US Election reporting is dependent on a few suspicious companies that provide results that are arguably manufactured. 

By on Jul 3, 2023

Guest post by Erin and David Clements

The number of Americans who believe Biden did not legitimately win the 2020 election has grown to 62 percent, with another 6 percent who say they don’t know if he won or not.  An enormous mountain of evidence, including whistleblower testimony, expert analysis, or proven machine vulnerabilities has awakened a huge majority of American voters to the realization that our elections are largely rigged.

As independent analysts and auditors improve their understanding of the centralization of election system architecture, the evidence revealed during the initial aftermath of November 3, 2020, is taking on new and vital importance.  Case in point, millions of Americans that watched election results witnessed impossible changes to their vote tallies.  Millions saw the “F-Curves” demonstrating an inexplicable injection of hundreds of thousands of votes in key races across the country.  The F-Curve has come to represent rigged elections and corruption.

See “Fraud Curve” Memes from Social Media above.

A trusting public relies on “Election Night Reporting” or “ENR” to find out who won on election night. But ENR broadcast on election night is not a report of real tallies but a tool to shape perceptions about election results. What was reported on television was perceived as reality and races were called by the media – outcomes that that any political or judicial institution was loath to challenge after the fact for fear of media bullying.

Election Night Reporting is another smoking gun that proves our elections are centrally manipulated.

Sources of Election Night Reporting Data

There are several places to find ENR data.  All secretaries of state (SOSs) report results on their websites and some counties join in reporting efforts. But the real powerhouse in election night reporting comes from a foreign owned, Spain-based company, called Scytl.

Scytl collects all state ENR data and provides it to Edison Research.  Edison works in tandem with all the legacy news agencies that televise election results.  The vote tracking features you see on the bottom of the screen during an election on Fox and CNN all come from Scytl-Edison.   Two corporate entities with no Congressional oversight, one with foreign ownership—have a monopoly on what the public sees on election night.

This self-evident national security threat wasn’t lost on the media during the Trump administration.  The Guardian warned of the threat posed by Scytl having total control of ENR, and Dominion Voting Systems possessing a near monopoly on tabulation. Those concerns, however, evaporated after Joe Biden was installed.

The Fraud Curve

Examples of ENR anomalies include one viewer noticing an exact 19,958 vote switch from Trump to Biden that occurred in Pennsylvania, and another observing 351,000 votes disappearing in the vote totals during the recall election of California governor, Gavin Newsom.

But what became the most famous and unexplainable ENR phenomenon were the datasets showing thousands of votes being injected for one candidate, while simultaneously showing a negligible or nonexistent increase for the other candidate.  This became known as the “F- Curve” or “Fraud Curve” in election integrity circles because of the shape the curves made in the plot of total votes for each candidate over the election count:

“F – Curve” or “Fraud Curve” appearing on a plot of cumulative votes for the 2020 Presidential Race in Wisconsin from this article.

Unsurprisingly, after the F-Curve materialized in the ENR datasets, the losing candidate would leap ahead to victory – typically in the middle of the night.  The swing states of Wisconsin, Michigan, Georgia, and Pennsylvania all exhibited an F-Curve after each had stopped counting at around 10 pm on November 3, 2020.

At the TCF Center in Detroit for example, Michigan, surveillance video showed a white van delivering thousands of ballots in the middle of the night, well after the delivery deadline, at the same time an F-Curve showed up in Michigan’s ENR data.

The “F Curve” has since been found in dozens of races in the 2020 and 2022 elections.  This includes the January 5th, 2021, runoff-election in Georgia where the fate of two U.S. Senate races were decided.  After the runoff-election, Democrats controlled the White House, the House, and the Senate.  The following year, candidates like John Fetterman, who campaigned in a hoodie and shorts with glaringly obvious cognitive impairments, defeated a well-known public figure in Pennsylvania.  Katie Hobbs, who ducked the debate stage and hid in a bathroom from reporters, was also able to ride the fictitious perception wave provided by ENR to a stolen victory in Arizona.

ENR Exhibits Centralized Manipulation

Most local election officials take criticism over the 2020 election personally.  They vigorously defend the handling of their elections.  And most have a good-faith belief in the vendor-provided talking points of how decentralized their election processes are.  Evidence of centralized control of election night reporting, however, has shattered those talking points.

