SC vote counts and exit polls - how odd is this?

[seaclyr]

Exit polls in South Carolina showed Bush winning by 8% but final vote counts had a much higher margin of 18%. Vote count percentages, however, varied tremendously over election night and did not reach 18% until well into the evening. Taking a look at the numbers as they first came in, early percentages reported by cnn (first image) were strongly pro-Bush but they dropped off quickly so that by 8:16pm EST the Bush margin(%B-%K)was down to 8%, matching the exit polls at this point. Shortly after this, however, there was another big shift to Bush that continued through the evening, bringing the final margin up to 18%.




Differences in actual votes, rather than percentages, provide another take on this. Vote differences (B-K), i.e. votes for Bush minus votes for Kerry, were calculated from the cnn numbers reported over the evening (2nd image; the y-axis shows the difference in votes measured in 100,000s). It's clear that vote differences early on favored Bush but the increase was relatively slow and linear up until 8:16 - 8:24pm. If the slope of the line up until this point had been maintained, the final Bush margin would have been 119,000 votes (the number obtained by least-squares analysis of all points up to and including 8:24pm EST), or 7.9% of the total, which again would match the exit polls. Instead, the slope abruptly increased beyond this point as the vote difference climbed to 276,000 (i.e 18% of the total).




Does anyone know if this pattern is indeed odd - on the surface it seems a curious, quite abrupt shift from numbers reflecting a 8% Bush advantage to a 18% Bush advantage (the difference between exit polls and final vote count) - or might it simply reflect some idiosyncracy in SC regional patterns and voter preferences?

[StephanieMarie]

1. Was the hacking "real time" ?

This makes it look a lot like there were real time decisions being made as the day went on to hack more votes, rather than the theory that the machines were pre-programmed to, say, give every 10th Kerry vote to Bush. This chart looks like someone was watching and making the decision as the day went on. Very suspicious.

[Imagevision]

26. A clearer picture to what happened as the days continue on...!

[StephanieMarie]

2. Can you get this to TruthIsAll?

And ask if the pattern is repeated in other states as well? Real time hacking should be easier to discover, maybe, because phone calls would have to be made, or emails sent, from the central decision maker (Rove?) to the techies at each tabulator.

[BlueDog2u]

5. This appears to match, in general outline, the pattern seen

in the national totals as well. The pattern is the same" Early Bush lead tapers off, and then a big surge at the end.

It does look like real time manipulation.

[Pachamama]

10. Was wondering the same about other states and patterns...I was also

thinking we should have someone who has expertise in detecting patterns and numbers (think alomg the lines of code crackers)...

[VirtualTruth]

25. please send all info about this to me

Hi I am closely following all items that involve the vote count. Please send all relevant information to me.
Thanks

Time will tell all the Truth

VT

Sean Lewis [e-mail]

Founder [e-mail]

[newyawker99]

27. Hi VirtualTruth!!

Welcome to DU!! :toast:

[autorank]

3. Correlate the precincts reported with the timing of the vote shifts.

I agree that this looks odd and my first assumption is that there was something amiss with the results. If the timing of precincts reported does not correlate with the shifts in vote percentage for *, then there is clearly a problem. For example, if his spikes came when mostly white districts came in in a group, that might explain things. If there was an even distribution of reporting (white versus black precincts) and this spike occurred, I think you're onto something. This is very interesting. Thanks for sharing.

[berniew1]

4. Would absentees be added in late and do this?

absentees might have been heavy to Bush

[autorank]

13. That's one possibility. This is a great analysis in the ?'s it poses.

If there was nothing in the way of pro * precincts or voting groups (e.g., absentees) at the time of these shifts, this is really hot stuff. It's the type of thing that is going on all over the country by bright people who understand statistics. Great stuff. Who knows, berniew1 may have stumbled on a path to a smoking gun. It sure makes sense that this would be done in SC. The politics there rivals Chicago and Louisiana in its pure intensity.

[jbnow]

19. deleted...I answered on wrong state's data

[dzika]

6. On a scale of 1 to 10...

...that is very odd. Off the charts odd.

[StephanieMarie]

7. kick. n/t

[Pachamama]

8. What strikes me, amongst a lot of things, is the # 119,000 - isn't that

what * "supposedely" won Ohio by?

:eyes:

[KoKo]

9. SC Absentees have to be counted 7:00 p.m. Election Day...here's link to

site which lists when all states must count their absentee ballots. SC is on the list. Hope that helps you...

http://www.uselections.com/faq_absentee.htm#3

[rdmccur]

11. This is very interesting.

One comment I made when looking at the results of the
statistical analysis by TruthIsAll was that it would be most interesting to look at a time history of the deviation (by state) between exit poll and vote tabulation results. What you have done for SC gives some idea of what a time history might look like for that state. There is a clear early spike in favor of Bush which then vanishes (because of other small precinct or county results coming in I suspect) and then a gradual rise (this is coming from mostly larger precincts or counties). Now perhaps much of this gradual rise comes from latent early votes (which aren't recorded here until the precinct or county depending on how the state reports)? My premise is that election fraud may have occurred predominantly at early times because once Bush's victory is assured there is less need to continue the fraudulent activity. A time history might bear this out. I would very much like to know if counties have records from which could be gleaned the time histories.

