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.

(Rasmussen Reports)   Rasmussen Reports: We're sorry that our own polling data is showing that President Obama is winning. Starting Monday, we'll change our methodology to show that the race is really a tie   (rasmussenreports.com) divider line 85
    More: Stupid  
•       •       •

3220 clicks; posted to Politics » on 28 Sep 2012 at 11:10 AM (2 years ago)   |  Favorite    |   share:  Share on Twitter share via Email Share on Facebook   more»



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

Archived thread
 
2012-09-28 11:11:42 AM  
Unlike most conservative propaganda outlets, they at least tell you they are lying.
 
2012-09-28 11:13:19 AM  
Neither the new or the old method is outside the margin of error, subby.
 
2012-09-28 11:14:18 AM  
This is hilarious.
 
2012-09-28 11:14:37 AM  

RolandGunner: Neither the new or the old method is outside the margin of error, subby.


Meaning: it's a statistical tie either way. No need to get your panties in a bunch because of one point.
 
2012-09-28 11:14:58 AM  
Or you could employ legitimate polling techiques among statistical representative data sets and ask legitimate questions.

hey just saying...
 
2012-09-28 11:15:38 AM  
Its a little late to change their polling policies when they hafta change them back a week out from the election to once again match reality.
 
2012-09-28 11:15:51 AM  
Rasmussen was already the only good poll and now it will be better!
 
2012-09-28 11:16:23 AM  
FTA: Romney is supported by 86% of Republicans, while Obama gets the vote from 85% of Democrats. The GOP hopeful has a four-point edge among voters not affiliated with either major party.

I distinctly recall making a joke about this exact methodology about a week ago.
 
2012-09-28 11:17:37 AM  
Wow, this is just sad. They just admitted they will fudge the numbers on purpose to make the GOP look better.
 
2012-09-28 11:17:40 AM  
Yeah, it's totally tied right now you guys!


/just keep sleepwalking to election day, it's all in the bag for Romney because magic...
 
2012-09-28 11:18:25 AM  
Poor Tea Partiers, even Fox News and Rasmussen are failing to support their version of reality.
 
2012-09-28 11:18:54 AM  
If you take out all samples of registered Democrats and Independents, Romney is an overwhelming favorite.
 
2012-09-28 11:19:24 AM  
 
2012-09-28 11:19:47 AM  
Seems like there's a huge market for simply inventing polls that certain demos want to see.

Don't even bother with the polling part, just make charts that tell them they're right.
 
2012-09-28 11:22:15 AM  

HotWingConspiracy: Seems like there's a huge market for simply inventing polls that certain demos want to see.

Don't even bother with the polling part, just make charts that tell them they're right.


There sure is!
 
2012-09-28 11:22:33 AM  
I heard a report yesterday that said Obama's internal polling shows AZ closer than expected, and they're debating whether to use some resources there.

Romney needs every single state he currently leads in; if true, this would be a great opportunity to make him waste time & money somewhere he should be safe.
 
2012-09-28 11:22:46 AM  
They've consistently been 4-6% off most of the other major polls. That doesn't necessarily mean they're wrong, but if this keeps up, it does make this election make-or-break for them: if Romney wins, they were the lone voice of reason; if he loses, they may as well ask to join Fox, because they'll definitely be dismissed as a partisan tool.

Interestingly, this seems to indicate that they know that, and that they're backing away to keep credibility. You don't often see conservative-leaning media blink like that these days...
 
2012-09-28 11:23:35 AM  
Does this really surprise anyone? Making shiat up to reinforce their version of reality is half the battle.
 
2012-09-28 11:23:35 AM  
Rasputin Reports: the Republican polling company
 
2012-09-28 11:24:19 AM  
 
2012-09-28 11:25:18 AM  
FWIW, Nate Silver has been using the "leaners" version of Rasmussen all along.

Still not enough to keep Obama from kicking Romney's ass from here to next Tuesday.
 
2012-09-28 11:25:36 AM  
i2.kym-cdn.com
 
2012-09-28 11:26:05 AM  
Fox News has also made a huge push to discredit polling numbers over the last week. The more they can convince their base that Romney should be winning now, the easier it will be to delegitimize Obama's second term after he wins.
 
2012-09-28 11:26:29 AM  

spelletrader: There sure is!



That looks legit.
 
2012-09-28 11:27:36 AM  
Nothing freaks a republican out like a brush with reality.
 
2012-09-28 11:28:03 AM  

NateGrey: Bad news....for Obama.



Obviously, their polling data is skewed.
 
2012-09-28 11:28:43 AM  

whistleridge: They've consistently been 4-6% off most of the other major polls. That doesn't necessarily mean they're wrong, but if this keeps up, it does make this election make-or-break for them: if Romney wins, they were the lone voice of reason; if he loses, they may as well ask to join Fox, because they'll definitely be dismissed as a partisan tool.

Interestingly, this seems to indicate that they know that, and that they're backing away to keep credibility. You don't often see conservative-leaning media blink like that these days...


They're already a known partisan tool.
 
2012-09-28 11:28:53 AM  
TBH, I don't see the issue with including "leaners" in the results when we're this close to the election. That just seems natural.

Rasmussen has a lot of problems, but this isn't one of them.
 
2012-09-28 11:28:53 AM  

HotWingConspiracy: Seems like there's a huge market for simply inventing polls that certain demos want to see.

Don't even bother with the polling part, just make charts that tell them they're right.


That's the FOX/Rush/Rasmussen/American Thinker/AM hate radio/etc. business model.

1) Lie up an alternate reality to get the audience worked up. Hit all the buttons. Say anything.
2) Get them nice and outraged to keep tuning in to your outlet.
3) Reenforce and profit!

The people that consume this stuff aren't smart enough to realize what is happening, and it suddenly makes your news outlet relevant and you can sell advertising to repeat viewers. I would not be surprised if FOX secretly wants Obama to win so that they can continue, Obama being Bleh has really helped their viewership.
 
2012-09-28 11:30:27 AM  

crab66: spelletrader: There sure is!


That looks legit.


It's freeper troll math. Just read about their methods of "unskewing" and you'll know it is a load of crap.
 
2012-09-28 11:30:45 AM  
You know how I can tell no one read the article?


Just astart posting boys! To hell with the facts!

You guys crak me up.
 
2012-09-28 11:31:31 AM  

imontheinternet: [i2.kym-cdn.com image 500x512]


imageshack.us
 
2012-09-28 11:31:45 AM  
Oh they're finally starting to include cell phone polls?

*reads article*

Cheap bastards.
 
2012-09-28 11:32:06 AM  
Sorry, Libmussen, but I don't listen to your libbo lies. I get THE TRUTH from the mustache of Joseph Farrah.
 
2012-09-28 11:32:53 AM  

Lord_Baull: NateGrey: Bad news....for Obama.


Obviously, their polling data is skewed.


lol All of them Katie.

Matthews had a guy on yesterday: "why would you pay someone for wrong/slewed poll results? these firms get employment based on their accuracy"

So basic hopefully even Teabaggers can understand. Maybe not.
 
sp
2012-09-28 11:33:04 AM  
It's pretty standard to include leaners a month from the election.
 
2012-09-28 11:33:46 AM  
Polling bias in action:

1* Think for a moment about the health care plan passed by Congress and signed into law by President Obama. If Mitt Romney is elected president and Republicans win control of Congress, how likely is it that the health care law will be repealed?
2* If President Obama is reelected and Democrats regain control of Congress, how likely is it that the health care law will be repealed?
3* If Mitt Romney is elected president and Republicans win control of Congress, is the economy likely to get better, get worse, or stay about the same?
4* If President Obama is reelected and Democrats regain control of Congress, is the economy likely to get better, get worse, or stay about the same?
5* In terms of its impact on your own life, how important is it whether Barack Obama or Mitt Romney wins the presidential election this year?

If the Republicans WIN control of Congress? Nice way to say that the Republicans AREN'T CURRENTLY THE SOURCE OF CONGRESSIONAL INACTION.
 
2012-09-28 11:36:08 AM  
Eh... The Electoral College numbers are the only ones that matter, and they look even worse for Romney than the poll numbers.
 
2012-09-28 11:36:46 AM  

2 grams: You know how I can tell no one read the article?


Just astart posting boys! To hell with the facts!

You guys crak me up.


A Republican whining? Thats so unlike them.
 
2012-09-28 11:38:28 AM  
The Rasmussen Reports daily Presidential Tracking Poll for Friday shows President Obama attracting support from 47% of voters nationwide, while Mitt Romney earns the vote from 46%.

