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.

(The New York Times)   Nate Silver does a recap of all the polling firms to see how accurate they actually were. CNN? Pretty good. Rasmussen? Pretty bad. Gallup? It might be time to get out of the polling business   (fivethirtyeight.blogs.nytimes.com) divider line 166
    More: Cool  
•       •       •

6137 clicks; posted to Politics » on 11 Nov 2012 at 11:03 AM (2 years ago)   |  Favorite    |   share:  Share on Twitter share via Email Share on Facebook   more»



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

Archived thread

First | « | 1 | 2 | 3 | 4 | » | Last | Show all
 
2012-11-11 05:01:27 PM  

quatchi: So the one polling group that the allegedly left-biased Nate Silver tossed out as an unreliable outlier was biased towards the Dems by 2.5 pts?


Not quite. It wasn't tossed for being an outlier, it was tossed because of questions regarding their methodology (not their results). Basically, he does a check into each polling firm and he could not get the required information from Pharos in order to satisfy himself they were using proper, scientific methodology and reporting accurate results. A similar case with JZ Analytics except that he had issues with a specific part of their selection process (basically, it was self-selected).
 
2012-11-11 05:03:13 PM  

mjjt: herrDrFarkenstein: I respect the work of 538, but to say that the most innacurate pollsters in your sample population should be discarded reeks of jackbooting thugs.

my understanding of 538 was that it worked on exactly that basis

1. get the results of all the polls
2. look at how reliable each has been in the past
3. take more notice of the ones that have a history of being right

you do the same in real life: if someone offers an opinion and you know in the past they've been wrong (for whatever reason) you take less notice of him than someone who's been right many times

nothing to do with thuggery, everything to do with rationality/science


There were also adjustments based on 'house effect' of the pollster, the state of the nation's economy and the 'state fundementals' (party ID numbers, etc.) but yeah.
 
2012-11-11 06:47:39 PM  

neenerist: That explains why I always feel above my station at a Starbucks.


You should: apparently you don't know what "academician" means.
 
2012-11-11 06:54:02 PM  

James F. Campbell: neenerist: That explains why I always feel above my station at a Starbucks.

You should: apparently you don't know what "academician" means.


like a scientician but with beard and sandals?
 
2012-11-11 07:15:23 PM  

mjjt: herrDrFarkenstein: I respect the work of 538, but to say that the most innacurate pollsters in your sample population should be discarded reeks of jackbooting thugs.

my understanding of 538 was that it worked on exactly that basis

1. get the results of all the polls
2. look at how reliable each has been in the past
3. take more notice of the ones that have a history of being right

you do the same in real life: if someone offers an opinion and you know in the past they've been wrong (for whatever reason) you take less notice of him than someone who's been right many times

nothing to do with thuggery, everything to do with rationality/science


actualhuman: mjjt: herrDrFarkenstein: I respect the work of 538, but to say that the most innacurate pollsters in your sample population should be discarded reeks of jackbooting thugs.

my understanding of 538 was that it worked on exactly that basis

1. get the results of all the polls
2. look at how reliable each has been in the past
3. take more notice of the ones that have a history of being right

you do the same in real life: if someone offers an opinion and you know in the past they've been wrong (for whatever reason) you take less notice of him than someone who's been right many times

nothing to do with thuggery, everything to do with rationality/science

There were also adjustments based on 'house effect' of the pollster, the state of the nation's economy and the 'state fundementals' (party ID numbers, etc.) but yeah.



http://en.wikipedia.org/wiki/Resampling_(statistics)#Jackknife

The joke ...

Jackknifing?
Bootstrapping?

Am I the only statistics joker here?

Sigh ...
 
2012-11-11 08:00:27 PM  

GardenWeasel: nogudnik: Doc Daneeka: DubyaHater: A couple of my right-leaning friends are still having teenage girl-level hissy fits over the election. They're looking at the overall popular vote map and wondering how Obama could possibly win "with that sea of red" in the middle of the country. I don't have the heart to explain it and I'd rather not lose a friend over the election.

They honestly don't understand the concept of population density?

Show the this. 2012 results with the state sizes adjusted for population

[www-personal.umich.edu image 300x200]

From: http://www-personal.umich.edu/~mejn/el ection/2012/

Or this, which shows results by county instead of state,

[i192.photobucket.com image 850x585]



Wow, that map is psychedelic, it moves in 3-D in the west, and the pie from Ohio to Georgia spins if you stare at it long enough.
 
2012-11-11 08:48:09 PM  

James F. Campbell: You should: apparently you don't know what "academician" means.


