Just like polling can be polluted by biased stratified sampling, machine failures at a precinct level can be disproportionately affected by maliciousness.
Hopefully Lake’s team brought strong evidence showing a pattern of intent vs random machine failures. Meaning the machine failures occurred disproportionally in R heavy precincts than in D ones. A true “ bug” in the system would behave without a discernible pattern .
Just like polling can be polluted by biased stratified sampling, machine failures at a precinct level can be disproportionately affected by maliciousness.
Hopefully Lake’s team brought strong evidence showing a pattern of intent vs random machine failures. Meaning the machine failures occurred disproportionally in R heavy precincts than in D ones. A true “ bug” in the system would behave without a discernible pattern .