The charts below are derived from The New York Times' real-time election feeds (e.g., here). They show "DNA-level" evidence of vote fraud that was systematically used to overcome massive Trump leads with "vote flips" to Biden.
The twin charts below depict the shifts in votes starting on election day. The X-axis is the date/time and the Y-axis represents the change in votes (positive values denote shifts for Trump, negative values represent shifts for Biden, in hundreds).
Notice the similarities in PA and GA? How the right sides of the graph show virtually no movement for Trump; and very predictable vote movements to Biden. How predictable?
You have to see the data to really understand the magnitude of the scam.
Below are excerpts of spreadsheets that show what was happening on the right side of each chart. Vote flips in the same-sized bundles (6,000 in PA and 4,800 in GA) were injected into the system to overcome Trump's lead in both states. You can click either image above to see all of the data.
The highlighted cells show where the vote counts -- stunningly obvious in retrospect -- were manipulated to benefit Biden.
Note the vote flips, represented by the highlighted cells, that occurred in both PA and GA. In PA, late vote flips in bundles of around 6,000 were clear anomalies to slowly overcome Trump's lead. In GA, the bundles were in 4,800 vote swaps.
Again, these are just excerpts. You can see the workbooks for yourself here: just click for Pennsylvania and Georgia.
Scroll down until you start hitting the highlighted cells.
Sorry, Democrats: this is what we call DNA-level statistical proof of fraud.
And there's a lot more where this came from. These are just the excerpts.
p.s., can someone who knows Sidney Powell or Joe DiGenova get this info to them?
Hat tip: BadBlue Uncensored News.
Holy crap. Nice work.
ReplyDeleteThe Kraken has been released.
ReplyDeleteLove it!
DeleteI don’t know if you mentioned this, I went straight to the data, but they were subtracting votes from Trump in real time also!
DeleteThat may only be its shoulders
DeleteI noticed the subtraction too
DeleteSidney Powell has received it and posted it on twitter.well done.
DeleteThis shows nothing mailed in votes for Biden. We knew this would happen. Get some real evidence.
DeleteIt definitely shows something dummy.
Delete🤣
DeleteLin Wood is a G.
ReplyDeletea straight up G
DeleteOG!!!!!!!
DeleteIt's is a blessing from above that this proof has surfaced. They, should dismantle these dominion voting machines and go back to old school and have observers and cameras everywhere. This is when spying is necessary. This company should be fined and jailed for a federal voter fraud scam along with the officials that allowed this in each state.
DeleteWord
DeleteThere is a contacts form at the bottom of the following webpage, where you can submit data to her.
ReplyDeletehttps://ldfftar.org/
How is this read? What does any of it really mean?
ReplyDeleteRead the article where it explains the sheet. It’s easy. But the minus means votes stolen from Trump basically. Time stamped and all.
DeleteIt suggest that votes were being flipped from Trump to Biden. The highlighted cells at the end are using clusters of 6k for PA and 4.8k in GA. Clusters are really common in data science, as it makes trends and anomalies in data more clear. The main column that is highlighted is the NetChange column, which is tracking the changes made to votes within that cluster. So if we call the NetChange value n, positive n in this spreadsheet would suggest a positive net change for Trump, while a negative n results in a positive net change for Biden. Meaning those highlighted cells in the NetChange column of -n thousand were votes for Trump that got changed to Biden.
DeleteThat is if I understand what the software itself is doing, which I don't actually yet have info too. It's possible it was just adding votes that were never real, multiplying votes, etc. Regardless this spread is comparing the net change between PA and GA. How it was done would require the source code for anyone to be sure.
Donate to Attorney Linn ! Why ? FREEDOM VS COMMUNIST! He’s saving us from Hell ! I donated again and I’m happy it went to FREEDOM! God bless Attorney Linn ans May God watch over you and give you strength to fight the evil empire! 🇺🇸
DeleteOur Constitution our liberty our freedom is at stake.
DeleteIt suggests voter being flipped? Couldn't it also just simply be that the late counted votes were mail-in votes? Votes, that for obvious reasons, came in from democratic voters instead of republican voters? Additionally, late counts came from the largest counties, which, once again, are primarily democratic. Please explain why this very simple reason does not equally explain the data?
DeleteOh yeah! release the kraken! we will win!
ReplyDeleteWow...!! Excellent Work!!! ��
ReplyDeletedo maricopa, az next!
ReplyDeleteYes!! Something very fishy in Maricopa County going on!
DeleteThe Kraken!
ReplyDeleteLet's goooo!!
ReplyDeleteThe president told his base for months to only vote at the polls. Biden recommended mail in voting to stay safe. This looks like how any reasonable person would expect these groups to behave.
ReplyDeleteCheck the bottom of the charts (marked red). They are taking the votes from Trump, and giving them to Biden.
DeleteI'd agree with you but for the fact that in terms of statiscal relevancy, the President received essentially no votes, that is then statistically anomalous, it should not look that slanted if it was in fact organic
DeleteUnfortunatel you're being lied to by MSM/Social media - Biden got crushed and the evidence is irrefutable.
DeleteDid you really think Biden won? He & his son laundered millions through China & Ukraine. The deposits into Hunter's bank account are public knowledge, but MSM/Social Media suppresses all the corruption & lies through their teeth about Trump & his supporters.
Pray for this guy. Hes a brainwashed sheep. Biden is considered a felon in Ukraine btw
Delete"Vote flips in the same-sized bundles (6,000 in PA and 4,800 in GA) were injected into the system"
ReplyDeleteThe "same-size bundles" would be EXTREMELY suspicious if they were real, but I no longer believe they are. I believe they are an artifact of the analytical process: specifically, the unavoidable result of multiplying by a number, vote_shares, with 3 digit precision. The bundles are all 1/1,000 the total vote count at any given time. They're a mirage.
To nail the Dems as the frauds they are, we need to focus elsewhere. For more info see the PA post: my comments and RW's comment at bottom.
Best regards, Steve
Thank you Good Americans for the real evidence and explanations of the horrid unfair voting system machines Now how are we to get MSM to tell the public? And when do we switch to BlockChain technology? And no mail ins ever
Deletehttps://www.smartmatic.com/us/smartmatic-fact-checked/
DeleteFact-checked by Smartmatic, hmm. Whose CEO just got picked for "President-elect" Biden(can't say that with a straight face)Cabinet. Nice reward, eh?
DeleteKeep fighting~! This AIN'T over!!
ReplyDeleteIf I'm reading this correctly, the second/third columns are the *total* number of votes for Trump/Biden, respectively. The real evidence, to me, is not the "consistent change of ~6000 votes", but rather, the times where Trump's vote total actually goes down, indicating the software was *removing* votes from Trump. That simply shouldn't happen, right? The "6000 vote swing" indicates how bad it was. Once we know it's ~6000 votes, we can see how often it happened.
ReplyDeleteThats so true! We need to keep an eye out for 6000, everywhere. I Really hope they manage to get a solid case Before the middle of december
DeleteThis has MADE my day!
ReplyDelete1. I'm not getting the GA link to work. The link might be broken. Please check.
ReplyDelete2. Do you have the same data for the other close swing states: NV, AZ, MI, WI?
3. You need to document where you got the NYTIMES data with a link so others can duplicate your analysis and cite to original sources.
Thanks.
I hope you're presenting this evidence to judges and that they'll see Trump won Georgia and Pennsylvania and Michigan and Wisconsin. We need the FBI and U.S Marshalls to haul these fraudsters out and put them in jail.
ReplyDeletePrayers🙏 for our Good whistle blowers that reveal all the FRAUD illegal voting machines and illegal mail in ballots and illegal machine tampering It is illegal to change data change votes Where are the cuffs?
DeleteThis is awesome news we all knew that they were cheating I'm so glad you guys are on top of your jobs keep up the great work thanks
ReplyDeleteI have three questions for the genius who made this:
ReplyDelete1. Why is this not explained by the difference in partisan breakdown for mail-in voting?
