comments to this entry are closed.

My Photo

February 2021

Sun Mon Tue Wed Thu Fri Sat
  1 2 3 4 5 6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 26 27
Blog powered by Typepad