It’s still not over, no matter what some “news” source is telling you about the media-announced president-elect and his plans to lock down America for another year or so.
In fact, several states are still counting, and recounting.
And there may be more of that to come.
This post is long. I would break it up, but too much is
happening too fast. So I piled it all in one post. After the lawsuit section comes the data scientist heroics.
More data scientists trying to save the country
screenshot from Dr. Shiva's Nov. 12 video
Those
Dismissed Lawsuits
If you’ve heard that the President’s lawsuits are being
dismissed—or maybe you’re hearing it as “laughed out of court”—that would be
yet another example of media disinformation. The only lawsuits of the President
that have been dismissed were because the damage was already done, so relief in
the manner requested in the lawsuit was no longer possible. That means there
will be other lawsuits to address the damages, to replace the lawsuits intended
to prevent further damage.
As for other, private citizen lawsuits, they’re not as
hopeless as you’ve been told. There’s a lawsuit I covered in my last post, in Michigan,[i] asking
for injunctive relief—that is, a specific request to stop something underway.
To grant an injunction, a request must meet certain requirements, one of which
is that there are no viable alternatives. In this case, the essence of the judge’s
decision was to dismiss because there were alternatives available, which he
actually listed for the plaintiffs, meaning the expectation is that the
plaintiffs will indeed file another suit looking for those other remedies.
The judge in that case could have simply dismissed and said
nothing more. Instead, he gave away his bias. He said he listened to the
defendants’ explanations and liked them better. He said the issues brought up
by the plaintiffs were simply misunderstandings about the process, which the
plaintiffs could have avoided if they had attended the pre-counting walkthrough
provided by the defendants. Except—they had not been invited to any such
walkthrough; they had not been informed that any such walkthrough or
instruction would take place. They were excluded from learning any such procedural
details—but we’re not certain the excuses by the defendants were accurate anyway.
And the judge shrugged off the detail about a truck with
out-of-state license bringing in box loads of ballots in the middle of the
night. Because it was a rental truck; out-of-state licenses are common on
rental trucks. The judge refused to see any reason for concern about a rental
truck being used for the government purpose of ballot delivery—in the middle of
the night, after counting had been paused, and the announcement had been made
that all ballots had been received. Nothing to see here.
What the judge has done, inadvertently, is identify where
the plaintiffs will need to shore up their evidence for the next lawsuit, which
I believe was taken on by True the Vote.
Those various dismissed lawsuits are mainly concerning mail-in
ballot counting issues, which are numerous.
More Data
Scientist Evidence
Let’s talk about the data science issues, adding to the ones
I talked about a week ago.
More on Benford’s Law
In case you’ve heard that the use of Benford’s Law[ii]
for voter fraud was debunked, that’s not true either. There’s a mathematician
in the UK, Matt Parker, who spent some time showing why Benford’s
Law may work fine for financial fraud issues, but not for election fraud. The
reason, he explains, is that Benford’s Law only works with a larger numbers. If
you deal with numbers under 1,000, which many precinct counts would be, you
don’t get the numbers you expect.
What he doesn’t say is, all you have to do is use either
large enough precincts, or do the data by pairing or combining precincts. You
could combine even up to 5 precincts and still have quite an array of data in a
large county. You could do it by area or randomly. You could do it by state
house district, or zip code, or some other combination. By the debunking guy’s
own explanation, Benford’s Law would then be a viable check for fraud. I do not
know if that has been done, but the data is available to do it.
Dr. Shiva Ayyadurai and Team
I became aware of another data science method of identifying
fraud just after my last post, which you may have heard of by now. I’ll lay
that out briefly, and then cover the debunking (by the same British guy).
Dr. Shiva Ayyadurai has a PhD in computer science from MIT,
among other credentials. He is also currently (since results have not yet been
certified) a candidate for senate in Massachusetts. He makes the presentation.
There were two other data analysts on the presentation team, Phil Evans and
Bennie Smith. Phil Evans has been looking at these patterns for a couple of
decades.
Their study looks at four counties in Michigan: Oakland,
Macomb, Kent, and Wayne. They are making the claim that, in three of these
counties, a minimum of 69,000 votes were taken from President Trump and
switched to Biden, for a total difference of 138,000 votes.
