"Autditability"

[eridani]

There is an old joke about an engineer, a physicist and a mathematician each discovering a fire in their hotel room wastebaskets. The engineer uses the room fire extinguisher to put out the blaze, and empties the rest of its contents onto the wastebasket and the surrounding area as an extra safety measure. The physicist does a quick test blast of the fire extinguisher, does a series of calculations on his laptop, and then aims a precisely timed squirt at the base of the fire which is exactly enough to put it out. The mathematician looks at the fire, goes into the bathroom, runs the water, dabbles his hand in it, looks at the fire again and says "Aha! A solution exists!" and then goes back to bed.

Elections officials currently act like the mathematician, when they really need to act like the engineer. IT IS UTTERLY USELESS TO HAVE TABULATION PROCESSES THAT ARE AUDITABLE, IF THEY ARE NOT ACTUALLY AUDITED.

[MH1]

1. Tabulation - interesting point

I was having a conversation about this recently and I am really curious what folks in this forum have to say (I am not a regular here, apologies if this has been answered ad nauseum).

I hear people talking as if fraud could have occurred in the "tabulation" presumably meaning at the county or state level.

To be blunt, that makes no sense in my own experience as a precinct election worker in Pennsylvania. Maybe other states and counties are different though...that is what I am curious about.

Here in my county, there is a tally procedure conducted at the precinct on election night, that includes creating 4 official copies of the tally sheet, plus however many unofficial tallies as are created by poll-watchers representing the parties and candidates.

One official tally sheet is posted publicly at the polling place. One is given as part of a sealed packet of election records to the Minority Inspector. The other two - at least one goes to the county obviously; I think one goes in a sealed packet to the Judge of Election for the precinct.

The unofficial tallies from each precinct make their way to the various interested parties ASAP (often being called in by cell phone), like to the Democratic and Republican party HQ's.

The tabulation at the county is merely a matter of adding up the tallies from the various precinct. Since all the precinct numbers are public, and known to the parties and election monitors of various stripes, how can the numbers be tabulated fraudently at the county or state level, without someone catching it? It seems like a very simple process for any election watcher organization to validate the tabulation.

Can someone tell me what I am missing here? Why is "tabulation fraud" even a possibility? Is it the case that the process I described is fine and fraud-proof, but not all states/counties follow such a process? If so what are those states?

Thanks for any light you can shed. This has been bugging me for a while and seeing this post triggered me to finally ask. (please don't throw things at me if I'm being stupid!)

p.s. I am only talking here about tabulation of the precinct counts. I am aware of things that can happen at the precinct level. If the numbers coming out of the precinct are fraudulent, that is a different question - I only don't understand how the tabulation can fail to reflect the tallies received from the precincts.

[Boredtodeath]

2. Here's just two examples

While the canvas at the precinct is supposed to reconcile - the number of "times counted" on the DREs is supposed to equal the total number of voters who signed in.

You'll note the first page of this precinct "recap" sheet (the tabulation process in this precinct) is off by 2 votes. State law says the pollworkers cannot leave or report the precinct until it reconciles, you'll find that reality dictates otherwise. There are 43 precinct recap sheets in the file below:
http://www.countthevote.org/open_records/dekalb_dre_recap3.pdf (see the notes near the bottom)

In the 2004 General Election in Dekalb County, Georgia 50% of the precincts could not reconcile.

Note also - here, while the "times counted" is the same in every race, note the Blank Voted numbers. So, here, you can see the possibility that the machines are "blanking" votes in specific races.

These are just 2 examples of how the counts and tabulation at the precinct level can be fraudulent and precinct workers be innocent or unaware.

[MH1]

3. "Blank voting" is a precinct / DRE question, right?

I haevn't digested your full post yet - just want to eliminate this point, unless I misunderstood.

After re-reading your post, let me clarify mine - I am concerned about allegations of fraud in the tabulations at county level and higher - not the precinct.

Back to blank voting - In the PA primary just held, Santorum got significantly more "blank votes" then Swann (gov race) in the Republican primary. This is an entirely expected outcome. (Not that Casey supporters like me aren't having lots of fun with it, ha ha).

Either way, let's assume it is true that the black box inside each machine could be recording the votes inaccurately - my question is not about that. So I am going to ignore this point for now. (I'm not saying it isn't a problem. I am just compartmentalizing.)

