If you can read this, either the style sheet didn't load or you have an older browser that doesn't support style sheets. Try clearing your browser cache and refreshing the page.

(Mashable)   Unskew this: Nate Silver's algorithms got 50 out of 50 states correct in 2012   (mashable.com) divider line 117
    More: Obvious, Nate Silver, nate, obama, big data, coin toss, electoral colleges, highest mountains  
•       •       •

3986 clicks; posted to Politics » on 07 Nov 2012 at 10:06 AM (1 year ago)   |  Favorite    |   share:  Share on Twitter share via Email Share on Facebook   more»



117 Comments   (+0 »)
   
View Voting Results: Smartest and Funniest

Archived thread

First | « | 1 | 2 | 3 | » | Last | Show all
 
2012-11-07 07:01:25 AM  
(cue Dick Morris)....."but, but, but, but it could have just as easily gone the other way and all 50 would have been found wrong. Amiright or amiright?"
"those are some nice shoes by the way" 
 
2012-11-07 07:51:57 AM  
I haven't seen Florida called yet, but I think this is the correct conclusion.
 
2012-11-07 08:26:18 AM  
I love that the "is Nate Silver a wizard?" website went from 'don't be stupid, it's math' to 'we can't rule out the possibility' last night.
 
2012-11-07 09:16:38 AM  
I was told there wouldbe no math.
 
2012-11-07 09:18:02 AM  
If I am reading his projections right, I think he only got 1 thing wrong - the ND Senate race.
 
2012-11-07 09:29:22 AM  

ManateeGag: I haven't seen Florida called yet, but I think this is the correct conclusion.


He has Obama "winning" at 50.2% or so. It'd hardly be "wrong" if Romney got the state.
 
2012-11-07 09:48:54 AM  
Political sabermetrics FTW.

So I guess the entire RW blogosphere including more than a few Fark TrollsIndependentsTM owe Nate a huge apology.

They'll be along any minute now I'm sure.

*looks at watch* 

Any second now.
 
2012-11-07 09:59:59 AM  
As usual science trumps "but, my feeeeeeeeeeeeeeeelings."
 
2012-11-07 10:06:59 AM  
Latest polling suggests a 98.5% chance that Nate Silver is in fact a wizard.
 
2012-11-07 10:09:44 AM  

DamnYankees: If I am reading his projections right, I think he only got 1 thing wrong - the ND Senate race.


He might've gotten the Montana senate race wrong as well. Not sure if its been called yet.
 
2012-11-07 10:10:11 AM  

DamnYankees: If I am reading his projections right, I think he only got 1 thing wrong - the ND Senate race.


Montana Senate race as well. Nate had Tester losing.

So that's 1 national poll, 50 state polls and 30 senate races in the correct column and 2 Senate races in the wrong column.

81 out of 83 (97.59%), not bad.
 
2012-11-07 10:10:44 AM  
In all seriousness, it's going to be extremely fascinating how this affects elections down the road. Will the GOP embrace solid mathematical predictions or will they just keep fleeing in the other direction? If the public buys into this "The math geek knows all" thing, is it going to have noticeable repercussions on voter turnout?
 
2012-11-07 10:11:17 AM  
DamnYankees

Looks like he projected Montana wrong too at this point.

Ironic that after all the Republican gnashing of teeth, his only two misses were senate seats which he had going Republican going Democratic instead.
 
2012-11-07 10:11:43 AM  

max_pooper: DamnYankees: If I am reading his projections right, I think he only got 1 thing wrong - the ND Senate race.

Montana Senate race as well. Nate had Tester losing.

So that's 1 national poll, 50 state polls and 30 senate races in the correct column and 2 Senate races in the wrong column.

81 out of 83 (97.59%), not bad.


To add more butthurt to the Republicans: If Nate Silver gets those Senate races "wrong," he was actually "wrong" in favor of the Republicans.
 
2012-11-07 10:12:05 AM  

quatchi: Political sabermetrics FTW.


I haven't followed 538 other than catching references to it on Fark, and I don't come down this stinking hallway of the P-tab too often. However, since Silver describes himself as a liberal libertarian and not a Democrat, would it be safe to conclude that:
1. Nate Silver (and Drew Linzer, et al) did well because they took a wide body of data and followed it to a conclusion
2. Partisan Analysts failed because they started with their conclusion and looked for friendly data

Or, 3. Nate Silver is a witch who owes us all a televised appearance at the duck pond.
 
