Wednesday, November 27, 2024

Election 2024 was marked by errors across the map, with vote-tabulation snafus being a common problem; no wonder voters have doubts about who won

Programming errors with vote-tabulation machines were present in 2024 (Getty)
 

We have no reports that Kamala Harris has asked for a hand recount in the 2024 presidential race, despite a growing mound of evidence that the election was maliciously hacked in order to favor Donald Trump at most every turn. (Note: Reports are starting to appear that the Harris csmpaign is setting aside funds for a recount.) Will Harris join the fight to show that she was the likely winner in a voting process that has been described as "willfully compromised?" We hope to know shortly, but if she refuses to take part, it will be a terrible "optics" for her and the Democratic Party -- the kind of voter betrayal and lack of fight that might take decades to wash away.

In a spurt of good news, at least one Democrat has decided that recounts matter, especially in a close election that shows smelly signs of being hacked. Election-integrity expert Stephen Spoonamore explains in an update at his Substack page -- ""And When You are a Hammer." Under the headline "A Hand Recount in MI Race, Stats, County Flips, and very disturbing tips about signature gathering/straw voters.
Still a hammer. Still hitting nails
,"Spoonamore writes:

Stephen Spoonamore
Nov 26, 2024
Tuesday

I have formally withdrawn a theory of bullet ballots in my Duty to Warn Letter to VP Kamala Harris. I was wrong on my estimate of the likely number of bullet ballots; it was a good-faith attempt to understand the data 48 hours after the election. I missed the mark on that specific theory. I stand by the remainder of the letter, including my certainty that there are tabulation manipulations and access to ePollBook check offs. The statistical improbabilities and use of bomb threats against our election all indicate underlying hacking. The conclusion of my letter stands: Demand Hand Recounts, which finally one candidate in MI is doing.

Unfortunately, as we have reviewed more and more tips, and public reports, my tabulation concerns are appearing in both swing-state and non-swing state counties. More below.

In my view, and in the view of a growing list of IT and statistics professionals, the present results of this election were not produced by voters. It has been manipulated. Hand recounts are absolutely necessary to restore confidence.

Mr. Jim Haadsma is the first candidate as far as I know to demand a hand recount for his race -- a race the tabulators have counted three times, with drastically different results.

Summary: In Calhoun County MI, (Traverse City area) a “programming error” was discovered with the high-speed tabulators which excluded 1,000s of votes. Incumbent Jim Haadsma was initially reported losing to a GOP challenger by 1,482 votes. The erroneous tabulation excluded substantial portions of absentee votes. The ballots run twice since then have shifted all the races. The margins shifted to 58 votes in the first run after the error was resolved and 61 votes in a second run. At the link below is a Detroit Free Press report about the latest on the Haadsma race:

https://www.freep.com/story/news/politics/elections/2024/11/25/lawmaker-jim-haadsma-requests-recount-battle-creek-state-house-election-steve-frisbie/76569420007/

Spoonamore says programming errors have been reported in other jurisdictions, and statistics increasingly indicate the results were manipulated by a machine, not the result of human actions. He writes:

In multiple counties, including here in Centre County PA, a “programming error” in the tabulation has specifically excluded some ballots. After the vendor reprograms the tabulator and ballots are run, the races have all swung substantially toward Democratic candidates.  How many places have this “programming error”? Were these tabulation errors that swung votes AWAY from Democrats intentionally programmed into the systems? If so, how was that done and who led that effort?

THE GROWING LIST OF STATS INDICATING PROGRAMMED RESULTS

The first thing to understand is that current results claim Harris lost to Trump in a close election 1.5%.   49.9 to 48.4.  If true, it would be the 16th closest election in history and should show all the markings of a close election in the underlying statistics. It does not. As posted today, Trump did not win a majority, yet won 7 out of 7 swing states? And all just beyond mandatory recount levels?  I don’t believe it. Nor do professional stats people.  And as they dig into the numbers they have found a growing list of absurd outcomes, all of which indicate these results were generated by a machine, not humans voting.

Among the  “This is not possible” tips coming in a number of them are pointing to the county-level flips. In 2024 there are 88 counties flipped vs 2020. That is a pretty normal number. What is not normal, every flip went from Biden to Trump. None flipped the other way.  By comparison in 2020 there were 82 counties that flipped.   19 Flipping Red to Blue, 63 from Blue to Red. There are detailed discussions about the subject going on at Reddit

https://www.reddit.com/r/somethingiswrong2024/comments/1gzxmmp/comment/lyzsbz9/?context=3&share_id=PH4b5tmZZRuaMUPmAIqVV&utm_content=1&utm_medium=ios_app&utm_name=ioscss&utm_source=share&utm_term=1

I note that one portion of this thread looks at county-level flips over the past 20 presidential elections.  All of them have flips each way, except this one.  

Hand recounts will resolve this. Ms. Harris. Be like Mr. Haadsma. Demand hand recounts.

We also have multiple quantitative people and voting geeks alerting us to another non-human pattern, which some of them claim is appearing in every county of several swing states. Noted here is Arizona,  where they have sent complete data.

In every county in Arizona 2024, HARRIS' vote share as a percentage never exceeds that of Ruben Gallego, the Democratic U.S. Senate candidate, while TRUMP always exceeds the percentage vote share of Republican Senate candidate Kari Lake. AND the gap between these two differences is nearly identical in every county of every type. It is the sort of output a machine produces, not humans. None of the tipsters nor Snus discussing this on Reddit sites believes this result is human created, and no one appears to have an example of this happening in any prior election. I will post MI and WI bar graphs and comps when they are generated.  

Next to AZ 2024 are the same calculations from AZ 2020, which has nominal human-voter results.

It appears to have the same machine-like output, indicating a force balance appears to be in WI and MI as well. I am waiting for bar graphs of those states.

Again, Hand Recounts will clear up what is really on the ballots.

TIPS UPDATE: ELON MUSK'S TEAMS DID MORE THAN DUPLICATE  ePOLLBOOK REBUILDS?

Three tipsters, who identify themselves as working in the signature-gathering profession, have provided overlapping concerns - here is a composite: 

All three believe the plebiscite petition Elon Musk ran ahead of Nov 5 could be used to hack the vote in swing states. One thinks it’s the only reason it was done.

This year, campaigns paid around $3 per signature for signature gathering. Musk announced his plebiscite (a petition that has no legal effect) would pay $47 each, and offered payments in some places to both the signature gatherers AND to the signers. It was run in all swing states and was unusual, as several swing states do not allow citizens' initiatives and forbid collecting signatures electronically, for the very real concern that collecting people’s complete voting data and signatures could be used for identity theft, straw voting, or other forms of hacking. Musk got around this by claiming he was holding a lottery in some places, or non-binding plebiscite in others. It is unclear what laws govern such a project.

No comments: