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.

(Some Guy)   Is Nate Silver a Witch?   (isnatesilverawitch.com) divider line 406
    More: Obvious  
•       •       •

8821 clicks; posted to Politics » on 05 Nov 2012 at 8:38 PM (2 years ago)   |  Favorite    |   share:  Share on Twitter share via Email Share on Facebook   more»



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

Archived thread

First | « | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | » | Last | Show all
 
2012-11-05 10:17:20 PM  

JW: In fact, Romney can still win the election and Nate can be right -- at the same time. It just means that the less-likely-to-represent-the-data scenario occurred.


In fact, Romney can still win the election and Nate can be right -- at the same time


In fact, Romney can still win the election and Nate can be right -- at the same time
 
2012-11-05 10:17:30 PM  
Holy crap. After 8 years, my first greenlight. =D
 
2012-11-05 10:19:32 PM  

LouDobbsAwaaaay: mayIFark: He is making me nervous now. His model just updated for a huge jump for Obama giving him 314EV.

By this time tomorrow, either he is going to be a super hero, or be ruined for ever.

I sure hope for the 1st, for his shake, and for the country.

The thing that really bugs me about Silver's anointed status in political polling is that he's not even doing anything particularly special. It's just a Monte Carlo ensemble system where he runs simulations with random perturbations representing errors in the polls. We could have had every major pollster doing this on their Gateway desktop 12 years ago. What makes Nate Silver so smart is that he's kind of smart, but he exists within the institution of journalism - the biggest pack of inbred, mouth-breathing yokels this side of a Toby Keith concert. It's the degree of choice for the person who is a little too intimidated to get a degree in physical education. Sarah Palin managed to scrape together a degree in broadcast journalism after failing out of half a dozen community colleges. It's people too dumb to be trusted to do anything else dumbing-down the day's news for people even stupider than they are. Journalism is a shrine to intellectual in-curiosity. So much so that some kid who cut his teeth on poker and baseball stats spends a three-day weekend tooling his spreadsheet program to accept political polling data, and his analysis is so much better than the norm that the entire institution of journalism doesn't have the slightest idea what to do about it anymore.


Except Silver didn't come from journalism. He ran his site independently as a labor of statistical love. NYTimes was just first to recognize (and have the bucks to buy) Silver's expertise.
 
2012-11-05 10:19:42 PM  

Metalupis: JW: mayIFark: I am guessing this is Nate's last and final call before the election: and he is basically called FL for Obama. In fact, he is called all the swing states for Obama besides NC.

/I can't take it anymore. I know, we should win, but the possibility of a Romney presidency scares me so much that I can't relax. The stake is way too high.

This is a misunderstanding of his analysis. He's not necessarily "predicting". He didn't "call" anything for anyone (well, not specifically, that is). He's given Obama a 52% chance of winning Florida, if all of the assumptions in his models hold true and there are no errors in the data. He's not a soothsayer or a prognosticator, he's just an analyst who built a model (well, a bunch of models).

Even in that scenario of perfect data and perfect assumptions, it's still essentially a 50/50 for Florida. In fact, Romney can still win the election and Nate can be right -- at the same time. It just means that the less-likely-to-represent-the-data scenario occurred.

this is what a lot of people fail to realize about his statistics and percentages in general


Still, if he's putting Ohio in the "Safe Obama" category. That really narrows the possibilities for Romney to win.
 
2012-11-05 10:19:52 PM  

tinyarena: 200 comments and not one Connie Booth pic?
Fark is full of old men tonight
Here,
[img706.imageshack.us image 324x323]
oh, and the election thing is interesting too, I guess


tomorrow, there is going to be a massive amount of butthurt displayed on internets regardless of who wins. yes, there will still be attractive women on there too, but the fark politics tab is probably not the best place to look for the women. feel free to continue posting them if that is your thing, but the butthurt is the most likely thing you will see here tomorrow
 
2012-11-05 10:19:59 PM  

RevMercutio: Holy crap. After 8 years, my first greenlight. =D


Gratz!
 
