Mega-Thread: 2020 United States General Election

[arandomuser]

https://twitter.com/NateSilver538/st...polls-as-crazy for context on trafalgar their methodology is entirely flawed

[Elenos]

Trafalgar's only claim to fame is 'predicting' the 2016 race.

Even that alone would give them a lot more credit then many other races, and they didn't call the 2016 election: They called out Michigan and Pennsylvania. They called a lot of races correctly in 2018 for both Democrats and Republicans, crucially calling out Florida for both the governor and senatorial race:

https://www.nytimes.com/article/trum...gar-group.html

Regarding people who quoted Clinton winning the Midwest by double digits in 2016 is concerned, they have the same partisan bias issues as Traflagar. So it's a false equivalency.

Yeah well when it's a pollster associated with the Washington Post and New York Times and gets treated completely seriously to this day despite that while Trafalgar is mocked...well then.

I don't really care really what the biases of someone is if they are accurate with their work. This is the rather infuriating aspect about US elections: The damned pollsters are biased, not just in their personal views but also their work. If I look at X European country ( and for Americans don't think we don't have deep political divides here in many countries ) in general the polling tends to be fairly accurate. In the US? It's like a trip to Wonderland.

So far Trafalgar has been mostly accurate in the results they've given. If they are wrong they are wrong, if they are right they are right, and when someone who has a proven track record for Michigan tells me Trump's going to win there I'm listening to that guy until he is proven otherwise.

[DingDongKing]

That's not true tbh. They called a lot of races correctly in 2018 for both Democrats and Republicans, crucially calling out Florida for both the governor and sinoatrial race:

https://www.nytimes.com/article/trum...gar-group.html

I didn't say they haven't called anything else correctly. Per the 48 polls counted by 538, they've gotten 75% right. Which is a pretty mediocre return.

[szechuan]

That's not true tbh. They called a lot of races correctly in 2018 for both Democrats and Republicans, crucially calling out Florida for both the governor and sinoatrial race:

https://www.nytimes.com/article/trum...gar-group.html

He's going to be wrong, he also called out 2018 races that were safe for Republicans. 2018 was also a win for the Democrats, it's going to be a win for Biden.

[DingDongKing]

-snip-

Can I be honest? And don't be offended. I really don't care what you think is true or isn't true.

[Elenos]

I didn't say they haven't called anything else correctly. Per the 48 polls counted by 538, they've gotten 75% right. Which is a pretty mediocre return.

New York Times/Sienna - 69%
Washington Post/ABC - 72%
CNN/ORC - 72%
Fox News - 72%
NBC/Wall Street Journal - 72%

On and on it goes on 538 with many of the "respected" pollsters.

Can I be honest? And don't be offended. I really don't care what you think is true or isn't true.

Feel free to care about what you care or not, as long as people understand what could very well happen today. That's all there is to what I'm saying here: Trump can win.

[Bepples]

So far Trafalgar has been mostly accurate in the results they've given. If they are wrong they are wrong, if they are right they are right, and when someone who has a proven track record for Michigan tells me Trump's going to win there I'm listening to that guy until he is proven otherwise.

The problem with this way of thinking about polling is it treats it as a binary X will win v.s. Y will win. The margins are important too, and Trafalgar consistently overestimates Republican margins even when Republicans do win. When you have an election with Republican upsets like 2016 of course a pollster that consistently overestimates Republicans is going to look good. But it's not like they have a perfect track record. They got Arizona wrong in 2018, for example, and their margins were also way off in races where the Republican won such as Texas in 2018.

[DingDongKing]

-snip-

Can you find any post where I've said that he can't win? The vast majority here are very very aware that he can win.

As far as Trafalgar Group is concerned, their results are extremely wonky. You can apply selective evidence to anything and claim a case.

RE: Polls in general, I have the exact same issue with 538. Their grading is really odd. I go by one basic principle - bias exists. And it shouldn't.

Whether a poll has bias for or against a party, it's not a good thing. Places like Atlas Intel, Insider Advantage, Susquehanna & Trafalgar are ALL setup by ex GOP hacks. They'll call a set of races right, sure they will. It's hard to get polling entirely wrong, or you'd be out of business.

But their bias is evident. Hell, Rasmussen's Twitter is basically now just a pro-Trump feed. It's hilarious how brazen they are.

[postman1782]

Just in general.

Here's a graph on RCP to give you an idea. You can also look at it state by state. People wanna smack talk Trafalgar but they have gotten important races right in the past and they are certainly more believable then those laughable jokesters who thought Clinton was going to win the Midwest by double digits in 2016 in some states.

https://www.realclearpolitics.com/el...ground-states/

Trafalgar is rated as a C- pollster for a reason. They claimed to have Kemp winning by 12% in Georgia in 2018, and he won by less than 1% to Stacey Abrams. Sure they may call a race correctly, but be wayyyy off percentage wise.

Literally the only 3 pollsters are heavy right leaning pollsters, and have been wrong most of the time. Like Rasmussen, Sasquatch and Trafalgar. That is the ONLY reason they are bringing the average down.

They also don't give their methodology for their polls, which means they are highly questionable.

[szechuan]

Trumpers can't spell.

[kaelleria]

Trumpers can't spell.

That started trending around 7 am Moscow time...

[Inuyaki]

Maybe if this wasn't 2020. But...you know...it's 2020. And we can't have nice things this year.

We got a really good final season 3 for Dark. Does that count?

[Felya]

Boo!!! 538 polling map is no longer updating... Trump never did hit single digits... ending with 10/89...

[Belize]

Good luck out there everyone.

P.S. that fence around the WH looks pretty climbable to me.

[lockybalboa]

I'm so nervous guys, how'd you deal with the election jitters?

[Machismo]

Just in general.

Here's a graph on RCP to give you an idea. You can also look at it state by state. People wanna smack talk Trafalgar but they have gotten important races right in the past and they are certainly more believable then those laughable jokesters who thought Clinton was going to win the Midwest by double digits in 2016 in some states.

https://www.realclearpolitics.com/el...ground-states/

I know you are really excited to see if a white supremacists is going to be re-elected.

[david0925]

I'm so nervous guys, how'd you deal with the election jitters?

block all sites that updates the count. We can save the drama for tomorrow

[DingDongKing]

block all sites that updates the count. We can save the drama for tomorrow

He's trolling. He's a Trump supporter.

[Felya]

Trump again stranded thousands of his supporters in dark and cold Georgia night. That’s a 3rd time he has done this within a couple of weeks.

No mater what we say about them, no one hates Trump supporters, more than Trump.

[szechuan]

Trump again stranded thousands of his supporters in dark and cold Georgia night. That’s a 3rd time he has done this within a couple of weeks.

No mater what we say about them, no one hates Trump supporters, more than Trump.

More proof that right wing tribalism is cancer, those Trump Rally attenders are literally poorly educated Low IQ voters who supports Racists and Anti-Science. Some of them even think they're voting against Obama.