I want to digest your post more thoroughly, but I'm thinking you misunderstood my question, so let me re-phrase: Does anyone here believe that (significant) fraud can happen at the county / state tabulation level (or wherever tallies from multiple precincts are added together)? If so what is the mechanism that this is believed to happen by?

Thanks for posting!

[Boredtodeath]

4. OK, let me try again

Yes, I did misunderstand your question. Sorry.

Let me talk a little about the "tabulation process" at the county level.

First, let's talk about the paper absentee ballots which are tabulated at the county office. Your location may do this differently, but not in Georgia.

All of the paper absentee ballots are scanned and turned into electronic ballots to be added to the DRE electronic ballots. The DRE electronic ballots are returned to the county for tabulation as described by a pollwatcher:

As poll workers bring the reports and PCMCIA cards back to election center:

1. HQ Poll workers receive the envelopes, checking the seal
2. HQ Poll worker registers the number of PCMCIA cards indicated in handwriting on the envelope.
3. HQ Poll worker breaks the seal, counts the PCMCIA cards, checks off task on list. The HQ poll worker passes the PCMCIA cards to the next station where the cards are counted again.
4. At no time during this process does any HQ poll worker check, monitor or verify the serial numbers or any other identifying information on the PCMCIA cards.
5. HQ Poll workers return to the envelopes and check off that it contains "tapes." The tapes are not removed from the envelope, they are not read, they are not documented and the envelope is dumped into a bin with other envelopes. These "tapes" contain the Zero Total Report and the End of Day Precinct Totals.
6. The 3rd station HQ Poll worker searches for the marked accumulator card and pulls it from the stack of cards because it has the accumulated total database file on it and that database is prepared for upload to the GEMS computer. In this district, the VIBS machines and associated card are designated the accumulator unit. We did not observe them actually uploading these files.
7. If there is no "accumulated database" on the designated card, the PCMCIA cards are located, taken to a touchscreen machine and the accumulation process begun. Once again, separating the accumulator card from the remaining cards which are returned to the stack.

There's simply no point in this process that verifies the data from the precinct (the end of day totals and Recap sheets are tossed aside).

During the county canvas, the paper absentee ballots and the provisional ballots are added to the DRE totals. This is accomplished on the Diebold GEMS system through manual entry:

http://www.countthevote.org/manual_entry/manual_entry.htm

I'll stop here so you can digest this information.

[MH1]

6. Thanks - some clarifying questions:

1) when you say "the DRE electronic ballots" you mean the PCMCIA cards in the voting machines? If I understand you correctly, the absentees are opened at the precint, and each absentee is fed into a machine as if it was from a walk-in voter.

2) you are saying that the votes are NOT tabulated at the precinct across the machines and posted publicly somewhere? The precinct election workers do not have any way to know the totals that are on the PCMCIA cards being delivered to the county?

Having started to look at this I am coming to a simple conclusion: use a friggin' spreadsheet, people. Have the precincts tally like they do where I live. Let the county manually enter the precinct values into a spreadsheet and have the election judge validate that the numbers for his/her precinct are correct. Post the whole friggin thing publicly so that the Minority Inspectors and all the poll-watchers can make sure each election judge did his or her job. And Election Integrity people can traipse around looking at numbers posted at polling places if they want to.

And spreadsheet software is essentially free. (okay, so you have to implement some security. Big whup.)

</rant>...off to try to understand this (in and around chores and just enjoying the beautiful weather today)

[Boredtodeath]

10. Answers

1) when you say "the DRE electronic ballots" you mean the PCMCIA cards in the voting machines? If I understand you correctly, the absentees are opened at the precint, and each absentee is fed into a machine as if it was from a walk-in voter.

Not in Georgia. In Georgia, absentee ballots are sent to the county election office. There, on election day, the outer envelopes are discarded and each ballot run through the OpScan. This is known as Central Count OpScan.

2) you are saying that the votes are NOT tabulated at the precinct across the machines and posted publicly somewhere? The precinct election workers do not have any way to know the totals that are on the PCMCIA cards being delivered to the county?

No, there are tabulated and a Precinct Recap Sheet (as shown in the original example) is posted at the precinct. So, do a degree, they DO know what the machines SAID at the closing of the precinct.

