[E-voting] Odd S'Indo Article

A.J.Delaney at brighton.ac.uk A.J.Delaney at brighton.ac.uk
Mon Aug 21 09:22:21 IST 2006

On Mon, 2006-08-21 at 06:32 +0100, Colm MacCarthaigh wrote:
> On Mon, Aug 21, 2006 at 03:48:24AM +0100, bobb wrote:
> > He maintains that an electronic system of counting votes was required because
> > of the difficulty in applying the Weighted Inclusive Gregory Method to manual
> > counting.
> So difficult they could never do it in, say, Northern Ireland. What
> a stupid article.
It does have a certain logic though.  The academic in question seems to
be advocating wholesale electoral reform which would require the
implementation of electronic counting (though not recording) for
practical purposes.  Having spent a few days looking at the Meeks method
for fractional transfers the proposition looks reasonable, though I've
not looked at the "Weighted Inclusive Gregory Method" (could be the same
thing for all I know).

What is being questioned here is vote counting, not vote recording.
Thus the proposition is to change the vote counting software (something
the CEV have recommended). I'm not sure why they are advocating the use
of the existing direct electronic recording equipment.  Which leaves us
with two issues
a) VVAT on the recording side (a must from our point of view) and,
b) cost of verification of the electronically counted votes.

I'll expand slightly on (b).  There seems to be general international
consensus that the count in a proportion of constituencies must be
verified manually.  That, say, 10% of all constituencies must be tallied
manually to ensure that the electronic count is accurate and, above all,
transparent.  Using our current STV count method the manual count will
rarely tally exactly with the electronic count given the randomness at
certain intervals.  However using a less-random count method could
increase the manual verification time to something crazy as all the
randomness in the current system is there to shortcut the count time.

Aidan Delaney
School of Computing, Mathematics and Information Sciences,
University of Brighton.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
Url : http://lists.stdlib.net./pipermail/e-voting/attachments/20060821/153607a8/attachment.pgp

More information about the E-voting mailing list