Absolute mathematical proof - Step 4a
This is different from the previous work. The following argument is mathematically stronger than the previous ones by several orders of magnitude.
Counting votes in a ballot is like taking marbles fron an urn until you've got them all out. Take an urn with 10 marbles, 5 reds 5 blues. You draw the 1st: probability if it being red? 50% Let' say it is red. Pick the second. Probability of being red? 4 reds left in there, so you know it is now 4 chances out of 9.
Now put a 10,000 marbles in there, 1/2 reds 1/2 blues. Pick one. Now make a chart of the cumulative % of the drawn balls that are red. Your first data point on the chart will be 0% or 100%. Draw the second ball. The 2nd data point will be 100% (drawn 2 reds), 50% (1 red, 1 blue) or 0% (no reds). As you know that your final result is 50%, the line will oscillate up and down randomly but rapidly converge towards 50%, something like this:
The mathematical law describing this process is call hypergeometric. It describes in particular the statistics of partial ballot counting! And if you tell me that Romney has got 50% of the votes in that ballot, I can tell you that he needs to be real close that 50% by the time we have counted 90% of the votes, a bit less close at 80%, etc... Poll science shows that 10/20% might suffice for the oscillator to turn into a complete flat line. But maths allow another nice trick as well. If you tell me after 25% of the votes counted what Romney's score is, I can actually tell you the probability that he will get to a score of 50% at the end of the vote tally. Isn't hypergeometry nifty?
As we have seen, Romney's lines do not converge flatly towards the final result. It does not everytime the ballot is relevent to the final outcome. His score starts by oscillating and flattening but then shoots up in a straight line in dozens of counties. That is mathematically impossible.
Let's look at the numbers of Allendale County, SC. 311 votes.
Here is a chart you are now familiar with: Candidate score vs Cumulative votes sorted by ascending vote tally
Normal patterns, nothing special apparently. Now the hypergeometric distribution allows me to caluclate the following probabilities:
What does it say? Take Santorum. His final score is 12%. Don't forget: this is not an estimate, it is actual. His cumulative score MUST get there. People in Fairfax No 1 are known for their superior political wisdom. Santorum gets no vote. So 23 votes have been cast. Santorum has only 311-23 votes left to go from his current 0% to 12%. Probabilitity of him getting 0% out of 23 votes cast? Hypergeometry says 5%. But he does very well in the Martin precinct (guys, get a grip, his a fake). His cumulated share of votes goes back to 10%, close to his final score of 12%. The oscillator has sent him close to the final target and the probability of that is high: now 39%. Etc... Gingrich does well early, so he then needs very little to reach his final 52%. The number reflect that. Paul was looking good all the way but is trashed in Fairfax No 2. His odds from reaching the final 7% fall suddenly from 77% to 7%. Romney needed a big last precinct to end up 29%, which he got.
So expect big volatility in the numbers as candidates outperform/underperform locally and oscillate before reaching their final score.
If we plot the last table, we obtain that cute doodle:
X axis cumulative vote, Y axis Probability of having the score X or lower.
When a Y value is at 50%, it mean that the oscillator is currently aligned with the final score. So expect to go across that line frequently. In a real world.
Ok, more example of untampered counties.
Now. Think, Winnie, think. What does it mean to be at 0%. It means that your interim cumulative score is so low that you have no chance to get back up to the final result. On the charts which will follow, 0% is always something like 0.4%, or 0.002%, or 0.0000000000000000001%. Take this one.
VOTE FLIPPER ON ALERT !!!
How do I know? Notice a difference with the previous charts? I guessed so. What are the odds when Romney is at 27% with 23% of the votes counted to end up where he did, at 32%? Hypergeometry, a law of the universe, says: 1 out of 267,385,153.
Checkmate.
Copy, paste, disseminate. Debunk, but good luck with that one. You'll need it.
(To be continued in Step 4b)
. Get on chat and arrange this.
