I think I figured out why King County was holding out for so long to release The Binder. They were creating a bunch of companion documents to "explain" the raw source documents to the public. The source documents in The Binder confess a ton of errors, but the "explanatory" companion documents try to paper them over. (Master Explanation, Reconciliation Worksheet and Notes)
I'll illustrate with one example polling place, which is dealt with similarly to several others that I've looked at: Denny Terrace, 100 Melrose Ave E, a facility of the Seattle Housing Authority.
Denny Terrace had one of the highest differences between polling ballots counted and polling voters credited of any polling place in the county. (869 ballots to 830 voters). As it turns out, this discrepancy is mostly attributable to unverified provisional ballots that were illegally inserted into to the ballot machines. These provisionals, however, were NOT included in the 348 provisionals that were acknowledged back in January. The county recently released a hardcopy list of the polling places where the 348 provisionals were cast. I'll have more on that list later. But Denny Terrace is not on that list. These are newly recognized unverified provisionals. Nevertheless, the companion documents try to sweep the problem under the rug.
Here is the actual Binder entry for Denny Terrace:
Notice all the notes that say "no label". That refers to the number of unverified provisionals that went into the machines. I looked at Denny Terrace poll books today. Click here for a picture of the Ballot Accountability Sheet for for Precinct 2541: 89 "regular poll ballots issued" and 102 ballots went into the Accuvote machine. The poll book signature count appeared to read "88". Other documents showed 102 ballots counted and 87 voters credited. The main difference is the "no label" provisionals.
I found the following totals for the entire polling place:
Signatures in Poll book: 836
Poll Voters credited: 830
Ballots counted in Accuvote: 867
Poll Ballots counted in canvass: 869
Here is how the companion documents spun this.
The "Notes" document summarizes in more readable form the notes from the source document that there were 42 "no label" ballots in the entire polling place. That strikes me as a bit of an overstatement, but fair enough.
The "Reconciliation Worksheet" calculates a "reconciliation variance (Plus/Minus)" (defined in "Master Explanation document") to be 30, which is roughly the sum of the provisionals that went into the ballot machine plus other errors. This number is adjusted downward by an "Adjusted #" of 25 to give a "Net Difference" of 5. (They don't explain how they came up with the Adjusted # of 25 and it's not at all obvious from the documents I've seen, including the poll books)
The "Reconciliation Summary" sums all of these "Net Difference" numbers to get a total of 216 positive difference and 158 negative differences, so they're claiming a grand total discrepancy of 374. (In 2000 the total discrepancy was 20). The total of the "Adjusted #"s is 660. That appears to be an approximation of the only-now-acknowledged unverified provisional ballots that went through the machines.
They're not willing to state that provisionals went into the machines except there is a specific report from an election worker to that effect (although it's hard to imagine what else could explain such a difference between signatures and ballots). Even when they do acknowledge that provisionals went into the machine illegally, they don't include that number in the "variance". So instead they set this "no label" number aside as an explainable discrepancy that's not considered a real discrepancy, or something. Its a half-assed concession that there were serious problems. At least they're not in full-bore cover-up mode anymore. But a more forthcoming acknowledgement and explanation of the problems would be better.
Stay tuned. I have plenty more to report.
Posted by Stefan Sharkansky at March 12, 2005 02:10 AM | Email ThisIf the canvasser’s examination of a precinct revealed a variance, steps were taken to investigate the issue. These included but were not limited to:
• checking the bins in the ballot counter;
This action supposedly
Starting the day after Election Day,
yet a Dec. 15 article by King 5 reports that during the manual recount King County "found" 22 votes in a bin in the ballot counter. How did these "slip" by their thourough checks for discrepancies in the poll books for not only the first but also a second count?
Posted by: ChuckJ on March 12, 2005 03:44 AMOf the 348, the percentage of "valid" ballots was 72% (252 of 348).
To my knowledge, it is possible to figure out which voters had improperly cast provisional ballots directly into the AccuVote machines.
This must be done countywide.
Step One: Create a list of all provisional voters from the poll books.
Step Two: Find every single provisional envelope (those that were not empty--there were some of those, indicating that the voter had put the ballot into the machine). Create a list of the names from these envelopes.
Step Three: Subtract the list of envelope names from the list of provisional poll book names.
Next, check the registrations, check the signatures, check to make sure these voters did not vote twice (e.g. absentee ballots, multiple provisionals, etc.) and then report how many of these ballots were "valid". (This process has already taken place for the 348 ballots.)
