X-no-Archive:yes
I question this interpretation.

Suppose for a moment that you are that custodian, and you count some
results, and find that they differ from the machine. What would one of your
first actions be? I assume that pretty quickly, you would count again. Now,
there are some counting techniques that can be used to ensure the accuracy
of the manual account, by assuming its inaccuracy. That is to say, should
you question your manual count, certain techniques would allow you to
isolate any error to some small and easily countable subset of the thing
counted. One obvious example is counting money, where some number of coins
or bills are grouped together, and can quickly and easily be recounted, so
that if you come up one short, you can find that one set of coins or bills
actually only has ninety-nine pieces and not one hundred as your first count
indicated.

However, machines fail, get out of calibration, and so on. Or some bright
fellow decides that the stock that the ballots are printed from can be of a
substantially different weight than that specified. So, I certainly see the
validity in double checking a representative sample of the machine-counted
ballots by human effort, as an audit check, without necessarily implying
that hand counts are de facto more accurate. They are just another means to
the same end, otherwise the one could not be used to validate the other.

Of course, if the machines are inaccurate, repeating the machine count on a
representative sample should also show this inaccuracy. I understand from
broadcast news that Bush is asking for machine recounts. If these cannot
agree twice in a row, then somebody has some explaining to do.

Greg Stigers
http://www.cgiusa.com

-----Original Message-----
From: Bob J. [mailto:[log in to unmask]]
Sent: Monday, November 13, 2000 2:52 PM
To: [log in to unmask]
Subject: Re: OT: Anatomy of a screw-up

 If I am interpreting 331 correctly, Texas law is declaring manual
counting
to be more accurate and to be used to validate electronic voting.
<snip>