So..I've been thinking...

Trump's internal polling sounds bad, given he's whining already about PA being rigged. Harris's polling sounds good, with rumors coming out that the last minute crowd is breaking for Harris by double digits (something like 55-44 IIRC...). The selzer poll predicts a harris landslide if applied to the rest of the rust belt. And even if it's off by its full margin of error (about 3.4 points, so almost 7 points both ways), it still has harris overperforming by like 3, which is enough to comfortably win PA, MI, and WI. I'm not sure if the poll translates to sun belt demographics well as southerners are just built differently politically, but quite frankly, we don't need the sun belt if we win the rust belt. It's nice to have, but it was always less essential. Still, if we see a harris +3 trend nationally, we should comfortably carry NV and NC, with AZ and GA being more tossups. 

And you see where I'm going with this? This is where things stand now. 

Trump is ahead, but nominally. I've been suspecting the WI and MI numbers are too conservative for Harris, and they should be 1-1.5 points further left or so. PA seems...accurate based on the polling, including the anti herding measures I've attempted to use to estimate where the states really are (basically averaging the most extreme polls in each direction). That puts PA at exactly 50-50 (R+3 and D+3), but WI and MI should be D+1.5 or so here. Still, if the internal polling is leaning toward Harris and say, the real Iowa number is like R+5 (assuming R+7 is the baseline for a rust belt tossup), we could be seeing a real situation of like D+1-3 or something. Idk. If I had to guess, and be bullish on Harris, im not gonna do it to insane levels. But I would do it on a level of D+1-3 to my current forecast. D+1 flips PA, D+2 flips NV and NC. D+3 flips AZ and GA. 

Let's look at that in practice:

D+1:

D+2:

 D+3:

 When i think about it, I guess the D+1 is the most likely scenario. Why? Because let's face it, these effects may be limited only to rust belt states. Sun belt seems to be a different beast and polling is far less hospitable there. if I do my anti herding measures there, if anything, I might expect them to be just as or more red than my official forecast. So...given D+2 would flip NV and NC, I would say that's out, even if the margins up north are more blue. And D+3...well, I'm not gonna dare say this flips AZ or GA.

But could this mean Harris wins the rust belt? Sure. It might even mean she wins it quite comfortably. We could see D+3+ in the rust belt and then the sun belt is what we expect. NV is always a wild card, I could see that maybe flipping blue too, although not with current polling pulling it so far right so fast. 

So yeah. If I had to guess, while my official forecast currently shows 251-287, I think the real result is likely to be 270-268. 

I will make my final determination on Monday/Tuesday, but yeah. I'm leaning toward actually predicting 270-268 in Harris's favor here. I am increasingly bullish on a harris overperformance in the rust belt based on last minute factors, and those last minute factors matter. I said it the whole time, the side that has the energy at the end of the campaign is the side that's gonna win. That's why i was freaking out over trump surging all of october, but in november, we're starting to see the rust belt shift back left, even as the sun belt continues to shift right. And the selzer poll could actually be correct. Keep in mind the herding situation is making polling less reliable than it is. Maybe that poll is that asch experiment person who tells it like it is while everyone else just herds to avoid being called out if wrong. You know? So...yeah. That's how I see it. I dont actually think Harris is gonna win iowa, although this poll does bring the polling average down there. It's possible. I'd actually kinda give it low key swing state status based on the fact that that poll changes iowa from a likely to solid R to lean R. 

So yeah. That's where I'm at.

I think in my final forecast, I'll do one prediction just based on polls, but much like my earlier 2008 and 2012 predictions, which actually ended up being far better than i remember and actually better than the polls only version of my forecasts, I think I will do a second forecast of what I think what will happen, and unless something changes my mind in the next 2-3 days, I'll probably go harder for Harris, while still remaining cautiously optimistic.

I mean, let's be real here. I don't see a D+10 relative to my forecast tsunami happening, heralding the age of blue texas and blue florida and 400+ electoral count victories. But something like D+1-3 in the rust belt seems very plausible. Could mean we hold the rust belt seats in the senate better as well.

We'll see. I do just wanna say, I think Harris is gonna win regardless of what the polls overall say, because polls can be wrong in either direction by several points, and the signs are pointing toward a Harris overperformance. My formal forecast has Trump at a 54% chance. That's a coin flip. Harris still has a 46% shot, and again, the signs seem to be indicating that a pro Harris outcome is likely here. 

So yeah. 270-268. Seems to be where I'm heading with this.