John D. Giorgis [mailto:[EMAIL PROTECTED]] wrote:
>
> I personally don't think that a manual recount is a good
> idea. The prospect of having humans pull 14 hour days
> reviewing 6 * 10^6 ballots is a poor way of calling an
> election. Machine recounts are very efficient, and moreover
> are unbiased in the number of ballots that they miss.
While I don't dispute that humans are poor at doing repetitive tasks,
machine counts are not necessarily unbiased. I've heard mention that the
punched holes usually leave a flap of paper attached which can randomly
cover or uncover the hole during the process. While one would hope that
this incident is small, and further that it is completely random and
therefore wouldn't affect the overall results, it is entirely
conceivable that the *position* of the punch on the paper is a factor in
whether or not this occurs - for example, punches towards the center of
the card could be more likely to exhibit this covering effect than
punches towards the outside of the card.
Pure speculation, of course, and I doubt that's the case. But never
assume blindly that machines are unbiased.
Joshua
