Re: The 11 days of Florida

From: Nils Paz <Nils_Paz_at_raytheon_dot_com>
Date: Thu Dec 02 2004 - 11:37:52 CST

Adding more days to vote may make the lines shorter but it would provide a
larger window for behind the scenes shenanigans. It means that the OVC
system must have cross checks at every point of ballot addition and at
every point of the system threads start and finish.
I have not heard of anyone discussing how the system should have checks to
see whether it has been altered or changed since initialization.
Sophisticated hacking is done using root kits that steal the basic system
applications that allow you run system things such as what threads are
running. They are supplanted by the hackers own version of those system
apps and you can never now that you have a trojan file monitoring,
changing or sending information out.
The only way that I know of is to ensure that as the machines are loaded
with an operating system, that system has byte check on all its basic
system applications and you compare those against a byte check of the OS
running at the time of an election and afterwards.

"Kathy Dopp" <>
Sent by:
12/01/2004 06:56 PM
Please respond to kathy; Please respond to Open Voting Consortium
discussion list

        To: <>, <>
        Subject: Re: [OVC-discuss] The 11 days of Florida

Keith Copenhagen said:
> Florida's election supervisors, impressed by the success of early
> voting, proposed dramatic reforms Tuesday that would eliminate Election
> Day, replace it with an 11-day election season and do away with
> precincts.


Interesing Keith,

FYI, A few of USCountVotes statisticians (off list but cc'ing me) have
been discussing the Miami Herald "recount" of 3 opscan FL counties and
have noted several things:

1. The Miami Herald did not report the method they used to recount the
ballots (did they only examine over-voted ballots by hand & use machines
to recount the rest or count them all by hand?)

2. The Miami Herald did not report which of the precincts they recounted
of the 60% precincts in Suwanee County, and that there isn't a set of
precincts in Suwanee whose votes add up to that reported by the Miami
Herald as the official vote total in Suwanee. It would be easy to slant
the count by selecting the wrong precincts, so the Suwanee county recount
is meaningless.

3. The Miami Herald chose op-scan counties to recount that would not
likely (according to what some of these statisticians thought) yield big
errors. (I myself would have selected Dixie county and a couple of others
instead of any of the 3 the Miami Herald selected.)

4. The Miami Herald reported that the results of its recount did not
indicate that the election results would be overturned, yet mathematical
extrapolation of the % change in votes that Kerry picked up in those 2
counties times the total number of votes cast for president in FL "Would
have changed the results of the FL election". Of course anyone realizes
that this gross calculation is not accurate or meaningful because you
can't extrapolate like that.

Another member of our group noted that the reported recount of New
Hampshire by the Boston newspaper also claimed that the NH recount didn't
show anything significant but failed to report any actual numbers at all
that would allow one to judge that analyses.

The full mathematical study at the precinct level, as soon as can get the systems in place to obtain all the election
results and recount data and normalize it into a database, will be very
interesting indeed.

The newspapers' recount reports are more enlightening for what they're
"not" reporting. It will take a project like to
truly shed light on this election.

I wonder how the early voting will affect the accuracy of future FL

Kathy Dopp
US Count Votes
OVC discuss mailing lists
Send requests to subscribe or unsubscribe to

OVC discuss mailing lists
Send requests to subscribe or unsubscribe to
= The content of this message, with the exception of any external
= quotations under fair use, are released to the Public Domain
Received on Fri Dec 31 23:17:01 2004

This archive was generated by hypermail 2.1.8 : Fri Dec 31 2004 - 23:17:22 CST