What are you proposing to compare to the online results? The paper copies of their online votes? Oh wait.
Yes, you could have meat space sample groups. You could also break the constituencies or regions into sub sets and compare return on those.
That would be totally useless.
Letâs say you find a discrepancy between the overall result between online voters and paper voters. Now what?
- Pundits would line up to say âOf course they voted differently - they have demographic differences that led them to vote one way or the other. Those who vote online tend to be younger, and to live in rural areas where the polling station is less convenient to get to, to name just a few differences.â
And theyâd be right. You canât prove anything by finding differences between the votes of one group of people, and the votes of a different group of people. The only comparison that is useful is between the votes of one group of people as counted by a machine you canât fully trust, and the same exact votes of those same people counted by a trusted method.
You canât even compare how they voted on Monday to how they voted on Tuesday. The only thing that proves the untrusted vote counting method was right, was to count the exact same body of votes by a trusted method.
- Even if you could prove malfeasance rather than just random difference between the two groups (which, I emphasize, you could not), what would you do? Having proved that the only copy of their votes is unreliable, you have nothing to go back and count.
Itâs funny because you can get a representative sample from 500 people. If you can wrap your head around that then you will be closer to understanding market research. Just ask Nate Silver.
Sending results to multiple servers could be a good way to approach your problem. But, half of the US states donât even bother to check if they match up. So this doesnât seem like itâs a real problem to begin with.
All of this is pointless though, as voting for a person (vs an issue) is basically giving away your democratic rights for 4 years.
1 Like
I will reiterate: None of this would help in the least.
- A representative sample can show if there is a statistically significant deviation between internet voters and your sample. It does nothing to explain the reason for the difference.
Did the people just vote that way because the demographics of internet voters differs in some way from that of your sample?
Did 20 or 30 of your sample voters realize they were in the âcorruption detectingâ control group and collude to change their votes and produce a false positive for corruption (or to cover up actual corruption in the software that handled millions of votes)?
Is the software counting the votes lying?
- Keeping the results on two servers does absolutely nothing. The initial vote comes in a single message, from the voterâs client device (which could be hacked) running voting software (which could be malicious) to the election bodyâs server (which could be hacked) running vote counting software (which could be malicious). By the time you have your hands on the vote to distribute it to two different servers, it could have been altered by a hundred different complex pieces of software you have no hope of auditing.
A piece of paper with ink on it, in a locked metal box that has been observed at all times by at least four people representing rival parties, is trustworthy in a way no software system ever can be.
1 Like
Your arguments can also be applied to the physical voting space.
My reality involves moving past physical voting in order for democracy to flourish.
I guess we will have to agree to disagree here. Thanks for your input!
How can my arguments be applied to physical voting? What undetectable computer software method will destroy the original paper ballots such that they cannot be hand counted to validate the computerized count?
Not a sample statistically similar to the voters in the particular precinct - the exact votes present in the box underneath the scanner, that can only have gottern there by going through the scanner.
1 Like
Here in San Francisco we have had many scandals where people give their mail in ballots to neâerdowells and get paid $20 or so. Any internet scheme is equally vulnerable.
Links? Recent?
SF voter fraud example:
Something more than allegations? Anything systemic? Anything that can be compared to Florida-style voter suppression?
1 Like
To be devilâs advocate, how about this:
Letâs say a mail voting system is contingent on some personal information.
Letâs also say that, I dunno. The year is 2016, and you can purchase huge volumes of personal information off the black market.
And that a cheap laserjet printer can reproduce the mailing form for voting mailers.
Reckon that voter turnout is realistically in the 30th percentile on non-presidential election years.
I figure that could swing a close race in a midterm election. Hell, we should be checking for that now!
You would dupe a bunch of ballots and a giant shitstorm would result. To systemically fuck the system is much more difficult than with softwareâŚ
Yeah, an error log is checked less often than physical mailers. This assumes the mailroom clerks canât be bought.
Thatâs probably just as cheap as the data.
Really the hard part would be figuring out who they were. How hard is that do you figure?
/tinfoil hat
/not really
/not that hard.
Basically, hereâs what would happen with a zillion fake ballots and ganked personal information.
-
Some percentage of those ballots would be duplicates, as the unlikely-but-not-impossible voter they were attached to decided to vote. (Which happens more often in mail-vote states anyway.) One duplicate ballot is an anomaly, two is suspicious, three starts an investigation. Youâve got some portion of a zillion. RESULT: massive investigation.
