Nevada paper trails work without a hitch

From: Charlie Strauss <cems_at_earthlink_dot_net>
Date: Mon Sep 13 2004 - 22:44:46 CDT

Well nevada just held the first large scale actual primary using
Sequoia AVC edge touch screens equipt with a paper trail. It worked
flawlessly by all reports. Indeed the only two glithces where both in
the electronic part of the voting. Clark county had some problem
getting the vote collection software to display it on the real-time
"precints reporting" unofficial vote count (so the TV just said 0 votes
all night. And Nye county reported a defective memory card.

I called up the clark county registrar of voters (their name for
"clerk") and had a nice chat with him. He said everything with the
paper went well. A few precinct judges misfed the paper rolls at first
but that was quickly sorted out in the few cases it happened. No
delays were encountered. I repeat no delays. Did I mention no long
lines or delays. Good cause Denise Lamb, head of NASED, insisted there
would be long lines when paper trails came to be. None Nada Zipola.
(Please phone denise up and let her know. :-)

The cost for the ink used was considerably less than the $1 million
dollars Denise Lamb likes to quote. the ink costs were, well zero,
because it was thermal printer just like those reliable ones you see at
the airport. The paper roll's well the clerk said they cost a dollar
or two apiece, less than they spent on ball point pens or flashlights.
   And what about the printer's themselves? well they increased the
cost of the machines about 10%. (~500, sequoia says this was below
costs on the printers themselves, though they made money on the total
package)

There was a one time cost they did have to bear however. that was
about $200,000 for a new vault they purchased to store the printed
paper rolls in securely. ( the way these things work, the paper
collection unit is a locked ballot box that the precint workers cannot
get into. This ballot box is taken to the vault)

I note that since they now have a paper trail they can recover from
errors and if they save a single election well that would pay of the
difference many fold

What the clerk liked best was he said usually five of six hundred
people always complain about the election. This time almost nobody
did. Every one had confidence. He felt that was the main reason for
the paper and he was really happy about it.

Now what about re-counts. Well Denise Lamb has said this was
impossible on "tape" style printers like the sequoia.

And yet as I was speaking to the clerk, in the next room people were
re-counting 2% of the paper ballots and confirming the electronic vote.
   In fact that's the Law in Nevada (2% or 20 boxes).

one interesting point: what happens if the records disagree? well
right now in nevada the laws are still the old laws (paper trails were
a gift from Dean Heller the SOS, not the legislature). So that means
the electronic ballot is THE ballot. But this could change now that
there have been no problems. this was the first run with these printer
systems and they have been using the pure electronic system for a few
years. So it may have been a reasonable thing to do the first time.
after all THEY WILL NOW KNOW IF THERE IS A PROBLEM or not.

The ballot was a slate of only 20 contests and voter turnout was light
on election day. More people voted in early voting (71,000) then on
election day (61,000). The early voting was 100% touch screen/paper
trail. the election day voting was a mixture of touchsreens/paper
trail, and older style AVC-advatages (aka shoup-style) puch button DRE.

according to the clerk, in nevada you sign in at the poll. Now it was
not quite clear if you sign a log book or your name gets cross off a
list. I suspect from the way he phrased it your name goes in a log
book which implies serial order retention. not sure however. But
there are multiple machines at each polling place and which machine a
voter uses is not recorded. And as I said precint worker cannot access
the printed tapes. So there is a plausible and deliberate procedural
safeguard against reverse engineering the vote order from the tape.
Shuffling ala OVC is obviously better for permenantly destroying the
vote order (unless we count GIDs). However one can also see that for
debugging errors there might be some utiltiy in being able to recover
the vote-order later.

==================================================================
= The content of this message, with the exception of any external
= quotations under fair use, are released to the Public Domain
==================================================================
Received on Thu Sep 30 23:17:07 2004

This archive was generated by hypermail 2.1.8 : Thu Sep 30 2004 - 23:17:11 CDT