Trump leads in PA, OH, FL. What now, Hillarykins?

*sips tea*

Other urls found in this thread:

projects.fivethirtyeight.com/2016-election-forecast/updates/#plus
twitter.com/SFWRedditImages

DELETE THIS

Cool I guess he'll be president then. You win!

NOOOOOOOOOOOOOOOOOOO

This is cherry picking though, as he's losing Kansas and tied in South Dakota and Arizona in some of those google polls. I don't trust any of them.

Look at the national Google Consumer Survey. What a sample size. 22,673 people. Crazy.

SurveyMonkey has 32,000 but it's proven to be a shitty pollster though

>$0.05 has been deposited into your account

>believing fake polls

Time for the Great Crusade.

Source?

Those numbers look too good to be true desu familia.

DELET THIS

It's the Google consumer surveys, but OP's pic is fivethirtyeight.com's poll aggregates

Polls are irrelevant, all that matters is the voting machines.

>google polls
>giving trump a lead
what the fuck. Also how the fuck is Trump in the lead in New Mexico?

Look in 538's recent update: projects.fivethirtyeight.com/2016-election-forecast/updates/#plus

They're google consumer surveys. But they're bullshit because one of them has Hillary leading in Kansas by 17 and in South Dakota by 8. Sorry, but I'm not buying it. Even if it has favorable numbers for Trump in the swing states.

Polls only truly become accurate a week or so before the actual election.

You can't exactly only follow one pollster, that's why there is poll aggregates to begin with.

The google consumer surveys are ultimately flawed because they only utilize online surveying, not live interviews and such. They're good, but not that good.

I agree about polling aggregates, but I think these google survey results should be tossed out and not included in the polling aggregates. Their numbers are just insane

Factoring it into the picture of things as a whole.

It's not gospel, but it's interesting. I'd like to take a look at their methodology.

Well look at the sample sizes. Fucking 59 for SD. That's definitely bullshit.

True, but some of these surveys included as few as 59 people. That's not representative of any population anywhere

They've been getting a lot of hype because compared to past online-only polls google's survey method is more well-proven. Their sample size is insanely large for a rolling daily poll as well.
Google Consumer Survey's has had a lot of past success with things other than elections, and there's a lot of research and development on the process compared to Reuters and SurveyMonkey's polls.

There will always be outliers but discarding these results would be a huge mistake, outliers are easy to discard anyway.

But how do we reconcile Clinton's 17 point lead in Kansas? Especially when most of the other surveys conducted in the state have Trump with at least a 10 point lead?

They don't explicitly take state-by-state polling. They take a national one, then take statistics based on which states each individual came from. So one poll might have a sample of 200 from one state and 10 from another, but quickly reverses the next day.
It's a brand new method which is why 538 ranked them lower than other pollsters, but still ranks them high for the fact that they have a huge sample of 20k+ individuals.

Google does not do state polls, they take national polls and split it into state polls. Similarly Reuters, Ipsos, and SurveyMonkey also have done the same.

>What now, Hillarykins?

Really highlights the faggot tone of anti trumpers. Adding the suffix "kins" is the type of shit the gay guy from 30 rock would do. Make your lips bleed after I punch them into your teeth kind of feel.

You could easily just do a half-assed outlier assessment and say it lies outside the 95% confidence interval in aggregate so you just toss it out. Not exactly pro-statistics there but nobody would argue with it.

NC, PA, and OH line up with current polls, but FL? Is the shilling really so bad in florida that they skewed the polls that far? And NM is never polled for some reason.

Trump will probably win. There just aren't enough smart and educated people to elect Hillary. America is 90% trash.

Can someone explain this to me? Looks in this "google consumer survey" for Florida Trump has a commanding lead. Yet the typical MSM pollsters show Hillary leading or him tied with her. Is the google consumer survey more accurate?

No it's not more accurate. It's a 50-state-poll not a poll of 50 states.
It's more accurate if you average their results over say a week or two, not by the day. They poll 20k-30k people a day, their state-by-state polling varies wildly and is only accurate as averages over longer periods of time to account for that.

I don't trust google, ipsos, or surveymonkey. I usually go with the more established pollsters like Quinnipiac, PPP, and Mason-Dixon. From those, you can see that the race is very much tied.

>It's more accurate if you average their results over say a week or two, not by the day. They poll 20k-30k people a day, their state-by-state polling varies wildly and is only accurate as averages over longer periods of time to account for that.

This makes sense, but a lot can change in politics in a week or two. That's why I dislike those surveys. I could be wrong.

Well over a 2 week period the average for their poll for Florida is Trump +9. That would fantastic if true because Florida is crucial.

Anyhow, polls are all over the place right now. Still interesting to look at because any poll that shows Trump is winning (considering the heavy media bias and stuff) is a feat in itself.

>tfw debate in 2 weeks.

It has a lag sure, similar to what the LA Times/USC poll does, which lags exactly 1 week according to their methodology.
The idea is by using huge sample sizes you're bound to get at least 1 person from each state or voting region so that technically makes a "sample" of that state.
It's not a balanced state poll like traditional pollsters but it is accurate in a moving average, it just lags. The lag isn't quite known simple sample size varies too much so it can lag even more than 1 week like the LA Times/USC poll, it can even have a varying lag period.

midwesterners are cucked

Trumps grand piviot was away from corn cramming muh jesus flyover states to the rust belt

It's so depressing thinking about how rigged voting is just something that everyone basically accepts.

>projects.fivethirtyeight.com/2016-election-forecast/updates/#plus

When you look at the whole page of states, it's actually not incompatible with what RCP is showing us, etc. Things are very much in flux. I do see--barring any dumb fuckups--Trump taking Florida, Ohio and maybe even Pennsylvania. Certainly Indiana, Iowa, Idaho etc.

What's interesting is that looking at the past versions of this survey week by week, Clinton's lead in CA keeps trickling downward. Not fast or considerable enough for Trump to win CA in 8 weeks, but IF things continue as they have been, CA will be surprisingly close, likewise OR where he's only down by ten-ish points.....even if he loses those states it will be close enough to constitute a hell of a message for the Dems.

For perspective, several months ago the Field poll was showing Trump down by like 30 points in CA; now, it's 16 points as per both this Google survey and the most recent PPIC poll in mid July. The shift continues. Most interesting.

There are only two polls extant of NM besides this as per RCP, both from mid August. He was down by 8-9 points in those but we know polls are all over the place this year. And that was 7 weeks ago, more or less. NM does have a Republican governor so from this you know there must be sufficient Republicans there.

>New Mexico
>Trump +3

Fuck everything else, this is the REAL news. How the fuck is this happening even?

Nevar forget, Kansas is full of Cruzmissiles and NeverTrumpers. We found this out in the primaries. Luckily, they are virtually useless.

I still think he will probably take Kansas though.

Based legal mexicans

All of the liberal boomers in Albuquerque realized they doomed the world and are gonna go out the right way.

DELET THIS NOW YOU DRUMPFKINS

The message will be that they won. By 2020 the demographics will be even worse.

>Plus Hillary will appoint the next Supreme Court.

$0.05 has been deposited into your America First™ Account.

Naw fool, you are comparing polls with Gary Johnson to polls without Gary Johnson.