However, once the cards and reports are turned into the county office - those figures can, and do, change. So, what you saw at the precinct can be radically different after the county, and then state, canvass is complete.

And, yeah, you're right - this process has been made too difficult for most to understand. My guess is that there is a reason for that - they don't want US sticking our nose in.

on edit:
I forgot one more thing....early voting in Georgia ALSO has it's own precinct. In other words, voters who cast their ballot during the early voting period on DREs are a precinct unto their own in each county.

[Febble]

5. Thanks for asking this

as my understanding is that in most systems and jurisdictions, precinct level "tabulations" are available at the precinct on election night. It has always struck me that tabulator fraud in such systems is unlikely because it would be so easily caught. Sure, it might not be, but the risk of detection seems to me to be high for the perpetrator. Perhaps someone can put me right on this.

However, more to the point: to bring up the exit poll issue, in about 60% of precincts, the vote-totals that are compared with the poll totals are collected at the precinct. In the remainder they are collected at the county. Thus, in the majority of NEP precincts, if fraud were to have been responsible for the observed precinct-level discrepancies, this would have had to occurred at the precinct tabulation not the county tabulation.

Does anyone know, if there is an answer to this, which voting technologies are associated with the non-availability of precinct level totals at the precinct at close of polling?

[Kip Humphrey]

7. It depends on the state/county procedures and subsequent system design

If there is a clear human-visible audit trail from machine-to-precinct total-to-county tabulator, any digital vote total manipulation (flipping) must occur in real or near real-time on precinct machines.

If machine or precinct totals are known and public (maybe just known if the allegiance of local personnel is a variable), digital vote manipulation (flipping) must occur at th precinct machine level (accumulator, where no voting machine tapes exist; voting machine, where they do exist).

If precinct totals are not tabulated and known or made public (or, at least where this does not occur until its way too late, such as in New Mexico where precinct count discrepancies were found to no effect), digital vote total manipulation can occur at the county tabulator level. See also Alaska, where precinct counts were declared to be proprietary information of the digital voting system vendor (otherwise known as a cover-up!).

Networking configuration also plays a role in malicious design and execution. In most systems, county tabulators are networked to the state tabulator, and in some, networked to precinct machines too. Where networked, the state tabulator can be used to manage and direct algorithm variables dynamically up until vote totals are certified (confirmed by human eyes). A state tabulator functioning as a central vote manipulator, can direct networked state tabulators by supplying dynamic variable values and/or rebuilding county databases to smooth vote switching across counties and ensure a desired outcome for state races. For example, in Ohio 2004, Blackwell boasted he had a personal tabulator in his office with bi-directional communication capability down to the precinct level (one might ask why?).

If, as in Pennsylvania, precinct totals are manually tabulated (or verified) at the voting machine level and precinct totals are tabulated prior to any network connection to the county and/or state tabulators AND those precinct totals are posted publicly on election night such that those precinct totals can be collected and compiled manually, digital vote count manipulation must be performed at the voting machine level by running a real-time digital voting machine algorithm producing vote flipping at the voting machine level. There's plenty of anecdotal evidence of this occurring in various systems across the country in 2004.

Two side notes: 2004 gave us the infamous excuse, "It was just a computer 'glitch'". These 'glitches' popped up in areas where voting machine level vote human-visible tabulation occurred (machine tapes existed and were used to tabulate precinct totals). With poor machine configurations, such as forgetting to turn off "screen refresh", some voters witnessed vote flipping before their eyes and a few reported such. "Oh, it was just a computer glitch" was the resounding excuse. More likely, it was incorrect configuration such that the flip didn't remain hidden. Second side note: In the case of the Hart Intercivic eSlate, the game (con) was configuring the voting machines so as to fool voters who voted "Democratic Slate" and, in so doing, denying Kerry uncountable numbers of votes. One would need to be trained in human-to-computer interaction to identify this exploit - I am one so trained and I tested for and identified this exploit when I voted on the eSlate.

[Febble]

8. OK, thanks

Could you give me a list of systems that in your view are least vulnerable to tampering?

Eg, hand-counted paper ballots, older technologies where votes are tabulated at the precinct before networking.

Also, if you know it, states/technologies where these conditions prevailed in 2004.

