Hacker News Reader

Comment by TheAdamAndChe | Hacker News Reader

As long as potentially hackable voting machines are used in elections, the anxiety about hacking will be there and will eventually be exploited by politicians. If we want to fix the issue, we really just need to move back to hand counting paper ballots.

View on HN

Dominion voting machines leave an auditable paper trail and thus do support hand-counting of paper ballots, which was done as part of the recounts in disputed states. No discrepancies beyond the margin of error of hand counting were found in those recounts vs the electronically reported results.

You can see how they work in this video: https://www.youtube.com/watch?v=b9BDiO3JGTs

I'm not saying they were hacked. I'm saying the layman needs to be confident beyond a reasonable doubt. A normal non-security professional needs to understand and have confidence in the system they use to vote. Most people don't even know how to troubleshoot computer issues beyond rebooting, let alone understanding complex IT security or auditing measures put in place for these machines.

If you watch the video you can see that a layman would easily understand how it all works. It literally prints a ballot where you can verify your choices, and you drop it off.

This is pretty much the gold standard: a paper trail that can be audited every step of the process.

Still not sufficient though, you only need to hack enough to get outside the automatic recount margin. You also need hand-recount of random samples in all elections.

Of course, if someone hypothetically hacked an election even more brazenly to get outside of that margin, the result would be even more unexpected and likely to be manually contested.

You only need to recount heavily when the election is near 50%, because the higher the margin of victory, the fewer random samples you need to detect fraud. This is what Georgia actually does. The automatic margin is for runoff elections.

Nothing can "fix the issue" when the issue is primarily that some candidates will not accept a losing outcome.

Had this election been entirely hand-counted, the same claims of fraud would have arisen. When a candidate declares in advance of an election that he can only lose if there is fraud, you can be sure he will act in bad faith if he loses.

Incredulity was all he needed to claim fraud - the numerous conspiracy theories explaining the fraud would inevitably surface after the fact. If it's not Dominion voting machines, it's dead voters, boxes of ballots showing up, and hand wavy claims of the opposition being in charge of the counting and abusing that role.

Hand counting can be manipulated too [0], so the idea that we can just move back to it and save our democracy is not accurate (or at least not a complete solution). In other words, dishonest politicians will find ways to exploit our fears even if we use paper ballots.

[0] https://www.sciencedirect.com/science/article/abs/pii/S02613...

You linked to an abstract of the Russian vote manipulation behind a paywall. That abstract says nothing specifically about hand-count being manipulated.

Hand counting remains the golden standard for voting.

All allegations against Dominion voting machines were pretty much disproved by subsequent manual vote courting. Again, the machines leave a paper trail that allows manual counting for verification. That makes simple handcounting without any assistance more prone to errors and less paper trail as well.

In what world would manual counting have less paper trail?

What is for sure - with voting machines, there is less validation. Hand counting means - all ballots are counted always (by volunteers with all parties invited) - whereas voting machines means a spot-test (in some states), and only a full manual count is when a recount is demanded.

All this to say - I think Dominion did a thorough job - but this all only happened because the election was close. What about those where the election isn't "close"?

Voting by hand takes days, which is why it is only done automatically for close elections.

However, a losing candidate may request a recount after the election. In most states, if the margin is outside of a stated threshold, that candidate must pay for the recount. If there is a discrepancy between the two counts, then a hand recount is performed.

If an election isn't close, then the respective Secretary of State will audit (by hand) a random selection of precincts to verify vote totals match, but this is a security check and generally takes place weeks after the election.

Voting by hand takes days - in the US. Not in Finland, France or the many other countries where voting is done by hand 100% successfully for decades with timely results.

Here's the conundrum - what constitutes what's "out of threshold"? If a voting machine, either by design or flaw, pushes a close vote outside the threshold - then that's a way to bypass the checks & balances and can be exploited by the unethical.

Scale matters. 160 million votes were cast in the U.S. in each of the past 2 elections.

36 million votes were cast in France in the last election.

The U.S. is capable of counting 36 million votes by hand overnight as well, and most votes were once counted by hand in the U.S.

What constitutes what's "out of threshold"? If a voting machine, either by design or flaw, pushes a close vote outside the threshold - then that's a way to bypass the checks & balances and can be exploited by the unethical.

That paragraph is disingenuous. A vote that is outside the threshold for an automatic recount is not a close vote; the margin between candidates is thousands of votes (or more).

If a voting machine has a design flaw or other flaw, that would have been discovered during one of the several inspections and trial runs it was put through before being certified for use. Moreover, many states now require paper receipts of all ballots cast (as a result of Russian hacking of election machines in 2016), so if there is any suspicion of manipulated results, the human-legible ballot receipts can be tallied. States with these types of printed ballot receipts will audit the electronic tallies against hand-counts of the printed ballots on a random precinct-level basis.

You scale by district. You just get more people involved in the counting per ballot location.

It's a scalable method. Yes, the reconfirmation at the regional levels are important.

Your faith in machines is questionable. I leave you with this obligatory xkcd - it's still completely valid: https://xkcd.com/2030/

In my state, the machines print out paper ballots in legible English saying exactly who I voted for.

I don't trust the machines, but I do trust the people doing the counting, especially since the counting process is observed by both sides.

No one ever claimed that hand-counting is perfect. Is isn't. For example, all of the Chicago ballot boxes that famously ended up in Lake Michigan in the middle of the last century.

But hand-counting works because it's virtually impossible to coordinate fraud widespread enough to actually make a difference. Whereas, if your entire voting system is computerized and all the computers are linked together, causing widespread changes are much more possible.

