June 18, 2024 - The Truth Central - Dr. Jerome Corsi
01:31:03
Why is there an Algorithm in the New York State Board of Elections Voter Roll and WHO Put it there? with Andrew Paquette
There has been a series of significant issues with a voter roll algorithm in the state of New York, interfering with a fair election process. Yes, you read that right: an dangerous algorithm was found in the New York State Board of Elections voter roll. Dr. Jerome Corsi talks with Andrew Paquette, formerly of the New York Citizens' Audit, who discovered the algorithm: "Voters in New York State are identified by two identification numbers. This study has discovered strong evidence that both numbers have been algorithmically manipulated to produce steganographically concealed record attribute information. One of the several algorithms discovered has been solved. It first utilizes a mechanism nearly identical to the simple ‘Caesar Cipher’ to change the order of a group of ID numbers. Then, it interlaces them the way a deck of cards is arranged to create a ‘stacked deck’. The algorithmic modifications create hidden structure within voter ID numbers. The structure can be used to covertly tag fraudulent records for later use." (Paquette, Research Abstract, https://www.researchgate.net/publication/370835885_The_Caesar_cipher_and_stacking_the_deck_in_New_York_State_voter_rolls).The question is: WHY is there an algorithm in the New York State Board of Elections voter roll, who put it there and to what end?Dr. Corsi and Paquette go in-depth to answer the question on The Truth CentralIf you like what we are doing, please support our Sponsors:Get RX Meds Now: https://www.getrxmedsnow.comMyVitalC https://www.thetruthcentral.com/myvitalc-ess60-in-organic-olive-oil/Swiss America: https://www.swissamerica.com/offer/CorsiRMP.phpGet Dr. Corsi's new book, The Assassination of President John F. Kennedy: The Final Analysis: Forensic Analysis of the JFK Autopsy X-Rays Proves Two Headshots from the Right Front and One from the Rear, here: https://www.amazon.com/Assassination-President-John-Kennedy-Headshots/dp/B0CXLN1PX1/ref=sr_1_1?crid=20W8UDU55IGJJ&dib=eyJ2IjoiMSJ9.ymVX8y9V--_ztRoswluApKEN-WlqxoqrowcQP34CE3HdXRudvQJnTLmYKMMfv0gMYwaTTk_Ne3ssid8YroEAFg.e8i1TLonh9QRzDTIJSmDqJHrmMTVKBhCL7iTARroSzQ&dib_tag=se&keywords=jerome+r.+corsi+%2B+jfk&qid=1710126183&sprefix=%2Caps%2C275&sr=8-1Join Dr. Jerome Corsi on Substack: https://jeromecorsiphd.substack.com/Visit The Truth Central website: https://www.thetruthcentral.comGet your FREE copy of Dr. Corsi's new book with Swiss America CEO Dean Heskin, How the Coming Global Crash Will Create a Historic Gold Rush by calling: 800-519-6268Follow Dr. Jerome Corsi on X: @corsijerome1Our link to where to get the Marco Polo 650-Page Book on the Hunter Biden laptop & Biden family crimes free online:https://www.thetruthcentral.com/marco-polo-publishes-650-page-book-on-hunter-biden-laptop-biden-family-crimes-available-free-online/Become a supporter of this podcast: https://www.spreaker.com/podcast/the-truth-central-with-dr-jerome-corsi--5810661/support.
Do you have a single-person company or a small business?
Then you're probably tired of hearing me talk about how easy it is to deliver the tax return with FIKEN.
So we give up here.
Because we like it simple.
FIKEN.
Super Simple Accounting.
In October last year, we locked the Kiwi price on healthier goods.
The response was fantastic.
That's why we at Kiwi have decided to continue to lock the price on many healthier goods until November 1, 2024.
At 300 grams of folk food, shrimp and pork with mango and chili, we lock the price to 49.90.
And you, prices marked with pricelist can't go up.
Only down.
Because we in Kiwi never give up.
This is Dr. Jerome Corse and this is the truth central.com We're doing podcasts every weekday.
You can find me on AXAT at CorsiJerome1.
You can find me at Substack.
It's JeromeCorsiPhD.Substack.com.
We have a really special guest today, Andrew Paquette with us.
And Andrew, we're going to be discussing the possibility of voter fraud in New York State.
And this is a study that Andrew has been doing now for a considerable period of time.
Andrew is with the all-volunteer Citizens Group, the New York Citizens Audit.
And he's found substantial evidence of the possibility of fraud in multiple elections held in New York State.
Now, this analysis is different than many of the analyses we've seen because it involves the voter roll in which an algorithm has been placed into the New York State voter roll.
The algorithm is designed to change the various information in the database to manipulate, to produce what they call steganographically.
Steganography.
Steganography.
It's a hard word to pronounce.
Steganography is a term that means you are mapping information from one source onto another source.
Okay, so it's a mapping of information, and this algorithm manipulates the records to conceal the record attribute information.
It's going to be a little bit difficult to follow, but we're going to try to make this as simple as possible.
First of all, to get an algorithm in the voter database is itself suspicious, because the algorithmic modifications create these hidden structures within the voter ID numbers, and the structure can be used to covertly
tag fraudulent records for later use.
Now, Andrew's done a great deal of study on this, and the evidence certainly is that the
algorithm imposed on this New York state voter database, which is still there,
as far as we know, is pretty much evidence that someone's manipulated the database.
That's not supposed to happen.
Database is supposed to be kept pristine, not modified.
You come in and you register, you get an ID number, and that number should be sequentially given so the next person coming in gets the next number.
And your information is recorded into the voter file, and then you stay in the voter registration file unless you've moved out of state or have died.
And those records have to be purged periodically by the New York State Board of Election.
So again, that's a normal process.
And Andrew, first of all, thank you for joining us.
We're glad to have you with us today, and I'm looking forward to this discussion.
Thank you.
And by the way, I do want to make an important correction.
I left NYCA almost a year ago now, so I'm not affiliated with them at this point.
Right.
And also, it's very interesting, your background, tell us about your background, because you don't come from a background of cryptography or anything of that nature, but yet you have great skills at it.
So give us what your background is.
Well, the funny thing is I always wanted to be a scientist or a writer, okay?
But because my mom was moving around all the time, I kept on losing all the credit from the schools I had been to in the past as I went to the new schools.
So I found that art went with me everywhere, because they always based their grades on whatever I could do at that moment, as opposed to what I knew from the previous school's textbook.
So I got into art, and I found out that art actually was quite scientific in a lot of ways.
It all has to do with observations, optics, geometry, trigonometry, things like that, that I actually was interested in.
So I wound up becoming a professional artist.
I worked for a long time in commercial art as a comic book artist.
I worked as a video game artist, art director.
I worked on movies, TV shows, video games.
And then I helped found a school for game developers in Europe, Golly, almost 20 years ago now.
It was very successful and I earned a PhD from King's College London while I was there.
And when I got here, I was in New York, I was planning on doing commercial photography, which is a skill I'd picked up while I was in Europe.
But then we had the lockdown and this political craziness, and suddenly I found myself interested in elections in a way I'd never been interested before.
Primarily because on the evening of November, I guess it was the 4th, Of 2020, I was looking at the results, and I was remembering all my statistical training in my PhD, and I was thinking, this doesn't look right.
I kind of wonder what actually happened, because this does not look possible at all.
And what I was primarily thinking of was, I had calculated that Trump could lose four of the five remaining states that had to announce their numbers, and he would still win.
And so for him to not just lose all four of them, but the fifth as well, I was thinking this is not possible, especially Pennsylvania.
So that's how I got started on this.
Well, and it's interesting because your skills in, you've worked on some high budget movies, right?
Which ones?
Can you tell us?
Sure.
Spider-Man, Space Jam, Daredevil.
A tiny amount of uncredited work on X-Men 2, and actually it's uncredited because I asked them not to give me a credit because I was kind of embarrassed at the tiny amount of work I did.
I was actually on the credit list and I said, no, come on, give it to somebody else who actually spent more than a few days on it.
But it's very sophisticated work because you're doing computer animations that are quite complicated.
Yeah, I was actually kind of one of the more technically inclined CG artists at the various studios I worked at.