Thanks!

[Kip Humphrey]

11. Hand-counted paper ballots and

Voter's Choice (VC). I'm not a donor so couldn't search and pull up my posts on VC. I won't give Skinner a stinkin' dime until he re-instates ERD to its place of visibility it once enjoyed. Even in this election year, ERD remains in the backwater of DU (at least he could put it on the home page's Featured Forum List!). Do a search on Voter's Choice and Kip Humphrey in ERD.

When it comes to digital vote processing systems, least vulnerable is like "a little bit pregnant". See Digital Voting for Dummies (4.65Mb .zip) for my take on Digital Vote Processing.

Being more focused on the 'how' in 2004, 'safe' states in 2004 haven't been my area of expertise. If I were to make such a search I would look to the subset of states with fewer than two counties employing digital voting. Re: most digital voting system vendor contracts stipulated a requirement that centralized digital tabulators be used to count all statewide races. Secondly, even in light of your position on the 2004 exit polls, I would examine those states where pre-election polls and exit polls jived with the official results.

[Febble]

13. Well, I am certainly

focused on the "how" of 2004, because without a plausible "how" we don't have anything like a ballpark on "how much".

What I need is an estimate of variance, not at state level, but at precinct level. Eg. in what proportion of precincts is the election-night precinct level unlikely to be tampered with. I have no real idea what is postulated here. If I know that 40% of precinct vote counts were obtained at the county level (because unobtainable at the precinct) what does that mean for the remaining 60%?

And what does it mean for the possibility for online adjustment during election day?

This does matter to the forensics, because, as you sound as though you are aware, my position on the 2004 exit polls is that while I do not share view that a biased poll is implausible (nor the view that there is no evidence for it) I find the non-correlation between "swing" to Bush and "redshift" difficult to reconcile with any postulated mechanism for massive, widespread fraud. If there is variance in the amount of fraud, or if there are limits to how much fine tuning to the amount was possible during election day, I do not see how fraud can be indexed by the exit poll discrepancy. And if it is not, then it means that there is very little evidence to back the case that digital voting resulted in massive theft, although I completely agree that it is an absurdly insecure way to vote, whether or not the insecurities were in fact used to steal millions of vote for Bush in 2004.

And speaking as one who has been "a little bit pregnant" at least eight times in my life but only successfully pregnant once, I'd say the degree of vulnerability matters. I only produced one child because I was only a little bit capable of carrying a child to term. The more heroic the efforts required to steal a vote, the fewer are likely to have been stolen. The more I find out about the mechanics of digital fraud, and the implications of those mechanics for the postulated impact on fraud on the exit polls - the prime piece of evidence cited for massive digital theft - the harder I find it to make any plausible case for anything other than an absurd risk.

Out of interest: how many votes do you think were stolen, digitally, in 2004, and why? (Not a trick question - I really want to know what numbers are being postulated here).

[Kip Humphrey]

15. My postulations

Q: If I know that 40% of precinct vote counts were obtained at the county level (because unobtainable at the precinct) what does that mean for the remaining 60%?

A:Assumption: We (regulars on ERD anyway) generally know Digital Vote Processing Systems (DVPS) are unsecured and vulnerable to malicious intrusion and manipulation of the vote.

Assumption: someone or some group is interested in using those vulnerabilities to ensure a desired outcome (I leave aside the real question of why the prevalent DVPSs ignore security in their requirements and execution of those requirements - something, as an Information Systems professional, I find indicative of intent BTW)

The 40% could be manipulated at the county tabulator as well as at the precinct level (accumulator or voting machine depending system design). The 60% would have to be manipulated at the precinct level to avoid detection. The determining factor is where in the system human eyes see the data.

Q: And what does it mean for the possibility for online adjustment during election day?

A:In a tabulator driven scenario (where human eyes do not see the data at the precinct level), variables driving vote flipping algorithms can be both preset and manipulated dynamically by human operators. In the precinct level scenario, variable variance would most likely be hard coded in the program and run independently of human intervention.

Q (implied): If there is variance in the amount of fraud, or if there are limits to how much fine tuning to the amount was possible during election day, I do not see how fraud can be indexed by the exit poll discrepancy.