-
Some percentage of your internet-bought personal information conflicts with what the registrar believes to be true about a given voterâs information. RESULT: massive investigation.
-
The Post Office notices that a weird number of ballots are coming in from non-household pickups. (Alternatively: you provoke a million police calls trying to sneak ballots into residential mailboxes for pickup.) RESULT: massive investigation with special guest The Postal Inspectors.
-
The bean-counters notice that a 30% election turnout has suddenly become a 50% election turnout, and correlates the increase to other weirdnesses. RESULT: massive investigation.
-
Some percentage of a zillion voters are confused by the receipt mailed to them saying that they have voted, when they did no such thing. RESULT: massive investigation.
-
Your cheap laserjet printer isnât quite up to the task of defeating the modest but non-trivial anti-forgery devices built into the official ballot. (Not all of which youâre guaranteed to have noticed.) RESULT: massive investigation.
-
Your forged witness identification data (which many jurisdictions require) is faulty, or is fantastically improbable from a metadata standpoint. (People donât have their ballots witnessed by someone who lives fifty miles away. Except when they do. How many people witness exactly one ballot? How many people witness fifty? What is the distribution of those numbers? How many witnesses themselves fail to vote? Etc.) RESULT: massive investigation.
-
Probably a lot more that can go wrong with a crime of this complexity. But really, even just the first thing will doom you completely.
-
It would have been easier to campaign.
As for buying off the guy in the mailroom, sure, except itâs fifty guys in fifty mailrooms. Or for a federal race, five hundred guys in five hundred mailrooms. (Every person added to your criminal conspiracy is another point of failure.) At this point, you might as well just wait to see who wins fairly and then bribe that person.
The reason vote-by-mail is a good system is not because it canât be fucked withâit can. Itâs not even because it canât be beaten at the level of a single vote here or thereâit can. Itâs because itâs vastly less vulnerable to vote-stealing shenanigans than the octogenarian glancing at your driverâs license and waving you towards a black-boxed Diebold⢠machine running Excel 95 with a taped-over USB port. And as a bonus, the massive investigation can be ongoing before the election is done, which minimizes the collateral damage. By contrast, you could much more easily disrupt a walk-in election by calling in a bomb threat (or actually bomb) to the single point of failure that is a polling place.
And in the meantime, mail voting makes the ballot effortlessly accessible to anyone who registers for one. There are people who will tell you that voting should be a test of your logistical skills and personal fortitude, but fuck 'em.
2 Likes
The forms are not on standard paper stock and have non standard pre creased folds as well. I imagine it would not be hard to find who ordered the paper stock to do it when done in volume to be significant enough to make a difference.
1 Like
Youâre basically making the case for mail-in voting, though. Look at what supposedly happened in that link:
- Shadowy figures coerced or stole ballots from defenseless senior citizens
- Said shadowy figures identified themselves as being from a particular campaign
- Complaints were registered by an organization affiliated with the other campaign
- Investigations commenced by both the voting officials and the District Attorneyâs office
- Nothing came of it (http://www.sfchronicle.com/bayarea/article/How-and-why-Peskin-won-race-for-District-6616261.php)
- The race turned out not to be close (by a factor of about 10,000 times more than the shadowy figures reported in one and only one place could account for)
- The challenger didnât contest it
In other words, an investigation happened in real time to identify any lawbreaking. Political measures were proposed to remedy the situation before the election was even over, while there was still time to do something about it.
Now, if you want, you can assume that all allegations are true no matter what the DAâs investigation turns up, but at that point youâre not really saying that the problem was mail-in ballots, so much as a totally corrupt or incompetent government.
Again and again you have to come back to this single point: the test isnât whether the system is absolutely bullet-proof, because no voting system is. Crimes that are physically impossible to commit donât need laws punishing them in the first place! The test is whether the system is better than the alternative. If youâve lived in the Bay Area for a while, youâll know that ominous murmurs about bribes and chicanery and electoral misdeeds coming out of Darkest Chinatown with its inscrutable voters and weird alien cultural practices (etc. etc. etc.) are nothing new. Theyâre about as valid when applied to vote-by-mail as any other issue.
2 Likes
What if the ballots went through two op-scanners in order to count? One provided by dems another by reps. Since they donât trust each other they would be forced to keep a legit count as possible.
After the voting ends, compare the results from both scanners, and if thereâs a discrepancy, not only can you do a recount, but youâll know exactly which side cheated.
This topic was automatically closed after 5 days. New replies are no longer allowed.