My specialty was getting things to be efficient, and also a really weird specialty that most people probably have no idea what it is, but It has to do with the way an image is placed on an object.
It's mapping, just like algorithms are mapping ID numbers in New York's photo rolls.
So to put this image of these stripes onto this shirt in the most efficient possible way, I have to figure out how to arrange what's called a local coordinate system of the three-dimensional object to match the two-dimensional coordinate system of one or more images.
And what I got very good at doing was dealing with the work of other people after they'd fouled it up really badly.
So you'd get something that would look like a cobweb in a cellar, and I'd have to turn that into something that looked like the Empire State Building.
And that was something that I didn't enjoy.
Well, actually, after a while, I did enjoy doing it because it was like a big puzzle, trying to figure out, OK, does that vertex belong to that window or that doorknob?
Because you'd see it all squashed and you'd have to figure it out.
I think these skills are what made you able to recognize patterns and be effective in doing what you were going to show the audience here in terms of the voter records.
Now, to begin with, I want to make it clear that what we're showing is evidence that we believe is pretty serious evidence of malfeasance.
Now, we're not going to say that this proves that the 2020 election was stolen.
We're not going to make a lot of... What we're going to say is this bears further investigation.
We want to take it seriously.
We think that the authorities that be should look into these issues.
And we're raising them here as questions that we're going to basically saying, why would this be done?
Why would anyone do what we're going to show you?
Because some rather remarkable and apparently nefarious things have been done with the voter record.
Certainly things that don't belong there, because you start out from the premise that the voter record ought to be pristine.
In other words, nobody ought to manipulate it.
And I would think, Andrew, that's the first premise from which you began.
Is that correct?
Um, actually, no, it's not.
I mean, it makes sense, but it's not where I started.
I started thinking, number one, I don't have any data to look at, so I have no way to determine whether what I saw on election night made sense.
But then I started thinking, well, I'm in New York, so maybe I can look at local data.
And I was thinking that it didn't make sense from a risk assessment point of view to do anything illegal or corrupt in New York, because New York was a solidly Democrat state, as I understood it.
And for that reason, the potential for a bad result in New York, for Democrats anyway, was very low.
So I didn't see any reason why they would want to take the risk of being caught doing that.
But I thought, what the heck, I'll see if I can find some data to look at and see, just see what I would see.
Now, what happened was, this is when I got associated with NYCA.
They managed to secure some voter rolls, which I looked at.
And right away, I started finding records that, based on my reading of the law, should not have ever been created in the first place.
Because what they were was, like, people who had a record, had an ID number, And then they had another ID number and another one and
another one there as many as one guy had 25 Separate ID numbers. It's definitely the same person
So I was thinking well if there's a lot of people like this and there were at that time
I didn't know how many there were in total At that time, I think the number I was looking at was around 700,000 records, but that's pretty significant.
And you can do a lot of bad things with 700,000 bad records, because what's going on in New York is you get a mail-in ballot sent to you whether you ask for it or not.
So, if we have, like, if my record, for instance, although I know this isn't true, because I checked, But if my record had another one with my name and address and birthdate information, but a different ID number attached to it, then I should be getting two ballots instead of one in the mailbox.
And if I send both of those in, which I could have done if this had happened with me, the people on the other end would have no way of knowing That one of those ballots was bad, especially if I sent them to different precincts where I put them in different drop boxes.
They wouldn't get them together, so they wouldn't look at them together, and they wouldn't be as suspicious.
So I knew that there were a lot of people who were probably getting these because I could see it in the voter rolls.
I was seeing all the justification necessary for someone to say, okay, here's a ballot I have to send out, and here's another one I have to send out.
Even notice that they were the same names because the ID numbers might be so different that they'd be in completely different regions of the database.
So I saw that and it occurred to me, okay, this is either an accident, The holidays are over, and everyday life is back.
And we at Kiwi want to make it as cheap as possible, with a lot of party goods.
At Norway 1.25 kilos, we set the price to 119.
At 1.26 kilos of juicy carbonara from Gilde, we set the price to 99.
And at Grandiosa without paprika, we set the price to 39.90.
Because we are the price pushers.
And we at Kiwi never give up on price.
Or it's on purpose.
And if it's on purpose, it has no value to anyone if they can't track these records.
And the reason is because if you're making fake records for the purpose of abusing them, you would have to know which records are fake.
And so I started with that idea and I was thinking, you know, there may be a tracking mechanism buried in the information in the voter rolls somewhere that would allow people to do that.
So what I was looking for initially was some kind of a tagging mechanism, right?
So you have all these fields like first name, last name, birthdate, etc.
And I was keeping an eye open for something that would let me identify which records were bad and which ones were good.
But to do that, I first had to find a sizable number of bad records, right?
So while I was focusing on the bad records, I was trying to keep my eye open for a tagging mechanism.
Now, the other people I was talking to, two people in particular, they didn't believe I'd find something like that.
And the reason is because they said it'd be too risky to open, you know, to tag the files.
They said that anyone with a reasonable ability looking at the records could possibly find that, and that would compromise that particular tool.
But I kept looking nevertheless, and eventually I found out that the records weren't tagged What happened was far more sophisticated and far more insidious.
Instead what happened was they took the county ID numbers and the state ID numbers and they married them in an operation called mapping.
So they're saying this belongs with this.
And what they did with this mapping scheme, which is extraordinarily complex, is they totally hid the fact that they had done it in the first place, right?
This is steganography.
Steganography is where you hide the fact that there's a hidden message, okay?
So they don't encrypt it.
It's not encryption.
It's a hidden message of some kind.
So for instance, In the Trojan War, I believe it was the Spartans did this, back then they had these clay tablets they'd write on.
So it's like a piece of wood and then they'd put clay on it and they'd write on the clay.
So what they did was they'd scrape the clay off, write their message on the wood, and then they'd put a fresh pile of clay on top and smooth it all out, and that would get through the guards.
So they wouldn't know that secret military messages were being passed back and forth because they couldn't see that there even was a message.
So what we were looking at, or what I was looking at in the roles, was this very complicated pattern based on something called a rep unit, which in math is any number that has, where all of its digits are the number one.
So 11, 111, 1111, et cetera.
Those are all rep units.
And it was very, very complicated.
It was obviously intentional, and it was obvious that it was in there for the purpose of obfuscation.
And the thing about that that puzzled me is that there are legitimate reasons to obfuscate data in databases.
OK, for instance, for privacy, for security, for efficiency.
That's actually efficiency.
You don't want to obfuscate data, but there's reasons to modify it so that it can be more efficient.
But for privacy and security, certainly social security numbers, for instance, are frequently masked as our credit card numbers.
So this is why a lot of times you don't have to give the whole number.
You just have to give the last four digits because the remainder is masked so that the people you're dealing with never really know what the full number is.
But the problem with the New York State voter rolls is that they're public document.
Anyone, you, any person.
I know researchers in other parts of the country who, after learning of my work, they requested New York's rolls and they got them and they were able to confirm what I found.
So the thing is, there's no privacy advantage here.
You don't really Actually, if they did make the data more private, or they did make it more secure by hiding, obfuscating, masking it in any way, as I understand it, it would violate the public disclosure components of the Help America Vote Act.
And actually, I think the NVRA also requires that this stuff be public.
So as a result, what's going on?
is they're doing something that is commonly associated with hiding data
for either security reasons or privacy reasons that is not hidden.
And the other thing about it that's really peculiar is it's not even really hiding the kind of data that you would
normally hide.
So normally what you might do is you'd hide a name.
So if you've got an ID number and you've got a name, you don't want people to know which name goes with the ID number, okay?
But the masking tool that they used doesn't change that relationship.
The name is still associated with the same ID number, it's still associated with the same DOB, same address, all the rest of it.
It's not changing any of that information.
The only thing it's really doing It's creating a very unique relationship between two numbers that have no privacy or security implications whatsoever.
So, as part of that process, it has to create another ID number, which is kind of strange, because what it does is it re-ranks all the numbers.
It puts them into a completely different order that, you know, unless you're really studying this like I did.
I spent over two years studying this thing.
Um, you would never know what that order is, okay?
But this rank order is the equivalent of a third ID.
So I can refer to, like, anybody's file in New York using that third ID.
And we're gonna, let's demonstrate some of this too, because I want to get too far ahead.
