Technical study by an international group of forensic analysts indicates strong divergences between new voting machines and old voting machines. A regional analysis.
By Matthew Tyrmand & An International Consortium of Forensic Analysts (Who Wish To Remain Anonymous)
The recent Brazilian presidential elections are becoming an Interpol-level affair given massive criminality with much already proven and much more to be uncovered. What we know is already proven to be a world-class statistical scandal.
Every day that passes, new independent analysis appears that increasingly show the fragility of the corrupt system that has been installed in Brazil for counting votes and fabricating electoral results.
A group of forensic scientists made a new analysis of the electoral data of the second round, and showed new evidence of fraud, falsehood and artificial tampering of results. Technically the expression is “inconsistencies” or “anomalies” but when taken in totality- the only logical conclusion is outright fraud.
Considering all the recent facts, and that one of the judges of the Brazilian Electoral Court declared that “elections are not won, they are taken”, what might be called “anomalies” starts to acquire a more explicit status- “manipulation.”
Not long ago, the leader of the Brazilian judicial militia- behaving as judicial dictators more than the ministerial judges that their constitutional role mandates, Alexandre de Moraes, ordered the inclusion of Brazilian journalist Allan dos Santos on Interpol’s “red notice” list, for political reasons and without an official conviction, or even a formal charge.
Interpol refused to comply with the capture and extradition order (as is mandated by such designation) of this judge given this insane overreach. They saw what we observers of good faith also see: Allan dos Santos is a journalist and not a criminal.
From the look of Brazilian electoral data, and considering all the evidence of fraud and the falsification of results that has been uncovered in recent days, it is rather Moraes who should end up on international criminal capture list.
Trial at the International Criminal Court (ICC) in the Hague would be an appropriate venue if the Brazilian judiciary is still too infiltrated with those who abetted his crimes against the Brazilian constitution and citizenry.
HIGH POPULAR REJECTION
In Brazilian public opinion Moraes already figures as “serial criminal” and “public enemy number one” to most citizens. Neither Moraes nor his Supreme Court colleagues can walk the streets or go to restaurants, whether in Brazil or in the United States, without being approached by Brazilians who question the lack of integrity and impartiality in the conduct of elections by the electoral court (TSE- the subsidiary court to the STF- the Brazilian Supreme Court…he heads both “deliberative” bodies).
On the day of the electoral certification (Monday December 12th, illegally pushed up one week by court fiat from the 19th in order to rush the process through), Moraes in the role of President of the Electoral Court, went to a celebration party for the candidate they had proclaimed winner, the ex-convict and ex-president Luiz Inacio Lula da Silva (“Lula”) that they had freed after he had served only 580 days of a 12 year prison sentence for public corruption (known as “Operation Carwash”. He had been tried in three courts and convicted of 12 charges unanimously by three panels of judges totaling 19 jurists. Upon his release they annulled his case, vacated his sentences, and expunged his record (given Brazil’s law that no convicted felon can run for public office).
It would be like the NFL’s refereeing team showing up at the Super Bowl winning team’s party. A notorious display of conflict of interest, partiality and dishonesty.
In analyzing the data from the second round, the forensic scientists found serious discrepancies.
The data took as reference the voting of the candidates in the second round, considering the electoral zones, with the analysis of the second digit of the numerical series.
It is evident from the analysis of these two graphic groups that if the results were corrected within the parameter of the discrepancies found, the winner would be the candidate Jair Bolsonaro, current President of Brazil. This is because the anomalies found indicate an artificial decrease in Bolsonaro’s votes against an artificial increase in Lula’s votes. Considering that the final election result was a difference of less than 1.8% in favor of Lula, a minimal correction would be enough to reverse the results.
In the example of the Southeast region, Bolsonaro’s votes are close to normal within the universe of new electronic voting machines. However, within the universe of old electronic voting machines (without adequate certification) Bolsonaro’s votes suffer a high incidence of abnormality.
In 10 of the possible sets of votes, 6 present inconsistencies greater than 10%, that is, an anomaly greater than 60% of the universe of votes, which indicates the possibility that their votes were possibly emptied within this geographic base (Southeast).
In the example of the Northeast region, Lula’s votes present anomalies in 7 of the 10 possible universes.
Considering the high level of anomalies in digits 7 and 8, this indicates a possible artificial alteration in the frequency of these digits. Digit 7, for example, has an anomaly rate of around 40%
The technique used in the audit is Benford’s law, a method widely used in investigations and audits by the FBI, CIA and various international investigative agencies.
