A report from another friend who's been observing the King County vote count:
At the end of the day Friday, November 12, the validation process was not complete at the 1st S facility, and the canvass board was still reviewing ballots at the County building. But other than the ballots the workers would validate on Saturday, when they supposedly finished the process, and those still awaiting canvass board review, the exact number of ballots waiting to be run through the counting machines should have been known. Or at least the count certainly could have been known if the County Elections officials had any simple systems in place to track it.I'd be very reluctant to allege explicitly organized fraud on the part of the election officials. On the other hand, this and other reports suggest that the process is sufficiently sloppy and disorganized to permit all kinds of fraudulent over-voting: individuals who vote multiple times, ballots of dubious validity that are aggressively accepted, ballots that are doubly counted, non-citizens who vote, etc. I have no specific evidence of any of these. But the sudden appearance of the 10,000 magical mystery ballots suggests a lack of proper oversight and controls that would inhibit the ordinary cheating that would tend to favor Democrats in King County.
By the end of the day Saturday, every validated ballot, except perhaps some of those at the canvassing board, had been recorded at one point or another in the system - either in validation, processing, or actual vote counting. At least that was what the workers at 1st S believed. So one explanation of the additional 10,000 count is gross mis-management of the County Elections office. Another, as I mentioned earlier, is that they were telling a lie. Even if the Election officials can't count without taking off their socks, it seems that they still should have known within a few thousand how many were to be processed by Friday night.
If fraud is involved, then the ballots could have come from one or more of several places -
* Absentee ballots mailed and received after the deadline, but having valid signatures* Provisional ballots originally ruled invalid
* Absentee and provisional ballots marked NSOF, No Signature on File, and therefore ineligible
There were a large number of NSOF ballots put in boxes and warehoused at the 1st S facility so they would have been handy. However they would have the NSOF notation on the envelope and I would think the observers would pick that up. To be "valid" some additional notation would have to be on the envelope and, typically a copy of the signature from the registration form is stapled to the ballot envelope. The NSOF notation would indicate that the voter isn't registered or registered after the deadline and would be ineligible to vote in this election.
Jim Miller, who named the phenomenon "distributed vote fraud" in a comment on my earlier post, has a good blog entry on this topic, here.
Posted by Stefan Sharkansky at November 16, 2004 11:07 AM | Email This* Provisional Ballots - No tight controls to prevent cheating. Someone with a fake ID can probably get another ballot in to vote twice, as well as non citizens. Signature matches are conducted by humans and are inherently innacurate.
* Election Board - The fact is that the election boards and canvasing offices are generally controlled by the party in power of any given county. Even with observers, the true objectivity of these boards is questionable, especially in dense population centers that can easily sway elections by gaming using timing, see below.
* Timing - This is the key to the game. By controlling when and what they report, counties can seriously effect the outcome of the election. For example, even if the 10,000 additional ballots found in King County are completely legitimate, the fact that they were "suddenly" discovered near the end of the counting process deprives the opposite side of the time they need to investigate the legitimacy of all ballots, or to respond in kind with their own gamed timing of a large set of previously unreported ballots. It turns the counting process into a card game with everyone waiting for the most opportune moment to play their hand.
What we really need is reform that eliminates provisional ballots, builds a truly bipartisan control of the elections process and requires that all results be reported at exactly the same time, and with an earlier hard deadline to prevent gaming. This will also encourage counties to get their votes counted quickly, or be cutoff. Delays in any election provide the time needed for the possibility of fraud and should thus be minimized.
It's truly pathetic that we hear these empassioned cries that every vote must be counted, etc. If a voter truly wants to be counted, he will have the dedication to register himself 30 days before the election, educate himself to the candidates and issues and show up at the correct precinct and vote. If one cannnot manage this simple process, then one should not be voting at all. The whole concept of provisional ballots is welfare for voting that creates opportunities for fraud and error and turns our formerly great elections the way of a Banana Republic.
-JB
Posted by: Jeff Boly on November 16, 2004 12:07 PMEach ballot has a unique control number, as well as a precinct code. The vote counting machine requires this (as well as the correct positioning of the ovals and other things) in order to process the ballot.
It would not be too difficult to create blank ballots with a color laser printer and the right software, which would exactly duplicate an existing ballot, including the control number and precinct code.
This could be done by taking a completed valid ballot, and reading it in a machine. The control number and precinct code are read, and a new blank ballot is printed from the color laser printer. This new blank ballot, once completed, will be counted as a valid vote by the King County vote counting machines.
On Friday, November 5, 2004, Gregoire ends up only 4,001 votes ahead of Rossi statewide. There are 360,000 ballots left to count, with 119,000 ballots being from King County. That same day, King County processed 59,000 ballots, which were Gregoire 54%, Rossi 44%, and Bennett 2%.
Obviously, at that point, Gregoire would lose by 15,000 or more votes statewide, since she was only pulling a 10% margin in King County, which would easily be offset by Rossi's margin from the other counties.
So someone takes 20,000 valid completed uncounted ballots from King County Elections some time Friday 11/05/2004 after close of business. Preferably from a fairly randomized sampling of the uncounted ballots.
This leaves 99,000 uncounted real ballots in the warehouse, instead of the total of 119,000 that was reported on the afternoon of Fri 11/05/2004.
