Statisticians have told us the numbers need to be increased both in the L & A AND the mandated hand counts after elections. The percentages stated in the mandates are minimums and too small to catch the errors. (Funny how they provide ‘Best in Class” by actually doing the minimum)
I noticed this in Maricopa County’s contract with Dominion
This is what it does.
Maricopa, after testing it, bought it.
I have no idea if other jurisdictions use them, just thought it may be pertinent. I have been thinking about the possibility that if someone knew a program well enough to know what triggers specific actions, it is possible to program a test deck to be generated that does not contain that trigger. That would enable the device to pass the test but only because the trigger did not occur in the test.
It could also explain the recent election results that were announced before the election.
p.s.
Maricopa County did not surrender BMDs ballot marking devices under the subpoena for the people’s audit conducted by Cyber Ninjas. The Countys’ position - the subpoena specified tabulation equipment. BMDs are not used in the tabulation of the vote. Just like archives were not specified. “Cyber Ninjas did not ask for them.”