,
4 tweets,
1 min read
Read on Twitter
Much to dislike about this voting architecture. The barcode isn’t the biggest problem. The ballot has only a summary of the selected candidates, in a different format from how the vote was entered. Particularly hard for voter to confirm, as noted by the Natl Academies report.
A proper risk limiting audit would use the text, not the barcode, but that text has to actually reflect the voters’ intent, an assumption that this design makes especially dubious.
Bottom line when examining a proposed voting product: Does it conform to the recommendations of the 2018 National Academies study? If not, that’s bad.
I should add that I’m not as convinced as some of my colleagues that BMDs are inherently bad; we still have a lot to learn here about the tradeoffs. But current BMD systems are clearly not consistent with the National Academies report, and that should carry SIGNIFICANT weight.
