Romney, Obama, Rasmussen

[Kaosium]

Except that Rasmussen did incredibly poorly in 2010 and just average in 2008. There's evidence all over the thread for this; I don't understand this need on your part to counterassert established fact.

Apologies, I spent a day having my life 'enriched' by three (adorable!) little rugrats, not much time for posting today.

I was unimpressed by the quality of that 'established fact' and offered this instead where Nate Silver's own methods suggests Rasmussen has been one of the most accurate pollsters. As for the other one from '10, are they attempting to compare (mostly) 4-500 respondent tracking polls with 1200 or 2700 sample polls? I wasn't paying any attention to that election--you know what I was doing online and it wasn't politics--but just looking back it does appear that Rasmussen was amongst--if not the first--to see the historic '10 election coming which is a helluva lot more valuable than any single poll.

Bet big, then. There's a predictions thread and a betting thread where people are just dying for someone to take Romney's side.

Just trying to help you out here.

*sigh* Seriously? This is weak sauce, dude.


It was 7.8 when he took office, and it was 7.8 last month. Of course, in Feburary of '09 it was 8.3 and spiked to 10 in October '09 as the recession hit full swing; you are welcome to assert that in a single month Obama was terrible for the economy, but let's be serious, shall we? We'll grant the non-representative cherry-pick of "7.8" for now.

Now, more importantly: do you know why it went up 0.1%?



People are starting looking for work again. It's a very good sign for the economy, in fact, because it represents confidence instead of discouragement, which is important. And the unemployment trend, of course, looks excellent for Obama.

What doesn't look good is the fact unemployment is still around where it was when he took office and there's still a huge number of LTE. When is the last time that number was so high? Even if some are now looking for work and being added back into the unemployment rolls, in the context of the government spending we added there's no reason to be happy about 7.9% unemployment at this juncture.

That money was spent to stimulate the economy, and had it given us ~4% growth rates and brought the unemployment rate back down to ~5% or whatever was projected I can see how it would have been worth it. Had he not insisted on that stimulus and we didn't double the (public) debt then it would be another thing entirely. However I just thought it funny that people were celebrating the fact that Barack Obama will go into the election with the highest unemployment rate I can think of off the top of my head (right before an election) since the Great Depression despite all the government spending and debt we invested in attempting to lower it.

 

[KoihimeNakamura]

Uhm... do you u.. Nah, I don't think you understand preventive measures.

 

[Kaosium]

For you, Kaosium:



http://fivethirtyeight.blogs.nytime...simple-case-for-saying-obama-is-the-favorite/

Actually the impression those polls in the past were biased was created by news media (at the time) not being very meticulous about 'likely voter' and 'registered voter' and the populace themselves didn't understand the difference either. Thus a registered voter poll (or even all eligible adults) would be put out there and it would look like the Dem had a bigger lead as that wasn't trying to account for turnout as well as a 'likely voter' survey. Silver is not going to find that if he's just going back and looking at 'likely voter' results.'

My point here is that a number of these polls market themselves as 'likely voter' surveys, however they don't appear to be doing a very good job at projecting the population that will actually turn out and vote. Thus it's very much like there's a number of (effectively) 'registered voter' polls being mushed in with 'likely voter' polls this election cycle, and that effect may very well disguise the likelihood that Obama won't get the numbers those polls are projecting for him.

I'm not talking about one here and there, but what appears to be a systematic overestimate of certain demographics being at the voter booths that quite simply might be unrealistic. Do you expect that the electorate this year will perfectly resemble those who voted in '10? I don't either, it doesn't work that way, but it seems (too many) of these polls are assuming (or getting) '08 on steroids and that's not very damn likely either.

 

[Kaosium]

The prediction markets/sites make an excellent measure of confidence.

That wouldn't be the case if they were being manipulated.

 

[A'isha]

Actually the impression those polls in the past were biased was created by news media (at the time) not being very meticulous about 'likely voter' and 'registered voter' and the populace themselves didn't understand the difference either.

Silver isn't looking at "news media" reports or confusion among the populace. He looks at the numbers reported on the polls themselves, and compares those numbers to the way the elections actually turned out.

Thus a registered voter poll (or even all eligible adults) would be put out there and it would look like the Dem had a bigger lead as that wasn't trying to account for turnout as well as a 'likely voter' survey. Silver is not going to find that if he's just going back and looking at 'likely voter' results.'

Actually, he not only is, his model specifically takes that into account.

I'm not talking about one here and there, but what appears to be a systematic overestimate of certain demographics being at the voter booths that quite simply might be unrealistic.

No, there is no systematic overestimate of certain demographics being at the voter booths.

 

[Foolmewunz]

That wouldn't be the case if they were being manipulated.

But then one would have to make that case. You haven't even successfully shown to my satisfaction the accuracy of your contention that all the LV polls (or actually the ones you disagree with, I think) are really RV polls. Now you're going to go off and show us how ALL the bet/prediction sites are being manipulated?

