I am debating about Nate Silver's predictions. What was

Help a brother out.

>I am debating about Nate Silver's predictions.
I suspect you will be debating his shit tier analyst or straight out lies... depending on how you look at it... until the next election.

"Silver has described himself as "half-Jewish"."

https://en.wikipedia.org/wiki/Nate_Silver

>>132343982
Of course he is jewish. But I am talking about his predictions being shit.

>>132343700

that photo is a classic

he was actually moving the odds towards Trump at the end but it all kinda got lumped in with Huffpos 98% chance for Hillary. He was technically the closest out of the cuck pollsters

>>132344239
>But I am talking about his predictions being shit.
Yes and I am saying that he was just one of the many lying Jews shilling for Hillary.

If you are trying to make logic out of Jewish lies you are in for a endless journey of confusion.

>Jews makes false predictions about a candidate that all his fellow Jews are making false predictions about.
What do you except him to do? Stand up against that wave of kike lies and be destroyed by the kikes in the media for suggesting that Trump could win.

At some point to goys will learn to stop taking these lying Jews serious and start realizing everything they have is based on these type of lies and unprovable claims.

Anything else? You done?

70-30 favor hrc

File: 1478893526104.jpg (2MB, 845x8546px) Image search: [Google]

2MB, 845x8546px

>>132343700
here, have a quick rundown

He predicted 70-30 for Hillary, which sounds about right. Just because Trump won doesn't mean those weren't reasonable odds. Nate spent a good amount of time of shitting on the people who openly dismissed Trump.

Thanks guys. I remember the 70 / 30 now.

He did feel the hillary wins popular while trump wins electoral was a plausible scenario

File: Yeah Nah.jpg (161KB, 1024x768px) Image search: [Google]

161KB, 1024x768px

>>132343700
Everyone was doing the same corrections and never stopped to think that it didn't matter how many beans they crammed into California.

Obama spent tons of time busing niggers & refugees out into the middle of red states trying to help change those demographics.

He was unironically the most accurate of all the mainstream data guys and actually took heat on Twitter for saying Trump had a 30ish percent chance. He's memed on because of the primaries mostly and for everything he said/says that isnt pure statistics.

The thing about prediction models is they can only predict based on the things you tell them to look at

He did have Trump at like 30% which is within the realm of possibility but I also think his model (and the pollster's data he was basing it on too) was probably calibrated for a race between a run-of-the-mill Democrat and a run-of-the-mill Republican, and got caught off guard like everybody else by Trump being... well Trump instead of a normal Republican

I know its fun to shit on smug "analysts" who get it wrong but honestly I can't fault him too badly, he was more on the ball than most people and was also one of the first to admit Trump was gonna win if I recall correctly

He said the Cubs would win the World Series and then 8 days later Trump would win

He predicted the whole thing.

https://twitter.com/natesilver538/status/730251094614528000?lang=en

/pol/ is just filled with Silver haters

>>132343700
On the eve of the election Nate had Trump at 28%. Over the course of the final two weeks he had him at ~35% for most of that time.

Don't know why the /pol/ hiveminds shits on Nate so much. Out of all the projectors Nate always had Trump at a reasonable chance to win and always said that it was very possibly for Trump to win.

File: r87.png (183KB, 680x591px) Image search: [Google]

183KB, 680x591px

>>132343700
Use his tweets, here he is being all sarcastic and both things wind up coming true

>>132348441
>>132348622
>>132348866
So the kike gets it wrong but because he didn't have Trump in the negatives and Clinton above 100 he gets a pass?

Nate Silver and the rest of the kikes who peddled their polls in the MSM were wrong. The polls were shit and everyone knows it and the fucks like Nate bandwagoning right along with the media will be labeled as the lying kikes they are.

File: Nate Nostradamus.png (39KB, 581x479px) Image search: [Google]

39KB, 581x479px

>>132348622
This.

Not to mention this outstanding prediction shortly before the election where he analyzed recent sport events to re-confirm his point even further.

>>132349433
>>132349213

Also, a modern day German defending a Jew. Why am I not surprised.

>>132349403
The polls weren't actually wrong on a national level.

>>132343700
>I am debating about Nate Silver's predictions
Why? What the fuck is the point of wasting your time with that?