2012-11-07 10:12:42 AM  

birdboy2000: DamnYankees

Looks like he projected Montana wrong too at this point.

Ironic that after all the Republican gnashing of teeth, his only two misses were senate seats which he had going Republican going Democratic instead.


Damn it. 26 seconds.
 
2012-11-07 10:12:59 AM  
While impressive, I would also like to point out that there were "others" who did similar or better job than him as well.

It is a win for Nate and I will be gloating for him, but at the end of the day, it is actually a win for math and science.
 
2012-11-07 10:13:13 AM  
i.imgur.com
 
2012-11-07 10:13:52 AM  

max_pooper: DamnYankees: If I am reading his projections right, I think he only got 1 thing wrong - the ND Senate race.

Montana Senate race as well. Nate had Tester losing.

So that's 1 national poll, 50 state polls and 30 senate races in the correct column and 2 Senate races in the wrong column.

81 out of 83 (97.59%), not bad.


It would have been more surprising had he got them all right. He made a few predictions with only 60-80% chance of a person winning.
 
2012-11-07 10:14:40 AM  
Even if FL goes red, his model had it at almost 50% which basically means "I think it can go either way". SO his model would be still right.
 
2012-11-07 10:14:50 AM  
Well, "my gut" got had a 120% success rate +/- 120%.
 
2012-11-07 10:15:13 AM  

jayhawk88: In all seriousness, it's going to be extremely fascinating how this affects elections down the road. Will the GOP embrace solid mathematical predictions or will they just keep fleeing in the other direction? If the public buys into this "The math geek knows all" thing, is it going to have noticeable repercussions on voter turnout?


Maybe we will see a growing interest in mathematics among young people, after society tells them that studying it will make them a wizard.
 
2012-11-07 10:15:44 AM  

jayhawk88: In all seriousness, it's going to be extremely fascinating how this affects elections down the road. Will the GOP embrace solid mathematical predictions or will they just keep fleeing in the other direction? If the public buys into this "The math geek knows all" thing, is it going to have noticeable repercussions on voter turnout?


if you think the GOP campaign has been ignoring him, you're insane. they just had to publicly denounce him this cycle because they were fighting such an uphill battle. but if you look at where they spent time and money, it was right in line with where the numbers said they had to.
 
2012-11-07 10:15:57 AM  

jayhawk88: In all seriousness, it's going to be extremely fascinating how this affects elections down the road. Will the GOP embrace solid mathematical predictions or will they just keep fleeing in the other direction? If the public buys into this "The math geek knows all" thing, is it going to have noticeable repercussions on voter turnout?


There is now, more than ever, WAY too much money on the line for them to ignore the methodology. These billionaires dumping 8-figure checks into SuperPACs aren't foolish and didn't get where they are by accepting sh*tty ROIs. They'll not be so kind to having their egos assuaged while in reality their money is being dumped into a furnace. That's the difference between the traditional, party money and dark money.

His prediction models aren't just a betting line, they have the potential for being very powerful tools to focus campaigning, message/platform moves, et cetera.
 
2012-11-07 10:16:51 AM  
I thought he was a little high in some of his predictions. Not that I thought Romney would win by any means, but I thought he was a little too generous especially in Virginia, but....

Can't argue with results. That settles it. He's a witch.
 
2012-11-07 10:16:58 AM  

jayhawk88: In all seriousness, it's going to be extremely fascinating how this affects elections down the road. Will the GOP embrace solid mathematical predictions or will they just keep fleeing in the other direction? If the public buys into this "The math geek knows all" thing, is it going to have noticeable repercussions on voter turnout?


I think it's going to take much harder beatings before the GOP tosses the "virtuous ignorance" camp under the bus and accepts that intellectuals-who-know-things should be in the driver's seat.
 
2012-11-07 10:17:05 AM  
What's even more impressive is Sam Wang - he once again hit the EC exactly. That means in the last three Presidential elections, he has been off by a grand total of one electoral vote.
 
2012-11-07 10:18:21 AM  
Considering anyone who used state level polling model pretty much nailed it, if any "credit" should be given to predictions it should be the pollster collective who do a pretty damn remarkable job of projecting what the electorate will look like, particularly on the state level.