After the 2020 Election impossible anomalies were identified in multiple states.  We found in Georgia and numerous other states that there were patterns in their Edison data that made no sense.

We labeled this “The Drop and Roll”:

Multiple analysts have confirmed the above analysis and found them in multiple races in dozens of states. Former systems consultant and MBA, Todd Buffington, is one of these. He found evidence of the F-Curve in at least 20 states. Buffington also confirmed that the Scytl and Edison datasets are identical. (Comparison of Nebraska SCYTL vs Edison Research Data (

Buffington’s analysis of Georgia further highlights the impossibility of the ENR data. ( Up until Line 371 in the data, Trump was ahead of Biden. At Line 372 of the data – 5,972 votes were added in a batch weighted 91% for Biden, bringing Biden’s total to exactly match Trump’s total of 2,447,591 votes.

Buffington Analysis 1. Multiple Batches in a Row with Same Ratios Between Candidates

While a batch coming in 9 to 1 for Biden to exactly match Trump’s total is alarming, what is observed next in the ENR data is impossible.  The next 55 batches of votes added to the tally are split exactly 50/50—such that Trump and Biden’s totals remain equal for the next 55 batches.  Even more bizarre, is that in many of these batches, there were an odd number of votes, making it mathematically impossible to split batches equally between the two candidates.

The likelihood of any of this happening randomly is likely zero.  Election night reporting has a tenuous tie, if any, to the reality of what is occurring in each election jurisdiction.  The only explanation for the phenomena observed is that votes are being fractionalized and automated through an algorithm. The algorithm being witnessed throughout the country is known as a proportional-integral-derivative control function, or “PID control.”

PID Control

The discovery that ENR data is little more than perception-shaping theater is demonstrated by the work of Draza Smith.  Smith is a control system expert and was a senior cyber engineer at Sandia National Labs focusing on Cyber and Grid Security. Smith’s extensive research into election night reporting and the data produced by the election system has proved that ENR in all 50 states is automated.  Smith’s work can be found here discussed in detail here. (

When Smith looked at the ENR data, rather than looking at the repeating ratios of vote distribution in each batch as discussed in the previous section, she looked at the ratio between the cumulative distribution of votes for each candidate. For the presidential race, that would be the number of Trump votes divided by the number of Biden votes at each timestamp provided in the ENR dataset – or what she calls the “Trump to Biden ratio.”

When she looked at the data this way, she noticed that the resulting curves resembled a function she was very familiar with as a control systems engineer – the function she saw in the ENR data is called a proportional-integral-derivative control function, or PID control.

PID control is a commonly used function in car cruise control systems and home thermostats. It is used to keep a system in a steady state or gradually move it from one state to another. For example, let’s say it’s the middle of winter and is 55 degrees in your house. You turn your thermostat up to 68 degrees. Your furnace will turn on at its highest setting and start to warm up your house.  As the temperature approaches the set point of 68 degrees, the system switches to a lower setting until it warms your house to a temperature slightly above the set point, and then it turns off. Wait awhile and your house will cool down again to a temperature below the set point of 68 degrees. Then your furnace will fire up again on a low setting and warm your house slightly above 68 degrees and turn off again. This process will repeat indefinitely until you move the set point of your thermostat.

The exact same behavior in the Trump to Biden ratio was found in the ENR in every state. Plotting the Trump to Biden ratios over the “count” of the election looks exactly the way a plot of the temperature inside your home would look.  Below are three plots of the Trump to Biden ratio from three states that are representative of what most states looked like over the election “count.”  Notice that the Trump to Biden ratio gradually rose until it exceeded an apparent set point, then it decreased gradually into that set point and remained there through the end of the count.

Smith Analysis – “Trump-to-Biden Ratio” in Three States Showing a PID Control Algorithm (1 Set Point).

Now recall the previous section (The Drop and Roll) describing successive batches with impossible repeating percentages in the Georgia presidential race.  Draza Smith’s work perfectly reconciles with Buffington’s independent analysis.  The period at the end of the count where the results are settling into their set points explains the number of repeating percentages awarded to each candidate.  At the end of the count, after the set point had been reached – the function (or algorithm) controlling the ENR data wasn’t really keeping track of any additive process, it was just updating the total number of votes cast and distributing the percentages to each candidate as needed to hold the set point, even if it meant it had to split single votes between candidates.

