Some algebra of the Ayyadurai Vote Video

This blog is for those specialists who’ve been trying to understand what Ayyadurai is saying in his video. The algebra of the Ayyadurai scatter diagrams will be laid out. My earlier blog was simplified.
To simplify, I’ll avoid subscripts and superscripts. Every variable should really have a subscript for precinct, like p, where p = 1…N, the number of precincts in a county. Consider a single county, so as to avoid notating counties.

ALthough I think that Ayyadurai’s analysis falls far short of locking in on the truth of the vote, I think his contribution is worthwhile in getting us to think about how to detect election fraud that’s hidden on machines. If we’re lucky, we may actually find out something about this if seizing servers leads to a real inquiry.

Let T(p) = total number of votes cast in precinct p of this given county. Write it as T to simplify, omitting (p). It also omits the county subscript, and county influences T(p) too.

T is itself a variable that depends upon the number of potential and actual voters, the number of ballots issued, and ballot frauds, and counting accuracy, and types of ballots issued (mail-in, absentee, polling place, etc.). But we’ll simplify and consider T as the result of counting votes for Democrat D and Republican R. Third party candidates are omitted.

The realized count of votes for D and R is further divided into straight ticket (S) and individual (I). T has 4 components: T = SR + IR + SD + ID, where SR = straight Republican, IR = individual Republican, SD = straight Democratic, and ID = individual Democratic.

From these counts, percents are defined by Ayyadurai. One problem is why he defines them as he does, or how his theory of these proceeds. There is first the proportion of all S votes that are SR votes. This is X, and X = SR/(SR +SD). Multiplication of this by 100 converts it to a percent. Skip that step and notation to simplify. Why we should be interested in X is unclear. If SD = 0, then X = 100%, which means within the straight ticket voting that went to the two parties, Republicans secured all such straight-ticket votes. I can think of a variety of explanations or models of why this might or might not occur, or why X might vary from 0% to 100%. Without a theory, we are at sea. Ayyadurai’s analysis is lacking a vital component, the theory he’s using.

In the attempt to understand his work, I will define an analogous variable for the individual votes (not straight ticket votes), Y = IR/(IR + ID).

Bear in mind that TR = SR + IR, not percents, and TD = SD +ID. Also, TR + TD = T.

TR/T = proportion of all votes that Republicans got. Changing to percents, which is no problem or issue, 100- TR = TD/T = the percent of all votes that Democrats got.

At 28m 10s of his video, there is a slide with a numerical example. T=2,000. TR= 700, TD = 1,050. TR/T = 0.35 or 35%, and TD/T = 0.65 or 65%. The slide shows “45% Votes from Republican S.P.) There were 1,000 S.P. (straight party) votes, broken down as SR = 450 and SD = 550. This 45% number is gotten from 450/1000, i.e., it’s precisely my definition of X = SR/(SR+SD) given above. The precincts are ranked by this X variable, and it’s shown as a horizontal red line starting from 0%. This is what Ayyadurai calls %RSP votes.

It’s important to ask the author for an interpretation of this particular X-variable. Why should we focus on the straight ticket R vote as a percent of all straight ticket votes of the two parties? What’s the theory? Why bring in the D straight ticket voting at all here? If we want a measure of the Republican straight ticket voting, why not just use SR/TR? In this case that’s 450/700 = 0.6429. However, that is not what Shiva wants. He wants individual voting to compete against a straight ticket benchmark. The implicit theory is that individual R voting (as a percent of all individual voting) should keep up with straight ticket R voting as a percent of all straight ticket voting, and if it does not, then he infers an algorithm run by invisible hands and brains is shifting votes.

What is the Y variable? At 20m 09s of his video there’s a slide and we read that the Y-axis is the difference of % Trump Individual Candidate Votes minus %RSP votes. The question raised about the use of %RSP instead of just plain SR/TR crops up again. Exactly what is Shiva theorizing about implicitly?

His excess Y variable is now going to involve two variables, each of which has brought in Democratic Party votes. I defined Y = IR/(IR+ID) in keeping with his definition in this slide. He is now defining what we can call an Excess Y variable as Y – X, both definitions being his and both being given exactly above.

Hence, his excess Y = [IR/(IR+ID)] – [SR/(SR + SD)] = Y-X.

It is this excess Y that is his Y-X variable. Using his Oakland County example at 28m 10s, we have excess Y = 25% – 45% = -20%.

With X being assumed as a benchmark for Y, and with X being ranked from low to high, the excess Y variable goes down as we move left to right along the X axis, ceteris paribus. The derivative of excess Y with respect to X is dY/dX -1.

The problem faced by Shiva that goes unaddressed is that there is no theory of Y and X and therefore no theory of excess Y = Y-X. If Y is simply a normally distributed variable with mean E(Y), then observations of Y will be distributed randomly around E(Y). The excess Y will be heavily influenced by the fact that the precincts are being ranked on the realized values of X from low to high. Excess Y will tend to decline with X. But this is only one among many possible scenarios. How can they be distinguished? I conclude that it’s extremely premature to start talking about votes being shifted and mysterious algorithms being executed to do the shifting. If, on the other hand, we ever actually get behind the veils of secrecy that hide the voting mechanics, if seized servers ever amount to something meaningful, or vote recounts are ever done honestly, etc. etc., then we might have more solid a basis for understanding such voting phenomena as straight ticket versus individual voting. And actually I’d hope that political scientists have been studying this kind of thing for many years and can weigh in with their insights and findings.

I know of no strong reason to expect one party’s proportion of all individual votes to be strongly related to that party’s proportion of all straight ticket votes. In fact, it could be that a greater intensity of straight ticket votes of one party is accompanied by a much less intense individual vote proportion, because the true believers have more or less exhausted much of their fire by voting the ticket. I emphasize that I do not know what’s going on with these variables and I have no theory.

Beyond this topic, I note that the relations between good truthful and honest science and politicized science is tenuous. Climate science and medical science provide two recent and strong examples. Science applied to voting is another. Even when we understand what a fair vote entails scientifically, we do not implement it. Even when we understand the true costs of lockdowns, the power structures override our knowledge. Our experiences in many areas tells me that it is a big mistake to place technocrats in charge of law-making and regulation-making. Progressivism depends in large part in the assumption that experts know better than we do what’s good for us. Sure they do. Look at how wonderfully the CDC has handled a new virus. Look at how well the smart CIA people handled Afghanistan. First they created a mujahideen force led by Osama bin Laden. When it became clear to these geniuses that bin Laden hated America and was mobilizing terror attacks against us under the title of Al-Qaeda, these intelligence technocrats decided to fight a war on terror against bin Laden. That and their attack on Iraq produced a new outfit known as ISIS, and then their brilliant idea was to go to war in Iraq and Syria against that group. All the time these CIA technocrats kept screaming that they were protecting national security. It was they who couldn’t even coordinate their intelligence with the FBI around the 9/11 tragedy.

Share

10:53 am on November 15, 2020