The 2020 Presidential Election - Convention & General Election

[RedDenver]

10 minutes ago, knapplc said:

We absolutely need to be vigilant and VOTE, but Mr. Moore is cherry-picking polls.

 

 

https://projects.fivethirtyeight.com/polls/president-general/minnesota/

 

 

 

 

https://projects.fivethirtyeight.com/polls/president-general/michigan/

 

 

 

 

It's less on "the party" and "the candidate" than Trump being an incumbent and having JUST gotten done with his national convention. 

Moore is referencing the RCP average of polls in the first tweet, so I'm not sure how that's cherry picking. You could maybe make that argument for the second tweet, but it's the most recent CNN poll for Minn.

 

We'll have to wait and see if this is a just a short-term bump for Trump due to the RNC or a longer term trend. Biden continues to trail Trump in voter enthusiasm, which IMO will end up hurting him, but Biden has been trailing in enthusiasm the entire time so maybe it doesn't end up mattering.

[knapplc]

3 minutes ago, RedDenver said:

Moore is referencing the RCP average of polls in the first tweet, so I'm not sure how that's cherry picking. You could maybe make that argument for the second tweet, but it's the most recent CNN poll for Minn.

 

 

It's cherry-picking because he's referencing one poll. The averages don't support his outrage porn over "the party" and "the candidate."

 

This time in 2016 Hillary had been actively campaigning. Biden has barely left his home since he became the de facto nominee.

 

Trump was just another candidate in 2016. Now he's the incumbent. 

 

There are so few parallels we can draw from that election to this one that any kind of conclusion like this is going to be colored by the biases of people like Moore. 

[Danny Bateman]

Moore's got a point, but it's not particularly well supported by the evidence he supplied. I liken him to a weather vane rusted in one direction - occasionally it's going to right, but it's not through any particular operational efficiency or savvy. It's just the way he always points.

 

I've said before but 2016 polls vs. 2020 polls are apples to oranges given the different timeframes and events around each election. Polling professionals will also readily tell you 2016 polls systemically undersampled non-college white voters (AKA white working class) which led to most battleground polling inaccurately inflating Cliinton's lead.

 

From what I've gathered, the same polling professionals think sampling voters across multiple battleground states (as in the second tweet) is a very bad idea rendering such figures borderline useless.

 

The race is very different in Maine or Michigan than it is in Texas or Georgia. Lumping voters from all of them together muddies things up. That's why it's better to just do each state individually.

[RedDenver]

9 minutes ago, knapplc said:

 

It's cherry-picking because he's referencing one poll. The averages don't support his outrage porn over "the party" and "the candidate."

 

This time in 2016 Hillary had been actively campaigning. Biden has barely left his home since he became the de facto nominee.

 

Trump was just another candidate in 2016. Now he's the incumbent. 

 

There are so few parallels we can draw from that election to this one that any kind of conclusion like this is going to be colored by the biases of people like Moore. 

Or - and hear me out - we might want to listen to the person who was one of the only voices in 2016 that predicted Trump would win.

 

And yes, there are reasons to think 2016 and 2020 are not identical, but that doesn't mean you just pretend there are no comparisons or lessons to be learned. You might consider that you also have biases that color how you view the comparison between 2016 and 2020.

 

And your cherry-picking argument doesn't hold a lot of water since Moore referenced RCP polling average in 6 states in addition to the one CNN poll from Minnesota.

[RedDenver]

2 minutes ago, Danny Bateman said:

Moore's got a point, but it's not particularly well supported by the evidence he supplied. I liken him to a weather vane rusted in one direction - occasionally it's going to right, but it's not through any particular operational efficiency or savvy. It's just the way he always points.

 

I've said before but 2016 polls vs. 2020 polls are apples to oranges given the different timeframes and events around each election. Polling professionals will also readily tell you 2016 polls systemically undersampled non-college white voters (AKA white working class) which led to most battleground polling inaccurately inflating Cliinton's lead.