I want to make a couple points that you made.
The stenography is when, for instance, you get a micro dot that has information contained in it.
Okay, that's another form of concealing the information in something that looks like a pixel.
Yeah.
Okay, so you, it's an obfuscation system to hide information that you don't want seen, but it's still there.
Now, This mapping system, we're going to get into it right away, but the key is what Andrew's just explained, is that the key to what New York did is all your information can remain the same.
In other words, your name, your date of birth, all this can remain the same,
but yet there's a, as it were, a cloned record of your record
with maybe some modifications in it, but not necessarily the information of who you are,
where you live, et cetera.
It can be altered, but the point is it's got a different ID number now.
With a different ID number, it looks like two separate records,
especially if they're not consecutive in the database order.
Okay, now this is gonna get a little bit difficult to follow.
We're going to try to make it simple.
And I want to show you the first one of the important slides in Andrew's paper.
I'm going to share screen.
And Andrew has done a paper here now, which has been published.
It's been published in a legitimate journal.
And as I'm sharing it here, we're going to see the paper.
The paper is called the Caesar Cipher.
And the Stacking the Deck of the New York State Voter Rolls.
So, Andrew, you can see this on the screen, right?
Yes.
Okay, this was published in the Journal of Information Warfare in 2023, so it is a published article.
Okay, now I want to go to the scheme here in terms of the numbering, and this table one here is very important.
And I'm going to have Andrew explain it, but we're going to be talking about mapping from a CID number, which is a county identification number, to a state board of elections ID number, which is created at the state board level.
So this is the modification and we're going to be talking about these ones if you follow my cursor here about these ones that are added in and those are the re-units and I may be able to see if I can get this to be a little bit larger let me just see if I can get it well it's going to be zoomed to here and I can make it maybe 200 percent.
Get it a little bit larger.
Okay, there we go.
Now, this is the table we're going to talk about right here.
Maybe a little bit too much.
Just do again.
I'll do zoom to, um, zoom to, I'll do 150.
Make it a little bit easier to see.
Okay, that should be good right there.
So Andrew, you want to walk us through this?
I'll try to follow you with my cursor.
Okay, so the first field is record.
Now, that is a number that I assigned based on the order the record appeared in the original database as it was handed to me.
So, on the disk, they have a bunch of records that are not in a database format.
It's in what's called a TXT format.
So, it's just a whole bunch of information that's strung together.
I'm not sure if it was comma delimited or tab delimited, but either way, I needed a way to separate these records so that I could have a unique number.
And the reason is because I knew that the algorithm is highly dependent on the exact position of each number.
So if I got the position of any of the numbers wrong, then it would throw everything, the solution off.
So what I did was I made this record ID so that if in my various sorting operations
I ever lost the original sort, I could go back to it by using that record ID number.
The record gap tells you what the difference is between the current record and the previous record.
Now, the first one, you see a record gap of zero because there's nothing above it.
But below it, you see that the two numbers there, they end in 739 and 850.
So the difference between those numbers is 111.
And then the next number that you see is also separated by 111.
And then what we do is, and by the way, you might be thinking, well, so what?
You just picked out those two records that happened to be 111 farther down.
Well, no, I didn't.
It's actually because of the way those records were selected that they came out that way.
So if you look at the CID, that's the county ID.
So now those are all sequential.
So that's 18,465, 466, 467, okay?
And if you look at the CID numbers, that's 100,136, 7 and 8.
So those are all in order.
So if I've got my CID numbers in order, my record ID numbers come out 111 apart.
And if I go to the SBO ID number, That's the State Board of Elections ID number.
Those mirror the gaps that you find in the records.
Now, by the way, the reason I don't substitute the SBO ID numbers for the record numbers is because they don't always align like this.
there are deleted SBO ID numbers, and that throws off the alignment
with the record and the SBO ID numbers.
So the important thing, though, is that when you put the CID numbers in consecutive order,
the SBO ID numbers fall into a different order based on rep units.
And that's what those 111 numbers are that you see in the SBO ID gap.
And then if you look at the, boy, now I'm trying to remember what I meant with those last two
ones.
I'm not sure what I meant with the last two at the moment, and that'd take me two minutes to remember.
So let's just move on.
You're making the point here is that, you know, you've got a sequential number of rows and CIDs, and yet because of the additions here, you're Putting the record now, 111.
The next record is 111 down in the rank, and 111 down further for the next one, and the SIDs are equally incremented by 111.
Yeah, and by the way, actually I should stop there for just a second.
So let's put this in context.
If you walk into the County Board of Elections, and let's say it's three people in a row.
It's you, me, and I forget your producer's name, But if we were all in line, one after the other, and we got our county ID numbers, we're going to get CID numbers 100,000, 136, 7, and 8, just like that, all consecutive.
But the thing is, the county is required to send those to the state electronically as soon as those numbers are done.
So they go straight to the state.
So what I would expect to see at the state level is sequential SPID numbers.
Okay, so for the three of us, they'd be either 130, you know, the equivalent, it'd be like 891, 892, 893.
Or if they had a couple of transmissions come in from other parts of the state at the same
time, they'd be off by a little bit, but they would not be off by 111 exactly between each
person.
And then that certainly isn't going to shift based on a number of other parameters I discovered
to different powers of 10, so like 1111 or 11111, et cetera, which also happened.
So this is in fact the algorithm.
It's putting in these additions by these rep units, and down here further you do define all the rep units and show how the different rep units, at some point or other you list them here, I believe, but you've just listed them out and said what they were.
Here, these are rep units.
Full rep units like 1,111, 111, you could have 11,111, etc.
comma 111, 111, that you could have 11,111, et cetera.
And we're gonna talk about, to some extent, we're not gonna get deeply into it,
rep units that add one, that's kind of like end a sequence of those rep units,
25% rep units and 75% rep units, which end up in 11111 being 8333.25,
three quarters of the rep units.
Now, those are a little bit harder to understand, but don't get confused at this point, because what we're doing here is we are adding, Creating a new number, the state board of identification, state board of election ID number, that is not in sequence.
And it's not in sequence, not randomly, but by a scheme that's predictable and you can retrieve the scheme.
So therefore, this kind of additions of these numbers is the imposition of the algorithm.
Okay, now Andrew, do you want to take it to the next step?
Where do you want to go next?
Go back to that image again.
Yeah, okay, so what Jerome is talking about here is what's called a deterministic algorithm, which means it can be reversed.
And that is part of why this is a security problem for the state, except for the fact this is all public data, so supposedly there's nothing to secure.
But in this case, this deterministic quality is what makes it so unusual, also because it's so complicated.
This part of it doesn't look particularly complicated, but it gets very complicated when you get into it.
Okay, let's go to the next image.
I only have an idea what the next image is because you just scrolled past it.
Ah, there we go.
Okay, so go down a little bit so we can see the headings.
So here, let's see what I'm looking at.
The point of this was that you were talking about... No, I want to see the caption.
Let me see the caption.
Right.
Okay, so let's go down a little bit so I can see the headings now.
I'll actually make this a little reduce it again so it's more so you can talk about the.
Okay, well, the point of this slide is that I wanted to look at the different ways these numbers could be ranked.
So you can sort them by the registration dates, you can sort them by the CID numbers, and you can sort them by the
SBO ID numbers.
And you get a different result depending on what you're doing.
The registration dates, which are the one that's, in my opinion, the most sensible way to sort these things.
And if you look at the county voter roll databases, that's how some of them do it.
They'll sort either by registration date first, or more likely, they'll sort by last name, first name, and then registration date.
So most of the county rolls I've seen are like that, and it would make it absolutely impossible to recover the sort order here if you're basing it on record number, because they don't sort that way.
The CID sort gives you some kind of, let's see, This is Jefferson County.
I'm not seeing the rep units here.
You know, this is the wrong slide.
Let's go.
I'd have to read the paper again.
It's been so long since I wrote it to remember what this was about.
You're basically showing that, you know, when you sort here by the state numbers, your registration dates are all over the place.
Oh, right, the ranks are different.
The ranks are different.
Yeah, that's what I said.
So your point is, you're saying that, you know, when you sort here by your date of registration, will the State Board of Education, of election, ID numbers aren't in order?
Right.