The concern is over any counting that uses a particular type
of software system—Diebold, which I believe is a product of Dominion. There are
86 counties in Michigan. You can extrapolate from there. And, of course, there
are many counties in many other states that used this same counting system. The
goal, I think, is to show that there is a need to look into the count wherever
there is the possibility that this vote switch was done.
Dr. Shiva says his goal is to educate viewers enough so they
can explain the concepts to their friends. So I’ll give that a go.
There are two types of voting available in Michigan:
straight ticket or regular voting (choosing a candidate for each individual race). Up until this election, we also had straight-ticket
voting in Texas, which is convenient but tends to invite uninformed voters and
makes voter harvesting easier. Anyway, they have that in Michigan, so we can compare straight-ticket to regular.
Data sets for these two types of voting include early voting and election day
voting. In some they graph these separately, and in others they combine them.
What is straight-ticket voting? It is by party; you press
one button, and all the races on the ballot go to the party of choice. On ours,
we had the option of going over the ballot afterward and changing individual
choices. I don’t know whether Michigan had that, but I think that would no
longer count as a straight-ticket vote, even though the convenience was there
for the voter. Other voters may go all the way down the ballot, doing each race
individually, voting all for one party. (That’s what I usually did.) That
wouldn’t technically be straight-ticket voting, as far as convenience for the
voter, but I think in the analysis it might count as such.
You can put each party’s straight-ticket voting on a line—the
x-axis of a graph—indicating how strongly Republican or Democrat a precinct is.
It may not be an exact identifier of party strength in a precinct, but it’s a
pretty good indicator. For example, if a precinct’s straight ticket voting is
60% Republican and 40% Democrat, that’s a fairly strong Republican precinct. If
a precinct’s straight ticket voting is 7% Republican and 93% Democrat, that’s
an extraordinarily strong Democrat precinct. So the x-axis runs from 0%
Republican to 100% Republican, and each precinct has a location based on that
percentage.
Then there’s the y-axis. The specific thing Dr. Shiva’s team
is comparing is, of those who varied from a straight-ticket vote, how many made
a change only in the presidential race? In other words, how many preferred a
different-from-party presidential candidate over the other candidates of that
party on the ticket?
What you would expect to find is something relatively close
to the x-axis, with some precincts giving a few more or a few less to a
particular candidate. Precincts where more regular voters preferred the
Republican candidate in a higher percentage than the straight-ticket voters
would be above the line, meaning Democrats or others chose the Republican candidate
over their own. Precincts where fewer voters preferred the Republican candidate
than the straight-ticket voters would be below the line.
A normal distribution would look something like this, with some above and some below.
 |
| simple example of a normal case screenshot from Dr. Shiva's Nov. 10 video |
Each square dot is a precinct. It’s direction on the x-axis
is the percentage of straight-ticket voters of, say, the Republican Party. It’s
placement on the y-axis is the difference, either positive or negative, of regular
votes who preferred a candidate other than the chosen party. Again the 0% is
the x-axis, the straight ticket voting percentage.
In this actual example, a few Democrats in heavy Democrat areas voted for the Republican. In the center you get more independents and swing voters, and you see a natural scatter. As you get into the strong Republican areas, there are fewer Democrat/other voters who can defect toward Republicans, so the percentage goes down. The red line is the average of precincts at a given location on the x-axis. The pattern is something of a parabolic curve.
 |
| A normal real example, showing parabolic curve screenshot from Dr. Shiva's Nov. 16 video |
In Dr. Shiva’s presentation, we’re seeing how many Republican voters there were who preferred all Republican candidates—except for President. And how many regular Democrat/other voters there were who preferred only the Republican candidate for President. If the near-straight-tickets veer away from Republican, the precinct is below the line. If the near-straight tickets veer toward Republican, it’s above the line. You’re seeing whether President Trump was more or less popular than other Republicans on the ballot.
What you would expect is, in less strong Republican areas, regular
Republican voters might be more likely to veer away from their candidate, and
in strong Republican areas, regular Republican voters might be more likely to
stick with their candidate, even if they veer on another race or two, and you’d
even see some of the other party’s votes joining in.
