tallgrant said:

---

Corporal Punishment said:

---

What do we know about uncounted vote by mail ballots? Are there a substantial number and are they being accounted for in this board's optimism?

---

You referencing that mailroom tweet that ended up being 7 unopened ballots?

---

No. To clarify, when is the deadline to count mail-in ballots and are there still many to count?

Corporal Punishment said:

---

What do we know about uncounted vote by mail ballots? Are there a substantial number and are they being accounted for in this board's optimism?

---

DeSantis said they're already been processing/counting VBM for a while now so I don't expect some magical crates to arrive without some heavy scrutiny. I also think, as I stated earlier, if R's lead is 300,000 or more you can't just cheat your way to that many ballots.

Prosperdick said:

---

Corporal Punishment said:

---

What do we know about uncounted vote by mail ballots? Are there a substantial number and are they being accounted for in this board's optimism?

---

DeSantis said they're already been processing/counting VBM for a while now so I don't expect some magical crates to arrive without some heavy scrutiny. I also think, as I stated earlier, if R's lead is 300,000 or more you can't just cheat your way to that many ballots.

---

I think it is group think rather than straight up willful negligence. They don't know how to properly model an electorate that might be a different ideology. They've been failing all over the world to predict how conservative voters behave.

This year of all years with multiple black swan type events how could anyone be confident in anything they were trying to model? Arrogance is also part of this.

Still the votes must actually be counted to prove this out but it doesn't look good in Florida for the Democrats. What does this mean for the rest of the country? We'll see.

Am I reading this right? It's now showing a 92K lead for Trump?

https://joeisdone.github.io/florida/

Prosperdick said:

---

Corporal Punishment said:

---

What do we know about uncounted vote by mail ballots? Are there a substantial number and are they being accounted for in this board's optimism?

---

DeSantis said they're already been processing/counting VBM for a while now so I don't expect some magical crates to arrive without some heavy scrutiny. I also think, as I stated earlier, if R's lead is 300,000 or more you can't just cheat your way to that many ballots.

---

It kind of seems like you are suggesting that competently run states won't have problems with vote counts at the end of the day today.

I think that is just party affiliation, but generally a strong correlation between their vote...

atmtws said:

---

Am I reading this right? It's now showing a 92K lead for Trump?

https://joeisdone.github.io/florida/

---

Yes you are reading it correctly and expect that number to continue to go up.
ETA - as BoerneAg11 stated, it's party affiliation but just ask yourself one question. How many Republicans, who aren't afraid of the Covid and standing in long lines, would pull the lever for Scientist Shutdown Biden? Then ask yourself how many Democrats, who are sick of lockdowns, would go out to vote today and crossover to Trump (and likely keep their down ballots Republican).

Ulrich said:

---

Florida has been a high profile swing state for the entire age of big data analytics and has nice, distinct voter groups. If the pollsters screw up Florida really badly, I think it is a pretty damning piece of evidence in favor of willful negligence.

---

Almost like the point of their polling is something besides accurately predicting the ultimate outcome?

No it means the Republicans have cast 92k more ballots than the Democrats. In 2016 Republicans cast ~57k more ballots than Democrats and won by 113k votes. Votes still must be counted but trend is not your friend if you are the Democrats.

txags92 said:

---

Ulrich said:

---

Florida has been a high profile swing state for the entire age of big data analytics and has nice, distinct voter groups. If the pollsters screw up Florida really badly, I think it is a pretty damning piece of evidence in favor of willful negligence.

---

Almost like the point of their polling is something besides accurately predicting the ultimate outcome?

---

It's disinformation. They are trying to demoralize republicans so they don't vote.

My prediction is the MOV will not be more than 3%.

Florida is pretty reliable in that regard. If it's more than 4% for Trump, I think you could probably feel insanely comfortable about the rust belt.

aginresearch said:

---

No it means the Republicans have cast 92k more ballots than the Democrats. In 2016 Republicans cast ~57k more ballots than Democrats and won by 113k votes. Votes still must be counted but trend is not your friend if you are the Democrats.

---

It's almost a 200k voter submission swing in favor of Reps in less than 5hrs today.

Prosperdick said:

---

---

That's overly strong. There is a pathway but it's a lot less likely.

Prosperdick said:

---

---

That's an interesting statement coming out So early in the day.

nortex97 said:

---

Prosperdick said:

---

Refresher on how accurate polling is in Florida...

---

Exactly. The amazing thing though is that it is actually pretty precise. Average Florida polling is generally always off by around 5-7 percent, in the same direction.

CNN is just off by an extra 10 or so. I don't remember any of the 'major' polling firms wrongly calling a big margin for the GOP, somehow.

---

I've said it before and I'll say it again, pollsters are a LOT like TOS recruiting service. They change their predictions at the last second to try to maintain their crystal ball accuracy and as you mentioned, the bias is always in one direction, burnt orange. It's why they always have top 5/10 classes and are stumped as to why they suck.

Yup. Florida very republican today.

Mostly Foggy Recollection said:

---

My prediction is the MOV will not be more than 3%.