I keep wondering why we keep pretending national polls are relevant. The best I can come up with is since the vast majority of the population is not in a swing state, we like to pretend we matter.
 
2012-09-28 11:39:38 AM  

DeltaPunch: Fox News has also made a huge push to discredit polling numbers over the last week. The more they can convince their base that Romney should be winning now, the easier it will be to delegitimize Obama's second term after he wins.


either that or they know they are going to steal it and this is just preconditioning the masses to accept it.
 
2012-09-28 11:40:22 AM  
Not that I am defending Rasmussen, but that's not a change of 'methodology', subby
 
2012-09-28 11:40:24 AM  

imontheinternet: [i2.kym-cdn.com image 500x512]


I would argue that .999... = 1 because:

1) .999... and 1 are on the same numeric system
2) This numeric system is continuous
3) Yet nobody can name me a number that is between .999... and 1
 
2012-09-28 11:40:52 AM  
This is what happens when you get paid for the results, not the process
 
2012-09-28 11:43:43 AM  

rufus-t-firefly: AZ


InTrade is polling it at 76% Romney.

538 has Romney up by 7-8% in AZ

While that's certainly closer than I expected from Arizona, that's a hard battle to win for 11 EVs.

But, it would certainly be interesting - just like Obama campaigning here in Texas. Lay the groundwork for a blue takeover of the entire Southwest in 8-12 years.
 
2012-09-28 11:45:45 AM  

crab66: spelletrader: There sure is!

That looks legit.


I love that one of the polls they rely on is web-based. One that has Romney up by 11.
Link

Romney/Ryan 55%
Obama/Biden 44%

The QStarNews Quick Poll is the first trial run preceding the upcoming QStarNews poll on the presidential race and current events. This election season has seen so many polls heavily skewed and poorly weighted, that show skewed and inaccurate results. The purpose of this poll is to present the most accurate results possible based on the latest of what is known about the electorate and voter behavior and other statistical information available.

The QStarNews poll works with the premise that the partisan makeup of the electorate 37.6 percent Republicans, 33.3 percent Democrats and 29.1 percent independent voters. Additionally, our model is based on the electoral including approximately 41.6 percent self-described conservatives, 32.6 percent self-described moderates and 25.8 percent self-described liberals.

Republicans are 68 percent conservative, 27 percent moderate and 5 percent liberal. Among Democrats, 14 percent are conservative, 36 percent are moderate and 50 percent are liberal. Independents include 39 percent conservatives, 36 percent moderates and 25 percent liberals.

Our polls about doubly-weighted, to doubly insure the results are most accurate and not skewed, by both party identification and self-identified ideology. For instance, no matter how many Republicans answer our survey, they are weighted at 37.6 percent. If conservatives are over-represented among Republicans in the raw sample, they are still weighted at 68 percent of Republicans regardless. This system of double weighting should insure our survey produces very accurate results, not skewed either way for the Democrats or for the Republicans.
...
The web-based survey included 2075 responses surveyed between September 10-15. The poll has a margin of error of three percent.


So, allow your sample to select itself, apply your own "weighting," and pull a margin of error out of your ass.

Seriously, this shiat is considered legit by unskewedpolls.com.
 
2012-09-28 11:47:15 AM  

sprawl15: imontheinternet: [i2.kym-cdn.com image 500x512]

[imageshack.us image 680x481]


That was always one of my favorite puzzles as a bartender, but I couldn't remember how it went. Thanks for posting this.
 
2012-09-28 11:50:12 AM  

Hobodeluxe: DeltaPunch: Fox News has also made a huge push to discredit polling numbers over the last week. The more they can convince their base that Romney should be winning now, the easier it will be to delegitimize Obama's second term after he wins.

either that or they know they are going to steal it and this is just preconditioning the masses to accept it.


I find this conspiracy hard to believe.

They would need to rig Ohio, Florida, Colorado, and Virginia.

If Obama takes ANY of those (not even counting North Carolina, which is starting to trend back toward Obama), he wins.

They need to manage to rig Pennsylvania, Florida, Virginia, AND another swing state to pull it off. And even Pennsylvania's voter ID law isn't expect to flip PA.

Obama has too many paths to 270 for that level of vote rigging.
 
2012-09-28 11:50:42 AM  

Vodka Zombie: sprawl15: imontheinternet: [i2.kym-cdn.com image 500x512]

[imageshack.us image 680x481]

That was always one of my favorite puzzles as a bartender, but I couldn't remember how it went. Thanks for posting this.


