THE FORMULAS
WHAT THEY MEAN AND WHY THEY MATTER
I derived four deterministic linear formulas from the ballot drop data:
Raman% = 27.87 (3.19 Ă drop number)
Pratt% = 22.05 - (1.17 Ă drop number)
Bass% = 40.20 - (1.86 Ă drop number)
Other% = 9.88 - (0.16 Ă drop number)
The Starting Points
The constants - 27.87, 22.05, 40.20, 9.88 - represent each candidateâs baseline. Where they genuinely stood when post-election counting began. Real votes. Organic support. These are the numbers before anything unusual occurs.
The Slopes - How the Formula Runs
The slope is the increment applied to each candidate every single drop. It gets multiplied by drop number - 1, 2, 3, 4 - which means the effect escalates automatically with each drop.
So for Raman it isnât just plus 3.19% every drop. Itâs:
Drop 1: 27.87 (3.19 Ă 1) = 31.06%
Drop 2: 27.87 (3.19 Ă 2) = 34.25%
Drop 3: 27.87 (3.19 Ă 3) = 37.44%
Drop 4: 27.87 (3.19 Ă 4) = 40.63%
Drop 5 predicted: 27.87 (3.19 Ă 5) = 43.82%
Each drop pushes her further from her baseline. It is built in. Automatic by design.
The slopes also sum to exactly zero.
3.19 - 1.17 - 1.86 - 0.16 = 0.00
Every percentage point Raman gains comes precisely from the other three candidates/groups combined. This is a closed system. Conservation of votes. The formula doesnât create votes - it redistributes them.
The R Values - The results were remarkable
To validate these formulas I ran linear regression analysis. The Pearson correlation coefficients, R values, came back as follows:
Raman vs Pratt: R = 0.9966
Raman vs Bass: R = 0.9934
Raman vs Batch: R = 0.9984
Raman vs Other: R = 0.9794
R values measure how perfectly data fits a straight line. They run from 0 to 1.
0 means completely random.
No pattern whatsoever.
1 means a perfect straight line.
Every point exactly where predicted.
For context:
0.70 is considered strong in social science research.
0.85 gets researchers excited.
0.90 is extraordinarily rare in human behavioral data.
Squaring them to get R² values:
Raman vs Pratt: R² = 0.9932
Raman vs Bass: R² = 0.9869
Raman vs Other: R² = 0.9592
Raman vs Batch: R² = 0.9968
That last number, 0.9968, means that 99.68% of Ramanâs vote share movement across these drops is explained by a single variable. Drop number. Nothing else. Just counting to five.
You get R values like that in physics experiments. In controlled laboratory conditions. Measuring the expansion of metal under heat. Not in elections. Not in a major American city with millions of diverse voters casting ballots across weeks.
The Slope Relationship
The slope of 3.110 between Raman and Pratt is particularly significant.
It means for every percentage point Pratt lost Raman gained 3.110 points. Every drop. Without variation. Without noise.
Candidates in a democratic election donât move in mathematical opposition to each other at a fixed ratio across 200,000 ballots. Four variables in an equation do.
Why The Shutoff Had To Exist
Because drop number keeps increasing, the formula keeps pushing percentages further from baseline. Left unchecked by drop 7 the math produces:
Raman: 27.87 (3.19 Ă 7) = 50.2%
Pratt: 22.05 - (1.17 Ă 7) = 13.9%
Bass: 40.20 - (1.86 Ă 7) = 27.2%
So the formula was never intended to run to completion. It was designed to run until a specific objective was achieved, Raman leading Pratt by a sufficient margin to secure the runoff and then stop.
The stopping condition appears to have been triggered at approximately 38,000 votes into drop 5. The precise moment Raman crossed 3,000 votes ahead of Pratt.
After that point the remaining 9,800 votes in drop 5 distributed at approximately the baseline that existed after drop four.
The formula completed its task. Then it stopped. And the numbers went back to looking normal.