4 ms·

No you wouldn't. Many people have full-time jobs trying to predict elections, and they are not nearly accurate enough to zero in on a small number of essential

No you wouldn't. Many people have full-time jobs trying to predict elections, and they are not nearly accurate enough to zero in on a small number of essential districts. They don't even do very well on the state level. Also, the election landscape shifts constantly, so you can't do this work very far in advance. How long will it take you to set up a system to "hack" Michigan's paper ballots? Even if you somehow perfectly predict the 12 counties necessary to flip the result a year in advance, which again is impossible, can you really infiltrate those counties thoroughly enough to be undetectable? Edit: "your only full time job was to guarantee an election win for a candidate" What you're describing is a campaign manager. You'll notice that during presidential elections, the candidates always have different theories of the electorate and will pick different places to focus their efforts. So either: the losing campaign is always so massively incompetent that they don't even know which states to target, even though you say predicting counties is pretty easy when "we have computers", or elections are actually quite difficult and unpredictable.

everdev 8y ago
Sure, but I think in this theoretical example of someone trying to manipulate an election, they'd pick the top X swing states, which are pretty well known. Of those, they'd pick the top X most populated cities in each swing state. Again, I have not heard anyone say this happened, but I think it's reasonable to assume that the potential attack surface might be fairly small (a few heavily populated counties in a few states) if someone were to want to try. FiveThirtyEight has county by county election data and it's widely discussed on election nights how certain counties vote heavily R or D. You wouldn't be flipping any counties political preference, just minimizing the win % in urban areas might be enough.
- burkaman 8y ago
  Now you're trading a prediction problem for an implementation problem as I explained in another comment. A few heavily populated counties is not a small attack surface. You need to hack hundreds of airgapped optical scanners in each state (and they all use different machines), not too much or too little, because then you'll trigger a recount using the real paper ballots. You still need basically perfect prediction to win and stay undetected, and you're up against big election departments with tons of employees and lots of integrity checks.

everdev 8y ago

Sure, but I think in this theoretical example of someone trying to manipulate an election, they'd pick the top X swing states, which are pretty well known. Of those, they'd pick the top X most populated cities in each swing state. Again, I have not heard anyone say this happened, but I think it's reasonable to assume that the potential attack surface might be fairly small (a few heavily populated counties in a few states) if someone were to want to try. FiveThirtyEight has county by county election data and it's widely discussed on election nights how certain counties vote heavily R or D. You wouldn't be flipping any counties political preference, just minimizing the win % in urban areas might be enough.

burkaman 8y ago
Now you're trading a prediction problem for an implementation problem as I explained in another comment. A few heavily populated counties is not a small attack surface. You need to hack hundreds of airgapped optical scanners in each state (and they all use different machines), not too much or too little, because then you'll trigger a recount using the real paper ballots. You still need basically perfect prediction to win and stay undetected, and you're up against big election departments with tons of employees and lots of integrity checks.