Current Polls - Maybe Rallies Don't Matter Edition

It just keeps getting better !

Other urls found in this thread:

longroom.com/polls/
youtube.com/watch?v=n6LfJdvB4-o
en.wikipedia.org/wiki/Overfitting
twitter.com/SFWRedditVideos

>A

>FUCKING

>LEAF

Yeah unless some huge scandal breaks out or hillary dies after the deadline that the party can change the candidate Trump has lost.


rip usa

One day we're going to be in an alternate universe where your leaf is upside down. But no one will remember.... Maybe its already happened.

Yea I can't wait till trump loses in a historical landslide so all the bro nazi bigot fags who post here an hero so this board can get back to being a civilized place for political discussions

longroom.com/polls/


This really makes me think...

>inb4 ur poll is phony >:^(


Also you cunts, it's not even a real leaf it's some fuck in DC with a leaf proxy.

Here's my counter
>1 post by this ID

>+6.8
hng, I came.

Trumpkins are tired of winning already I guess. Good thing for us Democrats, we never get tired of it.

+6.8?! If this continues, will Trump have any votes at all by November?

Really makes you ponder

Don't move the goalpost. Work hard to take back the lead and MAGA. The first step to fixing a problem is acknowledging that there is a problem.

...

Don't believe his lies.

>Look guise, if we don't count democrats, Trump is in the lead!
Math is liberal sport, retards. Leave it to us.

WTF IS THIS JPEG

MY EYES!!!!!!

Good job retard. Any fleas here to read it?

Never was such

lmao, counter.
this isnt a fucking card game, ahaha

Huge scandals have been breaking out pretty much every other week for Hillary, though. The MSM doesn't care, and would rather spend 3 hours of prime time about Trump "attacking" some Muslim literal who.

Your country is fucked.

I know this is b8, but
>oversampling Democrats makes for fair polls
>Liberal logic

Current polls assume the demographics are going to be the same in 2016 as they were in 2012. They're not.

>way less blacks are going to vote
>way less young people are going to vote
>way less dems are going to vote
>way more reps are going to vote

Proof of this is found in the primaries. The fact that the media honestly thinks there's going to be more dems than reps in this election is hilarious.

found the New Fag, do i get a reward?

>CTR shill saves thumbnail

I CANT EVEN

It is if you get paid per post : ^ )

>let's just give him a 10 point bonus for no reason whatsoever

>Poll aggreate that accurately predicted 2008 and 2012 within 0.3%
>Left out Democrats
>Pick one

If only they knew how to use the website, then maybe they would be more effective

>this flag
How many rubles do you get per post?

+6.8 HOLY SHIT TRUMP IS GETTING SCHLONGED

When they did that lame retroactive prediction model to predict a Romney blowout in 2012, at least they fitted their data all the way back to 1980. These people just fit their data to 3 elections and called it a day. It's like Trump hired illegal immigrant data scientists to make this one up and they needed to be done by siesta.

yes i agree fellow redditor onion

unskewed polls all over again

hope Sup Forums bites this one too, will make the assblast even funnier

>LongRoom Unbiased
>longroom.com literally created in 2016 as a pro Trump website

Reminder LongRoom Unbiased has correctly called the last three presidential elections with a maximum 0.3% margin of error.

They are by far the most accurate polling service.

Only 9% of registered voters voted in the primaries so I cant really trust relying on them also its not like republicans completely destroyed dems in votes I dont think it was that big of a margin

aaaand we're back where we started

lol at genetically inferior hamburger unable to read 5 micro font

Yeah like the super unbiased Reuters poll right?

A poll which has registered Democrats at 48% of their sample

they didn't exist in 2012 or 2008, they say their methodology says they had the last 3 general elections predicted correctly.

Yes, predicting the past is easier than predicting the future.

I agree. The question is whether it will be by enough to have a Republican in the White House.

Or maybe they know where the poll happens and are actively sending people there?

he's right, this board has gotten exceptionally stupid in the past year.

...

>2 posts by this id
they are catching on

You morons are falling for the demoralization.

But if Sup Forums isn't real, then how can polls be real?

theyre evolving

What's a "fair poll?" Equal number of Democrats and Republicans? So 50/50?

But what about Independents? We need an equal amount of them too. So 33/33/33.

But wait, there are Libertarian and Green Party candidates, so we need to sample them equally as well. So now it's 20/20/20/20/20.

Yay! Now everybody's winning equally! Hurray for fairness!

It's a right wing news site, not necessarily pro-Trump. They also normalize the demographics to match the most recent major election, to factor out oversampling of one political party.

Yes that's called fitting a model.