2012-11-05 10:20:35 PM  

Biological Ali: SweetBluebonnet: Nate is going to have a lot of explaining to do if Obama doesn't win.

Not really. Let's say a guy has a coin and declares "I'm going to flip this three times and get three heads!", and you try let him know that he only has a one-in-eight chance of that happening. If he flips the coin and gets three heads anyway (which is very possible - just not likely), would you "have a lot of explaining to do"?


Not to anyone who understands probability...but remember that most people don't. Most people think that if you flip a coin three times, you have only a 50-50 chance of it coming up heads any time, AND that the odds of it coming up heads three times in a row are actually astronomical because a flipped coin must come up equally heads or tails. In other words, in the mind of the average person, if a coin comes up heads once, then it must come up tails next time because statistically that's how it works, right? (wrong)

So Nate Silver saying "Obama has an 80% chance of winning, and Romney has a 20% chance of winning" (or whatever the current probability is) in the average person's mind means that Obama is GOING TO WIN and the odds of Romney winning are almost off the charts. Now, you and I know that what that means is that in Scenario A, Obama will win and in Scenario B Romney will win, and A is turning up 8 times out of 10; but that's not how the average person understands it. So yeah, if Obama loses, Silver is going to look like an ass to the hoipolloi.
 
2012-11-05 10:20:49 PM  

RevMercutio: Holy crap. After 8 years, my first greenlight. =D


And 8 years ago Bush won. THANKS A LOT ASSHOLE!

/and congrats on the greenie
 
2012-11-05 10:21:17 PM  
Oh man! Nate is on a one man war against the entire Media and a Major political party of USA. He is calling Ohio safe Obama.

Took some balls.
 
2012-11-05 10:21:18 PM  

nekom: I also just noticed, I feel he is overstating Romney's chances in North Carolina. Most of the the other polls I've seen have that a lot closer, yet he has it 75% Romney. So I guess it seems to me that he's overstating just a bit for both sides in various states.


Eh? Looking at the polls for the last 2 weeks, Romney is ahead by .05-2.0 points. Given that this generally falls within the margin of error, the reason it's at 75% is because even though they're in the margin of error, they've been fairly precise in favor of Romney by one amount or another. Granted, Silver is admittedly highly cautious in terms of this, but it's not unreasonable to give Romney a 75% chance in NC.
 
2012-11-05 10:24:07 PM  

vygramul: Except Silver didn't come from journalism. He ran his site independently as a labor of statistical love. NYTimes was just first to recognize (and have the bucks to buy) Silver's expertise.


That's kind of my point. Outside of journalism Silver is a stat-junkie in a sea of stat-junkies. Inside journalism, he's a mythological figure using math-voodoo to destroy the concept of free will. His prediction skills didn't increase appreciably when he started applying them to political polling. He just ended up in a field where the next best guy at the job has a degree in journalism and a nice haircut instead of the least bit of knowledge or experience with what he was talking about.
 
2012-11-05 10:24:23 PM  
This...needed to be said?
 
2012-11-05 10:25:40 PM  

Blue_Blazer: There's also the fact that Floridians, for all their faults, know a good handling of a hurricane when they see it.


Assuming that handling is in accord with Foxslow's heirachy of needs:

1) Free hand crank AM radio to hear Rush's show.
2) Power restored to watch Fox News.
3) Less government regulations so that the gasoline/water/food in disaster areas is priced to what the market will bear.
4) All available Search and Rescue teams must be diverted to save Fluffy, your 18 year old miniature poodle.
5) Demanding reparations from FEMA.
6) Your neighbor heeded the evacuation notice but you didn't. Therefore, all their base belong to you.
7) Your right to BBQ may not be superseded by Federal or State government order because severe gas leaks in the area.
 
2012-11-05 10:25:49 PM  

Jairzinho: JW: In fact, Romney can still win the election and Nate can be right -- at the same time. It just means that the less-likely-to-represent-the-data scenario occurred.

