I finally figured out how King County came up with their announced "final" discrepancy of 1,217 more ballots than voters: they used a junky data file and they're off by over 600.
The file they distributed on Friday contained 1,003 duplicate records, 636 of which were people who were recorded as voting on Nov. 2.
The real discrepancy is therefore closer to 1,800 (not including the 348 illegally cast provisional ballots). Here are the actual numbers:
|
Voters
|
Ballots |
||
| Records in File | 897,107 | Manual Recount Canvass Total |
899,199 |
| Duplicates | (636) | Illegal Provisional Ballots | (348) |
| Cancelled Voters | 253 | ||
| Address Confidentiality Program | 69 | ||
| Federal Absentee | 251 | ||
| Total Known Voters | 897,044 | Total Ballots Believed Legitimate | 898,851 |
All of the above gives us a net discrepancy of unexplained ballots = 1,807.
If you take out the number duplicate records, you get to 1,171, which is close enough to 1,217, that I can believe that there would have been another as yet untold story to explain the remaining discrepancy of 46.
But this all looks pretty funky. I hope they will at least announce that they made this mistake and revise their number to 1,800(+/-). Unless of course there are other bizarre errors that I haven't figured out yet. Which there may well be. I've been seeing indications that the bad news for King County Elections is only going to get worse in the next few days.
Posted by Stefan Sharkansky at January 10, 2005
01:37 AM | Email This
If I weren't in Japan teaching, I would be in Olympia at the rally. I will miss the chance to thank you in person for your hard work and effort.
When a Japanese is competing in any kind of race or contest or personal challenge, friends and spectators encourage that person to strive and fight for success, even in the face of great obstacles to that success. The word in Japanese is "gambate" and can be loosely translated as "fight".
Gambate, Stefan. We're cheering you and all the rest from across the Pacific. I'll be wearing orange tomorrow as I teach my students.
Posted by: JRR on January 10, 2005 04:52 AMI found you through an article at WSJ Online.
Here is a link to it:
http://www.opinionjournal.com/diary/?id=110006139
Keep up the good work.
Posted by: More Than Right on January 10, 2005 06:20 AMHow can you tell the difference between a duplicate entry in their database and a voter credited with having voted more than once?
Posted by: Micajah on January 10, 2005 08:57 AMHow many employees does Dean Logan have? And yet it takes the singular exploits of a private citizen to clue them into the fact their final tally, which by the way they took eight weeks to get us and which is already admittedly off by 1200... is actually off that 600 because these conscientous public servants who administered such an impeccably august excercise in democracy didn't do something as Cro-Magnon as check for frickin duplicates!
Anyone who has the least bit of confidence in these people should really be locked up for their own protection. Of course the people who seem to have the most confidence may well be getting locked up anyway.
KC's official website still reads "the remaining difference in the number of votes cast versus the number of registrations credited with voting in the 2004 General Election is 1,217 – an accuracy rate of 99.99% based on close to 900,000 ballots cast."
It's a good thing these folks aren't in charge of computing anything important like property taxes or elections.
1217 out of 900,000 is not 99.99% you witless wonders! 121.7--10 FRICKIN TIMES LESS--would be closer to 99.99%
And we trust these a__h___s to count votes?
Could this mean that 636 people voted twice? Such as getting and voting two absentee ballots in the mail, mailing in an absentee ballot and then voting at the polls by regular ballot, mailing in an absentee ballot and then voting at the polls by regular provisional ballot, mailing in an absentee ballot and then voting at the polls by stuffed provisional ballot, or casting provisional ballots at more than one polling place (or perhaps more than once at the same polling place)?
If there were 1,003 duplicate records, and 636 of them were recorded as voting twice, what were the other 367 duplicate records about?
Posted by: Richard Pope on January 10, 2005 09:42 AMI'm not sure why you subtracted the 636 duplicates from the count of known voters. It seems to me that each set of duplicates either represents two distinct voters who share a name, or they represent a single person who voted twice. In either case, you would have two ballots. Since you are trying to compare counted ballots with voter database entries, you should leave the duplicate records in the voter count.
Just to be clear, I'm not suggesting that the second votes of people who vote twice are legitimate. I'm just trying to get clear on the accounting between the voter database and the number of counted ballots.
Posted by: scottd on January 10, 2005 10:07 AMConsidering the large the Military presence in our state......this is bizarre and plain negligent imo.
I don't know about others, but I hate being last at anything.
M
http://www.soundpolitics.com/archives/003435.html#003435
Posted by: Scott in Carnation on January 10, 2005 11:47 AM"Huennekens' boss, county Elections Director Dean Logan, said Friday many of the unexplained ballots probably were cast by registered voters who failed to sign in when they went to polling places."
When I went to my polling place to vote I had to sign a book before I received my ballot. I think I would like a more detailed explanation of how you can get a ballot without signing first.
Posted by: jaybo on January 11, 2005 08:18 AM