Okay, and then if you sort by the CID number, again, your SBOE ID numbers are not in order.
And your CD rank, however, of course, because you've got the CID here in sequence, is in order.
But when you sort here by the state order, again, your registration dates are all off.
Yeah, the point is that none of these things correlate with each other.
They don't correlate.
And that was what I was getting at.
And the reason is because of the algorithm.
The algorithm is telling us that, or actually it's forcing a lack of correlation between each of these sort mechanisms.
Yeah, out of order.
They're out of order because of the, the only thing that's in order is the State Board of Education numbers.
Everything else is out of order now.
State Board of Elections, yeah.
State Board of Elections.
Everything else is out of order.
That's what the algorithm does because, again, it's moving the State Board of Election number up by 1 or by 1-1, 1-1-1, or 1-1-1-1-1.
It's the addition factor here throws its ranking off what you would expect if it was just sequential and normal.
Next one gets the next number.
Yeah, speaking of efficiency, that actually makes the database in some ways less efficient.
Because if they hadn't done that, if they had assigned these on a first come first serve basis, then clerks working in the County Board of Elections office, or actually I suppose the state because the state numbers aren't available at the county level, But they would have a rough idea of which numbers were assigned in which years.
So they could look at a number and have a pretty good idea when it was assigned.
That could narrow their search parameters, thus making search more efficient for them.
Doing this actually makes those types of search activities more difficult.
Okay, let's go on.
Let's just go to the next image.
Those seem to be the, okay, stop here.
Yeah.
Okay.
So this is talking about the partitions and the partitions are super important because the algorithm is hidden within a partition that is hiding the fact that there's an algorithm to hide.
Okay.
This is where steganography comes into play again.
And this is actually the first thing that you have to find in order to discover the algorithm.
So at the top under table three, you see a minimum and a maximum, and then you see the words out of range, in range and out of range again.
So what it is, is that they have a total of 99,999,999 numbers available to assign.
So that's the number space.
You can think of it as a square box into which they've got almost 100 million numbers.
Now, of course, they're only assigning about 21 million, so they've got a lot of extra space left in there.
But they assign them to the ranges that you see here.
So the in-range numbers, between roughly 8.5 million and 40.5 million, that's where you find the algorithm that we're talking about.
And the out-of-range areas, you've got different algorithms that are being used there, and they have a pseudo-random quality to them.
Which basically disguises the presence of the in-range numbers.
So if you're looking for any county's records, like for instance Onondaga County, you'd be getting the out-of-range and in-range numbers.
And notice that the out-of-range numbers appear at the below and above the in-range numbers.
So for instance, if you were to sort all the numbers in the database by SBO ID number and then just search through it, You would give up out of boredom before finding or running into the in range numbers.
And if you looked at it from the other side, the same thing would happen because there's so many of these numbers to sort through.
It takes ages to do.
So, the likelihood of discovering the in-range numbers is really low.
Now, the thing is, the in-range numbers are assigned by county, and this actually kind of makes sense.
So, let's say you have 10 million numbers total to assign to five counties, and you want to make sure there's no overlap.
One way to do it is you could say each county gets 2 million numbers, you know, 0 through 2, 2 million and 1 through 4, etc.
And just give them those ranges.
And that way, you wouldn't have to worry about overlap.
But in this case, it looks like what they did was they took whoever was already registered on the day they implemented the algorithm, which, as far as I can tell, was mid-2007, and they gave them these highly specific ranges with really odd numbers attached.
So, I mean, like, why would they start with the very first number in range?
It's 8,502,559.
To this day, I don't know why they would do that other than to obfuscate something.
They certainly had enough room in here without unassigned numbers that they could have made it an even eight and a half million or eight million or even two million or something like that.
But nevertheless, they start with this oddball number.
And then they've got a buffer of 500,000, and they've got all these other ranges, and they're all assigned to these different counties.
There's 62 counties, and there's five buffers.
The organization of them is kind of strange also, because if you look at the county codes, you see these are all out of order.
But if you look at the SBID numbers, they're all in order.
If you put the numbers in order, the counties are out of order.
The counties, by the way, these codes are assigned by the state and they are alphabetical.
So the first, county number one is Albany, because that's the first in an alphabetized list.
And Allegheny is number two, because that's the next one.
Bronx is number three.
Broome, I think, might be number four or five, something like that, and so on.
What they're doing here is they're making these partitions, which hide the algorithms, and then within the partitions, they're actually scrambling the county codes, which by the way, cost me quite a little while to figure out.
I think it was I might have spent three or four days on this because I had started with Allegheny and
After I got to this number twenty million three twelve one one eight. I was expecting the next number to be
County three which was Bronx and it wasn't the Bronx It was something else.
It was totally random.
I have no idea what it was offhand, but that threw me off.
And so I was thinking maybe I was wrong about these partitions.
So I decided to persevere and I eventually worked it all out.
But these partitions are important.
Okay, let's continue.
Okay, this is interesting also.
So this has to do with when the The records were given these numbers and which algorithm they're using.
And although I made this particular image, it's based on something made by somebody at NYCA, so I'll give her credit for that.
Um, but in any event, what it is, is that the, uh, the light colored bars here represent in range numbers and the dark bars represent out of range numbers.
Um, and then if you, if you look, well, actually, I guess you see that in the legend right up here, but the thing is, is that you see in range numbers pretty much stop at 2007 and more specifically around June 15th.
And the out-of-range numbers start at about that date, okay?
Now you will see in-range numbers that penetrate into this upper zone, but you don't find the out-of-range numbers going in the other direction except for Those numbers that are assigned to what I call the shingle algorithm.
And every single one of those files is suspicious.
They are either, well, first off, they're almost 100% purge.
So why are you giving them an ID number?
And if you wonder about that, you might be thinking, well, what are you talking about?
You get an ID number and then you purge the record after something happens that causes you to purge it, right?
Well, the idea that almost 100% of all the numbers assigned using a certain algorithm are purged tells me that they knew they were purged at the time they gave the numbers and that those numbers were reserved for purged records.
Now, if they were taking, for instance, all of the purge records that existed as of June 15th, 2007,
and then partitioning them just to keep them separate from all the rest of the numbers, I could understand that.
But if that was the case, I would find no purged records anywhere else.
And that's not true.
So in the in range section, I'm making hand gestures and you can't see them, so I moved my camera.
Actually, let me, I can quit sharing the screen if you want me to for a minute.
No, no, no, no, it's okay.
So anyway, so the in range numbers have something like, I think around 40% of them are purged
in those dates before 2007.
So they actually had a certain set of purge records that for some reason were different from,
or exceptional related to all the rest.
And so they gave a completely different, also complicated algorithm to those
in order to map the CID numbers to those SPOID numbers.
But in addition to being almost all purged, and when I say almost all,
I've been able to identify 99 point, I think it's five, 4% of them as purged.
But the pattern is actually really intricate and visual, and it takes a lot of time to fully extract the numbers.
So I wouldn't be surprised if it's actually 100%, but I haven't fully extracted
the numbers that don't belong yet.
So But in any event, a lot of them are what appear to be illegally duplicated records as well.
And when I say illegal, I don't necessarily mean nefarious.
What I mean is that they belong to people who already have ID numbers.
Okay, so how that came to happen is a mystery.
But the fact that the law says it shouldn't happen at all is a fact.
And actually, I've had several commissioners tell me exactly that.
I think I'm comfortable saying it.
Okay, so let's go on.
And I want to make a comment here that, so first we talked about how the the rep units were added in to change the order of the state board of election IDs in relationship to the mapping from the county IDs.
So now you've got a new set of reordered state numbers that you're going to work with.
Now, Those are going to be the numbers, those are very important numbers, but you've got to also figure out how to hide them further.
So let's say you're essentially planning here to mark cards.
We're going to come to that in a minute.
You mark cards, you want to order them so that unless you're the card shark, you can't see that they're there for a particular reason.
But if you're the card shark, you've got to know where to find them.
So you want to hide them in the deck so nobody else can find them.
There's no real systematic way anybody else can know where they are, but you know where they are because you've identified them one way or the other that you can find them.
So, for instance, even when you say, you know, here's three decks of cards, pick a card, any card, I'll tell you what you picked.
And you pick a card and you hold it up so the That the magician can't see it.