2. Why is it "4000 and 6000 vote bundles" when none is quite the same and why is the size of each precinct not sufficient explanation for it? (You say "it's obvious" but it doesn't seem obvious to me. Maybe I'm the dumb one)
3. What do you make of the manual Georgia recount showing similar results for all county bar that error in a mostly republican county (Floyd) made by probably republican voting clerks? Currently used Dominion machines in Georgia do not have shading ability, so that doesn't explain it, and "the signatures weren't verified the way we want them to be" still does nothing to explain the lack of any major discrepancies between hand-counted and machine-counted ballots—it just moves the goalposts
These are good questions that i was also wondering.
DeleteYou are so innocent, when they flip the vote for Biden they thrown away/destroy Trump's Ballots already. Where other place they print multiple hundred thousand ballots for Biden only to put in the clerks files them with the regular ballots vote. That's why Trump had no ballots after 4am in the morning until Biden had more vote than Trump then they stop. Unbelievable cheat.
DeleteI noticed the huge round whole even numbers of votes for Biden Which I said were impossible I noticed the no Trump ballots getting processed times too Trump had a landslide coming in but they illegally did everything they could to increase the numbers for Biden🤮 Where are the cuffs?
DeleteBIDEN ABSENTEE VOTE LEAD IN EACH MIDWEST STATE
DeleteLA - 15.9%
OH - 16.3%
MI-37.9%
PA - 57.7%
IL- 24.7%
IN-.13.7%
Now..... Looks suspicious to me
And that right there tells us something is wrong in PA and MI
DeleteDemonRats on mask-wearing: "Follow the science!" DemonRats on DNA-level voter-fraud evidence:"Not my science! This data is racist!" Etc.
ReplyDeleteAnyway to cross reference the reported data to the counties that found votes to see if there is any matching of switching votes?
ReplyDeleteGood question
DeleteWhat person do you suggest was actively monitoring this and executing the inputs? Was this one person playing in all states or did each state have to have their own person on site?
ReplyDeleteIf this is the hypothesis, wouldn't a manual recount like in Georgia identify it?
ReplyDeletenot if they ran the same ballots through the same system without any checking, basically just duplicating what they did the first time around. Thus why the Republicans are frustrated ballots weren't scrutinized closely for fraud and there was no audit of the computer system against manual counting.
DeleteThey have eyewitness testimony of this in Detroit. They were refeeding batches of Biden ballots into the machine scanner. I watched the video testimony.
DeleteWhat time were the mail in ballots counted?
ReplyDeleteTHEY SAY HE MAKES EVERYTHING A REALITY TV SHOW... I CAN'T WAIT TO SEE THE LOOK ON CNN'S FACE WHEN THEY REALIZE THEY MISSED HIS VERSION OF "UNDERCOVER BOSS" LET'S MAKE SURE THAT OUR CHILDREN ARE PLEDGING ALLEGIANCE AT HOME TOO!
ReplyDeleteMade me laugh, but I agree can't wait to see their faces.
DeleteThank you for reporting this!
ReplyDeleteCan you explain why Biden also loses ~608,000 vote?
ReplyDeleteDid anyone notice that it is done in increments of 300 votes per electoral college vote? That is - Pennsylvania has 20 electoral college votes, so the vote change increments are in rough groups of 6000 (300 x 20). Georgia has 16 electoral votes - so the changes are in 4800 increments roughly. Now, the curios thing would be to look at other states and see if the pattern persists there, with the same ratio (ie, Nevada with 6 votes would have increment changes of 1800. Wisconsin I believe has 11 votes, and so would have increment changes of 3300). Interesting to check if anyone has the data.
ReplyDeleteWI has 10.
DeleteThis strongly suggests that the good guys were onto them the whole time and hacked their code.
DeleteA competent programmer would have hidden the alterations without allowing the Trump lead over Biden to grow like it did, requiring counting to stop for more fake ballots to be added.
This, in turn, suggests that the good guys were onto them the whole time.
This was one massive sting operation.
Now we know why John Durham was so slow to act.
What? Is the result ratioed because of reporting method and electoral college number or was the fraud midstream manipulation through a back door and they ratioed it to make sure Biden won the electoral college vote in that State. Where is data from non dominion states ?
DeleteCareful, my friends, there's good stuff here but I am worried about the part that is not airtight, the same-sized bundles.
ReplyDeleteLefties will certainly focus on the weakest link, then claim the fraud argument is "discredited" in its entirety.
The same-sized bundles are the weakest part of this argument. I no longer believe they are real. I believe it invites trouble to keep talking about them, absent further study.
Best regards, Steve
Unless their code was hacked by the good guys for the purpose of exposing them.
DeleteNotice they even made it obvious:
“increments of 300 votes per electoral college vote”.
This is hacker humor!!!
They are mocking them.
Beast but this might just be because Edison is not reporting actual numbers. Where does this show that a vote was added midstream that did not correspond to an actual physical fed ballot. If the latter it is just old fashioned ballot box 13 fraud, right?
DeleteContact info for Sydney Powell, defendingtherepublic.org
ReplyDeletehttps://www.sidneypowell.com/contact
Anonymous 6:52 PM:
ReplyDeleteYes, I did notice, and there's a very good reason for that. Unfortunately it's not a reason that helps us. It's connected to a fundamental systemic problem in Doug's analysis. It's the same systemic error that created the MIRAGE of same-sized bundles.
PLEASE let's not rush into making a weak case. Let's prune the weak stuff and focus on the strong stuff. The same-sized bundles are a mirage.
Best regards, Steve
fyi, steve's a master troll. ignore and use the channel for comments
ReplyDeleteWas this manipulation of the ballots or manipulation of the tabulation of the ballots?
ReplyDeleteNote that the hand recount in Georgia is reportedly "spot on" with Biden winning. If it was just a computer program messing with the ballot count the hand count would not be in agreement.
a hand count is not really a hand count
DeleteI believe that was the reason for the overwhelming push for unprecedented voting by mail. That process makes it very easy to forge ballots and evade audits and recounts, since you cant verify the validity of the ballots counted
DeleteData like this besides being evidentiary of fraud should give support to not certify the election results, i.e. this evening Wayne County MI deadlocked in trying to certify results meaning not certified. But with the total corruption of PA other alternatives may be necessary.
ReplyDeleteSorry to say, but these events are based on only hypothetical. In the NYT dataset there is only one decimal accuracy for the parties percentage (ex: 51,5%). So counting millions of votes using this inaccurate percentages, sure will result in some anomalies. And I haven't mentioned the rounding errors...
ReplyDeleteHere is a correct analysis of the situation, mentioning rounding errors also on the red elephants:
https://trp.red/election/electionanalysis.php?state=pennsylvania
Nevertheless, this dataset is still very disturbing with it's huge jumps. But only if it's anywhere near to valid and relevant...
Thank God for you ! ✝️
ReplyDeleteIn Pennsylvania, look at lines 46, 49 and 54. Line 46 Trump's total votes go down 42,325. But Biden's go down 196,432. Why? In line 49, Trump's total votes go down 127,916, Biden's go up 12,401. Line 54, Trump's total's go down 145,180, Biden's go down 416,518. Why would anyone's vote totals DECREASE like this. I can't agree with the author that this analysis demonstrates fraud when the data looks like this.
ReplyDeleteOne possible explanation that would have to be eliminated is whether, with a backlog of ballots, they were processing them in fixed batches, reporting the results for the batch, and then moving on to another batch.
ReplyDeleteI was at first thinking the same. Note however that the constant 6000 / 4800 number is the **difference** between trump and biden votes. Therefore it is not a batch size.
DeleteThe charts say it all.
ReplyDeleteIt would be an interesting exercise to evaluate these consistent numbers and their consistent multiples based on how many ballots fit into the trays used in those jurisdictions.
ReplyDeleteThe consistency suggests to me that the bundles of ~6,000 in PA and ~4,800 in GA let us hypothesize about how many ballots it takes to completely fill a tray that's been prepared for processing.
That wouldn't prove fraud - but it would suggest that there were full trays of Biden votes prepared for processing when trays for Trump were not similarly full at the same time.
You can try Newsmax 1-800-485-4350. I couldn't find a number for OAN.
ReplyDeleteOr the FBI 1-800-225-5324
Try contacting Newsmax 800-485-4350
ReplyDeleteI couldn't find a number for OAN.
