Hillary Clinton's lead over Donald Trump doubled Thursday (Sept...

Not sure why everyone parrots this around. It's not a straightforward poll - it weights responses by the respondent's confidence that they'll vote for their chosen candidate. So it's not comparable, and to me seems inferior.

Most Trump voters probably report 100% confidence for obvious reasons. There are probably a lot of hillary voters who hate her (but hate trump more), so they might put 70% or something to make themselves feel better about doing it. In a word, there aren't as many "reluctant" Trump voters.

No it doesn't. There is a separate section that weighs average confidence of either outcome of the elections. The poll is straightforward. Sorry, I used to be a Hillary fan, but after seeing that poll its convinced me that Trump is going to win. Hillary is just too much of a globalist shill.

He's also hit a 3 month low in prediction markets.

predictit.org/Market/1234/Who-will-win-the-2016-US-presidential-election

>Literally one outlier poll that probably uses mostly landlines
>in any way proves Trump will win

No the LA Times/USC poll for each day is a weighted average for the past 7 days.
Their main poll is a percentile of who you are planning to vote for, they have another poll for percentile of who you think will win, and one other one, can't remember.
So you could be 90% positive you're voting for trump but say 75% sure Clinton will win. It's two separate polls.

538 did an article on it and said it might lean trump, but maybe only like 2-3 percent. But it's not that their methodology is bad, it's easy enough to account for with simple adjustments.

I guess when you have an uberbourgeois neolib as a prime minister you get pretty desperate and start imagining things

But Trump is going to be crushed in the American elections. He was terrible in the debate, a fact even the New York Post acknowledged - and they're endorsing him for president. The electoral college is going to devastate Trump. The only swing states he might take are Pennsylvania and Ohio. But that won't be enough, and Ohio is looking doubtful seeing as he couldn't even take it in the primaries and Kasich fucking hates him/won't even endorse him despite being asked to take the VP slot.

You're wrong.

Typically, polls ask people which candidate they favor or lean toward. Those who say they don’t know or are undecided don’t get factored into calculations of candidate support.

The Daybreak poll, by contrast, asks voters, using a 0-to-100 scale, to rate their chances of voting for Clinton, for Trump or for some other candidate. As a result, everyone who responds to the survey has some impact on the results. Because that approach gathers information from everyone in the poll sample, it should give a better read on the many voters who remain ambivalent about their choices.
...
Using the 0-to-100 scale, however, almost certainly makes the Daybreak poll differ somewhat from other surveys.
latimes.com/politics/la-na-pol-polling-differences-20160809-snap-story.html

As others have pointed out, it is not only weighted by the chance that people say they will vote, but also it is a 7 day trailing average.

Now I see where the confusion comes from. You are confusing the section that asks voters who they THINK will win (which is interesting in its own right).

>there are people on this board who unironically believe Cheeto Benito's campaign isn't imploding like his self-restraint at 3 AM after a coke binge
all of my wew

t. cuck

Trump is winning Florida, Ohio, and Colorado and that is all he will need

>Read Post
>look at flag first
>(((leaf)))

Thank God I didn't have to read your post

Most polls are not straightforward. They like to do statistical tricks to be more efficient.
The standard double blind poll is rarely done by modern political pollsters.
LA times just does statistical tricks that are Trump leaning.
Others do statistical tricks that are Hillary leaning.
At the end of the day, error bars are so large that it doesn't really matter - just wait 'till election day.

>so they might put 70% or something
You have no idea what weighed average is, do you?

You look retarded because you dont realize why it's over sampled.

Well . . . I'm demoralized. My "countrymen" are severely retarded if they're voting for more of the same.

Yeah it's not weighted by their vote choice, it's a weighted moving average. I think they didn't do a great job of explaining that on their page. But it is a weighted 7-day moving average so it takes 7 days to erase results pre-debate.

Two things are happening in this poll.
1) They are asking people two things. Who they plan to vote for, along with their personal and subjective confidence that they will actually vote for that person. They then average the response's with each response weighted by the subjective confidences.
2) They then take the results of the past seven days, and do a weighted trailing average.

The key is that they are doing BOTH weighted averages.

Correct

They aren't weighting it by their vote choice though, they're weighting it as a moving average, as in they're treating it as a time-series so the most recent day is weighted more heavily.

No. Again, it is BOTH.

