Don White wrote:

While flippin' channels yesterday, I breezed through CNN.
There was some king of panel with Wolfe Blitzer, that tall lanky guy, Jack
McClafferty? and a couple others.
They were talking about some voting irregularities in Washington (state?) by
the Republicans. Apparantly, the people counting the ballots figured they
could predict the outcome at a certain point in their counting..so packed it
in before finishing. What the &*^?
Canada may have to start sending down observers to make sure things are run
on the up & up. Afterall someone has to help protect that fledging democracy
you'all call the U.S.of A.

Statistics. You don't need to know how *everybody* voted. So if you
have a statistically significant sample, that's enough.

So, if you took a statistically significant sample of the *original*
sample, you wouldn't have count all of *them*, either.

So, if you took...well, you see where I'm going with this. We really
only need to count one vote, then we can extrapolate from there.

DT