3 ms·
First you need to get all the states to hand them over. Many seem unwilling. As for analyzing, you don't necessarily know what you're looking for before examin
by auganov 6y ago
First you need to get all the states to hand them over. Many seem unwilling.
As for analyzing, you don't necessarily know what you're looking for before examining them. For example - if there is a distinct pattern of markings suggesting ballot stuffing by the same person, you'd need to identify those first. Then you could train a program to identify all occurrences. But without ground truth (knowing for a fact you really spotted fraudulent ones) this wouldn't be strong proof. Absence of anomalies also wouldn't prove absence of fraud, might merely disqualify some forms of fraud.
- tmaly 6y agowould have an independent tabulation on different software apart from the software used by the official machines offer anything here?
- auganov 6y agoYes. But problems with tabulation have already been rendered unlikely by hand recounts. Many/most places haven't done them or have only had limited ones, so this would add some extra reassurance. But out of plausible theories of fraud, rigged tabulation without creating a trail [0] is far down the list. [0] and even if you do, it's still not that plausible