In fact, Romney can still win the election and Nate can be right -- at the same time

In fact, Romney can still win the election and Nate can be right -- at the same time


When you put somebody at 92%, I think not plausibly so. He's all-in - rightly so, imho, but nonethelelss.
 
2012-11-05 10:26:51 PM  
People making a big deal about FL becoming barely blue now in Nate's model do not get it. There is still no much difference. It's still a tossup.

It's like if your doctor said yesterday "You have 48% chances of survive this cancer" and you "OMG I'M GONNA DIE!!11!!!"
then the next day your doctor says "You have now 52% chances of survive this cancer" and now you "OMFG!!! I'M GONNA LIVE!!!!1111!!!! LET"S PARTY!!!!11!!!" 
 
2012-11-05 10:27:05 PM  

Jairzinho: JW: In fact, Romney can still win the election and Nate can be right -- at the same time. It just means that the less-likely-to-represent-the-data scenario occurred.

In fact, Romney can still win the election and Nate can be right -- at the same time

In fact, Romney can still win the election and Nate can be right -- at the same time


I think "right" is a stretch at 91.4 wtf 92.2% now. I think it's safely gone from probability to prediction.

If Romney wins tommorow at less than 8%, I will certainly call him wrong. if the odds were 25% I'd say otherwise.
 
2012-11-05 10:27:19 PM  
Updated again: forecast and nowcast are the same. GAME ON!!
 
2012-11-05 10:27:25 PM  
He's calling Florida about 50/50, but it's a very slight blue on the map. If he goes 50 out of 50 tomorrow the media's games with these elections will be changed pretty much forever.
 
2012-11-05 10:27:34 PM  

Paul Baumer: Jairzinho: JW: In fact, Romney can still win the election and Nate can be right -- at the same time. It just means that the less-likely-to-represent-the-data scenario occurred.

In fact, Romney can still win the election and Nate can be right -- at the same time

In fact, Romney can still win the election and Nate can be right -- at the same time

When you put somebody at 92%, I think not plausibly so. He's all-in - rightly so, imho, but nonethelelss.


...no, it's entirely plausible for Romney to win, even with only an 8% chance.
 
2012-11-05 10:28:07 PM  

nyseattitude: Nate is just running a "numbers racket"  according to this scholar.


YOOUUUUUUUU, made me click on that farktard's link.
 
2012-11-05 10:28:14 PM  

Modguy: Paul Baumer: Jairzinho: JW: In fact, Romney can still win the election and Nate can be right -- at the same time. It just means that the less-likely-to-represent-the-data scenario occurred.

In fact, Romney can still win the election and Nate can be right -- at the same time

In fact, Romney can still win the election and Nate can be right -- at the same time

When you put somebody at 92%, I think not plausibly so. He's all-in - rightly so, imho, but nonethelelss.

...no, it's entirely plausible for Romney to win, even with only an 8% chance.


Not incredibly likely, but it's certainly not off the table.
 
2012-11-05 10:28:24 PM  
Looks like Nate just updated again.

President Obama is now showing 315.3 EV and a 92.2 chance of winning.
 
2012-11-05 10:28:39 PM  

Blue_Blazer: Updated again: forecast and nowcast are the same. GAME ON!!


Yes indeed. Hard to believe the day has come.
 
2012-11-05 10:28:50 PM  

mayIFark: /I can't take it anymore. I know, we should win, but the possibility of a Romney presidency scares me so much that I can't relax. The stake is way too high.


Me too. My usual remedy is booze, but I have blood work being drawn tomorrow morning for my annual physical, so must forgo.

Speaking of which, since I'll be on base I plan to hit the Package Store for some celebratory hooch. Anybody have suggestions? Scotch is a family fav, but strikes me as too heavy for all evening drinking. Wine seems too light. Beer too meh.

Vodka? I haven't had any in months. Gin? I bet I haven't had any gin in several years. Ideas, people...I need ideas!
 
