Nate Silver

How accurate is Nate Silver's model?

Other urls found in this thread:

fivethirtyeight.com/features/the-odds-of-an-electoral-college-popular-vote-split-are-increasing/
fivethirtyeight.com/features/election-update-the-campaign-is-almost-over-and-heres-where-we-stand/
twitter.com/NSFWRedditVideo

Nate Aluminum

Nate Tinfoil

58% doesn't mean 100%

Haha lol he was wrong about trump haha remember that from two years ago guys

Education is for liberal cucks so I wouldn't know

Nate Dihydrogen Monoxide

40% was about what i figured this result would be.

I mean he wasn't wrong on the popular votes, its just the popular vote doesn't mean shit

>It's another Sup Forums doesn't understand that just because you have a

42% :^)

he wasn't wrong though, he was the guy who was screaming "TRUMP HAS A CHANCE GUYS" so much that batshit twitter libs started accusing him of being an anti-hillary shill

...

He only uses ELO rankings for American football right?
I that's the case I wouldn't say that's it's very reliable, because he doesn't takes in count thing like injuries or current form

He was the only pollster nerd who gave Trump a chance.

But 538 were the only ones who rated muh electoral collage win with popular vote loss

Either the Patriots win or the Eagles win, it's basically 50/50 retard.

He also said Roy Moore would win Alabama.

This. He was literally giving Trump a one-in-three shot while Reuters, CNN, Ipsos, Qunnipac, etc. all had Hillary at 90-95%

You're remembering it wrong, he gave him like 1% or something.
Are you sure? Got a screenshot or something?

I'm a data monkey at a hedge fund and even the rookie quants laugh at this guy. He's the equivalent of a 24/7 lip syncing pop star

Sup Forums is proof that IQ is meaningless

The American education system in full effect.

Nah. He the highest he gave Trump was like 30%, which was reasonable. Hate Silver is pretty accurate. It was HuffPo and NYT that had Trump at 1%.

Its was a 30$ chance everybody else had him at 1% or less.

>You're remembering it wrong, he gave him like 1% or something.
You're thinking of a different poll aggregator.

fivethirtyeight.com/features/the-odds-of-an-electoral-college-popular-vote-split-are-increasing/

Nate Whateverchemicalmakesafartsmellbad

>this many people getting assblasted over a meme

These are pretty reasonable odds.

Specifically gave him 20%, and was openly lambasting those that gave horrendous 1%-2% predictions. He also gave the Eagles 3 points more than Vegas. In both these instances he gave the underdog far more credit than other models.

nate silver never expects black people to vote

meant for

>dunning-kruger newbies laugh at a guy who's been doing it longer than they have
sounds about right

poop makes farts smell bad

fivethirtyeight.com/features/election-update-the-campaign-is-almost-over-and-heres-where-we-stand/

>Thus, while Clinton’s a 76 percent favorite to win the popular vote according to our polls-only forecast, her odds are more tenuous — 64 percent — to win the Electoral College.

Thanks for the red pill, Pepe.

Pretty sure poop smells bad because it's constantly covered in farts

Ask Hillary.

>hillary has a pretty good chance of winning the popular vote
>wins the popular vote
>has a significantly worse chance of winning the electoral college
>loses the electoral college
s-shillver btfo

>it's another Nate Gravel gets btfo and redditors try to damage control episode

He jinxed it. If you value your team or whatever, never let him cast a favorable prediction for it.

hmm you have a point

>PhD holders who are entrusted with millions laugh at a guy who has a BA in Economics, failed at proper journalism and has a glorified blog talking about baseball stats

That sounds more right

...

When is probability a 100% guarantee?

>t-that one d-doesn’t c-count
>if he didn’t give drumpf 99% he is shill!!!!

>PhD holders
any fucking monkey can get a PhD

Apparently Nate Pubes couldn't

the south legislated pi to be 3.14 over libtard whining see no reason we can’t legislate probabilities to be 100% too. keeps things easy

Nate Poop

More like constant proof of the confirmation bias.

Nate Slither

THE CURSE CONTINUES

>Nate Trashcan

if the models cant tell me anything important then whats the point?

>statistics is useless because its not a crystal ball

but what is the point?

Asking Trumpsters and Sup Forums to understand statistics is like asking a golden retriever to negotiate a corporate merger.