>>132343700
He wasn't wrong because his prediction was wrong. He was/is wrong because his entire methodology is flawed.

https://dl.dropboxusercontent.com/u/50282823/binary%20forecasting%20538.nb.pdf

File: nate-silver-following-nate-silver538-a-chicago-cubs-vs-cleveland-5282753.png (133KB, 500x661px) Image search: [Google]

133KB, 500x661px

>>132349213
This one too, back when he had the odds incredibly low for both

File: nElXDz.jpg (799KB, 690x6602px) Image search: [Google]

799KB, 690x6602px

Blumpf!

>>132349758
>He wasn't wrong because his prediction was wrong. He was/is wrong because his entire methodology is flawed.
>He was not wrong but he was wrong.
Pure insanity.

>>132349888
ah hell, i'm feeling good lads. Let's make this a liberal butt-hurt thread.

He was one of the most accurate other cunts said Trump had 2% chance
He got told off by other pollsters for inciting panic saying Trump had 30% chance

>>132350023
Had Hillary won, his methodology would have still been wrong. If I can claim to predict the outcome of a coinflip based on my magic model and get it right 5 times in a row, will you pay me $10,000 to place a bet on the next coin flip?

You personally. Probably but that's because you're a moron. But most reasonable people probably would know it's BS.

>>132350236
Scott Adams and George Soros were the most accurate. Nate Silver's analysis was useless.

>>132350354
Didn't Soros predict Trump would win the popular vote but lose the EC? Literally the opposite of what happened.

>>132350498
vise versa

Why study a loser who was wrong when you can go to /pol archive and see thousands of people who were right?

>>132344278
they got into a big bitchfight because he had Trump at 30 percent late. It was down to 20 by election day though

>>132350323
>Had Hillary won
kek

>Probably but that's because you're a moron.
See this is the problem with all your cucks trying to shill for (((Nate))). You claim I am a moron but at the same time you excuse yet another one of the kikes who peddled their fake polls to try and keep with the storyline the rest of the kikes in the media were pushing. They were all wrong but you keep trying to excuse it.

Honestly, I hope you and the rest of the kike cock suckers keep pushing the failing 2016 agenda and tactics and lets see how far it gets you.

>>132345808
>>132346315
>>132346426
>>132347465
>>132348441
For all the pride this board takes in the hard sciences, it gives me a chuckle whenever this comes up. 70/30 is basically a 1 in 3 shot in favor of a trump presidency.

The weekend before the election Silver compared Trump's odds to winning a tough NFL road game. He insisted it was not impossible, but Clinton was the clear favorite.

>>132343700
He gave Trump ~30% odds on election day, which is more than anyone else.

I ran my own election simulation, and Trump never exceeded 20% probability. Pennsylvania, Michigan, and Wisconsin were never even on the table. It was a serious stretch to imagine Trump winning any one of them, unless you were a hardcore Trump loyalist who only read Breitbart and Drudge. I figured Trump would need New Hampshire or Nevada.

>>132348866
>Don't know why the /pol/ hiveminds shits on Nate so much
He was a pundit during the primaries. Rather than using his models and data to predict the Republican primaries, he agreed with the conventional wisdom that Trump couldn't win because he's not a politician.

>>132349729
The state polls weren't wrong either. It's just that the undecideds went overwhelmingly for Trump. Hillary's vote total was in line with the polls, but Trump gained a lot.

That suggests
>there really were shy (hidden) Trump voters
>Trump gained a lot in the final few weeks

Probably both, but the Comey letter certainly didn't help Hillary.

>>132350323
Fucking terrible comeback. You should feel bad.

>>132348622
hey fucknut look at the date on >>132349213

he tweeted that in May '16 sarcastically and you're trying to claim he was making a legit prediction.

kys literal kike shill

>>132350856
>For all the pride this board takes in the hard sciences
>hard sciences
>/pol/
You do know what board you are on right?

>>132351003
Some of the state polls were absolutely wrong but that's what you get when state polls have lower quality methodology, lower frequency and smaller sample size.

>>132343700
He actually started yelling at other pollsters that Trump did have a path to victory. Nobody believed him.

>>132343700

Check this shit out.

I work in media.

I work for that news network.

You know...THAT news network.

I am that guy.

I will tell you without a hint of reservation that all the polls are skewed.

We always -- and I do mean always -- downplay the sampling sizes and demographic data.

These faggot fucking pollsters make something like 900 phone calls, with 60 to 70 percent of the people on the other end claiming to be Democrats...

