NYT Nate Silver Doubles Down: 75% Chance Obama Wins

Reading back through the thread and noticed this:

I noticed jopatmc in his farewell address never did give props to Nate Silver.

 

Delusional even 'til the end.

 

that's my point to Commodore. Only PPP deserves some credit.

and I'm sure he don't fully understand nate's methodology because weighted averaging is just the 1st of 7 steps in it.

 

I think you're being way too hard on the polls. No poll had enough of a sample size to actually predict a swing state beyond its margin of error. There's simply not enough money to do so. I don't think it's fair to penalize a poll because it was on the wrong side of the margin of error for its final pre-election poll.

It really depends on how you look at it. I mean, without accurate polling, Nate would be looking like an idiot right now. So he looks smart because all the polls used a pretty representative sample of the voting population. And it's certainly a credit to the polling companies that they're all incredibly good at finding a random sample without bias. In fact, any type of prediction has to be based on this work.

And to be honest, even if one were to simply do step 1 and just use weighted average of all the polls, that person would've gotten the election pretty much right. So in that sense, Step 1 is the most important for the final tally.

If anything, Nate's work was much more important 2 weeks or a month ago, since poll numbers back then are much less indicative of voting results without proper adjustments.

 

I am willing to give the pollsters which got the states correct partial credit.
Why should I give credit to a pollster who got the state results wrong?

I give RCP credit for averaging the polls and getting a pretty close picture of the EV map.

I give Nate credit for averaging the polls and getting a pretty close picture of the EV map and putting % chance on his prediction.

 

I agree with the latter point, but just looking at who won is only a small portion of his model's outputs. He also helps identify which states are most useful to focus on, looks at demographic trends which give states a built in margin for one party or the other (predicting how undecideds will break in a particular state), etc. that's all stuff that polls don't provide at all.

 

Just looked back briefly at the 2008 threads. Much the same dynamic. Here's a gem of a dialogue:

http://bbs.clutchfans.net/showthread.php?t=155854&highlight=obama+mccain+2008&page=3

 

Give the guy a little credit...

 

Wow. How fitting was that? He went out with the same arrogance, self righteousness, and delusions of grandeur as Rush, Rove and company. Another victim chewed up and spit out by radical right propaganda. They get you to vote against your own self interests, then leave you holding these bags of festering hatred for people who support policies that actually help you and your family.

<iframe width="420" height="315" src="http://www.youtube.com/embed/TXlvFpf56vU" frameborder="0" allowfullscreen></iframe>

 

For all you Nate Silver guys, he's having a live chat on deadspin right now

 

http://deadspin.com/5960364/skew-yourselves-nate-silver-is-here-to-answer-your-questions

 

<iframe width="560" height="315" src="http://www.youtube.com/embed/xUHZHmXlEqM" frameborder="0" allowfullscreen></iframe>

Wanted to put it in the hangout, but figured it would be safer in here. Some elbow drops.

Hour long podcast on grantland as well. I haven't had the time to listen to the full interview.

 

was nate silver right?

 

We need a top 10 list of completely wrong election threads of the year.

Mojoman or jopatmc must certainly be the winner.