(9 x 3) - 2 = 25.
 
2012-09-28 11:53:40 AM  

balki1867: imontheinternet: [i2.kym-cdn.com image 500x512]

I would argue that .999... = 1 because:

1) .999... and 1 are on the same numeric system
2) This numeric system is continuous
3) Yet nobody can name me a number that is between .999... and 1


If this is gonna turn into a troll thread about math, why not?

Step 1: one has to differentiate between numbers and numerals (or really better, numbers and representations of numbers). Let's say that numbers are points on a number line, and numerals are driving directions for how to get to those points. So for example, it doesn't bother anyone that the same point on the line has driving directions "2 + 2" and "5 - 1" -- those are both perfectly valid ways to get to the same point.

Step 2: What the hell does a driving direction like "1/7 = 0.142857142857..." mean?

cache.ohinternet.com
/hot
 
2012-09-28 11:54:17 AM  

Jackson Herring: Vodka Zombie: sprawl15: imontheinternet: [i2.kym-cdn.com image 500x512]

[imageshack.us image 680x481]

That was always one of my favorite puzzles as a bartender, but I couldn't remember how it went. Thanks for posting this.

(9 x 3) - 2 = 25.


Yeah. I know.
 
2012-09-28 11:59:40 AM  

Vodka Zombie: sprawl15: imontheinternet: [i2.kym-cdn.com image 500x512]

[imageshack.us image 680x481]

That was always one of my favorite puzzles as a bartender, but I couldn't remember how it went. Thanks for posting this.


It's showing two different sets and trying to add them together. What the three men paid, and where the money is.

Where the money is lines up fine, $25 to the hotel, $2 to the clerk and $3 to the dudes.

What was paid also lines up fine, $9 from each guy, a total of $27.
 
2012-09-28 12:02:54 PM  

Antimatter: Wow, this is just sad. They just admitted they will fudge the numbers on purpose to make the GOP look better.


www.myfacewhen.net
 
2012-09-28 12:03:13 PM  
I guess I have to take Rasmussen off the LOPCATGOPATA. I only added them this morning based of the Freepers freaking out.
 
2012-09-28 12:03:50 PM  
Something that worries me: what if Romney "wins"?

All of the polling data from everyone but Rasmussen is showing an Obama win. All of the respected data analysts (Nate Silver, Andrew Tanenbaum, Intrade, etc) are showing at least a 75% chance of an Obama win, with some data showing it as an almost certainty.

So if Romney wins, what do we do? It would be such an unlikely thing that there would almost have to be fraud going on. Would we investigate? Would Congress refuse to certify the vote? Would the courts step in? There's no real Constitutional solution here; Congress could refuse to certify the electoral vote, but that would be unheard of and cause a huge loss of legitimacy for whatever candidate eventually won.
 
2012-09-28 12:04:39 PM  
sprawl15:

I'm losing my farking mind over here. I've got a sheet of paper and everything. What the hell is this? What am I missing?
 
2012-09-28 12:08:57 PM  

moriarty23: I'm losing my farking mind over here. I've got a sheet of paper and everything. What the hell is this? What am I missing?


The guys didn't pay $27, they paid $25. $25 in the hotel's cash register, $2 in the clerk's wallet, $3 refunded.
 
2012-09-28 12:09:30 PM  

Mager: Vodka Zombie: sprawl15: imontheinternet: [i2.kym-cdn.com image 500x512]

[imageshack.us image 680x481]

That was always one of my favorite puzzles as a bartender, but I couldn't remember how it went. Thanks for posting this.

It's showing two different sets and trying to add them together. What the three men paid, and where the money is.

Where the money is lines up fine, $25 to the hotel, $2 to the clerk and $3 to the dudes.

What was paid also lines up fine, $9 from each guy, a total of $27.


Thank you for this, my ears were starting to smoke.
 
2012-09-28 12:09:43 PM  

2 grams: You know how I can tell no one read the article?

Just astart posting boys! To hell with the facts!

You guys crak me up.



FTFA:

Beginning this Monday, October 1, Rasmussen Reports will be basing its daily updates solely upon the results including leaners.

Care to elaborate?
 
2012-09-28 12:12:03 PM  

Lord Dimwit: Something that worries me: what if Romney "wins"?