For the sake of convenience, let's say that the total number of provisional ballots that were found to have been fed through the AccuVote machines comes to 1000. Chances are, the percentage of "valid" ballots among the 1000 would be somewhere around 72%. That would mean that about 280 "invalid" ballots were included in the certified totals.
However, it is possible that the percentage of "valid" ballots is less than 72% since the 348 ballots does not represent a random sample.
Posted by: Tim B. on March 12, 2005 05:52 AMWhen first they practice to deceive.
The next issue is:
Did they accurately report identify these discrepancies, investigate them and report them to the Canvassing Board?
What precisely did Election Official Logan report to Canvassing Board member Logan (and his faithful sidekick bloodhound Larry Phillips).
The "Adjustment Number" of 25 should NOT be a mystery. In fact, it probably should have been reported to the Canvassing Board. It probably has something to do with Provisional Ballot envelopes or signatures on pollbooks...something like that.
Stefan--another important question are DATES when changes to these Polling Place Canvass Reports and companion documents were made and by whom. I see this document has initials...but no DATES.
I believe we will find that Logan and Huennekens were guilty of PREMATURE CERTIFICATION!!!!! All or almost all of this work should have been done before certification of the ORIGINAL 1st COUNT which was November --??
I'll bet they did not report the magnitude of DISCREPANCIES at that 1st count BECAUSE they were negligent in their reconciliation process.
Logan and Huennekens are obviously not the election "guru's" they have portrayed themselves to be. There should have been a very CLEAR accounting of discrepancies. A clear investigation. And a clear report to the Canvassing Board. The CB could have withheld certification pending further investigation. They certified DESPITE this mess. Excellent Work!
Posted by: Mr. Cynical on March 12, 2005 05:55 AMGood work my friend. I want to echo Mr. Cynical's comment that the modifications to the records do not seem to have dates, some do not have initials - which means there is no trace of custody. And some reports appear to be by the same person (Denny Terace and Bothell Region). I am not a handwriting expert by a long shot, but someone should check and see how many individuals actually did the red mark ups.
Personally, without dates and signatures for the modifications, the modifications should be disqualified. If these mods were used to get the second and final tally, they should be tossed out.
Edits to records with out a clear audit trail and sign off by a person with authority should never be considered valid over the originals.
Posted by: ajstrata on March 12, 2005 06:50 AMI think you are not understanding correctly the term "NO LABEL".
Read the notes at the bottom of each page of the "Polling Place Canvass Report" (pdf).
Here is what they say:
" ID = Inspector, No Label = Provisional ballot label missing, Accuvote = vote counting machine, AV = Accuvote" (Emphasis added.)
Notice that there is in fact some "label" that is "missing".
Then look at virtually any page in that same document and notice how the notes refer to, for example, "2 pb ran through Accuvote" immediately above a line which states, "7 pb no labels."
"No label" may have something to do with putting a provisional ballot through an AccuVote machine, but it clearly refers to something more.
It clearly refers to some "label" which is missing from the ballot.
What "label" should be on those ballots?
How could the "label" be missing?
Posted by: Micajah on March 12, 2005 09:08 AMWhat the heck happened? Is there any reasonable cause for this discrepancy, or is it time for Goldy to start eating major crow?
Posted by: Jeff B. on March 12, 2005 09:26 AMWithout an explanation for why so many of the precinct totals do not match, there can be no confidence in the result of an election that was so close.
And I think they should rename Bothell Regional Library to Bothell Moonbat Regional Lie-brary.
Posted by: Jeff B. on March 12, 2005 09:32 AMI didn't vote provisional so I'm not sure if a "label" is actually added, but the No Label could be refering to lack of a signed security envelope for the provisional ballot. Just a thought. Posted by: chuckj on March 12, 2005 09:35 AM
Thanks.
Posted by: C. Oh on March 12, 2005 09:48 AMIn other words, I don't know what No Label means, either.
Posted by: Huckleberry on March 12, 2005 09:58 AMCould Have
Might Have
Thinks
May Have
Is possible
Isn't their some type of "validation" for these people???
Posted by: chris on March 12, 2005 10:03 AMGood questions. I'm busy for the rest of the day. I'll do my best to address them later.
Posted by: Stefan Sharkansky on March 12, 2005 10:11 AMThe Democrats were trying to claim that the provisionals would have, most likely, been accepted at a rate of 92%. Two thousands questionable ballots would have, under that assumption, meant about 160 illegal ballot were cast. Apportioning those ballots according to the overall percentages in King County would have only yielded Rossi about 30 votes.
