[extropy-chat] An odd discrepancy in exit polls in touch-screen electronic voting states

[J. Andrew Rogers]

On Nov 6, 2004, at 2:25 AM, Gennady Ra wrote:
>     The hot story in the Blogosphere is that the "erroneous" exit 
> polls that showed Kerry carrying Florida and Ohio (among other states) 
> weren't erroneous at all - it was the numbers produced by paperless 
> voting machines that were wrong, and Kerry actually won. As more and 
> more analysis is done of what may (or may not) be the most massive 
> election fraud in the history of the world...


Tinfoil hat nonsense and propaganda.

As has been hashed out in slightly more technical forums, the polling 
methodology and internals on those early exit polls were badly broken.  
The supposed "Kerry lead" turns into a slight Bush lead when the 
internals are normalized, something that was pointed out by other 
bloggers in less than an hour of those exit polls being released.  Not 
only were those exit polls very questionable in terms of sample 
population, no attempt was made to normalize the results of the skewed 
polling data to provide a more accurate picture.

But more generally, that still doesn't explain how Bush managed to get 
significantly more votes than 2000 in many precincts run and controlled 
by the Democrat Party and where Diebold was nowhere in sight.  Did the 
Democrats rig their own machines against themselves?

Any voter fraud that went on this year clearly falls below the noise 
floor.  It is not constructive to assume that Democrats automatically 
should win every election, and that if they don't it simply must be due 
to Republican fraud.


j. andrew rogers