Nate Silver was Dead on Again its Crazy

[rh287]

I really thought Nate Silver's reputation would be severly damaged this election cycle. I mean like so many others, I never dreamed this President could possibly get re-elected. So I when I looked at Nate Silvers 538 forecasts, I thought no way. Seems he truly is a statistical genius and the only one of countless poll analysts that gets it right everytime...so far. Left wing asshole. Forget exit polls or pundits, Nate Silver is one I'll follopw from now on, no matter how bad it hurts.

 

[Quantum Windbag]

He was not dead on, all he did was run 1 million elections, slightly changing the parameters for each one, and post the results. He would have been just as accurate if Romney had won in a landslide, and you would look just as stupid for thinking he is a genius.

 

[Qantrill]

Can someone please keep Dick Morris out of the FOX newsroom?

 

[Truthmatters]

being correct is a crime to some

 

[candycorn]

I really thought Nate Silver's reputation would be severly damaged this election cycle. I mean like so many others, I never dreamed this President could possibly get re-elected. So I when I looked at Nate Silvers 538 forecasts, I thought no way. Seems he truly is a statistical genius and the only one of countless poll analysts that gets it right everytime...so far. Left wing asshole. Forget exit polls or pundits, Nate Silver is one I'll follopw from now on, no matter how bad it hurts.

I heard today in the last 2 elections he's 99 and 1 in predicting the winner of each state.

 

[Quantum Windbag]

If you dig into his work to see what he actually predicted verses the results you will see what I am talking about. His model predicted the margin of victory in each state, added in a fudge factor of over 10% in some states.

In other words, he fudged his numbers so much that he was all but guaranteed to get it right, and produced a 96% success rate.

 

[rh287]

If you dig into his work to see what he actually predicted verses the results you will see what I am talking about. His model predicted the margin of victory in each state, added in a fudge factor of over 10% in some states.

In other words, he fudged his numbers so much that he was all but guaranteed to get it right, and produced a 96% success rate.

I don't know, he clearly explains most of his methods, some are prioritarty though. I do remember before he started doing politics, he predicted the winner of the next World Series a year ahead ...its was a super long shot...some people made a fortune on that bet.

 

[Liberal]

In other words, he fudged his numbers so much that he was all but guaranteed to get it right, and produced a 96% success rate.

So what you are saying is that his method was correct.

Got it.

 

[Quantum Windbag]

If you dig into his work to see what he actually predicted verses the results you will see what I am talking about. His model predicted the margin of victory in each state, added in a fudge factor of over 10% in some states.

In other words, he fudged his numbers so much that he was all but guaranteed to get it right, and produced a 96% success rate.

I don't know, he clearly explains most of his methods, some are prioritarty though. I do remember before he started doing politics, he predicted the winner of the next World Series a year ahead ...its was a super long shot...some people made a fortune on that bet.

He doesn't make predictions, he uses a complicated algorithm that projects odds. The reason bookies always win is they hedge the bets to they don't pay out as much as they take in, he is a political bookie, and cannot lose as a result.

 

[Quantum Windbag]

In other words, he fudged his numbers so much that he was all but guaranteed to get it right, and produced a 96% success rate.

So what you are saying is that his method was correct.

Got it.

Only if you think statistics lets you change the odds of being wrong whenever you want.

 

[NYcarbineer]

Can someone please keep Dick Morris out of the FOX newsroom?

Don't conservatives get sick of being lied to by Foxnews?

 

[NYcarbineer]

The election was easy to call.

Realclear politic's final state by state averages, aka the electoral map with no toss-ups, called it 303 - 235 Obama. They only missed by one state assuming Florida goes to Obama.

btw, that same final map called it for Obama in 2008, within a couple states.

 

[g5000]

If you dig into his work to see what he actually predicted verses the results you will see what I am talking about. His model predicted the margin of victory in each state, added in a fudge factor of over 10% in some states.

In other words, he fudged his numbers so much that he was all but guaranteed to get it right, and produced a 96% success rate.

Providing odds is not "fudging" numbers. There is a big difference between saying "this guy is definitely going to win" and "the odds favor this guy". The definite person is the one "fudging" their numbers.


You want to see people who got their odds all wrong so you can see the difference between real science and confirmation bias?

Here: Just How Bad Was It?

.

 

[rh287]

The election was easy to call.

Realclear politic's final state by state averages, aka the electoral map with no toss-ups, called it 303 - 235 Obama. They only missed by one state assuming Florida goes to Obama.

btw, that same final map called it for Obama in 2008, within a couple states.

No, this election was not easy to call. Most professional pundents, left and right leaning got their asses creamed on this election. Many, Many people, (me included), were calling Nate Silver a fucking fool because he was adamant that he was right, like it or not. He even did a TV interview about it and from what I hear, he hates interviews.

 

[Quantum Windbag]

If you dig into his work to see what he actually predicted verses the results you will see what I am talking about. His model predicted the margin of victory in each state, added in a fudge factor of over 10% in some states.

In other words, he fudged his numbers so much that he was all but guaranteed to get it right, and produced a 96% success rate.

Providing odds is not "fudging" numbers. There is a big difference between saying "this guy is definitely going to win" and "the odds favor this guy". The definite person is the one "fudging" their numbers.


You want to see people who got their odds all wrong so you can see the difference between real science and confirmation bias?

Here: Just How Bad Was It?

.

I did not say providing odds is fudging numbers, I said that using a different projected margin of error for every single state is fudging numbers. Maybe you should stop pretending I am arguing from partisan politics and learn a little about statistics so we can discuss the actual problems I have with his work.

 

[rh287]

He was not dead on, all he did was run 1 million elections, slightly changing the parameters for each one, and post the results. He would have been just as accurate if Romney had won in a landslide, and you would look just as stupid for thinking he is a genius.

Man fuck you. I really don't care for Nate Silver, he's a fucking bleeding liberal. But I give credit where credit is due. The fucking guy is more accurate than anyone else in the business. He developed PECTA and sold it for a mint. He sold his blog to the NY Times. Yeah, I like a lot more people after this election think the fucker is a genius. Asshole!

 

[Mr. Shaman]

Can someone please keep Dick Morris out of the FOX newsroom?

C'mon....c'mon......he's still good for......​

....a LAUGH!!!!!