All of the polling data from everyone but Rasmussen is showing an Obama win. All of the respected data analysts (Nate Silver, Andrew Tanenbaum, Intrade, etc) are showing at least a 75% chance of an Obama win, with some data showing it as an almost certainty.

So if Romney wins, what do we do? It would be such an unlikely thing that there would almost have to be fraud going on. Would we investigate? Would Congress refuse to certify the vote? Would the courts step in? There's no real Constitutional solution here; Congress could refuse to certify the electoral vote, but that would be unheard of and cause a huge loss of legitimacy for whatever candidate eventually won.


If polls maintain their current status, the level of electoral fraud needed to pull out a Romney win would be phenomenally high. We're talking about "losing" a couple million votes in a half dozen states. Unlikely.
 
2012-09-28 12:12:25 PM  

KarmicDisaster: HotWingConspiracy: Seems like there's a huge market for simply inventing polls that certain demos want to see.

Don't even bother with the polling part, just make charts that tell them they're right.

That's the FOX/Rush/Rasmussen/American Thinker/AM hate radio/etc. business model.

1) Lie up an alternate reality to get the audience worked up. Hit all the buttons. Say anything.
2) Get them nice and outraged to keep tuning in to your outlet.
3) Reenforce and profit!



4) Audience gets outraged that liberal MSM media isn't covering the REAL stories/ in the tank for Obama.
 
2012-09-28 12:14:38 PM  

LazarusLong42: Lord Dimwit: Something that worries me: what if Romney "wins"?

All of the polling data from everyone but Rasmussen is showing an Obama win. All of the respected data analysts (Nate Silver, Andrew Tanenbaum, Intrade, etc) are showing at least a 75% chance of an Obama win, with some data showing it as an almost certainty.

So if Romney wins, what do we do? It would be such an unlikely thing that there would almost have to be fraud going on. Would we investigate? Would Congress refuse to certify the vote? Would the courts step in? There's no real Constitutional solution here; Congress could refuse to certify the electoral vote, but that would be unheard of and cause a huge loss of legitimacy for whatever candidate eventually won.

If polls maintain their current status, the level of electoral fraud needed to pull out a Romney win would be phenomenally high. We're talking about "losing" a couple million votes in a half dozen states. Unlikely.


Right. I'm asking what we'd do if that happens.

More likely than losing the votes, though, is voter suppression combined with missing votes. Close polling places in minority districts. Reject voters without "sufficient" ID (where "sufficient" is different for different people). Reject ballots due to extraneous marks. That sort of thing.

But yeah, what do we do if Mitt wins through obvious fraud?
 
2012-09-28 12:18:14 PM  

moriarty23: Mager: Vodka Zombie: sprawl15: imontheinternet: [i2.kym-cdn.com image 500x512]

[imageshack.us image 680x481]

That was always one of my favorite puzzles as a bartender, but I couldn't remember how it went. Thanks for posting this.

It's showing two different sets and trying to add them together. What the three men paid, and where the money is.

Where the money is lines up fine, $25 to the hotel, $2 to the clerk and $3 to the dudes.

What was paid also lines up fine, $9 from each guy, a total of $27.

Thank you for this, my ears were starting to smoke.


This cartoon is one of the best illustrations I've seen about how the power of suggestion can be used to deceive people. The cartoon suggests that after all the transactions are completed, you should account for the $30 that was initially paid. And many people, even smart people, will go nuts trying to do so. But without this suggestion, I think most reasonably smart people would immediately see that after all transactions are completed, you should be accounting for the final $27 that was paid. Then, of course, it's easy: $25 was kept by the innkeeper, and $2 by the clerk.
 
2012-09-28 12:21:36 PM  

moriarty23: That was always one of my favorite puzzles as a bartender, but I couldn't remember how it went. Thanks for posting this.

It's showing two different sets and trying to add them together. What the three men paid, and where the money is.

Where the money is lines up fine, $25 to the hotel, $2 to the clerk and $3 to the dudes.

What was paid also lines up fine, $9 from each guy, a total of $27.


I had an english teacher in high school give the class this puzzle and then asked the class to explain it. Any time anyone raised their hand and got close to explaining it correctly, he would interrupt us, fume a little bit, and then reiterate the puzzle.

The lesson that day was, "this is how to win an argument. you distract someone when they start to make a legitimate point."
 
