Slashdot Mirror
Why Does a Voting Machine Need Calibration?
New submitter Shotgun writes "I heard on the radio that there were some issues with voting machines in Greensboro, NC (my hometown), and the story said the machines just needed "recalibration". Which made me ask, "WTF? Why does a machine for choosing between one of a few choices need 'calibration'?" This story seems to explain the issue."
TheBlaze (i.e. Glenn Beck) is not a credible news source. Please delete this article.
This was a problem with electronic voting machines during the 2008 elections:
http://www.techdirt.com/articles/20081029/0131342676.shtml
https://www.google.com/search?q=2008+voting+machine+screen+calibration
I'm speaking from a perspective of someone that regularly works as a poll worker during elections in the state of California.
One of the first things I do once our touch screen system is set up is confirm the calibration of the LCD panel. It's typical for the registration to be off by a few pixels, as our fingers are not perfect pixel-sized points. However, I have yet to experience an issue where the calibration is so bad that the wrong selection is made on behalf of the voter. Remember there are a whole host of perfectly valid reasons why this may be more of a problem for some voters than others, certainly including finger size and physical impairment affecting fine-motor skills.
If a voter did report a problem of this nature, recalibrating the touch screen would be one of the first things I would try.
Anyone that's ever worked with touchscreens before knows that those things need frequent recalibration
Touchscreens—particularly resistive touchscreens—often need recalibration. On a poorly calibrated screen, tapping on one button could select the one adjacent. Not good in a voting machine with a column full of candidates in densely packed rows.
Note: I haven't read TFA, this is just the first thing that came to mind.
How can I believe you when you tell me what I don't want to hear?
For either/both sides to call shenanigans when the vote does not go their way. I wonder if someone has done a study on the amount of press voter fraud gets vs. party election outcome and if there is as stark of a difference as I perceive. And if people really think that one party only wins when they "cheat", does that just reinforce myopic visions of political views (i.e. Most people think the way I do and so the only explanation is fraud)?
Put one at the left, the other at the right, and make them so far apart that they CANNOT POSSIBLY BE CONFUSED even if the system is out by some number of pixels (or even some fraction of an inch)!
Why is this so complicated?
Needs to be calibrated sometimes. I work elections for Clark County, Nevada. I've worked every election the last 10 years. And yes, the touchscreens can fall out of calibration and make it difficult to select the correct candidates. I can't speak to other election districts, but here in Clark County we're trained on how to perform this calibration on site (it's very simple) so that any problems reported by voters can be handled right away.
It usually refers to the coincidence between what the coordinates reported by the digitizer (touchscreen) as the center of the contact area, and the display coordinates underneath it:
“He played around with the field a little and realized that in order to vote for Romney, his finger had to be exactly on the mark,”
Still, the piece is biased starting with the title ("MORE ELECTRONIC VOTING MACHINES CHANGING ROMNEY VOTES TO OBAMA"), and the issue could be down to the active rectangle being different from what's displayed:
Nancy wrote in an email. She said “the invisible Obama field came down about 1/4 [of an inch]” into what should technically have been the Romney area. In a phone interview with TheBlaze, she explained further that her husband said he felt the area on the touchscreen that could be pushed to vote for Obama was larger than that for Romney.
This post contains no rudeness or derision of any kind. All arguments are friendly. Terms and exclusions may apply.
Even before reading the article, I knew what the answer was. This is because at my workplace (a public library), we deal with a very similar thing on a regular basis. We have several self-checkout units at each branch, which are basically all-in-one Windows PCs running special software. They have RFID pads for scanning the books, and they take input via a touchscreen. The capacitive touchscreens on tablets and smartphones are generally of good quality, but these are different. They are crappy resistive touchscreens, designed to keep costs down. Accuracy is poor, and a calibration utility must be run regularly or the screens will start to drift. Calibration entails running a program designed for that purpose, then touching targets displayed in each corner of the screen in sequence.
If calibration on a low-quality resistive touchscreen is off, then the mouse click may register at a location as much as 1 full inch away from where the user pressed. I have personally seen this happen many times on our self-checkout units. So if you hear a story that someone on a voting machine pressed the box for the Democratic candidate and it checked the Republican, or vice versa, I'd be willing to bet money that this is what happened. If they were deliberately tampering with the votes, why would they show that to the user?
