A very interesting analysis, Peter. I’m looking forward to further discussion of how data anomalies might be produced by injecting disinformation (i.e. vote fraud) or by natural phenomena.
Regarding “drop-off analysis”, the logic you present assumes that it is somehow easier or more probable for fraudsters to rig presidential vote without rigging other contests, i.e. that high drop-off suggests fraud. I had assumed the opposite, and have used the argument that the fact that in 2020 Trump lost while Republicans in general won (high Republican drop-off) suggests lack of fraud because a blue-fraudster would likely have switched votes on multiple contests, not just the presidency. You do a good job of pointing out that anomalies are subject to multiple interpretations, and this is an example.
Thanks for pulling all this together Peter! Still trying to wrap my head around all the data in here!
A very interesting analysis, Peter. I’m looking forward to further discussion of how data anomalies might be produced by injecting disinformation (i.e. vote fraud) or by natural phenomena.
Regarding “drop-off analysis”, the logic you present assumes that it is somehow easier or more probable for fraudsters to rig presidential vote without rigging other contests, i.e. that high drop-off suggests fraud. I had assumed the opposite, and have used the argument that the fact that in 2020 Trump lost while Republicans in general won (high Republican drop-off) suggests lack of fraud because a blue-fraudster would likely have switched votes on multiple contests, not just the presidency. You do a good job of pointing out that anomalies are subject to multiple interpretations, and this is an example.