HOW DIEBOLD TOUCH SCREEN VOTING FAVORED BUSH

[Bill Bored]

Enough with the exit polls, the anecdotal reports, the hacking theories, etc. Here's a simple FACT about how Diebold Accuvote TS machines could have favored Bush. It's based on the GEMS user manual (used to remotely program Diebold Touch Screens and Op Scan voting machines from the GEMS server) and I believe it. No hacking or any other malfeasance is necessary. All you have to do is TRY TO VOTE.

First, let's assume there really were Republicans for Kerry, just like those websites, the Op Eds, and the endorsements before the election said there were. After all, we had the worst president in history and before the election it sure looked as if a number of apparently well-meaning conservatives and Republicans were going to take their chances with John Kerry while maintaining their loyalty to a GOP Congress -- just to keep Kerry in line. You know, checks and balances and all that founding fathers stuff, right?

But if they voted on Diebold Accuvote TS touch screen machines, unless they selected their candidates in EXACTLY the correct sequence, their votes would have gone to Bush, or they would not have cast a vote for president at all. Here's why:

With these machines, voters MUST make straight party choices BEFORE proceeding to change individual candidate selections. If the voter decides to make a candidate selection first, like for president maybe, BEFORE making a straight party selection, the presidential selection is lost, gone, history, asta la vista, toast! At this point, the presidential vote would either SWITCH to Bush (does this sound familiar truth seekers?) or perhaps just be an undervote. The manual isn't clear on that point, but either way, there's no way in hell the vote would have gone to John Kerry. Of course the voter could probably CORRECT his or her mistake, but come on, we're talking about Republicans here! What would be the odds of THAT happening? (Sorry but I had to throw that in.)

Now, take the same voter, same machine, same undisclosed software (hacked or not), and vote for the straight party GOP ticket FIRST and THEN change your presidential choice to Kerry, and voila! -- no problem! Just like that, you're a card carrying Republican for Kerry!

Now, I know there were a few Dems who voted for Bush: Fox viewers spooked by 9/11, Zell Miller, etc. But given the polls and the press and the endorsements and the publicity before the election, there were probably a hell of a lot more Republicans for Kerry than Democrats for Bush! And unless I'm wrong, a bunch of 'em were downright disenfranchised by Diebold. Imagine that; oh the irony; oh the humanity!

There is another way to program the Accuvote TS to actually be...well...ACCURATE. It's called Pennsylvania Straight Party Voting. This option allows voters to make candidate selections BEFORE straight party selections, without the candidate selections ending up in the bit bucket. And guess which swing state Kerry actually won? Does PA ring a bell? By any chance, do they have Diebold in PA, programmed with the PA Straight Party option so Republicans there can vote for Kerry without having to consult a GEMS user manual? Maybe that's how Kerry was able to carry the state. They have their very own configuration option in GEMS!

Would anyone care to refute this theory? Are there big flashing warning signs on these touch screens that tell Republicans for Kerry how to vote for him along with the straight party ticket for the other races? Or is it just like that old speed bump in the parking lot that someone forgot to paint yellow?

I wonder how all those Florida Dixiecrats were able to vote for Bush though. Oh that's right -- they used Op Scans, didn't they?

[Wilms]

1. Wow.

You said:

"At this point, the presidential vote would either SWITCH to Bush (does this sound familiar truth seekers?) or perhaps just be an undervote. The manual isn't clear on that point, but either way, there's no way in hell the vote would have gone to John Kerry. "

Cite from 'the manual', please. I want to make sure I understand.


But wow.

[Bill Bored]

5. GEMS User's Guide; Revision 3; July 3, 2001; GEMS version 1.17.15

I'm still trying to figure out how the Accuvote OS (Op Scans) work.
But of course, they inherently have VVPBs so there is less chance of voter error. There are however a plethora of ways to count and reject these ballots, all remotely programmed by GEMS. The question is whether any of them can actually switch votes. It will take me a while to sort out.