It is necessary to say that Benford’s Law is not absolute, as it does not indicate the frauds themselves, but rather points out paths to be followed as investigation trails. After the Benford’s Law audit on digital data has indicated the avenues to be investigated, there should be a second stage of physical verification of evidence and documents. And here there is a serious problem, as the process led by Moraes and his cronies does not allow any document for physical verification. That is, in Brazilian elections it is impossible to carry out recounts.
We applied Benford’s Law to both the first and second round of this cycle’s presidential election and the anomalies that would indicate results are unlikely to be as they were presented in the initial post-election data- before the TSE took the data down two days after the second round and scrubbed it before putting “the data” back up- but not before it was downloaded by many forensic experts including Argentine expert Fernando Cerimedo.
The main logic of Benford’s Law is the following: If a fraudster wants to avoid reaching the threshold number of 100,000 Dollars to evade inspections, he will try to disguise this number by making an entry of 99,000 Dollars. With this he will change the natural frequency of the numbers.
There is no reason to refute the application of Benford’s Law to this data, as two identical data sets with the same patterns are being compared. If the data were generated naturally, the frequency patterns would be similar. But they aren’t similar. On the contrary, the two datasets analyzed have strong discrepancies between them. Considering that they were generated on the same day and on the same geographic basis and at the same times, they should show similar behavior. But that’s not what happens.
Bolsonaro’s votes in the Southeast region show signs of being reduced. As for Lula’s votes in the Northeast, there are indications that they have been increased.
If we consider a frequency path that starts from number 1 and goes to number 7, it can be seen that in Lula’s votes this frequency increases, that is, there are more numbers 4, 5, 6 and 7 than numbers 1, and number 7 is 40% above normal. This is a strong and clear indication that “high” numbers have been thrown in where there should have been “low” numbers. The norm within the frequency patterns of Benford’s Law is exactly the opposite, the number that appears the most is the number 1, and then the other numbers have their frequency reduced.
In the case of Lula’s vote in the Northeast, the opposite occurs, the frequency increases for the larger numbers. The frequency of numbers 4, 5, 6 and 7 is higher than the frequency of number 1. And numbers 4, 5, 6 and 7 are above the expected average, all of them.
In Bolsonaro’s voting in the southeast region, the exact opposite occurs: Numbers 4, 5, 6 and 7 were withered, that is, their frequency is reduced to below the expected average. All of them are below average. Within the same universe, one of the groups oscillates downwards, and the other oscillates upwards. With the same constancy.
This is nearly impossible from a forensic statistics point of view. Which makes this statistically questionable.
NO PHYSICAL VOTES
The lack of physical elements that allow an audit is a very serious inconsistency of the Brazilian electoral process. Especially considering that Electoral Court (TSE) judges interfered directly to block the attempt to formulate a law creating a duplicate paper ballot of the vote receipts to ensure an audit-ability. Judge Barroso, a member of the Brazilian judicial mafia, declares in his lectures that he personally went to the National Congress to prevent this statute from being approved.
It was not uncommon the last two years to hear Bolsonaro supporters, gathered en masse at rallies, chanting: “give us paper ballots, give us auditable ballots.” They knew what was primed to happen. They also remember Bolsonaro, when he was a legislator in the lower house, known as a deputy in the Chamber of Deputies (like a member of the House of Representatives in the USA), spearheading legislation to put printers at the polling precincts so as to have this paper ballot back up. This legislation passed the lower house and was vetoed by then-president Dilma Rousseff (who took over when Lula was booted from office and charged with public corruption crimes, she was later impeached and convicted in the same public corruption scandal- the aforementioned “Operation Carwash”). She cited fiscal costs as to what compelled her to exercise this veto. As if socialists are ever dissuaded from public spending due to pedestrian concerns such as expense incurred by taxpayers. Let alone the de minimis costs of printers. When it was sent back to the Congress the veto was overrode by a super-majority of members (a rare occurrence in the fractured and fractious Brazilian legislature). Upon the veto override, the judiciary stepped in and extinguished the law on supposedly constitutional grounds. They cited “privacy concerns.” The fix was in.
Ensuring no back up paper ballots and a lack of physical votes and the impossibility of recounts is intentional, and is part of the dirty project of the judicial dictatorship by this judicial cohort- with the majority of their ranks being populated by Lula, Dilma, and her successor, Temer’s judicial appointees.
