NYT Nate Silver Doubles Down: 75% Chance Obama Wins

Why is that so difficult to understand?

Silver gives Obama an 80% chance of winning Ohio, and therefore, essentially an 80% chance of winning the election. He gives Romney a 20% chance of winning Ohio, and essentially a 20% chance of winning the election. Yes, "if Romney takes one state," Obama's odds go down, but the chances of Romney taking that one state are low themselves. That's not an out. That's statistics.

It's the equivalent of the Rockets having James Harden at the free throw line with 1 second left, down by 1. He is a good free throw shooter, and has an 80% chance of making them. If so, the Rockets have a 97% chance of winning the game, barring a last second miracle for the other team. If he misses them, the other team has a 97% chance of winning the game barring a last second miracle for the Rockets. Yes, either could happen, but the odds are in the Rockets' favor when you look at Harden's free throw percentage.

Now, I get that you are arguing Silver's model is wrong, and the shooter the Rockets have at the free throw line is not nearly as good a free throw shooter as Harden. (Since this is a make believe scenario not based in reality, and we want a poor free throw shooter for this example, let's just use... oh, I don't know... Dwight Howard.) Now the Rockets need two free throws from Dwight Howard to win the game. If he makes them both, they almost certainly will win. If he doesn't, they will almost certainly lose. But since he is such a terrible free throw shooter, his odds of making the free throws and the Rockets' odds of winning both go down together. They are directly related.

Is Silver's model correct or wrong? That's up for people to debate. But comments like the one above claiming that Silver has created some sort of "out" show a general inclination to avoid actual analysis and understanding.

 

Looks like someone's daughter is going to fail a class.

 

Where is that "unlikely" image?

 

That would be a terrible class.

No, you've consistently demonstrated otherwise. You don't understand inputs vs outputs or anything about poll methodology.

 

Spin it however you want. Nate is wagering Obama will win. Call it probabilities. He's flat-out wrong. 90% of the data going into his model is based in 2008 election turnout, which isn't going to happen...and he hasn't adjusted for that. That is why his probabilities are dead wrong. The true mathematical probability of this election right now is somewhere between 80-100% Romney. How can I say that? Because if you unskew the polls and take into account voter turnout during different times in this nation's history, during times of great financial crisis, you will see the worm turns. Nate is not accounting for skew. If he were, he'd have it at least at 50/50.

It will be interesting how he spins everything next Wednesday.

 

that's only the 1st of 7 steps.. can you read this? http://fivethirtyeight.blogs.nytimes.com/methodology/

can you list all the polls that nate is using and show which of them are extremely left leaning?

 

No, it's not. Example #1 of you not understanding even the basics of how polls work.

 

GIGOty
GIGOty
GOO

 

FIFY!

 

prove it

 

Yes but I'm not going to.

GO

DO

YOUR

OWN

HOMEWORK






or remain willfully ignorant. I don't care. But I'm not gonna sit here an spoon feed you and change your statistical diaper.

 

That's just you not understanding "very advanced mathematics."

 

so you don't have it thanks

 

What exactly was spin in my post? I said people can disagree on Silver's model's accuracy, but that doesn't change the fact that his conclusions are based on his model, and he is not creating an "out" for himself like you claim he is.

Oh wait, you didn't really read it.

 

jopatmac, two bits of constructive criticism:

1. if Obama wins, you've made yourself look like a huge douchebag & fool (again)

2. if Romney wins, you've made yourself look like a huge douchebag & fool (again)

Just stop. When did you go so far off the deep end? What happened?

 

This thread is hilarious. Thank you, jopat, I needed the entertainment.

 

Shockingly, jop is pretending that Major didn't just prove that he was blatantly lying about his prediction that Silver would tweak his analysis to reach 50/50 by election day and only added his ridiculous caveat after he was being proven wrong.

 

The poster in question stated yesterday his only objective is to troll the board. He is not interested in honest discussion.

The best course of action is to ignore him until he leaves next week.

 

Problem is... he has already proven he has no integrity, so expecting him to honor his part of the bet and leave is unrealistic.