A:In general, it can't. However, one may be able to spot aggregate trends if malicious algorithm variables comply with upper or lower programmed variance limits or settle to a mean (less likely to show up at the precinct level; more likely to show up at the county level). For example: At the lowest level (digital voting machine), a program keeps a running total of the Kerry/Bush/3rd party votes (each machine does this for its individual running totals). The variation parameters follow the following rules: 3rd party vote < 8% (or stay within pre-election MoE, whatever that is 2 weeks before the election), undervote < 10% (to avoid triggering a mandatory audit in 21 states), total flip from Kerry to Bush <= 4% (to maintain viability within MoE of pre-election polls. I predict this would be increased to 5% in 2006). Interesting 2004 facts: 11 southwest Texas counties (all 80# hispanic and historically 70% Democratic) all had identical 9% undervotes using Optiscans. Interestingly, before the election, BOTH parties urged Democratic voters in those counties to vote using the "Vote Democratic Slate". Also, interesting is that the Texas Republican party urged Texas Republicans to NOT use the "Vote Republican Slate" option when voting in 2004.

Q: how many votes do you think were stolen, digitally, in 2004, and why?

A: Because the software code is unavailable and most precinct tallies remain unpublished and/or uncollated, any total election numbers ascribed to digital vote flipping are necessarily speculative. Each piece of evidence we have is therefore circumstantial, anecdotal, and therefore not definitive. That said, it is my view that in the aggregate, the body of evidence supporting the charge that the 2004 election was stolen is presently sufficient for a verdict in the affirmative at least by the preponderance of evidence and, in my view, meets the requirements of beyond reasonable doubt:

That 2 standard deviations of reported voting "irregularities" favored Bush over Kerry, that what you voted on (digital or analog) coincided with the degree of discrepancy (High vs Low) between exit polls and official tallies

That DVPSs were engineered without any common sense industry standard security measures, that the top tier DVPS vendors were all operated by Republican loyalists

That these same loyalists lied about the security, networking, certification, and reliability of their DVPSs

That I personally tested for and found (on the Hart Intercivic eSlate on which I voted) an HCI (Human-Computer Interaction) exploit utilizing well established Human Factors responses and engineered to deny Kerry votes

That Karl Rove and GW returned to the White House at 8:36 PM EST to oversee operations of an "Election War Room" (the White House's name) inside the Oval Office (not to be confused with a similar set up in the White House dining room), staffed by RNC (not White House) employees (12), with communications equipment and networked PCs (12), with White House photos of same released on 11/3 and broadcast by at least CNN (Wolf Blitzer) and MSNBC (Keith Obbermann), which corresponded in time with the exit poll red shift

That obstruction of EVERY inquiry and investigation and court case has been consistent and ongoing

That the accusations and evidence of fraud have been largely ignored by corporate media with no substantive MSM investigations (even the Carter-Baker Commission complained to me they couldn't get MSM coverage! - count our blessings I guess)

Throw in (in almost no particular order):
Blackwell's actions pre and post-election day
Republican dirty tricks in 2004
Massive voter disenfranchisement only where it hurt Democrats
The subjective experience of everyone I've talked to (including Republicans) of election day (100% thought a landslide for Kerry was happening)
The take-over of the Republican party by a minority coalition in the 80s and 90s
The pronounced imperatives and agenda of the Neo-Cons (see PNAC)
The Max Cleland "miracle" loss in Georgia 2002 (a suspected beta test for 2004), The Chuck Hegal "miracle" win in Nebraska
The lack of public support for GW's agenda and actions since the 2004 election
The stolen election of 2000 proven but not covered by the MSM (one would think such a finding would result in typical obsessive MSM coverage - NOT!)
The Ohio recount used pre-selected precincts to recount, involved vendor technicians resetting tabulators, and remains incomplete to this day
That I strongly suspect I was investigated (person pursued us seeking personal information and emailed from a server named FED1 in Atlanta which contained files providing step-by-step instructions how to hack various networks and network protocols - definately not a hacker's server by content and language; definately intended to instruct non-hacker types) and wiretapped (stripped wires at the pole terminal block per the telephone repairman who repaired the damage) from 12/8/04 to 1/19/05, presumably for my election fraud protest efforts
Oh yes, and the discrepancies between exit polls and official tallies in 2004

I could go on and on and on...