[jsascj]

12. It would be interesting to see similar charts for other states

to see when other shifts occurred. I still think something fishy went down in VA.

[breadbox]

14. don't know if this is suspicious

It is certainly the case, however, that there are huge regional
differences in this state: the majority of the region in the north wes
is extremely strong * country, while the south east corner is much
more southern democrats (and willing to vote kerry over bush even though
they might be dixiecrats).
And then there are pockets --- for example, there is a town in the NW
just outside greenville where kerry only lost by 6 votes, and inez
tenenbaum beat out jim demint.

Regional differences, poor rural versus cities, and NW versus SE
might be enough to produce a graph like the one above.

However, one would have to explain why the regional differences let to
the timing differences, and rural versus city probably wouldn't be
enough for that.

[breadbox]

15. really easy electoral fraud

upstate SC had the easiest election system to beat that I know:
on arrival one queued according to A-K, L-Z first letter of last
name. After signing in, they gave one a piece of card stock,
3"x4" or so, burgundy, with a hole punched in it (there were
different colors, but the dimensions all looked the same: I am
assuming that these were precinct related). One then lined up in
a third queue, this time to vote.

The thing is that the lines were so long that it would have been
easy to sneak out without voting, duplicate the card stock exactly
and have 20, 50 or 100 people come back and just surreptitiously
join the back of the third queue (all three queues started in the same
basic physical location). Things were so busy that I don't think that
anyone would have noticed until the voter numbers were compared to the
number of signatures at the end of the day --- by then, far far too late.

[Pierre de Fermat]

16. Been there, done that.

I 'm new here but have been reading DU for quite some time. We spent 6 weeks talking about exit poll discrepancies, to no avail. Bottom, line: Exit polls are useful to the people (MSM) who pay for them, and are useful for the purpose they are intended - demographic analysis.

Exit polls can be used for valid demographic analysis even if the exit poll raw number are inaccurate. This is why exit polls are still popular even after their recent failures (2000, 2002) to match actual vote count (use google for background.)

The types of questions that are asked in a fraud analysis, cannot be answered by analyzing exit polls.

See here:

http://www.mysterypollster.com/main/2004/12/the_raw_exit_po.html

[Timebound]

17. kick n/t

[bemis12]

18. South Carolina's final totals

are much like the 2000 election results.

http://www.cnn.com/ELECTION/2000/results/index.president.html

They also pretty well match every poll done for the state, pre-election.

http://www.electoral-vote.com/states/south-carolina.html
http://www.rasmussenreports.com/election_2004.htm

Indeed, for anyone who was watching at all, the South Carolina results are singularly unsurprising. That exit poll appeared wrong from the beiginning.

[seaclyr]

21. Thanks for the information

there may be nothing odd about this, but reliance on final totals can still be deceptive. I think that if the 2000 pattern over time pretty much matches the 2004 pattern over time then I'd agree there's no reason to build a case for anything unusual. You'd probably agree that the profiles would be harder to match than the endpoints, right? And that, barring anything strange, the profiles should be about the same? I'm working on getting the numbers from election night 2000 to compare. So, you may well be right, but we'll see!

[IndyOp]

20. Seaclyr - There IS evidence of REAL TIME manipulation in Ohio...

Can you make these graphs for Ohio?!?!

Three cases in which votes appear to have been moved from the Democrat to third party candidate(s) to weaken the Democrat’s performance as precincts were reporting during election night: Ohio (2004) – Hamilton and Lucas Counties and Florida (2000).

OHIO (2004): The Ohio Secretary of State’s website showed that with 11.25% of Hamilton County’s precincts reporting, 34,804 votes had been cast George W. Bush; 39,541 for John F. Kerry; and 39,541 for David Cobb <http://www.oliverwillis.com/node/view/1152 >. The Ohio Secretary of State’s website now shows, (with 100% of precincts having reported) 215,639 votes for George W. Bush; 190,956 for John Kerry; and 0 for David Cobb < http://election.sos.state.oh.us/results/RaceDetail.aspx?race=PP >. Hamilton County used ES&S voting systems according to VerifiedVoting.org.

OHIO (2004): The Ohio Secretary of State’s website showed that with 6.06% of Lucas County’s precincts reporting, 1,917 votes had been cast George W. Bush; 0 for John F. Kerry; and 4,685 for David Cobb <http://www.oliverwillis.com/node/view/1152 >. The Ohio Secretary of State’s website now shows, (with 100% of precincts having reported) 85,405 votes for George W. Bush; 128,874 for John Kerry; and 0 for David Cobb < http://election.sos.state.oh.us/results/RaceDetail.aspx?race=PP >. Lucas used Diebold voting systems according to VerifiedVoting.org.

