Re: determining vote count audit sample size to detect outcome-altering miscount

From: Kathy Dopp <kathy_dot_dopp_at_gmail_dot_com>
Date: Sun Jul 23 2006 - 15:55:39 CDT

This is SOOO Exiting!!

Frank Stenger tested his first function in maple and maple coughed, so
Frank wrote a program in mathematica using a different algorithm to
calculate the sample sizes that uses an algorithm similar to what I'd
originally envisioned for honing in on the correct value, but Frank's
algorithm and computer program is worlds better than what I had
envisioned - Incredibly elegant and efficient, but then Frank writes
books on numerical methods - such a genius. He still uses the gamma
ln function but within an iterative loop and mathematica.

Now we really need help now doing the months-long project of finding
math libraries we can install on our server to make this calculation
publicly available, writing this up in scientific paper with charts
and graphs - so that laymen and other scientists can understand it;
and contacting NIST, project ACCURATE, Brennan Center Task Force, and
others to get this formula adopted as the standard for vote count
audit sample size. The nice thing about math is that there is usually
only one right answer, despite more than one correct way of obtaining
it, so that once the one right method for vote count audit sample
sizes is determined, there is really no question but to adopt it, at
least for the scientifically-minded.

Argh, I desparately desparately need some help at my home office, if
you know of anyone competent who might be able to devote a few months
to living in Park City, UT and working on this issue, helping to
manage an office and nonprofit, or can help raise funds to hire
someone... please tell them we need help.

Thank you to those of OVC who sent me video clips or links to video
clips for my upcoming presentation to the Green party convention.
I've got a slew of great videos now though not enough time to play
them all. I love it!

Was anyone of OVC able to answer the gentleman who needed an
explanation of why testing voting machines would not detect or prevent
vote tampering or errors?

Thank you. Please let me know if you want to work on this audit paper
and have access to math libraries that include the gamma ln function,
or want to help regardless.

We have now derived two mathematical methods to ensure that election
outcomes are accurate:

1. A formula for exit poll discrepancy in terms of vote miscount and
partisan response bias that lets us evaluate exit poll data for
consistency with each - so no more time-wasting arguing for months
over exit poll discrepancies IF they publicly release the data for
standard statistical analysis, it will also give us the data for
evaluating consistency with either vote miscounts or partisan response

2. This equation for calculating the amount of vote counts to audit in
order to have any desired degree of probability for detecting
outcome-altering vote miscount. It will virtually ensure that future
elections have honest accurate outcomes, and is certain to become
adopted once people understand it.

This vote count audit work indicates that fixed 1 or 2% audits are
inadequate for detecting outcome-altering miscount in close races.
This may even eliminate the need for a 100% automatic hand count in
close races when a 50% or 75% audit will do - although perhaps there
would be a point near say 50% when it would be easier just to hand
count 100% as go through the process of randomly selecting so many

There are so many questions yet to answer.



On 7/22/06, Kathy Dopp <> wrote:
> Determining Vote Count Audit Sample Size to Detect Any
> Outcome-Altering Vote Miscount
> -----------------------------------------------------------------
> Everyone (NEDA and OVC)

> function to calculate
> the sample sizes using the following inputs:
> N = # total vote counts
> C = #minimum corrupt counts to alter any election ( C = (minimum
> candidate margin)/(maximum margin swing per machine))
> P = desired probability for detecting at least one vote miscount
OVC-discuss mailing list
= The content of this message, with the exception of any external
= quotations under fair use, are released to the Public Domain
Received on Mon Jul 31 23:17:07 2006

This archive was generated by hypermail 2.1.8 : Mon Jul 31 2006 - 23:17:09 CDT