More Votes Lost By Diebold; Discovered By Unique Voting Transparency Project
from the reliable,-huh? dept
For years and years Diebold Election Systems (now Premier Election Solutions) had always vehemently denied that its e-voting or optical vote scanning machines had any problems — despite mounds and mounds of evidence of problems. We were shocked, this past summer, when the company finally admitted to a glitch with some of its machines, but the company still downplayed the significance of this, claiming that it didn’t believe the glitch (which loses votes) had actually impacted any elections.
Yet, even after this glitch was officially revealed, in the election just last month, we’re now finding out that Diebold machines caused 200 lost votes in an election in California. Even worse, no one would even know about this at all if it weren’t for a highly ambitious and very unique program set up by some voting activists to ensure there was real transparency. They convinced the local government to let them scan every single ballot and put it online for anyone to view. It was that separate process where they discovered the ballot counts didn’t match, and that Diebold seemed to show absolutely no records of the missing ballots, despite having scanned them.
Makes you kinda wonder how many other areas lost votes that absolutely no one knows about because they didn’t have such a system in place, huh?
Filed Under: accuracy, california, e-voting, elections, lost votes, transparency
Companies: diebold, premier voting
Comments on “More Votes Lost By Diebold; Discovered By Unique Voting Transparency Project”
Just look at the elections where a few hundred votes count like Tom Perriello and Virgil Goode in Virginia. The margin there is 745 votes.
Re: Close elections
How about Minnesota. Last I saw (12/6/08), they were STILL doing the recount and the gap was under 200 votes. Depending on which party you listened to and which interpretation of the challenges you took, either Norm Coleman or Al Franken could be in the lead by as few as 40 votes. 200 missing votes there could make a HUGE difference.
Voting Soviet Style
As I have said in the past – those who vote decide nothing; those who count the votes decide everything.
Re: Voting Soviet Style
those who vote, and count votes, decide who gets to sell their vote to Washington interest groups for campaign financing. the system is designed to be corrupt.
This is absurd. Diebold makes ATMs for banks (or at least my bank) and you know banks are pretty serious about people not taking their money and I’ve heard no stories of people ripping off ATMs so they must be fairly secure and they definitely count the money correctly, I’ve always received the correct amount.
On top of all that, if you want to deposite a check at the Deibold ATM at my bank, you have to put the check in the ATM and it scans the check and uses OCR to determine how much it is made out for. How can they read hand writing on a check but can’t read bubbles on a ballot?
Re: Re:
I hear about them being hacked all the time and I’ve seen, first hand, ones that count wrong. For like 2 weeks ours kept giving out 10s instead of 20s no matter how many times it was worked on. They eventually took it away in a big metal van and never put in a new one.
Re: Re:
Do a Google search. ATMs malfunction quite a bit. A vote that’s incorrect because of a malfunction is just as bad as a vote that’s been changed maliciously. The end result is the same, yanno.
Re: Re:
That’s simple! No one is PAYING THEM to misread your checks! On the contrary, the ARE getting paid to read them clearly and correctly.
If you think you get the politician you voted for, you’re a damned fool. You get the politician who was arranged by the people who pay for and count those votes.
Re: Re:
The answer is simple: by intention.
The CEO of Diebold has a few quotes you could use to insinuate this further.
The fact is there is NO valid reason for this to be happening. Even only halfway competent programmers and engineers can design a system that can do this accurately.
So that leaves me wondering why a company with experience in a related field is ‘having issues’ with this.
Re: Re:
Uh…an ATM is not a voting machine. ATMs are simply a console into the bank’s systems, the backend systems are what process your transactions and deliver your money. The ATM is simply an interaction point. Unlike voting machines which accept input and tally votes based on the total independence of the software running on a simple machine…not so simple, eh?
Re: Re: It's really that easy
A voting termnial is exactly as simple as an ATM. It runs the software installed, and for each vote, it increments a variable. Say ‘OBAMA’ or ‘MCCAIN’ were the variables. A vote for Obama, you say?
