Thread by @ThyConsigliori

21 October #BetweenTheLinesDotVote Analysis

Analysis Works! Check This Out!!

I started this analysis a few days ago, and knew I'd gotten my hands on something, but wasn't quite sure what. Now I know, and show in the chart below. They call this their Periodicity Problem.

2) It takes mental discipline to read this chart. I call it the Go On 3 problem. Do we go the instant we hear 3? Or, do we wait until the word "three" is completed? Do you count from the start of something, or from its end? Believe it or not, its actually an important problem.

3) To read this chart, you need to understand some fundamental information. 1, 2, 3, 4, 5, 6, 7! That's one week. Seven days. And, it marks roughly one half of Dornsife's respondents. I call them Group 1. And I denote their thoughts at the END of the week.

4) I do the same, obviously, for Group 2, the 2nd half of the Dornsife sample. On this chart, there are 9 weeks' data. In EVERY SINGLE INSTANCE Group 1 resists Trump, Group 2 supports him. Look and see. It's as predictable as if it were a law.

5) A review is needed. Back in 2016, Dornsife cycled through its entire sample every single week. Thus, there was NO Periodicity Problem! What does this mean? It means they have utterly failed the RANDOMIZATION requirement for their poll. It is NOT RANDOMIZED.

THIS IS HUGE.

6) I wont bore you, I promise. Random sampling is simply the sine qua non, the ultimate mandate of scientific analysis. If you allow your sample to be biased, you have failed. I know whereof I speak. My own BetweenTheLines.Vote sampe is still NOT randomized.

7) I keep intending to present my data, but as I assess it, I can't get over my fear of a NON-RANDOMIZED sample. I fight this with all I am. And, that makes me hold back. I fear my sample is BIASED, not RANDOMIZED. A biased sample is simply of little value.

8) I don't know - I think they do... - if the Dornsife people follow my analysis. If they do, they MUST do something about this problem, far more than they have done so far.

Let's come at this another way.

9) The data is NOT supposed to be predictable. If it is predictable, then why do we need it? We make a simple prediction. We're right. All done. The sun will rise tomorrow, just like it did today. This is not a matter of question. No polling will help us understand this.

10) Let's go deeper now. To my eye, as you know, Dornsife was the best out there. They have surrendered their right to title. They did it to themselves, but my theory is that they did it in mere compliance with their finance source, the LA Times. Biden will win. Period.

11) This is NOT science. It is the abrogation, the treason to science. No outside analyst such as myself should be able to predict outcomes which therefore FAIL to predict America's will. They have surrendered the torch of their mission. Shame on them. Shame on you, Dornsife.

12) Can we therefore predict a Trump victory? I have. But my qualifier is that much as I tried, I NEVER found a source of data - or built one - adequate to my own scientific standards. So, my call reflects my desires NOT as supported by data. I hate that more than words can say.

13) But there may be a way for me to relax a little. If the best in the business HAS TO falsify their own data, then maybe that tells us just how powerful the MAGA Movement really is. If they showed the true data, Trump's easy victory would be obvious, right there.

14) Believe it or not, this is this morning's short thread. I have another analysis to pick up on, first posted yesterday, here:

https://twitter.com/ThyConsigliori/status/1318437521336958977?s=20

15) If I can get that second part completed today, then, I plan on a full chart update tomorrow. I suspect we're seeing the implosion of Biden's run over his criminal and treasonous actions, now coming out. But, no need to jump the gun. We'll see, soon enough.

Thread ends at #15.