1) "We ask voters what the chance is that they will vote for Trump, Clinton or someone else, using a 0-100 scale. The overall level of support for each candidate reflects the weighted average of those responses."
2) "We update the data each day based on the weighted average of poll responses over the previous week"

so 7 days ago Trump was somehow doing worse than he is doing right now, and then the average goes up

It's over, the voting is only about a month away.

Donald Trump is going to win. There is a lot of stigma supporting Trump in public.

Kerry in 2004? I remember being very surprised that he lost, and it had to do with gay marriage referendums bringing Republicans to the polls.

Yet his rallies draw 20k people.

Trump has no shot.

See my recent posts. Some more evidence:

"Respondents are asked three predictive questions: What is the percent chance that (1) you will vote in the presidential election? (2) you will vote for Clinton, Trump, or someone else? and (3) Clinton, Trump or someone else will win?"

Everyone is right that they are doing a trailing time series average. 100% correct. But the components of that average are the results of a daily weighted average based on respondents' confidence that they will vote for their choice.

You can't weight support levels, it's impossible.

You can't say I support X candidate 70% but Y candidate 30% and weight it off that. Where's the weight coming from?
They're weighting by time while using the percentile support levels as data points inside the weighted moving average itself
en.wikipedia.org/wiki/Moving_average#Weighted_moving_average

That time-period has long since passed.

>Donald Trump is going to win. There is a lot of stigma supporting Trump in public.

lmao i have 2012 flashbacks

See above posts.

You're misunderstanding the process. Here is an example:

Person A: Who do you plan to vote for?
Trump.
How likely are you to ACTUALLY vote for that person on election day, as opposed to stay home?
100%

Person A: Who do you plan to vote for?
Clinton.
How likely are you to ACTUALLY vote for that person on election day, as opposed to stay home?
60%

They explicitly stated they're doing a 7-day weighted average though which makes your point moot though.
Okay, maybe they weight for voter likelihood but it explicitly states they do a 7-day average, which suggests a 7-day weighted moving average.

>Where's the weight coming from?

Americans have A LOT of weight though, I fail to see the problem here

>graph with no trump lead at all

seems legit

Again, see my above posts. I never denied that they are doing a seven-day weighted average. That's clear. It's also clear that, on each day, they are weighting the response of each respondent by the self-reported confidence that they will actually go out and vote.

"Okay, maybe they weight for voter likelihood"
very jewey

>breitbart
HAHAHAHAHA

It's not jewey at all, 538 showed that it actually gives trump an easy 4-5% bias when he's leading and a 1-3% bias when clinton is leading (based off trendline corrections off other polls). That bias is toward Trump though. Usually when the USC poll is at 5% expect to knock 2-3% off that to match it up in the poll aggregate.
They're more biased than Rasmussen reports, but it's an experimental polling technique anyway so who cares, they never claimed anything to begin with.

But the entire of idea of aggregating to begin with is you don't give a shit about the bias unless it's a ridiculously absurd bias because the aggregate easily knocks out outliers. Ipsos have had quite a few polls knocked out of aggregates due to being very far outside usual confidence intervals.

No, what I meant is that your way of admitting fault was very jewey. You said that they "might" be weighting for likelihood (even though I showed that they are) instead of owning up to being wrong.

youtube.com/watch?v=MnbEZjaZdy8

It's not that bad, ideally your polling group hits 100% likelihood they'll vote but the USC poll is just taking into account lazy voters.

Do you honestly think just cause someone picks up the phone and says they're voting for someone that they'll actually vote? Unlikely.
Pollsters have always been off by easily 1-2% in every election.

My original point was the following:

While I agree that weighting by subjective confidence in personal turnout is a good idea, I think that in this case it is probably overstating Trump's chance of winning. My reasoning for this is that more people are guiltily voting for Hillary. Most Trump voters, when asked by a pollster what the likelihood they will vote is, will scream ONE HUNDRED PERCENT. On the other hand, take your basic bitch bern victim who doesn't like Hillary but is voting for her because of scary racist trump. When asked the chance that he will vote Clinton, he might just say a small-ish number (say, 60%) to assuage his guilt for voting for the hillo bitch, EVEN WHEN his actual likelihood of voting is 100%. So basically, human psychology could turn what is supposed to be a weight of self-reported turnout probability into a weight of how-much-the-voter-actually-likes-the-candidate. This would understate Hillary's vote turnout and thus the results of this poll.

I'm not sure how big of an issue this is. But I think it's interesting.

Aww you're just adorable.

Trigger warning

Obama got destroyed in the first debate and he lost the election too.

