Fascinating. So for the final vote total, Democrat = Republican + Libertarian + about 2250 votes. Almost as if it's partitioned.
About a week after the election, I ran some numbers. Take before-electionknown party registration percentages, for each county, for mail-in votes only from the Pennsylvania data. Assume all registered Libertarians and Republicans voted for R or L, and all registered Democrats and all other party registrations -- 100% -- voted D. I found, when you sum it up for the entire state, that this predicted the mail-in vote except Democrats came out 2400 votes ahead.
It'd look realistic in retrospect, that the number of R+L voters and the number of R+L votes matched, for mail-in. Until you realize it means 100% of independents supposedly voted for Biden. 100%. Plus 2400.
I have this analysis in SQLite format (with accompanying CSVs and scripts) if anyone is interested.
Fascinating. So for the final vote total, Democrat = Republican + Libertarian + about 2250 votes. Almost as if it's partitioned.
About a week after the election, I ran some numbers. Take before-election known party registration percentages, for each county, for mail-in votes only from the Pennsylvania data. Assume all registered Libertarians and Republicans voted for R or L, and all registered Democrats and all other party registrations -- 100% -- voted D. I found, when you sum it up for the entire state, that this predicted the mail-in vote except Democrats came out 2400 votes ahead.
It'd look realistic in retrospect, that the number of R+L voters and the number of R+L votes matched, for mail-in. Until you realize it means 100% of independents supposedly voted for Biden. 100%. Plus 2400.
I have this analysis in SQLite format (with accompanying CSVs and scripts) if anyone is interested.