The Tale of Joe Biden’s Election Night Spike
In the early morning hours of election night, the release of final vote counts was delayed in some battleground states. There were various reasons for the delays, including a suspicious report of a burst water pipe in an Atlanta polling site, but generally the reason given was the huge amount of mail-in ballots due to the COVID pandemic, which delayed the normal counting process.
In four of these battleground states, Trump was leading Biden before the delay, but Biden’s vote count spiked upwards after the results reporting resumed. When President Trump addressed supporters in the early morning hours of  Nov. 4, he had a significant lead above Biden in four of these states, with two-thirds or more of the votes counted in each case:
  • In Pennsylvania, Trump was up nearly 700,000 votes at 56 percentage points against Biden’s 43 points.
  • In Michigan, Trump was up nearly 295,000 votes at 53 percentage points against Biden’s 45 points.
  • In Georgia, Trump was up by nearly 120,000 votes at 51 percentage points against Biden’s 48 points.
  • In Wisconsin, Trump was up by more than 116,000 votes at 51 percentage points against Biden’s 47 points.
After reporting resumed, Biden’s vote total spiked upwards in all four states, which he went on to win. Had they remained in the Trump column, Trump would have won the election.
In each of the four states, the votes that came in during the early morning hours were heavily slanted toward Biden:
  • In Pennsylvania and Georgia there were about five Biden votes for every two Trump votes.
  • In Michigan there were five Biden votes for every three Trump votes.
  • In Wisconsin there were two Biden votes for every Trump vote.
Is it plausible that the late vote counts would be so skewed toward Biden? Yes. The late counts came mostly from mail-in ballots and from heavily Democratic urban counties that contain the cities of Philadelphia, Pittsburgh, Atlanta, Detroit, and Milwaukee. That a delay and a pro-Biden spike would come from these places is not necessarily unusual on its face. But hold on to your hat, because we’ll come back to this point.
Voter Registration And Turnout, Up Bigly
What is unusual and unexpected is the tremendous over-registration of voters during the 2020 election cycle. Before the election, the conservative legal watchdog organization Judicial Watch reported that more than a tenth of all the counties in the United States had more registered voters than citizens eligible to vote.
“The new study shows 1.8 million excess, or ‘ghost’ voters in 353 counties across 29 states,” Judicial Watch President Tom Fitton said in a release. These counties include Fulton County, Georgia, which includes the city of Atlanta (at a 109 percent registration rate) and the entire state of Michigan (at a 105 percent registration rate).
Even in places where the number of registered voters was below the number of eligible voters, registrations reached their highest level in decades during 2020. Philadelphia County had more than 90 percent of its eligible voters registered, its highest level in 35 years.
Excessively high registration rates are concerning because they are a convenient cover for fraud. “Dirty voting rolls can mean dirty elections,” Fitton said. This is especially so, since most states and local governments used COVID as an excuse to relax voter registration requirements, and to send blank ballots to badly outdated lists. The excessive number of states with registrations above eligible populations “highlights the recklessness of mailing blindly ballots and ballot applications to voter registration lists,” Fitton said.
Indeed, there were several reported instances of dead people voting in Michigan and other places. For instance, the conservative commentator Austin Fletcher used data from the Michigan Voter Information Center to show that a 118-year-old named William Bradley, a 120-year-old named June Aiken, and a 119-year-old named Donna Brydges cast votes during the 2020 election. All of them have been dead for decades.
Biden-supporting organizations have countered that the instances of dead people voting is probably only a small-scale problem. But dead people voting is only one expression of the consequences of over-registration and a flood of absentee ballots. Project Veritas uncovered a significant ballot-harvesting fraud operation in the Minneapolis Somali community, which involved hundreds of unofficial campaign operatives paying Somali residents cash in exchange for requesting and filling out ballots in their names.
Nationwide, citizen voter registration reached 72.7 percent in 2020, and voting turnout reached 66.8 percent—these were at their highest levels in 28 years, according to Census Bureau data.
Past elections with voter registration and turnout rates as high or higher than 2020 have tended to feature an enormously charismatic presidential candidate who sweeps into the Oval Office in a convincing popular vote victory. The two most recent election years with higher registrations and turnout than 2020 were 1992, when Bill Clinton trounced George H. W. Bush by nearly six million votes, and 1984, when Ronald Reagan swept the table against Walter Mondale, winning nearly 17 million more votes and 49 states.
Now consider: Joe Biden hardly campaigned at all, sitting in his basement while Trump filled arenas. Biden also committed one embarassing senile gaffe after another, such as telling the host of a syndicated radio show that blacks who don’t vote for him “ain’t black.” You’re telling me that Joe Biden, who had a lackluster performance in his own party’s primaries, really lit a fire of enthusiasm in the hearts of voters, to the greatest extent since Clinton and Reagan? For some reason the numbers say he did.