2012-11-05 10:30:33 PM  

mak3_7up_y0urs: Looks like Nate just updated again.

President Obama is now showing 315.3 EV and a 92.2 chance of winning.


Yep. and the peak on the EV-probability chart is now at 332 EV for Obama, with over 20% probability.
 
2012-11-05 10:30:59 PM  
List of People Conspiring Against the GOP, and therefore, America
(LOPCATGOPATA for short):
Liberals
Democrats
Socialists
Community Organizers
Geologists
Biologists
Meteorologists
Climatologists
Atheists
Muslims
Jews
Satan
ABC
NBC
CNN
CBS
PBS
All of cable news except FNC
The New York Times
The LA Times
The Washington Post
The Associated Press
Reuters
BBC
The Guardian
Black People
Mexicans
Human Rights Activists
SCOTUS
Europe
Movie Industry
Television Industry
Environmentalists
ACLU
The United Nations
Labor Unions
Colleges
Teachers (including kindergarten teachers)
Professors
ACORN
National Endowment for the Arts
Gays
Judges
NPR
Paleontologists
Astrophysicists
Museums (*except Creationism Museum)
WHO
WTO
Inflated tires
The Honolulu Advertiser
The Star Bulletin
Teletubbies
Sponge Bob and Patrick
Nobel Prize Committee
US Census Bureau
NOAA
Sesame Street
Comic Books
Little Green Footballs
Video Games
The Bible
CBO
Bruce Springsteen
Pennies
The Theory of Relativity
Comedy Central
Young People
whatever the hell a Justin Beiber is
Small Business Owners
Math
CPAC
Navy SEALs
The Economist
The Muppets
Iowa Republicans
Low-Flow Toilets
Breast Cancer Screenings
Chrysler
Clint Eastwood.
Robert Deniro
Tom Hanks
Glenn Frey
Norman Rockwell
James Cameron
Dr. Seus
Nuns
Supreme Court Justice John Roberts
Jonathan Krohn at age 17
Fact Checkers
Australia
Mitt Romney
Rasmussen
Fox News
Lockheed Martin
Bureau of Labor Statistics
Paul Ryan
Debate moderators
Ben Stein
Soup kitchens
Chris Christie
Nate Silver
 
2012-11-05 10:31:03 PM  

LouDobbsAwaaaay: vygramul: Except Silver didn't come from journalism. He ran his site independently as a labor of statistical love. NYTimes was just first to recognize (and have the bucks to buy) Silver's expertise.

That's kind of my point. Outside of journalism Silver is a stat-junkie in a sea of stat-junkies. Inside journalism, he's a mythological figure using math-voodoo to destroy the concept of free will. His prediction skills didn't increase appreciably when he started applying them to political polling. He just ended up in a field where the next best guy at the job has a degree in journalism and a nice haircut instead of the least bit of knowledge or experience with what he was talking about.


This. This is the value of specialists who are actually experts, as opposed to people who read a book or took one class calling themselves experts in things like economics or statistics.

I would have liked Nate to point out (and maybe he has and I missed it) that pundits can't talk about the states and election models interchangably. In other words, it is illogical for someone to say "Even if Romney loses Florida, he can win by taking Virginia" (for example. one can substitute Virginia with Pennsylvania for a better example). It is illogical because if Romney loses Florida, he's not going to win Virginia. Or Pennsylvania. It's like in baseball (to take a cue out of Nate's old expertise), how someone who plays shortstop can transition to first base, but not vice versa - it's a spectrum. In this election, the spectrum basically goes - North Carolina, Florida, Colorado, Iowa, Nevada, Virginia, Ohio, New Hampshire. I expect Obama to win all of those but North Carolina, and I think Florida and Colorado are in play.
 
2012-11-05 10:31:23 PM  

rthanu: Jairzinho: JW: In fact, Romney can still win the election and Nate can be right -- at the same time. It just means that the less-likely-to-represent-the-data scenario occurred.