Oh, and I owe you an answer on Why Fox News. Simple. I am an underprivileged person. I was ten years in Hong Kong and all I had was CNN International, BBC Int'l, MSNBC, and Al Jazeera. I used to be able to watch YouTubes of Fox but had never experienced it first hand. It's my only American news network at my house in Pattaya, so I tuned in once figuring it couldn't be all that deranged and I do like to keep some semblance of contact with things going on back home. Well, my face froze over and they managed to pry me out of my armchair and feed me liquids after about the first twelve hours, and I've been watching ever since. It's nice to see where the loony 5% get their information. (That's their maximum viewership! The only reason we see so much of them is that people flip their content to their blogs and websites and of course onto YouTube.)

But the good news is that I'm tiring of them. I do have to turn off Hannity quite often, but I want to see what the demagogues (not demi-gods as they like to think of themselves) are spinning so I'll often turn down the sound.

Great Moments in Freudian Slip History: Sean Hannity, about an hour ago, was plugging his pre-election special on Sunday.

And I quote...

"And for the election, we're going to do a one-time special... an hour of sanity... er Hannity, Sunday at 9 PM."

Karl Rove was on the next segment as he's sounding a little more sane than when he was acting as a bobblehead doll to Dick Morris yesterday. But he still sees OH, WI, CO, NH, and NV in the Romney column. When asked individually, he weasels so as to protect some degree of credibility after the election, but one by one he and Hannity agreed that all of the aforementioned are Romney wins. (Plus, of course, locks in VA, NC and FL).

 

[BenBurch]

mmmmmmmmm..............toast

Don't burn a DVD without it!

 

[Kaosium]

Silver isn't looking at "news media" reports or confusion among the populace. He looks at the numbers reported on the polls themselves, and compares those numbers to the way the elections actually turned out.

However if those numbers are generated by poor methods, perhaps in part because of what he noted in the link you provided, (that some of us have been politically abused ) this cycle and are no longer answering) then his modelling will be effected negatively because of it.

Actually, he not only is, his model specifically takes that into account.

That article supports the point I am trying to make: that 'likely voter' results are more likely to be accurate and registered voter results are generally more favorable to Dem candidates. What I am further noting is that from what I've seen a lot of these 'likely voter' polls are giving results that look like they were generated by 'registered voter,' or even 'eligible adults.' Those polls in Nevada I posted predicting ~40% minority turnout instead of a ~30% result which would be much more likely are one example

No, there is no systematic overestimate of certain demographics being at the voter booths.

Partisan ID is just shorthand for noting the demographics are skewed towards those which historically vote more democratic. Look, if I did a poll that (regularly) sampled 60% Texans and 40% Californians and claimed it represented the population when (pretend here for a moment) historically there were equal numbers voting and showed a 10 point lead for Romney with 10 percentage points more Republicans would you wonder if there was something wrong? Or think that all of a sudden there were a lot more Republicans in California?

Yes partisan ID can change, but in this election independent polling suggests there are more people that identify as Republicans than in '08 and for some reason a lot of these polls aren't getting anywhere near that result.

 

[Kaosium]

"Publicly" is the operative word here. There's zero chance that the Romney campaign thinks the polling is "skewed" even though they have an interest in fooling people into believing that they are. Trust me, exactly no one in that campaign will be surprised on Tuesday (unless they miraculously win, then they'll poop in their magic underwear)

I disagree, here's an article where the director of the NRSC is wondering (basically) the same thing, that GOP internals don't match what's being bandied about in the press--to a significant degree. Here's another one showing how some are wondering about what the actual demographic turnout will be in '12. Here's Michael Barone, who probably knows more about each and every congressional district and their voting habits than any man who ever lived, predicting Romney winning handily, and echoing some of the points I have been trying to make here.

I don't expect you to like that, but am just pointing out there's a divergent view of this election which is damned curious. Internals from both camps ought to say the same thing, whatever is in the press gets spun by either side, but in a campaign usually both sides are seeing the same thing privately. That doesn't appear to be the case this year.

 

[Kaosium]

But then one would have to make that case. You haven't even successfully shown to my satisfaction the accuracy of your contention that all the LV polls (or actually the ones you disagree with, I think) are really RV polls.

It's not that I disagree with the result but that the data in the polls or the methodology employed doesn't suggest it very likely to be an accurate predictor of the election.

Now you're going to go off and show us how ALL the bet/prediction sites are being manipulated?

Nah, I was just noting that if they were being manipulated, then they wouldn't be an accurate predictor of the election. As for whether someone might, some 4 billion (or whatever--I didn't verify that figure I just read it somewhere) has been spent on this presidential election and it appears some have trumpeted it as an indicator of Truth^TM and (apparently) some media are following it. I found it damned curious that huge drop in Romney's fortunes coincided with the release of the '47%' video from David Corn that was supported by the meme that it 'ended Romney's chances.' That was kooky nonsense and it's remarkable that so many 'penny stock' investors chose that moment to cut and run...