Well, the magician's got coded on that card what the card is.
So the magician can read the card from the dots or however they've coded it.
So you've got to have older numbers that you scramble into a deck so that only you can find them.
Now that's the next part of the exercise and why it becomes so difficult and why these partitions become important.
Actually, one thing I want to comment on what you just said before you leave that other image.
Could you go back to the previous one?
I can, sometimes without even looking at this list, because I've actually memorized some of these number ranges kind of accidentally as I did this work, but if you came to me and told me that you lived in Allegheny County and you had an ID number that was 22 million something, I would know that you moved there from another county.
Right.
And it's just because I know what number Allegheny gave out.
Either that or you got your number after 2007.
But I actually can say things about people based on their ID numbers.
Okay, continue.
Okay, and we've seen that one.
Seen that one.
OK, so let's see what this is.
This is the sort methods.
OK, so when I first saw this, I saw it as you see it on the left.
OK, because I I was looking at it sorted by CID number.
Now one thing to keep in mind, I put the three most relevant pieces of information right next to each other.
The registration date, the county ID number.
And this data.
Right.
Did we lose?
I'm here.
OK, good, keep going.
OK, so in the actual database, these fields are spaced rather far apart.
OK, so you're not going to have the columns right next to each other.
So if you sorted by county ID, you would not be able to see what happens with the state IDs, and most of the time you're not going to have a reason to do that anyway.
While looking at the filtered numbers where you're only looking at the numbers that are in range, which is what we're looking at here.
Um, so you normally wouldn't see this anyway.
So what I did was I looked at them and I was looking at it.
This might have even been the range I was looking at because these look like Allegheny to me.
Um, so I was looking at this and I thought, huh, that's funny.
These all start with 13,000 and then there's a break.
And that's back to 13,000.
And there's another break.
And what do you know?
That number is right after this number.
So it's 1340 and 1341.
I don't understand the one at the top, but I don't care because I'm going to just keep on looking through this list and see if this pattern goes any further.
And indeed it did.
I found that every 10 records would be followed by an 11th record.
that had a that belonged to a different number sequence okay so it's like somebody took two ribbons that had numbers on them and cut one of the ribbons up into little slices so you only had one number per slice and then the other one they let have 10 numbers at a time then they inserted one number after every 10th And then I found that they also inserted another number
every 100th and every 1000th and every 10,000th, okay.
And you can see that these numbers are, so one and then 1340,
and then the next break number is 1341.
But those are the break numbers because they're totally out of the sequence,
especially from 214 through to 26, 26. Yeah.
And these numbers, too, are a reasonable sequence.
But the 1341 is only four digits, where the other numbers above it are five digits.
So that comes from some other series of numbers.
Yeah.
And actually, another thing I want to pay attention to here is the registration dates.
If you look at these registration dates, they're all over the place, okay?
The number one is 1971, then the 1340 is 1967, and the one on the bottom is 1972.
And if you look in between, you've got all these different years.
And, you know, like, look at this.
The second one is 1995, which is quite a bit.
Is that the latest one?
Yeah.
And then you've got a couple from 1992 in here.
They seem to be scattered all around, meaning the registration date didn't make any Yeah, they're not ordered by registration date.
The registration date has no relationship to the ordering of the CID or the State Board of Election ID.
Right.
So anyway, so to achieve this result, you have to sort by the state ID.
Which, by the way, I found out... The second group of numbers, this column... No, the first group of numbers is sorted by... Oh, yes, yes, yes, yes, that's right, okay.
Yeah, and it turns out this is the best way to look at it, because although if you sort by the county ID, you get the rep units visible in the SBU ID gaps, it turns out that if you do it this way, as you see in the middle chart, And you have a county that uses alphanumeric ID numbers.
It screws it up completely, and pardon the language.
It makes it very hard to detect the algorithm.
But in any event, the usefulness of this is when you're in a county that doesn't have alphanumeric ID numbers, you can see the rep units, because now you see what the separation is between each of these state board of elections ID numbers is.
Um, so when you and I and Chris are standing in line at the County Board of Elections and we get numbers one, two, and well, I don't know why this goes straight to four, but let's just say three there.
Okay.
Um, we're going to get these state ID numbers and they're going to be spaced by 2778, which is one quarter of a rep unit.
And then 11,111, which is the largest rep unit that fits in that, um, county.
And then it's going to be followed by the remainder here, which is 8345.
And then it's going to be, because it's switching, it's downshifting from the power of 10,000 to the power of 1,000, right?
So whenever it downshifts, you get a big, you know, an odd number like this, because that's the remainder.
So then you get to the 1,111s.
This is when it hits the ceiling.
So this is near the cut line.
So that's actually the highest number, and then it goes to lower ID numbers.
So then you go through the rest of the 1,111s, and it's going to drop to the 111s, etc.
So that was kind of interesting.
The reg state sort, we already went through that, so I'm not going to talk about it again.
Let's keep going.
And we already went over the rep units, so let's go to the structure.
Now, this is important, okay?
So that 2778 gap that you saw in the SPID numbers in the previous image was one quarter of a rep unit, right?
So why does that happen?
The reason is what they're doing is they're taking the list of numbers and they're shifting them by 25% of a rep unit.
And that's what you see going on with the Caesar cipher.
The Caesar cipher is a cipher that was used in Roman times and actually used by Caesar himself.
So what you do is you take all the letters of the alphabet and you shift it three characters.
So all of a sudden what and what by doing this what they do is they are remapping the values.
So now A equals X, B equals Y, C is Z, D is A, etc.
Okay, so they would write their messages this way.
So this is the same principle as what's going on in New York's photo roll algorithm.
Okay, because what they've done is they've shifted all the numbers down and By not using a full rep unit for the first number, okay?
So in other words, just to explain a little bit more detail, so your first line, you've got the alphabet A through Z. By shifting, and then you take the last three letters and you put them at the beginning, A suddenly becomes X.
B becomes Y, C becomes Z, D becomes A, etc.
So you've got here a remapping of the alphabet onto the cipher.
Okay, and it's an ingenious cipher because all you do is move it over three spaces and you take what the last three numbers are and you put it at the beginning.
And that's how you're creating, you're shifting of the numbers to further obfuscate their placement within the deck.
Okay, and this chart here is going to explain that in a little bit more detail, but I'm going to let you go into that, Andrew, but I'm correct in this, right?
That's how this works.
Yeah, actually, I'll say one other thing.
A lot of skeptics, and by skeptics, I mean people who don't like Donald Trump and Don't like the fact that he got elected in the first place and don't like any comments that suggest he might not have lost the last election.
Tend to complain about this aspect of the algorithm saying, oh, the Caesar cipher is so easy.
Everybody who's into cryptography knows that.
Nobody in their right mind would use that.
Well, in this case, they did use something very similar to it, in addition to some other things that made this very, very complex.
So, to suggest that this is simply a Caesar cipher is simplifying this beyond the point of recognizability.
So, if you were an example of what they did, then the kind of thing they did, then precisely what they did.
Yeah, yeah.
So in this particular case, this is a preliminary version of how I was describing these things, where I called the different powers of 10 a strip, right?
So you'd have a strip of numbers that would all be separated by 111.
So that would be the power of 100, right?
So that would be the power of 100, right?
So, and then you'd have the 11,000s, those are the power of 10,000.
And so basically what it was is that the higher powers of 10 had the smallest number of numbers
So you see, for instance, in Allegheny County, you've got these nine numbers in strip one.
And what those represent are the 10,000s and possibly the 1,000s, because at that time, I didn't realize that that's how they were breaking it up.
And so I was breaking it up a little differently here, but the idea is the same.
So you have the vast majority of numbers are in the ones, which is all the way over here on the end, where you've got almost 100,000 numbers attached to it.
And by the way, I don't mean because you've got 100,000 people in Allegheny County.
I mean, those are the number values that are assigned to that range.
Between 999 and 100,000, you've got this number of numbers.
Yeah, so what they did in Allegheny, and they didn't do this in every county, but what they did was they actually used a power of 10 range of numbers For each of the power of tens.
They don't do that always.
Sometimes they're more or less continuous.
So those kinds of local variations between counties also obfuscate this a little bit because you can't use the same solution from one county to the next.
