You mean the election where 500 voters fucked up their ballots and didn’t log the vote they wanted to log, or where 200 voters didn’t get to vote because they screwed up their registration? I actually supposed to care about 17 voters out of 5+million? The number is so far below the margin of error it doesn’t make logical sense to care about it. You would need an accuracy rate of 99.9997% to make those 17 votes the driving force behind the result. I wonder what has a similar accuracy rate. Wow, this does. Well, not quite. Six Sigma looks for 99.99966% accuracy, which I suppose is close enough.
Do YOU think that random people using machines once every 4 years in a church basement is going to match the defect rate of the highest quality controlled manufacturing processes in the world? I don’t, and it isn’t even a race.
The slop factor around an election is so high that the tiny percentage of illegal votes is swallowed up 1000x over by people just screwing up.
What you want to do is trade 17 statistcally meaningless votes for thousands of legal voters getting turned away because their ID expired, and thousands more having to sit at their kitchen table trying to figure out how to manage a trip to the DMV for an ID they never needed before. A trip that’s especially hard without a car because the DMV doesn’t particularly cater to the needs of pedestrians, being the Dept. of Motor fucking Vehicles. And it costs millions of dollars.