Also (if you trust them), the FBI
1-800-225-5324
These are statewide numbers for Pennsylvania? It would be very telling if data was available for Philadelphia and Allegheny and maybe Luzerne as opposed to state wide. If tranches of ballots were being force fed to swing the lead to Biden it is unlikely to have occurred in more than one place, Philadelphia, with some possibility of the other two counties. If the bundles are coming as a result of totals entered from all over the state it's hard to see it being deliberate fraud at the election board level. It would be more likely the fraud occurred through the software. Or am I missing something?
ReplyDeleteAntrim County Michigan. 6000 switch. But they caught it. Glitch? Antrim County is like 20000 people total. They knew it was for Trump but ended up with Biden so they had a recount and caught this. Is this what happened there???? If so proves glitch and perhaps fraud.
DeleteI have yet to see an explanation for how that specific flip of votes happened. They said user error. I want to know the exact user error that results in something like that.
DeleteShe found you. https://twitter.com/SidneyPowell1/status/1328839465255702528?s=19
ReplyDeleteCorrect me if I misunderstand something, but isn't it true that large counties typically provide their results much later than the smaller counties (since it takes longer to count more votes). Isn't it also true that large counties (ie: Philadelphia and Atlanta) are far more likely to skew democratic. So if the early returns were a composite of many counties, some more republican and some more democratic, and the later returns were only from large counties, are the first plots really that surprising? Furthermore, if all the later votes are coming from the same county, or very similar counties, would it really be surprising if the breakdown (ie: Biden +6000 or +4800) was similar between tranches?
ReplyDeleteThe comment about the percentages and rounding errors is not clear to me. The data feed looks to have exact vote number and a percentage displayed.
ReplyDeleteNTMA
ReplyDeleteWell, what does this look like for other states? let's have more to compare to please.
ReplyDeleteTHIS is the Kraken? Are you kidding me right now? Is this a joke? Statistical anomalies indicate fraud may be present. You've still gotta find the actual fraud. This gets us nowhere. It is utterly useless.
ReplyDeleteSteve is right.
ReplyDeleteThe packets of 6,000 equal a tenth of a percent of the vote share for Biden or Trump. So a move from 48% for Biden to 48.1% would be about 6,000 votes, and a move at the same time negatively for Trump would be about another 6,000 votes.
Doug, it would have been better to post more about the actual data, not your version of it. I wasted a couple hours analyzing the data believing that a count was a count, and I wrote an email to a contact of mine at the Wall Street Journal, which I had to retract with a "NEVER MIND"!
Running sums that spontaneously decrease are an issue. How does that work? It's not a checking account.
ReplyDelete1+1+1+1+1-4+1-2-1+1+1+1+1+1-555+1+1+1...
We need to move faster, evidence is being destroyed as we speak.
ReplyDeleteposting in an ebin thread
ReplyDeleteKryptonite! Check out VA, too. Corrupt take over by democrats...
ReplyDeleteIsn't the key to this that Biden is getting votes when Trump seemingly isn't gaining any after a certain period of time? For ex, in GA Biden is getting 4k votes when Trump is only getting 50 or less. Is this a wrong interpretation of this? What am I missing here?
ReplyDeleteWith the Dominion Hardware/Software on hand, and the engineers/whistleblowers that designed it, they are going to be able to show exactly HOW the manipulation was done, match it up with multiple different sources of the live vote feed count, make graphs like as shown above, and easily prove this in court.
ReplyDeleteThe owners manual for the Dominion machines actually shows you exactly how to manipulate the results as you see above, giving detailed instruction on how to do so. There are multiple different methods as well, votes even having the ability to be cast in "weights".
Furthermore, evidence of the votes being stripped away from Pres Trump and has actually be captured on live feed CNN broadcast. And this is not the first time. It's also been captured in other elections. Hammer/Scorecard is the name of the program. Apparently the dummies that programed it didn't think enough to realize that a vote switch/manipulation could be caught on live feed of the vote count.
The suggestions that this is some type of artifact of the process is absurd. Several different highly qualified groups have analyzed all of the live feed data, of which there are several different sources, and ALL show the same ridiculous statistical impossibilities. This will be a slam dunk in an unbiased court once all the data is gathered and presented. The graph above is just the tip of the iceberg.
If not for the seriousness of the situation, the sheer incompetence of the whole crew behind this and the blatant cheating would be laughable.
The cheating or fraud is the work of a psychopath. It's like a bad James Bond movie. Right?
DeleteLink for Georgia just points back to this page.
ReplyDeleteMichael's correct. You have a bad link to the Georgia data.
ReplyDeletePenn data opens.
you all are fucking suckers
ReplyDeleteThis is as distracting as the See Vee hoax - both make me sad to see so many wannabe Americans stuck on stoopid.
ReplyDeleteRefute this if you can, but I bet you can't.
JUST IN FROM A FRIEND OF MINE IN VIRGINIA:
This is a forward from a High School friend who’s married to a lawyer and works for a law firm. What do you think????
info from some of the lawyers from my group:
"Ok - in a nutshell:
This is going to the Supreme Court where they will rule that the election is invalid due to fraud or mistakes on a countrywide scale. It will go one of two ways: either they will rule that all the unconstitutional mail-in ballots will be removed and the states ordered to recount without them or they will simply rule the election is invalid due to mass voter fraud and at that point, it will be sent to the congress and senate for a vote. This is where it gets good. The house/congress votes on who the President will be. It has nothing to do with what party has power. Every state gets one vote and 30 States are held by Republicans and 19 by Democrats. They have to vote down party lines - they have no choice due to the 12th Amendment of the Constitution - and the Senate votes for the Vice-President where a similar event will take place. This is The law. This is why the Democrats are so mad at Nancy Pelosi. This will all happen in January. The only way President Trump won’t be President is if he concedes the election and that will never happen. So stop watching the fake news and don’t let your heart be troubled and live your life knowing this will all work out.
President Trump will remain President.
I have researched all of this and it is Fact!
Another fun fact: they call Gore the President Elect for 30 days in 2000
until the courts ruled against him and declared Bush the winner.
And two people that were part of that decision were none other then the new Supreme Court Justices, Brett Kavanaugh and Amy Coney Barrett. Why do you think the Democrats tried so hard to keep them from being confirmed?"
"THERE IS A REMEDY AT LAW FOR EVERY PROBLEM; and any inference with the lawful use of a remedy is permission to create other remedies at law."
per Judge Anna.
Forest Queen, Re:Judge Anna's message to you
DeleteI thought the Gore vs Bush decision was up to the Supreme Court in that case of the hanging chads fiasco. How did Kavanaugh & Barrett have any hand in that decision? Thanks if you or anyone can reply.
Actually, this has nothing whatsoever to do with the Senate. It will be presented to the House who will be allowed one vote each per State. The founders were genius. Whomever presents this argument to students, Communists, and the left in general already know the outcome. Trump wins overwhelmingly, and possible names attached to indictments begin floating around. The less guilty will tell all they know to avoid prison time (they may already be turning on each other like starving hyenas, we'll see. Ideslly, the Supreme Court tosses the election and declares Trump as POTUS. Otherwise, we keep the courts bogged down and as long as Trump refuses to give a concession speech, on 01/20, the matter is turned over to the Houses' single vote, and Trump wins that scenario as well. After I began but to see the extent of the fraud, I was willing to fight for my President (disabilities and all), Putin-style, but with the promise that after he received his initial 4 years,maybe 8, he'd step down having saved our lives as Giuliani helped do after 09/11. The Left spent SO much of our Money investigating fraud, collusion with foreign powers, attacking his nominations and filing false impeachment claims, that the so-called "Party of, for, and by the people" means absolutely nothing to said people. We- the right, are for the freedoms, for capitalist businesses,for smaller government,less taxation, etc. We are what they have pretended to be since the Civil War, how they turned it around was sinister, racist and abominable. I hope the Socialist Left is torn apart and forced to begin again. Pure Evil, my fellow Americans, Pelosi wants to shred the Constitution! I think she may need it to keep warm at night, but I hope it gives her little comfort. God bless all of you working tirelessly to right this egregious wrong, and above all, Thank You Sidney Powell, for everything. How long did you sit on the 'Kraken' line?? And how do we donate directly to Powell?