[rigel99]

7. great digging BB

can you post the link to this GEMS manual.... I'm being asked to put an ACLU technical person in charge of reading all this stuff and want to point them to the best source of Diebold technical info...

thanks! and great work....

[Bill Bored]

11. rigel, I already did this for you in another post, remember?

If you can't find it, PM me.

This stuff is dynamite. It's just very time consuming to wade through. Like giving oneself a training course without an instructor.

Your case should/could be based on how easy it is to manipulate the system without having any knowledge of actual programming. Then you don't have to prove the system was hacked -- just that it sucks to begin with.

[Andy_Stephenson]

9. Bill...

look to the central tabulator. The opscans are pretty reliable. The weak point is the transmission of the votes. Are the xmitted via modem? Or are they uploaded in a secure environment? In bigger counties they are generally xmitted via modem. If they are, there is a problem right there.

Another issue is wether or not the tabulator is hooked to a phone line or to the interent. In some cases they may not transmit votes via modem but the tabulator is connected to a phoneline so diebold can upload "updates" to the system.

Generally speaking the optiscans are very reliable. The issue with vote switching is in the transmission and tabulator.

[Bill Bored]

10. Andy, I am not opposed to OpScans

But Accuvote OS is programmed from GEMS! There are zillions of options that determine how overvotes, undervotes, straight party votes, etc. will be counted, whether or not the ballot is rejected by the Op Scan, giving the voter a chance to fix it, etc. I get a headache just reading this stuff.

I started with the touch screens because they are the least transparent, and most often cited as problematic, but Op Scan programming via GEMS may be just as bad. I'm still investigating.

If this work has already been done by anyone, let me know. It might save me some time, but with all the talk about hacking, etc. which is certainly possible, but very difficult to prove, I thought it would be more productive to look at how screwed up the system is WITHOUT the hacking.

[Colorado Blue]

2. Holy &*)T. Can we demonstrate this???

[vpigrad]

3. In SC...

the straight party tickets did not select the vote for President. That is the law in that state. You have to manually select the candidate for President. I worked with a group that tested voting machines for SC, and they worked correctly.

In GA, there are no laws on how this should work so the Repukes could have gotten away with anything they wanted. Knowing the type of people that are repukes around here, we all know they did.

[Wilms]

4. Do you know what the SC 'undervote' was for President? n/t

[Bill Bored]

6. This is another variation.

You set up GEMS to keep the presidential race out of the straight party definition. While this requires the voter to vote for pres. separately, it eliminates the possibility of the vote being switched due to voter error. On the other hand, it's possible to have undervotes by not voting for president. If there's an alarm to call this to the attention to the voter, it's not too bad but I'm not sure right now if that's an option or not. Suppose you could turn that alarm off!

In any case, I'm trying to look for things that would favor one candidate over another, and at least on the surface, this feature does not seem to do that.

[Bill Bored]

8. Wait a minute....

Any technique that increases the percentage of undervotes for president could be used to either candidate's advantage by targeting precincts in which the opponent has the lead! So let's say in Kerry-heavy precincts, you disallow straight party voting, but in Bush-heavy precincts, you allow it. You also set up the Bush-heavy ones without the PA Straight Party option, so that a vote by a Republican for Kerry will go away (see top of thread). This is the default setting for straight party voting, BTW.

The interesting thing about all this is, it barely scratches the surface of what's possible with these machines, without any malicious code or hacking whatsoever. It's fraud by design.

Where are the lawyers?

[Melissa G]

12. Kick for BB! Great research! n/t

[Senator]

14. Another minute

Any technique that increases the percentage of undervotes for president...

...could also be helpful if the machines are "miscalibrated" to "default to bush" for under/overvotes.

But that would produce shockingly low "residual" rates.

Just like we had (and were spun as a "big success").