If there were any biases, looks to me they were maybe ever so slightly biased against Obama/Democrats. My guess is they're still slightly undersampling first time latino voters and cellphone only/no landline demos.
 
2012-11-07 10:19:00 AM  
For the Left... Math = Math.
For the Right... Math = Magic.
 
2012-11-07 10:19:06 AM  
 
2012-11-07 10:19:09 AM  

47 is the new 42: max_pooper: DamnYankees: If I am reading his projections right, I think he only got 1 thing wrong - the ND Senate race.

Montana Senate race as well. Nate had Tester losing.

So that's 1 national poll, 50 state polls and 30 senate races in the correct column and 2 Senate races in the wrong column.

81 out of 83 (97.59%), not bad.

To add more butthurt to the Republicans: If Nate Silver gets those Senate races "wrong," he was actually "wrong" in favor of the Republicans.


If I am not mistaken, he in whatever cases he has been wrong, he has been wrong in favor of republicans.

/Given he has a liberal bias in him as a person, it would only make sense that he will try to over correct himself in the model.
 
2012-11-07 10:20:15 AM  
imageshack.us
 
2012-11-07 10:20:38 AM  

DamnYankees: If I am reading his projections right, I think he only got 1 thing wrong - the ND Senate race.


Every newspaper endorsed the Republican candidate (Berg), there wasn't a single independent poll that put him behind, Romney won the state by 20 points, the minimum margin of victory for Republicans in statewide elections outside of the Senate rate was 13 points, and Berg still lost. I'm actually looking forward to seeing how this election gets interpreted, because in ND it's going down as one of the biggest upsets in state history.
 
2012-11-07 10:23:04 AM  
I believe there should be a mea culpa thread where all of those posters predicting a Romney win or questioning Nate Silver's methods can come and admit to being either a) ignorant of how statistics work, b) partisan wishful thinkers c) trolls. Anyone who comes in a self-shames can be forgiven their posting sins. Anyone who doesn't show up in the thread can be shamed by others.
 
2012-11-07 10:24:32 AM  

factoryconnection: These billionaires dumping 8-figure checks into SuperPACs aren't foolish and didn't get where they are by accepting sh*tty ROIs. They'll not be so kind to having their egos assuaged while in reality their money is being dumped into a furnace.


Rove must have been downing bourbon and Tums this morning before his Crossroads conference call. He made a lot of promises he failed to deliver on.
 
2012-11-07 10:25:15 AM  

thomps: jayhawk88: In all seriousness, it's going to be extremely fascinating how this affects elections down the road. Will the GOP embrace solid mathematical predictions or will they just keep fleeing in the other direction? If the public buys into this "The math geek knows all" thing, is it going to have noticeable repercussions on voter turnout?

if you think the GOP campaign has been ignoring him, you're insane. they just had to publicly denounce him this cycle because they were fighting such an uphill battle. but if you look at where they spent time and money, it was right in line with where the numbers said they had to.


Well I guess the "publicly" part is what I meant. To me, it doesn't make much sense in trying to rally the base/get out the vote by telling them that your man is going to for sure win by 30-50 EV's. Wouldn't it have been better to point to the 538 results and say "We're in real trouble here folks, you need to get yourselves motivating"?
 
2012-11-07 10:25:54 AM  
He's clearly a smart guy but he's far from the only one who made basically the same call with similar methodology. Is he a wizard because he's the best known? I think he's a little overrated. Still, well played.
 
2012-11-07 10:26:17 AM  
media2.fanup.com
 
2012-11-07 10:28:15 AM  
imgs.xkcd.com
 
2012-11-07 10:29:17 AM  

you have pee hands: He's clearly a smart guy but he's far from the only one who made basically the same call with similar methodology. Is he a wizard because he's the best known? I think he's a little overrated. Still, well played.


He became the face of the methodology and took all the hits from the right wing punditocracy. Now that mathematics has defeated superstition yet again, he should take a very public victory lap.
 
2012-11-07 10:31:06 AM  
When it comes to mathematical probability modeling, Nate Silver is Hari Seldon.
 
2012-11-07 10:31:32 AM  

HotWingConspiracy: Rove must have been downing bourbon and Tums this morning before his Crossroads conference call. He made a lot of promises he failed to deliver on.


I was pondering something similar: Mitt Romney, capping 6 solid years of presidential campaigning with a disappointment going late into the night can't have a got-damn stiff drink nor even a cup of coffee.