If their model fits so well to 3 past elections there's a very high probability it's a correct model with good predictive powers.

lrn2statistics

3 recent +9-15 Hillary polls aren't in that average

But sure, I suppose if you throw out half your opponent's lead and then give your candidate a free +4.2% bonus, you can make polls look like your winning.

>(((Real Clear Politics)))

One that corresponds to a likely percentage of voters in the general. This would depend on the area being polled.

That's not how sample size works.

>>way less blacks are going to vote
>>way less young people are going to vote
>>way less dems are going to vote
>>way more reps are going to vote
Something tells me there's a new, highly motivated segment of voters out there this time

What happened? Wasn't Trump at the top after wikileaks revealed Clinton's mail? Americans already forgot about that after a few days?

You said they correctly called the last 3 elections, not that their model retroactively applied predicts it.

Whatever helps you sleep at night, I'm sure whoever's behind this is making the same good Google AdSense money that the unskewed polls guy was.

You don't need a big sample size when your model fits the data with a 0.3% margin of error.

>I never took a Stats class: the post

A fair poll would represent party affiliation based on voting patterns.

A poll wouldn't be correct unless it sampled 30% Democrat, 29% Republican, & 41% Independent.

nice thumbnail medikitfag

Isn't Trump leading by like 10 points with Independents?

>Huge scandals have been breaking out pretty much every other week for Hillary, though. The MSM doesn't care, and would rather spend 3 hours of prime time about Trump "attacking" some Muslim literal who.

Couldn't agree more. It's actually concerning at this point.

Even the e-mails, something we KNOW happened and that the MSM DOES acknowledge, is now dismissed as a 'boring' or 'tired' topic, regardless of the fact that it was never appropriately covered by the MSM in the first place.

MADMAAM PRESIDENT

>
>>
>

>If their model fits so well to 3 past elections there's a very high probability it's a correct model with good predictive powers.

A quadratic function can always be found to fit 3 results perfectly. Having a 'model' any more complicated than a straight line that fits only 3 data points approximately is completely meaningless.

What makes you think Dems will be only 30% of vote come November?

Of course rallies don't matter you can't get a crowd of people for anything.

If you stick to the formula without tampering with it, yes you can indeed test it on previous election cycles to verify your methodology. Proof testing for standard models is all part of the scientific method.

No, a fair poll would choose people randomly. If more happen to belong to group A than group B, that tells you that more people might belong to group A than group B. You don't make a assumptions about those numbers and you don't choose people based on those numbers. Doing so does absolutely no good, except to give you hilariously inaccurate results like the ones you shills keep spamming to make yourselves feel better.

Numbers don't care about your feelings.

The graph has gone suddenly way too erratic.

This. Noone knows what the turnout will be. Any model is biased including RCP and Longroom. The polls tell us nothing

It's all about the cycles, do you guys even know how to read charts?

Jesus. Trump should quit before he turns Texas Blue.

primarymodel.com

Just gonna leave this here.

He's way behind for him to realistically win, even if it's adjusted for bias and voter registration

>Something tells me there's a new, highly motivated segment of voters out there this time
And they are?

we're reaching stalin era levels of propaganda, i think im gonna be sick

>What makes you think Dems will be only 30% of vote come November?
Because

>I am popular with imbeciles whit a lot of free time on their hands, who love to go to wrestling matches and other hate-based events in droves
>Rally actually looks like a wrestling arena filled to the brim with rubes

lol fucking shill, go back to le reddit faggot

Sage goes in all fields. Poll threads are shill threads. Everyone knows they're cooked.

For the record, the PRIMARY MODEL, with slight modifications, has correctly predicted the winner of the popular vote in all five presidential elections since it was introduced in 1996

>slight modifications
i.e. when our data was shown to be wrong, we changed our shit to fit the results.

that's the amount of REGISTERED DEMOCRATS YOU MONGOLOID

Is that stat even a good thing ?
Why is he spending so little to get his idea across.
That stat looks like total dumbassery considering where he is now in the "Biased" polls.

JESUS!

REMINDER

youtube.com/watch?v=n6LfJdvB4-o

Not if all 3 points are collinear you fucking moron.

No he's not even close to over. They're asking 800 registered voters. There's no guarantee all these people will even vote

...

en.wikipedia.org/wiki/Overfitting

>Having a 'model' any more complicated than a straight line that fits only 3 data points approximately is completely meaningless.
Well when I was in engineering school 3 points for an anchor were good enough to graph a straight line, and for statistical analysis a straight line is good enough to graph a TREND.

That just means the coefficient for the square will be 0. Learn to math or stfu.

When are people gonna realise polls don't win shit? Remember when polls gave Remain a 10 point advantage IN THE DAY OF THE BREXIT REFERENDUM? It's over when it ends. Besides, Bush won with a couple of voters less lmao

This. It's probably less than the republicans now though because of the demexit.

This was intended for this comment