FLORIDA (2000): Black Box Voting < http://www.blackboxvoting.org > details what happened in the 2000 Presidential election in Florida. At10 p.m. on election night, a Democratic Party election official, Deborah Tannenbaum, called the county elections department and found that Al Gore was leading George W. Bush 83,000 votes to 62,000 votes. But when she checked the county’s Web site for an update half an hour later, she found a startling development: Gore’s count had dropped by 16,000 votes, while an obscure Socialist candidate had picked up 10,000 - all because of a single precinct with only 600 voters.

Additional research reported by Black Box Voting revealed that Precinct 216 gave Al Gore a –16,022 when votes were uploaded at approximately 10:30 p.m. on election night. On January 17, Volusia County employee Lana Hires asked the technical staff at Global Election Systems for help. “I need some answers!” she wrote. “Our department is being audited by the County. I have been waiting for someone to give me an explanation as to why Precinct 216 gave Al Gore a minus 16,022 hen it as uploaded.” Talbot Iredale of Global Election Systems replied: “Only the presidential totals were incorrect…The problem precinct has two memcory cards uploaded. The second one is the one I believe caused the problem. They were uploaded on the same port approx. 1 hour apart. As far as I know there should have only been one memory card uploaded.”

What was the impact of Volusia ‘error’ on the election? Again, Black Box Voting reports: At some point between 10:16 pm and 1:12 am Bush took the lead in the state and the gap between Bush and Gore widened by an amount sufficient to cause FOX, NBC, CBS, and ABC to call Florida for Bush. By 4:00 a.m. the Volusia ‘error’ had been corrected and CBS News retracted the call for Bush. According to a CBS internal report, “the call for Bush was based entirely on the tabulated county vote. There were several errors that were responsible for that mistake. The most egregious of the data errors has been well documented. Vote reports from Volusia County.”

Important: In the cases above it appears as if votes were moved from the Democrat to third party candidate(s) to weaken the Democrat’s performance as precincts were reporting during election night. In Florida (2000) the votes initially given to the Socialist candidate were apparently returned to Gore. In Ohio (2004) it is not clear that the votes were returned to Kerry.

[seaclyr]

29. Ohio is done - check your e-mail

Will post results on DU soon.

[seaclyr]

22. Other states aren't hard to do, except

results from some states may be less informative than others because of the constraints that come from pulling numbers from the cnn ticker; there are typically more numbers (reflecting more frequent updates) for states that closed early than for those that closed later. Also, for the small state of Delaware the numbers came in in a large block - the first cnn numbers represented 40% of the total vote and there wasn't a lot of useful information that could be extracted from the later numbers. It's possible to get a few more numbers from the msnbc ticker which jbond56 also has where the times are offset from cnn, but one would need a more complete set of numbers from AP (anyone have any idea on how to get these?) to track all the states adequately. I will have to check but I do believe there are enough numbers for Virginia and Ohio to carry out the same kind of analysis.

[seaclyr]

23. A general point

If anyone wanted to sway the election by adding votes for a favored candidate (and its variants, switching votes, subtracting votes from an opponent, etc.) it might not be too hard to figure out how many votes would be needed and where to find these votes (large urban areas, for instance). I think it would be harder to know exactly when to insert these extra votes. Inserting early would have the advantage of providing a comfort zone and perhaps a way to persuade people, if necessary, that exit polls are off. Inserting late would definitely look suspicious. Most sophisticated would be to match the time profile from the previous election, feeding votes in gradually over time. But that gets harder once you've chosen an area to get your extra votes from and perhaps a preferred voting method (one with no paper trail for example). Also, you couldn't match the prior election too well, because that would be suspicious too!

[bemis12]

24. Or even NOT match the previous election.

I can see it now. Speadsheets showing that the votes came in at DIFFERENT times, and being used to show the fraud.

"Wouldn't we expect the precints to report in approximately the same times and support the same parties? Fraud I tell ya!"

[Rhapsody in Blue]

28. Couple of observations

That very early spike up and then down in the % margin *might* be due to the fact that the actual numbers it's based on are still very small, so the corresponding percentages would be exaggerated. In other words, a 1,000-vote difference is a 25% margin when the total votes are 4,000, but becomes a 5% margin when the total votes are 20,000. If you have the actual numbers before 9:00 pm, it may help to explain it.

What really bothers me is the steady INCREASE in the % margin, consistent throughout the evening. Shouldn't the margin have reached a kind of point of stability (say 15% or whatever), and then stayed right around that number through to the end of the night? Why the slow and steady progression? (Note I'm talking about the % margin here, not the actual vote numbers, which are supposed to increase as the total votes are accumulated.)