Some Tennessee voters say they touched one candidate’s name on a voting machine but saw a different name selected. State election officials say unintended touches can cause that to happen.
Just use a scheme where it selects the closest candidate to the “touch”, then offset the calibration such that it thinks a touch on #2 is #1. Then tapping on #1 will still select #1 because it’s the closest, while tapping #2 will select #1 because the tap thinks it’s closest to it.
So if the Republican candidate is always #1, then I’d bet it’s deliberate. Actually, all it needs is for the Democratic candidate to be #2 to split the Democratic vote, though then it runs the risk of splitting the Republican vote as well (assuming the “miscalibration” is linear, if it’s non-linear, then it could get “more correct” the further along the screen it goes).
If it was unintentional, you’d expect to see no specific party benefiting from the errors overall. Ie, in some locations it favours red, some blue, and in others one of the other candidates. Though even with this, if the distribution ends up such that it favours blue in blue strongholds and red in battleground areas, then that’s still fishy.
Seems like a good reason to maybe have random ordering? Then it would need to be reading the content to be able to properly change things instead of something a little less obvious like screen calibration.
Absolutely, then you could do statistical analysis on the positions to see if any position is biased (for or against) and detect the problem and recalibrate the screen and no one candidate would gain that much of an advantage (well, candidates that weren’t going to get many votes would benefit, but not to the point where they’d be able to steal an election).
But that assumes they are being run in good faith.
Wouldn’t be hard to game.
Just use a scheme where it selects the closest candidate to the “touch”, then offset the calibration such that it thinks a touch on #2 is #1. Then tapping on #1 will still select #1 because it’s the closest, while tapping #2 will select #1 because the tap thinks it’s closest to it.
So if the Republican candidate is always #1, then I’d bet it’s deliberate. Actually, all it needs is for the Democratic candidate to be #2 to split the Democratic vote, though then it runs the risk of splitting the Republican vote as well (assuming the “miscalibration” is linear, if it’s non-linear, then it could get “more correct” the further along the screen it goes).
If it was unintentional, you’d expect to see no specific party benefiting from the errors overall. Ie, in some locations it favours red, some blue, and in others one of the other candidates. Though even with this, if the distribution ends up such that it favours blue in blue strongholds and red in battleground areas, then that’s still fishy.
Seems like a good reason to maybe have random ordering? Then it would need to be reading the content to be able to properly change things instead of something a little less obvious like screen calibration.
Absolutely, then you could do statistical analysis on the positions to see if any position is biased (for or against) and detect the problem and recalibrate the screen and no one candidate would gain that much of an advantage (well, candidates that weren’t going to get many votes would benefit, but not to the point where they’d be able to steal an election).
But that assumes they are being run in good faith.