optical scan ballots fall short

Today's Plain Dealer blog has a story critical of Cuyahoga County's new optical scan voting system.

http://www.cleveland.com/plaindealer/stories/index.ssf?/base/opinion/120...

This article confirms my experience as a Board of Elections member in Greene County. Although I refused to vote for the touch screen (Diebold DRE) machines that we did buy, their performance in the five recounts that we have had has been consistently perfect. That is, the count on the electronic memory cards has been exactly the same as on the paper roll that also records the vote. I recognize that there are many problems with these machines, but once a vote is recorded accuracy seems to be assured. This was not the case with the punch card system that we had before nor is it with the optical scan system that we have used for mailed absentee and early voting ballots. The punch cards, essentially old IBM cards, were counted by a computer that recorded light shining through the punched holes. The famous hanging chads would flip in and out of place. Thus, with the punch cards we considered a variance of two or three votes per precinct (400-800 votes) as routine "chad error" not meriting further hand recount. Now with the paper ballots that we optically scan there is a similarly high rate of error, either over votes where people vote for more than they are allowed to, and thus are voided, or where they inadequately mark the ballot. The count by the optical scan machine often does not pick up what a human eye would recognize as an intended vote. And then, there are the people who just needed glasses when they marked and we can't figure out what they intended. Thus, there has been a variance in the hand recount compared to the initial electronic optical scan count. So, should all ballots be hand counted? Well, have three different people count the same three inch high stack of index cards. Do they all come up with the same result? Try larger cards with 30 different either-or choices and work with a stack of 50,000.
That optical scan story again:

http://www.cleveland.com/plaindealer/stories/index.ssf?/base/opinion/120...