I guess he'll just have to fall back on being an unemployed grandfather that rakes in ~$15M/year and get the little kids that miniature giraffe they've been begging for.

you have pee hands: He's clearly a smart guy but he's far from the only one who made basically the same call with similar methodology. Is he a wizard because he's the best known? I think he's a little overrated. Still, well played.


Wizard, not Highlander... there can be more than one.
 
2012-11-07 10:31:39 AM  
From Rasmussan Reports' website this morning....


Last night's results raise significant questions about whether the Republican establishment in Washington is relevant. Nearly four years ago, Scott Rasmussen wrote that the Republican Party would "sink further into irrelevancy as long as its key players insist on hanging around Congress or K Street for their ideas."

This race was very likely the last presidential election of the telephone polling era. While the industry did an excellent job of projecting last night's election, entirely new techniques will need to be developed before 2016. The central issue is that phone polling worked for decades because that was how people communicated. In the 21st century, that is no longer true.
 
2012-11-07 10:32:58 AM  

DamnYankees: If I am reading his projections right, I think he only got 1 thing wrong - the ND Senate race.


Like I said. He is a complete failure.
 
2012-11-07 10:33:56 AM  

jayhawk88: thomps: jayhawk88: In all seriousness, it's going to be extremely fascinating how this affects elections down the road. Will the GOP embrace solid mathematical predictions or will they just keep fleeing in the other direction? If the public buys into this "The math geek knows all" thing, is it going to have noticeable repercussions on voter turnout?

if you think the GOP campaign has been ignoring him, you're insane. they just had to publicly denounce him this cycle because they were fighting such an uphill battle. but if you look at where they spent time and money, it was right in line with where the numbers said they had to.

Well I guess the "publicly" part is what I meant. To me, it doesn't make much sense in trying to rally the base/get out the vote by telling them that your man is going to for sure win by 30-50 EV's. Wouldn't it have been better to point to the 538 results and say "We're in real trouble here folks, you need to get yourselves motivating"?


maybe, but it's hard to get anyone excited enough to contribute to a campaign that has a 70-80% probability of losing. the "it's neck and neck" narrative is much better for rallying the troops. the "it's going to be a romney win by a landslide" predictions, on the other hand, were just moronic.
 
2012-11-07 10:34:17 AM  

Jairzinho: Even if FL goes red, his model had it at almost 50% which basically means "I think it can go either way". SO his model would be still right.


Any thing he got with 2% is basically a correct call.
 
2012-11-07 10:34:22 AM  

factoryconnection: I was pondering something similar: Mitt Romney, capping 6 solid years of presidential campaigning with a disappointment going late into the night can't have a got-damn stiff drink nor even a cup of coffee.


Romney is the least of Rove's worries. That was always a long shot. A year ago he was saying the GOP had a better than 50/50 odds of flipping the Senate, and instead the Democrats gained seats. They also said they expected to flip about 6 Democratic State Governover races, and instead only got 1.

A billion dollars don't buy what it used to.
 
2012-11-07 10:34:43 AM  

Rann Xerox: When it comes to mathematical probability modeling, Nate Silver is Hari Seldon.


isn't hari seldon the hari seldon of mathematical probability modeling?
 
2012-11-07 10:38:15 AM  

Mr. Coffee Nerves: As usual science trumps "but, my feeeeeeeeeeeeeeeelings."


Dear Pundits,

Fark your feelings.

Love,

SCIENCE!
 
2012-11-07 10:39:40 AM  

imontheinternet: He became the face of the methodology and took all the hits from the right wing punditocracy. Now that mathematics has defeated superstition yet again, he should take a very public victory lap.


He's hasn't defeated superstition just because he was right and they were wrong. If there were another election tomorrow they'd be right back to touting their preconceived notions in the face of mathematical probability - these are literally people who accept the existence of microevolution but think Darwinism is a tool of the devil. The cognitive dissonance is astounding. He's welcome to his victory lap, though, he did make good calls. I just feel like people are making more of this than what it really was. I'm probably just being cynical and contrarian.
 
Displayed 50 of 117 comments

First | « | 1 | 2 | 3 | » | Last | Show all

View Voting Results: Smartest and Funniest


This thread is archived, and closed to new comments.

Continue Farking
Submit a Link »






Report