[internetworkers] Are You Registered To Vote At YourCurrentAddress? Register Online!
clubjuggler at gmail.com
Thu Aug 7 00:46:10 EDT 2008
On Wed, Aug 6, 2008 at 2:10 PM, S B <merlin180 at hotmail.com> wrote:
> Hi Tanner
> Actually, now that you mentioned it, there are several things about the
> initial physical process I've often wondered about - maybe you can educate
I can try. :-)
> Assuming Durham's process is the same:
> The reading machines look like giant shredders - an unfortunate
> co-incidence :)
> The ballot page itself is a series of small roundish boxes that the
> voter fills in to indicate their intent
> The ballot is placed in the reader, which whirs for a bit, increments a
> counter on the front, and drops the ballot into the hopper
> The voter collects their 'I voted' sticker and leaves
> The counter appears to indicate a ballot was received but beyond that, what
> checks are in place to ensure that the machine:
> - read a *valid* ballot (would it spit it back out if it detected
> something inconsistent like 2 votes where there should only be one ?)
The reader will spit the ballot back out if it detects an overvote. It will
then give the option of pulling it back out and revoting on a new ballot
or just letting it go through and not counting the race with an overvote.
It will specifically not, however, complain about undervotes.
(Definitions: overvote - voting for more people than you're supposed
to on a given race
undervote - not voting for enough people on a given race)
> - correctly interpreted the voter's intent ? - AFAIK the voter gets no
> feedback as to what the machine 'saw' on the page - maybe a bent ballot or
> something else caused a misread
You are correct. There is no feedback on this. However, if it cannot read
anything on the ballot it will spit it back out again (had that happen to me
the last runoff election)
> - increments the correct totals ? (e.g. something like open kimono code
Once again, there is no feedback for this, but it would be caught by
the random recount if there were a problem.
> - Ditto the downstream hardware/software
Again, no way of knowing, which is why they have the random recount.
> The random manual recounts you mention might catch some issues - at least
> for the machines that were recounted:
> What percentage of ballots/machines/stations/districts are randomly
> recounted, and how random is it. ?
I don't remember the exact numbers at this time but enough to statistically
tell if there are problems.
> Do such recounts ever *exactly* match the machine numbers ?
I don't know for certain, but from what I understand it's common to be off
by 1 or 2 votes either way. However, I believe NC law allows a challenger
to request a full recount if the actual vote count is fairly close anyway.
> What happens if random recount indicates a "significant" disparity with
> the machine totals - does the recount widen ?
>From what I understand, yes.
Note, btw, that if the counter machine spits a ballot out the voter is allowed
to press a button saying to just take the ballot as is. If for some reason the
voter has already left the precinct, election officials are instructed to press
that button for the voter so that at least some of their vote will count.
Also, the voter has to be the one that places the ballot in the machine.
This is to guard against an election official swapping out the voters ballot
with one they pre-prepared.
(fieldless) In fess two roundels in pale, a billet fesswise and an
increscent, all sable.
More information about the InterNetWorkers