There are indeed serious concerns with the lack of source availability for voting machines, and the ownership of voting machine companies by individuals with partisan ties. But calibration is not some kind of conspiracy – it's the inevitable result of using cheap touchscreen hardware.
http://xkcd.com/463/
Free the Quark 3 from asymptotic confinement! Bring your charm! Don't get down! All colours and flavours welcome!
A lot of these voting machines still run on Windows CE, similar to Siemens WinCC Flex HMI. They typically come with calibration software built in, once you launch the calibration you have to tap on several cross hairs that appear one after the other. The touchscreen is measuring resistance, when you run the calibration software it adjusts the amount of resistance it looks for to determine where you're tapping on the screen.
... and in the DRM, bind them.
First that letter was all about setting up a legal and public relations basis to question the election later.
Second, yes voting machines need calibration. Different types require different kinds.
For example the touchscreens, usually older resistive touch screens get mis calibrated on position. You have to remeber these things get locks in closets and sit in non-temperature controlled ware houses for a couple years at a time between elections, then they are jostled in trucks, cleaned with cleaners, and sometime run off various power sources. Empirically they do go out of calibration.
I personally have a ballot I saved from an AUtomark paper ballot printer in which all the votes are off by one oval width. that is 100% of the votes are incorrect and you can tell because a few are printed past the range of ovals.
Opscans are fairly easy to allign since they have relatively few degrees of freedom but they do get misalligned and become sensitive to printing tolerances.
Old lever machines used to have the gears wear down.
The solution to all this is not to require perfect everything but to have ways to check things. hand marked Paper ballots and some sampled recounts of those paper ballots such as is done in New Mexico is I believe the best compromise between transparency, robustness and simplicity. It's robust against human and machine errors so mere mortals can carry out very transparent elections. It's also robust against voter turnout variations too since it only takes more pencils to let more people vote, and if a machine breaks, you can still gather the ballots, so you dont get long lines at the polls.
Some drink at the fountain of knowledge. Others just gargle.
“He played around with the field a little and realized that in order to vote for Romney, his finger had to be exactly on the mark,”
welcome to the age of tablet computing.
sysadmins and parents of newborns get the same amount of sleep.
WTF? Why does a machine for choosing between one of a few choices need 'calibration'?
Because Rich Daley is not on the Chicago ballot anymore for mayor.
For us carnivores, "Sucking the marrow out of life" isn't a transcendentalist philosophy but a practical instruction.
Dec 15th 2009, claimed that Galileo proved the earth was round and that it revolves around the sun, and that the Dems/Obama are just like the evil people that tried to shut him up (I guess Obama is a Muslim Christian then, or Christian Muslim or something like that).
As soon as I read the title, I knew this had something to do with touchscreens. My question is, or something as important as voting in an election, why would anyone trust something as inaccurate as a touchscreen? Wouldn't it make more sense to just list the names with a physical button next to each, similar to what you'd see on many ATM's?
As for many people here saying they never need to re-calibrate their modern phones and tablets, is it possible that they do some type of self-calibration upon startup? I have an old, old Nexus One and on occasion the touchscreen will begin behaving erratically. Simply pressing the power button to lock the screen, then unlocking again resolves the issue.
we should always use paper ballots
you can cheat with paper ballots, but it's hard and you need a lot of effort and cooperation between many saboteurs
with electronic voting, magnitudes of order more attack vectors are introduced, because it's more complicated, unnecessarily. and one well-placed hacker can untraceably and silently cheat in milliseconds over a broad swath of votes
if people don't believe their government represents the popular will, then we have all sorts of problems
so paper voting only. now and forever, no matter how rich or technophilic the society. the voting in finland should be the same as in bangladesh as in brazil as in the usa: paper ballots only. to preserve the integrity of the process, people trusting their vote matters
intellectual property law is philosophically incoherent. it is your moral duty to ignore it or sabotage it
Claimed Sean Smith was a CIA operative sent to Benghazi to cover up Obama's involvement in the Libyan uprising.
Why not have physical buttons displayed down the left (or right, or top, or wherever makes sense) that correspond to the location of the screen next to them?
May 26, 2009 Beck claims that Hitler's "empathy" was the cause of the holocaust.
... is, as we say every time this comes up on /., paper ballots marked by the voter with a pencil.
If anyone wants their tablet recalibrated, they can send it to me. In my experience, the 64 gigabyte ipad 4 4g model is particularly prone to miscalibration. Typing errors can be a sign that recalibration is neccessary.
