Reply #32: Paper ballot good, 2 percent random audit not good [View All]

They don't get it. They keep trying to set random audit techniques, designed to catch accidental error. With fraud-prone accounting (financial, voting) you have to do a different type of auditing. A random audit of 2 percent means you have a 98% chance of success, if you rig the vote, and if you happen to get caught, you just say it is a "glitch" or a "software error."

Nope. Robust auditing is not a tiny random count. There are two ways to do it, but because one is a little complicated, and therefore will open itself up to political shenanigans and manipulation, I think we need to opt for the 100% comparison of paper to machine, which can be done inexpensively and speedily if set up correctly.

That is, if there are machines. I hear the proponents of all paper, hand counted, and the biggest thing that argument has going for it is simplicity and low cost. However, it doesn't address # of questions on ballot (we vote on about 70 things at once, each with several options; hand-count elections tend to put just one or two questions on the ballot).

Good, paper ballots.
Two more things needed:
1) Immediate interim plan in place by the primaries, for manual count, including optical scan counties.
2) Long term permanent auditing solutions that are appropriate to the task at hand.