Update 10/27
[Thread]
Today was perfect for Rs. VBM was WAY down from what I expected. I expected more than 2x from yesterday, since that's what happened last Tue. But it only increased 35%. And despite the increase, Rs lost VBM by LESS than yesterday (6k vs 7k). Amazing.
1/ https://twitter.com/JeffersonJaxson/status/1320527500410298368
[Thread]
Today was perfect for Rs. VBM was WAY down from what I expected. I expected more than 2x from yesterday, since that's what happened last Tue. But it only increased 35%. And despite the increase, Rs lost VBM by LESS than yesterday (6k vs 7k). Amazing.
1/ https://twitter.com/JeffersonJaxson/status/1320527500410298368
More great news, Rs continued to dominate IPEV, netting +60k for the day: the best day yet!
Now part of that may be in prep for the hurricane that will clip the panhandle tomorrow. We'll see if tomorrow's down. Pensacola is going to be stormy, but it looks like Panama...
2/
Now part of that may be in prep for the hurricane that will clip the panhandle tomorrow. We'll see if tomorrow's down. Pensacola is going to be stormy, but it looks like Panama...
2/
... City won't be, so the storm doesn't hit too much of FL, but it does hit red counties. I'm adjusting my model for a dip tomorrow.
Onto my projection. Assuming VBM trends down for the rest of the week (and I'll give a slight uptick to Ds in the final few days), and...
3/
Onto my projection. Assuming VBM trends down for the rest of the week (and I'll give a slight uptick to Ds in the final few days), and...
3/
... assuming Rs pump out a few more good IPEV days (and I'll give Ds an IPEV gain for the final 2 days of EV since many red counties aren't open), I'm projecting 8.7M early votes, with Ds holding only a 136k vote lead over Rs (it was 88k in 2016).
I'm projecting a total...
4/
I'm projecting a total...
4/
... of 10.6M voters, so about 1.9M to vote on ED. With an ED split of 50/30/20, we would see Rs submit 383k more ballots than Ds on ED. This nets out to R+247 when you account for the early vote deficit. That's massive. Rs were only +65k in 2016 (and Trump won by 113k).
5/
5/
Note that I'm not projecting who is receiving the votes. Only what party the votes are coming from. But it would take a significant shift in crossover votes for Biden or a significant shift in Other voters toward Biden for him to beat Trump. Could it happen? Maybe…
6/
6/
... if a lot of Rs and Os are mad at Trump for COVID, it's possible.
But CBS/YouGov had an encouraging poll (yes, these polls are bad, but it's a good sign when they don't significantly favor Biden).
In that poll...
7/
But CBS/YouGov had an encouraging poll (yes, these polls are bad, but it's a good sign when they don't significantly favor Biden).
In that poll...
7/
... voters that had already voted before 10/23 had the following Trump/Biden splits:
R - 87/12
O - 40/58
When you consider that the majority of these votes were VBM, it's a good sign. We know Biden is going to do better with VBM… and most Rs and Os who vote Biden...
8/
R - 87/12
O - 40/58
When you consider that the majority of these votes were VBM, it's a good sign. We know Biden is going to do better with VBM… and most Rs and Os who vote Biden...
8/
... are likely to do it by mail.
Without diving into the details, I think Trump will easily fix these splits to winning levels b/c of all the IPEVs coming in btwn 10/23 - 11/2, and of course ED voters. And remember, those polled splits are from…
9/
Without diving into the details, I think Trump will easily fix these splits to winning levels b/c of all the IPEVs coming in btwn 10/23 - 11/2, and of course ED voters. And remember, those polled splits are from…
9/
... Fake News. Reality is likely better for Trump.
In summary, amazing day for Trump. As long as a windfall of VBMs don’t come through, and as long as Rs and Os don't significantly abandon Trump, he's going to win FL. My model has him winning by 3%....
10/
In summary, amazing day for Trump. As long as a windfall of VBMs don’t come through, and as long as Rs and Os don't significantly abandon Trump, he's going to win FL. My model has him winning by 3%....
10/
... I'll continue to update as more numbers come in. No promises that my VBM and IPEV projections hold, but I feel good about them!
Thanks for following and making it to the end. I'm a statistician who's interested in politics on the side. If I have time I'll look at...
11/
Thanks for following and making it to the end. I'm a statistician who's interested in politics on the side. If I have time I'll look at...
11/
... other states, but FL is definitely the most interesting due to the high number of early voters as well as voters being registered by party. Makes the analysis much more accurate.
Great day for FL and great day for Trump's re-elections prospects as a whole!
12/12
Great day for FL and great day for Trump's re-elections prospects as a whole!
12/12