Washblog

Also -- If you wonder who edited your story -- that was me.  I shortened long URLs, as they distorted the site....  

Locally
As for us, we're still facing Diebold's new highspeed tabulator (DRS hardware, Diebold Assure Suite software), Diebold's automatic signature verification system (VoteRemote), and Diebold's computerized voting touchscreens (the notorious AccuVote TSx).

The Diebold AccuVote TSx computerized voting touchscreens are what King County has. My copy of the docs (from my open records requests) are at home, but from memory we own 200 and lease 400.

These machines are crap.

The current plan, along with all the other odious changes, is to have regional voting centers (RVCs) and dropoff locations. Last I heard, there will be 17 RVCs. There's a whole separate committee and decision making process that I haven't had the bandwidth to follow. The details are bound to change over time.

At the RVCs, King County Elections intends to eliminate paper ballots and rely solely on virtual ballots cast on the touchscreens. Including provisional ballots. (Which makes no sense to me.)

King County Elections is estimating that 42,000 voters will use the RVCs for early voting, who will vote over two or three weeks. I think they're inappropriately optimistic. We had 32,000 provisional voters during the 2004 presidential. I think we'll have at least 42,000 voters on election day, wandering around the city, trying to figure out where to vote.

(The great irony is we're replacing our supposed "hybrid" election system, because it's so gosh darned complicated, with a completely untested "hybrid" election system that is way more complicated.)

Additionally, they'll replace our paper poll books with electronic poll books. This is a horrible idea. Experiences reported from other jurisdictions are cause for great concern. I've asked for more information about that. But that particular open records request got a brush off.

Here are two corrections to my post.

From the original posting: "there is a 1 in 4 chance (26% actually) that an invisible ballot stored in the flash memory of the AccuVote Tsx DRE will be accurately copied to the removable memory cards."

We're indebted to Kathy Dopp for pointing out that this isn't what the ESI report actually says. Corrected, the text would read: "there is a 26% (1 in 4) chance that an entire Vote Center will have one or more discrepancies between the DRE memory card and the flash memory on the AccuVote TSx." [In the ESI report a "Vote Center" is synonymous with a "polling place."]

Also, it's notable that in 6 percent of the cases the discrepancies were by 25 votes or more.

For these data see Figure VI-16 of the ESI report:
"DRE Memory Card vs. DRE Election Archive - by Vote Center - (absolute value of difference between DRE Memory Card vs. DRE Election Archive)"
ESI Report: hardcopy, page 118; pdf file, page 123.

Also please note one further minor correction: the machine internal storage for votes is referred to in the ESI report as "Election Archive;" and ESI says it believes but has not confirmed that the medium is a hard disk drive rather than flash memory.

See "DRE Election Archive" in the Glossary of the ESI report:
ESI Report: hardcopy, page 229; pdf file, page 234.

The machine malfunction, then, was less frequent than as first reported.  But that it occurs at all is depressing.

Kathy Dopp has kindly agreed to my quoting from her message. Her main web site is:

http://electionarchive.org

Link for the ESI report:
http://bocc.cuyahogacounty.us/GSC/pdf/esi_cuyahoga_final.pdf

Link for John Washburn's post:
http://washburnsworld.blogspot.com/2007/03/diebold-accuvote-tsx-like-those-used.html

Somehow, people aren't using the recommend feature very often on Washblog. Maybe I'll have to try to make the link more visible....

Noemie
(one of the Contributing Editors.)