Bernie would've won (with new charts)

 So I did my previous Bernie would've won article but given I'm on a roll with my election prediction charts, I decided to make charts for that too.

2016

It's kind of wild how good Bernie's map here is. He starts out with 254 electoral votes, with >12 point margins, and even at 8 points, he would get 281, putting him over the top. Many swing states aren't swing states at all, and he had a near 99% probability of winning. Of course, this was in June 2016, months before the election, but as I covered in my original article, Clinton actually had a polling crash in June and did about as well as she would have on election day. Clinton did have a polling crash every few months showing her odds dropping precipitously to around the 50-50 mark, only to recover again. And she had one around election day, which was why she lost.

Sanders though? No, he had a solid polling lead. He was ahead by double digits in many rust belt states and if it ain't on the chart, and it's deep blue, it's because his polling was that good. Amazing the dems overlooked him, but that's what happens when you're a political machine. Sanders had wild potential in 2016 and if he won we would be in a completely different, and better timeline than today. Seriously.

2020

2020 was a bit of a different map for Sanders than 2016 was. As I stated in my article, he actually did worse in a lot of ways. Some states that on the Biden map weren't even swing states like New Hampshire and Virginia were competitive, the sun belt was gone, outside of maybe North Carolina, and while Sanders held the rust belt (minus ohio and iowa), if there was similar polling error to what occurred for Biden in 2020, he would have lost. 

Still, in an alternative history scenario, it's possible a lot could've changed polling wise from April on. It's possible that Sanders' numbers would've improved, or did worse, who knows? On paper his probability was higher, but idk what an April 2020 Biden would've looked like probability wise either. Many of my predictions had bonkers numbers for Biden that ended up panning out, because I started weighting the polling averages, although Sanders does slightly overperform my corrected Biden 2020 prediction.

Still, I admit that Sanders was potentially a weaker candidate than Biden was in 2020. he still could have won, and he would've been favored in my forecast, but the 2020 race was just...different. 2016 really was sort of a realignment. I was hoping at the time it wasn't a full realignment but just a 1968 style DEalignment where the left could've won those voters back, but I'm guessing a lot of the voters who Sanders had appeal with in 2016 but not 2020 realigned to become MAGA voters. And I suspect that starting in 2020, we started getting pigeonholed into this suburban strategy a la Clinton. And that explains a lot of the weaker polling in a lot of states. VA and NH on the table, WI/MI/PA still disputed and not solid blue like in 2016 for Sanders, yeah. Sanders still had a path to victory, but it was admittedly narrower and Trump could've possibly closed it off with enough luck and effort. 

I still argue on the graphic itself that Bernie would've won by virtue of him being favored in my model, even slightly more than Biden was on election day by this same methodology, but whether the maps are comparable is debatable, and it wasn't certain that Sanders could've pulled it off and won. 

Still, he wins on paper, and this is an on paper victory, so I'd call it for sanders, but with a ton of footnotes and asterisks about everything that could've gone wrong.