DeleteTo check and see for yourself that this is bs:
ReplyDeleteDownload the actual json data (or use an online viewer), scroll down to the time series, do you see actual vote counts for trump/biden? No, only the total of votes and the fraction for each candidate of that total. The split counts in the spreadsheet therefore are not original, but were generated by multiplying this fraction with the total count. The smallest change in that fraction of only three digits is 0.001, this multiplied with the end total gives you the smallest change in vote count per candidate that you can "detect" from this data (..try it, guess what number you get..).
There are actually two series of data in each .json.
DeleteThe first series is actual vote counts. And the problem with these is they report too much vote coming in and for Pennsylvania Biden receives over 4MM votes. It should be explained why surplus vote is reported in the vote counts and how they determined which portion of it was discarded for reporting purposes. The actual vote counts do not match the results reported by states or the news stations, but are only evident in the .json files.
The second data series in the .json is the % of vote series, where there are possibilities that rounding to three digits of precision can create rounding errors, but the total vote count is also reported in each row of the % series, and it should be explained why the total vote can decrease, which it does in several cases.
Also it seems there are some examples where the shift in ballots calculated exceed the size of a rounding error.
For example, the .json reports the following leaps in Penn:
11/4/2020 @2:13:11 - Total vote reported 1,111,586 (64.1% B 35.1% T)
11/4/2020 @2:14:32 - Total vote reported 871,782 (59.20% B 39.9% T)
11/4/2020 @2:17:03 - Total vote reported 774,021 (68.9% B 30.1% T)
11/4/2020 @2:18:59 - Total vote reported 1,288,457 (67.7% B 29.9% T)
11/4/2020 @2:21:59 - Total vote reported 1,325,632 (66.4% B 31.2% T)
11/4/2020 @2:22:45 - Total vote reported 739,443 (62.7% B 36.3% T)
Can any of the trolls trying to refute these election results please explain why President Trump WENT NEGATIVE in votes many times?
ReplyDeleteExplain how that works.
Can you explain this?
ReplyDeletehttp://graphics.thomsonreuters.com/data/election-assets/img/NEP_ExpectedVoteMethodology.pdf?fbclid=IwAR09u5sGbt4HTQ21CWfFuRUSY3uSz3aESD2GhQ5GKoT28d4MBE_JUdkCrxw
It's an piece from Edison data telling how they are changing the way they present the results this year.
The data taking from the NYT site says Edison on it which is usually connected to exit poll data?
It looks like Biden also loses votes throughout the day...how can the running total go down for either one?
ReplyDeleteNot only did they take voted from Trump , they subtracted those votes from his tally; and added the votes twice to Biden. The number is added in Biden’s tally and added to the total tally.
ReplyDeleteNice Job! One quick note: In PA spreadsheet, TrumpDelt had so many negative rows, why? They are real false or intentionally deduction?
ReplyDeleteThe statistics is the proof. In a random bundle of ballots at a Biden 70% clip; the odds of all being Biden is .7^100. That proves these weren’t random batches coming through the mail.
ReplyDeleteNow add affidavits of poll watchers being denied access, threatened, stopping counts, 140,000+ votes counted in 5 minutes; adding the circumstantial evidence of payoffs and personal connections (Biden’s Chief of staff, for example) are enough for the Court. It’s not a criminal trial. Proof beyond reasonable doubt isn’t necessary. It’s “given the reasonable person, would they presume fraud happened. Then the bigger question is what to do about it.
Yes. Exactly. This plus old fashioned traditional fraud.
DeleteThe courts already thrown out the lawsuit about poll watchers, judge said everything was ok.
DeleteThrown out is a big word still have appeals and tons of affidavit that are indicative of violation of the statutes and will be reviewed by groups of judges not just one partisan judge but you are right some of the suits were dismissed not over yet and some suits are being allowed to proceed like in the pa. Case
DeleteThis Must Be Brought to Light.
DeleteCheaters are still Losers. Our Founding Fathers have prepared us for such a time.
If an honest judge will let it in the court...
ReplyDeleteCan someone please send contact info for you. I would like to put some of the spreadsheet in an amicus brief in Michigan with Michigan dataset run. Antrim County caught a 6000 vote dump tranche. Is this backdoor midstream manipulation. Need to know where at between input point and output point. Please let me know. Cjtucker@lexfori.org
ReplyDeleteIf you want to show the difference between using the Voting system vs not, you need to show another State result that are not using this Voting system as a reference. That is what a normal pattern should look like and tell us how this pattern is different from normal. You possibly need more than one other reference State to be able to call it normal.
ReplyDeleteHow does a data feed published by the very source you claim to distrust, prove anything?
ReplyDeleteThis data was not directly published by the states holding the elections, it was published by the NYT.
Especially when the reports are not based on actual votes tabulated but statistical probability. Again maybe good old fashioned ballot box 13 fraud like the Democrats have been doing forever, but not nefarious computer midstream hack injecting votes for biden and removing votes for Trump. If latter how can this NOT be obvious?
DeleteTRUMPY WIN
ReplyDeletei want to believe, but this data is a mess, large and small oddities all over the place, in both directions. It's clearly not a running scorecard, or if it is, not one that can prove anything to anyone. This sort of thing has been all over the internets, and I thought, and hoped, it had legs, and maybe it does, but I'm profoundly sad after seeing the tables. I think if it had legs we'd know it by now, and see it somewhere else than the xyxtrumpwins.com
ReplyDeletesites...
Data for Georgia?
ReplyDeleteSpreadsheet for PA only available, Georgia link leads to article...
It's a result based on statistical reporting results not actual. So how does this help to explain anything other than statistically insignificant fraud events at the ballot box that have been occurring for years and years. Just Covid enhanced.
ReplyDeleteI've been studying the numbers and I am confused. Why are there places where even BIDEN drops in total votes? What kind of anomaly causes people to LOSE total votes? It happens in the numbers for both Trump and Biden, but it happens to Trump way more often.
ReplyDeleteStill, I'd like an explanation for why even Biden's TOTAL vote count DROPS a few times.
I spent hours analyzing this and making a graphic to help people understand it.. and then I saw this comment above
ReplyDelete"Anonymous said...
To check and see for yourself that this is bs:
Download the actual json data (or use an online viewer), scroll down to the time series, do you see actual vote counts for trump/biden? No, only the total of votes and the fraction for each candidate of that total. The split counts in the spreadsheet therefore are not original, but were generated by multiplying this fraction with the total count. The smallest change in that fraction of only three digits is 0.001, this multiplied with the end total gives you the smallest change in vote count per candidate that you can "detect" from this data (..try it, guess what number you get..).
9:27 PM"
He is right.
For example, 0.001 * 6,900,000 total votes = 6900, which is about what the shift is toward the end as expected (it increases as it goes on because total votes increase).
People are analyzing these data as if they are counts of votes. It appears that they are not.
ReplyDeleteDon't use these data in a lawsuit or you will embarrass yourself.
Someone took the percentage of votes for Trump or Biden and multiplied that by the total number of votes.
For example, if Biden has 48.44% (rounded in the NYT data to 48.4%) of the votes at one minute and the next minute he has 48.46% (rounded in the NYT data to 48.5%) of the votes, he will get a bump of about 6,000 votes in his margin: 0.1% times the total number of votes cast. If Trump's % stays the same, he would get a few more votes, but Biden would still get his extra bump of about 6,000 votes.
If Trump's percentage of the total votes went down from one minute to the next (eg, from 49.0% to 48.9%, then because of rounding, his total number of votes would go down as well, about 6,000 votes.
The weird pattern is just a rounding error.
And note that the earlier bumps average a bit less than 6,000 while the later bumps are a bit more than 6,000. As the total number of votes rises, then 0.1% of the total votes cast will get slightly larger.
Please, this is an error. It's a dead end.
If someone has the actual votes for each candidate by time, not the percentages for each candidate to the nearest 0.1% multiplied by the total votes, then one could analyze things.
If they were real counts (rather than essentially counts estimated to the nearest 6,000 votes), they would mean something.
Jim Lindgren, JD-PhD
Professor, Northwestern University
Great work! Sharing.
ReplyDelete"I spent hours analyzing this and making a graphic to help people understand it.."
ReplyDeleteAdding the graphic I mentioned above anyway because I spent hours making it.. I believe those comments + the one from the professor above are correct, but like I said I spent hours on this, maybe someone will find my graphs interesting or something. I noted in the graphic that it's probably debunked.
https://i.imgur.com/DzlOttT.jpg
This comment has been removed by the author.