2012-09-28 12:24:06 PM  

moriarty23: sprawl15:

I'm losing my farking mind over here. I've got a sheet of paper and everything. What the hell is this? What am I missing?


It isn't 3 x 27 + 2 = 29 because $30 is meaningless. It doesn't exist anymore as a sum.

It is 3 x 27 -2 = 25

/or if you insist 3 x 9 - 2 + 5 = 30
 
2012-09-28 12:24:29 PM  

Hollie Maea: This cartoon is one of the best illustrations I've seen about how the power of suggestion can be used to deceive people. The cartoon suggests that after all the transactions are completed, you should account for the $30 that was initially paid.


The version I originally heard was slightly better as the clerk actually brought the full $5 to the men but they would rather tip the guy $2 than worry about how to split $5 three ways. That solidified the idea that they paid $9 each plus $2 as a tip.
 
2012-09-28 12:29:12 PM  

sigdiamond2000: 2 grams: You know how I can tell no one read the article?

Just astart posting boys! To hell with the facts!

You guys crak me up.


FTFA:

Beginning this Monday, October 1, Rasmussen Reports will be basing its daily updates solely upon the results including leaners.

Care to elaborate?


I'll elaborate, even though I didn't make the original statement, since I've been following the numbers closely for some time.

Rasmussen might be trying to make numbers look better for Romney than they really are. We don't know. We DO know that he is making the numbers look better for Romney than they really are--he has a consistent and predictable Republican "house effect"--but we don't know if it is intentional or not. He might just honestly believe in the demonstrably bad assumptions that he makes that leads to this house effect.

All this said, moving to a "leaners" set definitely is not a ploy to help Romney. The "leaners" numbers dance around on each side of the "likely voters" numbers--today they would "help" Romney's numbers by a point, but a few days ago they would have "hurt" him by a couple of points. It is pretty common practice for pollsters to move from a Registered Voters model to a Likely Voters model and then to a Leaners model as the election gets closer.

Yes, Rasmussen has a GOP house effect. It's predictable, and therefore harmless. It's nice for people who want to find some straws to grasp. But this "change in methodology" (which isn't actually that, just a change in top line reporting) won't change this for good or ill. In fact, for those who actually care about predicting the election, such as Nate Silver, this will change nothing at all, since he's been using the "leaners" numbers from Rasmussen all along. He's also priced in the house effect, for what it's worth.
 
2012-09-28 12:32:28 PM  

sigdiamond2000: FTFA:

Beginning this Monday, October 1, Rasmussen Reports will be basing its daily updates solely upon the results including leaners.

Care to elaborate?


"Leaners" are undecided Mexicans.
 
2012-09-28 12:35:10 PM  

rufus-t-firefly: The QStarNews poll works with the premise that the partisan makeup of the electorate 37.6 percent Republicans, 33.3 percent Democrats and 29.1 percent independent voters. Additionally, our model is based on the electoral including approximately 41.6 percent self-described conservatives, 32.6 percent self-described moderates and 25.8 percent self-described liberals.


dgt1.net
 
2012-09-28 12:40:28 PM  

RolandGunner: RolandGunner: Neither the new or the old method is outside the margin of error, subby.

Meaning: it's a statistical tie lie either way. No need to get your panties in a bunch because of one point.


Obvious typo.
 
2012-09-28 12:41:24 PM  

spelletrader: HotWingConspiracy: Seems like there's a huge market for simply inventing polls that certain demos want to see.

Don't even bother with the polling part, just make charts that tell them they're right.

There sure is!


well of course, it's all explained here. what? you don't think examiner.com is a creditable source?
 
2012-09-28 12:43:58 PM  

RolandGunner: RolandGunner: Neither the new or the old method is outside the margin of error, subby.

Meaning: it's a statistical tie either way. No need to get your panties in a bunch because of one point.


That's - not how MOE works.
 
2012-09-28 12:52:04 PM  

Curious: spelletrader: HotWingConspiracy: Seems like there's a huge market for simply inventing polls that certain demos want to see.

Don't even bother with the polling part, just make charts that tell them they're right.

There sure is!

well of course, it's all explained here. what? you don't think examiner.com is a creditable source?


With proper suspension of disbelief any source is credible.
 
2012-09-28 01:01:53 PM  

Hollie Maea: sigdiamond2000: 2 grams: You know how I can tell no one read the article?

Just astart posting boys! To hell with the facts!

You guys crak me up.