I've never had any calibration issues with my iPad. This kind of thing is a hallmark of older touch-screens, modern devices don't have this problem.
The justification sometimes used for using touch screens is because they have the ability to scale up the screen for accessibilty for visually impaired voters.
That's probably the only passable reason for using these peices of crap, and even at that, one could be set aside per precinct for use on-request by the people that need them, while the rest of us vote on a more verifiable system.
We should not settle for anything less than hand-marked paper (computer processing of the paper e.g. opscan is ok for a first count) with mandatory random hand-count audits of a statistically sound number of machines, with an automatic trigger of a full recount if those spot checks fail.
Someone had to do it.
In his book "Arguing With Idiots" (alternatively titled "My Inner Dialog"), Beck claims that Article 1, Section 9 Clause 1 of the constitution put a $10 entrance fee on immigrants coming to this country because the founding fathers "actually put a price tag on coming to this country: $10 per person. Apparently they felt like there was a value to being able to live here."
In actuality, Article 1 Section 9 Clause 1 was intended to prevent congress from ending the slave trade.
and I wanted to write a comment that maybe they was technically, scientifically and religiously retarded but I couldn't comment without signing up so I do so here instead.
Also they wouldn't have to use touch screens. Though any switch can fail of course.
But seriously, this hits close to home. I live in the Piedmont Triad (greensboro, nc area), and have voted on these very machines - this year and years past. The hype about this was way overblown and far too political. Most of the stories I've heard of potential fraud from the great north state has been on local media, and from those who may not have the best vision or may suffer from tremors due to age. It's not calibration if you can't choose the right region of the screen, due to medical conditions. That said, I work in Senior living and have not heard any complaints from residents that have voted early. In fact, they loved how easy it is - as most of them have not voted in years in a polling place. The poster seriously needs to stop listening to Clear Channel radio stations (rush radio, I'm assuming?) and perhaps some healthy NPR or our two fine college stations. Also, though it's been said: The blaze? Really, Slashdot?
There's a reason we don't replace our keyboards on computers with touch screens, and it isn't because the technology doesn't exist. It is because a keyboard is way better for typing than any touchscreen.
Why do you have machines??
If you read the statistical analysis showing the vote flipping that was done to get Romney a win in the primaries:
http://www.themoneyparty.org/main/wp-content/uploads/2012/10/Republican-Primary-Election-Results-Amazing-Statistical-Anomalies_V2.0.pdf
The rigged precincts showing the rigging were 'Central Tabulator' systems (Diebolds), paper vote districts didn't show any 'flipping' for anyone let alone Romney.
http://www.themoneyparty.org/main/stolen-election-2004-plus-the-voter-fraud-scam-series/wisconsin-no-tabulator-versus-tabulator-counties/
It turns out Diebolds Central Tabulator, lets the operator change the vote via a manual override screen!
http://www.youtube.com/watch?v=eRFtYGJtOEQ&feature=relmfu
So they didn't even need to do anything special, they just changed the numbers on the central counter.
How different can it be? Can an election be complicated enough that voting cannot be done via on screen prompts and some text next to 8 buttons?
Change the text on the screen next to the button, but don't pretend that something as simple as a "Select which candidate you want to vote for" can't be done with a few buttons. Surely something which for the last couple of hundred years has been as difficult as tick this box can only be done with a touchscreen right?
Yeah, sure you could. Which is of course why you're glibly saying you could instead of, er, actually doing it.
It's remarkable how simple many, many things appear to be if one is ignorant of how they actually operate and how much work goes into designing them. I used to be routinely guilty of this in machine shop, but I quickly learned to strip the phrase "just" from my vocabulary in light of how damn long it takes to get things right with a mill or lathe.
Most people learn to hold their tongue rather than spout off about how "simple" something they don't understand is because they don't want to look like idiots to those who do understand.
That all voting machines are shit? Because that's the GIST I'm getting.
I'm pretty sure I could build a better voting machine in my garage for under $100, bet these things cost tens of thousands.
It's been done. See Open voting consortium or Open voting solutions. But the problem is vastly harder than you estimate, to get it right. It's not that it has to be complex. It's the many pitfalls most people fail to anticipate, even one of which, destroys the whole concept.
Some drink at the fountain of knowledge. Others just gargle.