ReplyDeleteI assume this is based on the NYT dataset, and the totals are derived from the % Tump and % biden fields?
ReplyDeleteBear with me, I am not too familiar with this.
My concern is that the 6000 vote steps you are seeing is an artifact from the significant digits of the % numbers it is derived from. (6,000,000 x 0.001 = 6,000) of percent is rounded to nearest 0.001, the result will always step as increased like seen.) Same with GA, but the total is lower, so step is smaller.
If you look at it on a graph, the step slowly increases as the vote total increases. I do not have the dataset you started with, so i may be wrong and can not confirm anything, but I wanted to warn you so you can look out for it.
I cross posted over at Insty as "jcp" if you want to contact me. I have other problems with that dataset and think it has little value to reveal the truth.
You've just gone viral on the news. https://theamericanconservatives.net/sidney-powell-rest-assured-the-kraken-is-on-it-dna-level-statistical-proof/
ReplyDeleteSeveral commenters, including Jim Lindgren, have noted this, but let me emphasize: the votes counts are not real. They don't appear explicitly in the original data.
ReplyDeleteIf I understand Edison's data and Doug's analysis, they are Doug's effort at reconstructing the (missing) actuals implicitly using the perfectly reasonable and correct formula
actual = total votes X vote share
Thing is, Edison's vote_share variable is only three significant figures. Nothing you do with it will have precision greater than one part in a thousand. And unaccounted changes of one part per thousand will mysteriously begin appearing everywhere you look.
Other parts of Doug's analysis may be correct. I certainly believe the Dems cheated. But this thing about same-sized bundles of votes mysteriously rolling in? That's a mirage, an artifact.
As this story catches fire in conservative media, we need to be very careful about what's real and what's a mirage. Cause you know the Lefties will focus on that one little thing we got wrong, then say the whole thing is "discredited."
You know they will do that.
Best regards, Steve
Yes, exactly. ACtual votes input by ballot should equal actual votes counted, should reflect actual voters registered and voted. This data is all supposed to be kept by law. 52 USC 20701 and 20702. But is it ? Probably not. However, there are still anomalies in the input data because of the weighted race feature in the machines. If there is a weighted race feature and no ballot images, then the data "can" be actually manipulated - meaning fake vote dumps can be made midstream (by someone other than the voter / elector) and then a weighted race feature is fixed to ensure that the favored candidate will coast to victory. If you don't have scanned ballots and retained paper ballots though you may not be able to prove this - UNLESS you more votes cast than there were actually voters voting. And this appears to exist and if so, that is an obvious fraud on the system and you don't need to prove midstream data manipulation by a ROGUE agent acting within the system through a backdoor, or programmed, and you also don't need to prove fraud in the ballots because more votes counted than votes actually cast is impossible. Right? What am I missing here. If the conservatives go with the Edison data analysis here alone it will be torn apart. They need to also look at the Dr. Shiva work on the major republican counties that had huge throttles of Biden votes, and then a weighted vote ration applied, because those counties had a straight republican favor but a statistically impossible anomaly in drop in votes for Trump.
DeleteEverybody: Do the math for yourself. It's not hard.
ReplyDeletePA: Multiply 6,000,000 votes by, oh, 0.492 (49.2%). Multiply it by 0.493. What's the difference? Huh, a bundle of 6,000 votes. One part per thousand.
Now do GA. Multiply 4,800,000 votes by 0.492 and 0.493. What's the difference? Huh, a bundle of 4,800 votes. One part per thousand.
Multiply any large number by any two 3-decimal numbers that differ by 0.001. The difference must be about one part per thousand. We can't see changes smaller than that. The vote_shares variable doesn't have the precision to let us.
We can't see any changes larger than that, either, unless they're large enough to register as two parts per thousand. That's why Doug also found bundles that were small integer multiples of 0.001. He couldn't find anything else, given the inputs he had.
Now dive deeper. Look at Doug's numbers, for instance
PA at 11/5/2020 16:58Z:
Votes = 3,230,103+3,114,512 = 6,344,615
1/1000 of that is 6,345
The "Bundle" = 6,265
They match, to within 1.3%.
PA at 11/8/2020 19:58Z:
Votes = 3,311,556+3,358,864= 6,670,420
1/1000 of that is 6,670
The "Bundle" = 6,767
They match, to within 1.5%.
But in 3 hours the size of the magic bundles grew by hundreds of votes. All while the bundles remained 1/1000 of the total vote. Huh. Did the cheaters start sneaking in with larger baskets? No.
I believe that Doug introduced a systemic error when he tried to reconstruct the detailed voting data from Edison's sparse data. (I assume he used Actual = Total x Share.) The rest of his analysis may be great - I hope it is, cause I believe the Dems cheated somehow - but the appearance of same-sized bundles of 4800 or 6000 votes is simply a mirage. I wasted a couple hours Monday slowly realizing this.
Don't believe me. Do the math yourself.
Do it yourself.
Best regards, Steve
ou guys could be right about the decimal places. All I can add is the original people who started this caught onto to this about the decimal places asap and had a long discussion about it. I don't think they could fully decide whether the decimals were causing the irregularities but they have been working on it for about a week now and the last time I checked they said, 'We've got them' but overall it all gone a bit quiet.
ReplyDeleteThis is the original post, you have to read through all the comments also as that's where the discussion is.
https://thedonald.win/p/11Q8O2wesk/happening-calling-every-pede-to-/
This is an updated post of the original.
https://thedonald.win/p/11Q8XQIWRs/-happening-ive-updated-the-switc/
Also I'd recommend this video analysis, he reckons he's worked out how they did it. Somebody above mentioned a weighted method of disguising the changes well this video explains that, you take votes from Independents, Biden and Trump and switch them around so that it not just Trump getting all the extra votes taken from him but ultimately Biden's votes increase the most.
https://streamable.com/vd158c?fbclid=IwAR03ExM3yRo228Kq1gbKeBamVx856UMnWfucU9eFO1xlW8XzSoFtmQ6sulg
That just links to a video, it's perfectly safe!
This comment has been removed by the author.
ReplyDeletethe error term is 0.1% of total vote. On the surface of it, it looks like some rounding issue.
ReplyDeleteEdison data only gave vote share % and total vote, so when you reconstruct vote numbers, the total % win by biden and trump could move up and down by 0.1% due to rounding.
BUT if the real vote count is recorded properly at raw data level, you would expect the number jump back when say vote share moved from 48.13% to 48.16% as instead of round down you would round up. (or at lest it moves to 48.2% when more votes are accumulated)
Looking at data presented, it seems on all these rounding lines, trump just simply lose that 0.1% and he is not getting it back. This deserves more attention in my opinion but simply looking at this you can not conclude 100% it is vote shifts.
You really don't need many of them to be faulty to cause election result to change. as each line is 0.1% of total vote, while vote share are very close.
Also, what we are looking at is just rounding at state level. Such rounding issue can happy at county level too. I don't know the details but if counties round up and down % number but not reporting actual vote number, it can shift a few thousands of actual votes here and there too. Then you can have it down to ward level to shift 10 votes here and there. You might think that does not matter but when you have thousands of wards, it adds up.
Two more issue I found in addition to rounding:
ISSUE 1: line duplications
Anyone looking at the edison data line by line?
In PA file:
line 340 and 344 ( seems line 344 = line 340 *2 exactly), how likely you have two pile of votes both having 95% for biden with total votes 11k and 22k?
line 394 and line 436 are almost identical
there is another pile of 15k vote it seems ( line 297, 308, 330)
ISSUE 2: lines shifting votes with no clear offset
line 188 324 337 etc are real line reducing trump vote add to biden but no line around it to indicate it is a clear error correction.
Bear in mind, no one would be so stupid to use EXACTLY the pile of votes, 100 for trump and 10000 for biden and put it through the scanner 10 times tell you what is happening. The worst can be done to hide such activities would be using a pre-checked pile, and then mix and match with a small amount of randomly selected votes to make the number look real. Then when doing it again, add/change it slightly.
The data can be 100% for real, but it is a fair question to ask for an audit.
The fact no one is even going through these again in a transparent setup is a joke.
