Reply #35: The Odds: 1 in 110,000

[tinfoil_beret]

I made some quick calculations based on a small sample of previous elections. As my sample, I chose the Florida presidential elections of 1984, 1988 and 1992. I chose these elections because in all of these a Republican had won the previous election.

In 1984, Ronald Reagan ran for his second term. As I recall, Reagan had strong approval ratings and won the Florida contest against Democratic candidate Walter Mondale easily.

In 1988, Ronald Reagan's V.P., George H.W. Bush ran for his first term. George Bush won in Florida against Democratic candidate Michael Dukakis.

In 1992, George H.W. Bush ran for his second term. Bush's approval ratings had fallen sharply due to the poor economy. Bush won Florida but lost his bid for a second term to Bill Clinton.

I excluded the 2004 Florida presidential election from my calculations because I believe the data is tainted.

First, I checked the sample of Florida counties with over 200,000 registered voters.

All Dem. Shift
-1% to 1% Ratio 1 in
Voters > 200K 42 8 19.05% 5.25
Voters > 500K 21 4 19.05% 5.25
Voters 200K to 500K 21 4 19.05% 5.25

You can see that the ratio is consistent among all ranges, at 19.05%.

Next, I calculated the probability that different numbers of counties would have Democratic support within plus or minus 1% of the Democratic support in the previous election.

Number
of
Counties Ratio 1 in
-------- ----- ----
1 0.19 5.25
2 0.04 27.56
3 0.0069107 144.70
4 0.0013163 759.69
5 0.0002507 3988.38
6 0.0000478 20938.99
7 0.0000091 109929.72

In 2004, as shown in the original post of this thread, Florida had 7 out of 14 counties with more than 200K registered voters and Democratic support within plus or minus one percent.

This puts the odds that Florida had a legitimate election at 1 in ~110,000.

Of course, when you consider all of the factors involved in this election, the odds are probably much slimmer.