[Wilms]

15. Great observation. n/t

[Bill Bored]

18. I hear you but...

2 things:

I am currently researching ways all this could occur through simple configuration errors or deliberate mis-configuration. This is sort of like FRAUD LITE. It's not hacking in the sense of altering code. I think you get that Senator, but I just wanted to clarify it.

It seems to me that there has been little or no research into this, but I might be uninformed. Someone please inform me if I am, and save me some time!

There is also an issue with the way Diebold counts undervotes. They only CALL them undervotes if a race has several possible winners, such as for elected judges, and has less than the maximum number of choices selected. In the case of president, where there should be only one choice allowed, an undervote is called a blank vote. Someone would have to be able to distinguish to produce correct stats.

[Wilms]

13. kick n/t

[Peace Patriot]

16. Some resources that might help...

Myth Breakers (2nd edition) at: http://votersunite.org/
(The second edition includes reports on 2004 election machine malfunctions, and has a thorough rundown on electronic voting systems--70 booklet--pdf)

Johns Hopkins Analysis of Electronic Voting System (Diebold)
http://www.scoop.co.nz/mason/stories/HL0307/S00196.htm#5

More on Diebold: http://www.chuckherrin.com/hackthevote.htm

------

Re: the evidence that Kerry won

Dr. Steven Freeman adds up Gore 2000 repeat voters + huge Dem success in new voter registration in 2004 (57% to 41%) + big jump of Nader voters to Kerry = 4 to 8 million votes for Kerry that just somehow vanished on election day (and changed the outcome).

This new Dem voter reg in '04 is an interesting item--because people just don't go out and register to vote for the first time in huge numbers suddenly out of nowhere. It means family, friends, co-workers, volunteers were pushing them to do so. So the repeat voters were very enthusiastic--for what? for voting for Bush? And were all these people flocking to the Dem party to vote for...Bush?

This section of Freeman's 2nd paper is entitled "The numbers don't add up."

Boy, does he have that right!

(His 2nd report is no longer available on the internet--he's publishing it as a book in May--but he will send it to you if you ask.)

(Freeman's reports: http://www.appliedresearch.us/sf/epdiscrep.htm)

[Peace Patriot]

17. I am not getting the picture very clearly...

"With these machines, voters MUST make straight party choices BEFORE proceeding to change individual candidate selections. If the voter decides to make a candidate selection first, like for president maybe, BEFORE making a straight party selection, the presidential selection is lost, gone, history, asta la vista, toast! At this point, the presidential vote would either SWITCH to Bush (does this sound familiar truth seekers?) or perhaps just be an undervote. The manual isn't clear on that point, but either way, there's no way in hell the vote would have gone to John Kerry. Of course the voter could probably CORRECT his or her mistake, but come on, we're talking about Republicans here! What would be the odds of THAT happening? (Sorry but I had to throw that in.)"

How could this happen WITHOUT hacking? I just can't seem to get what you're saying.

If the program defaults to Bush or to undervote, in a situation in which the voter picked prez first, then went to straight party vote, doesn't that require a hack? Why would it default to Bush or undervote, and not Kerry? Why wouldn't it default randomly (if it's flawed and is going to default)?

Or are you saying, ok, here's this flawed program that eats votes for prez, and by setting up this flawed program, ahead of time, they created the backdoor by which to LATER hack it and make it default to Bush or undervote?

????

[Bill Bored]

19. NO.

I am saying that this is the published documentation for the product.
It's not a hack. It's how it's supposed to work.
Still working on all this. Will post more later.
623 pages of documentation to go through.

[Bill Bored]

20. More:

Let me reiterate that this would favor the candidate who had the fewest cross-party votes. I.e., it would favor Bush if there were a lot of Republicans for Kerry. I believe there were a significant number of these, judging from all the press they got before the election. These would be the voters who would have had the most difficulty voting for Kerry. And this presumes the machines were set up for straight party voting, without the PA Straight Party Option. (This is the default for straight party voting on Diebold.)

