[thecodont]

But the tabulation of undervotes tells me that they kept had to have a way to know which page 1 goes with which page 2 - which suggests some kind of serial/ballot number that matched page 1 to page 2. And if that was the case, it makes no sense for the software designers (literate people being paid to do a specific, detail-oriented job) to use the number of pages cast to calculate the “% Turnout” on their summary report.

For my three-card ballot I cast on Tuesday, here's how each card was numbered (per the strip torn off each card I got to keep as a sort of receipt):

Side One: A-Card 1, [number1] [(number2)] [number3] BT:3 (right side)
[number4] - POLL, [number5] [barcode square graphic]
Side Two: [number4] - POLL, [number5] [barcode square graphic]

Side One: B-Card 2, [number1] [(number2}] [number3] BT:3 (right side)
[number6] - POLL, [number7] [barcode square graphic]
Side Two: [number6] - POLL, [number7] [barcode square graphic]

Side One: C-Card 3, [number1] [(number2}] [number3] BT:3 (right side)
[number8] - POLL, [number9] [barcode square graphic]
Side Two: [number8] - POLL, [number9] [barcode square graphic]

There were three numbers consistent between the three ballots, and two which were unique to the ballot. I believe the "BT:3" indicated there were three cards per ballot.

[thecodont]

There were three numbers consistent between the three ballots, and two which were unique to the ballot. I believe the "BT:3" indicated there were three cards per ballot.

Ack. I meant to say, there were three numbers consistent between the three CARDS, and two which were unique to each CARD.