Now, just as I cannot "prove" Rove oversaw the flipping of tabulator votes on election night in the Oval Office, I cannot prove some 3.8 million votes were flipped in the 2004 election. However, motive, means, opportunity, and a boatload of circumstantial evidence leads me to those conclusions. A preponderance of evidence? Beyond a reasonable doubt? This citizen's verdict is "guilty as charged".

[Febble]

16. Thanks

A:Assumption: We (regulars on ERD anyway) generally know Digital Vote Processing Systems (DVPS) are unsecured and vulnerable to malicious intrusion and manipulation of the vote.

I agree, from the evidence I have seen from a distance.

Assumption: someone or some group is interested in using those vulnerabilities to ensure a desired outcome (I leave aside the real question of why the prevalent DVPSs ignore security in their requirements and execution of those requirements - something, as an Information Systems professional, I find indicative of intent BTW)

I am happy to work with hypotheticals, if they make testable predictions, and this one does

The 40% could be manipulated at the county tabulator as well as at the precinct level (accumulator or voting machine depending system design). The 60% would have to be manipulated at the precinct level to avoid detection. The determining factor is where in the system human eyes see the data.

Yes

A:In a tabulator driven scenario (where human eyes do not see the data at the precinct level), variables driving vote flipping algorithms can be both preset and manipulated dynamically by human operators. In the precinct level scenario, variable variance would most likely be hard coded in the program and run independently of human intervention.

It is the hard coding part that presents the problem but of course it depends on the code.

However, one may be able to spot aggregate trends if malicious algorithm variables comply with upper or lower programmed variance limits or settle to a mean (less likely to show up at the precinct level; more likely to show up at the county level).

I think I get you, but again we have the problem that 60% of the NEP vote-count totals were collected at the precinct.

For example: At the lowest level (digital voting machine), a program keeps a running total of the Kerry/Bush/3rd party votes (each machine does this for its individual running totals). The variation parameters follow the following rules: 3rd party vote < 8% (or stay within pre-election MoE, whatever that is 2 weeks before the election), undervote < 10% (to avoid triggering a mandatory audit in 21 states), total flip from Kerry to Bush <= 4% (to maintain viability within MoE of pre-election polls. I predict this would be increased to 5% in 2006). Interesting 2004 facts: 11 southwest Texas counties (all 80# hispanic and historically 70% Democratic) all had identical 9% undervotes using Optiscans. Interestingly, before the election, BOTH parties urged Democratic voters in those counties to vote using the "Vote Democratic Slate". Also, interesting is that the Texas Republican party urged Texas Republicans to NOT use the "Vote Republican Slate" option when voting in 2004.

Thanks for all this detail. In short, the algorithm needs the pre-election estimate, and then flips a proportion of the Kerry vote to below a certain maximum plausibility. So we would expect the greatest discrepancies in high Kerry precincts in states where Bush's vote came in at the top end of his state pre-election poll MoE, and where there was either county input into the precinct count (digital systems only?) or where the count was only obtainable from the county tabulator, right?

Q: how many votes do you think were stolen, digitally, in 2004, and why?

A: Because the software code is unavailable and most precinct tallies remain unpublished and/or uncollated, any total election numbers ascribed to digital vote flipping are necessarily speculative. Each piece of evidence we have is therefore circumstantial, anecdotal, and therefore not definitive. That said, it is my view that in the aggregate, the body of evidence supporting the charge that the 2004 election was stolen is presently sufficient for a verdict in the affirmative at least by the preponderance of evidence and, in my view, meets the requirements of beyond reasonable doubt:

OK, let's consider the supporting evidence:

That 2 standard deviations of reported voting "irregularities" favored Bush over Kerry,

well, this isn't quite kosher, as there is no reason to suppose that the reporting system was known equally well in Republican as compared with Democratic precincts. I agree it is suggestive, but the binomial theorem is not applicable here, so your st dev is not legit. You can only use the binomial theorem if you are sure of a random sample, and this was a "volunteer sample" and there is no evidence that the volunteers were even drawn from comparable populations. But, as I said, and you say, it is suggestive.

that what you voted on (digital or analog) coincided with the degree of discrepancy (High vs Low) between exit polls and official tallies

The evidence suggest the opposite of what you imply. Discrepancies were greater in urban precincts using older technology (levers; punchcards) than newer systems (optical scans; DREs) which is not that surprising given the fact that Gore lost the presidency because of over-votes on punchcards. In places with population <50,000 there was no difference in discrepancy between any voting method, including paper.

