[jimsug]

Not sure where 15,000 to 45,000 comes from, as the report itself concludes only c.8500 cases of duplicate voting.

I'm also not sure about the methodology there, so perhaps someone could explain it to me.

From what it looks like, GAI started with 60,000 matches from the state data. Then they... added additional identifiers and confirmed c.7000 of them? How do you get from uncertain data to more certain data in this way?

There seem to be c.15,000 instances of prohibited addresses being registered, which I don't believe alone indicates voter fraud.

[lenkite]

"Extending GAI’s conservative matching method to include all 50 states would indicate an expected minimum of 45,000 high-confidence duplicate voting matches"

GAI was unable to conduct a comprehensive review since a complete data set of state voter rolls is currently unobtainable. (it was denied)

[jimsug]

I don't quite understand why the expect that there would be ~6x the number detected, though, assuming that the ~8500 cases detected is accurate. It would be very (and probably statistically naive) if the minimum total cases was simply because they have only ~1/6 of the total number of state pairings.

I think the other major concern I have, other than the methodology, are the definitions - I still don't know whether 8500 represents 8500 people who voted twice (17000 total votes cast), or 4250 people who voted twice, or something in between, or some thing completely different. Perhaps I missed this.