Official 2022 Congressional Election Results Thread

[cult]

You can actually access NYT internal data quite easily. I took a look at CO-03 and these are the votes they expect to be left in each county. Obviously they underestimated Pueblo county so there could be other underestimations as well

```
Mesa (R+15.6):   648
Pueblo (D+6.3):   -829
La Plata (D+26.2):   297
Garfield (D+13.6):   238
Montrose (R+25.9):   205
Delta (R+29.1):   158
Montezuma (R+14.6):   124
Pitkin (D+58.4):   700
Gunnison (D+38.5):   90
Archuleta (R+8):   156
Otero (R+17.3):   71
Las Animas (R+2.8]:   66
Alamosa (D+1):   179
Moffat (R+59.5):   54
Rio Grande (R+14):   50
Eagle (D+43.4):   41
San Miguel (D+59.1):   40
Huerfano (D+5.2):   38
Ouray (D+27.1):   35
Conejos (R+5.7):   34
Rio Blanco (R+64.8]:   30
Saguache (D+15.2):   27
Costilla (D+34.3):   16
Dolores (R+49.2):   13
Mineral (R+12.6):   7
Hinsdale (R+7.6):   5
San Juan (D+38.3):   5
```

If we abstract this out and expect that all remaining votes are in line with the existing margins, we end up with the remaining votes being only D+9.2. Which is not enough to dump Boebert

```
DEM: 160,918 + 1,817 = 162,735
GOP: 162,040 + 1,510 = 163,550
```

(this ofc doesn't include the negative numbers)

[cult]

Here's what the last few vote drops for CO-3 have looked like:

2,343 new votes at Nov 9, 6:59 PM
--> R+0.4 1,167:1,176 (Frisch:Boebert)
1,672 new votes at Nov 10, 9:26 AM (+14h 27m 37s)
--> R+26.9 611:1,061 (Frisch:Boebert)
77 new votes at Nov 10, 10:20 AM (+53m 35s)
--> R+61 15:62 (Frisch:Boebert)
2,087 new votes at Nov 10, 12:24 PM (+2h 3m 43s)
--> R+17.3 863:1,224 (Frisch:Boebert)
2,371 new votes at Nov 10, 3:56 PM (+3h 32m 42s)
--> R+18.3 968:1,403 (Frisch:Boebert)
131 new votes at Nov 10, 5:19 PM (+1h 22m 17s)
--> D+71 112:19 (Frisch:Boebert)
3,192 new votes at Nov 10, 8:39 PM (+3h 20m 6s)
--> D+0.4 1,603:1,589 (Frisch:Boebert)

A small part of me is hopeful that there's a blue shift happening, but looking at where the votes are left and how many are left takes all that hope away lol

[cult]

Was Oregon called or was it premature?

Local journalist's thread on why they felt comfortable calling it:
<tweet>

Still logic a bit sketch when reviewing comments.

Oregonian not traditionally known for a rag which calls election results, but still that doesn't necessarily make their call wrong, but could be a bit of a stretch for a "scoop".

Oregonian should "discuss with maps", show precinct swings, and effectively explain their data points and analysis for all of us rubes...

OR-05 was absolutely callable to anyone paying attention. Just look at NYT's numbers on this race:
```
OR-05 (lean-rep)
• Mcleodskinner (D) vs Chavezderemer (R)
• Chavezderemer leads by 6,792, 53,960 expected remaining votes
• 84.5% reporting. 5 of 6 counties have expected votes left
   - Clackamas (D+2): 30,244
   - Deschutes (D+5): 11,167
   - Linn (R+36): 4,760
   - Multnomah (D+59.7): 3,120
   - Marion (R+37.2): 4,668
• Abstracted result: Chavezderemer leads by 7,205
```

[cult]

Was Oregon called or was it premature?

Local journalist's thread on why they felt comfortable calling it:
<tweet>

Still logic a bit sketch when reviewing comments.

Oregonian not traditionally known for a rag which calls election results, but still that doesn't necessarily make their call wrong, but could be a bit of a stretch for a "scoop".

Oregonian should "discuss with maps", show precinct swings, and effectively explain their data points and analysis for all of us rubes...

OR-05 was absolutely callable to anyone paying attention. Just look at NYT's numbers on this race:
```
OR-05 (lean-rep)
• Mcleodskinner (D) vs Chavezderemer (R)
• Chavezderemer leads by 6,792, 53,960 expected remaining votes
• 84.5% reporting. 5 of 6 counties have expected votes left
   - Clackamas (D+2): 30,244
   - Deschutes (D+5): 11,167
   - Linn (R+36): 4,760
   - Multnomah (D+59.7): 3,120
   - Marion (R+37.2): 4,668
• Abstracted result: Chavezderemer leads by 7,205
```

This is absolutely terrible abstraction, has very low confidence, and also implies that we should "obviously" call AZ-1 and AZ-6, and that the call in CO-3 wasn't premature.

Abstracting like this requires that samples be representative. We don't know that the existing sample is representative at all.

EDIT: No, this cannot be vindicated as a "good call" if it ends up being right any more than using a coin toss to predict it can be. Making this call look valid would involve demonstrating that the sample is reasonably representative.

Sorry I posted this without explanation and confused people. I use the "abstraction" as a quick base level scenario. Definitely wasn't trying to "make a call" or whatever tf. The main thing I was trying to draw attention to was simply where the remaining votes are coming from

I'm working on a simulation that will look at the remaining vote and analyze the most recent vote dumps and see what those margins are and try apply a bias to the remaining votes based on that.

But that's also irrelevant to my point. Which is, if you just took a look at what the current margins are and where the remaining votes are from it'd be very obvious that a flip is a pretty extreme stretch