Thursday, November 29, 2007

Recounts & Other Touchscreen Messes-the literal unwinding them to solve

Yesterday I wrote about the (yet unsourced) encoder problems in North Olmsted, and the mess that caused. The board began the day with yet another meeting - or resumption of - from the day before. ( They figuratively "stay in session" during the recounts.

All 3 candidates for Council, previously prepared for a recount yesterday, were there.
It was found and fully documented with copies of original 11/6 documents, that 19 regular voters had voted on paper, to access the school board race, and had ballots put into provisional envelopes that were then rejected for lack of information.
One of those voted said she voted half Tsx, and asked for a paper ballot for school board. Hers was the only one of the 19 not accepted, for it could not be known if she had voted a whole ballot on the Tsx, which would have made her paper ballot a second one. The school board race had a solid outcome.
All the other 18 were accepted and were to be included in the certified count. (That was after sorting out the fact that the poll worker had noted the wrong "Gary" as having voted provisionally. The two "Gary J.'s" (last names were given) were called, affirming the info of the Gary J on the envelope, and that the other, noted Gary had indeed voted on a touchscreen.) What a mess.
I actually worked today (!) so did not go to the board meeting at noon, (nor call, nor to any recounts, thank goodness - they are truly tedious with the VVPATS and very long - see below) to find out the new result of that council race to be certified, and to find out if a recount would still be mandated. All candidates by the way, even the winner, affirmed what the board was doing was the right thing - expressed that they wanted every voter's vote counted.

2. After the Candidate and Voter Services Dept. found the problem in North Olmsed, they went back to see if there had been any similar numerous provisional rejects, for missing information, from one polling location that demanded further study, in case something similar had happened - regular votes on paper at the polls, being misconstrued as and rejected as provisionals.

They found two other instances.
The board accepted 11 mistakenly rejected paper ballots in Lyndhurst 4D that were given to voters when early on 11/6 the Tsx machines were "having technical difficulty." These 11 would not affect the outcome of the race, but could make it close enough to trigger an automatic recount. At the time of this writing, I assume that question was answered when the board certified those new totals today.

They also moved yesterday to accept 14 ballots, previously rejected in Cleveland 16E for the same reason. Those paper at poll ballots were due to a "faulty encoder."

In all of the above cases, the board checked all notations, the poll book etc. to make sure that the ballots they accepted were from registered voters who had signed in at their voting location.

The Literal Unwinding of Paper Trail Mess of the First Recount Day - complete with 15 damaged (from DRE printer) of 70 paper trails total
Is this what a recount should look like?
Board members determine damaged ballots to reprinted per Memo 07-21
This is a copy of a message I sent to a few election integrity groups yesterday morning. (Too tired right now to create a new one about the lonnng, tedious one at the recount the day before.)
In a message dated 11/27/07 4:17:21 PM, a member of Citizens Alliance for Secure Elections wrote:


Why not print out a copy of the VVPAT, two observers canvisually spot check ballots to make sure they match, or else inspect all ballots side by side, (would be easy to do on a large long table).

They actually kind of did that yesterday in the CCBOE recount, after reprinting the board-approved damaged tapes - as well as showing and explaining every step.
There were 2 recounts - one 5 precincts - but a race of picking 9 of 14 candidates!
The other - I think 8 precincts. Don't remember. Will get.
Since when I left at 7:30 pm - with one recount still going on - and other observer and all election workers still there - and I estimate still at least 3 hours to go... ( the problem is the VVPATs!) I was exhausted... and "looking forward to" another observer day today! The CCBOE was excellent in transparency.

BUT the absolute RIDICULOUSNESS of this whole messy, tedious, lonnnnggg, detailed day - and any future similar AUDITS, too - with the VVPATS - even if well done, and people can see every step, and they've demonstrated that cards and papers had originally been used together - as it was very well pointed out by Jason Parry, as below. All the process does is shows that the machines can add "SOMETHING" and that people can also add-up THE SOMETHING that the machine already added.

This closed-loop, circular summation however, completely leaves out the question of: do these duplicating messy, tedious, lonnnnggg, detailed "feats of addition" have anything to do with what VOTERS intended to cast?

