Posting this under a throwaway for obvious reasons.

I volunteered to do some analysis on the election data because I'm more interested in finding the truth than being partisan. During the counting process, the relative advantage of a ballot reporting batch toward Biden has strong correlation to certain external regressors that are meta to the counting process itself. These correlations absolutely should not be there, and yet they occur in swing states with p-values in excess of 10^-40. I can't say which external regressors, because this is part of an active legal battle, but suffice it to say, something strange is afoot. It's better to perform an audit and prove no fraud than to be ideological about the impossibility of fraud.