Most states behaved as the representative plots from Colorado, New Mexico, and Arkansas shown above. But the swing states that all stopped counting in the middle of the night behaved differently. In fact, they behaved exactly as a PID control algorithm would behave if it had its set point changed in the middle of the count.

Smith Analysis – “Trump-to-Biden Ratio” in Three Swing States Showing a PID Control Algorithm (2 Set Points).

The plots for the swing states of Wisconsin, Michigan, and Georgia all exhibit PID controlled behavior.  All settled into a Trump-to-Biden ratio that would have resulted in a Trump victory. Then the swing states “stopped counting.” When they started counting again, they were approaching a new set point – one that would result in the swing states going to Biden.

The Edison Zero

 In the time that elapsed between the change in set points in the swing states, a very strange thing shows up in the ENR data.  All fifty states show a total of zero votes for a period. This event is called the “Edison Zero” in election integrity circles.

Below is a dual plot from the website of analyst Jeff O’Donnell. The yellow line is the total cumulative votes across the entire country and the green line is the Trump-to-Biden ratio. The yellow and blue vertical lines are the points when the states of Florida and Texas were called for Trump respectively.

Plot of the Scytl-Edison Data Showing Removal and Replacement of Entire National “Vote”

In simple terms, after Florida and Texas were called for Trump, it is believed that all the votes in the national Scytl-Edison ENR system were pulled and reset back to zero. It took almost two days to put those same votes back in the system.  Prior to 10 pm, when all the swing states decided to “stop counting,” Trump was in the lead. After the Edison Zero and re-insertion of votes that had supposedly already been counted, Biden lept ahead and stayed there until the end of “counting.”

Several experts in election analysis circles believe that Florida and Texas were originally slated to go to Biden, but when Trump received more support than expected, “Plan B” was deployed. This required seven smaller states to: (1) stop counting, (2) recalculate the algorithms, (3) reboot the ENR system, and (4) retabulate the votes in different percentages to give the election to Biden.

To make it seem normal and expected, “Plan B” was floated out to the public a few days before the election by Bernie Sanders on late night television when he made an exact prediction of which states would stop counting at 10 pm:

Several states exhibited a clear reset in their ratio control after the Edison Zero. Virginia is one of these:

Smith Analysis – “Trump-to-Biden Ratio” in Virginia Showing Reset in Algorithm.

More Proof of an Algorithm Reset

More evidence that the ENR is controlled by an algorithm was discovered by Jeff O’Donnell in some county-level ENR data that was reported by Scytl-Edison on election night in Chester and Philadelphia Counties in Pennsylvania; Fulton, Macomb, and Bibb Counties in Georgia; and Chatham and Surry Counties in North Carolina.

In these seven counties, the ENR data is more detailed. It shows the total ballots cast and breaks those into in-person and mailed ballots. The videos below graphically depict the numbers of ballots in the in-person and mailed categories performing impossible feats of increasing and decreasing how many were in each category from the evening of November 3rd and for several days following. The data is showing an algorithm that is attempting to solve itself but cannot do it with the original constraints:

All the evidence points to the fact that the ENR data Americans see on election night is not a straight report of a real tally, but rather an algorithm. And that algorithm is set up ahead of time with a pre-determined set point of percentages to each candidate that every state is expected to reach by the end of election night. Support for Trump on November 3, 2020, overwhelmed whatever “Plan A” was, necessitating a complete re-boot of the reporting of election results, which required the swing states to all “stop counting” at 10 pm so their set-points could be changed to give Biden a win.

Where does the Plan to Steal an Election Come From?

To steal a national election, it would have to be modeled ahead of time. Estimates would have to be made as to how many votes needed to be fraudulently inserted into certain jurisdictions to swing the election in the desired direction.

Campaign strategists model elections all the time to create campaign strategies for their candidates. Consultants study historic elections down to the county and precinct levels and they estimate how many people in which locations need to be won over to their candidate to give them a win. They plan campaign stops, phone calls, and door knocking accordingly. Then during early voting and on election day, the candidate’s campaign will follow up with the people who were favorable to him or her during campaign season and remind them to go out and vote.

