I wanted to represent CVR data in a way that helps people who are not mathematically inclined to visualize the data.
This is a single county of CVR data. Think of this as having 2,000+ people waiting in line to vote with either a RED shirt on or a BLUE shirt indicating how they are going to vote.
RED is Trump. BLUE is Biden.
Now, let’s say we start with the first 100 people and you count the number of RED and BLUE shirts. Then you go to person 2 through 101, then person 3-102 and so on until you get to the back of the line. Each time, you record the number of RED and BLUE shirts in that group of 100 people.
As you look at 100 people at a time, you would expect the number of RED and BLUE shirts to behave according to the law of large numbers.
Meaning this in my own words:
The count of RED and BLUE shirts in each successive group of 100 should be about the same. The reason for that is because the order in which people vote is random human behavior.
Now look at the graph below.
You can see that up until person 375 (the x axis value) that appears to be the case. The ratio is hovering a bit above 50 for RED and a bit below 50 for BLUE.
One could assume this is a slightly more RED county. Trump will win. (But, Trump ends up losing in fact.)
Then the data gets weird.
Now look at what happens at say person ~520. Between person 520-620 we have a count of about 27 RED shirts compared to 73 BLUE shirts.
Meaning this: The BLUE voters all decide to go to the polls at the same time and clump together in-line. This doesn’t make sense.
Random data would not have 73 BLUE votes clump together in the voting que.
There are other wild swings in the data.
Look at person ~1360. We count 32 RED shirts and 68 BLUE shirts between person 1360-1460. This doesn’t make sense.
Then look at the big picture, The RED line has an overall downward trend.
Meaning this:
For these 2000+ people waiting in line, you find less and less RED shirts on average waiting in the line the further towards the back of the line you go.
In fact, the last 100 you count 24 RED shirts and 76 BLUE shirts!
This is not what random human behavior looks like.
It is also the reason why obtaining the CVR record is so important. It shows the fraud.
The take-away is this.
The entire graph should look like the first part between person 1 -200. The rest does not follow the Law of Large Numbers. Here is an example.
So why does my RED-BLUE chart actually look this way?
You have one of two choices to pick from:
People in this county coordinate their voting behavior so that it can no longer be considered random data and does not follow the Law of Large Numbers.
This is not random data which indicates a voting machine algorithm is being used to cast votes for the We The People.
Does that help folks?
I think it's really interesting how the two lines are almost exact mirror images
I love this analysis...its very illuminating and adds yet another data analytics point to the support of nonuniform, aberrant anomalies. It would be helpful to have a baseline for folks to see even clearer...for instance, a county or another year (same county) that shows normal, random behavior where we know there were no shenanigans.
I also loves this bc it shows that they injected large #s to make the aggregate competitive while later on they trickled in showing that they were trying to cover it up after they won the aggregate total count. Great job. This would be great to have for all the counties over successive election years, but that's a big ask.