I think the concern is how do you isolate bad ballots(having come in after the deadline dates) from good ballots when they are all mixed together when all other numbers match. You can’t but if the numbers are enough that can significantly throw an election then the whole batch has to be disqualified.
Physical ballots and their counts must match the data tape/file/numbers. If they don’t then there was a programming issue with the tabulators. Physical ballot numbers must match the polling logs of the precincts of those registered voters of the precinct who are recorded as having voted that particular election. There will be some who may not have voted. Any number over 90 percent starts to look suspicious and anything over 100 percent not supported by same day registration additions shows over voting.
You are getting away from the discussion. It has nothing to do with isolating ballots and everything to do with the total number of ballots counted. You can still have 100 physical ballots, even if you isolate them, but if the count shows 225 ballots counted then the question is where is the missing 125 ballots. If you then combine the missing 125 ballots with the accusations that multiple ballots were scanned, this becomes evidence in the litigation proving that fraud did indeed occur.
You say there is a “programming issue” with the tabulation if physical ballots don’t match up. I would proffer this would also be the same digitally with the number of voters who actually voted. However, digital records are much more easy to manufacture and to manipulate than physica ballots. Now back to this “programming issue”. I started out by saying that poll workers were accused of running ballots multiple times through the scanner. How does this become a “programming issue” when the scanner is counting ballots presented to it. There is no “programming issue” when the scanner was “programmed” to only count ballots presented to it. The multiple runs is no more a “programming issue” than the sky is blue. In this case it is a “user issue”.
Its funny, in the business world c-level execs are more apt to not believe what is in their database versus their gut. Yet when it comes to elections, everyone must trust the data even when the data has been compromised. In an open and fair election the data should not, in theory, be compromised but when there are actions of multiple scans of the same ballots then the data has been compromised and human manual counting should be done. This includes an independent poll worker, a demoncrap and repube eyeing each and every ballot and then agreeing on the vote for the candidate. Any ballot in question should be placed aside and investigated further. This counting process shooud in theory match the count from the machine scanner. If not, we have an issue of the fourth kind happening.
Like I said a total overall ballot count should be done not worrying about the vote cast at this point. This would indcate the numbers don’t add up. The reason for this is that if multiple ballots are run multiple times through a machine then numbers would not match to total ballots cast. This is before looking at what is on any ballot. Everyone is concerned wit what is on the ballot neglecting the total number of ballots cast and acting as if there were that many total ballots cast.
Again, the way they check the machine is pull random ballots from the pile of ballots and run it through the machine. If the machine comes up with the correct number they certify the machine is correct. If a person runs multiple ballots through a machine that has been certified as correct, this still does not address the accusations ballots were run through the machine multiple times as the machine will give a count whether it is fed 100 unique ballots or 10 ballots run through 10 times. Each time the machine will determine it saw 100 ballots and will have reported it has counted 100 ballots.