Re: Fw: new voting system

[Paul E Condon <pecondon@xxxxxxxxxxxxxxxx>]

I'm not sure I understand this proposal correctly, but it appears to
substitute "Ballot Interpretation Reports", as defined by Paul
Walmsley, for true ballot images because images are deemed too hard to
handle images at the county level. But Humbolt County, which is not
one a center of web expertise, seems at easy with the idea of handling
them. I think Humbolt can handle the data set, and Boulder can also
handle it.

It then introduces a new obfuscation: When a voter goes to the web 
to verify his vote record he is presented with what appears to be
a marked ballot, but is actually a computer reconstruction of a marked
ballot that is based on the 'ballot interpretation report'. 

2) eSlate does not actually record a 'ballot interpretation report'.
It tallies votes 'on the fly' as the election progresses. For this 
to comply with the recent recommendation of NIST for 'software independence',
the entire roll of paper from the eSlate printer would have to be read
and the data entered into the 'ballot interpretation report' database.

3) There is no way for a concerned citizen to view the whole body of
voting data and test it for statistical validity, as would be the case
under the Humbolt proposal.

4) The heavy use of random numbers is a problem. Understanding a proof
that the system is adequately protecting voter privacy requires a
level of mathematical sophistication that is not commonly available in
the electorate.  What is the point of all this stuff, if people are
left having to trust 'authority'?

On Tue, Dec 19, 2006 at 07:50:22AM -0700, stevepom335 wrote:
> 
> ----- Original Message ----- 
> From: stevepom335 
> To: humboldt_elections@xxxxxxxxxxxxxxxxx ; info@xxxxxxxxxxxxxx ; bbvreport@xxxxxxxxxx 
> Cc: Liss, Josh 
> Sent: Tuesday, December 19, 2006 7:48 AM
> Subject: new voting system
> 
> 
> To the folks working on the Humboldt County ballot system:
> 
> I am a former city council member in Boulder CO, and have been thinking about voting systems using publicly accessible data bases for a while. Yesterday I discussed my ideas with Josh Liss, the Boulder County CO election coordinator, and he said that just yesterday morning he had received an email about the proposed Humboldt County system using scanned images of paper ballots. Apparently good ideas pop up in multiple places at once.
> 
> In any case, I have written up my approach in the attached paper. I think it could make the idea of public data base voting a bit easier to implement than the approach apparently under consideration in Humboldt County. Thus I thought you might find this brief paper of some value.
> 
> Please contact me with any thoughts on this matter, and feel free to forward the paper to others who might be interested.
> 
> Regards,
> 
> Steve Pomerance
> 335 17th Street
> Boulder CO 80302
> 303 447 8026
> 



-- 
Paul E Condon           
pecondon@xxxxxxxxxxxxxxxx