3 ms·
Is the problem "people not filling in circles properly" or is the problem that there are other errors. If I'm not mistaken you're referring to the hanging chad
by godelski 6y ago
Is the problem "people not filling in circles properly" or is the problem that there are other errors. If I'm not mistaken you're referring to the hanging chad issue. The reason it was an issue is because it is hard to see the chad. We're in tech, this should be unsurprising that machines don't work with 100% efficiency. If your machine (human or computer) is marking a ballot it would be unsurprising if the mark is transposed or warped or incomplete. Instead we should design a system that accounts for these errors in a clear and concise manner.
- the_snooze 6y ago>Instead we should design a system that accounts for these errors in a clear and concise manner. We already have these systems in deployment. My precinct uses optical-scan paper ballots where any mark inside the bubble is valid. You can fill, dot, cross, check, whatever---the machine will count it as a mark. If you have a stray mark that results in an overvote, the machine will reject your ballot, then prompt you to either correct the overvote or override the error. The scanner will also accept ballots fed in any orientation as long as it's not folded or wrinkled. It's surprisingly robust and user-friendly.
- godelski 6y agoYes, basically my response was about how the parent's issues are solvable and don't make things a non-starter.
- owenmarshall 6y agoAccounting for errors isn’t the problem. Teaching people to use and trust an insanely complex system is. I’ve yet to hear any proposal for an electronic voting system that offers any advantages over current systems AND doesn’t require a PhD in mathematics to grasp.
- godelski 6y agoI'm still mixed on electronic voting, but I do think it should be further researched. The simple answer I've seen is giving people a hash or code that can be used to verify their vote in a database. Obviously you have to trust the database, but you also have to trust the people counting. I do think there are enabling technologies like zero-knowledge proofs and locally differential privacy that do help with many of the problems, but my understanding is that neither of these is mature enough for use in voting, yet. But as to general advantages, I think there's two major ones I see. 1) Electronic voting better enables access to voting since people can vote in the comfort of their homes (especially relevant in a situation live covid). 2) Enables better research about candidates. I live in a state with mail in voting and how I vote is with my ballot in front of me while I research all the candidates. In fact, I spend several days voting (sure, not everyone will do this but it makes it easier). Links to official campaign pages or voter guides (we have this in my state with the ballot) would be helpful (and encourage candidates to create them! Because often they don't even have a website, at least on local levels). As a minor advantage I do see having the ability to perform different voting testing and better answer questions to things like ordinal and cardinal voting systems (by participation not forced A/B testing). Just because enabling technologies aren't there yet doesn't mean we should shutdown the conversation about how to solve those technological challenges.