"It's still 12 months out", "It's still 11 months out", "It's still 10 months out"...

[Wrong about 2024 Ghost]

When does this rhetoric end?

Good question - when do people stop pretending polls well in advance of election have any solid predictive power? President Clinton (2008) and President Clinton (2016) would sure like to know, and so would the McCain campaign's very close race, along with the Biden campaign's easy 2020 win.

[BG-NY (permanently retired)]

When does this rhetoric end?

Good question - when do people stop pretending polls well in advance of election have any solid predictive power? President Clinton (2008) and President Clinton (2016) would sure like to know, and so would the McCain campaign's very close race, along with the Biden campaign's easy 2020 win.

RCP only has February going back to 2008, but taking a look:

Feb 2008: Obama led McCain by 3-5 (final: Obama +7.2)
Feb 2012: Obama led Romney by 3-4 (final: Obama +3.9)
Feb 2016: Clinton led Trump 2-3 (final: Clinton +1.9)
Feb 2020: Biden led Trump by 4-5 (final: Biden +4.5)

This seems pretty accurate. All four leaders in February ended up winning the "popular" vote in November of that year. Margins the last three cycles have been extremely similar to the February RCP averages.

[Annatar]

When does this rhetoric end?

Good question - when do people stop pretending polls well in advance of election have any solid predictive power? President Clinton (2008) and President Clinton (2016) would sure like to know, and so would the McCain campaign's very close race, along with the Biden campaign's easy 2020 win.

RCP only has February going back to 2008, but taking a look:

Feb 2008: Obama led McCain by 3-5 (final: Obama +7.2)
Feb 2012: Obama led Romney by 3-4 (final: Obama +3.9)
Feb 2016: Clinton led Trump 2-3 (final: Clinton +1.9)
Feb 2020: Biden led Trump by 4-5 (final: Biden +4.5)

This seems pretty accurate. All four leaders in February ended up winning the "popular" vote in November of that year. Margins the last three cycles have been extremely similar to the February RCP averages.

Accuracy of polling has changed a lot post 2000 due to increased polarization. When people say polls 6 months out or 8 months out don't mean much, they would have been correct in the pre 2000 world. There were huge polling swings in the 1970's and 1980's and into the 1990's when there were many more swing voters and the electorate moved around a lot. The predictive power of polls taken early has increased substantially since 2000 due to there being less swing voters.


This is why RCP has been accurate in the past few cycles in its Feb polls compared to the final result, most of the electorate has made up its mind and there are only a few swing voters left. My own view is that Biden's best hope is to hope Trump gets convicted on a serious case, not the business one but the DC case.

[BlueSwan]

Accuracy of polling has changed a lot post 2000 due to increased polarization. When people say polls 6 months out or 8 months out don't mean much, they would have been correct in the pre 2000 world. There were huge polling swings in the 1970's and 1980's and into the 1990's when there were many more swing voters and the electorate moved around a lot. The predictive power of polls taken early has increased substantially since 2000 due to there being less swing voters.

Exactly, and this is even more the case in 2024 than in the previous elections. The electorate has never been as polarized as this and the candidates have never been as well known as these two. People are seriously deluding themselves. It is not over for Biden and stuff can definitely happen when you are dealing with two candidates who could basically die any minute, but it is very much an uphill battle and anybody denying that Trump is the favourite at this point are engaging in wishful thinking.

[DrScholl]

Accuracy of polling has changed a lot post 2000 due to increased polarization. When people say polls 6 months out or 8 months out don't mean much, they would have been correct in the pre 2000 world. There were huge polling swings in the 1970's and 1980's and into the 1990's when there were many more swing voters and the electorate moved around a lot. The predictive power of polls taken early has increased substantially since 2000 due to there being less swing voters.

Exactly, and this is even more the case in 2024 than in the previous elections. The electorate has never been as polarized as this and the candidates have never been as well known as these two. People are seriously deluding themselves. It is not over for Biden and stuff can definitely happen when you are dealing with two candidates who could basically die any minute, but it is very much an uphill battle and anybody denying that Trump is the favourite at this point are engaging in wishful thinking.

But that's the problem with this sort of logic. "At this point" doesn't matter because no one is voting right now. Even if the election was tomorrow the picture could look very different if left-leaning voters who are undecided got off the bench if they felt Trump was at risk of winning. The same would be true if Biden was leading right now.

[Birdish]

When does this rhetoric end?

Good question - when do people stop pretending polls well in advance of election have any solid predictive power? President Clinton (2008) and President Clinton (2016) would sure like to know, and so would the McCain campaign's very close race, along with the Biden campaign's easy 2020 win.

RCP only has February going back to 2008, but taking a look:

Feb 2008: Obama led McCain by 3-5 (final: Obama +7.2)
Feb 2012: Obama led Romney by 3-4 (final: Obama +3.9)
Feb 2016: Clinton led Trump 2-3 (final: Clinton +1.9)
Feb 2020: Biden led Trump by 4-5 (final: Biden +4.5)

This seems pretty accurate. All four leaders in February ended up winning the "popular" vote in November of that year. Margins the last three cycles have been extremely similar to the February RCP averages.

Accuracy of polling has changed a lot post 2000 due to increased polarization. When people say polls 6 months out or 8 months out don't mean much, they would have been correct in the pre 2000 world. There were huge polling swings in the 1970's and 1980's and into the 1990's when there were many more swing voters and the electorate moved around a lot. The predictive power of polls taken early has increased substantially since 2000 due to there being less swing voters.