They're very similar, but they're not the same.
They usually have they introduce like a new rule of one kind or another.
It's kind of like my name.
You can spell it with two T's and an E, but some people in France spell it with one T and E. It's the same name, but it's spelled differently, so if you're doing a computer search for it, you might find it one way and not the other.
Okay, so let's go on to the next section.
Okay, this looks like a mapping of Allegheny County's schematics.
Okay, so This is a little tricky to read, but let's see if we can explain this, okay?
So, the first strip of numbers starts with 9, ends with 17.
The second strip, notice the 10, which comes after 9, is not the first number, right?
So this is the first number in the segment is 17, right?
It goes all the way up to 99, and then it goes to 10.
But 10 is the number that follows 9, okay?
And then you look at the end number is 16, which is just before the first number, which is 17.
So this just shows the shift.
Okay.
That we were talking about before.
So what they've done is by, by shifting the numbers by one quarter of a rep unit, they've pushed the, what should be the first number, which is right here towards the end of the list of numbers.
Okay.
And then they do the same thing for all the rest of these.
So this ends at 99 and then over here, It starts at, oh, in this case, they have an alternating sequence plus a straight sequence.
So it stops here at 9.11 and starts at 1.03, which is probably, there's a couple of missing files after 100.
That's why it's 1.03 instead of 100.
And the 9.13s continue until you get to the full normal sequence.
Which starts at 115, it goes to 145, and then wraps around itself to 146.
The alternating sequences, which are in grey, are a little tricky, and I think that if you looked at these using my new method of analyzing it, it'd be a little clearer than this.
But this still shows you the principle of how they're shifting the numbers and how they wrap around each other.
Now, the deck stacking, which is what we're seeing next, It shows you how a card cheat would cheat a card.
Okay, in poker anyway.
So let's say he had a confederate playing poker and he wanted to get a Royal Flush.
And you want that to be player 3.
So what you do is you want the 3rd, 8th, 13th, 18th, and 23rd card to be these cards.
So what he does is he opens up the deck, he counts the cards, and places those cards exactly where they belong.
That way, our player is going to be dealt the cards that he is meant to have so that he can win.
But nobody else knows that this is going on, and he doesn't even have to mark the cards.
They're just going to be dealt straight out to him.
Now, these player segments that you see here correspond to the strips I was showing you before, or the power of 10 groupings.
What it really means is they've introduced predictability into these These numbers that they're mapping to each other.
And that predictability is what makes them volatile in some ways, okay?
Because I can now... Oh, here's my favorite diagram, actually.
So this tells you where they start and end.
This is Yates County.
So you can see the ID numbers are here.
This is the state ID number.
And this is the lowest one and the highest one for each one of the strips.
And each strip is defined by a power of 10.
So you see 1, 11, I don't understand why this also says 11.
That looks like it's a mistake.
This looks like it's, this is a hundred, and yeah, this is a hundred and eleven, and this should be a thousand eleven, and this should be eleven thousand one eleven.
So these are off by a power of ten.
And you know that because 83 plus 28 is 111 and 833 by 278 is 1111, etc.
So anyway, but the point is is that they all start on a three-quarter rep unit and end on a one-quarter rep unit, okay?
And then they reverse that with the highest power of 10, which I've seen in some counties, but not in all.
But again, the point though is it's predictable.
It's very, very predictable.
So if you know the algorithm, You would be able to tell which state ID numbers are matched to which county ID numbers.
And that's valuable information, because now what you can do is you can separately reference those without anyone knowing what you're looking at.
So this is basically constructing an algorithm assigned ID, which I call an AID.
And so let's say somebody's AID number is 45, their SBO ID number could be something like 22 million something or another.
Their CID number might be something else.
It's totally different.
So if you have another database that's using the AID instead of these numbers, and somebody finds your database, they'll have no idea what it means because you don't have any of the numbers that are found in the normal database, and yet you're still working with files that are part of the state voter rolls database.
But no one would be able to tell that.
So it's a perfect way of hiding what you're doing.
Can you go down farther?
I'd like to cover this part right here with this paragraph here on the strips where you come up with a example 0110060311.
Yeah.
That paragraph.
Uh-huh.
Well, are you asking me to talk about it or did you want to say something?
Yeah, no, I'm saying, what I'm saying is that when you follow How the numbers are arranged in the different strips.
You can identify, so you've got AID, which is the new number that you've created to identify that record.
It does not exist in this database, it's in your database, and it references the number in the database.
But you also created a positional number to tell you where it's stacked in the array.
Okay, so the AID does not appear in any other field of the voter database, making it inaccessible for normal use.
But you've created an AID that essentially, positionally, tells you where to find the record that's important to you, the one that you've marked, the one that you've altered, the one that is a fake, a clone.
Actually, one thing I want to add about this is that at the time I had published this article, this idea was a theory based on what I was seeing, but I have since determined what those IAIDs actually are, and I do not consider it a theory anymore.
Could you go back up to that paragraph?
So, you see that number, the 0110060311, that's based on, you know, which position each part of the numbers appear in, in the strip diagram that I showed.
The actual AID is much simpler than that, and it's smaller and more compact than the actual SBOID numbers.
So more likely the AID would be a number between 1 and, in this case, I think 20,000 or so.
So it'd be a much shorter number.
But that's... You're just identifying the marked records, not all the records, just the marked records.
Yeah, within the county.
So if you know the county and the AID, you can find it, and you can find it with fewer numbers than if you were looking at the state ID.
Right.
Okay, now if we take a look, let's stop at this for a moment and not go into it any further.
It's pretty much the end of the paper.
I don't want to go into the rest.
That's a little bit, let's just stay at this much.
You're missing my favorite slide.
Okay, I'll give you your favorite slide.
Go ahead.
All right, so this is really important.
So if this is a plot, and actually it's two images, but this is a scatter plot of records with the CID numbers on the on the bottom axis, the x-axis, and the SPOID numbers on the y-axis.
Okay, so each one of these x marks represents a number of records that correspond to You know, these, these number values.
Now, when you look at all of the records in the out of range partition, you have all these vertical bars and then you have some horizontal bars.
Okay.
But if you go down to the next image, all the horizontal bars are gone.
And that's because these are all the active records.
Um, which is kind of interesting because all the, the horizontal ones belong to the shingle pattern and they're all purged, which means to me anyway, They knew they were purged at the time they assigned the numbers, and there's no point in assigning a number to somebody who's been purged because the purge status means you're ineligible to vote.
So if you're ineligible to vote, why are you getting a number?
Now, again, I do think it's possible that they had existing records that were already purged when they maybe changed the system, and so they decided to partition the space in such a way that these guys got a special algorithm.
But if That was what was going on.
It would be what they did for everybody, and it's not.
So there's something different about these records versus the records that are handled with the other algorithm.
OK, but that's that's all I wanted to say about this.
OK, and now I want to go to what this results in, because it's very beginning of the paper.
What you make clear is that there's a lot of anomalies in this database that you're trying to explain to begin with.
So, for instance, What potentially is trying to be done here?
And again, this is the best construction we can come out of.
Every time you create an algorithm, you create it for a particular reason.
You want something to happen.
And that algorithm is supposed to produce a manipulation of the data to produce that result.
Now, what we're saying is, if this is an algorithm that, if you wanted to have false records introduced into the voter rolls, So the records belonging to false voters were covertly tagged via an algorithm for easy retrieval when needed.
And the absentee ballots were requested by the false registrants.
So the ballots and ballot envelopes were gathered at central collection points.
Fraudulently generated ballots were cast and they fraudulently obtained ballot envelopes.
False voter records were updated to reflect false votes and after certification false voter records were manipulated to disguise their purpose in history.
In other words, in a sense I call this kind of like a certification scheme because if you just invent votes and there's no way to tie those votes back to people who voted, you may have far too many votes You don't know who's got the benefit of that, but you know that there's this, you can't have far more votes than people who are registered.
That's one kind of anomaly.
But here, the more effective way, which I kept wondering in 2020, you could see these apparent, not proven, but apparent Um, irregularities, ballot, no, the voting stops and then these ballots, mail-in ballots appear and they're run through the machines.
And then Biden's numbers go above Trump numbers and that pattern persists.
And you say, well, how could that be?
