Arcane Demesne
banner
arcanedemesne.bsky.social
Arcane Demesne
@arcanedemesne.bsky.social
Below is my 2026 GA midterms prediction, saved here for posterity. Very interested to see if it holds up. Has a lot of data that went into it based on VA.

https://bsky.app/profile/arcanedemesne.bsky.social/post/3m57yfdj6y22u
That's what I'm hoping. As per my PM to you, Cheatham could give Dems a big boost if R turnout fell, Dem rose, and there was also persuasion.
November 27, 2025 at 3:55 AM
Also in the Nashville primary Rs did fucking terrible. Lost by like 67% lmao. They better hope *that* doesn't happen by the end of election day.
November 27, 2025 at 3:37 AM
Van Epps did terrible in the geographical core of the state and then did better on the border from Wayne up the west and across the north to Robertson. Did terrible in Williamson and Dickson and Hickman, which is the exact place he needs to do well. Kinda mid in Nashville, too.
November 27, 2025 at 3:37 AM
How much bluer compared to the primary do you think? that was like a 9% spread right? So you see more of a 4%-5% in this data?
November 27, 2025 at 3:19 AM
Sorry I retyped and missed a line. 2022 total votes. %EV is nice too though.
November 27, 2025 at 3:09 AM
Think comparing vs 2022 total is most useful. Don't see that one on there?

Guess we'll see how useful the primary comparison is.
November 27, 2025 at 3:02 AM
Certainly having Jody Barrett lick boot and beg for votes while Republicans amp up the mailers, texts with Trump's face, and tweets is not ideal for Democrats. Really hope they just give up and accept that the race is nationalized now.
November 27, 2025 at 2:26 AM
Then we need 8,000 to 9,000 out of the3 medium population partly suburban counties.
This seems pretty plausible as Dickson has very low turnout and Cheatham is up in comparison.
But we can't very easily be sure what the margins will be in these seats. Need a pretty big swing to get there. We'll see.
November 27, 2025 at 2:25 AM
Dems need to gain, essentially, 40,000 votes off the margin to flip the seat if the total vote share is the same.

Let's say we have 20,000 vs 27,000 from these 2 counties based on the projection.

So can we get 4,000-5,000 of the 8 less populous rural/exurban counties? Most probably.
November 27, 2025 at 2:25 AM
It seems like Davidson could net Dems anywhere from ~6,000 to ~10,000 more votes than in 2022, assuming the same election day vs early voting split in the county ~50%. Taking 21,000 votes to 42,000 votes.

Montgomery also seems on track for a 10,000 vote swing in my projection. 6k-7k from Williamson
November 27, 2025 at 2:25 AM
Could be more like 1,000, but basically the project turnout this year minues the total in 2022.
November 27, 2025 at 2:16 AM
But will they do 22% better? Probably but it is hard to be sure.
November 27, 2025 at 2:15 AM
Yeah but suppose turnout drops12.5% in the rural counties from 15.4% to 13.5%. Were all the people who dropped out Republicans? Was it 50/50? Was it by vote share?

So Dems will do better than10% better, which is around what you get comparing differences in bigger counties as well as rural counties.
November 27, 2025 at 2:15 AM
And then Davidson is 9.25% more of the total votes. So that gets you around a 6,000 vote reduction in the 39,500 vote margin just from Davidson alone.
November 27, 2025 at 2:13 AM
A big issue we have in estimates is we can't really separate turnout from persuasion. So we can't look at the turnout changes and predict the persuasion differences. They are to some degree "dependent variables".
November 27, 2025 at 2:06 AM
For early vote specifically it was low? Where are the daily early vote numbers for Davidson in 2022 and 2024?
November 27, 2025 at 2:03 AM
Why a bigger turnout edge? Davidson lost 2.5 hours. Is this the 8 less populous rurals or the other 5 medium/large counties that did better than you thought?
November 27, 2025 at 2:02 AM
Don't see how. You were just expecting a lot more dropoff in the other counties?
November 27, 2025 at 1:59 AM
The 8 less populous rural/exurban counties account for 12.8% of the early vote so far. Based on their EV/E-day variation it'll be something like 14.2% on e-day, so 13.5% overall. They were 15.4% in 2022. A 12.5% drop in their share of the overall vote. We can expect them to swing Dem by some margin.
November 27, 2025 at 1:58 AM
Dickson is way down, should be 15% higher than Cheatham. Even accounting for their EV vs E-day shares it isn't there. Williamson is down over 2,000 votes vs 2022 while Davidson is up ~1,000 when correcting for EV/E-day shares. Montgomery is up 2,000 in the projection. Great news for Dems all around.
November 27, 2025 at 1:49 AM
Cheatham is typically ~63% of Robertson in the final tally, cross both 2022 and 2024. Robertson is typically 45% of Davidson. If you account for theirEV vs E-day that's about what we see here. That means Cheatham is way up on Robertson and it is the most Dem friendly county of the 3 medium pop ones.
November 27, 2025 at 1:49 AM
These numbers are actually better for Dems than they appear. Williamson as a county is heavily early voting typically while Davidson is in the middle and Montgomery more election day oriented. 55% EV for Williamson, 50% for Davidson, and 45% for Montgomery. 2024 higherEV but same order of intensity.
November 27, 2025 at 1:49 AM
I know on Reddit we suggested people go to Hillwood or Metro. Could account for the low total at Lentz and the higher ones at those two.
November 27, 2025 at 1:37 AM
Total EV Share By County(84,356 Votes):
Davidson | 20,951 | 24.84%
Montgomery | 19,110 | 22.65%
Williamson | 14,777 | 17.52%
Robertson | 6,823 | 8.088%
Cheatham | 6,241 | 7.398%
Dickson | 5,654 | 6.703%

Very good news for Dems in my view. Analysis to follow this.
November 27, 2025 at 1:36 AM