In fact, Romney can still win the election and Nate can be right -- at the same time

In fact, Romney can still win the election and Nate can be right -- at the same time

I think "right" is a stretch at 91.4 wtf 92.2% now. I think it's safely gone from probability to prediction.

If Romney wins tommorow at less than 8%, I will certainly call him wrong. if the odds were 25% I'd say otherwise.


Oh god da--

morbo.jpg
 
2012-11-05 10:31:39 PM  

rthanu: If Romney wins tommorow at less than 8%, I will certainly call him wrong. if the odds were 25% I'd say otherwise.


This is the lack of understanding of probability this country faces. Somehow it's okay for Romney to win with a 25% chance but not an 8% chance? WTF?
 
2012-11-05 10:31:56 PM  

Blue_Blazer: Updated again: forecast and nowcast are the same. GAME ON!!


they're not identical yet- the Now-cast has Obama a tenth of a percent more likely to win than the forecast.
 
2012-11-05 10:32:05 PM  

StoneColdAtheist: mayIFark: /I can't take it anymore. I know, we should win, but the possibility of a Romney presidency scares me so much that I can't relax. The stake is way too high.

Me too. My usual remedy is booze, but I have blood work being drawn tomorrow morning for my annual physical, so must forgo.

Speaking of which, since I'll be on base I plan to hit the Package Store for some celebratory hooch. Anybody have suggestions? Scotch is a family fav, but strikes me as too heavy for all evening drinking. Wine seems too light. Beer too meh.

Vodka? I haven't had any in months. Gin? I bet I haven't had any gin in several years. Ideas, people...I need ideas!


Jager? A nice pina colada maybe? How about Black&Tan in honor of our President? Or perhaps a mean White Russian, Jackie?
 
2012-11-05 10:32:52 PM  

organizmx: He's calling Florida about 50/50, but it's a very slight blue on the map. If he goes 50 out of 50 tomorrow the media's games with these elections will be changed pretty much forever.


No they won't.

The media's game is to sell advertising and get you to watch it.
 
2012-11-05 10:33:30 PM  

LouDobbsAwaaaay: vygramul: Except Silver didn't come from journalism. He ran his site independently as a labor of statistical love. NYTimes was just first to recognize (and have the bucks to buy) Silver's expertise.

That's kind of my point. Outside of journalism Silver is a stat-junkie in a sea of stat-junkies. Inside journalism, he's a mythological figure using math-voodoo to destroy the concept of free will. His prediction skills didn't increase appreciably when he started applying them to political polling. He just ended up in a field where the next best guy at the job has a degree in journalism and a nice haircut instead of the least bit of knowledge or experience with what he was talking about.


Here's a good rundown of why pundits don't like him
 
2012-11-05 10:34:35 PM  
If this is gibberish to you...you prob should not be commenting too much on nate's methodology.

with that said, just remember "GIGO". I'm not saying his model is wrong, just that it COULD BE wrong, due to bad data inputs or missing important variables.
 
2012-11-05 10:35:07 PM  

GWSuperfan: Blue_Blazer: Updated again: forecast and nowcast are the same. GAME ON!!

they're not identical yet- the Now-cast has Obama a tenth of a percent more likely to win than the forecast.


Yeah, Im not seeing that, they are identical on my screen. Maybe switch from dial-up?
 
2012-11-05 10:36:13 PM  

Lenny_da_Hog: organizmx: He's calling Florida about 50/50, but it's a very slight blue on the map. If he goes 50 out of 50 tomorrow the media's games with these elections will be changed pretty much forever.

No they won't.

The media's game is to sell advertising and get you to watch it.


I think the "and get you to watch it" part is the part that changes if Nate's predictions hit the nail on the head. It's going to be hard to convince people to watch as long as there's this guy with a free blog just giving away predictions that are consistently much better than what the 24-hour news guys can squawk about.
 
2012-11-05 10:36:16 PM  

cameroncrazy1984: nmemkha: I predict the liar will win.

So, Romney?


