What are the audits like? Do they pick a certain number of ballots from certain races to recount by hand?

so long as there is a paper trail... agreed, it probably works pretty well most of the time. The fraud concerns I have are with electronic voting where no permanent paper record is simultaneously created and confirmable by the voter.

WaGOP: "We could probably get that down to 0.01% if you would just let us require photo ID's from the blacks and Mexicans"

Amen, Goldy.

As you're well aware, I've argued ad infinitum that hand recounts are more accurate than machine recounts. That's because of precisely what Mr. Likness pointed out -- humans are better at pattern recognition than machines. Those 12 ballots were marked, even if optical scanners couldn't identify those marks.

Yes, sometimes humans make mistakes. But when there are two (or more) humans looking at the same ballots, the probability is vanishingly small that they'll both make the same mistake; it's a system that always corrects those few errors.

Unfortunately, all the accuracy in the world doesn't do much good if the voters put a shithead like Don Benton into office. Might as well raffle the seat off and give the proceeds to charity, since any random dipshit would be far better than that blowhard idiot.

@1 In King County, ballots are tabulated in batches of several hundred each. Random batches are counted by hand both during the counting process, and afterwards, and then compared to the machine tally. Scanners are also tested before counting begins.

Any technology can be used to do no good. Ultimately, what matters is the people's involvement in the vote and beyond. Laws guaranteeing the fairness of the vote are always a good idea.

I don't care what you say Goldy, Rush and the gang at Fox told me that it is ACORN's fault that Chink, Nazi, Socialist, Islamic, black guy got elected and now we are all doomed to driving off cliffs.

...it's a test that no Washington election has ever failed.

Phrasing it that way naturally leads me to believe that you are aware of cases somewhere in America where recount tests have failed... Florida 2000, perhaps?

If you are in fact aware of recounts that have shown the initial count to be a failure, the question you ought to ask is not "did this happen in Washington?" but rather "could this happen in Washington?"

More pragmatically, when you look at the election recounts you've studied that have revealed inadequate vote-counting (if any!), what characteristics might the systems used in those cases share with Washington's vote-counting systems?

So, in Washington, the problem is just the private money and the media influence and advertising it buys, not the vote counts. Although there's still that tricky business of disenfranchisement for felony convictions -- regardless of the type of felony -- until sentence, parole, and probation are completed, and the prosecutor's right to revoke restored voting rights if reasonable efforts to pay court-ordered fines haven't been made. Nationwide, one in nine African-American men of voting age are disenfranchised at any given time because of criminal convictions, but the statistics are probably skewed by the permanent disenfranchisement of felons practiced by many confederate and red states. It's probably not as bad in Washington, but it's still troubling.