 

From what I've gathered, the same polling professionals think sampling voters across multiple battleground states (as in the second tweet) is a very bad idea rendering such figures borderline useless.

 

The race is very different in Maine or Michigan than it is in Texas or Georgia. Lumping voters from all of them together muddies things up. That's why it's better to just do each state individually.

Good points, but the same polling professionals might not have correctly diagnosed or correctly fixed the issues from 2016 polling. We won't know if they are correct until after the election. I find it weird for them to claim that comparisons shouldn't be drawn because then why even have polling at all? It's meaningless if there's no historical context from which to evaluate the data.

 

Moore is referencing polling from individual battleground states, so I'm not sure how your "lumping" argument applies.

[knapplc]

5 minutes ago, RedDenver said:

Or - and hear me out - we might want to listen to the person who was one of the only voices in 2016 that predicted Trump would win.

 

And yes, there are reasons to think 2016 and 2020 are not identical, but that doesn't mean you just pretend there are no comparisons or lessons to be learned. You might consider that you also have biases that color how you view the comparison between 2016 and 2020.

 

And your cherry-picking argument doesn't hold a lot of water since Moore referenced RCP polling average in 6 states in addition to the one CNN poll from Minnesota.

 

Per RealClearPolitics, Biden is currently +5.3 in Minnesota. 

 

Only one poll, Trafalgar, shows Biden & Trump tied. All the rest - including Fox, which has Biden +13, show Biden with a lead, and the polling average does not show a tie. Trafalgar also shows Trump leading in Michigan, but the RCP average has Biden +2.6.

 

Micheal Moore was not the only voice, or even one of the only voices, saying Trump could win in 2016. Nate Silver was saying his models all showed Trump with a significant chance to win, especially after the Comey letter. 

[Danny Bateman]

5 minutes ago, knapplc said:

 

Per RealClearPolitics, Biden is currently +5.3 in Minnesota. 

 

Only one poll, Trafalgar, shows Biden & Trump tied. All the rest - including Fox, which has Biden +13, show Biden with a lead, and the polling average does not show a tie. Trafalgar also shows Trump leading in Michigan, but the RCP average has Biden +2.6.

 

Micheal Moore was not the only voice, or even one of the only voices, saying Trump could win in 2016. Nate Silver was saying his models all showed Trump with a significant chance to win, especially after the Comey letter. 

 

I'd just point out Trafalgar is a trash poll. They weight for using "social desirability" which means they assume people are lying about supporting Biden because they're embarrassed about supporting Trump - the shy Trump voter thing - and adjust numbers accordingly. The reason it's ridiculous is Trafalgar collects responses by people hitting keys on the phone or speaking to a recording machine, so there's no human interaction and no one to be embarrassed to!

 

Pretty niche point but with methodology like that, if there's one poll to actually just ignore, it's them.

[knapplc]

5 minutes ago, Danny Bateman said:

 

I'd just point out Trafalgar is a trash poll. They weight for using "social desirability" which means they assume people are lying about supporting Biden because they're embarrassed about supporting Trump - the shy Trump voter thing - and adjust numbers accordingly. The reason it's ridiculous is Trafalgar collects responses by people hitting keys on the phone or speaking to a recording machine, so there's no human interaction and no one to be embarrassed to!

 

Pretty niche point but with methodology like that, if there's one poll to actually just ignore, it's them.

 

 

Exactly. Which is weird, because that seems to be the poll Michael Moore is referencing, if he's referencing any specific poll. No others show what he's saying is going on in MN or MI.

[BigRedBuster]

Michael Moore selectively using information for an agenda. 
 

color me shocked. 

[BigRedBuster]

I guess maybe Trump should be scared of mail in voting. 

[RedDenver]

26 minutes ago, Danny Bateman said:

I'd just point out Trafalgar is a trash poll. They weight for using "social desirability" which means they assume people are lying about supporting Biden because they're embarrassed about supporting Trump - the shy Trump voter thing - and adjust numbers accordingly. The reason it's ridiculous is Trafalgar collects responses by people hitting keys on the phone or speaking to a recording machine, so there's no human interaction and no one to be embarrassed to!