Somehow different counties reporting their results in favor of one candidate or another is fraud? This is just a graph showing results as they came in. A rural county will not be the same result as an urban county. Maybe next time you can label the counties so we can verify the county results ourselves.
ReplyDeleteAnyone who posts this kind of data analysis without also posting data for a "normal case" or "unaltered case" is not doing enough due diligence to earn our trust. We are desperate for answers but stuff like this and Gupta are opportunists exploiting what more and more appears to be false hope. If this is the Kraken then its over.
ReplyDeleteThe public will not understand this. More importantly, the media won't understand it. Or they will misunderstand it as they are mostly incompetent fools. And the courts and judges are not going to rick getting doxed by deciding against Biden.
ReplyDeleteSad, but true.
DeleteI am more interested in vote losses (minus numbers in Trump delta in spreadsheets) than highlighted net change numbers, because the vote loss numbers show that the machines were moving Trump's votes to Biden's. It was too obvious and even the first grade can tell!!!
ReplyDeleteIs someone looking into NEW JERSEY!!! everyone i know cant seem to log in and see if votes were counted ... including me ! as we sit in a Blue blob!
ReplyDeleteHow do they compare to other states, FL, SD, WY, CA, NY, etc.?
ReplyDeleteThere would be the need to be able to verify the data, then chart it for each state.
Just curious if the Dominions was used in Miami... Dade County.... Fla was first state completed and yet there was great relunctancy to call.
DeleteIs there any legitimate reason that either candidates vote total would go down over the course of the election night counting? The first instance is at time 11/4/2020 1:24, where Biden loses 144 votes, followed a few cells down by 88 votes at 1:31 in the PA data.
ReplyDeleteNow look at PA's first highlighted cell, I don't see how this supports the vote switch theory. If the PA Multiple is 6k, At this time period, Trump only loses 1226 votes, whereas Biden gains 11679 votes, showing a 1.91x multiple of 6k. I'm assuming there is some aggregation of actual votes + flipped votes being reported at each time stamp. Essentially, if this were a 2-batch swap, Trump should have had about 11k votes to Biden's 0? Or if this is just a pure swap? If so, how do you explain the non-standard multiples?
The theory is they are also taking votes from Independents, a lot .
DeleteFor Dr. Lindgren and Steve...
ReplyDeleteYou are mistaking the highlighted cells for counts. They are not. They are deltas of deltas along a time-series continuum.
Using PA as an example, the time-series and deltas work like this:
Column F: the difference in reported vote count from the prior array element (for DT)
Column G: the difference in reported vote count from the prior array element (for JB)
Column H: the difference between F and G
Looking at counts along won't tell you much. We need to look at the swings along the time-series, ignoring the counts altogether.
The standard-deviation of the anomalies alone indicate massive vote manipulation, IMO. And it appears completely consistent with the reports out of Venezuela with Maduro's election using the Smartmatic system.
Amazing data. Where did it originate from? How does that information become available? I need more of it!
ReplyDeleteIt was scraped from the NYT website. On election night the big news agencies and a few other groups receive the ongoing vote count in real-time so they can display it on TV and on their website. It looks like they saved these values in a JSON file as each time a new count comes in they have to refresh that section of the webpage graphic that displays the vote and display the new totals, hence previous counts are saved and JSON is the popular file storage format atm. Personally I'm doubtful what this data actually is as it mentions Edison in it and Edison are a company that only deal in exit poll data. So there are questions as to what this data is in the first place. If you go to the thread on www.thedonald.win that explains most of it. I've posted the link above. I assume that's the info trump is tweeting about, you guys in here are just going through the same cycle of discussion they started around a week ago.
DeleteIt's impossible to know where they are at atm, they may have realised they made a big mistake and don't want to admit it hence it's all gone quiet.
Personally I'm very sceptical and think these irregularities can be explained other than fraud.
looking at Florida data, the round issue discussed happens less often, and you can see it affects both candidates in a more balanced way in the sense net effect nets off in pairs. Not all nets off because there are still real votes coming in, but it is definitely not like what you see in PA data.
ReplyDeleteIt would help us non-data-heads to see a "normal" graph (from a non-swing state?) alongside the fraudulent graphs
ReplyDeleteUnfortunately, this proves nothing, and Steve is not a troll.
ReplyDeleteThat doesn't mean there wasn't fraud; obviously there WAS. But this doesn't prove it, and I hope to discourage anybody from wasting the legal team's time with it. The problem is that this data is of low quality - there's just not enough information there.
As I understand it, any vote switching or dropping happens at the vote machine level - it wouldn't show up in the aggregate "chunks" reported in the data used here anyway.
EXAMPLE
Let me spell it out with an example. Look at the 0:12 and 0:20 data on the PA side (7th and 8th rows, on the left). It gives Trump's votes as 3267121 and 3260785. Take the second minus the first, and you get -6336. Simple.
The problem is where those two numbers come from. Specifically, they come from the total votes, times Trump's share, which the only have to 3 digits. Specifically:
6547337 x .499 = 3267121.163
6547763 x .498 = 3260785.974 (apparently he's truncating instead of rounding)
These numbers come from the original data file, but are not shown on the table above.
But the .499 and .498 here are obviously ROUNDED, so:
.499 is really anywhere between .498500 and .499499
.498 is really anywhere between .497500 and .498499
So the actual number of votes can be anywhere from:
3263847 to 3270388, for .499
3257512 to 3264053, for .498
That means the DIFFERENCE between the two vote counts can be anything from
3257512 - 3270388 = -12876 up to
3264053 - 3263847 = 206
Which means the real number CAN be positive!
What's the real number likely to be? Well, the increase in total votes is
6547763 - 6547337 = 426
In this counting session, Trump was getting 23.0% of the vote, so the real number is likely around:
426 x 0.230 = 98 (which is within the -12876 to 206 interval calculated above)
So while -6336 is what we get with our very imprecise numbers, the REAL number is probably about 98.
A single number for how many votes can be off by +/- 0.0005 x the total vote; the difference between two such votes can be double that. So when the total vote is around 650000, the error can be +/- 6500!
The reason why it only seems to effect Trump is that after the first pause in counting, Biden got about 3/4 of the votes. Trump's share was slowly counting downwards the whole time, while Biden's was increasing. Biden only has ONE such drop during that time period.
So the "packets of 6000" thing is bogus.
Still, the dataset DOES give us some suspicious hints. It's just that they're not "a smoking gun", unto themselves, and not discussed in this blog post.
-FJS, aka "Optimizer"
These blips are just reflections of manual and/or batch processes occurring in either the vote counting process or some later process that is aggregating results from the states, e.g., whoever is the source for the NYT data.
ReplyDeleteS
Makes sense to me.....however not believing the source NYT
Deletereally??
This is easily debunked. The original data stream had candidate vote percentages, rounded to 1 digit after the decimal, and current vote totals. THe data above is derived from this. If you take any change in votes from timestamp to timestamp for either candidate, the maximum associated error is 0.1% of the vote total at that timestamp. Check yourself - did either candidate at any point lose more than this amount in the PA data? NO!
ReplyDelete...with 3 exceptions. Trump lost 42325 at 2:14, 127916 at 2:17 and 145180 at 2:22, or 315421 total. Biden lost 196432 at 2:14 and 416589 at 2:22. So BIDEN lost more than Trump.
If this is all Trump's lawyers have (i.e. this is the Kraken...) then they are going to be ridiculed and will lose all credibility. I certainly hope this isn't the case, but they did Tweet out this link...
BAHAHAHAHAHAHAHAHA
ReplyDeleteTrump lost. It's over. Stop it. You all are making yourselves look as stupid as he has been making himself look. His next stop is a courtroom to defend against charges by SDNY. Then it's off to prison. Or Russia.
At 11:23, above, I explained how the 6000 vote "packets" were NOT a real thing.
ReplyDeleteBut the dataset DOES tell us some interesting things.
(1) The counting stopped after about 7.23 hours, and did not resume for 6.05 hours. However, in the middle of the time it was stopped, about 68912 votes were added, 90.0% of which went to Biden.
(2) The next day, everything was suddenly different from the previous day. About 956026 votes were counted, with 73.5% going to Biden. Were these supposed to be mail-in? And/or "Biden counties?"
