45 ms·

How do you tell if something goes wrong? Well, if you hand-count a voting machine, and it gives you a different result from the machine, then you know something

by beat 8y ago

How do you tell if something goes wrong? Well, if you hand-count a voting machine, and it gives you a different result from the machine, then you know something went wrong. And if you hand-count a bunch of machines and they all have the same misbehavior, then it's time for throwing out the machine count altogether.

XorNot 8y ago
In Australia it's all hand-count, with machine count for verification. Specifically ballots are sorted by volunteers on election night to do the preferential voting, and then the stacks are counted and checked by essentially bill-counting style machines. Discrepancies mean someone else does a count and recount. The key to all of this is involving as much of the public as possible though: democracy is government by the people for the people, the people need to be involved in its operation. The only way you dispel the idea of rigging is by ensuring a large plurality of your population was involved each time and would attest to their neighbors that they saw no misbehavior. Electronic voting is based on trying to take people out of the process - it misses the point.
- Klathmon 8y ago
  This is the point I always try to hammer home. The inclusivity of paper voting is amazing. It doesn't matter if you are young or old, black or white, highly educated or illiterate, rich or poor, just about everyone can count ballots. And if you think someone is trying to fuck you over, you too can go count along or watch that ballot box all day to ensure nobody is tampering with it. Get a few hundred people in the same room that all don't trust one another and all count along. That's how you run an election.
  - briandear 8y ago
    ..and we verify that everyone actually voting is who they say they are.
    - beat 8y ago
      False voting is a severely overrated problem. It doesn't scale. One person might be able to make a fake vote/registration, or even a few of them. But enough to materially affect the outcome of an election? That requires an organized effort with likely thousands of people involved. Such conspiracies are impossible to keep secret. Don't let something that feels wrong steer you away from much more severe and likely problems.
      - travisoneill1 8y ago
        A large conspiracy would be caught instantly. But it doesn't have to be a conspiracy. If people know they can vote multiple times and can't get caught (and w/o checking ID I don't see how they can) a certain number will decide to do so. I can't say how often this occurs, but how can you even measure it to determine if there is a problem without some sort of ID verification system at the polls?
        richmarr 8y ago
        I don't think it's a problem that happens, though. If you're going to fraudulently rig an election it's vastly cheaper to do so by rigging the polling station than it is by faking the voters. But by introducing voter ID systems you are very likely to skew elections away from the people least likely to possess accepted ID, i.e. the lowest socio-economic groups.
        travisoneill1 8y ago
        > I don't think it's a problem that happens, though But how do you know. That's the issue. If P is the probability of getting caught, and C is the number of people caught then N ~= C/P. With our current system it seems like P is close to 0 so even with C = 0 there could be a lot of successful attempts. Rigging a polling place is not viable because it would require coordination between multiple people and leave a clear mathematical trail after. So I would say that, including detection after the fact, that P > 0.5 and that the fact that it hasn't been caught means that N = 0. > But by introducing voter ID systems you are very likely to skew elections away from the people least likely to possess accepted ID, i.e. the lowest socio-economic groups. This is pretty much separate from my point but I am also skeptical about this. If someone can get to the polling place, stand in a line, and do paperwork to vote, then why could they not do pretty much the same thing to get an ID?
- foobarbazetc 8y ago
  As an Australian living in the US now (where voting is ... interesting) I now recognize how awesome the Australian system is. The other great thing not mentioned here is that elections are run by the AEC, a non-partisan commission. They are also in charge of redistricting. So you don’t have gerrymandering and you don’t have BS going on like what happened in Georgia where they “accidentally” wiped all the data 3 times when subpoenaed by the FBI. Add to that compulsory voting, elections always run on weekends, and a thousand other small things and it’s like one country wants you to vote and the other does not.
pjmorris 8y ago
> Well, if you hand-count a voting machine, and it gives you a different result from the machine, then you know something went wrong. And you'll probably have to ask someone who's not in the room about what went wrong. With human hand-counts, the person you need to ask is already sitting there. No mystery, no confusion.

by beat 8y ago

XorNot 8y ago

In Australia it's all hand-count, with machine count for verification. Specifically ballots are sorted by volunteers on election night to do the preferential voting, and then the stacks are counted and checked by essentially bill-counting style machines. Discrepancies mean someone else does a count and recount. The key to all of this is involving as much of the public as possible though: democracy is government by the people for the people, the people need to be involved in its operation. The only way you dispel the idea of rigging is by ensuring a large plurality of your population was involved each time and would attest to their neighbors that they saw no misbehavior. Electronic voting is based on trying to take people out of the process - it misses the point.

Klathmon 8y ago
This is the point I always try to hammer home. The inclusivity of paper voting is amazing. It doesn't matter if you are young or old, black or white, highly educated or illiterate, rich or poor, just about everyone can count ballots. And if you think someone is trying to fuck you over, you too can go count along or watch that ballot box all day to ensure nobody is tampering with it. Get a few hundred people in the same room that all don't trust one another and all count along. That's how you run an election.
- briandear 8y ago
  ..and we verify that everyone actually voting is who they say they are.
  - beat 8y ago
    False voting is a severely overrated problem. It doesn't scale. One person might be able to make a fake vote/registration, or even a few of them. But enough to materially affect the outcome of an election? That requires an organized effort with likely thousands of people involved. Such conspiracies are impossible to keep secret. Don't let something that feels wrong steer you away from much more severe and likely problems.
    - travisoneill1 8y ago
      A large conspiracy would be caught instantly. But it doesn't have to be a conspiracy. If people know they can vote multiple times and can't get caught (and w/o checking ID I don't see how they can) a certain number will decide to do so. I can't say how often this occurs, but how can you even measure it to determine if there is a problem without some sort of ID verification system at the polls?
      - richmarr 8y ago
        I don't think it's a problem that happens, though. If you're going to fraudulently rig an election it's vastly cheaper to do so by rigging the polling station than it is by faking the voters. But by introducing voter ID systems you are very likely to skew elections away from the people least likely to possess accepted ID, i.e. the lowest socio-economic groups.
        travisoneill1 8y ago
        > I don't think it's a problem that happens, though But how do you know. That's the issue. If P is the probability of getting caught, and C is the number of people caught then N ~= C/P. With our current system it seems like P is close to 0 so even with C = 0 there could be a lot of successful attempts. Rigging a polling place is not viable because it would require coordination between multiple people and leave a clear mathematical trail after. So I would say that, including detection after the fact, that P > 0.5 and that the fact that it hasn't been caught means that N = 0. > But by introducing voter ID systems you are very likely to skew elections away from the people least likely to possess accepted ID, i.e. the lowest socio-economic groups. This is pretty much separate from my point but I am also skeptical about this. If someone can get to the polling place, stand in a line, and do paperwork to vote, then why could they not do pretty much the same thing to get an ID?
foobarbazetc 8y ago
As an Australian living in the US now (where voting is ... interesting) I now recognize how awesome the Australian system is. The other great thing not mentioned here is that elections are run by the AEC, a non-partisan commission. They are also in charge of redistricting. So you don’t have gerrymandering and you don’t have BS going on like what happened in Georgia where they “accidentally” wiped all the data 3 times when subpoenaed by the FBI. Add to that compulsory voting, elections always run on weekends, and a thousand other small things and it’s like one country wants you to vote and the other does not.

pjmorris 8y ago

> Well, if you hand-count a voting machine, and it gives you a different result from the machine, then you know something went wrong. And you'll probably have to ask someone who's not in the room about what went wrong. With human hand-counts, the person you need to ask is already sitting there. No mystery, no confusion.