W. Edwards Deming, a scholar and statistician once said, "In God we trust, all others must bring data."
One of the first posts on this blog debunked the earliest of 2020 election "fact-checks" from DHS/CISA by showing the evidence of vote switches in Pennsylvania and Michigan.
Those of us who are alleging (proving) vote manipulation do so based on the election data.
Conversely, the reporters who write the fact checks trust the words of election software companies, polling organizations, and government officials who rarely, if ever, provide data to support their claims of a secure election and an accurate vote count.
What is truly amusing, however, is when those rare instances occur and the media sources attempt to use data to support their claims. These instances reveal not only their own ignorance of data but the ignorance of the fact checkers (who obviously didn't pay attention in math class).
On December 2, 2020, a Reuters fact-check attempted to debunk or otherwise discredit an individual who used Edison Research's election night reporting data (NEP) sent to the New York Times (NYT) to demonstrate incidents of vote manipulation in Virginia.
From Reuters:
"Social media users have been sharing an article which claims to use spreadsheets of election data from the New York Times (NYT) to show evidence of voter fraud in Virginia. The (Gateway Pundit) article says that vote entries are not in whole integers, some entries are removed and re-added, and the proportion of votes for each candidate is the same for most entries. However, the NYT confirmed to Reuters that it did not make the spreadsheets and its own data is in whole integers."
Failed Fact Check #1: It is not relevant if NYT made the spreadsheets.
Here is the link to the NEP data that was transmitted to the NYT:
Next are three sequential transactions from the raw NEP data from the state of Virginia showing percentages of votes for Biden and Trump expressed as percentages.
{"vote_shares":{"bidenj":0.542,"trumpd":0.44},"votes":4456303,"eevp":99,"eevp_source":"edison","timestamp":"2020-11-19T14:52:42Z"}
{"vote_shares":{"bidenj":0.542,"trumpd":0.44},"votes":4456303,"eevp":100,"eevp_source":"edison","timestamp":"2020-11-20T14:18:12Z"},
{"vote_shares":{"bidenj":0.542,"trumpd":0.44},"votes":4456310,"eevp":100,"eevp_source":"edison","timestamp":"2020-11-20T21:30:53Z"},
Percentages are not integers. They are irrational numbers.
Failed Fact Check #2: The NYT data is not in integer form for Biden and Trump.
The strings also show that the source for the NYT is "edison" and that the percentages are, in fact, not changing with each transaction.
Failed Fact Check #3: The proportion of votes are the same -- and Reuters ignored that.
In order to further analyze the results, those investigating election fraud typically import the raw data into an excel spreadsheet or to a statistical software program such as SAS/JMP. Once the data is in the spreadsheet, numerous computations can be derived to show votes added and lost.
A spokeswoman for the NYT told Reuters that their company did not make the spreadsheets in the Gateway Pundit article. The data used by the NYT comes from the National Election Pool and Edison Research (here , here). In this data, the number of ballots appears in whole integers, unlike the figures in the Gateway Pundit data.
Failed Fact Check #4: The first link is to a NYT graphic that is derived from the NEP data that again shows percentages with decimal places.
Failed Fact Check #5: The second link unwittingly confirmed that the analysis is using actual vote counts from Edison (NEP).
From Reuters:
Next, the article (ref Gateway Pundit) draws attention to three vote dumps totalling over 900,000 votes for Biden and almost 200,000 votes for Trump, two of which were subsequently subtracted. The article says, “These reversing entries look like adjusting entries […] There really is no logical reason except for fraud.” Lastly, the article claims that nearly every entry from 8AM ET on Nov. 4 to 9:30PM ET on Nov. 7 had the same proportion of votes for Biden (55%) and Trump (45%), saying, “This too is not reasonable and impossible. […] This too indicates fraud.”
The transactions highlighted below (from my spreadsheets) are the ones mentioned in the Reuters fact check that show the "adjusting entries." The figures are the same as Gateway Pundit's (GP), however my spreadsheet contains more detail that shows vote manipulations other than those pointed out by GP.
Note that the second entry (Row #143) does not correct the first entry (Row #133), but instead removed 2,894 of the "Leftover votes" (2,984-5,978) while leaving residual votes for Biden and Trump.
This is part one of a two part vote shift for Biden.
Part two of the vote switch occurs in Row 147, where -2,610 Leftover votes are moved to Biden. Doing the math: New Votes = 48,769 - Trump 20,714 = 28,082 + Leftover 2,610 = 30,692 (Biden).
Rows 145 and 146 are legitimate adjusting entries that follow back to back.
Entry #5 (line 152) is a corrected entry that disadvantages Trump by a net 2,564 votes.