(3) The next day, after a 9.75 hour gap, about 290544 votes were counted, with 73.2% going to Biden. It being basically the exact same percentage is suspicious. Are these supposed to be mail-ins that had not been counted yet, or not arrived the previous day? 3rd parties did better this day, 3.5% vs just 0.6% the day before. (Why?) Trump was still about 19918 ahead.
(4) The next day, after a 5.88 hour gap, about 93613 votes were counted, with 78.3% going to Biden. Biden pulls ahead, about 33659 votes More mail-ins still arriving? Maybe absentee ballots? They must have known they were nearing the end, maybe that's why they went with an even higher percentage for Biden.
(5) Over the next 5 days, a total of about 72208 votes are counted. "Curiously", only 68.4% went for Biden. One might conclude that the previous days' fractions, which JUST got Biden over the finish line a little bit ahead, were fixed, and that 68% is a much more reasonable fraction for these kinds of votes.
So that might give you an idea about how this fraud went down, but it's not sufficient evidence that could be proven in court.
This was NOT supposed to be "the Kraken", BTW. I'm sure that refers to testimony from the Dominion whistle-blower and other witnesses, plus hopefully data from the server that was seized in Frankfurt.
-FJS, aka "Optimizer"
73.5% sounds right I read somewhere for one state the ratio of mail in votes was 75/25% D/R I can find the link if you want?
DeleteIt's the 90% 9nes that are suspect.
DeleteI’ve been a numerical analyst my whole career, recently retiring. I look at numerical sequences for patterns to be able to approximate the behavior mathematically and algorithmically. Numbers in nature can be quite noisy and hard to predict. This pattern sticks out like a sore thumb. In both cases a target gain in votes is divided over the selected precincts to modify to achieve the needed small win. The difference in the Trump and Biden votes are compared for each of these precincts reporting. If this is not enough to get the target gain, The remained of the needed votes are changed from Trump to Biden to get desired gain. In the case that Trump did not get enough votes to cover the switched vote count, his delta votes goes negative. We all know THIS SHOULD NEVER HAPPEN. This is happening in the software and database. It is really a stupid and easily discoverable algorithm. You just need the data complied like this to see it. Kudos!
ReplyDeleteRounding errors go away when you show a range for the % given. Ex. 56.4 Add a couple more columns for 56.45 (minimum votes) and one for 56.44 (maximum votes) - and you'll start to see there is no vote switching. It's just rounding errors.
ReplyDeleteIt's important to include the 3rd party values, too. They only provide total votes, Trumps % and Biden's %. You have to calculate the 3rd Party by subtracting Trump and Biden from 1. So, your variance is higher simply because you've omitted 2.1 to 2.5 % for the third party.
Also, the percentages for each candidate and third party levels off after the first day. The number of votes added per time slice go WAY down. In some states, you might see 1 or 2 votes added. That, too, starts to make the total votes per candidate look odd - again, this is just rounding.
This data shows nothing nefarious - I was hoping to see the smoking gun, too, but it's not there.
What good is any of this when they aren't presenting it to any judge or court.
ReplyDeleteBecause they probably don't think a judge will understand it and hence throw it out, they need to have it presented in a way that undisputable.
DeleteIt's weird that you have professional data analysts here that are saying totally different things.
ReplyDeleteAh here you go, from this, thread,
ReplyDeletehttps://thedonald.win/p/11Q8O2wesk/happening-calling-every-pede-to-/?sort=old
I do a worst-case and best-case rounding error to generate Max T Vote, Min T Vote, and Max B Vote and Min B Vote that shows if you have 0.578 share, the max number that would represent would be 0.57849999999, and the min number would be 0.5775. The specific pic highlighted here allows no possible way whatsoever that Trump did not lose votes. The smallest number of votes he could have lost assuming in 27 there is a worst case rounding error and in the 28 step there is a best case rounding error (extremely unlikely), Trump lost a 4846.353. In worst case, he lost 5994.627 votes. There is literally NO way anyone can explain this away. They can also not explain that away as an asynchronous issue, as the timestamp itself is later (1), and (2), the total vote count increased. This single time step is proof alone.
Comments about rounding errors do not explain vote totals going down. The totals should always be rising.
ReplyDeleteThe rounding errors do explain the vote totals going down. Here's a real example:
ReplyDeleteAt 10:54 UTC, 11/4/2020, T share was 0.551, B share was 0.436
At 14:16 UTC, 11/4/2020, T share was 0.550, B share was 0.436
So T share went down while B share stayed the same. This doesn't (necessarily) mean that Trump's votes went down and Biden's stayed the same. If, for example, Trump's actual share was 0.5505, which was rounded up to 0.551 in the "before," then went to 0.5504 in the "after," he still would've gained votes in actuality because of how much the total amount of votes counted went up in that time.
See example calculations in image below
https://i.imgur.com/4UxHzId.png
(this is of course just a hypothetical to show an example, we don't have the real numbers)
Here is data for all 50 states, presented with a bit more info and some graphs:
ReplyDeletehttps://docs.google.com/spreadsheets/d/1gkXXr8mnHfLzBbMKQSA-J402sGZuETSVgw_qM_IQUls/edit?usp=sharing
However, adding to what I said above, the example given above from thedonald.win should not be possible. (example is from Michigan btw)
ReplyDeleteAt 01:51:26 UTC, 11/4/2020, T share was 0.578, B share was 0.401
At 01:51:52 UTC, 11/4/2020, T share was 0.568, B share was 0.406
There were only 560 votes added in that update.
So even if Biden got all 560 and Trump got 0, the results would not even be close to possible, as the guy said.
So what we need is a way to run through every instance of an estimated net vote change being negative and calculate, given the total number of votes, if that instance was possible or not, based on best case scenarios and worst case scenarios. If both values are negative, then its not possible without the removal of votes.
I assume you don't read any of my replies?
DeleteFrom thedonald.win
I do a worst-case and best-case rounding error to generate Max T Vote, Min T Vote, and Max B Vote and Min B Vote that shows if you have 0.578 share, the max number that would represent would be 0.57849999999, and the min number would be 0.5775. The specific pic highlighted here allows no possible way whatsoever that Trump did not lose votes. The smallest number of votes he could have lost assuming in 27 there is a worst case rounding error and in the 28 step there is a best case rounding error (extremely unlikely), Trump lost a 4846.353. In worst case, he lost 5994.627 votes. There is literally NO way anyone can explain this away. They can also not explain that away as an asynchronous issue, as the timestamp itself is later (1), and (2), the total vote count increased. This single time step is proof alone.
As a conservative, this data set is meaningless. As others in the comments have pointed out, this is due to the underlying data being imprecise. Snippet of the data format is shown below.
ReplyDeleteData link for those that missed it: https://static01.nyt.com/elections-assets/2020/data/api/2020-11-03/race-page/pennsylvania/president.json
{
"vote_shares": {
"bidenj": 0.498,
"trumpd": 0.49
},
"votes": 6812866,
"eevp": 99,
"eevp_source": "edison",
"timestamp": "2020-11-12T20:10:47Z"
}, {
"vote_shares": {
"bidenj": 0.498,
"trumpd": 0.49
},
"votes": 6813048,
"eevp": 99,
"eevp_source": "edison",
"timestamp": "2020-11-12T20:20:48Z"
}
When generating the excel spreadsheet, the votes for each candidate seem to have been calculated by multiplying the "vote_shares" value for each candidate against the "votes" field. Can you see the problem? The float data only has 3 significant digits at best.
Because we are effectively rounding to nearest 0.1% (0.001) of total vote with this setup, we inevitably see the jumps of 0.1% of total vote as the "vote_shares" percentage distribution number ticks up or down for each candidate. So when a candidate shows up as suddenly having negative votes, whats really happening is that they were being over reported in previous intervals of this data due to the stupidly low floating point precision, which eventually "corrects" itself as the percentages round up or down after enough votes are gained.
Important to pay attention to your precision. You would need the actual raw event counts for each candidate over time to try and make this analysis -- not low precision floats being multiplied by the total vote count. You can't pull high precision analysis out of low precision data.
People have been through all this, they start off with, " it's just a rounding problem nothing to see here" then after more analysis they end up with "it's vote switching, he was robbed!!!"