If elections can be modeled honestly, they can also be modeled dishonestly. Just like a candidate, bad actors that plan to steal an election would need to know: (1) how real voters intend to cast their ballots (pre-election polling), (2) they would need to know what the turnout was through early voting and election day (collected by internet-connected electronic pollbooks and distributed by SOSs), and (3) they would need to know how those people were casting their ballots (exit polling). They would also need (4) a censorship regime to crush all discussion of what citizens noticed going wrong during the election.

Turnout data can easily be collected by internet-connected electronic pollbooks – data which is regularly distributed by SOSs across the country. The censorship regime also exists and was recently revealed by the U.S. House Oversight Committee to be headed up by the Department of Homeland Security colluding with election officials, social media giants, and leftist organizations.

Is there anyone out there who is collecting sufficient pre-election and exit polling in every county in the U.S. that would fill in the data gaps left by electronic pollbooks and the DHS’s censorship regime?

Yes, there is.

The aforementioned Edison Research – the entity working with Scytl we just discussed as having a complete monopoly on Election Night Reporting. According to their website, “Edison provides the National Election Pool [ABC News, CBS News, CNN, and NBC News] with pre-election polling, early voter exit polling, Election Day exit polling, and tabulation of the national vote across every county in the United States.” (emphasis added)

They go on to explain that Edison handles “election polling for international clients, most recently Venezuela, Iraq, Ukraine and the Republic of Georgia.”

Venezuela and Ukraine are two countries where widespread election fraud is commonly believed to be occurring. In 2018 a phone call between Victoria Nuland, U.S. State Department Assistant Secretary and Geoffrey Pyatt, U.S. Ambassador to Ukraine, discussing their plan to overthrow a democratically elected president in Ukraine. Interesting that Edison happens to have a presence there.

With the data collected in all fifty states by Edison, it is perfectly possible to make an estimate of how many ballots need to be inserted in specific jurisdictions during early voting to swing an election. Then using the real-time vote count as it comes in on election day, those estimates can be updated and revised in real-time. Since Scytl controls the flow of ENR data nationwide, they can slow it down when adjustments need to be made – for weeks if necessary.

Is there any evidence that elections were being manipulated during early voting? Yes, there is a lot of it. Most notable is the research effort by an election integrity organization, True the Vote, to gather video surveillance and cell phone data which showed “ballot mules” illegally inserting thousands of ballots into drop boxes all over the country during early voting.

There was also the truck driver that reported that he was directed to drive hundreds of thousands of ballots from New York to Pennsylvania in late October of 2020, with a recent postal service report validating his account of what occurred.

Gaslighting About ENR

Hostile “fact-checkers” and election officials have offered two insufficient explanations for why ENR is not solid proof that elections are centrally manipulated and fraudulent.  The first, and most common explanation election officials offer, is that the ENR is third-party data. Thus, with no refutation of the actual data, they merely exclaim: “We don’t know where it comes from! We are not responsible for what is in it!”

This is ridiculous because states pay vendors thousands of dollars every election for software to take their election data and populate their SOS websites with the results. The Election Assistance Commission (EAC) expressly provides that states are responsible for their ENR data and that it needs to be an accurate representation of the results, as it is perceived by the public as the official outcome of the election.

The second explanation is that ENR data are “projections” and subject to change as claimed by a Fox News reporter below when the number of votes that had been processed in the Pennsylvania 2022 Primary inexplicably dropped from 99 percent down to 95 percent on live TV:

The truth is that with the almost universal deployment of electronic pollbooks across the country, the number of votes that have been cast should be known within 1 percent as soon as the polls close in most jurisdictions – a five percent drop in total ballots processed is not believable unless fraud is underway. And if five percent is to be the new norm for ENR accuracy, Fox News had no business calling the Presidential race for Biden in 2020 based on such flimsy projections.

More importantly, the EAC leaves no room for using projections or polling to report the election results to the public. Neither is there any mention of projections or polling in any of the contracts ENR vendors have with states. The reporter in the video above accidentally revealed an incriminating truth. He accidentally told the American public that the data Scytl-Edison feeds the public are not real election results.

Election Night Reporting is a show—political theater—set up ahead of time to fool the American people into believing that they are watching an honest process.  And if the American people don’t wake up to the truth now, they will watch another election cycle of leaders being selected rather than elected.