OBAMA = OBAMA + 1
Either at the end of voting, or in real time, the vote counts are transmitted to a database. It’s a remarkably simple system that someone with only basic programming knowledge could whip up within a week. It won’t be perfect, but that’s what testing is for. Testing that these companies get paid enormous amounts of money to do.
Re: Re: It IS that simple.
An ATM is simple? Not from my point of view, but okay, let’s go with that. You have an interface that interacts with the database that tracks the ‘money’ in the bank. You gain access to your account, can do some rudimentary account maintenance, check balances, so forth.
A voting machine is much more simple than an ATM.
All a voting machine does is track variables that contain vote counts. It then transmits these counts to a database that holds the sum of all voting counts. Sometimes, the machines don’t even hold the data on the machine, they send the data in real time to the database.
Honestly, it’s perfectly realistic to think that these companies are paid off. It’s even reliably likely. However, you’d need to catch them at it. ‘Mistakes’ like those seen with these machines would never make it to release software, BECAUSE of how simple the system should be.
Re: Re: Re: Sorry...
Pardon the double post.
Re: Re:
ATMs cost $250,000 and have had major investments made to make them secure and accurate. No county can afford a $250,000 voting machine, much less 10 or more (per polling location).
Prove a negative
It’s like my IT department asking how many emails I have missed since our email was updated….genius.
This is an example of people not doing their jobs. The programmers at Diebold, or Premier, or whatever they’ll call themselves after this, are being lazy. When you work on projects like this, with such an important function, you need to make sure you get it right. And scanning cards isn’t truly that difficult. I’m currently working on a project that’s similar.
Or you could offer up some form of conspiracy, but that would mean pointing the finger at a side, and it’s not fair for an arbiter to do such a thing.
Er, but they already admitted it
The story says Diebold admitted to the problem in 2004 and tried notifying election officials of the workaround. The problem appears to be not that the error was unknown, but a failure of due diligence on either Diebold’s or the county’s part in not either keeping track of the workaround or making sure a patch was distributed.
I'd pay for that
Disturbing. Four (4) years between problem admittance and lack of problem solution. There is no problem here, this is a for pay on-demand “feature”.
No, it's *not* that easy
I don’t think anyone here realizes it isn’t easy. Try writing software that always keeps a 100% accurate total in the presence of hardware errors, operator errors, and outright malicious intent. We’ve been writing operating systems for 60 years, and can’t them to work with malicious system admins, why would voting systems be able to? As previous posters pointed out, ATM’s *do* get defrauded, banks just write it off – I *KNOW* I’ve done ATM withdrawals that never appeared on my bank statement (I have the receipt). And they have it easy – they’re counting bills – physical pieces of paper that *often* get counted multiple times just as a doublecheck. It’s not a matter of incrementing a counter – that’s the trivial part – it’s the cross checks, the secure once and only once data transfers, the verifications that someone loaded all the machines *exactly once in the presence of errors*, no matter what the intent or skill of the election judges. The problem is counting individual human anything to +-.05% accuracy (200 votes in 400,000) is actually hard.
Re: No, it's *not* that easy
Yes it is. Paper ballots, x marks the spot. At least that takes away the “software” access to defraud an election. Search for GEMS software. Hanging chads….was BS. If there was a “dimple” in the spot next to a candidate…
“I don’t think anyone here realizes it isn’t easy.” Tsk tsk. You need a new job. Vending machines accurately count product and cash intake. Now when you open the cash box and there is a fake dollar, or slug, they still get counted. How easy would it be to construct a machine to punch holes through paper? No, dont put anything in it that is not purely mechanical. Same with the machine to read it. Nah…thats too easy. As far as getting money without it being removed from your account, I am sorry, but I do not believe you.
“In the next century, nations as we know it will be obsolete; all states will recognize a single, global authority. National sovereignty wasn’t such a great idea after all.”
Strobe Talbot, President Clinton’s Deputy Secretary of State, as quoted in Time, July 20th, l992.
vote counting
Apparently the open source software that was used to “audit” the vote count , and that uncovered the error, worked just fine.
Why not just get more jurisdictions to use that?