On 11/23/2016 02:14 PM, Kurt V. Ullman wrote:
On 11/23/16 3:48 PM, (PeteCresswell) wrote:
Per Kurt V. Ullman:
Past things I have seen indicate that the computer machines have a
lower error rate than the other two (especially to the extent that
things like creases, etc., make it necessary to hand count the optical
machines). It could just as easily be that the computer is more accurate
and the problem is with the paper ballots and optical scanners.

I did not notice any reference to whether-or-not the suspect districts
had paper backup.
They didn't. They were comparing districts who had machine voting to
those with the other two (hand counting and optical readers) and saying
that the voting machine districts were wrong because of the difference.
I was noting that the differences could have just as easily been because
the machines are more better.



http://fivethirtyeight.com/features/...ction-results/