Considering Romney has taken both sides of any policy position, he will always be rated half-true so you can't really call him a liar.
 
2012-11-05 10:36:47 PM  

vygramul: LouDobbsAwaaaay: vygramul: Except Silver didn't come from journalism. He ran his site independently as a labor of statistical love. NYTimes was just first to recognize (and have the bucks to buy) Silver's expertise.

That's kind of my point. Outside of journalism Silver is a stat-junkie in a sea of stat-junkies. Inside journalism, he's a mythological figure using math-voodoo to destroy the concept of free will. His prediction skills didn't increase appreciably when he started applying them to political polling. He just ended up in a field where the next best guy at the job has a degree in journalism and a nice haircut instead of the least bit of knowledge or experience with what he was talking about.

Here's a good rundown of why pundits don't like him


I thought it was his weight and overall feminine mien.
 
2012-11-05 10:36:47 PM  

vygramul: LouDobbsAwaaaay: vygramul: Except Silver didn't come from journalism. He ran his site independently as a labor of statistical love. NYTimes was just first to recognize (and have the bucks to buy) Silver's expertise.

That's kind of my point. Outside of journalism Silver is a stat-junkie in a sea of stat-junkies. Inside journalism, he's a mythological figure using math-voodoo to destroy the concept of free will. His prediction skills didn't increase appreciably when he started applying them to political polling. He just ended up in a field where the next best guy at the job has a degree in journalism and a nice haircut instead of the least bit of knowledge or experience with what he was talking about.

Here's a good rundown of why pundits don't like him


Ahhh, the parallels to the Catholic Church/Science can be inferred from that. It truly is the old tug and pull that defines human history.
 
2012-11-05 10:37:00 PM  

SlothB77: Same answer to "will he have any credibility after this election"


No, I am pretty sure he is not a witch.
Even if he loses a lot of credibility in political circles, he is still a top statistician and his reputation in other fields (namely baseball) and with people who understand statistics will still be great. Worst case scenario he ends up back somewhere like Baseball Prospectus, and considering he plans on leaving politics for something new after these elections anyways....

BKITU: The odds against that are slight. Silver is, far and away, the most generously-conservative in that respect, saying it's about a 16% chance of being the case. Other aggregators, like Sam Wang at Princeton, put the odds of that at 1% or less.


That was a couple days ago, it has been steadily shrinking (as more polls came in and as the national polls swung for Obama) at 10:30pm EST he is giving Ohio a 9% chance and any combination of states to win the election for Romnet an 8.5% chance (with a 0.1% chance events between now and the close of polls change the outcome).

nekom: That makes sense. There would be little he could do to curb that, other than imagining that it would probably be rare for the polls to be inaccurate, which may be true. He appears to have a very good track record, I guess I'm just naturally suspicious when anyone is so very sure about something, even though it certainly does seem likely.


Part of his analysis includes looking at historic polls and results as well as demographics and other events to determine the likelihood of this sort of scenario.
 
2012-11-05 10:37:26 PM  
imgs.xkcd.com
 
2012-11-05 10:37:35 PM  

rthanu: If Romney wins tommorow at less than 8%, I will certainly call him wrong. if the odds were 25% I'd say otherwise.


i.imgur.com
 
2012-11-05 10:37:39 PM  

LouDobbsAwaaaay: rthanu: If Romney wins tommorow at less than 8%, I will certainly call him wrong. if the odds were 25% I'd say otherwise.

This is the lack of understanding of probability this country faces. Somehow it's okay for Romney to win with a 25% chance but not an 8% chance? WTF?


That's what people do not get. The only way Nate can be technically wrong here is if he said the chances for Romney are ZERO% and Romney still wins. Probabilities, how do they work???
 
2012-11-05 10:37:46 PM  

StoneColdAtheist: mayIFark: /I can't take it anymore. I know, we should win, but the possibility of a Romney presidency scares me so much that I can't relax. The stake is way too high.