And then it certifies.
I was sure it wouldn't certify.
Because if you were just producing ballots and running them through the machines to get the results you wanted, that could be detected.
But here, you're saying, OK, I need 20,000 votes.
This is hypothetical.
So of the pool of clones I've created, which are false votes created from real records, I can pull 20,000 of them.
And I can have those 20,000 request a mail-in vote.
So that when the ballot's printed, it has that voter's number on it, State Board Elections ID, and when they look at the end of it to certify it, I've got a voter with that false ID, I've got a vote with that false ID, I don't know that they're false because they appear in the database to be legitimate numbers, and that vote's certifiable.
So that's, I think, one of the ultimate Geniuses of this system is that you can cast certifiable votes that may have been created to be fraudulent in the beginning, and that only you knew where they were, only you knew how to find them, and yet when they're requested, they can't be detected that they're false votes.
And we have a lot of different, it results in There are hundreds of thousands of illegally generated registrations in the official New York State Board of Election voter rolls.
The exact number is unknown, but it's not less than about 338,000 for registrations active for the 2020 general election.
It's according to the NYCA in 2022.
If other elections are included, the number apparently of illegal registrations jumps between 1.2 and 2.4 million.
Yeah, actually, I kind of want to comment on these numbers.
These are generally accurate, but one thing to keep in mind is that there are so many examples of anomalous data in the voter rolls, that is to say erroneous data, that any attempt to count how many people voted is prone to a decision tree problem, because you have to decide Am I going to count this record or not as a legitimate record of a vote?
So, if the voter history says the voter voted, but we know that the voter history has been deleted in some cases, does the absence of a vote mean the person didn't vote?
We know for sure that that's false in some cases, actually quite a lot of cases.
If, on the other hand, it says the person voted, but we have heard that this person didn't vote, or we know that because we're the person or something, Is that a real vote?
And also, if we have a guy who has 11 records, and we know he can only vote once, and let's say four of those records show a vote, do we count that as four, or do we count that as one?
Or do we look at the missing votes as being possible 11 votes for this guy?
So there are a lot of different ways to count it, and depending on how you count it, you can wind up with a discrepancy, that is to say, a shortage of voters who voted, or you can have a surplus of voters who voted, relative to the certified vote totals.
So when New York Citizen's Audit is talking about the 338,000 number in their publications, they're referring to their methodology for counting them, which is, you know, it's exactly accurate for the way they did it.
But there's more than one way to do it.
None of them reconcile with the certified vote totals that I know of.
In fact, I think the closest you can get is about 70,000 off, something like that.
But on the other end, for the 1.2 to 2.4, At the time I wrote that, that was before I did a little bit further research that has narrowed that number down.
I would say it's a fairly solid 2 million right now.
It's not 1.2.
It's not much less than 2 million.
And it might be a bit over 2 million as well.
Of a total of how many?
21 million.
It's about 10% of the records are affected.
Okay, so 56.93 of all voter ID numbers were assigned based on the primary algorithm discussed in this paper.
The algorithm allows a hidden attribute tag to be added.
On the NYCA is to recover documents related to a fictitious identity with 22 registrations that request multiple absentee ballots sent to the same Reddit mailbox.
And NYCA has identified other fictitious identities like this.
Actually, can I stop and make a comment on both these items?
Absolutely.
On the hidden attribute tag, this is important because what it means is that the algorithm is adding data to the database that doesn't appear in any of the fields.
So this is the opposite of efficiency.
They're actually creating Um, the equivalent of a lag in the database that would impair its function.
So all by itself, that I think is very interesting.
It's against normal IT protocol to do something like that, to add an unnecessary or superfluous type of data to 21 million records.
It's, you know, the funny thing is, is it might not seem like a big deal to add a field to a database, but then you have to multiply it by the number of records in that database.
And it can become a very big deal indeed.
And as far as the fictitious identity with 22 registrations, I later discovered three more for the same person that were slightly misspelled.
So it's actually 25.
And in this case, they've also been able to verify that in the state rolls, it doesn't show him voting, but some of the counties do show him voting.
So his votes are missing.
And now in his case, And I'm not convinced it's a real person, by the way.
I'm actually fairly strongly convinced he's not a real person.
But some of his records have been purged, and they've been purged at different times.
It looks to me as if those records were set up to commit voter fraud in a certain year.
But the thing is, they would have continued to get absentee ballots, because all of them had this tick box marked for absentee ballots.
And if you look at the years when they were first registered compared to when they were purged, I'm guessing that about 300 ballots were generated for that identity alone.
Something like 300.
It's a lot.
Okay, go on.
So the canvassing, going back and checking these records, has uncovered cases where false votes were added to false registrations or genuine votes were erased.
Okay, so that's both directions.
A comparison of four versions of the New York State Board of Election Voter Database.
Roles were created over a 13-month period.
Shows hundreds of thousands of modifications to multiple fields belonging to the same voter ID numbers.
Although there are apparent valid reasons to update these fields, none of the reasons apply
in situations for, in Greene County, a voter with the name initials CS. It's a date of birth of 5-5-1925,
the 2021 database, 2022 database, the date of birth is 8-18-1971.
Both records have the same ID and numbers, addresses, and registration dates. It's the
same person, but birth dates are immutable, they do not change, both records cannot be correct. By the
way, one thing I want to point out about this, this.
There are, if I recall correctly, there were around 10, maybe 11,000 records like that, that I know of anyway, where the date of birth got altered.
And that particular finding was some other guy on the NYCA team.
But the thing is, is that Anytime someone says to you that the database is accurate, and yet you have examples like this where the data has changed, the answer might be, well, we fixed it, so now it's accurate.
And I'm thinking, well, it wasn't accurate before, so I no longer trust you on this point, okay?
And that is true for the rest of these.
Okay, go ahead.
A voter with an issue with RV has three records in the 1021 and the 1221 databases.
The 2021 database, he has two State Board of Education ID numbers.
The 2022 database, he has one.
An examination of the records show that one record with a registration date of 6-9-21 was retroactively altered to change the State Board of Education number.
This is illegal because no voter is allowed more than one State Board Education number And changing this after the fact conceals the prior existence of an illegal record.
Before you go on, I want to first say it's Board of Elections, not Board of Education.
I mean, I continue to do that.
I apologize.
That's okay.
And the other thing is that when they delete that number, There's no way to know what else is deleted with it, okay?
So when there are separate records, the data fields are separate from each other, right?
So you can have the name and the address and the voting history and all the rest of it are attached to each number, but it's all separate.
So when you delete one of the numbers, you may be deleting a voter history that is completely different from the voting history that remains.
You might be merging them.
In fact, I would imagine that if they're smart, what they're going to say is, We're merging these records, so we're not losing anything.
But since they've deleted the other number, there's no way to prove that that's true, because they've deleted the evidence of it.
So they might say, we've got the changelogs, so we can show you that we changed it properly and we documented it.
But so far as I know, they have never disclosed those changelogs to anybody.
They just claim they exist.
So as far as I'm concerned, until I see it or until somebody sees it, I'm not going to trust that.
Go on.
A voter with the initials MP has two state board of election ID numbers, one of which is illegal, and both have a vote recorded for the 2020 general election.
When MP was canvassed, she said she did not vote in that election, nor was she registered at the time of the election.
The voter rolls confirm that she registered almost three weeks after the election on 11-23-2020.
Her voting history does not reflect her actual voting behavior.
And by the way, on this, this is kind of an important point because we have commissioners and other officials who have tried to explain some of these findings and sometimes they are legitimate explanations for a very tiny fraction of The findings or the, you know, the affected records.
And sometimes they don't work at all.
And sometimes they're just confusing.
And that's what's going on here.
This person was interviewed with her mother, and they were very excited and happy to be able to talk to somebody about this because they thought the whole thing was strange.
And this was because on voting day, they didn't like what happened.
But the later explanation from I want to say it was the county commissioner, but I'm not totally sure because I wasn't involved.
But as I understand it, they gave an explanation that differed completely from what we were told by the voter herself.
Okay, so.
And I didn't participate in that conversation either.
So I'm getting the second hand.
But the idea was, she said she went there to vote for the first time in her life, didn't know she had to be registered to vote.