That DVPSs were engineered without any common sense industry standard security measures, that the top tier DVPS vendors were all operated by Republican loyalists

Outrageous.

That these same loyalists lied about the security, networking, certification, and reliability of their DVPSs

ditto.

That I personally tested for and found (on the Hart Intercivic eSlate on which I voted) an HCI (Human-Computer Interaction) exploit utilizing well established Human Factors responses and engineered to deny Kerry votes

Details?

That Karl Rove and GW returned to the White House at 8:36 PM EST to oversee operations of an "Election War Room" (the White House's name) inside the Oval Office (not to be confused with a similar set up in the White House dining room), staffed by RNC (not White House) employees (12), with communications equipment and networked PCs (12), with White House photos of same released on 11/3 and broadcast by at least CNN (Wolf Blitzer) and MSNBC (Keith Obbermann), which corresponded in time with the exit poll red shift

Well, here we have a problem. I don't buy the "coincidence in time". The red shift in the exit poll simply reflects reweighting to the vote count - the common variable is much more likely to be the closing of the polls and the availability of precinct counts. There is really nothing mysterious here, although I realise it was widely misunderstood. The issue is not the time the polls were reweighted, which would have happened anyway, but the magnitude of the reweighting which gave at least some of us an early indication of the size of the discrepancy, which we now know much more precisely from the E-M report.

That obstruction of EVERY inquiry and investigation and court case has been consistent and ongoing

OK.

That the accusations and evidence of fraud have been largely ignored by corporate media with no substantive MSM investigations (even the Carter-Baker Commission complained to me they couldn't get MSM coverage! - count our blessings I guess)

OK.

Throw in (in almost no particular order):
Blackwell's actions pre and post-election day
Republican dirty tricks in 2004
Massive voter disenfranchisement only where it hurt Democrats
The subjective experience of everyone I've talked to (including Republicans) of election day (100% thought a landslide for Kerry was happening)
The take-over of the Republican party by a minority coalition in the 80s and 90s
The pronounced imperatives and agenda of the Neo-Cons (see PNAC)
The Max Cleland "miracle" loss in Georgia 2002 (a suspected beta test for 2004), The Chuck Hegal "miracle" win in Nebraska
The lack of public support for GW's agenda and actions since the 2004 election
The stolen election of 2000 proven but not covered by the MSM (one would think such a finding would result in typical obsessive MSM coverage - NOT!)
The Ohio recount used pre-selected precincts to recount, involved vendor technicians resetting tabulators, and remains incomplete to this day
That I strongly suspect I was investigated (person pursued us seeking personal information and emailed from a server named FED1 in Atlanta which contained files providing step-by-step instructions how to hack various networks and network protocols - definately not a hacker's server by content and language; definately intended to instruct non-hacker types) and wiretapped (stripped wires at the pole terminal block per the telephone repairman who repaired the damage) from 12/8/04 to 1/19/05, presumably for my election fraud protest efforts
Oh yes, and the discrepancies between exit polls and official tallies in 2004

All those are what got me into this too.

Now, just as I cannot "prove" Rove oversaw the flipping of tabulator votes on election night in the Oval Office, I cannot prove some 3.8 million votes were flipped in the 2004 election. However, motive, means, opportunity, and a boatload of circumstantial evidence leads me to those conclusions. A preponderance of evidence? Beyond a reasonable doubt? This citizen's verdict is "guilty as charged".

Well, this foreigner's verdict: fraud looks to have been far less of a factor in the 2004 result than I feared at first; voter suppression remains outrageous, well supported by evidence, and neglected; lack of confidence in the system is justified, and will continue to corrode American democracy until voting protocol is completely overhauled, making it transparent, auditable and just.

Cheers, and thanks again

Lizzie

[Kip Humphrey]

17. Preferring to not belabor our disagreements, I think we can both agree

The current crop of Digital Vote Processing Systems deserve a well-founded vote of "no confidence".

Kip

[Febble]

18. Absolutely

I think we probably agree on quite a bit more than that too!

Good luck.