...AND THEY STILL CLAIM THE POLLS ARE INDICATIVE OF THE FEELINGS OF THE ENTIRE FUCKING COUNTRY.

"Oh, you just polled 750 Democrats who said Trump is doing a bad job? Cool. Our new headline is 'America Dumps Trump!'"

It's a fucking joke.

Every pollster should be punched in their stupid little fucking mouths.

They get paid to do this shit, too.

Companies pay them to make polls for them.

Polling is fucking cancer.

File: nate silver is fake news.jpg (2MB, 2760x1775px) Image search: [Google]

2MB, 2760x1775px

>>132343700

Nate Plastic (and everyone else at (((((((((((538)))))))))))) for that matter) is a fraud that is only there because of jewish nepotism.

>>132351003
Trump filled major league sports stadiums several times a week while Hillary couldn't even fill middle school gyms. It's your own fault if you put your head in the sand and made simulations based on data you got from Huffington Post.

>>132351375
Portubro called it the whole time

Nate Bronze

File: 1486083393514.jpg (54KB, 571x438px) Image search: [Google]

54KB, 571x438px

>>132351375
I used data from realclearpolitics.

A small enthusiastic base helps, but total votes are more important. It doesn't matter if you rile up your voters or drag them on a leash.

>>132351816
Doesn't change the fact that people who based their predictions on data got it wrong, nearly 100% of the time. I think LA Times was the only one with polls showing Trump would win. Meanwhile, people who made their predictions based on reality and what was actually happening in the country got it right. No matter how much you insist that you should have been right, you will always be wrong. Doesn't help that they were consistently oversampling Dems as much as 70%, or that they actually deluded themselves into thinking that Hillary would get the same black turnout as Obama. But hey, that's what happens when your whole entire worldview revolves around rejecting reality.

30% is a 9/4 bet or +225 in US sportsbooks basically its not far off Trumps proper probability he may be a weirdo but hes pretty good with the sums

>>132351063
Wow you are an angry elf. They predicted low odds for both right up until it happened. 538 put cubs at 15%

https://fivethirtyeight.com/features/the-cubs-have-a-smaller-chance-of-winning-than-trump-does/

>>132352402
Archive clickbaiters
https://archive.is/WAqGJ

>>132351375
>Trump filled major league sports stadiums several times a week while Hillary couldn't even fill middle school gyms
and yet still more people voted for her so this means fuck all

he was looking at the same polls everyone else was looking at, I think he was the only one, or one of the few, who pointed out that she has a better popular vote lead than Obama did before the 2012 election but that her EC numbers are much more shaky and that a point or so swing towards Trump due to all the undecided voters could make her loose some close states

>>132350780

I posted a link explaining why Nate Silver is more wrong than simply having an incorrect prediction but it has maths in it so proceed with caution to avoid being triggered. You don't have the reading comprehension to see that I'm calling Nate Silver a moron just like you. Sad.

File: deep pondering.gif (2MB, 300x168px) Image search: [Google]

2MB, 300x168px

>>132352626
>and yet still more people voted for her so this means fuck all
I wonder why 22 of the states, including California, are refusing to share voter data so the Trump administration can't investigate potential voter fraud? Surely nothing shady is going on here?

Does 538 do its own polls? Or was he reliant on the fake ones? When all the polls are fake its hard to have decent accurate predictions.
Brexit was worse there was £80 million bet +110 on Betfair alone

>>132353699
538 aggregates polls and weights them based on how accurate their previous predictions are.

And they don't do predictions, they do forecasts. More similar to "30% of rain" than "30% of a trump victory"

File: Nate-Silver-4.jpg (227KB, 690x2622px) Image search: [Google]

227KB, 690x2622px

>>132343700
>Do you guys remember any of his tweet or articles?

something about a ceiling

File: Nate-Silver-5.jpg (46KB, 600x391px) Image search: [Google]

46KB, 600x391px

>>132343700
>Do you guys remember any of his tweet or articles?

Math doesn't work.

Its important to stress just how dumb and gullible pollsters like Nate Silver are. The MSM spent a year fear mongering about how Trump supporters were violent, dangerous Nazis, while at the same time insisted that there was no way there could possibly be a "shy Trump voter" effect. And pollsters were actually dumb enough to believe that and account for it in their predictions.