Debunking the MITRE Report

0
280

A team of MITRE data scientists calling themselves the “Bald Eagle” produced a report over here which provided the following conclusion: In summary, multiple types of analysis found no evidence of fraud, manipulation, or uncorrected error in the eight states included in this research.

We had a quick glance of the report and would like to point out a significant flaw in the methodology which renders the conclusion to be inconclusive.

The Gateway Pundit also argues the case over here, but their argument is not quantitative as we are.

The primary issue can be framed with this simple question.

“Suppose you are in charge to validate a consolidation process which involves thousands of data points, and someone tells you that they have discovered an error. What is the first thing that any one with any experience in data analytics will do?”

The answer is simple:

You will try to create a method to detect the error which had been surfaced and then apply that same method to find out all such errors which follow the same pattern.

The reason is that the entire downstream analytics relies on the principle that the count as reported by the voting machines is authentic. If there is an issue with the count, then the downstream analytics will not be able to generalize that the entire process was sound, because the base data, namely the count itself, was suspect.

In the case of the US Election, the most obvious filter is the negative incremental vote count which we have reported over here. This signal capture mechanism works by checking against each timestamp the running total of the votes received by each candidate for a particular state and isolating cases where the incremental difference is negative.

It is so basic, so intuitive and not too difficult to estimate that cannot be debunked by anyone else because it is even obvious in the New York Times json file over here.

The validity of this method was justified because it detected the case of the vote switch in  Antrim county, which was later found out to be  a suspected case of electoral fraud. The follow on question will be how many other counties followed the same filter.

Well, we have already provided even video evidence of screen grabs that this happened in Pennsylvania at 04:08 AM over here:

It was quite surprising that if a small-time kaki lima  (‘street vendor’) outfit like ours can flag this out, but it was missed by the great Mitre.  This form of signal analysis was not explored in the report and our primary criticism is that if you cannot have a process to detect known errors, how can you make general statement about the entire process as all?

Lets be clear one thing: If the Rembau Times puts a hard number out there there can be no doubt on the accuracy of the number. None at all, because we argue with hard facts not narratives.

If you want to detect small changes, you cannot use a measuring tape, you must use a micrometer! And Biden beat Trump in most swing states by less than 1%.

Mitre examined voting patterns at the aggregate county level, which is not sufficiently granular enough to detect voting irregularities in a competitive 2 horse race. More pertinently, the analysis used cannot debunk My-Pillow guy’s analysis because My-Pillow guy is arguing based on time-stamped indexed data and not aggregated data. If you argue based on time-stamped indexed data, where you allege that the vote count was changed by a specific IP address at a specific point in time, the easiest way to debunk this is to

  1. Find out the vote count just before the timestamp, at the timestamp and just after the timestamp
  2. Debunk (1) if the allegation does not stand up.

This is so simple, so point-in-time that it begs the belief why the Mitre did not use this? Why go into statistical distribution tests when the argument is not that the distribution was skewed but that a vote count a particular point in time was changed?

After all it is not that Biden got 90% and Trump got 10% and any method which seeks to find a signal indicating as such will suffer from Type 1 error , namely failing to reject the Hypothesis of No Election Fraud because the signal is not large enough to be detected. Methods to detect voter fraud , where the winning margin is greater than 5% cannot be applied to the United States where the winning margin is smaller than 1%.

Even the talking dog at the Shell Shack can tell you that!

We thus have debunked the Mitre report under the category of “Improper Modelling Techniques.” Interestingly, the Mitre report made this conclusion:

This conclusion is consistent with Georgia’s risk limit audit (“the hand recount”) and machine recount, and with a joint statement from the elections infrastructure government coordinating council and the election infrastructure sector coordinating executive committees.

Is there any Audit Principle that support this conclusion?

Usually, when there is any suspicion of fraud, the principle is to conduct a full audit.

Perhaps, the Bald Eagle team knows better.

But we are not calling for anything neither advocating for anything, just exercising our First Amendment right to debunk, but providing a sound basis to do so. You make your own conclusions, whether to believe Mitre of The Rembau Times.

As we discuss, Chairman Joe is the Chairman of the United States and that is the fact. As we also discuss, we now embrace his Chairman-ship, stutters and all.

Why we like Chairman Joe

Now we have come along way from our initial Trump supporting ways prior to the election. We now also know that there existed a Cabal who intended to get the “desired outcome”, which was disclosed by none other than Time Magazine.

Currently, our interest is in the continuation of Chairman Joe’s rule.

We see great benefit as it reveals the hypocrisy of the Wokism, Fox News, the media and the Cabal members of the Republican party, which on balance , provides greater benefit to the world at large than Trump remaining in the White House.

Don’t talk about MAGA, if the Kentucky guy is still a Senator and the Cheney girl is a representative and the Brad guy is a Secretary of State.

All these folk should be recalled as soon as possible.

If there is a choice between Trump remaining in the White House and Disney going bankrupt, we will take the latter option every single time.

Our only hope is that the Chairman will feature more pictures of his beautiful German Shepherds.

There is also an element of morality that we would like to address, and it goes with the current theme of the day, which is Racism.

Our position is clear: Any person who hires Gina Haspell or the Wray-fella to head the CIA and the FBI respectively deserves everything that came their way. Trump hired based on race, and those who fit the profile stabbed him in the back. Interestingly, one of the few African Americans of any influence during his Presidency, the Hon. Supreme Court Justice Clarence Thomas was one of the few voices out there, both Black and White, who remained steadfast in the belief that a proper investigation was warranted.

 

Perhaps Trump should not have practiced such race-based hire approaches,  but heck, I am not the guy who got the most votes in Presidential history and still lost the race. Maybe he knows better.

More pointedly, several outlets are reporting that Trump is now groveling to be reinstated on Facebook.

The same Facebook, whose owner poured hundreds of millions of dollars into funding Black Lives Matter, and in securing the election for the “desired outcome.”

All we say is you want to beg,

go down on your knees.

 

De-Platform Today.

De-Platform Tomorrow.

De-Platform Forever

 


Warning: A non-numeric value encountered in /home/customer/www/rembautimes.com/public_html/wp-content/themes/Newspaper/includes/wp_booster/td_block.php on line 326

LEAVE A REPLY