Lizzie

[eridani]

20. Re how many votes were stolen on 2004

Howthefuck can anybody tell if the primary data (votes and tabulation systems) are considered to be private property that citizens are not allowed to look at?

[Bill Bored]

22. Well, I like this one:

"Interesting 2004 facts: 11 southwest Texas counties (all 80# hispanic and historically 70% Democratic) all had identical 9% undervotes using Optiscans. Interestingly, before the election, BOTH parties urged Democratic voters in those counties to vote using the "Vote Democratic Slate". Also, interesting is that the Texas Republican party urged Texas Republicans to NOT use the "Vote Republican Slate" option when voting in 2004."

The only thing I don't like about it is that this can easily be set up so that one party could in fact vote straight ticket and not be screwed, while the other party could vote straight ticket and be screwed. It should not be necessary to tell the Repubs to avoid voting straight ticket, unless there is something I haven't thought of yet. Good catch though.

[WillYourVoteBCounted]

9. True - and here is an example

You hit that nail on the head -

example

cash register full of cash, coins, checks, credit& debit card slips.
cash register journal tape (of each transaction).

The manager of this "store" would never dream of trusting the journal tape only.
The journal tape lists what "should" be in the "till".
It is not a guarantee of what is in the "till".
It was never meant to replace counting the contents of the drawer, but
was intended to be a measure to check the contents against.

The cash, coins, checks, credit & debit card slips are all counted, every bit.
The purpose of the journal tape is the record of what should be in that "till".
The two should reconcile.

[Kip Humphrey]

12. The ONLY vote you can count on is your vote of "no confidence" in

Digital Vote Processing.

[eridani]

14. With all mail-in balloting--

--tabulation is centralized, which is an argument against it. Precinct level tabulation is a partial check against this, but precinct auditing in my county is only checking the number of voters signing in against the number of ballots in that precinct tabulated by the scanner. This can't by itself be a check of whether or not the totals were manipulated in such a way as to have the total number of ballots remain correct.

[Stevepol]

19. I'd also like to ask about the central tabulators. My impression is that

in those states where the computers are pretty much in control now as an overwhelming percentage of voting procedures used (and were in 04), that is, FL, GA, OH, AL, TX and others, the lack of a precinct level tally being done and published at that level (as is apparently done in PA) makes it a piece of cake for a central tabulator to flip the total vote in any way desired, of course depending on access, insider information, and other factors.

If that's not true, then why the elaborate efforts to build into the tabulators all these "back doors," such as the recent one in Diebold that Avi Rubin and others are complaining about?

It seems obvious that, as stated earlier, if audits are possible but aren't done at any level (and not just audits to check the total vote against the total number of people checked off as voting), you can't by definition have a democracy.

This is the whole point of the ERD forum and all the people interested in this: paper has got to be used to verify the accuracy of the machine vote count or it's just not possible to have a democracy. That's why, in my estimation, and it seems obvious to me, we don't have a democracy today. And lacking that you have to look at the evidence on either side, not just demand conclusive evidence of fraud, but look at the evidence on either side and ask this: Is it more likely that the vote alleged is actually the way people voted or that it's not? To me it seems, based on what I've seen, the evidence on the side of fraud is far greater than the evidence for a fair election. And I'd also say that it's far beyond a reasonable doubt.

[eridani]

21. To quote David Dill again--

--the problem is that we can't possibly know one way or another. "It is not enough that elections be accurate--we have to KNOW that they are accurate, and we don't."

And you are right--as long as we don't know, we don't have a democracy.

[Boredtodeath]

23. NOT TRUE in Georgia

in those states where the computers are pretty much in control now as an overwhelming percentage of voting procedures used (and were in 04), that is, FL, GA, OH, AL, TX and others, the lack of a precinct level tally being done and published at that level (as is apparently done in PA)

This is simply not true for Georgia. Precinct level tallies ARE PUBLISHED at the precinct.

[Febble]

24. Do you know where else?

I'd like to know in which states precinct tallies are published at the precinct (or in which states they aren't).

[Boredtodeath]

25. I really have no idea

However, I was at a speech a long time ago (a couple years at least) where Doug Jones mentioned that Georgia's election law was some of the most detailed he had ever seen.

Based on that, I have to assume it's not many.