This is why RCP has been accurate in the past few cycles in its Feb polls compared to the final result, most of the electorate has made up its mind and there are only a few swing voters left. My own view is that Biden's best hope is to hope Trump gets convicted on a serious case, not the business one but the DC case.

It seems coincidental.

There were multiple events during all those campaigns that shifted the polling averages. Had polling been static, perhaps i'd agree.

[emailking]

When does this rhetoric end?

1 month out.

Not snark. That's when I think we basically know what happens.

[AltWorlder]

This year I'm waiting for the October Surprise, so 1 month out works

I wonder if a lot of other voters are doing the same. When given two unpalatable choices, while all sorts of chaotic things happen in the background, people are adopting a wait and see attitude.

[Mr.Bakari-Sellers]

Rs think Trump is gonna win and it's not even time to vote yet, get over it

[All Along The Watchtower]

We might as well not have an election then and just install Trump as dictator now FFS.

Seriously, it's clear that "Talk Elections" (RIP Atlas Forum) has utterly succumbed to horse race coverage and obsession with polls and media narratives like every other place at the expense of actual discussion of elections and the (real, empirical) data that they provide. Furthermore, it's clear that a lot people either actively want or have otherwise resigned themselves to a future in which elections don't matter in the US and that there will be electoral-type events at best.


It reflects very poorly on the forum culture here.

[Mr.Bakari-Sellers]

We might as well not have an election then and just install Trump as dictator now FFS.

Seriously, it's clear that "Talk Elections" (RIP Atlas Forum) has utterly succumbed to horse race coverage and obsession with polls and media narratives like every other place at the expense of actual discussion of elections and the (real, empirical) data that they provide. Furthermore, it's clear that a lot people either actively want or have otherwise resigned themselves to a future in which elections don't matter in the US and that there will be electoral-type events at best.


It reflects very poorly on the forum culture here.

It's a 303 map females don't vote always in these polls that's why they are R leaning that's what I previously said but turnout in the Eday, it's computerized polls with male dominant voters

Just like the Red state Redban how many red state red bans are inI, WI and PA and 303 compared to the South very few, just like he is obsessed with Biden 37/60 Approvals even Xing said that. Buden@44 in Rassy that is not 36

[Tekken_Guy]

We might as well not have an election then and just install Trump as dictator now FFS.

Seriously, it's clear that "Talk Elections" (RIP Atlas Forum) has utterly succumbed to horse race coverage and obsession with polls and media narratives like every other place at the expense of actual discussion of elections and the (real, empirical) data that they provide. Furthermore, it's clear that a lot people either actively want or have otherwise resigned themselves to a future in which elections don't matter in the US and that there will be electoral-type events at best.


It reflects very poorly on the forum culture here.

Yeah a lot of people are writing off Biden early. And a lot of people are also basically saying that if Trump tried to end free and fair elections he’s guaranteed to succeed, and there’s no chance of it failing and/or backfiring on him.

[cg41386]

When does this rhetoric end?

Good question. People are really deluding themselves about this election. This is NOT a normal election. This is basically an election between two incumbents. People are not voting for "generic D" or "generic R". People are not going to learn anything new about these two candidates. The political climate is massively polarized. Trump is an 80% favourite at this point.

lol wut

[Wrong about 2024 Ghost]

When does this rhetoric end?

Good question - when do people stop pretending polls well in advance of election have any solid predictive power? President Clinton (2008) and President Clinton (2016) would sure like to know, and so would the McCain campaign's very close race, along with the Biden campaign's easy 2020 win.

RCP only has February going back to 2008, but taking a look:

Feb 2008: Obama led McCain by 3-5 (final: Obama +7.2)
Feb 2012: Obama led Romney by 3-4 (final: Obama +3.9)
Feb 2016: Clinton led Trump 2-3 (final: Clinton +1.9)
Feb 2020: Biden led Trump by 4-5 (final: Biden +4.5)

This seems pretty accurate. All four leaders in February ended up winning the "popular" vote in November of that year. Margins the last three cycles have been extremely similar to the February RCP averages.

Accuracy of polling has changed a lot post 2000 due to increased polarization. When people say polls 6 months out or 8 months out don't mean much, they would have been correct in the pre 2000 world. There were huge polling swings in the 1970's and 1980's and into the 1990's when there were many more swing voters and the electorate moved around a lot. The predictive power of polls taken early has increased substantially since 2000 due to there being less swing voters.


This is why RCP has been accurate in the past few cycles in its Feb polls compared to the final result, most of the electorate has made up its mind and there are only a few swing voters left. My own view is that Biden's best hope is to hope Trump gets convicted on a serious case, not the business one but the DC case.

It seems coincidental.

There were multiple events during all those campaigns that shifted the polling averages. Had polling been static, perhaps i'd agree.

On Feb 4th, when the thread was started:

In 2008:
Clinton leads Obama for the nomination by a little over 3 points.
McCain and Obama are virtually tied.

In 2012: Obama up by 2 over Romney

In 2016: Clinton leads Trump by 2.5 points.

2020: Biden leads Trump by over 5 points.

Comparing these slightly different numbers with final results just makes my point: the polls are volatile. And the final result has far more to do with the precise details of the turnout than national polls

In 2008, Obama and McCain were all over the place between February and November

In 2012, Romney was tied with Obama early in September, and his October numbers were higher than February.

Trump spent most of 2016 doing worse than his February or final numbers, but was ahead briefly in July.

In 2020 Biden's lead between Feb and Nov was regularly over 8, and peaked over 9.

I don't think it's possible to usefully link the highly variable national polling to turnout in decisive counties. And the latter is what actually decides elections.