If straight party voting was not allowed for president, then you'd have to choose your president separately, and this should work. You see, each race can either be included in Straight Party Voting or not, and the vote can therefore be set to vanish (if you select the included race first).

Furthermore, there are two ways to set up how a vote can be CHANGED by the voter. They are 1-click, like on a web browser, or 2-clicks, which forces the voter to click the incorrect choice first, and then click the correct one. 2-click is the default setting.

Consider the scenario where Straight Party races are set up that include POTUS and (by default) WITHOUT the PA Straight Party Option:

With 1-click voting:
The voter picks Kerry, votes Republican, the Kerry vote disappears. The voter clicks Kerry again, and presumably, this time, Kerry is selected because the selection happened AFTER the Straight Party selection and only one click is required to change a vote.

But suppose the system is configured for 2-click voting (the default setting):
The voter picks Kerry, votes Republican Straight Party, the Kerry vote disappears. The voter clicks Kerry again. Now, the voter has NOT removed the Repub Straight Party selection, so would he or she be allowed to select Kerry? Or would this require clicking the Straight Party again to get rid of it, selecting Straight Party again to get it back, clicking Bush to get rid of that selection, and THEN clicking Kerry? Can't say for sure, can you?

Now let's assume one more thing in the above scenario:
The voter does EVERYTHING "right", i.e., votes Repub Straight Party FIRST, then wants to change Bush to Kerry. How does she do that? At a minimum, it seems she MUST FIRST click Bush, and then click Kerry!

Piece of cake right? Of course, instead of clicking with a mouse, this all has to be done on a touch screen.

(Now I must say I'm not 100% sure about all of these scenarios, but you could see how it could be possible, given all the available options, most of which are actually set by default from the GEMS server itself and downloaded to the touch screens.)

Now, imagine if you set these things up differently in each precinct, because you knew which one would favor each candidate. Then you can really have some fun, and once again, there is no hacking or malicious code required at all. It's just part of the program.

If anyone has actual experience with this, please post here. It's fascinatin', ain't it?

[Wilms]

21. That is rather fascinating. n/t

:grr:

[Niche]

22. Absolutely true... look at the #s across the board

many districts/states had 69-70% Dem voter turn out vs. 100% repuke turnout... really? look at the registration vs. turnout... it's amazing.

[Bill Bored]

23. Well it is true that Mitofsky had to adjust his Repub turnout numbers

upward and the Dems downward to get the exit polls to jive with the vote counts, but what does this have to do with a Diebold configuration that would make it difficult for Repubs to vote for Kerry?

[BeFree]

24. Bill

The Scoop website had a dieblod manual which stated: If a presidential vote is selected first and then the voter makes a straight party selection the pres. vote is deleted, and the machine makes a default selection for the pres vote.

You keep looking, you'll find it somewhere. I posted it here, back about a month ago, but ol' befree is a pariah these days. LOL

[Bill Bored]

25. I remember that one

It's from a manual given to the techs. Also tells them how to look like they know what they're talking about. Yeah right. I'll look and see what it says about the "default" selection. I have not been able to find a way to program the default, but it could simply be the straight party choice, which in the case of a Republican for Kerry, wouldn't be Kerry!

Here's the post about this from Feb:
<http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x310796#313964>

Thanks to icehenge and BeFree for digging this up originally!

[Bill Bored]

26. Could it be that all those complaints about votes switching to Bush

were from Republicans trying to vote for Kerry?

This is clearly what would happen under the scenarios described above.

There were far fewer complaints the other way around, but there were some. The reason there were fewer has been attributed to the complaint numbers being given mostly to Dems and progressives. Will Doherty of Verified Voting made this comment in an interview.

But could it be that there were simply far fewer Dems who wanted to vote for Bush, and therefore fewer complaints of votes switching to Kerry?