But, if there were two thousand ballots fed directly into the accuvote machines, and 28% were invalid, that would be 560 illegal ballots. If those ballots broke in favor of Gregroire by 23% that would be the final margin in the election.
23% is only 6% more that Gregroire's margin county-wide. If these ballots were in fact entered in precincts that heavily favored Gregroire, Rossi can plausibly claim that he would have received the most votes on election day had the ballots submitted on election day been handled properly by election workers: malfeasance.
And that would before considering felons voting,or individuals casting ballots on behalf of dead people, etc.
I believe it would be a matter of proving *intent* to cover-up or intentional failure to act, and this is more difficult than it seems. Proving facts is one thing. Proving them to a court is quite another. Establishing all of the elements necessary to prove fraud is pretty tough, but the way these elections people are going, who knows? They seem to know this and are probably thinking for good reason that they can B.S. their way out of this like they do everything else.
If one can make the case confusing enough, linkeages of intent easily dissolve in the process. I believe that is what the King County Election people are doing now. Like the Hillary FBI files imbroglio, they will dog us, mis-direct, confuse, mix evidence, claim ignorance, accuse the accuser, and act as if they are reasonable all through it. In effect, they will bury the case in endless paper and technicalities. If they can look reasonable while lying through their teeth, they will get out from under this, but only if we don't stay cool and dogged like GW Bush.
Posted by: Inquisitor on March 12, 2005 10:52 AMThe circumstances surrounding the existence, discovery, and production of "The Big Book" , along with the first few items that Mr. Shark has published has a certain "odor" that should make all voters scream for a 'clearing of the air' and for justice.
The entire voting process, including balloting, reconciliation and certification and several other steps - along with vocabulary words I don't understand, such as "No Label", "Reconciliation Summary", "Adjusted #", put the analysis and discovery of error and/or fraud out of my reach. And all or this superimposed on a "count", a "recount" and a "manual recount". I must rely on Mr. Shark and others who appear to understand the nature and depth of the problem and to pursue it through the proper channels to a point where an eventual ReVote will take place.
When I was learning how to do a material inspection as a Commanding Officer in the Navy, I was told, "If you stand in one place long enough, you can always find a discrepancy." It takes time and it takes knowledge to do it right and it appears that Mr. Shark and friends are doing the right thing at the cost of much personal time, and I am grateful to them for their labors.
I don't know what the equivalent election terminology is for "A complete audit of the books"; but one needs to be done. This unknowledgable "Eastern Washington Retired Guy" is with you in any way I can be of help.
Posted by: Dave on March 12, 2005 11:26 AMThanks for the great work and keep it up.
Posted by: Inquisitor on March 12, 2005 12:04 PMYou may have hit on the source of the odd term, "No Label."
You stated above:
"A provisional ballot is issued by the election inspector when a person not listed in the voter rolls at that polling place demands a ballot. The voter fills out an explanatory and identification form, which is simply a label affixed to a manilla envelope. A ballot is selected from one of the precincts...."
Perhaps "No label" simply means they were unable to find an envelope with that label anywhere at the polling place. In other words, they have some indication that a provisional ballot was issued, but they don't have the envelope it should have been put into.
After reading their explanation of the Bothell Regional Library polling place, it appears that they simply decided that 31 of the provisional ballots issued at that polling place were inserted into the AccuVote machines -- and thus "adjusted" the difference between the number of signatures in the poll books and the number of ballots inserted into the AccuVote machines by 31.
They then reported a "variance" of only one ballot more than the number of identifiable voters who participated in the election.
I wonder if there is any rational basis for this "adjustment." Was there some way to identify 31 of the people who were issued provisional ballots but whose names didn't appear on provisional ballot envelopes with the "label" stuck to the envelopes?
It appears from Logan's explanation that "PBAV" was used in their notes to indicate when precinct workers said that provisional ballots were improperly inserted in the AccuVote machine. But, if that was simply the supposed explanation, rather than something said by the precinct workers, then "No Label" was used to note the conclusion reached by the canvasser.
In other words, when there was a shortage of signatures in the poll books, that shortage would be "explained" in one of two notes (both of which meant the same thing) -- PBAV and No Label. PBAV means provisional ballots were inserted improperly, and at least one precinct worker said "yeah, I think that happened." No Label means provisional ballots were issued but not returned in their envelopes (with the labels stuck on the envelopes), and they appear to have been inserted improperly into the AccuVote machines -- but no precinct worker said "yeah, I think that happened."
