|
I have posted this a few times already. But in light of today's Diebold bombshells, lets do it again.
Thanks to Bev Harris and others, we know that there was the possibility of a back door hack of the voting machine computing programs. They had the means, they had the motive, they had the opportunity. Especially in Georgia. But did they do it? Can we determine, based on the final polling and election results, the probability that the machines were hacked? No proof, just probability….
Of course we can. Take off your tinfoil hats and put on your thinking caps…
This is a 2002 Election Probability Analysis. The probabilities cover all cases, assuming 8 to 34 CRITICAL SENATE elections. The mathematical model used was the Cumulative Binomial Distribution.
Using the the published SCOOP analysis of the 2002 election results, and the fact that an improbably high number fell outside the latest polling margin off error, I prepared a probability analysis to determine the likelihood that 4 out of 8 critical elections would fall outside the 3% margin of error (MOE). In other words, there is a 95% (19/20) probability that each poll would turn out to be correct within the MOE. Thus, the chances of a given election falling OUTSIDE of the MOE is 1 out of 20 (5%).
Here are the results for 8 critical Senate elections: ..........Last Poll %...........Election % Result ........Dem.....Rep.....Diff....Dem.....Rep.....Diff...Chg OutMOE NC......42......48......-6......45......54......-9.......-3...0 MN......47......39.......8......47......50......-3.......11...5 AK......51......46.......5......54......46.......8........3...0 GA......49......44.......5......46......53......-7.......12...6 TX......48......49......-1......43......55......-12.....-11...5 MO......46......46.......0......49......50......-1.......-1...0 NH......46......40.......6......47......51......-4.......10...4 CO......45......47......-2......45......51......-6.......-4...0
The Red flag states: MN, TX, GA, NH (all exceeded the MOE)
I maintain that only 8 were really critical and hotly contested (the ones that the GOP focused on). All others were one-sided for the Dem or GOP incumbents. These should NOT be included in the analysis, but I have done so to determine the hypothetical results, which showed that 4 out of 34 elections falling outside the MOE is also unlikely to happen. Now 4 out of 8 CRITICAL elections fell outside the latest poll Margin of Error (MOE), a very weird (and rare) occurrence, indeed. So, let's see the way the probabilities are effected as we INCREASE the number of CRITICAL states in our sample. The probability that AT LEAST x out of the N CRITICAL senate elections will be outside the MOE (+ or - 3%) is:
Total sample Probability Matrix Elections Elections outside MOE ...........3...............4..............5..............6 8 0.005788 0.000371751 1.54049E-05 4.0082E-07
10 0.011504 0.001028498 6.36898E-05 2.75458E-06 13 0.024508 0.003102996 0.000286569 1.97497E-05 16 0.042938 0.007003908 0.000857312 8.08995E-05 20 0.075484 0.015901526 0.00257394 0.000329294 25 0.127106 0.034090601 0.007164948 0.001212961 30 0.187821 0.060771561 0.01563551 0.003282486 34 0.240651 0.088128675 0.02591563 0.006269405
From the table: The probability of 4 out of 8 elections falling outside the MOE is .000371, or 1 out of 2695.
Now, if ALL 34 elections were hotly contested and critical (obviously, not the case) then the probability that 4 out of 34 would be outside the MOE is .0881, or 1 out of 11, still unlikely.
But we are not done yet... Since ALL four elections fell for the Republicans, the final odds are 1/2*1/2*1/2*1/2*1/2695, or 1 out of 43,120.
Yes, Virginia, it probably happened. Big time.
What we have is a mathematical confirmation of an intuitive feeling. Something very strange happened in 2002. The election results seemed out of whack. Zogby himself said he was NEVER that far off. It makes you wonder. VNS exit polls were not published. And there were 22,000 Diebold computers installed in Georgia, which we now know were patched the weekend before the election.
|