Some people like to compare the Presidential results by State or by Congressional or State Legislative seats but that is boring and probably useless. That is why I dug into 4000 years of Solar Eclipse Data to get the people the real answers!
The first thing off the bat is that out of the 398 total eclipse paths I analyzed Joe Biden won 211 of them for a win percentage of 53.01% when comparing the 2-way vote.
Next, we are gonna take a deep dive into just the Eclipse paths Biden won and the ones Trump won. While Trump did win the most recent 2017 solar eclipse Joe Biden won the earliest one in 1994 BCE
The next most obvious place to look is the Solar Eclipse path that covered the most counties as well as the Solar Eclipse path that contained the most votes. Joe Biden won both of these paths even though one runs East to West and the other North to South
Since some Solar Eclipses can only be viewed from just a couple counties looking at the strongest paths for each candidate it is important to have a strong sample size so this is each candidate's strongest margin with at least 100 counties in the path.
Okay enough about margins give me raw vote winnings. Well you got it. Here are both candidate's paths that had the largest raw vote lead they achieved. Biden's largest was ~5.6 million votes while Trump's was ~2.4 million
The closest Solar Eclipse path between the 2 candidates is this one here. There were over 16 million votes cast in this path and they are currently separated by less than 2000 votes! This will be an important Solar Eclipse path to watch in the next election.
Finally here is the most recent Solar Eclipse that was viewable in America the 2017 Eclipse. Trump won this path by a strong showing of 57.55% in the two-way vote.
There will be another Total Solar Eclipse in America on April 8, 2024. I have to imagine that candidates will be doing everything they can to run up their margins in that path to make sure they can clinch the White House.
Shoutout to NASA for making 5000 years of Solar Eclipse data available and some random French website for turning that data into Shapefiles (even if I had to download them 1 at a time). Also thank you to @rudnicknoah and @Aseemru for helping me find the election data by county!