Hacker News Reader

Comment by salman89 | Hacker News Reader

Voting machines need to be a lot better than this or not exist at all, but does anyone actually think that IF this machine was altering votes, it would alter it in this fashion with a UI element tied to the alteration? Seems more like a crappy touch screen.

View on HN

This was my thought exactly. If someone was going through the trouble of altering votes, it would make no sense to make note of this within UI elements, it could just as easily "secretly" change the votes for 1/8 (just my theoretical magical number) voters and alter the results just enough without being noticeable.

Except this just frustrates the voter and maybe he doesn't bother voting for Obama at all. There's no accusation of "altering" the votes because how do you track, "Touchscreen didn't select Obama, voter just gave up?"

Donning my tin foil hat for a minute here -- I think if I were to try and attempt to commit voter fraud through the machines, I'd make it this obvious intentionally so it seemed like a legitimate accident to those who noticed; "It's too obvious to be fraud"

It would probably cause a small dent and shift by people who don't notice or are confused (elderly).

If you had complete control over a voting machine, it wouldn't make any sense to try to lower turnout or confuse anyone- you'd want people to vote and leave without having noticed anything, so you could change their vote.

The result here is that the voting machine was taken out of duty and will be investigated. If the strategy was to make the error so outlandish that nobody would believe it was happening, the strategy has failed.

Since it's pretty obvious that this would be the outcome, it's not a good strategy, and is therefore not likely to have been enacted.

"Cannot reproduce. user error"

I wonder though... maybe the inconvenience would be enough to either change a percentage of votes or make a few people give up.

Looks like a bug, actually swings a close election?

from the article (emphasis mine): > I initially selected Obama but Romney was highlighted. I assumed it was being picky so I deselected Romney and tried Obama again, this time more carefully, and still got Romney. Being a software developer, I immediately went into troubleshoot mode. I first thought the calibration was off and tried selecting Jill Stein to actually highlight Obama. Nope. Jill Stein was selected just fine. Next I deselected her and started at the top of Romney’s name and started tapping very closely together to find the ‘active areas’. From the top of Romney’s button down to the bottom of the black checkbox beside Obama’s name was all active for Romney. From the bottom of that same checkbox to the bottom of the Obama button (basically a small white sliver) is what let me choose Obama. Stein’s button was fine. All other buttons worked fine.

It sure looks exactly that way from the video.

I've used that exact type of machine to vote in North Carolina (but this ballot is not from North Carolina). You have all sorts of people pounding on the screen all day.

Remember the reason we all wanted electronic voting after the year 2000: because it gives absolutely clear answers, not necessarily better answers. (Even the best systems have errors because there are humans involved.) You don't have to debate whether that smudge counts or does not count as a vote.

Electronic voting should be allowed but not unless it prints out a ballot with all your selections on it so you can verify and have a real paper trail; not some bit in a database that can be changed and most likely has been changed in the past based on grand jury testimony from one of the diebold programmers. As a programmer knowing how insecure programs are I want paper for something like this.

> "Electronic voting should be allowed but not unless it prints out a ballot with all your selections on it so you can verify and have a real paper trail"

The lack of a receipt is deliberate and prevents vote-buying. A carbon copy of your vote will open as many avenues to fraud as it will close.

Not what I was meaning. A physical ballot that you can actually put into a box that then gets counted. You don't get to keep it.

With everyone and their dog carrying around a camera in their pocket by means of a cell phone? Hardly. Anyone interested in selling their vote could easily take a photo of their ballot (with some identifying info added to the photo as well to show it wasn't a copy from a friend) to prove they voted the way they were paid.

Even if this theoretically did increase the chance of vote-buying, I think it's still a worthwhile tradeoff for the massive increase in accountability we get from it.

What does it matter what the voter takes a photo of? That has no proven relationship to what bit is stored in the machine's memory.

People could falsify the ballot to the buyer.

I've used this machine in another state, and it absolutely generates a paper trail. There is a rolling receipt going on the left-hand side of the machine behind glass that records every single button press you do, and scrolls out of view when you complete voting so the next guy can't see it.

It's not perfect, and not my preferred system, but it's pretty decent.

That receipt is actually legally iffy.

In a small community you can keep track of who voted in what order, then look at the receipt and know who voted for who.

Seems about the same as having someone watch the order the paper ballots go into the box and then having someone go back through.

Of course, we have ways of securing access to ballot boxes to a reasonable degree, and those ways are also applicable to the paper receipts.

No, since they individual papers you shuffle them.

Probably a receipt without a timestamp, and with a paper cutter would work as well.

I'd prefer to keep a receipt for myself, with a randomly generated serial number on it, so at some point I could verify that the digital record of my vote #158FA134 matches the paper receipt and who I voted for.

As others have said, this is exactly how you can start buying votes. Pull any book about cryptographic voting off the shelf and this is treated as a design flaw.

I think I missed the discussions in the thread; how is verifiability (is that a word?) a design flaw?