Joe Biden is not even Barack Obama, and yet he exceeded his former boss’s turnout numbers handily. Biden got more people to vote for him in Milwaukee County than Obama did at the height of his popularity in 2008. This is all the more strange, since Milwaukee County has lost tens of thousands of residents in the years between 2008 and 2020, and yet there were more of them on hand to pull the lever for old Joe, as The Federalist’s Evita Duffy noted. Has anybody checked Milwaukee’s graveyards?
To be sure, Biden supporters can try to explain the high turnout number in the same way the Pew Research Center did: it was “fueled in part by the bitter fight between incumbent President Donald Trump and challenger Joe Biden.” There is a legitimate point here: tempers boiled over into civil strife in 2020 to the greatest degree since perhaps the 1968 election, which was also a year where registrations and turnout exceeded 2020’s. In that year, Richard Nixon barely eked out a win over Hubert Humphrey (though he probably would have won more convincingly if George Wallace wasn’t in the race).
Despite this, the weight of the evidence on the 2020 registration and turnout numbers and the potential for widespread ballot fraud, whether physical or electronic, is too great to ignore. Especially when you combine this with all of the other historical voting patterns that were overturned during the 2020 election, which Patrick Basham explores in his article in this issue of Chronicles.
Dr. Frank’s ‘Fantastical’ Algorithm Thesis
The mathematician and physicist Dr. Douglas G. Frank has a theory and a statistical model regarding those high registration numbers that is worth exploring. One of the most interesting guest speakers at Mike Lindell’s Sioux Falls Cyber Symposium in August, Frank said that he’s found that the voter registration rolls of thousands of counties across the United States have a strong statistical correlation with U.S. Census data.
Frank posited that in 2020 state voter registration rolls were inflated ahead of time by people who intended to manipulate the election results. This extra coterie of “ghost” registered voters provided a sort of buffer that could be electronically converted into cast ballots on a case-by-case basis.
Amortizing 2010 Census data to the year 2020 based on the Census Bureau’s own yearly formula and mortality rate creates an algorithmic line that correlates at 0.99 with the voter registration rolls, Frank said. Correlation coefficiencies are usually around 0.7 or 0.8 in any statistical dataset involving people. “Anytime you have a correlation coefficient near 1…that’s not natural. Something is going on that’s making the two agree with each other,” Frank said.
That was my first clue that they were using the census to inflate the registration rolls. Now if you think about that, it totally makes sense. Let’s say you want to add a bunch of people to a county—you want to add a bunch of phony voters. Are you just going to add a million 60-year-olds? No, that would stick out like a sore thumb. You have to add a certain number of each age. Well, how many do you add, what do you compare it to? The Census, and you use the best census available, the 2010 Census. You shift it 10 years, you attenuate it to the size of the county, and you fill up to there.
Frank believes that in 2020 this algorithmic model was duplicated in the voter registration roles in every county across the country, and is recognizable in the shape of the chart. “You’ll notice it’s going to take on the shape of the census. And a couple of the distinguishing features are these two peaks on the side and the peak [at the top],” Frank said. “It’s like a fingerprint in every county. I’ve done thousands of counties now across the country, and it’s in every county.”
above: Dr. Douglas G. Frank speaks on day 1 of Mike Lindell’s Cyber Symposium in Sioux Falls, S. D. on August 10, 2021. (
One of the left-wing “fact-checking” organizations, Politifact, attempted to debunk Frank’s theory in an article entitled, “No, Michigan didn’t use an algorithm to manipulate 2020 election results.” Politifact’s critique, however, is unconvincing, amounting to a complaint that Frank’s theory is hard to understand, followed by quotes from establishment experts who call him “fantastical” and a “conspiracy theorist.” 
For what it’s worth, I think Dr. Frank may be onto something. For one thing, his theory could explain the unusually high voter registration numbers, as well as account for all the historical precedents that were overturned in 2020, which Patrick Basham details. It may also explain the spike that Biden got during the lull in early morning reporting on Nov. 4. Were the modems in the voting machines turned on to give Biden an injection of “ghost” voters to push him over the edge?
I think of a panicked tweet that Minnesota Attorney General Keith Ellison sent on election night, begging Minnesota’s Democrat voters to get to the polls because “we don’t have all the votes we need quite yet.” This in a state that Biden apparently won convincingly. Perhaps Ellison needed a boost from the ghosts, too.
While I am not totally certain how a massive election fraud on the scale that Dr. Frank and MyPillow CEO Mike Lindell contemplate happened, I do think there were all the signs of a conspiracy to ensure Donald Trump did not win a second term, even well beyond what I can discuss in this limited space. Let’s not be so credulous as to adopt wild QAnon theories; at the same time, it’s not paranoia to believe the left really is out to get America.