Me too. My usual remedy is booze, but I have blood work being drawn tomorrow morning for my annual physical, so must forgo.

Speaking of which, since I'll be on base I plan to hit the Package Store for some celebratory hooch. Anybody have suggestions? Scotch is a family fav, but strikes me as too heavy for all evening drinking. Wine seems too light. Beer too meh.

Vodka? I haven't had any in months. Gin? I bet I haven't had any gin in several years. Ideas, people...I need ideas!


Rum and coke with a splash of orange soda. Personal favorite.
 
2012-11-05 10:38:41 PM  

Blue_Blazer: Updated again: forecast and nowcast are the same. GAME ON!!


This.

organizmx: He's calling Florida about 50/50, but it's a very slight blue on the map. If he goes 50 out of 50 tomorrow the media's games with these elections will be changed pretty much forever.


Even if he went 48 or 49 out of 50, he'll be what Lunardi is at ESPN with that hole Bracketology crud.
 
2012-11-05 10:39:00 PM  

Nebulious: StoneColdAtheist: mayIFark: /I can't take it anymore. I know, we should win, but the possibility of a Romney presidency scares me so much that I can't relax. The stake is way too high.

Me too. My usual remedy is booze, but I have blood work being drawn tomorrow morning for my annual physical, so must forgo.

Speaking of which, since I'll be on base I plan to hit the Package Store for some celebratory hooch. Anybody have suggestions? Scotch is a family fav, but strikes me as too heavy for all evening drinking. Wine seems too light. Beer too meh.

Vodka? I haven't had any in months. Gin? I bet I haven't had any gin in several years. Ideas, people...I need ideas!

Rum and coke with a splash of orange soda. Personal favorite.


I like my rum (or bourbon) with Vanilla Coke.
 
2012-11-05 10:39:15 PM  

LouDobbsAwaaaay: rthanu: If Romney wins tommorow at less than 8%, I will certainly call him wrong. if the odds were 25% I'd say otherwise.

This is the lack of understanding of probability this country faces. Somehow it's okay for Romney to win with a 25% chance but not an 8% chance? WTF?


get over yourself. I understand probability just fine, but there is a line that where Silver can become "wrong". I would argue that threshold is 85/15, but it's subjective of course.

Let me guess, if he said Obama +99.999999% you'd be saying the same thing.
 
2012-11-05 10:39:53 PM  

Jairzinho: LouDobbsAwaaaay: rthanu: If Romney wins tommorow at less than 8%, I will certainly call him wrong. if the odds were 25% I'd say otherwise.

This is the lack of understanding of probability this country faces. Somehow it's okay for Romney to win with a 25% chance but not an 8% chance? WTF?

That's what people do not get. The only way Nate can be technically wrong here is if he said the chances for Romney are ZERO% and Romney still wins. Probabilities, how do they work???


i.imgur.com
 
2012-11-05 10:39:58 PM  
The election day forecast and the nowcast have become aligned. It's like a beautiful electoral eclipse.
 
2012-11-05 10:40:30 PM  

Jairzinho: That's what people do not get. The only way Nate can be technically wrong here is if he said the chances for Romney are ZERO% and Romney still wins. Probabilities, how do they work???


He'd be wrong if none of his disparate predictions hold up. For example, right now he gives Obama a 92% chance in OH, 86% in NH, 86% in Ia, 82% in VA and 82% in CO. Now, based on those numbers, it would be reasonable for Romney to win one of those 5 states. But if Romney wins all 5, you'd have to say Nate's model is fundamentally flawed and he was 'wrong'.
 
Displayed 50 of 406 comments

First | « | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | » | Last | Show all

View Voting Results: Smartest and Funniest


This thread is archived, and closed to new comments.

Continue Farking
Submit a Link »
On Twitter





In Other Media


  1. Links are submitted by members of the Fark community.

  2. When community members submit a link, they also write a custom headline for the story.

  3. Other Farkers comment on the links. This is the number of comments. Click here to read them.

  4. Click here to submit a link.

Report