Grok the Vote

Now I feel terrible for using this title. Ech!

That aside, here is my solution to voting machine irregularities. As a non-trusting digital citizen, I don't want to place my life in the hands of machines. I don't trust them. I don't trust humans either, so it's really a toss-up as to whom I trust less.

The problem with most current computerized voting machines is that they don't leave any paper record. The problem with most current non-computerized voting machines is that they are hard to understand and hard to count because they have to be fed into some kind of tabulation machine. Let's address the second problem first.

For the record, after the 2000 election I was in a human interface design class (fancy word for usability class) and we looked at the "Butterfly ballot." It's poorly designed, really poorly, in fact, but there's more than that. It's a poorly designed example of a poorly designed system, because I've voted using ballots of the type in question with no problem, but the Florida ballot was really difficult to understand. If I, who knew what to expect, had difficulty with it, I can only imagine what people who were less savvy to the process might have thought.

Why the hell not design a good system? I don't know. Probably money. So the first problem is easily solvable: design a good system, using good design principles.

Ha ha. Okay, got you. If you think that designing a good system is easy, then no offense, but you need to look around at the Internet. Find a design company, a web design company even, look at their web site, and tell me all the ways it violates rules of good design and usability. I'll wait.

While the foolish are doing that, let's look at the second problem, counting manual ballots. Hanging chads should convince us of this problem too. So why not use computers?

Well, then there's the third problem: computer voting machines suck. The solution: use them like non-computerized voting machines. The computer helps the user vote. It prints out a copy of the vote, which the voter then takes over to the ballot box just like a non-computerized vote. Rinse. Repeat if you live in Chicago.

How will this solve the problem, you ask? Well, provided the design of the voting system is usable (the people who believed that would be an easy task have all either committed suicide or they're still looking), even marginally better than existing systems, it can help the user vote. It can then either simplify matters by counting the vote itself or leave that up to a tabulating computer somewhere else (probably a better solution, since it's easier to tamper with a computer in place, especially if it has to be hooked up to the network, than a central computer). But since the computer prints the vote, it can print both a human-readable representation and then a little bar code at the bottom for the big tabulation computer back at the base.

So we get the best of both worlds. We have computer standardization to keep hanging chads from fouling up the count, we have a (hopefully) slightly more easy to use system for confused voters, and we have a paper trail which will be easier to recount anyway because, like we said earlier in the sentence, no hanging chads.

I know I have problems trusting the vote. But we don't live in Zimbabwe, so I trust it enough to vote. And this system would not require me to trust it any more than I already do. It's the ideal system because it combines the positives of several systems which already work acceptably well.

And this (and money), children, is why it will never be implemented. There will never be any standard of voting. Sometimes I curse our Federal system.