Reply #26: But it can be selective

[Febble]

Here are the deviations between the exit poll margin and the
recorded vote margin (as of yesterday) for the Senate races,
according to downloads of the close-of-poll exit poll
crosstabs by OTOH (close-of-poll crosstabs are not - because
they can't be - adjusted to the incoming returns):


Maine	       -12.6%
Massachusetts	-8.1%
Minnesota	-7.6%
Connecticut	-6.7%
Montana	        -6.5%
Wyoming	        -5.2%
Arizona	        -4.7%
Washington	-4.7%
Ohio	        -4.2%
Utah	        -3.8%
Florida	        -3.8%
California	-3.2%
Virginia	-2.7%
New York	-2.7%
Nevada	        -2.4%
Wisconsin	-2.1%
Texas	        -1.6%
West Virginia	-1.5%
Vermont	        -0.9%
Missouri	 0.0%
Hawaii	         0.4%
Tennessee	 0.8%
Rhode Island	 1.0%
Pennsylvania	 1.4%
Michigan	 1.4%
Maryland	 2.9%
New Jersey	 3.4%
New Mexico	 4.1%
Nebraska	 7.7%



A negative value means that the discrepancy was a
"shift" in the vote count (relative to the exit
poll) in favour of the Republicans; a positive value means a
"shift" in the vote count in favour of the
Democrats.

According to my calcs, there are no outliers (the distribution
is not significantly different from normal on either
Kolmogorov Smirnoff test or Shapiro Wilk, nor is there any
significant skew or kurtosis) but the mean is significantly
less than zero - i.e. there was a significant
"red-shift" (1 in 92 probability of having occurred
by chance).  This does not, however necessarily imply fraud -
it may also reflect pro-Democratic bias in the poll.

But if these numbers are right, it is clear from this that
there is nothing especially anomalous about Montana and
Virginia.  Looks like you would be better checking out Maine
(or perhaps pro-Democrat fraud in Nebraska....?)

But better still, check out 

http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=457470&mesg_id=457470

and, of course, good old voter suppression in Virginia.

And great job winning both Houses!  Now you can get some real
Election Reform, and fraud investigation.  Just make sure you
are really looking in the right places!

Cheers

Lizzie