Tuesday, October 13, 2020

Mid October Election Predictions (10/13)

So it's been two weeks. A lot has happened since then. Trump had an awful debate where he just yelled over Biden and Biden seemed kinda weak in response. Then another debate where Pence tried a similar strategy with Harris and it didn't work well. And then Trump got COVID and recovered, and likely had it at the debate. It hasn't been a good two weeks for Trump. And polling reflects that. Biden is now in a very strong position to win, overtaking anything I predicted in 2016 with Hillary. While democrats shouldn't claim victory just yet because remember 2016, it's looking very good for Biden and I don't see how Trump can realistically recover. That said let's look at the data.

 General Election - Biden +10.0%

 4 Way -  Biden +8.8%

Now keep in mind, I don't think a spoiler effect exists here, rather the polling averages haven't shifted as much in the 4 way from last time due to less polling. Still, in both matchups, Biden has a commanding lead outside of a 4 point margin of error. The 2 way results report double digits between Biden and Trump. In both the 2 way and 4 way polls, Biden is at 51.6%, which means even if Trump won every undecided vote and third party vote at this point, he still can't win and would top out around 48%. Yikes. I'd say it's over, but hey, remember 2016.

 For the swing states, I'm removing New Hampshire from my analysis  because it's up by 10.5%, outside of the margin of error. It will be assumed "safe" for the purposes of this analysis as the chance for Trump to win it is under 2%. As for the rest:

States

Winner/Margin

Z Score

% Biden Win

% Trump Win

Electoral Votes if Biden Wins

Electoral Votes if Trump Wins

Pennsylvania

Biden +7.0%

-1.750

96.0%

4.0%

246

312

Michigan

Biden +7.0%

-1.750

96.0%

4.0%

262

292

Nebraska CD2

Biden +6.5%

-1.625

94.8%

5.2%

263

276

Wisconsin

Biden +6.3%

-1.575

94.3%

5.7%

273

275

Nevada

Biden +5.2%

-1.300
90.3%
9.7%

279

265

Florida

Biden +3.5%

-0.875

81.0%

19.0%

308

259

Arizona

Biden +2.7%

-0.675

75.1%

24.9%

319

230

North Carolina

Biden +2.7%

-0.675

75.1%

24.9%

334

219

Iowa

Biden +1.2%

-0.300

62.0%

38.0%

340

204

Ohio

Biden +0.6%

-0.150

56.0%

44.0%

358

198

Maine CD2

Trump +0.2%

+0.050

48.0%

52.0%

359

180

Georgia

Trump +0.4%

+0.100

46.0%

54.0%

375

179

Texas

Trump +4.4%

+1.100

13.6%

86.4%

413

163

South Carolina

Trump +5.5%

+1.375

8.5%

91.5%

422

125

Missouri

Trump +7.5%

+1.875

3.0%

97.0%

432

116

Ouch. Biden has a 94.3% chance of winning and Trump only has a 5.7% chance. A Biden win is just short of statistical significance. Now, again, does that mean he will DEFINITELY win? Well, many ballots are being cast now. And I doubt Trump can regain the energy he lost. Even if he regresses to the mean and gains say, 3-4 points, he would only be at 259 electoral votes with Florida. You'd literally need a Trump +6-7 scenario for him to WIN, which is extremely unlikely. Like, 4-7% level unlikely. Trump is toast. The electoral map is a commanding 358-180 win for Biden.

I mean, for reference, the highest I ever gave Hillary by this methodology is like a 85% chance, with Trump having a 15% chance. And Hillary's numbers fluctuated a lot with her going into the 50s. A bad day for Biden is a good day for Hillary. 

Now, how does the senate look?

Well for starters I added Texas into the mix as it's now technically "in play". I don't expect it to go red as it's barely in play but it's worth looking at. This brings the GOP's "safe senators down to 45. 

States

Winner/ Margin

Z Score

% Dem Win

% GOP Win

Senate Seats if Democrat Wins

Senate Seats if Republican Wins

Arizona

Kelly +8.0%

-2.000

97.7%

2.3%

47

54

North Carolina

Cunningham +5.7%

-1.425

92.3%

7.7%

48

53

Michigan

Peters +4.9%

-1.225

89.0%

11.0%

49

52

Iowa

Greenfield +4.8%

-1.200

88.5%

11.5%

50

51

Maine

Gideon +3.7%

-0.925

82.3%

17.7%

51

50

South Carolina

Graham +0.5%

+0.125

45.0%

55.0%

52

49

Georgia

Perdue +2.8%

+0.700

24.2%

75.8%

53

48

Montana

Daines +3.3%

+0.825

20.5%

79.5%

54

47

Texas

Cornyn +7.6%

+1.900

2.8%

97.2%

55

46

Based on this, the democrats have an 82.3% chance of winning the senate by the trend model. The republicans have a 11.5% chance by comparison. This means there's a 6.2% chance it's a 50-50 split. This is up from last time for the democrats, but is still far less certain than what will happen with the presidency.

All in all, it looks like the GOP are screwed. As I said I ain't even doing the house, but I doubt that is particularly going well for them either given they need virtually all swing seats just to achieve a narrow majority. Yikes. While we still have 3 weeks and a lot can change, this looks like a blow out for the democratic party. Bye Trump, you're fired! Or you will be a month from now, probably. 

For the record I'm probably gonna be doing these more frequently between now and the election. We're getting close so I'll likely be doing them once a week now unless literally almost nothing changes, and I will especially be paying close attention in the last week or two leading up to the election. If things change, I want it to be documented here so we can analyze what happens.

No comments:

Post a Comment