[ExI] FW: elections again with punchline

[spike]

http://neggie.net/vote2008/nh_primary.cgi?min_votes=300&max_votes=900

...I entered 300 lower limit and 900 upper limit, and discovered this, after
doing the calculations:


			Hand count(%)	Machine count(%)
(machine-hand)/hand
...

Clinton		34.0			40.0			+17.6
Obama			42.1			35.5
-15.7
Edwards		18.1			19.3			+6.6
Richardson		5.8			5.2			-9.9

Now if we swap the number of machine counted votes for Sen. Clinton and Sen.
Obama, we ge the following numbers:

			Hand count(%)	Machine count(%)
(machine-hand)/hand
...

Clinton		34.0			35.5			+4.3
Obama			42.1			40.1			-5.0
Edwards		18.1			19.2			+6.6
Richardson		5.8			5.2			-9.9

Now isn't that interesting?  The swapped results agree with the hand counts
much more closely than the results being reported.  If someone in the
Diebold company managed to somehow insert a line of code that would swap the
count for Sen. Clinton with those of Senator Obama, we might expect to see
results like what we see in the first set of numbers above.  Otherwise, the
second set of numbers.

Computer jockeys among us, how difficult would it be to write in a line of
stealth code to do this?  Could it be an innocent and undiscovered bug?
Statisticians among us, are these finding statistically significant?    

My friends in democracies all over this restless planet, do think about this
carefully: if there were no hand counted paper ballots, we never would have
seen this apparent discrepancy.  To me these are numbers that are screaming
for attention.

spike