It really does seem that straight party Republican voters who wanted to vote for Kerry were the most likely to be screwed by these machines! That's one way to engender party loyalty I guess. Now if we can find a way to reach these people, we might find some allies in this fight. Have the Republicans for Kerry "moved on" or "gotten over it" yet, or what???

[TruthIsAll]

27. You are on to something very important here.

Evidence of BBV fraud is out there in droves.
Here you have dsicovered documented fraud BUILT-IN to the design.

"Sorry, if you are a Repub, your vote for Kerry will register for Bush. You are forbidden to change allegiance".

You must make this digestable to the MSM.

[Bill Bored]

28. Thanks for your "endorsement" TIA!

I was just wondering what a math guy like you might be able to infer from this. I was going to look at the exit and pre-election polls to review the party affiliation numbers vs. who they voted for to see if there is any significant difference in the cross-party presidential vote between Dems and Repubs in 2004 vs. prior years. If we can prove that there were more Repubs for Kerry than Dems for Bush, it would be helpful.

Note that I was going to use the actual results from the prior year polls, not the responses about Y2K given in 2004.

I believe there is typically ~90% party loyalty for president but if it can be shown that this didn't apply in 2004, and that there was also a discrepancy in the exit polls, and that there were more Repubs for Kerry than Dems for Bush, it would support the theory.

This all depends on how the machines were configured, county by county, precinct by precinct. This is something that could only be found from GEMS itself, or perhaps the TS machines themselves, or by asking BOEs or whoever configured them. I'm sure there are times when they had help from the vendors, and unless this specific behavior of the machines is clearly understood by all concerned, I'm sure that some were duped into using it, or didn't know any better. Again, the hidden speed bump analogy.

As seen from other posts above, there are different laws in each state regarding straight party voting:

1. Include POTUS/exclude POTUS
2. PA Straight Party/No PA Straight Party (by calling it "PA", other states might NOT use it, even if it's not illegal!)
3. No straight party voting at all
4. Do whatever the F$&( you want!

These should be available online somewhere. I'd suggest focusing on swing states though.

If we knew these laws, we could look for correlations in the numbers to see if Bush did better than expected in states with the default straight party configuration.

Then you have the 1-click/2-click option.

What does seem clear is that the DEFAULT configuration, at least in the case of Diebold, is to keep the troops in line by making it difficult for them to vote cross party in any race defined as part of a straight party ticket. It takes a lot of effort to configure the Diebold machines to do otherwise, and it might even be illegal to do so in some states! Other vendors' software needs to be investigated too.

And, if you set them up to EXCLUDE POTUS from the straight party ticket, you risk having a lot of undervotes for POTUS (which Diebold calls "blank" votes). Are these being counted/reported correctly?

As you know, there were greater exit poll deviations in states with Republican Governors. Since GEMS is usually configured by only one or two people (for "security" reasons) it follows that those with top level access would be Repubs.

But TIA, as you know, I'm not totally convinced about the validity of the exit polls. HOWEVER, if they do fit into the larger picture and help to corroborate other evidence or theories, I'm open to citing/using them!

I think what we need to do is to get some Republicans for Kerry on our side. This could really start a ground swell of some kind if they can be convinced that they were disenfranchised. So where are they?

And one more thing: obviously, I can't do all this work myself! ;-)

[Wilms]

29. Uh. What did you just say??

http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x342985

[Bill Bored]

30. Wow! Could this be reaching some sort of critical mass?

We should send this article to Conyers and Sensenbrenner. Sensen is from WI and we want him to have those hearings right?

Where is Teddyk23 these days?

[Wilms]

31. Have you considered emailing him...

or the House Dems/Judiciary?

[Bill Bored]

32. Done nt

[Wilms]

35. Great! When will you follow that up with a call??

:)

[Bill Bored]

33. kick nt

[Name removed]

34. Deleted message

Message removed by moderator. Click here to review the message board rules.