Donald Trump

jim tolmies perm said:
Really anyone believing these recent Trump surrogate polls needs to give their heads a wobble. The real data, that is registration and early voting suggests that Trump is toast, which maybe one reason he's sped off to Florida where Harris now has the "big Mo"
But of course the corporate media is desperate for a race and to keep the Trump clown show going in order to keep the ad men intersted.
Click to expand...

In that case surely lump loads on Harris at 11/8 right now ?
 
KS55 said:
In a poll where the margin of error is, say 3%, a swing of 0.5% to Trump from poll1 to poll 2 is of no statistical significance whatsoever.
Click to expand...

Margin of error is 1 divided by the square root of sample size. So in a sample of 16 people the square root of 16 is 4 and 1 divided by 4 = 25% margin of error.

A sample size of 1000 is about 3.2%
 
bluenova said:
Very true - but I don't think that's what we're seeing in the US polls.

If it's the same polling company comparing their one weekly poll, then all those X is +1 and Y is -1, don't prove anything on their own.

However, where you have multiple polls combined, showing a trend over time, then statistically that's the margin of error being all but ironed out. Any poll watcher who sees a poll that's out of sync, or shows a new direction, should always wait for further evidence.

In the US, there has definitely been movement towards Trump - and that's showing in multiple state level polls, where the majority of the close races have flipped from Harris to Trump. Some of those changes will be well within the margin of error, but put together, it's a real trend.

Now, as @SWP's back has pointed out, there are problems with polling beyond the margin of error, which will skew all polling. I'm not sure I agree that it's definitely giving them a pro-Trump bias, as there are reasons why Trump supporters may also be under-polled. These effects also change between elections, with the polling companies having to be careful not to over-correct for the bias they showed in the previous campaign.
Click to expand...

I think people need to be really careful with the assumption that the majority of polls moving in one direction = a trend.

There’s a well known phenomena in polling known as herding where polls will tend to cluster in groups because as a pollster it is dangerous to your reputation to be a significant outlier.

The best example I can think of of this is the 2017 UK general election. All of the polls clustered around the same conclusion which was a comfortable victory for May. The only major pollster that got anywhere close to Labour’s 40% vote share was Survation. The rest were all out by at least 5%. They had herded themselves completely in the wrong direction.

I have shared research like this on here before but “perceived momentum” has a measurable impact on polling:


In short, people think that a party is more likely to perform well if polling momentum is in their favour, this makes them more likely to say they’re going to vote for that party even if they won’t (there are plenty of people who just want to be on the winning side). This creates falsity in the results because often these people are answering polls based on perceived social pressure and not their true voting intention. It is a small effect but it can make all the difference and it can generate these weird feedback loops where a candidate seems to be gaining steam in the polls but it is all a complete social illusion.

Election polls are more accurate if they ask participants how others will vote

business.leeds.ac.uk business.leeds.ac.uk
(This is super interesting btw)

I think the reality is that most US polling methods are simply not fit for purpose. The margin of error extends beyond their usefulness and it is really hard to draw anything concrete for them. They will be within 5% but when the race is going to come down to a 2-5% max lead that renders a lot of them of limited use. Polling in Europe and the UK is better (still not perfect).

What is of most interest to me is the Dems are not releasing polling they’ve funded with a few exceptions. GOP are releasing them several times a day. That’s weird to me and it suggests they are both seeing information that the rest of us aren’t.
 
Looking forward to the next Joe Rogan podcast. Rogan does not seem to be a fan of either candidate whatsoever so I don’t think Trump will have an easy time of it.
 
Longballutd said:
Looking forward to the next Joe Rogan podcast. Rogan does not seem to be a fan of either candidate whatsoever so I don’t think Trump will have an easy time of it.
Click to expand...

I’ve had Rogan down as a anti vax republican conspiracy nobhead for quite a while - so I suspect he will give Trump the easier ride.
 
Longballutd said:
Looking forward to the next Joe Rogan podcast. Rogan does not seem to be a fan of either candidate whatsoever so I don’t think Trump will have an easy time of it.
Click to expand...
Fuck all chance, he'll lick Trump's hole. Why do you think he's agreed to do it in the first place?
 
You must log in or register to reply here.
Unread Posts

Don't have an account? Register now and see fewer ads!

SIGN UP
Back
Top
✖
  AdBlock Detected
Bluemoon relies on advertising to pay our hosting fees. Please support the site by disabling your ad blocking software to help keep the forum sustainable. Thanks.