Florida is pretty reliable in that regard. If it's more than 4% for Trump, I think you could probably feel insanely comfortable about the rust belt.

---

Yeah, people on here talking about a 7 point Trump victory in FL aren't being reasonable. 3 points would be massive.

7 points would require all of: incredible ED turnout of all R voters, crossover strongly favoring Trump, and NPAs breaking unexpectedly high for Trump. Strains credulity.

3 points would be huge. If it gets as high as even 3.5-4 that's a great sign for Trump elsewhere.

All this is predicated on the polling being wrong, but:

I think all the data exists to do really good polling, but it's hard to link it all together, hard to go the extra mile to uncover shy voters, hard to do really good subgroup sampling, and really hard to update a mental model.

If doing things the old, easy way (phone calls with random sampling) gives you the answer you expect, why go to all that effort? The samples aren't random, not really, so acting like they are is going to fail.

You're right that groupthink is a factor, I'm just taking it one step further to try to understand the mechanism by which groupthink without malice applied to theoretically unbiased data turns into a very wrong answer.

I dont think there's much "I want to see Biden +10 so by damn I will find a way to get Biden +10." But I do think that of you do a phone bank you get a bunch of Biden voters because people who pick up the phone and answer pollsters have personality and demographic traits that correlate with blue.

Gillum is so brave he once got gangbanged by a bunch of pol boys in a hotel room full of drugs. So brave.

DTP02 said:

---

Mostly Foggy Recollection said:

---

My prediction is the MOV will not be more than 3%.

Florida is pretty reliable in that regard. If it's more than 4% for Trump, I think you could probably feel insanely comfortable about the rust belt.

---

Yeah, people on here talking about a 7 point Trump victory in FL aren't being reasonable. 3 points would be massive.

7 points would require all of: incredible ED turnout of all R voters, crossover strongly favoring Trump, and NPAs breaking unexpectedly high for Trump. Strains credulity.

3 points would be huge. If it gets as high as even 3.5-4 that's a great sign for Trump elsewhere.

---

Toss out looking at Florida as a whole. Draw a line from Jax to Tampa. Duval, Orange, Pinellas and Hillsborough will give you a good snapshot of how rust belt snowbirds are voting.

If Trump outperforms there versus 2016, then I'd get really nervous as a D.

Yeah 3 is high end and would be a massive victory. Just don't know how these NPAs are going to break overall. Today I'm confident they are breaking Trump. It's the early vote I'm concerned about.

FWIW, Pinellas county just flipped to overall Red for the first time. Many people say it is a bellwether county.

Ulrich said:

---

All this is predicated on the polling being wrong, but:

I think all the data exists to do really good polling, but it's hard to link it all together, hard to go the extra mile to uncover shy voters, hard to do really good subgroup sampling, and really hard to update a mental model.

If doing things the old, easy way (phone calls with random sampling) gives you the answer you expect, why go to all that effort? The samples aren't random, not really, so acting like they are is going to fail.

You're right that groupthink is a factor, I'm just taking it one step further to try to understand the mechanism by which groupthink without malice applied to theoretically unbiased data turns into a very wrong answer.

I dont think there's much "I want to see Biden +10 so by damn I will find a way to get Biden +10." But I do think that of you do a phone bank you get a bunch of Biden voters because people who pick up the phone and answer pollsters have personality and demographic traits that correlate with blue.

---

I recall a poster a few weeks ago who stated he was visiting his parents (long-time registered Republicans) and he happened to catch a pollster calling them and decided to play a prank. He proceeded to tell the pollster that his parents had both switched to vote for Biden.

For the next week they were swamped by polling services and they were super pissed at their son because they had to keep telling them they didn't switch. I think he made a joke that he was now out of their will.

It's anecdotal for sure but I think if they catch a registered Republican flipping they put them on speed dial, especially all the Never Trumper's. I also found out recently that many polling services use the same respondents over and over, which to me seems counter-intuitive UNLESS you are driving a certain result.

TL/DR - Polls are biased junk and should never be leveraged again if Trump wins in a landslide.

How do we know they aren't democrats?

Prosperdick said:

---

---

Ron Swanson is my spirit animal

It probably also doesn't help that the Lord of All Pollreaders and Gatekeeper for the Holy Realm of Election Punditry Nate Silver may declare your poll unreliable and throw it out of the averages if it shows a response he doesn't believe. Puts an economic incentive on the polling firms to stay dumb.

Legend said:

---

FWIW, Pinellas county just flipped to overall Red for the first time. Many people say it is a bellwether county.

---

Saw that the Dem lead was narrowing to a few thousand votes earlier this morning. So more good news! Thanks for posting.

ChrisTAMU said:

---

How do we know they aren't democrats?

---

Good point, I'm sure they're all Democrats, not afraid of the floating Ebola virus their Dem overlords warned them about and to VBM early (and often). Those rallies of 50,000 meant nothing and the +200,000 registered Republicans in the state since 2016 were all doing it as an elaborate goof on the Orange Man.

Sorry for the snarky reply.