FTFA:

Beginning this Monday, October 1, Rasmussen Reports will be basing its daily updates solely upon the results including leaners.

Care to elaborate?

I'll elaborate, even though I didn't make the original statement, since I've been following the numbers closely for some time.

Rasmussen might be trying to make numbers look better for Romney than they really are. We don't know. We DO know that he is making the numbers look better for Romney than they really are--he has a consistent and predictable Republican "house effect"--but we don't know if it is intentional or not. He might just honestly believe in the demonstrably bad assumptions that he makes that leads to this house effect.

All this said, moving to a "leaners" set definitely is not a ploy to help Romney. The "leaners" numbers dance around on each side of the "likely voters" numbers--today they would "help" Romney's numbers by a point, but a few days ago they would have "hurt" him by a couple of points. It is pretty common practice for pollsters to move from a Registered Voters model to a Likely Voters model and then to a Leaners model as the election gets closer.

Yes, Rasmussen has a GOP house effect. It's predictable, and therefore harmless. It's nice for people who want to find some straws to grasp. But this "change in methodology" (which isn't actually that, just a change in top line reporting) won't change this for good or ill. In fact, for those who actually care about predicting the election, such as Nate Silver, this will change nothing at all, since he's been using the "leaners" numbers from Rasmussen all along. He's also priced in the house effect, for what it's worth.


I wasn't expecting someone to actually elaborate with a considered response. What a buzz kill.
 
2012-09-28 01:01:53 PM  

EighthDay: Hobodeluxe: DeltaPunch: Fox News has also made a huge push to discredit polling numbers over the last week. The more they can convince their base that Romney should be winning now, the easier it will be to delegitimize Obama's second term after he wins.

either that or they know they are going to steal it and this is just preconditioning the masses to accept it.

I find this conspiracy hard to believe.

They would need to rig Ohio, Florida, Colorado, and Virginia.

If Obama takes ANY of those (not even counting North Carolina, which is starting to trend back toward Obama), he wins.

They need to manage to rig Pennsylvania, Florida, Virginia, AND another swing state to pull it off. And even Pennsylvania's voter ID law isn't expect to flip PA.

Obama has too many paths to 270 for that level of vote rigging.


I so hope you are right. I have nightmares about another Florida 2000. Just a little election fraud and look what the long-term effect was. A decade of wars and a farked economy.
 
2012-09-28 01:07:50 PM  

sprawl15: imontheinternet: [i2.kym-cdn.com image 500x512]

[imageshack.us image 680x481]


::little bald kid from matrix::
"There is no thirtieth dollar...."
 
2012-09-28 01:10:06 PM  

Curious: spelletrader: HotWingConspiracy: Seems like there's a huge market for simply inventing polls that certain demos want to see.

Don't even bother with the polling part, just make charts that tell them they're right.

There sure is!

well of course, it's all explained here. what? you don't think examiner.com is a creditable source?


It's some award-winning reality-creation. Or it should win some awards, at the very least. The guy behind Unskewed Polls posted a prediction over the summer as to what the 2020 election would look like. Cliffs Notes: The striking down of Obamacare leads to a resounding Romney/Portman win, they cut taxes and save the world, and in 2020, Ken Cuccinelli is well on his way to beating Debbie Wasserman-Schultz to maintain GOP control of the White House.
 
2012-09-28 01:11:33 PM  
So we've moved to completely substituting our own reality then?
 
2012-09-28 01:13:17 PM  

MrBallou: I so hope you are right. I have nightmares about another Florida 2000. Just a little election fraud and look what the long-term effect was. A decade of wars and a farked economy.


Well, it's certainly not in the bag for Obama yet. There's still a lot that could happen in a month, but with the way the polls are now, even adjusting for the Bradley Effect, which was minimal from 2008, Obama would have to lose ALL of the battleground states to lose (I'm counting CO, FL, VA, NC, NH, OH, and NH as battleground states).

But all he has to do is get Virginia (and NH) or Florida or Ohio and it's over.

If Obama didn't have so many different paths to 270, vote rigging would certainly be a major concern, but that level of concerted fraud I just don't think the Republicans are capable to pull off (I think that some of them would be willing to do it, however.)
 
2012-09-28 01:17:38 PM  

sigdiamond2000: I wasn't expecting someone to actually elaborate with a considered response. What a buzz kill.


Oops...sorry. Won't happen again.
 