So they told her sorry, you can't vote in this election because you're not registered, but you can go home and online you can register for the next one.
And that's what she did.
But she did it like three weeks later or so.
And, and She was expecting to vote in the next election and then she wound up getting two votes somehow recorded in her two ID numbers that she had.
Now what we were told later was what had actually happened was that she went there and they said you can't vote because you're not registered but we'll give you a, what is it called, an affidavit.
A provisional vote.
That's it, a provisional.
And then you can go home and register, and provided you register before a deadline of some kind, we'll go ahead and let that provisional vote count.
That doesn't explain how she would get two votes.
It also doesn't explain why it's completely different from what she told the canvasser.
So as far as I'm concerned, that explanation is not satisfactory.
Now, the reason I'm pointing this out, and the reason I stopped you, and I apologize for doing that, is I have heard, and in some cases read, Some people try to explain some of these findings, and I
want the people listening to this to be very aware that just because someone gives you an explanation doesn't mean
it's either truthful or an informed explanation.
What I found is quite often the person giving the explanation is giving an explanation based on what they
think should be the truth, but it isn't the truth.
So they may believe it, but it's wrong.
Or it may be true, but only true for a few files.
So in one case, an explanation applied to about 5,000 files, which sounds like a lot, but it was out of a group of a million files, which is not a lot.
It's an incredibly tiny number, and that explanation was extremely defective for that reason.
So if I were One of your listeners watching this podcast, and I was telling somebody about it, and they said, oh yeah, the explanation is X, whatever it is.
I'd be really careful about just believing anything you hear about these things, because I found that the people on the other end of this, either through genuine ignorance on their part, or quite possibly a desire to obfuscate the data, They're quite adept at dismissing these findings, despite the fact that their explanations are rather poor.
Sorry, go ahead.
Well, I just want to, just one more point here and then we'll begin to wrap it up, but this last paragraph here I do want to cover, which is, let me just share here, and it's unknown whether fraudulently gathered ballots were cast, but it's known the ballots were fraudulently requested and
the votes were recorded as cast by voters associated with fraudulent state board of
election ID numbers. These two types of evidence indicate the probability that physical
ballots were fraudulently cast.
Okay, now that's I think the most important conclusion of all this exercise is that
going through all this convoluted number, this algorithm, and cryptology is complicated.
There's no...
If it weren't complicated, it wouldn't work.
It has to be complicated because you'd want your code to be able to be uncracked.
And so therefore, this system is creating fraudulent state board of election numbers, Being able to stack them in the deck so that they are sorted in a peculiar way that's identifiable, has a logic to it, where the ID number, the AID, which is not in the database, positionally identifies where the ballot, where the vote, where the registration, the fraudulent one that you've cloned, is that you might want to recover.
So if you've got the codebook, You can say, give me 20,000 of the ones I've created as clones, and you've got them there, and the computer can then vote them, or request to have a mail-in ballot, which is printed with that ID number on it, such that it appears to be valid.
Now, all we're saying with this, we're not saying this is how the 2020 election was stolen, this is New York, New York is not a state that was even in contention.
But the point is, Andrew, you found this scheme in other states, correct?
Or similar schemes?
Similar.
I haven't found an algorithm that is identical to the one in New York anywhere else.
I found a very strange modification to New Jersey's numbers.
I found scatter plots that indicate multiple algorithms in use in different states.
As you say, Pennsylvania, Ohio, and North Carolina.
I believe There's another one in there someplace, but I'm forgetting it off the top of my head.
Another guy I know in Pennsylvania, a guy named Vico, found a literal tag on the state ID numbers in Hawaii.
That, ironically, is exactly what I was expecting to find in New York, but did not find.
But it would still function the same way.
And I want to reiterate something that Jerome has said a little earlier, and that is that The algorithm would allow people to do what Jerome was saying.
Whether they did that is something that we would need more evidence to be able to demonstrate.
The presence of the algorithm on its own is highly suspicious, and for people who aren't familiar with databases or IT work, you might not think it's all that suspicious, but every expert I've talked to in this field It tells me that there's no way something like this is sitting in there without a good reason, or a malicious reason.
But there's going to be a reason, because it's a lot of work, it's very complex, it's definitely designed to be obfuscatory, and that by itself warrants investigation.
Fair enough?
That's exactly right.
We don't want to make statements that are beyond what our evidence allows us to say.
So the conclusion of watching this is that something's wrong with the New York State database voter roll, and it appears to be wrong in several other states.
Now, if that's the case, and we don't have integrity of the voter rolls, we don't have elections that we can rely on as being accurate.
And that's the point.
So the point is we need to know, you know, how many irregularities there are, if they can be found in this state and other states, if they were used, why they were created, what their purposes were.
And there's a lot more investigation yet to be done.
I mean, I don't want to be running off, you know, saying we found the Kraken, we know the election was fraudulent.
We're not saying that today, but we are saying that You know, there's trouble in paradise here because you don't have algorithms written into databases that are voter dependent because you don't want to manipulate that data and you don't put an algorithm in unless there's a purpose for which you're manipulating the data and there does not appear to be a legitimate purpose because it doesn't add to efficiency, it doesn't add to transparency, it's not
Maximizing any of the legitimate purposes of maintaining a voter roll.
And this is a serious question in a country about to have another national election, which is important to us in the history of the republic, and outcome is going to be very determinative of which directions we go as many elections are.
This one perhaps more than most, or perhaps they just all seem that way.
At any rate, What Andrew has done is very brilliant work that has identified something that merits everybody's attention and much more serious examination and much more detailed asking of questions before we get to the conclusions, etc.
Now, Andrew also has a couple of blogs And the blog on which you do a lot of your work on, continuing the discussion of algorithms plus your cartoon work, etc.
Which is that blog?
Well, this is my website that you're looking at right now.
So that'll get you to my photography, drawings, paintings, commercial arts, a couple of books that I've written, and I'll talk a little bit about my teaching and so on.
The other one... What is your website?
This is P-A-Q-A-R-T dot com.
So that's PACART.
P-A-Q-A-R-T dot com.
So this just shows, oh, these are a bunch of headings.
If you click on any one of these, you'll get a whole bunch more things.
So, yeah, so these are just looks like sketches of people that I do when I'm waiting for my car to be repaired and things like that.
But if you go to commercial art, you'll see my finished stuff.
And if you go to paintings, you'll see some of the stuff I used to sell in Arizona.
Like if you go to the upper right, that's my comic book work.
And so you can click on, let's see the guy with the barbed wire.
Right there on the right.
Just below that.
Okay, well, this is the one I'm working on right now.
This is a science fiction story that I wrote.
This one has a very striking surprise ending that I'm not going to tell anybody because it's going to really surprise everyone.
Anyway, so this is my comic book work.
Go to paintings, you'll see something totally different.
If you feel like it, anyway.
But that's where I have landscapes of mostly the Arizona desert and California.
Your substack for the work you do on continuing on the algorithms, what is that substack?
That's right there.
It's zarkfiles.substack.com.
Z-A-R-K.
Yeah, and then the word files.
And I don't write as much about the algorithm these days as I did when I started the blog, primarily because I've essentially solved it.
And I don't have things to say about it every day these days, although last week has been different because of our conversations is making me think of it.
So I'm looking at things again.
More often these days I write about issues that are in the news.
Politics, for instance.
So like the lawsuits against Donald Trump, which I think are an absolute travesty that should be embarrassing to every American that these cases even went to court.
So I write about that kind of stuff also.
Okay, it's been a real pleasure.
Thank you for your time and all the work you've put in this.
I think you've shed some light on a very important issue in a responsible way, and I commend you for doing that, and we're happy to bring this to the attention of the various states and the federal government.
We cannot have algorithms in our Voting rolls.
It makes no sense at all, and a subsequent question is, why were they there and what did they accomplish?
Those are subsequent questions.
With Andrew Pickett today, and you now can see his website, this is TheTruthCentral.com.
I always say, in the end, God always wins.
God will win here, too.
It's not going to be easy this time around.
But if we join in spirit of 2 Chronicles 7-11 and ask God for forgiveness for letting our land get this messed up, God will hear our prayer and heal our land.
And so, thank you for joining us on TheTruthCentral.com.
We're doing podcasts every weekday, and we'll be back tomorrow.