It is also important to note that according to the Brazilian constitution, the military has a right and responsibility to audit the election if irregularity is alleged (which it has been since the conclusions of both the first round and second round). To that end the military has demanded to review the machines (known as “urns”), the source code, and the tabulation data. They have been summarily ignored by the Electoral Court. In the 65-page military audit report submitted to the court and to the public, they stated there are anomalies but we cannot prove anything as of yet as we have been blocked out by the illegal obfuscation of the court that oversees this process. Dutifully the global mainstream media has reported this as them having concluded “there is no fraud.” Needless to say the reporting media has failed to mention the illegal and unconstitutional obfuscation engaged in by this lawless court.
When the leader of the Liberal Party (PL in Brazil, and Bolsonaro’s party), Valdemar Costa Neto demonstrated irregularities in the machine tabulations and inconsistencies, relating to the different machine production provenances we have discussed here as well, Moraes fined the party 22mm Brazilian Reais ($4-5mm USD) and pronounced this elder statesman of Brazilian party politics would be put under criminal investigation in the infamous “fake news criminal inquiry” for what they term “assault on democratic institutions.” This is the same ongoing (and never-ending) criminal investigation used to censor and arrest journalists, elected legislators from the PL Party, and now protestors. It was also what was cited to justify the illegal detention and interrogation of this correspondent and my traveling party in September 2021. I wrote about this in detail in September of this year ahead of the first round of the election.
BENFORD’S LAW IN ELECTORAL MATTERS
There is an academic discussion in technical-scientific environments about the validity of applying Benford’s Law on electoral data. Professor Walter Mebane, renowned researcher at the University of Michigan, went deep into this issue and concluded that it is perfectly possible and feasible to apply Benford’s Law – or the Law of Anomalous Numbers – on electoral data. He has already published many works on this subject, stating that yes, there can be such an application, and that his results are very consistent when applied to the second digit of the numerical series of votes. A simple Google search will return reams of academic statistical study on this subject.
In the Brazilian case, the Court of Accounts and Audits (TCU) states on its website that yes, it is possible to apply Benford’s Law on public works data, and cites as an example a work by Professor Mebane on electoral data from the elections in Iran. There is no doubt, therefore, about the usefulness of this technical instrument for data integrity analysis.
‘NO AD’ subscription for CDM! Sign up here and support real investigative journalism and help save the republic!‘
In the case of the study carried out by our forensic scientist team we gleaned that the data was organized by electoral zones and grouped by regions, and later divided into 2 groups: new electronic voting machines and old electronic voting machines (within the same territorial base).
The result is that the index of inconsistencies in old electronic voting machines is much higher than in new electronic voting machines. Coincidentally, the official audit of one of the Brazilian political parties indicated that in the old electronic voting machines (prior to 2020) it was impossible to certify the veracity of the data, as they do not have digital certification within international compliance standards. By an incredible coincidence, the candidate Lula only won in these polls. But in the new electronic voting machines (2020 onwards), which have full and rigorous certification, the winner was the incumbent president of Brazil, President Bolsonaro.
The simple fact that these scientists cannot show their faces, and cannot publish their data under direct threat of arrest and heavy fines, is already a violent attack against the democratic process, as Judge Alexandre de Moraes conducts an interminable inquiry for more than two years, during which he has been handing out prison sentences and fines without due process and sans constitutionally mandated trials having been completed, without people presenting their evidence, and many times those accused don’t even know what are the actual accusations against them. Moraes recently ordered the arrest of a journalist who was beaten and as a result has suffered lower limb paralysis (Oswaldo Eustaqio). To crown this escalation of judicial chicanery and crimes against Brazilian constitutional democracy, last week Moraes ordered the arrest of an indigenous tribal leader of the Xavante tribe as well as a leading protestor and pastor, Fabiano Olivera, on the alleged charges of criticizing the electoral court and questioning the digital election results.
Brazil no longer lives under the Rule of Law, and its democracy has been destroyed by judicial gangsters who have taken over the Brazilian Supreme Court, eviscerated the constitution and eradicated due process, fixed the election, and obfuscated the process of proving the election fix. This leaves the only constitutional remedy- Article 142 of the Brazilian Constitution which is tantamount to martial law. If it is not invoked post haste, the electoral court taken physically to force constitutional compliance with the military audit, the machines procured and source code/tabulation data examined, the judges of these two courts (TSE and STF) arrested for their overt in plain sight numerous crimes, and elections fairly held or audited results presented to meet the expectation and approval of the citizenry, then we are afraid to say….Brazil will become Venezuela exceedingly quickly. Which is Lula’s expressed goal since he co-founded the incubator for Latin American Marxists, the Sao Paulo Forum, in 1990 with….drum roll please….Fidel Castro.
See here the graphs with samples taken from the Northeast Region (Lula vote) and the Southeast Region (Bolsonaro vote), comparing new and old electronic voting machines: