HE'S IMPLODING

>rasmussen

lmao this election is a repeat of 2012

>Release 2 NBC polls in 2 days
>don't count Rasmussen +5 for Trump
>t-this isn't to boost the average

really makes you think

>Outlier LA Times poll released daily

hmm really makes one ponder.

Consider this record corrected.

I'm stupid, and I don't have a college degree, so I need someone to explain this to me.

How does asking a few hundred people, give an accurate representation of what many millions of people will do? It doesn't make sense to me.

>LA Times maintains normal interval
>NBC feels the need to change on this one occasion when Trump is up

hmmm really fires those synapses

why are NBC polls so different from the rest?

You say that, but the countless spam threads don't really help prove it

go back to begging for a gf

CNN has decided to stop polling all together after their Trump +1

...

LA Times is questionable in results, but the methadology is clean.

It's like 2012 all over again.

rasmussenreports.com/public_content/politics/elections/election_2016/white_house_watch_sep22

>Trump 44%, Clinton 39%, Johnson 8%, Stein 2%

oh look

FOCUSED ON THE PAIN

NBS, starts shilling polls and trump polls go down

(((NBC)))

youtube.com/watch?v=zMGZtkMS3sQ

I agree, it's pathetic how desperate these TrumpTards are. They just can't stand the fact that their racist white supremacist ideology is a failed ideology.

just like brexit :^)

>Trump is a white supremacist.

Well, it's not that it *does* give an accurate representation, it's that it can do if you do it correctly.

Basically, you need to use representative sampling. If you're going to sample 100 people (this would be very low, but run with it), then you'd want the correct proportion of these people to be black, white, asian, and whatever else to represent the general population. The same would be true for things like where they live, their income bracket, and so on.

Obviously, there are so many factors to control for that you need to be pragmatic and decide that certain factors don't affect it - star sign, for example.

This means that you really need to look at the methodology of a survey in order to tell how accurate the results are - the correct factors must be controlled for, and they are often lazy or incompetent and fail to do so. As long as it's done properly, a comparatively small sample can estimate the population values quite well.

There's another big factor in these Trump surveys though, and it's the same one we saw with Brexit - there is immense social pressure not to vote for Trump, so there could potentially be quite a large underestimate of Trump's share of the vote right up until the actual results come in. With Brexit (where people who wanted it were similarly slandered as racists), the polls showed 60% in favour of remain, including the exit poll on the day itself - and we all know how that one turned out.

>current year
>not knowing how ids work

Oh, it's worth noting also that the presence of these factors which must be controlled for means that it is quite easy to manipulate survey results to give the result you want if you know what you're doing. For example, you could fail to make the survey representative of race, "accidentally" sample heavily from the black population, and conclude that almost no-one is voting Trump. This is too obvious a ploy to actually be attempted, but the nature of the "random" representative sampling used means that this is a possibility.

I hope MSM shows her winning polls all the way until election day.
No one wants to vote for her. Many libs will stay home if they think she's winning.
On the other hand, Trump supporters will crawl through broken glass to vote

Only way you leafs know to win
Is by losing

...

Does CTR still not understand there are IDs on every post?

>The polls were accurate for the past 35 years

Polls conducted in the same week by different people are vastly different.

And do you really think that mid-September polls, pre debate, for the last 35 years, accurately predicted the outcome of an election?

Well... Good luck, leaf!

I'd suggest you stop talking to yourself in public, makes you look like a loony.

yep, fucking corrupt banana republic still pretending that votes and voters matter

NOW THE POLLS ARE FAKE AGAIN! DAMN KIKES!!!

2 nbc polls yeah i don't think so, bias

>Canada agreeing to himself
how sad

>Hilary's worst month imaginable
>trump still can't pull away, and is even losing in some polls

WAKE ME UP

But call him a Leaf and you will be astonished at how he recoils, how injured he is, how he suddenly shrinks back, I’ve been found out...

...

Anyone who has even a basic grasp of survey methodology knows that social desirability bias is likely going to result in an underestimation of Trump's vote share until the actual results are in. Comparing this election to the elections of the past 35 years is obviously ridiculous.

The exact same thing happened with Brexit, too. It's like you're wrong not just from a theoretical standpoint but also from a recent historical one. Impressive, leaf.

This, just like every election, is a shit show to harvest donations and ad buys from people that are still in denial about how corrupt the US government is.

>wikileak emails confirm illary bought the MSM
>still believing their polls

As long as the Trump doesn't say anything stupid, he'll be fine. That dip in the polls is meaningless.

It's only because Hillary's negative press has dissipated, not because Trump has done anything bad.

...

>first day of fall
>leaves already fucking everywhere

You do realise that as long as he within 5 points, it's too close to call?

Don't believe the (((polls))). I predicted a week back that the (((MSM))) would be putting out polls showing Hillary BTFO'ing Trump in the run up of the debate.

>most polls are updated every one or two weeks
>LA Times in favor of Trump has been in free fall the past few days
Hmmmm

Think about it this way. Start by pretending you know the exact figures and working backwards.

> Assume you have special knowledge that;
> 40% of people in the country support candidate X.
> 60% support candidate Y.

Because you know this, you also know that each time you ask a random person on the street who he supports, there is a 60% chance he will say Y, and a 40% chance he will say X.

Ask one person who he is likely to vote for, and he will probably say Y.
>Y 100%
Ask two people, one will probably say X and one will say Y.
>Y-50%, X-50%
Ask three people.
>Y-66%, X-33%

As the number of people asked increases, the result converges on the true result. Obviously you'll only get the exact figure once you ask everyone - but after enough people are asked, it's a statistical certainty that the poll can get close enough to the real result to be meaningfully accurate.

>Trump leading by +4, +2, +1 in 3 polls
>Tied in 1 poll
>Hillary leading by +2, +1 in 3 polls

>NBC +5, +7 FOR HER

Yeah, the NBC polls are bullshit. This race is pretty much tied with Trump slightly edging a lead. NBC thinks that Hillary will get the 2012 Obama turnout and Bernie 18-24 millennial turnout.

DRUMPF BTFO

When the flying fuck were those taken? All the current polls show him either climbing or leveling off above her

Do the shills not understand that him performing badly in polls makes it more likely people will bother to go and vote for him?

Exact same thing happened with brexit.

>mfw President Hillary soon

can't wait for 24/7 burgerfat ass pain on Cred Forums

We were due for another dip anyway, it just means that if he regains the lead again it'll be an even bigger panic

They're hoping to make his victory chances seem hopeless, so turnout drops.

The election is mine Don!

...

JEB SHOT THE NIGGER IN CHARLOTTE CONFIRMED

It was a drive-by.