The claim in the last paragraph is a bold one, but not hard to understand for technical experts who have studied the data and realized there is no other explanation. But the final nail in the coffin for the “fact-checkers” and brainwashed election officials can be found in little Rhode Island.  A state so small that they process the entire state’s election on a single Election Management System (EMS) – a technical term for a central computer where the election data is aggregated and stored.

The figure below shows two graphs created by Draza Smith. The graph on the top is a plot of the Trump and Biden votes as they accumulated on TV screens across America on November 3, 2020. (Notice Rhode Island has a F Curve just like all the swing states.) The graph on the bottom are the votes as they are stored within the computer that houses the official election results.

ENR Data from Rhode Island and Actual Election Data from Election Management System

As anyone can see, these two graphs are not the same even though there is no legitimate reason why they should not be. There are only three explanations for this: (1) the data broadcast on election night is manipulated, (2) the data stored in the central computer is manipulated, or (3) both the election night reporting and official results are manipulated.  Whichever of these is the truth (and experts suspect it is the third), there is no other conclusion that can be reached other than that our elections are not honestly portrayed to the public.

Who is Responsible for Manipulated ENR Data?

Look to Georgia.

Now to shed some light on who is most likely responsible for showing the public at least partially fabricated data on election night and claiming they are real election results.

Recall that election officials brush off massive evidence of fraud in the ENR data. They claim it is simply third-party data and can’t be held responsible for it.  There are at least nine states for which this excuse holds no water: Arkansas, Colorado, Georgia, Illinois, Iowa, New Jersey, South Carolina, Utah, and West Virginia. These states contract directly with Scytl to provide their election night reporting. That means that the numbers that float across the TV screen are provided by Scytl, and Scytl alone, who get it directly from election officials. According to the EAC, this means that the state election officials are responsible for those numbers.

Since we shared Georgia’s data in this article, let’s look at the Georgia Secretary of State’s election results. Click on the election results for any election since 2012, and you will be taken to a Clarity Election website – which is owned by Scytl. Scytl even copyrights Georgia’s ENR data.

Screenshot of Georgia’s ENR Website Showing Scytl Ownership

Earlier in this article, we showed impossible numbers of batches of votes in a row came in at exactly the same percentage for two candidates, which also happened to match the overall percentage. Georgia’s data also shows all the hallmarks of being calculated by a PID control algorithm with an original setpoint that was revised after Florida and Texas were called for Trump.

Georgia’s Secretary of State and Scytl are directly responsible for the impossibilities in the election results in the 2020 General and Runoff elections. In other words, Scytl is easily identified as a major culpable party behind the famous “Fraud Curve” when examining the evidence.

The Rabbit Hole Goes Deeper

Election integrity activists have become aware how impossible it is to overturn a fraudulent election after a result has been announced, even when election fraud can be proven without a doubt and there is video surveillance of the steal taking place. This means that controlling the mechanism by which winners are announced gives unfathomable power over elections.  And it doesn’t matter what evidence comes out afterwards—because if history is our guide—nothing will be done about it.

Scytl doesn’t only have a monopoly on ENR reporting, but they also have their fingers on the official election data in 26 states after acquiring “100% of SOE Software, the leading software provider of election management solutions in the United States.” SOE Software owns a product called “Clarity,” a suite of 8 software modules spread throughout 900 jurisdictions across the United States.  Scytl-Edison’s publication of what looks like bogus ENR data on live television is a terrifying reality. With the ability to control the official election data, there are no limits on what they can do.

Scytl also has an unfortunate connection to Konnech through a man named Luis Nabergoi-Puente who lives in Spain and has been Konnech’s Worldwide Technical Director as well as a program manager at Scytl.

Konnech is an election company who claims to have distributed its software to “thousands of election offices across North America,” but have also been proven to be storing sensitive election worker information in China, as well as giving Chinese nationals full access to their software. (

Konnech’s illegal connections to China were discovered by True the Vote.  After the information reached the Washington D.C. FBI headquarters and after the D.C. office got involved, the FBI did everything possible to shut down the investigation into Konnech and implicate the researchers who discovered the breach.

There is a powerful force behind Scytl, Edison and Clarity. How they ever obtained influence in US elections is not clear.  Their activities in US elections are suspect at best as well. 

Listen to a discussion of Scytl and more at today’s Joe Hoft Show on with Professor David Clements.