In both cases, the apparent source of the "voterless ballots" was provisional ballots.
And, that apparent source was tendered as an explanation for the variance between the number of signatures and the number of ballots.
In other words, if anything other than magic or UFOs could explain the presence of "voterless ballots," then there is no "variance."
I hate to see election inspectors grilled by the county board or judges, but I think it might be necessary for the inspectors at some of the more aggregious polling places to be brought forward to give public testimony.
I have no problem believing that one or two provisional ballots (or pilfered ballots) could make there way into the Accuvote at a single polling place, but dozens is unnacceptable. The inspector at such a polling place was not doing a credible job.
Posted by: Huckleberry on March 12, 2005 04:07 PMRest ASSURED the poll workers and inspectors at aggregious polling places will be called in to EXPLAIN discrepancies.
I believe there are more than a few people who have seen their numbers MORPHED into something that does not represent what happened Election Day. In fact, each and every one of them should have been consulted about major discrepancies LONG AGO! Posted by: Mr. Cynical on March 12, 2005 04:52 PM
I wonder if you have obtained the cover sheets plus the ballot accountability forms for each precinct that used the Denny Terrace polling place.
The whopping difference between the apparent number of voterless ballots based on the post-certification updating of voter registration records (to include the date of the last election in which each voter participated) and the supposedly pre-certification effort reflected in the "big binder" may be little more than sleight of hand.
Notice that the summary of the notes for Denny Terrace shows a total of 42 "no label" provisional ballots run through the AccuVote machines.
But, then without apparent explanation, the numbers drop. Here are the precinct numbers and "no label" numbers:
SEA 43-2541: 13 no label (still 13 in summary)
SEA 43-2545: 3 no label (reduced to 1 in summary)
SEA 43-2546: zero (still zero in summary)
SEA 43-2547: 12 no label (still 12 in summary)
SEA 43-2548: zero (still zero in summary)
SEA 43-2549: 7 no label (reduced to 4 in summary)
SEA 43-2852: 7 no label (reduced to zero in summary)
What started out as a total of 42 provisional ballots run through the machines was reduced to 30 without explanation.
Then, the "adjusted #" of 25 was further subtracted from the total, reducing the so-called "variance" between the number of signatures of voters and the number of ballots cast and counted to 5.
From 42 to 30 to 5 -- with no explanation as to how those numbers were derived.
If you have the poll book cover sheets and ballot accountability forms, perhaps you can figure it out.
(One of these days, please post an image of the counterpart to the cover sheet for Bothell 01-0254 and the ballot accountability form for Bothell 3271. It's extremely frustrating not to have both documents from one precinct available as a possible way to figure out what the canvasser was doing. The two documents ought to fit together like hand and glove, but you've posted a right-handed glove from one set and a left-handed glove from another set.)
Posted by: Micajah on March 12, 2005 05:10 PMYour worksheet showed a big discrepancy at "Rainier Free Meth Church" (precinct SEA 37-1629): 114 votes in the first recount and 115 votes in the manual recount cast at the polling place, but only 75 voters credited with having voted.
The "big binder" indicates no such discrepancy. It shows one more voter than votes (or one fewer vote than voters), and indicates that only 102 votes were counted as having been cast on election day at the polls.
How could there have been 114 or 115 rather than 102 votes cast and counted?
King County's first recount report showed Gregoire receiving 87 poll votes plus 11 add-ons, Rossi receiving 13 poll votes plus 3 add-ons, and Bennett getting no votes at the polls.
That totals to 114, just as you indicated in your worksheet.
What's going on with that precinct?
Posted by: Micajah on March 12, 2005 06:18 PMThe King County Admin Bldg polling place also makes no sense to me.
According to the written explanation of "add-on" ballots, they should show up in the "big binder":
Election Day Machine Count – the number of ballots tabulated at the polling location on Election Day.
Add-ons – valid ballots not counted on Election Day because they did not go through the Accuvote machine. This fail-safe mechanism is designed to ensure ballots are counted and secured when there are technical glitches or when a voter does not want to use the Accuvote machine. These ballots are kept secure in the second side bin of the Accuvote machine base unit, recorded in the reconciliation binder and later added to the official count.