For sports its mostly betting and moneyball tactics. Beyond that its more important for risk management and insurance.

>bark bark!
>No you don’t understand!

He literally gave trump a ~30% chance to win, which is one of the most generous probabilities that anyone gave

and about the world series

IT WAS HER TUUUUUUUUUURN

He was literally getting called out by other people in the prediction business who were claiming he was falsely inflating Trump's numbers to get more clicks.

538 also wrote multiple articles and talked repeatedly about how Trump was a normal polling error away from winning, and that if he pulled it off it'd be through a collapse in the rust belt.

I mean, uh, nate gravel b-btfo!

...

Nate gets memed on because of how wrong he was about Trump in the primaries. There isn't a ton of useful data in the primaries so he went into subjective analyst mode and was hilariously wrong for like 9 months straight, even in the face of mounting evidence against his position.

He was pretty accurate in the G.E., called out early that Trump had a chance of winning without winning the popular vote, and got attacked by insane leftists over giving Trump a 30ish% chance. Being closer than most in the general election isn't going to erase his meme status though.

Do you even understand what probabilities are?
this is your brain on Sup Forums

You aren't answering, what's the point of a model that only tells you that the game is 50-50?

I predicted that the Eagles would win based on actual facts, like the team being younger, having more backups, putting pressure on Brady, etc.

>Having that hair
KEK

Beauty encapsulated

>Do you even understand what probabilities are?
>this is your brain on Sup Forums

So if your prediction turned out to be wrong, would you just assume that all predictions ever made are pointless?

the fuck does politics have to do with this?

At least it's based on actual player-to-player analysis. This faggot model just analyses the performance during the year as if that shit mattered. The style of play is so much more important, always has been. That's why betting agencies have to suspend people when they're winning too much.

here it is bros

is he right about anything?

whats next?
>99% doesnt mean 100%
so his website is just pointless?

>Nate Bronze

>Nate Tin

Nate would be broke if he had to put real money on the odds he gives.

you literally know nothing. take your proxy off.

according to 538 readers the site has no purpose, they just like to read it and feel smarter by reading something a "smart" person wrote

see nate was really just writing drivel his fans wanted to read

then
>Nate is correct in 1 election
>HES A GOD
now
>Nate shits the bed all the time now
>guyz they are just probabilities, its ok that upsets keep happening and my probabilities are made up bullshit

You have 99 blue marbles and 1 red marble in a bag, shaken up. You reach in and on your first try you pick red. Does that mean 100% chance?

He did fine this Super Bowl.

Now last year, when the Patriots were predicted to have only a 0.1% chance at one point, that was a statistical failure.

we get it, the website is worthless

its literally redddit trash

Guess education really is for liberal cucks.

Best come from behind story in politics ever! Why can't Canadian politics be this entertaining, we get to choose between a communist/globalist cuck and a Israel fetishist/globalist cuck.

Trump was 30-something % on election day according to 538

It's all the other idiots who insisted his odds were about 1 or 2 percent

The problem is that if he can't be wrong, he can't be right either.

take the communist every time
Israel is poison

Only Huffington Post was so low. I wrote a Matlab code in 15 minutes that got the exact same figures they did (98% Hillary, 1.2% Trump). Unlike real statisticians, HP considered all the states by their individual odds. Only the all female staff there could have been so naive.

at least the communist will tell israel to fuck off

There is no fucking right or wrong in this kind of analysis because it's a measure of a one-time event. If the same exact game was played 1000 times then we'd be able to see if the prediction was correct.

J U S T

>If the same exact game was played 1000 times

but thats impossible?

so the impossible needs to happen... for us to see if nates blog was correct?

i guess when hes right he right, and when hes wrong it doesnt matter as it was just a probability

and you cant really prove if the given odds were accurate in anyway

meh, pats woulda won if not for the russians

he's not right or wrong, he's interpreting data. "this team is statistically slightly more likely to win" does not mean "this team is guaranteed to win"

look at it this way: it's statistically improbable that you'll ever get hit by lightning. but you get hit by lightning anyway. does that mean the people who said it's statistically improbable that you'll be hit by lightning are wrong? no, because millions of other people still go through their whole lives without it happening to them.

you do realize nate got famous because his predictions were accurate right?

recently hes been "wrong" about most things