Consider some scenarios:

Jim is not a nice man and is very demanding and abusive and goes to vote. He wants his wife to vote for X, but he suspects that she will vote for Y. Today, she can vote for X or Y and tell her husband that she voted for Y. He has no way to prove her wrong. With receipts, he can tell her, on no uncertain terms, that she is to show him her receipt afterwards. If she doesn't produce a receipt showing she votes for X, he'll beat her. With receipts she does not have a free vote, without receipts she has a free vote.

I want to buy your vote, and will pay you €10 to vote for X. Today, you can vote for who ever you want, and tell me you voted for X and demand payment. I have no way to prove you wrong. Hence vote buying is hard to do, because you can't know if you're actually buying votes. With receipts, I can ask to see your receipt, and only pay out if you vote for X. With receipts, vote buying now becomes an actual thing that I can do.

Cell phone video of the whole process would probably be enough to placate the husband and buy a vote though.

Yes, no system is perfect, and the current system does have that flaw. But "record yourself marking the paper and putting it in the box" doesn't scale. If lots of people in an area do that, someone'll figure out what's going on.

Other scenarios that would be possible with "verify you can

Employers could now request that employees tell/prove/check who they voted for.

Unions could now tell/prove/check who their members voted for.

Churchs could now tell/prove/check who their members voted for.

Because if you can prove to a third-party how you voted, you can sell your vote, or coerce someone else to vote a certain way.

It's one of the biggest challenges in crypto-voting systems. And there are some solutions. But we need to hammer down on the "silly people, why not just print a receipt" idea really hard.

If the receipts are anonymous, they can't prove how you voted, only that you were able to obtain a receipt. So what if we increase the supply of receipts enough to destroy the value of an individual receipt? For example, voting machines could drop duplicate receipts in a bucket that voters have access to.

But I still would like to be able to have it proven to me that the way I voted matches what the system records. To me, that seems more important, but I also cannot say that it's more important than your points. Hm.

Solved problem: http://en.wikipedia.org/wiki/Punchscan

"We all" never wanted electronic voting. "We" wanted electronically-counted paper voting.

Yep. That's clearly what's happening, and given that there's instant feedback, the user would know there's a problem, and either ask for a different machine or adjust until the system represents their actual intent.

I'm not claiming it is malicious, but would they? If they can swing even 1% of the votes of people who just don't care enough, or would feel embarrassed at failing to use a machine, they can have a significant impact while having an obvious scapegoat.

Assuming it's miscalibration, how many people used this machine before the video? how many afterward?

>I then called over a volunteer to have a look at it. She him hawed for a bit then calmly said “It’s nothing to worry about, everything will be OK.” and went back to what she was doing. I then recorded this video.

That sounds like they were very likely the first person to complain. What did the rest do?

> does anyone actually think that IF this machine was altering votes, it would alter it in this fashion with a UI element tied to the alteration

I think you mean if the machine was intentionally designed to alter votes, it wouldn't do it in this fashion. The article doesn't even hint that this might be done on purpose.

> The article doesn't even hint that this might be done on purpose.

Altering is an action, not something passive. If a machine is altering votes then it must be doing that intentionally. What is happening here is the person interacting with this machine is interacting in such a way that makes the machine believe he is pressing elsewhere on the screen.

The title reads as if the machine is doing something, the reality is the user is doing something in such a way that the machine believes they're doing something else, due to poor calibration.

Except the guy is a programmer, understands calibration and tested other elements on the screen.

http://www.reddit.com/r/politics/comments/12q6wu/2012_voting...

> I initially selected Obama but Romney was highlighted. I assumed it was being picky so I deselected Romney and tried Obama again, this time more carefully, and still got Romney. Being a software developer, I immediately went into troubleshoot mode. I first thought the calibration was off and tried selecting Jill Stein to actually highlight Obama. Nope. Jill Stein was selected just fine. Next I deselected her and started at the top of Romney's name and started tapping very closely together to find the 'active areas'. From the top of Romney's button down to the bottom of the black checkbox beside Obama's name was all active for Romney. From the bottom of that same checkbox to the bottom of the Obama button (basically a small white sliver) is what let me choose Obama. Stein's button was fine. All other buttons worked fine.

Then the issue could be the system is damaged. There are these sort of touch screen systems all over England in train stations and every single one I've used has problems with the "next" button because of how much usage they see.

My point is "alter" is something that has to be done on purpose, by using the word alter they are stating the system KNOWS the user is voting for Obama and changes it to Romney.

His test is far from conclusive. Some subsection of the screen working incorrectly while other sections work correctly is entirely consistent with a miscalibrated or faulty touchscreen.

>Altering is an action, not something passive.

I disagree. If this person decided to walk away because of confusion/frustration/etc, then it would've been one less vote cast and would have altered that individuals vote. It wasn't intentional, but a software/hardware bug.

The article doesn't even hint that this might be done on purpose.

Cause it won't be spun like that at all.

But maybe the programmer didn't do this. Maybe it's something under the election worker's control like calibration settings on only part of the screen, or the spacing of the ballot, or something else user-configurable. Just a thought?

Highly doubtful that an election machine would be able to pass any standard tests that would allow it to be specifically rigged like that by an election worker. If it were built to calibrate on only part of a screen, it would be ripe for rigging by both sides.

Reply on news.ycombinator.com