West Virginia - Vote theft already underway

[AJRitz]

I'm surprised this hasn't been posted here yet. Locations in West Virginia where electronic voting machines are in the care of Republicans are reportedly showing problems in early voting. Voters select Democratic candidates on the touchscreen, but the machine registers Republican votes. And, of course, the machines are incapable of providing a paper receipt to ensure to the voter that the correct vote was recorded.

Full story in The Charleston Gazette (http://wvgazette.com/News/200810180380).

[mercurial]

I'm pretty sure this was already here somewhere, AJ. Maybe in one of the early voting or voter suppression threads.

[AJRitz]

I'm pretty sure this was already here somewhere, AJ. Maybe in one of the early voting or voter suppression threads.
Ahhh - sorry. I checked out of the early voting thread partway through - this didn't really seem to have anything to do with early voting, other than time frame. They're the identical voting machines that will be used on Election Day. I'd ask a mod to delete it, but I really think this issue deserves attention. It's totally absurd that a company owned by a Bush "Pioneer" can make voting machines that are babysat by Republican election officials, and then are put into service with no cross-check ability.

More stories like this one, coupled with polls that I've been arguing for the last two years are conducted using fatally flawed methodology, and we could see some real (and arguably even legitimate) civil unrest on November 5.

[mercurial]

Oh, I don't disagree. I hope stuff like this is mostly due to simple error and will be caught but, on the other hand, it makes me a little paranoid and very angry about the other possibilities.

[walters]

Sigh. If someone wanted to produce a fraudulent touchscreen voting machine, don't you think they'd do it by making it appear to work correctly?

These "I touched A but B got selected" is simple miscalibration of the touchscreen and nothing more. They keep calling them "ATM style" and I wish that they had actually designed them that way (with physical buttons next to the screen).

[TheIndependent]

physical buttons are actually more error prone than touch screens for the user - lining up the button to the candidate often results in errors.

and yes, if you design a machine to rob votes, you certainly don't give it away by telling the user you are duping about it - but maybe the designer/coder sucked :)

[JP]

How big are the screen buttons on these things? Anybody have an idea what a typical candidate-selection layout looks like?

I've programmed touch-screens and it's possible for the calibration to be off by an inch or two, but if you're touching the left side of the screen and it thinks you're touching the right side then it's not calibration that's the culprit. Could be a cracked or faulty screen, connector worked loose in transport, something like that.

As for wanting to produce a fraudulent machine and making it appear to work correctly, it depends on who's doing it and what access they have to the hardware and software. Anybody with authorization to calibrate them (election workers, etc) could purposefully miscalibrate them, and with a little trial and error could make them perform as reported. Making it look like you're voting one way while it secretly votes another, that requires A LOT more knowledge and access.

[DH]

physical buttons are actually more error prone than touch screens for the user - lining up the button to the candidate often results in errors.

and yes, if you design a machine to rob votes, you certainly don't give it away by telling the user you are duping about it - but maybe the designer/coder sucked :)

The machines are cheaply manufactured pieces of crap.

Optical scan, without being perfect (because nothing is), seems like a no-brainer as far as the way to go.

[mercurial]

physical buttons are actually more error prone than touch screens for the user - lining up the button to the candidate often results in errors.

and yes, if you design a machine to rob votes, you certainly don't give it away by telling the user you are duping about it - but maybe the designer/coder sucked :)

The machines are cheaply manufactured pieces of crap.

Optical scan, without being perfect (because nothing is), seems like a no-brainer as far as the way to go.

My preference was a touch screen that PRINTED an opscan sheet of your results. You could verify what it had marked on the sheet and then if there were any issues, the paper trail was easily human auditable.

[fake]

Any votes the Republicans steal will be equalled by the number the dems steal

[Mysteryman]

physical buttons are actually more error prone than touch screens for the user - lining up the button to the candidate often results in errors.

and yes, if you design a machine to rob votes, you certainly don't give it away by telling the user you are duping about it - but maybe the designer/coder sucked :)

The machines are cheaply manufactured pieces of crap.

Optical scan, without being perfect (because nothing is), seems like a no-brainer as far as the way to go.

My preference was a touch screen that PRINTED an opscan sheet of your results. You could verify what it had marked on the sheet and then if there were any issues, the paper trail was easily human auditable.

When my friend and I sat down to tackle the problem of electronic voting here's what we came up with:

Two computer systems linked mechanically. One of these systems would be running the proprietary software currently running on evoting machines, the other would be an open source software package.

Mechanical buttons would press electronic buttons simultaneously on each system. You would see the vote of both system on side-by-side screens as you make it, and the immediate result. If one system says "you voted A" and the other says "you voted B" you know there's a problem. Paper audit tape would exist and both systems would print their respective results.

This is still hackable, but much harder when one entire half of the system is open to public scrutiny.

[nataylor]

I like our electronic voting machines. No touch screen. You use a scroll wheel and a button to make your selections: http://www.hartintercivic.com/files/eSlate.swf

But yeah, I wish there was a paper trail, too.

[AJRitz]

This whole issue has bothered me ever since these machines started coming out. The whole thing just seems to be the opposite of what we should be expecting from an open democracy. The machines run proprietary code, which its developers zealously protect under the guise of the DMCA. Statutes actually require that the code be vetted by a 3rd party testing agency, but the 3rd party testers aren't given access to the actual code. If you've seen Hacking Democracy you know that the Diebold software that runs a majority of these voting machines has been demonstrated to be vulnerable to hacking. And there's no paper trail or ability to cross-check the electronic results? It's just fucking absurd.

I don't have time to search for them right now, but I even recall some midterm election issues when some data cards holding vote records were not turned in for counting (they were "misplaced"). With no audit trail, that would be an incredibly easy way to blunt high voter turnout in neighborhoods where one party or the other's voters are predominate.

[bryce1012]

And there's no paper trail or ability to cross-check the electronic results? It's just fucking absurd.This is what really bothers me the most. I'm all for technology being used to accelerate the counting process, or as a double-check, or whatever. But any system that doesn't have a permanent, physical record of your vote is absolutely and completely untrustable, IMO.

Touch the screen to mark your ballot. It records your choices, and prints out a tape with human- and machine-readable versions of your votes. You hold it up to the screen to verify that the machine and the tape match, and then you put that tape into a ballot box.

Initial results can then be tabulated instantaneously, with the "official" count coming from the paper tapes a few hours or days later.

Why is this so difficult for some people to understand?

[Mysteryman]

For the reasons that both of you just mentioned is why half (from front to back) of my voting system is open source. It may not be able to tell you what's going on on the commercial system, but it can tell you it's not what it thinks it should be.