So, I decided to jump the gun and do another election prediction. A few things have changed and I kinda got OCD over this.
2 way national polling- Trump +2.3%
5 way national polling- Trump +3.5%
So the 2 way vs 5 way has narrowed somewhat, but still slightly favors the GOP.
Going with the state polls:
I'm axing
Maine CD2
because it's only gotten worse since last time and I really don't wanna get in
the habit of covering states above 12 points. It's up to 17 now. Trump
literally has a 100% chance of winning that.
I'm also gonna change my estimation for Nebraska CD2, which I don't have data for yet. Basically, before I was just tying it to minnesota but Im gonna do a direct calculation by calculating how much things have deviated since 2020. In 2020, Biden won the popular vote with 4.5%, and Nebraska CD2 went blue by 6.5%. So I'm going to say that NE2 is just 2 points bluer than the popular vote margin. That would place it at Trump +0.3%.
Anyway the overall result is mostly the same. 294-244 Trump. The NE2 result is mostly due to a change of my methodology, rather than tying it to Minnesota I'm tying it to the national environment overall. I actually have no clue what NE2 is doing. Biden has a 18.4% chance of winning, Trump a 81.6%, with the deciding state being what I've been discussing recently, MICHIGAN. Seriously, get to working on that ceasefire, Joe. We kinda need it to win here. The pro Palestine people arent super strong on a national level but all you need is a few hundred thousand in a single state to decide the whole thing here.
A few other things to discuss here. Michigan did get a little less red, that's why Biden's chances went up. Still, Wisconsin, NE2, and Pennsylvania are in coin flip territory. Maine proper, yeah, they added some more polls including a pro trump poll that pulled the average way down to swing state territory. It remains unclear if this is accurate, but I just report the numbers as they are, and that's what they're saying. But yeah, beyond Michigan there is a massive cliff, similar to what Biden had in 2020 with the rust belt states that are going to be difficult to surmount. AZ, NC, GA, and NV are all just barely swing states. TX and IA have been that high. And yeah. That's the electoral map.
It gets worse with 3rd party candidates included. I ain't like to include them too much since the data is a lot more spotty at the moment, but I can do a smaller chart like last time showing JUST the 7 most covered swing states (which show the most interesting part of the data anyway).
All in all things don't change a ton in net going to the five way results. Wisconsin and Pennsylvania both become the deciding states, with Biden needing to win both of them to win, and Trump only needing one. Michigan becomes more likely to go blue slightly but not a ton. Again, better work on that cease fire, Joe. Biden's chances drop slightly to 15.9%, while Trump's go up slightly to 84.1%. It should be noted that NE2 is now technically in the red column for Trump here at 1.5% ahead, but it wasn't the deciding state. Still, Biden needs that or he would need to win another state on top of the three rust belt ones entirely. And that's where....Biden's chances drop to 3%. The wall is even harder to transcend here, with the next states being 7.5%+ in Trump's favor, giving Trump a 97%+ chance of winning them. God, I wish we had some NE2 data. That little district could decide everything. Keep in mind even if Biden wins Wisconsin, Pennsylvania, and Michigan, we could be stuck at 269-269 if that little district goes red. Just food for thought.
Anyway, that's my prediction. In net, a movement in Biden's direction in probability, although changing my methodology for NE2 in the absence of polling data puts it slightly in the red column. For now I'll still say the two way predictions are my official predictions, although the 5 way ones might be more accurate given those candidates will exist too. All in all, you can generally take my 2 way predictions and add 0-4 points for Trump depending on the state. About 2 on average.
No comments:
Post a Comment