Trump Approval Rating Reaches Highest Ever; Equal to Obama's At Same Point in Presidency (Gallup)

[salamander]

So this trend bodes well for the republicans during the incoming midterm elections then?

 

[DC23]

Anybody anti-Trump is anti-America.

 

[Rygu]

Should still be higher.

 

[Strychnine]

You know what would be wild?


IF President Trump gets primaried in 2020....

 

[Farmer Br0wn]

The average of polls right before election day was within 2%.

Apologies for taking longer than normal to get back to you.

That proves less than nothing.

That's just a confession by the polling firms that they really had no clue what was going to happen to begin with.

What a coincidence that right before an election, all of the polls suddenly find themselves within the margin of error.

This just proves that the pollsters really have no clue what they're talking about, and they're just using their polls as a political weapon to try and suppress political opposition.

 

[Grenz]

Apologies for taking longer than normal to get back to you.

That proves less than nothing.

That's just a confession by the polling firms that they really had no clue what was going to happen to begin with.

What a coincidence that right before an election, all of the polls suddenly find themselves within the margin of error.

This just proves that the pollsters really have no clue what they're talking about, and they're just using their polls as a political weapon to try and suppress political opposition.

Nope, it really just once again proves that you don't understand statistical analysis or quant research. If you're holding out to see a poll that predicts anything with 100% accuracy, you're going to be waiting a long time:

1) Seems obvious, but apparently it needs to be pointed out that you can't poll the entire population. So you go after a representative sample that matches the general population as close as possible. Obviously that will never be perfect, but the alternative - in this case polling every single voter in the US - is simply impossible. Thus, the idea of a representative sample will always introduce a margin of error. But even if you could poll every single voter....

2) Humans are not rational. Someone might tell you something on Monday and then do a complete 180 on Tuesday after they got some new information. Opinions aren't static, they're fluid, humans are highly emotional. This increases the margin of error even further.

If these sort of polls were totally off - like 10% or more off - then I'd agree with you in saying they're essentially worthless. But they aren't. They come pretty close in providing a picture of how the general population feels at a certain point in time. The fact that they do not achieve 100% accuracy isn't surprising, and currently, we have no better way of gauging popular opinion.

 

[Farmer Br0wn]

Nope, it really just once again proves that you don't understand statistical analysis or quant research. If you're holding out to see a poll that predicts anything with 100% accuracy, you're going to be waiting a long time:

1) Seems obvious, but apparently it needs to be pointed out that you can't poll the entire population. So you go after a representative sample that matches the general population as close as possible. Obviously that will never be perfect, but the alternative - in this case polling every single voter in the US - is simply impossible. Thus, the idea of a representative sample will always introduce a margin of error. But even if you could poll every single voter....

2) Humans are not rational. Someone might tell you something on Monday and then do a complete 180 on Tuesday after they got some new information. Opinions aren't static, they're fluid, humans are highly emotional. This increases the margin of error even further.

If these sort of polls were totally off - like 10% or more off - then I'd agree with you in saying they're essentially worthless. But they aren't. They come pretty close in providing a picture of how the general population feels at a certain point in time. The fact that they do not achieve 100% accuracy isn't surprising, and currently, we have no better way of gauging popular opinion.

This was post #43 of this thread.

https://www.politico.com/story/2016/10/clinton-trump-poll-abc-news-230199

This one from October 23rd 2016 was off by a few......

This poll had Hillary Clinton ahead by a ridiculous 12 points.

These are the kind of polls I was talking about. Mere weeks before an election we're given polls so wildly untrue that it seems almost comical in retrospect.

What other purpose do polls like these serve other than as an attempt to suppress one political opposition or another?

 

[Grenz]

This was post #43 of this thread.


This poll had Hillary Clinton ahead by a ridiculous 12 points.

These are the kind of polls I was talking about. Mere weeks before an election we're given polls so wildly untrue that it seems almost comical in retrospect.

What other purpose do polls like these serve other than as an attempt to suppress one political opposition or another?

This goes back to point number 2 I made. If you recall, subsequent polls showed a dramatic slide in Clinton's numbers, the hypothesis at the time being that Comey's public announcement that he was re-opening the investigation played a large part in changing some people's opinions. So if you looked at the polls from the date you linked (October 23rd) right up until the election (Nov 8th), you would have seen a clear downward slide in Clinton's numbers, right up until the point where all the polls I recall seeing basically had them head to head once you factored in standard deviation.

This election was extremely emotional, you had two of the most (if not the most) polarizing candidates since WW2, and Trump's unorthodox style, plus all the Russia shit (which affected both sides, Trump being accused of being in bed with Russia, and the Dem's insistence they were 'hacked') all made this an election where public opinion went through extreme swings. Then on top of this you need to factor in 'swing states' where voter opinion swings between both parties even without any significant events.

Again, this one was a tough one to poll, but show me a poll from a reputable firm that showed Clinton being ahead by 10 points mere days before the election.

 

[Greoric]

So now it's back to who gives a shit about polls, eh?

That was fast.

Or perhaps you're a partisan hack and you're projecting that hackery on someone else?

 

[Rational Poster]

Or perhaps you're a partisan hack and you're projecting that hackery on someone else?

Nope. I don't even vote.

 

[hillelslovak87]

Hollow policy when preceded by massive cuts to said agencies.