But you don’t see that.
This is a scatter graph of Kent County.
 |
| graph showing Kent County, MI screenshot from Dr. Shiva's Nov. 10 video |
You see the same pattern for early voting and election day voting separately for Macomb County.
 |
| graph showing Macomb County, MI early voting and election day voting screenshot from Dr. Shiva's Nov. 10 video |
Take a look at the first 20%, the least strong Republican areas. In the presentation, he does this with Oakland County. The average of precincts is a relatively straight line, showing that Trump is 7% more popular than other Republican candidates.
 |
| graph highlighting least Republican precincts in Oakland County, MI screenshot from Dr. Shiva's Nov. 10 video |
Then you see the graph drop. The stronger the Republican area, the more likely it is that regular voters choose Biden over Trump. And it’s a straight downward slope.
 |
| graph of Oakland County, MI yellow line shows average of precincts screenshot from Dr. Shiva's Nov. 10 video |
There is a direct proportionate link between strength of Republican areas and choice of Biden. Republican voters are choosing all or most of the other Republican candidates, but not choosing Trump for President. Greater Republican strength = less Trump popularity. That’s just the opposite of what you’d expect to happen.
Could this be happening naturally? If there’s a movement of strong
Republican voters who disapprove of the President, then it could happen. But
there’s no evidence of any such movement. And, if there were, it would happen
organically, mixed among Republicans all across the list of Republican voters,
not getting more obvious moving toward the more strongly Republican areas.
The linear nature of the data is evidence that it is not
natural. Remember, the natural graph above with the parabolic curve? That isn’t
happening here. What they speculate is that, up until about 20% Republican
strength, the vote is relatively natural, and then an algorithm kicks in,
giving more Trump votes to Biden the further you move into Republican strength.
Wayne County is an exception. It looks like this, which they think is natural—or, at least it doesn’t imply manipulation by an algorithm. What you notice is a messy splatter graph.
 |
| graph of Wayne County, MI screenshot from Dr. Shiva's Nov. 10 video |
Oddly, Wayne County is where so many of the lawsuits are
happening, based on observed counting irregularities. Dr. Shiva makes it clear
that what his graphic data does not show is whether some other form of fraud happened;
it only shows that an algorithm was used to alter data from one party to another.
Cleverly, the altered data happens in areas that still win a
Republican majority, just at much lower percentages than you would expect for
strong Republican areas. So the theft is less likely to be noticed by casual observation.
An interesting detail, especially obvious on the Wayne
County graph, is that, in some of the strongest Democrat areas, you see the
most movement toward Trump. The area shows much higher concentration of
Democrat strength, according to straight-ticket voting, but surprisingly heavy veering
toward Trump in those areas. Dr. Shiva and team speculate that, without the
algorithmic interference, you would have actually seen a landslide victory for
President Trump.
More Debunking
Now, back to the guy who tried to debunk Dr. Shiva’s
presentation. Parker is not looking at the same things. He quickly
dismisses the first 20% being parallel to x-axis and just draws a longer straight
line down; he doesn’t say he has actually measured the average of the precincts
in that section of the graph. Then he uses non-straight-ticket voting in total
for the data. If it’s not straight ticket, but is nearly entirely Democrat, he’s
using that. So of course deviation from straight-ticket Republican is greater,
if you’re counting all Democrat non-straight-ticket voting, all of which
includes a vote for Biden. The more non-straight-ticket-Republican votes, the
more deviation from Trump votes.
But, if you’re looking at the ballots that vote nearly fully
Republican except for the President—which is what Shiva’s team looks at—then there
should clearly not be a proportional relationship. And a straight line downward
is truly suspicious.
Could I be wrong? Certainly. I haven’t worked the data
myself; I have gone by what different people have said they used. But the data
is publicly available for anyone who wants to work it themselves. Plus, earlier
today, Dr. Shiva came out with another video addressing some of the debunking.
More Fraud
Details
Let’s add a few details.