The way they did it and the randomness of the appearance of the HIGH RATE of damaged tapes, along with poll worker signatures on the originals, would preclude any insider having secretely "recreated" cards and papers from them, along with new " numbered seals" on canisters - to "clean things up ahead of time for recount witnesses" - as was purported in past 2004 previously convicted Cuyahoga recount.


But with these machines - ALL of this - does still NOT affirm that they are recording votes as intended.
By the time it gets to recount/audit - the slippery slope attention is put to, not voter intention, but by then - the voter-disconnected question of only the addition ability of machine vs."man."

And from that, when they INEVITABLY match ( if the people are good detailed counters,) we're then told (not by the CCBOE on either of the below, but by most - even some of the candidate observers at the recount) it "proves the machines work." !!!!! ( I ask, for what???)

And from the messy, tedious, lonnnnggg, detailed day, we're also told, it also proves that machines are much "faster" than "man" at such addition - thus, to those people no reasonableness of hand marked, paper ballots!

Because of the invalid closed circularity - we get
completely invalid conclusions - dangerously affirmed by some direct experience with "SOMETHING" !

Further, it's been shown that even the experience of today in Cuyahoga, would hardly be what some politically-patronaged (thus, falsely trusted) BOE's in Ohio would actually do.

It's much, much easier - and actually just as ultimately (in-)effective in checking voters unified intent - to spill coffee.....and if some BOE's really want to extend themselves to keep up a show..."just re-print the recount results from the tabulator computer and call it a day."

At least the CCBOE is following the laws...and going out of their way to do what they can do to show people they really care about them and their concerns too - while struggling long and hard with the machine limitations and the ridiculousness that they impose.

( Recounting punchcards would have been MUCH faster. AND the human counting of the paper ballots from today was also - MUCH faster, as well as being originally much easier to reconcile, including allowing the people there to UNDERSTAND what they were doing and why...only partly because that process has inherent meaning.)

BTW, though I don't yet know how many tapes were counted in total today - I estimate between 50 and 75 total between the two recounts. I do know that 14 (actually it ended up to be 15) of those tapes needed to be reprinted. Percentage facts will obviously have to wait.
But it seems today that an estimate of 15-30% would be correct.

Throw the junk back..after getting our money back!


I have pictures, but right now, have to "pack my tent" for today... Promised a candidate I would observe for her. A whole different story...

Per Jason Parry:

"I mean really let's just cancel any recount where tapes are
printed from memory cards. You have to be a complete
moron to think this is a valid idea. Why not just re-print
the recount results from the tabulator computer and call
it a day.

A recount is intended to count paper results by hand and
compare that to the computer tabulation results from the
memory cards. If you print the paper tape, to be counted
by hand, from memory cards then barring human error,
OF COURSE they match. And anyone who has read any
of the detailed reports on VVPAT and election tally results
realizes that they are not the same and in many cases the
tabulation results do not match the VVPAT."

This is the PD report from the first recount day. The reporter, Joe Guillen also was an all day survivor. In it, the man of famous name Chris Riggall, Diebold-Premier spokesman said again, Diebold-Premier needed to do "further investigation." In another film yet to be published here, you'll see that Jane Platten, Director reporting on election night and official count tabulator crashes, on 11/21, that Diebold-Premier, even had that "chutzpah" to suggest that their "investigations" be done in their own offices! Thankfully, she declined the offer. A sure bet that Premier's conclusions will point blame at the CCBOE, the poll workers and anyone but themselves, for their woeful (to us) failures.

And I'm processing film from the day - only from the approximate 3 hours of reprinting procedures, not the counting itself. If you love watching paper being unrolled, you'll love the film. It will also serve as a good sleep aid.
But seriously, there are some interesting facts mixed in. And you'll get to see some of how careful everyone was to make sure the witnesses could see everything they wanted to (though limited by the rolls of tiny print.)
Stay tuned.

BTW, in the end, after a few human errors, and much human addition on the very human-unfriendly VVPATS, as predicted, the North Royalton count did not change.

I also understand that the CCBOE decided after exhaustion set in at about 9pm. to get legal permission to pause "the continuous count," of the Bedford race which continued to have errors, especially on counting ballots with 9 candidates to be chosen, hone their process, and go home to rest for the next day's marathon. In the end, the outcome of the Bedford race will not make any difference! The margin between #9 and #10 candidate was large. The automatic recount was triggered between something like #3 and #4 candidates, which will not affect who gets on the Commission.

No comments: