|
Edited on Wed Nov-10-04 05:41 PM by Petrodollar Warfare
..but who "submitted" them I do not know. That's why we need an INFOSEC forensic investigation.
"...how do you know the data was fictional."
I don't. But as an empericist I do know that the data is NOT based on the laws of mathematics. I think Spock once said that "..once you eliminate the impossible, whatever remains, however improbable, is in fact possible." Fraud on these auditless e-Voting machines is more than possible, it is practically invited.
This analysis supports fraudulent reporting of exit poll data, and "electioneering" is certianly nothing new in this country. Like I said, exit poll methodology is a mature science, but the level of sophistication in e-Voting is much more steathy due to the intrinsic technology - it's bytes over a wire. It's like sending your vote via an email to Diebold or ES&S, who then announces the results of the "election," but refuses to offer any type of paper audit or reconcillation process due to "trade secrets." Their response is nothing less than: "Its secret/proprietary, but you can trust us."
I don't trust the machine count because multiple exit polls in 6 swing states showed a different outcome well outside the MOE, and High School math refutes the possability that the AP's reported exit polls reflect the 4th/final exit polls based on sample sizes with the same gender composition as the earlier polls, and similar composition of "race." This is not rocket science, just observable facts and simple math. As for your next point:
"It's reweighting a base that's virtually identical."
Well, you now have the data, "re-weight" as you please. But given my review of the gender data, and race data as found on CNN, neither can account for such a significant shift in the precentages. I know that exit polling weighting is done on race and to a much lessor degree gender, but remember - the gender split stayed the same in both samples, so did the white male percentage (40%), and the white female percent went up one notch from 45 to 46%. The data sets relative to the sample size are essentially the same.
Unless you know of a "3rd gender" of some sort, you will not be able to construct a data set that matches what was reported. Good luck.
|