Weighted Ballots Are a Feature
The software used for counting—in these cases, they are looking at Diebold—there is a feature that allows for weighted voting. It’s a feature, not a bug. (See their user manual, below.) It’s possible, for example, to count a vote as something greater than “1 vote = 1 vote,” for example, something like “1 vote = 1.5 votes” for a particular area or category of voter, while another area or category is counted as “1 vote = .5 votes,” or any weight that gets programmed in.
 |
| in the Diebold user manual, red circle added by me screenshot from Dr. Shiva's Nov. 16 video |
There’s a question about all those places that stopped counting at around the same time in the middle of the night, when they were showing Trump winning, and then when they come back online to count hours later, the counting goes to Biden in such numbers that it seems—unlikely. Or maybe even mathematically impossible.
Then there’s that one county, Antrim County, Michigan, where
there was a Republican familiar with her area and could see the outcome was way
out of line with what was expected. So she did a hand recount. There were 6,000
votes that had been switched from Trump to Biden.
It was called a glitch. Dr. Shiva and his team say there is
no such thing as a “glitch.” The software does what it is programmed to do. A “glitch”
or “bug” is when the software, in testing and trial use, ends up doing
something unintended. It’s not something that shows up randomly in some places,
but not other places under the same conditions. The machine doesn’t make a
counting “error”; it counts wrongly when programmed to do so. That glitch in
Antrim County is, in itself, enough to question many other places where the
same counting system was used.
Then there’s this. The counting machine, at least for
Dominion, takes a picture of a paper ballot—and then counts the image of the
ballot. The ballots are set aside. It is federal law that all voting data and
materials must be held for at least 22 months beyond an election. But many
states are deleting the images—the things actually being counted. They claim
they don’t have to keep them, because the ballots themselves are saved. But by
deleting the actual images, they cannot show whether any alterations were made
to the images.
Every such location should have a hand recount of the actual
ballots.
Refusal to Reject Mail-in Ballots
Then there’s the matter of ballot rejection for absentee
ballots.
Back in August I referenced a NYT article from October 20, 2012, “Error and Fraud at Issue as Absentee Voting Rises/National Election Defense Coalition.” The difficulty with any type of absentee ballot is chain of custody. How do you prove that the person who filled out the ballot was the person entitled to that vote? If signatures don’t match, they can be rejected. If they don’t arrive with the appropriate signed outer envelope, they can be rejected. If they arrive late, they can be rejected. There are any number of reasons a ballot should not be counted, because there are so many ways these ballots are easy marks for fraud.
So you would expect that, anyplace that uses mail-in ballots
would need to put in place very strict rules to make sure the person entitled
to vote is actually voting, without pressure or influence—all things you can
watch for and guarantee at an in-person voting location that follows the rules.
Instead, this year you’re seeing many counties in many
states refusing to do even the minimal checks. Instead of a 2% rejection rate,
you have rates much lower, for example, 25 times lower in Pennsylvania.
Georgia Recount Failure
About that recount in Georgia: the Secretary of State came
out publicly, promising to do absolutely everything by the book, to get an absolutely
accurate recount, re-canvass, and full audit. And then they started recounting
main-in ballots without checking signatures. That absolutely can’t go on.
Release
the Kraken
Meanwhile, other stuff is going on. Rudy Giuliani tweeted
yesterday,
Stay tuned for big news tomorrow. @SidneyPowell1 and I have
substantial evidence of fraud and I can confirm that we have Dominion in our
hands and are analyzing the logs. It will expose fraud to such extent it will
be irrefutable that @realDonaldTrump won in a landslide. (2:13 PM, Nov 15,
2020)
 |
| Sidney Powell interview with Lou Dobbs screenshot from here |
There’s plenty of speculation about what that could be. Some
of it rampant. I’m trying to weed out truth from maybe over-the-top wishful
thinking. What I think is true is that Dominion and Scytl servers were seized
as evidence. In the case of Scytl, it would be done under an executive order
signed in 2018, allowing seizure of assets used in an attempt to influence the
outcome of a US election. Scytl servers were located in Germany and routed
through Spain. Rep. Louis Gohmert validated the claim that the servers were
seized, with Germany’s cooperation.
Things I think they would be looking for:
·
Source codes indicating an algorithm that caused
a change in how votes were counted—such as the weighting of ballots. It’s
there; are there indicators of when/where the weighting was turned on, and the
ratios? Because that could give estimates of actual votes changed.