Yes the USC poll is way way above average. But it's not determined by how much. You can't determine that until after the election, hence why 538 has no rating, no bias, etc. They accommodate for the poll via trendline correction (as in if it's way off other polls they'll downgrade the weighting in the aggregate)
The poll is not incorrect, it's not exactly accurate as far as we know, but it's a polling technique that's experimental. It also lags which is a bitch and half to begin with. I think everyone is too harsh on it and it's roughly accurate in backtesting, but we can't backtest with zero uncertainty until next election because this is the first election for this type of poll. It will only get better and better each election. Even in backtesting the 7-day lag still screws you though.

I do find it disturbing that people take it as de facto voter numbers but that's not what the aggregates are doing anyway. Anyone that readings 538 or RCP already gets their info properly weighted so it doesn't matter. If you're solely following the USC poll it's of course biased though.

I have a pole you foreign cuckfags can gloss over. You dipshits were wrong about the Brexit and you will be wrong about this election. You aren't from here thus have no idea about the general mood of the public. But keep believing what the kike media feeds you, like the good little goy you are, OP.

>a poll that isn't even used by aggregates, and isn't rated by 538

Yeah, and to be clear I'm pretty certain Trump is going to win. A lot of these "get out the vote" initiatives are scaring me, though.

The aggregate polling shows clinton ahead, statistically there has never been a single candidate to reverse a polling trendline this close to election, the R^2 value gets closer and closer to 1 the more you move toward election day.
Trump won't be elected. If he does it, not only would it be a world first it would be a punch in the face to every pollster in the entire world.

You DRUMPFAGS are so delusional.

Face the facts, Trump got utterly BTFO in the debates, and these polls prove it. This debate was his chance to turn things around and he blew it. Of course, I'm sure you dumbass drumpfags will give me the usual MUH RIGGED POLLS bullshit, but it's true.

It's time to give up, it's over. Trump was never going to win, the presidency belongs to Hillary Clinton. I would say you guys should have chosen a TRUE conservative, but there are no real conservatives anymore. The Jew has won.

Daily reminder that the polls are rigged to shit.

Even in a fantasy world where polls are fairly conducted, why would they be accurate for Trump supporters?

EVERYONE who supports Trump instinctively hides their power level by now. WE'VE ALL HEARD THE STORIES of what happens if you so much as wear a Trump hat in public. So when a Trump supporter gets an unsolicited call asking who they are voting for, 9 times out of 10 they are just going to hang up or say "I don't know who I am voting for". That leaves 1 in 10 who claim they are voting for Hillary (because they are turbocucks who have to lie in front of their shitlib wife who is in the same room if they ever want their penis touched again), OR have balls of steel and will say to a stranger, "I don't know who you are but yes, I live at this address and support Trump, please come burn my fucking house down".

Why perpetuate the pretense that these polls can be trusted? They aren't independently audited. They can simply make shit up and we're expected to take them at their word. If you trust these polls, you belong in a GOD DAMNED OVEN. Literally fuck your own face until you die.

The best indicator is OBVIOUSLY rally turnout - you know, that thing where people demonstrate they are capable of taking a shower and getting dressed and leaving the house, which ANYONE WHO WANTS TO FUCKING VOTE IS GOING TO HAVE TO DO.

EVERY. SINGLE. INDICATOR. points to Trump winning in a COLOSSAL FUCKING LANDSLIDE.

They bombard us with rigged poll results in a pathetic attempt to mass-gaslight us so when Hillary "wins", we won't question it. They WILL steal the election with closed-source electronic voting machines deployed nationwide; RIGHT NOW they are CONDITIONING YOU to accept it.

So tell me: When Hillary "wins", will you accept it as truth and go back to your slave life, you cowardly faggot? Or will you recognize it as THE FLASHPOINT and go forth and kill the enemies of light and truth in our midst once and for all?

Whatever the LA Times poll is doing, the general consensus seems to be that their formula is not good.

But - who knows... maybe they are right on the money and everyone else is off.

You know... just like that Rasmussen poll from 2012 that FOX and all the Republicans were hanging their hats on.

Fucking kikegun

You are listing the WRONG POLLS...

ONLY LA TIMES POLL IS VALID.

It wasn't anything anyone said during the debates. It was the fact that Hillary looked like a live human being and not a decaying corpse like she seemed to be in the days leading up to the debate. Trump needs to leverage Hillary's multiple felonies against her and hope for another Hillary health incident. He's still a live dog.

>kikefart.com
really fired up those neurons, i am now a #TrumpTrain