In a message dated Fri, 10 Nov 2000 9:20:08 AM Eastern Standard Time, Erik
Reuter <[EMAIL PROTECTED]> writes:
<< No joke. How is THROWING OUT SOMEONE'S VOTE any less likely to cause
problems than counting them as partial votes? Also, I used normalize in the
mathematical sense, I said "normalize to 1" not `normalize votes' as
you misquote me. >>
I understand what you're trying to do. My point is, most people seem to have
an enormous suspicion of using mathematical and statistical techniques to
accomplish important political tasks. That was why I gave the example of the
huge uproar caused by the Census Department's proposing to use sampling
methodology to increase the accuracy of the decennial census. That caused a
political battle on the Hill and the threat of court cases. Can you imagine
the uproar if anyone seriously proposed using statistical techniques to
estimate the percentage of a disputed vote different presidential candidates
received? It would cause a political upheaval not seen in this country in
generations.
There are ways to improve the accuracy of an actual count of presidential
votes without resorting to statistics. That doesn't help in this situation,
but it's something we should be considering for the future.
For now, you have to prove that people actually did have their votes
invalidated or thrown out. And if it happened in Florida, it probably
happened elsewhere. Of course, the margins elsewhere were nothing close to as
paper-thin as in Florida, so it doesn't matter quite as much.
It's for the people whose votes are at dispute to decide what to do. The
presidential campaigns should stay out of it. (And keep in mind, I'm a strong
Gore supporter.)
Tom Beck
