Problem (was: Re: need help with tool requirement)

[bmcculley at rcn.com]

---- Original message ----
>From: bscott at ntisys.com  

>On Wed, 21 Apr 2004, at 8:13am, travis at scootz.net wrote:
>> I always wondered about this sudden desire to audit the
voting system.
>
>  The issue with the close electoral vote for the US
President in Dade
>County in Florida during 2000 woke up the unthinking masses
to the fact that
>the the mechanics of the voting process are (like all human
endeavors)
>imperfect.

Statewide in FL, not just Dade.  But essentially correct.

Also the advent of pure electronic voting systems where the
entire process is virtual has led to a number of problems
peculiar to the new media.

>
>  A statistician will tell you that the vote in question was
so close as to
>be within the margin for error of *any* polling system.  That
effectively
>makes it a tie.  Of course, many people were not willing to
accept the
>facts, and played arbitrary legal games instead.  Denial is a
common theme
>in human history.
>

Also our electoral system does not allow for ties.  A one-vote
difference will determine a winner.  If there is an identical
vote count some method is prescribed for selecting a winner,
usually by pure chance (e.g. coin toss).

This tends to feed the desire (requirement?) for a system that
counts votes accurately.  Florida 2000 just provided proof
that the existing systems did not meet requirements.

Question is, how to meet those requirements?