So...one week until election day. Let's see where things stand.
Presidential
It seems like Trump's upward momentum has slowed to stopped. I wouldn't be too overly optimistic though. Things are still flipping back and forth. We're getting tons of polls in given we're now in the last week and things are flipping erratically. Michigan has flipped to the blue column, pennsylvania is trending blue slightly, but then wisconsin trends red. Arizona is all over the place. I forgot to resort the map for nevada, but that's trending blue. All in all, we're at 45-55 in trump's favor. Even though i give the advantage to trump, the real reason i was dooming last week is the momentum. But it seems like trump's overall momentum has slowed or even stopped.
I wanna make things clear. With probabilities like this, a stiff breeze can push the race in either direction. I'm serious, the MSG rally might've just done it for Harris. Latinos are OUTRAGED, and the environment here in my majority latino city is one of anger toward trump. There's a lot of mobilization to vote for harris here right now. Canvassers are out, convincing latinos to vote for harris, I've seen them reaching out to my neighbors already. Yeah.
Like, there are half a million latinos here in PA and while latinos vote republican by a 1 to 2 margin (meaning 2/3 vote democrats), i think insulting them is gonna drive dem numbers up. This might be the actual october surprise, the 2024 equivalent of the james comey thing. And the worst thing for trump was this was just a total unforced error.
But yeah. I wanna point out, if Harris overperforms by even half a point, she wins 276-262. Basically, you know simulation map #18 from my simulations? Yeah. That.
That's all it will take to win the election. Just half a point...
And I've seen articles about how polling methodology changes could shift outcomes by as much as 8. So we have no idea what's going on here. The election margins are razor thin and it's possible any outcome within my 8 point 95% confidence interval can happen. That can range from a 389-149 outcome for Harris to a 194-344 outcome for Trump. Anything is on the table just about. My simulation mostly gives me outcomes involving the 7 swing states but it can go all over the place.
Speaking of simulations, yeah, expect blue kansas to start happening on occasion. R+5 poll dropped for kansas yesterday, it's the only poll for kansas on both 538 and RCP, so...yeah. I don't expect it to be an R+5 state, more like R+15, I think this is another solid red state with an outlier, but I feel intellectually obligated to include it here given my methodology and model. We'll see if it actually happens but I doubt it. State's probably R+13 and had an outlier or something. Kansas never goes left of Missouri to my knowledge.
Anyway, that's the presidency. I'm feeling less doomerish today. Let's see where the momentum goes this week, because this right here is the make or break moment of the campaign. And if Trump falls apart and Harris keeps it together, boom, we win.
Senate
Ouch. So....the Nebraska race....RCP finally started covering it so I switched to their polling average which is R+2. Who knows if the other 538 polls were accurate? I was using their data just to get something, but yeah now I'm using RCP like I do for everything else.
That puts us back at 49-51, with it being an uphill battle to even break 50-50. Let alone win 51+. We need that Nebraska seat for a tie, and we need texas or florida to actually win. It can happen, it's just not likely.
On the other side what's looking increasingly likely is that we might lose some rust belt seats. Wisconsin is within a point, Ohio is at 1 point. Pennsylvania is just over a point. Michigan is lean D, but still, a republican overperformance could wipe out the rust belt. It's not a good situation. Republicans have a 69% chance to control the senate outright, with the other 31% split between a tie and dem control. Not really good.
House
So I just had to redo the entire house forecast since new polls came in after I posted it, and then a race I assumed to the red column had a D+2 poll and adding it changed my entire forecast. The house is now a tossup, with CO8 being a literal tie breaker. This brings the odds down to 50-50 and makes it dead locked at 217-217. I can no longer project a winner. The simulation still favors republicans though.
Simulations
I'm reverting completely to 100 simulations of each. Giving up on the big simulators. Too complicated to deal with.
Presidency- 47 Harris, 51 Trump, 2 ties
Senate- 1 Dem, 90 Rep, 9 Tie
House- 25 Dem, 75 Rep (EDITED)
Basically, congress heavily favors republicans in my model, but keep in mind that model uses a much different formula for calculations different from my actual forecast. it's good to see what can happen but not at what WILL happen.
Conclusion
Anyway, I'd say at least trump isn't gaining much ground. We are losing it in congressional races somewhat, but despite the simulations, anything can happen, and the simulations are flawed. The presidency is a tossup, the the house is a tossup, and the senate is only lean R. We are still in this. And given the MSG rally, i think that's gonna hurt him. So we can still turn this around. Just push for the last week and do the best we can.
No comments:
Post a Comment