They don't seem to show up in the "reconciliation binder" according to the summary King County did. Only 3 add-ons are noted in the big binder for the 5 precincts that used that polling place. [SEA 37-1823, -2685 (1 add-on), -3416 (2 add-ons), 43-1778 and 43-2862]
If I added them correctly in my head, 188 show up in the first recount. (I haven't downloaded the file of the initial count to see if it might show the same as the first recount.)
Your worksheet had shown 523 ballots cast (which agrees with the first recount I downloaded) and only 458 voters credited with voting -- for a difference of 65. But the big binder shows only 4 fewer voters than ballots.
Where did those add-ons come from? When were they added onto the initial election day count? Why wouldn't they be reflected in the big binder's reconciliation of those precincts? Puzzling -- how could the difference in voters and ballots be 4 in the big binder and 65 in the list provided to you?
If you were going to 'cook the books', why leave a difference between votes and voters of 4? Why not make it zero? Make the waters muddier? We have to make it look like a "human" process? Posted by: Dave on March 12, 2005 08:10 PM
BTW - Most of these errors came during the hand
recount- So there should be no doubt that machine counts are more accurate than hand recounts - especially in King County when the Republican is ahead. Actually all three vote counts had an element of a manual recount - which were the new ballots conveniently found during each count when it was discovered that more would be needed to put CG ahead !
"If they [King Co Election workers et al]can look reasonable while lying through their teeth, they will get out from under this, but only if we don't stay cool and dogged like GW Bush AND STEPHAN SHARKANSKY."
I second the notion above - because this is what day in court will come down to.
Posted by: KS on March 12, 2005 10:12 PM
Right now I would settle for understanding what King County claims to have done in their reconciliation effort and how their claims match their paperwork.
I'll leave the guessing about evil deeds and ulterior motives until later. First, we need to understand what appears to have happened based on the documentary evidence.
Posted by: Micajah on March 12, 2005 10:15 PMI looked at Aki Kurose Middle School polling place on the first page of the summary, just to see if I could tell whether "PBAV" and "no label" were being used the way I thought.
Strange: According to the canvassers' notes, there were 4 PBAV in 37-1618 and 2 PBAV in 37-1619; and there were "no labels" in the other precincts -- 7 in 37-1620, 3 in 37-1635 and 1 in 37-1636. But, when I looked at the summary I found that only the 4 and 2 PBAVs are noted.
There doesn't seem to be any indication in the summary that those other 3 precincts at that polling place had "no label" provisional ballots.
The summary simply shows 0 (where there were 4 PBAV), +1 (where there were 2 PBAV), -1 (where there were 7 "no labels"), +1 (where there were 3 "no labels") and -1 (where there were 4 "no labels) for a "zero" "Plus/Minus" balance.
If there were 4 "PBAV" in 37-1618, where did they go? Why aren't they reflected in the "plus/minus" column?
Ditto for the 2 PBAV in 37-1619: How could 2 PBAV be only a +1 in the plus/minus column?
If there were 14 provisional ballots issued that turned up as "no labels," where did they go?
I don't see any way to compare the notes to the summary, if Aki Kurose Middle School polling place is typical.
It makes no sense to note that 14 "no label" provisional ballots were issued, yet the polling place is found to have exactly the right number of ballots (which would mean that "crossovers" in ballot issuance explain the difference in each precinct).
It seems that a comparison requires at least the poll book and ballot accountability sheet to see if the summary and notes make any sense at all.
"Makes no sense" should be the King County elections office motto.
Posted by: Micajah on March 14, 2005 07:23 PMI've been able to figure out how to make sense of the reconciliation summary and notes for the Bothell Regional Library polling site and precincts.
It does work out the way the note on the BOT 01-0254 cover sheet says -- 31 provisional ballots seemed to disappear ("no label"), but presumably were inserted into the AccuVote machine.
It's a long explanation, but it you want a break from whatever you're doing (or for some reason want your eyes to glaze over), it's at this entry at Croker Sack.
Posted by: Micajah on March 15, 2005 04:27 PMDear Sir,
We have aded your link on our
http://www.top-entertainment-sites.us/admin_edit.php?c=91-1&s=11
Now it would be great if you add our link on your site in Mind Machine catagory with the following information.
Title:Mind machines UK - light and sound machines at Sensed UK
Discription:Hire and buy mind machines and light and sound machines
URL:http://www.sensed.co.uk/brainwave-entrainment-mind-machines/mind-machines-light-and-sound-technology.html
Thanks.
Posted by: Naima on April 11, 2005 03:31 AM