How do you know the under vote rate wasn't undercounted in the past and a better count is available when its electronic?
I would assume it should be relatively easy to run a diagnostic program by a third party developer both before and during an election on a spot basis to insure against tampering that could certify the machines working as expected.
(I had reported that the undervote rate went from 700-800 to 4,000-5,000 with the electronics.}
You said:
"How do you know the under vote rate wasn't undercounted in the past and a better count is available when its electronic?"
The way this is known is that, historically, the undervote rate ranges from 1-3%. The paperless electronic system to which I refer caused undervote rates to soar to as high as 70-80%.