 

Pretty niche point but with methodology like that, if there's one poll to actually just ignore, it's them.

 

20 minutes ago, knapplc said:

Exactly. Which is weird, because that seems to be the poll Michael Moore is referencing, if he's referencing any specific poll. No others show what he's saying is going on in MN or MI.

 

Come on, knapp, Moore has links in the tweets I posted to the polls he's talking about. Here's the image from Moore's tweet:

 

That could very well be a trash poll, but it's being shown on CNN as their poll conducted by ssrs. (Note the 5.4% margin of error supports the conclusion it's not a good poll.)

 

And here's the image from his tweet for the RCP averages (referencing Walker Bragman's tweet):

 

I get disagreeing with Moore's conclusions, but it continues to be weird to claim Moore is cherry-picking when his tweets are showing multiple polls in them. It's not like he's scouring the corners of the internet - these are RCP polling averages and images from CNN.

[Danny Bateman]

12 minutes ago, RedDenver said:

Good points, but the same polling professionals might not have correctly diagnosed or correctly fixed the issues from 2016 polling. We won't know if they are correct until after the election. I find it weird for them to claim that comparisons shouldn't be drawn because then why even have polling at all? It's meaningless if there's no historical context from which to evaluate the data.

 

Moore is referencing polling from individual battleground states, so I'm not sure how your "lumping" argument applies.

 

This is a bit in the weeds so bear with me...

 

My first point was that not weighting by education undersampled the white working class (non-college) folks in 2016 and thus underrepresented Trump's strength, esp in battleground states where that group is higher relative to other states. More on this here from Vox:

 

Quote

There’s nothing wrong with weighting your sample based on race, age, and sex to match the demographics of the state. That’s standard practice in the industry. The problem is what the poll didn’t weight on — educational attainment. Many state-level polls omitted this factor in 2016, leading them to underestimate Trump’s strength in key swing states. The most responsible pollsters responded to 2016 by making sure to improve their weighting. But many pollsters — especially those doing state-level polling — continue not to weight by education.

 

This failure to weight not only leads to errors (which could be compensated for by averaging), it leads to systematic bias against Trump and the GOP, meaning everyone who publishes or disseminates unweighted polls ends up contributing to misinformation about the real state of American politics.

 

The bolded is what I've found doing a quick skim of recent battleground polling from this year. So if methodology has largely been tweaked to account for more WWC people in battleground states, polling this go around should be more accurate and thus not apples to apples with 2016 polling.

 

Moore is referencing one CNN/SRSS poll that had a subset of 636 RVs in battleground states and 1108 respondents overall. The battleground subset shows 49% - 48%. Close, right?

 

The problem with the methodology is they pulled those 636 responses from 15 states and lumped them into one "battleground" bucket. So the 1% difference they found does not reflect current state averages like Biden +10 in Maine, Biden +7 in Michigan, Trump +1.5 in Texas or Trump +<1 in Georgia because it's pulling responses from places where the state is wildly different and lumping them together.

 

See Nate Silver here:

 

 

[Danny Bateman]

40 minutes ago, knapplc said:

Exactly. Which is weird, because that seems to be the poll Michael Moore is referencing, if he's referencing any specific poll. No others show what he's saying is going on in MN or MI.

 

And I'm not even saying the polls might not STILL underrate Trump's strength.

 

But I follow a lot of polling geeks and Trafalgar is a trash poll with horrible methodology. We should not draw conclusions from a poll they tweak based on dubious theories lacking strong evidence.

[Danny Bateman]

17 minutes ago, BigRedBuster said:

I guess maybe Trump should be scared of mail in voting. 

 

I also saw this today to file away in "the sky isn't following for Biden" file.

 

Here's hoping these voters and more like them reside in states where their vote matters.

 

 

[TGHusker]

21 minutes ago, BigRedBuster said:

I guess maybe Trump should be scared of mail in voting. 

We'll see if he has a coordinated effort to block the military vote