However, Reuters explanation of these transactions was provided by Edison Research and the first is not accurate based on the computations above.
A spokesman for Edison Research explained to Reuters that the subtraction and multiple entries occurred because Edison deleted two web reports from Fairfax County after they could not be confirmed: “At 11:42 PM on 11/3 we received a web report from Fairfax County and that went into our system. We could not confirm that report so we took it out of our system at 12:12AM on 11/4.”
Fact Check Fail #6: The first "correction" left votes in the system for Biden and Trump.
“Similarly, we receive a feed report from Fairfax County at 12:26AM. We could not confirm that report immediately so we took it out of our system at 12:30AM. We then confirmed the vote report from Fairfax County at 2:17AM and it was put back into our system.”
Confirmed by the data and is a legitimate adjusting entry.
Regarding the claim about the proportion of votes received by Biden and Trump staying at 55% to 45% for most of the vote entries, the Edison Research spokesman told Reuters, “That does not match the data in our log so I have no idea where the data they are referring to is coming from.”
Fact Check Fail #7: Edison feed confirms GP's claim of 54% to 44%
The Gateway Pundit column stated that proportion of votes was 54% to 44%. Those figures are confirmed in the NEP data.
{"bidenj":0.541,"trumpd":0.44},"votes":4457418,"eevp":100,"eevp_source":"edison","timestamp":"2020-11-22T14:25:25Z"},{"vote_shares":
{"bidenj":0.541,"trumpd":0.44},"votes":4457700,"eevp":100,"eevp_source":"edison","timestamp":"2020-11-22T14:27:28Z"},{"vote_shares":
{"bidenj":0.541,"trumpd":0.44},"votes":4457624,"eevp":100,"eevp_source":"edison","timestamp":"2020-11-22T14:29:17Z"},{"vote_shares":
{"bidenj":0.541,"trumpd":0.44},"votes":4457619,"eevp":100,"eevp_source":"edison","timestamp":"2020-11-22T14:30:50Z"},{"vote_shares":
{"bidenj":0.541,"trumpd":0.44},"votes":4457935,"eevp":100,"eevp_source":"edison","timestamp":"2020-11-22T14:33:08Z"},{"vote_shares":
{"bidenj":0.541,"trumpd":0.44},"votes":4457936,"eevp":100,"eevp_source":"edison","timestamp":"2020-11-22T14:34:39Z"},{"vote_shares":
{"bidenj":0.541,"trumpd":0.44},"votes":4458350,"eevp":100,"eevp_source":"edison","timestamp":"2020-11-22T14:37:08Z"},{"vote_shares":
{"bidenj":0.541,"trumpd":0.44},"votes":4458533,"eevp":100,"eevp_source":"edison","timestamp":"2020-11-22T14:38:43Z"},{"vote_shares":
{"bidenj":0.541,"trumpd":0.44},"votes":4458342,"eevp":100,"eevp_source":"edison","timestamp":"2020-11-22T14:40:37Z"},{"vote_shares":
{"bidenj":0.541,"trumpd":0.44},"votes":4459317,"eevp":100,"eevp_source":"edison","timestamp":"2020-11-22T14:42:56Z"},{"vote_shares":
{"bidenj":0.541,"trumpd":0.44},"votes":4459430,"eevp":100,"eevp_source":"edison","timestamp":"2020-11-22T14:45:02Z"},{"vote_shares":
Clearly, the Edison spokesperson is unaware of the data that it sends to the media and it's also obvious that the reporter did not fact-check Edison's claims.
Reuters (and most other major media outlets) simply take the word of their source when it fits the preferred narrative. This passes as not only acceptable journalism, but as exceptional journalism under today's agenda driven media world.
The New York Times and Washington Post were awarded with Pulitzer prizes for their reporting on Trump-Russia collusion -- all of which turned out to be false because their "Deep State" sources (i.e, Clapper, Brennan, Schiff, et al) were not telling the truth.
Pulitzer citation (my emphasis added)
For deeply sourced, relentlessly reported coverage in the public interest that dramatically furthered the nation’s understanding of Russian interference in the 2016 presidential election and its connections to the Trump campaign, the President-elect’s transition team and his eventual administration.
In a different fact check, Reuters claimed that a vote spike of 140,000 votes in Michigan was due to a clerical error and was later corrected. FiveThirtyEight made the claim that the spike was corrected.
FALSE!
That error remained in the vote count.
The bottom line is that media fact checks about the 2020 election are nonsense because they rely on unreliable sources who make statements that are not supported by the election data.
Ray Blehar, February 7, 2021, 12:38 PM EST
No comments:
Post a Comment