·
Any indicators instructing voting places to shut
down during the night, possibly allowing for the resetting of counting to be
weighted, or weighted further, from that point on.
·
Key players controlling beyond local and state
officials.
·
Enough evidence of fraud, or possible fraud, that
the election results in many states are clearly unreliable so other
constitutional procedures can take place.
There’s probably more. After so many years of seeing
wrongdoing that gets brushed aside, with lawbreakers not being held
accountable, I’d love to see all the corruption brought out publicly and
brought to an end. It might be messy, but a lot less so than having
anti-Constitution tyrants try take over and impose totalitarian rule over us.
Check my links and footnotes too, but I looked at these
sources. Use your discernment:
·
Dr. SHIVA LIVE video: “MIT PhD Analysis of Michigan Votes Reveals Unfortunate Truth of U.S. Voting Systems,” Nov. 10, 2020.
·
Dr. SHIVA LIVE video: “MIT PhD Continued Analysis of Michigan Votes Reveals More Election Fraud,” Nov. 16, 2020.
·
Bill Whittle, The Stratosphere Lounge video: “Episode265: The Smoking Gun”
·
The Epoch Times, American Thought Leaders
Interview, video: “Google Vote Reminders Only Went to Liberals, Not Conservatives for at Least 4 Days—Dr. Robert Epstein,” Nov. 12, 2020.
·
Viva Frei Vlawg video: “Michigan Voter Fraud
Lawsuit DISMISSED—Here's Why!” Nov. 14, 2020.
·
Viva & Barnes Live Stream, video: “Ep. 34:
Elections Lawsuits from Michigan to Georgia, Updates & MORE!” Nov. 15,
2020.
·
Viva Frei Vlawg video: “Dominion Voting Machines Issues—the New York Times!” Nov. 12, 2020.
·
Lou Dobbs interview of Sidney Powell on Fox, clipon Instagram by davidjharrisjr
·
Buck Sexton, The First video: “Trump's
Lawyer Gives Update on Legal Battle,” Nov. 12, 2020.
·
CDMedia video: “Interview with Source on Electronic Vote Fraud,” Nov. 5, 2020.
·
“Bellwether Counties Went Overwhelmingly for Trump in 2020,” by Petr Svab for The Epoch Times, Nov. 15, 2020
·
“2020 Rejection Rate of Pennsylvania Mail-in Ballots Over 25 Times Lower Than in 2016,” by Elizabeth Vaughn for the Dan
Bongino Show, Nov. 7, 2020.
·
“Error and Fraud at Issue as Absentee Voting Rises” by Adam Liptak for the New York Times, Oct. 6, 2012.
·
“True the vote sues Gov. Gretchen Whitmer to contest illegal ballots counted in Michigan,” True the Vote press release Nov. 12,
2020.
·
“The US Raided European Software Company Scytl, seizes servers with links to Dominion Voting System” StreetLoc, Nov. 13,
2020.
·
“What’s Kraken?” by Clarice Feldman for American
Thinker, Nov. 15, 2020.
·
“New federal lawsuit seeks to throw out 1.2million votes in Michigan, flipping the state for Trump” by Chris Enloe for The
Blaze, Nov. 14, 2020.
·
“Why 2020 US Election Votes Were Counted By A Bankrupted Spanish Company Scytl” Great Game India Journal, Nov. 13,
2020.
·
“Did Crown Agent Dominion Voting Systems Rig The US Elections 2020” Great Game India Journal, Nov. 9, 2020.
·
“The Extremist At Dominion Voting Systems” by
Darryl Cooper for The American Conservative, Nov. 16, 2020.
·
“How a Philly mob boss stole the election—and why he may flip on Joe Biden” The Buffalo Chronicle, Nov. 14, 2020.
[i] Viva
Frei discussed this dismissal in a vlog posted November 14. Also, Frei
and Robert Barnes discuss this on their Sunday night livestream, at around 1:17:00
in.
[ii] There’s
a good mathematical explanation from 2011 called “Benford's Law—How mathematics
can detect fraud!”
No comments:
Post a Comment