RE: Improving the LAT (Re: Bad ballot markings)

["Some Guy" <someguy@xxxxxxxxxxxxx>]

Sarcasm noted.

Of course Neal's test in the LP LAT stack was a partial test. It made people
in the clerk's office go apeshit.
I thought that it was interesting that the tiny spec in the middle of the
box registered as a vote, while one of the same size near the edge of the
box did not.

More discussions with Ralph at lunch revealed that despite Neal McClure's
assertion that the scanners don't need calibration and can discern marks as
small as a pixel - it's a marketing farce.

My notes from the trial (first day) have McClure saying that the resolution
was one pixel. I was alarmed by that statement, because there was no
definition of what size of pixel he was talking about.
The word 'pixel' in and of itself is not a definition of size. It is simply
a term used to describe a measurement. IE: I can say something is defined in
'inches', but not say how many inches, or fractions thereof.

Since we don't know what the software is doing, we can't know what is
acceptable and what gets tossed as noise or dirt.

The LATs as they were done were ineffective. Poor assumptions were made
about what voters might do, and therefore how the scanners and software
might respond.
Throughout the discourse about the LATs at the time they were performed
someone said (Joe I think on the Capitol steps) that the tests only show
success and not failure. In fact they do show failure, but the officials in
the clerk's office have defined failures as something humans do, not
machines. I can't recall the exact quote (not taking notes at the time), but
it was excellent.
Was this you Joe? If so, can you repeat what you said? It was good and we
should all be able to repeat it.

sg

-----Original Message-----
From: Joe Pezzillo [mailto:jpezzillo@xxxxxxxxx]
Sent: Friday, November 05, 2004 5:56 PM
To: Jan Kok
Cc: pshields@xxxxxxxxxxxxxx; 'KGNU news'; 'LPBC Board';
cvv-discuss@xxxxxxxxxxxxxxxxx; 'Ryan Morgan'
Subject: Re: Improving the LAT (Re: Bad ballot markings)


One of the arguments in favor of the "Vote Marking" machines that
Boulder County citizens asked the county to consider in the stead of
DREs in December of last year is that they could produce uniform,
unambiguous marks on paper ballots for voters of any ability.

Too bad no one in Boulder County thought in advance to test ambiguous
marks or other variable conditions on the ballots, eh? (sarcasm)

Joe
[|>]