2012-09-28 01:19:43 PM  

NateGrey: Matthews had a guy on yesterday: "why would you pay someone for wrong/slewed poll results? these firms get employment based on their accuracy"

So basic hopefully even Teabaggers can understand. Maybe not.


Depends on what you want to sue the numbers for. If you're using them to decide policy and strategy, then you're right, it's worthless.

But if you're solely planning on using the "semi-legit" skewed poles for propaganda purposes or to blow smoke up your donors asses...
 
2012-09-28 02:07:12 PM  

Satanic_Hamster: NateGrey: Matthews had a guy on yesterday: "why would you pay someone for wrong/slewed poll results? these firms get employment based on their accuracy"

So basic hopefully even Teabaggers can understand. Maybe not.

Depends on what you want to sue the numbers for. If you're using them to decide policy and strategy, then you're right, it's worthless.

But if you're solely planning on using the "semi-legit" skewed poles for propaganda purposes or to blow smoke up your donors asses...


Such polls are also useful for combating voter apathy that could otherwise adversely effect more local elections.

The creators of such polls may also believe their alternative weighting methodology to be valid. This could be a consequence of confirmation bias; alternatively, a possibility exists -- however small -- that their methodology is valid and that the majority of polling companies are (by mistake, rather than due to conspiracy) using incorrect weighting methods.
 
2012-09-28 03:27:32 PM  

Dimensio: Such polls are also useful for combating voter apathy that could otherwise adversely effect more local elections.

The creators of such polls may also believe their alternative weighting methodology to be valid. This could be a consequence of confirmation bias; alternatively, a possibility exists -- however small -- that their methodology is valid and that the majority of polling companies are (by mistake, rather than due to conspiracy) using incorrect weighting methods.


Regardless of your presentation of the issue, it is pretty farking stupid to change your weighting methodology mid-election only when you can no longer show your candidate winning or in a tie.

Rasmussen is finished after this election as a credible polling service. They were very weak on their 2008 election predictions and now they have just made it worse.
 
2012-09-28 04:48:53 PM  

rufus-t-firefly: Our polls about doubly-weighted, to doubly insure the results are most accurate and not skewed, by both party identification and self-identified ideology. For instance, no matter how many Republicans answer our survey, they are weighted at 37.6 percent. If conservatives are over-represented among Republicans in the raw sample, they are still weighted at 68 percent of Republicans regardless. This system of double weighting should insure our survey produces very accurate results, not skewed either way for the Democrats or for the Republicans.
...
The web-based survey included 2075 responses surveyed between September 10-15. The poll has a margin of error of three percent.

So, allow your sample to select itself, apply your own "weighting," and pull a margin of error out of your ass.

Seriously, this shiat is considered legit by unskewedpolls.com.



Even worse when you look into the polling company a little bit. The guy who runs it describes himself on Twitter as "Politically incorrect, member of the Vast right wing conspiracy, the TCOT community, working to make Obama a one term president."

Totally legit.
 
2012-09-28 07:51:00 PM  

Five Tails of Fury: rufus-t-firefly: Our polls about doubly-weighted, to doubly insure the results are most accurate and not skewed, by both party identification and self-identified ideology. For instance, no matter how many Republicans answer our survey, they are weighted at 37.6 percent. If conservatives are over-represented among Republicans in the raw sample, they are still weighted at 68 percent of Republicans regardless. This system of double weighting should insure our survey produces very accurate results, not skewed either way for the Democrats or for the Republicans.
...
The web-based survey included 2075 responses surveyed between September 10-15. The poll has a margin of error of three percent.

So, allow your sample to select itself, apply your own "weighting," and pull a margin of error out of your ass.

Seriously, this shiat is considered legit by unskewedpolls.com.


Even worse when you look into the polling company a little bit. The guy who runs it describes himself on Twitter as "Politically incorrect, member of the Vast right wing conspiracy, the TCOT community, working to make Obama a one term president."

Totally legit.


It's the same guy! That "polling company" exists only to provide the weighting for unskewedpolls.com. He is using his own "poll" to weight the others. He was using Rasmussen for a while until he realized that even Rasmussen was about to start showing Obama ahead.

TOTALLY Legit.
 
Displayed 85 of 85 comments

View Voting Results: Smartest and Funniest


This thread is archived, and closed to new comments.

Continue Farking
Submit a Link »






Report