Barista?
 
2012-11-12 12:22:52 AM  
Next Nate needs to do an analysis off all the pundit predictions.
 
2012-11-12 12:23:54 AM  

neenerist: James F. Campbell: You should: apparently you don't know what "academician" means.

Barista?


So what's wrong? Are you an engineer? Did I hurt your little aspie feelings? I didn't know you guys had feelings, frankly.
 
2012-11-12 12:38:03 AM  

mjjt: herrDrFarkenstein: I respect the work of 538, but to say that the most innacurate pollsters in your sample population should be discarded reeks of jackbooting thugs.

my understanding of 538 was that it worked on exactly that basis

1. get the results of all the polls
2. look at how reliable each has been in the past
3. take more notice of the ones that have a history of being right

you do the same in real life: if someone offers an opinion and you know in the past they've been wrong (for whatever reason) you take less notice of him than someone who's been right many times

nothing to do with thuggery, everything to do with rationality/science


FUNNIER STILL
the GOP did EXACTLY what this guy is talking about.
except they threw out all of science
they tossed out all of the polls which they disagreed with
they kept and worshiped the polls which fit their narrative.

so instead of science and logic, they went with PURE fiction
 
2012-11-12 01:04:31 AM  

namatad: mjjt: herrDrFarkenstein: I respect the work of 538, but to say that the most innacurate pollsters in your sample population should be discarded reeks of jackbooting thugs.

my understanding of 538 was that it worked on exactly that basis

1. get the results of all the polls
2. look at how reliable each has been in the past
3. take more notice of the ones that have a history of being right

you do the same in real life: if someone offers an opinion and you know in the past they've been wrong (for whatever reason) you take less notice of him than someone who's been right many times

nothing to do with thuggery, everything to do with rationality/science

FUNNIER STILL
the GOP did EXACTLY what this guy is talking about.
except they threw out all of science
they tossed out all of the polls which they disagreed with
they kept and worshiped the polls which fit their narrative.

so instead of science and logic, they went with PURE fiction


yeah I actually think this is one of the most interesting underlying lessons of the election - science-based reality beats faith-based imaginary reality.

I realize most will choose to ignore the lesson but l hope it will be just one more drip that will eventually erode the mount

Not sure whether it's in the category of anti-evolution belief ("I don't care what the facts are, l won't examine my beliefs")

or whether it's so blatantly obvious that some will reconsider
 
2012-11-12 01:32:04 AM  
So the bookies, Vegas, and Intrade got it right after all. Wow. Who needs all them polls?
 
2012-11-12 01:40:32 AM  
K. Rove said he corrected the polls back to the 2004 dem/repub/indie demographics, on the presumption that 2008 was a one-off, that Obama was simply riding an anti Repub wave, and that 2012 turnout would look like 2004 and not 2008.

Not to mention ignoring changing trends in voter demographics that are independent of political preference.

Similar logic to Unskewed. 

A whooole lotta assumption / potential error / echo chamber / Super PAC waste going on there.
 
2012-11-12 03:54:58 AM  

Doc Daneeka: DubyaHater: A couple of my right-leaning friends are still having teenage girl-level hissy fits over the election. They're looking at the overall popular vote map and wondering how Obama could possibly win "with that sea of red" in the middle of the country. I don't have the heart to explain it and I'd rather not lose a friend over the election.

They honestly don't understand the concept of population density?


This. As a "liberal" (actually social-democrat) Australian (bizarrely our conservative party is the Liberal Party), I'm not invested in your election but was fascinated to watch it unfold anyway. Even I know that "We the people" is a pretty important phrase in your constitutional affairs. You can explain it without any politics, just simple maths. The "sea of red" would mean a lot if acreage had votes. But it doesn't. The people do, hence "We the people". And the people don't all live in the "sea of red", they substantially live in highly concentrated population centres away from the sea of red. So you can rightly point out "If you don't like our constitution, yer can git out" (heavy Southpark accent on this phrase). Guaranteed to make any right winger shut the fark up.
 
2012-11-12 10:52:03 AM  

Weaver95: if he keeps telling it like it is and proving his methodology to be correct, someone is gonna put a hit out on his ass.


It's okay. He'll know when and where the hit is scheduled to happen with a +/- 0.1% accuracy,
 
2012-11-12 02:10:31 PM  

Aussie_As: Guaranteed to make any right winger shut the fark up.


LOL U
 
Displayed 16 of 166 comments

First | « | 1 | 2 | 3 | 4 | » | 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