Sorry to burst your bubble...
Strangely, the Ukraine election is affected by the same problems as the US election: voting irregularities and major discrepancies between official results and exit polls, almost always in favor of the official winner. Everyone is seeing that the Ukraine election was stolen. I have yet to see a plausible explanation other than vote-rigging for the irregularities and discrepancies with regards to the US election.
My concerns that the US election was stolen came after studying this statistical analysis of the Florida election results:
Per county detailed figures of voter registration, voter turnout and official results.
and:
Statistical Analysis of the difference between actual turnout and voter registration
The problem of Florida ballots is *not* that more Republicans turned out to vote for Bush than Democrats turned out to vote for Kerry. In short, the main finding of this study is that the additional Republican turnout, compared to expected turnout as per voter registration, was in average significantly higher in counties with optical-scan voting than in counties with E-touch voting. In other words, there is a significant statistical difference between E-touch counties and Optical-scan counties. So far, in my opinion, no sensible explanation other than vote-rigging was proposed to account for this rather strange difference.
Let's take for example this article of the Miami Herald: it says that "No flaw is found in Bush's state win". What the reporters did was to manually recount the ballots in three counties: Union, Suwannee and Lafayette. Well, does their article indeed contradict the findings of the study above? Definitely not. A more honest title for this article should have been "No flaw is found in Bush's state win *so far*".
Please notice that these three counties are all optical-scan counties with a maximum of 21,930 registered voters and therefore are not part of the statistical study cited above, which is carried out only on counties with between 80,000 and 500,000 registered voters. Therefore, the only way to prove the study right or wrong would be to find or not find major discrepancies between manual recounts and official results in a large number of counties, much more than the three small counties that were chosen by the Miami Herald. Maybe the vote-rigging did take place only in counties large enough for it to go unnoticed?
But there is more: why did the Miami Herald reporters stop counting ballots at "almost 60%" in Suwannee? Because they were in a hurry to go for lunch? Could it be because they knew that the final recount would only "nearly" match the official results? In fact, you can argue that their recount at almost 60% does not really "nearly match" the official tally.
First, the undisputable figures.
The official tally is the following: Bush 11,145 Kerry 4,513 Total votes 15,785
This gives percentages of 70.60% for Bush, 28.59% for Kerry and 0.80% for other candidates. The percentage of registered Democrats is 63.6%. Remember however that the problem is not that a Democrat county would swing Republican, the problem is that it would swing more Republican than if it were E-touch, OK?
The recount done by the Miami Herald is: Bush 6,140 Kerry 2,984
The total number of votes in the recount is not known. For the sake of accuracy, let's assume that, on this recount, there were also 0.80% of the votes for other candidates. This would give a total number of recounted ballots of about 9198, with 74 votes for other candidates. We can see that the Miami Herald recounted about 58.3%, indeed almost 60%, of the whole county. Our computation is quite accurate as the number of votes for other candidates is only 127 on the whole county, therefore not significant.
The recount gives percentages of 66.75% for Bush, 32.44% for Kerry and 0.80% for other candidates. Does this nearly match the official result? Well, yes in the sense that the difference between the partial recount and the final resultats is only -3.85 for Bush and +3.85 for Kerry. And no, because this is exactly the kind of difference that would swing the election for Bush if arising from a state-wide vote-rigging and that we should therefore be carefully looking for.
Now, this calculation is evidently *not* a proof that the election was rigged in Suwannee County. Maybe that, if the remaining ballots were also recounted, the totals would eventually match the official tally, as they did in Union and Lafayette.
What should the recounts yield on the remaining ballots? I see also a potential problem here.
On the remaining 6,587 ballots, uncounted by the Miami Herald, there should be 11,145 - 6,140 = 5,005 votes for Bush and 4,513 - 2,984 = 1,529 for Kerry. The percentages over the remaining ballots should be 75.98% for Bush and 23.21% for Kerry.
What are the odds of 66.75% for Bush over 58% of the ballots turn into almost 76% on the remaining 42%? I'm no statistician but I think that they are pretty faint, unless there were a significant statistical difference between the recounted and not-recounted ballots.
So, what does this article prove? Not much if you ask me. Only that Union and Lafayette counties (total 8,000 votes) did accurate counts of ballots and were not rigged. Mr. Bush, you will need much more than that to convince us that the election was fair.