DeleteI know a certified quality auditor who does data analysis for a living and he looked at the Michigan vote chart that CNN had on their source code and he laughed out loud. He said there is no way that graph was created naturally. The data was manipulated. No natural graph would look like it did. You could see where they added votes en mass and added to Biden while deducting for Trump. An algorithm was used. Clear as day. There is MASSIVE voter fraud for Biden but we all know there is NO way Biden got more votes that Obama. And now, after seeing this I wonder if Obama really won his second term. Would love to see the voter data on that.
ReplyDeleteOur God NEVER fails... Truth and justice will prevail
ReplyDeleteI've been a programmer since I was 14. By the time I was a sophomore, they ran out of math courses for me. I also became a rated master for the us chess federation. However, I failed mostly everything else because I was a total outcast to what they were teaching me. At 12 years old I listened to Ross Perot the first independent allowed in the national debate and realized "wow, this man is makes sense". He took tons do votes in the election". then I realized why they don't allow Independence to speak at a national debate because they get attention and take away from the power of the two parties.
ReplyDeleteIt was from that point on I became an independent seeing through the lies of the system.
I dabbled with statistical analysis it is what I use to find out truths when the media tells us something .
I voted 3rd party this time around. From everything I know and learned in my life, I can tell you that I was 1001% certain third party candidates would get more votes in 2020 than 2016.
Especially with Bernie Bros threatening to vote independent.
Nope! In 2016, Independents garnered about 7 million votes. In 2020 after a record turnout... Only about 2.6Million went to independent voters.
I would like to see the charts for independent votes as well I believe those votes were stolen and given to Biden.
Yep, watch this.
Deletehttps://streamable.com/vd158c?fbclid=IwAR03ExM3yRo228Kq1gbKeBamVx856UMnWfucU9eFO1xlW8XzSoFtmQ6sulg
Nobody is talking about third-party voters! In 2016 third party can and it's received about 7 million votes.
ReplyDeleteIn 2020, third party candidates received about 2.7 million votes.
4.3 Million less votes for 3rd party candidates after a record smashing voter turnout?
Time to pull out the charts for third party votes! I believe some of those votes were also flipped to Biden from the software.
nobody is paying attention because they are too focused on trump votes being flipped.
Publish those DNA statistics on 3rd parties NOW!
Watch
Deletehttps://streamable.com/vd158c?fbclid=IwAR03ExM3yRo228Kq1gbKeBamVx856UMnWfucU9eFO1xlW8XzSoFtmQ6sulg
Someone Needs to look into the 3 or 4 recycling plants that mysteriously caught on fire in the days around the election. It's really been bothering me. Along with the helicopter crash and 2 small planes crashes during those same days. It was the day before the boiler room exploded at the VA hospital. Too many things. Im not crazy..Burning evidence, ballots etc or killing people who had more proof to get to the good guys, is not something I'd put past these people.
ReplyDeletethe workbook link for GA is missing (leads back to article)
ReplyDeleteWhen this evidence is properly submitted it seems that SCOTUS could rule to just swap back the votes right? Also those swaps were mainly from republican precincts based on the algorithm analysis from Dr Shiva. Just put things back where they belong, and that includes the rioters that will follow...
ReplyDeleteHi, I have been replicating this excercise myself. Where is the proof that this was fraud? In the fact that Biden got more mail-in votes?
ReplyDeleteThere is absolutely no proof in this. The numbers you are highlighting only prove that Biden got more mail-in votes than Trump. Trust me, I would like it to be different, but this is very weak. BOTH candidates constantly lose votes. You could make the exact same argument saying that Trump cheated, because Biden loses a TON of votes in this strange process too.
Plus we have done this for far more states than PA and guess what? In some states BIDEN loses more votes than Trump...
1. The link to the GA dataset goes to PA. Where is the GA sheet? I suppose I could snag it from NYT (but they might be a Byzantine General)
ReplyDelete2. Comments on 'rounding' go more generally to the subject of 'quantizing'
3. Rounding down ('floor' function) for one candidate and up ('ceiling' function) for another is an old (no longer possible in banking) scam known as the "sausage scam" where little bits are shaved off, each one not very noticeable. But little by little a big pile is made. [popularized in movie "Office Space", but flick is technically wrong since banks have been using binary-coded-decimal (BCD) forever, not base-2 numbers]
4. Time-series analysis on this would be useful, but morese with both GA and PA dataset. For the naive, just run it through the "Auto-Modeler" in RapidMiner.
GA link is fixed, thanks
ReplyDeleteGeorgia spreadsheet.
ReplyDeleteBut, but, but DNCNN and MSDNC insist that there is no voter or election fraud. /sarc off :-)
ReplyDeleteIn its legal filing Wednesday, Trump’s campaign maintained that Pennsylvania’s election results fit that profile and argued that a series of policies enacted by Democrats had purposefully and illegally diluted GOP votes.
ReplyDeleteThe new complaint steers into *****unsupported****conspiracy theories laid out in court Tuesday by Trump lawyer Rudy Giuliani and accuses state and local leaders in Democrat-leaning counties of working in concert to steal the election through fraudulent mail-in votes
It's not the vote differences over time that stand out...because larger vote numbers came in earlier compared to later, so that's an apples-to-oranges comparison. What stands out more to me is the RATIO of Trump to Biden votes...a ratio is apples to apples, rather than raw vote numbers. It is THIS RATIO that is HUGE red flag. Every vote drop after Biden pulls head is pretty much 1:1, Trump to Biden vote. Only an algorithm could do that.
ReplyDeleteContact for Sidney Powell:
ReplyDeletehttps://www.federalappeals.com/contact-us/
LOL. You are chasing your tails, and the only person who explained it to you, Steve, you rudely blew off as a "master troll." These numbers read of in "batches" of 6000 and 4800 because the respective computers are simply pegged to data set numbers, and that is based on the fact that whoever made this chart is not working from actual raw numbers but from election PERCENTAGE data. Do you understand? They guy who made all this chart took election percentage data and tried to turn that into real numbers, but without all the rounding detail, which gives you these nice rounded blocks of numbers, and in addition the computers are REPORTING those percentages in chunks over time, because that is how they were set up to update. So you can, say, set up your computer to update the numbers with each .001 Percentage change in the votes, and you will get the wacky numbers above. It has nothing to do with anybody secretly shredding votes for Trump and replacing them in the files with votes for Biden. THAT IS VERY CRAZY TALK. REMEMBER THERE ARE CAMERAS EVERYWHERE. But regardless, this chart is barking at the moon.
ReplyDeleteJust post the information about Eric Coomer's arrest... He's the Antifa kingpin spearheading the fraud as a Security Agent of Dominion Software... a criminal for all to know!
ReplyDeleteI did an interval analysis (see FJS/Optimizer) on the entire PA dataset. With one exception,* at no point once things settled down** did Trump's vote count delta _clearly_ go negative.*** The other times it appeared to go negative can be explained as artifacts.****
ReplyDeleteThat one exception is important. The 04:07:43-04:08:51 drop is unambiguous. In 68 seconds Trump lost 14,892+ votes and Biden gained 14,945+. PedeInspector and others have reported on it too. I never heard a good explanation for it.
It's probably the result of poor quality data not outright fraud, but this is a good place to focus investigations. What happened in PA at 04:08Z (11:08pm EST)? Anything? To answer that definitively we need better data than this NYT/Edison junk.
There may yet be a smoking gun here. But it has been eight days. Doug needs to retract the weak parts of his analysis (same-size bundles, ambiguous drops) and highlight the strong parts (11:08pm unambiguous drop).
Best regards, Steve
* At 2020-11-04T04:08:51Z: B=0.426, T=0.560, V=2984522.
B=(0.4255,0.4265), T=(0.5595,0.5605)
VBmin=1,269,914, VBmax=1,272,899, VTmin=1,669,840, VTmax=1,672,825
DeltaBiden=14,945 to 20,914, DeltaTrump=-20,861 to -14,892
** Very early in the count, during a ten-minute period (2020-11-04T02:13:11Z to 02:22:45Z) when the total was jumping all over, both Biden and Trump had one or more unambiguous negative deltas.
*** That is, dVTmax=VTmax(n)-VTmin(n-1) < 0
**** That's not to say they're false, just that they're ambiguous. They can be explained another way.