These 20,000 stolen ballots are all scanned, and 20,000 blank ballots are generated, with the same precinct codes and control numbers as the stolen ballots. Each fake blank ballot, of course, has all the correct races and oval positions for the precinct it is supposedly for.
Then you have people fill in the ovals on all of these 20,000 blank ballots. They "vote" for Gregoire on these forged ballots, and presumably for the rest of the Democrat ticket as well.
Then you start feeding these 20,000 forged ballots back into the system over the remaining period before certification. The best thing is to add about one fake ballot for every five real ballots that are counted (20,000 forged ballots vs. 99,000 remaining real ballots)
So, on Monday 11/08/2004, 63,000 ballots are reported counted. But this is only 53,000 real ballots and 10,000 of the forged ballots.
Since 53,000 real ballots are counted on Mon 11/08/2004, and only 99,000 real ballots were left after the theft on Fri 11/05/2004, the Monday afternoon update reports that only 46,000 ballots are left to count. This is the same number as the number of real ballots left.
On Wed 11/10/2004, King County reports 21,000 ballots are counted. But only 17,000 of these were real ballots, and 4,000 of these were fake ballots. While 29,000 real ballots are actually remaining, King County reports only 25,000, so that the math does not look suspicious.
On Fri 11/12/2004, King County reports 14,000 ballots are counted. But only 11,000 of these were real ballots, and 3,000 of these were fake ballots. While 18,000 real ballots are actually remaining, King County reports only 11,000, so that the math does not look suspicious.
On Mon 11/15/2004, King County reports 17,000 ballots are counted. 14,000 of these were real ballots, and 3,000 of these were fake ballots. This leaves 4,000 real ballots -- the same number as King County reported being left on Mon 11/15/2004.
Obviously, there is a 10,000 ballot discrepancy that has to be explained. While it is easy to subtract 63,000 from 119,000 and get 46,000 when no one is noticing, it is quite another thing to subtract 17,000 from 11,000 and get 4,000 when the whole world is looking.
So King County Elections comes up with some phoney baloney story early on Mon 11/15/2004 to explain this 10,000 vote discrepancy.
As the first paragraph suggests, I invented this theory without any evidence whatsoever to support it. But it could explain a lot of things.
For example, the votes counted on Fri 11/05/2004 gave Gregoire only a 10 percent margin -- Gregoire 54%, Rossi 44%, Bennett 2%.
But everything counted since then out of King County has given Gregoire at least 20% margin, and in some cases, more than a 25% margin.
This can easily be accounted for, if 1/6 of the ballots were replaced with forged ballots that were all "voted" 100% for Gregoire.
Start out with 10,000 ballots -- 5,400 Gregoire (54.00%), 4,400 Rossi (44.00%), 200 Bennett (2.00%).
Remove 1/6 of these ballots at random. The remaining 8,333 valid real ballots should be -- 4,500 Gregoire, 3,667 Rossi, 166 Bennett.
Replace the 1,667 stolen ballots with 1,667 forged ballots voting a straight Democratic ticket -- 6,167 Gregoire (61.67%), 3,667 Rossi (36.67%), 166 Bennett (1.66%).
This process has now converted a 10% margin for Gregoire (observed on Fri 11/05/2004) to a 25% margin for Gregoire (observed starting Mon 11/08/2004 to the present date).
This also is a net gain of 18,000 votes for Gregoire. 20,000 real ballots would have given her a 2,000 vote margin -- Gregoire 10,800 (54%), Rossi 8,800 (44%), Bennett 200 (2%). On the other hand, Gregoire gets a 20,000 vote margin if these same real ballots are replaced with forged ballots.
Instead of Gregoire losing by 15,000 to 17,000 votes to Rossi, she instead miraculously pulls off a razor thin win of about a thousand votes.
Posted by: Conspiracy Nut on November 16, 2004 12:39 PMNothing sinister, and nothing new. And if these Democrat election officials are so bent on doing their party's dastardly business, why did almost every one of them endorse Republican Sam Reed for reelection as Secretary of State- and NONE of them supported the Democrat?
Posted by: Jim King on November 16, 2004 12:39 PMGregoire had a margin of well over 20% in the 63,000 ballots counted on Mon 11/08/2004 -- even though almost none of these were provisional ballots.
Almost all of the 63,000 votes counted on Mon 11/08/2004 were mailed-in absentees, just like the 59,000 votes counted on Fri 11/05/2004. But there was a sudden increase in Gregoire margin from under 10% on Fri 11/05/2004 to nearly 25% on Mon 11/08/2004.
Based on the samples from re-marking out-of-precinct ballots reporting by GOP election observers, Gregoire is polling only about a 10% margin from this portion of the provisional ballots.
Explaining the drastic change in Gregoire's margin based on the switchover to counting provisional ballots, instead of mailed-in absentees certainly sounds nice. But it is NOT supported by the actual daily vote counts.
Posted by: Conspiracy Nut on November 16, 2004 12:57 PM
This doesn't help their credibility.
But King County obviously erred by making their provisional ballots identical forms and color. By changing either the form or the color, we could identify ballots improperly fed into the machines without the eligibility of the voter having been verified.
I plan on going on this site and horsesass.org and scouring them for suggestions on how to improve the elections in the future. Any ideas, let me know.
Posted by: John Slyfield on June 8, 2005 04:08 PM