Comments on: How Nate Silver Predicted Obama’s Win

By: Merinda

Merinda — Fri, 06 Feb 2009 17:09:00 +0000

More and more I keep thinking maybe I should take a statistics class. Maybe Heinlein was right: It all comes down to numbers.

By: Sillstaw

Sillstaw — Fri, 06 Feb 2009 16:26:00 +0000

Yeah, it wasn’t too hard to guess that Obama would be elected. It seemed like everyone was in love with him from early on, and that certainly helped.

In my government class, we did a questionnaire about the incumbent party that, if it got a majority of “no’s,” would mean they would lose. Sure enough, they lost–but only by a slight margin.

All that is fine and well, but to prove it with statistics, and to come close to the actual results? That’s impressive.

reCaptcha: usually here

By: Natalie

Natalie — Fri, 06 Feb 2009 13:54:00 +0000

I can’t believe this. I KNOW Nate Silver– we grew up together– I was friends with his little sister, and my brother was friends with him. I guess I knew he was smart, but I had no idea. I just phoned my mother to tell her about this article and she said, “I always told his mom he’d be president some day.”

By: amiee

amiee — Fri, 06 Feb 2009 13:33:00 +0000

I watched his site all through the election. It was spectacular. Itâ€™s also possible that his own discussion of pollster methodology changed their methodology, though I donâ€™t recall a specific incident of that.

By: Jerad

Jerad — Fri, 06 Feb 2009 05:47:00 +0000

Mike he was off by 1 electoral vote, the nebraska split vote kept him from being perfect.

By: Chris Higgins

Chris Higgins — Fri, 06 Feb 2009 03:37:00 +0000

@Kikadee – yeah, that’s a good question. I think he was on TV a fair bit towards the end of the campaign, so it’s likely that his own quasi-journalism/analyst discussion influenced the electorate. I’m also really curious what his readership was — were they all lefties enjoying the good news? Or did he have bipartisan appeal?

It’s also possible that his own discussion of pollster methodology changed their methodology, though I don’t recall a specific incident of that.

By: Mike

Mike — Thu, 05 Feb 2009 21:47:00 +0000

I watched his site all through the election. It was spectacular. For any nay-sayers out there, he took statistical data from all the major polls, so there was no way he was slanted (toward his own political leanings).

Plus, unless I’m mistaken, didn’t he predict the exact number of electoral votes that Obama actually won? (He may have been off by one or two.) That is very impressive.

Check out the site. It’s really cool.

By: Kikadee

Kikadee — Thu, 05 Feb 2009 21:30:00 +0000

I’d be curious to know if Silver’s political statistical models take into account the influence (if any) of his published results on public opinion and, subsequently, on the polls themselves.

By: Ian

Ian — Thu, 05 Feb 2009 21:15:00 +0000

Not to mention he was picking Obama way before anyone else was picking him to win (when everyone thought Hillary was going to run away with it).

By: Chris Higgins

Chris Higgins — Thu, 05 Feb 2009 20:08:00 +0000

Jeff – it’s the details of the win that mattered. He predicted state by state totals, electoral college votes, and popular vote — all quite accurately.