Good thing hand counting can be (and is) used with the Dominion voting machines to verify the electronic result, then.

Certainly a good thing!

Do let me temper that slightly :

Some states didn't or didn't fully have a paper trail quite yet https://ballotpedia.org/Voting_methods_and_equipment_by_stat...

Also, be aware that Dominion voting systems actually acquired the old "Diebold Election Systems", who were not so secure a couple of years back. (though things may have changed by now, of course) https://en.wikipedia.org/wiki/Premier_Election_Solutions

Note that it looks like the states on that list that have “DREs without VVPAT” (electronic machines without a paper trail) are Indiana, Kentucky, Louisiana, New Jersey, Tennessee, and Texas on that list, along with some DREs without VVPAT in Kansas and Mississippi.

Since Trump won 7/8 of those states, a conspiracy theorist might suggest that Trump’s team successfully hacked the electronic machines in those states that have no paper trail.

However, there’s really very little reason to believe there was fraud in this election - but those republican states (and New Jersey) should replace their DREs without VVPAT by the next election so that they have a secure paper trail for auditing.

So the machines are redundant?

If you think that voting machine hacking is a genuine issue, then you should want Giuliani to shut up and go away ASAP. Having your arguments mixed up with crazy pants conspiracy theories is a one way ticket to people deciding that you’re also a crank to be ignored.

That doesn't give anyone the right to say baseless claims and present them as facts on front of millions of people.

Even that won't stop the lies about the election being stolen. Pennsylvania, Michigan and Georgia are all paper ballot states [0] counted by machine but preserved for manual recounts. The lie just became that there was either ballot stuffing at some point, a point which changed every time a claim was disproven, or that the Dominion tabulation machines changed some Trump votes to Biden votes. Even Members who were elected in that very same election were saying it was fraudulent, just not their votes of course.

[0] https://ballotpedia.org/Voting_methods_and_equipment_by_stat...

Yes, this election was the most secure due to the paper trail that mail-in ballots provide. I would trust that more than the bits of a closed, black voting box

For a community (HN) so focused on technology and the future, I'm surprised to see support for antiquated technology such as paper ballots.

Does the future of voting not belong to online remote voting, like almost everything else? You can even buy a house from your phone now. In the future, we will know the results of the election as soon as the online polls close.

The future is paperless.

No it isn't.

https://www.youtube.com/watch?v=LkH2r-sNjQs

I disagree with your conclusion. Voting does not need to be made more convenient beyond preventing obvious voter suppression and making election day a federal and/or state holiday. Additionally, a digital compromise of a fully digital system could compromise all votes. If the system has a nondigital artifact it can reference as a source of truth, then it can be validated, verified, and restored to truth.

A federal holiday would be fantastic. The main benefit is not getting more people to vote, but getting more people to watch the polls & counting. The more opportunity that people have to watch, the more faith people will have in the system.

With the election process, optics is as important as the actual security. It has to seem simple and secure for most people to accept the results easily. We saw recently that even that does not always work but I have hope it was an aberration. Can you imagine the fallout if everything was done digitally instead, considering the average person’s understanding of crypto and “hackers”?

I'm surprised to see support for antiquated technology such as paper ballots.

"Antiquated" is a loaded word that smacks of ageism. Restaurants are antiquated, too. Do you never go out to eat?

Does the future of voting not belong to online remote voting, like almost everything else?

Are you trying to say that every other country votes paperless? Because that's simply not true.

You can even buy a house from your phone now

You have obviously never bought a house, and certainly not one from your phone. Just because there's an ad on YouTube doesn't make it true.

In the future, we will know the results of the election as soon as the online polls close.

Why is that necessary? What benefit does it provide?

The future is paperless.

People have been saying that for over a hundred years. Yet, here we are, still with a durable, recordable medium that does things that paperless methods can't.

> Restaurants are antiquated, too.

The modern restaurants is kind of recent, right? It was invented around the same time as the bicycle.

> For a community (HN) so focused on technology and the future, I'm surprised to see support for antiquated technology such as paper ballots.

I'm not. No one understands as well as technologists how often “novel technology” represents a regression from the tried and true, and how much the cry of “new” is used to sell defective crap.

> You can even buy a house from your phone now.

This is technically partially true in that a lot (not all) home purchase and finance paperwork can be done via electronically signed documents, and if you have a tolerance for extremely bad UI you can sign those on a phone. Typically, you’ll still need some wet signatures (notarized even) at the end of the process.

But that’s, even if it was completely accurate, a different problem domain than voting. (Both have concerns for assurance, but only one also has nonattribution as a critical goal.)

Once you know how the sausage is made, sometimes you don't want to eat sausages.

I actually do think that digital voting is theoretically possible, if done cautiously and in an open source manner. (it needs to be fully open source to be verifiable at all)

A remaining problem is that -in an election system- every voter needs to be able to check and perceive that the election was honest. That could still be a bit tricky in a society where not everyone is a programmer.

One big barrier to online voting is that you don't want people to be able to be forced to show their votes which makes any kind of verification immensely difficult (you have to accomodate people needing to show a fake verification) because otherwise you leave things open to voter intimidation and vote buying.

Another is that by going online you instantly open the attack surface up massively. Currently to steal an election you have to physically move a large number of people to a broad set of voting sites or send in a lot of forged mail votes. Those both require physically doing something in the US where an online system lets anyone in the world potentiall attack the system.

Pretty much every noted cybersecurity expert disagrees.

Reply on news.ycombinator.com