Oh, and I owe you an answer on Why Fox News. Simple. I am an underprivileged person. I was ten years in Hong Kong and all I had was CNN International, BBC Int'l, MSNBC, and Al Jazeera. I used to be able to watch YouTubes of Fox but had never experienced it first hand. It's my only American news network at my house in Pattaya, so I tuned in once figuring it couldn't be all that deranged and I do like to keep some semblance of contact with things going on back home. Well, my face froze over and they managed to pry me out of my armchair and feed me liquids after about the first twelve hours, and I've been watching ever since. It's nice to see where the loony 5% get their information. (That's their maximum viewership! The only reason we see so much of them is that people flip their content to their blogs and websites and of course onto YouTube.)

But the good news is that I'm tiring of them. I do have to turn off Hannity quite often, but I want to see what the demagogues (not demi-gods as they like to think of themselves) are spinning so I'll often turn down the sound.

Great Moments in Freudian Slip History: Sean Hannity, about an hour ago, was plugging his pre-election special on Sunday.

And I quote...

"And for the election, we're going to do a one-time special... an hour of sanity... er Hannity, Sunday at 9 PM."

Karl Rove was on the next segment as he's sounding a little more sane than when he was acting as a bobblehead doll to Dick Morris yesterday. But he still sees OH, WI, CO, NH, and NV in the Romney column. When asked individually, he weasels so as to protect some degree of credibility after the election, but one by one he and Hannity agreed that all of the aforementioned are Romney wins. (Plus, of course, locks in VA, NC and FL).

LOL!

I don't watch any of them anymore--why encourage them?

 

[12AX7]

mmmmmmmmm..............toast

I can't help but notice you aren't as excited about this thread now as you were when you started it.

 

[Peephole]

What doesn't look good is the fact unemployment is still around where it was when he took office and there's still a huge number of LTE. When is the last time that number was so high? Even if some are now looking for work and being added back into the unemployment rolls, in the context of the government spending we added there's no reason to be happy about 7.9% unemployment at this juncture.

That money was spent to stimulate the economy, and had it given us ~4% growth rates and brought the unemployment rate back down to ~5% or whatever was projected I can see how it would have been worth it. Had he not insisted on that stimulus and we didn't double the (public) debt then it would be another thing entirely. However I just thought it funny that people were celebrating the fact that Barack Obama will go into the election with the highest unemployment rate I can think of off the top of my head (right before an election) since the Great Depression despite all the government spending and debt we invested in attempting to lower it.

US debt didn't double. Gross debt rose from about 70% to a little over a 100% since 2008. There's a good case to be made that if the US hadn't passed any stimulus (and not taken a couple of other actions like the financial industry bailout or the expansive monetary policy), the US economy would have had a Great Depression style collapse, which would eventually have led to even greater levels of debt accumulating. What the stimulus did, was limit unemployment, and return the US to growth. That unemployment is still high and growth to low, is a consequence of too little stimulus. Everything we've learned over the past five years, has just reinforced what we've known for decades about economic depressions: austerity makes the hole bigger, and only intensive fiscal policy helps countries climb out of it.

Low growth and high unemployment is an enormous problem, one that makes the US's financial position a lot worse, especially over the long term. Fixing unemployment and restoring growth is the key to keeping government debt low, and is the easiest and only path to actually bringing it down to lower levels again.

That being said, US debt is in no way a problem at the moment, since the US can borrow at pretty much no cost at the moment. So it's a very strange time to be concerned about debt when the US economy is still so bad.

One final thing, the US today is doing better than in previous financial crises:

 

[PhantomWolf]

I can understand the Repubs dissing the polling, it's the only hope they have left. Traditionally any candidate with a lead greater than 1.5% ten days out went on to take the state. If you take all the states with a difference greater than 2% (just to be fair) and assign those to the candidate who holds the lead, Obama holds 277 EC votes even without assigning the final 4 toss up states. Based on that, Romney could pull all <2% states, (Florida, Nevada, New Hampshire, and Virginia) and would still lose. Hence why they have to convince themselves that the polls are all 4% too high for Obama...

 

[remirol]

I can understand the Repubs dissing the polling, it's the only hope they have left. Traditionally any candidate with a lead greater than 1.5% ten days out went on to take the state. If you take all the states with a difference greater than 2% (just to be fair) and assign those to the candidate who holds the lead, Obama holds 277 EC votes even without assigning the final 4 toss up states. Based on that, Romney could pull all <2% states, (Florida, Nevada, New Hampshire, and Virginia) and would still lose. Hence why they have to convince themselves that the polls are all 4% too high for Obama...

There's a great deal of schadenfreude to be had in this for those of us who fell into the same trap in 2004. Eight years later and wiser, it's certainly reassuring that the polls _do_ match what we want to happen -- and it's amusing to see all the Republicans making the same exact mistakes and using the exact same excuses and rationalizations as we did then.

In January 2005 those of us who were "anybody but Bush, please" had to acknowledge that the Dems had, in the end, nominated someone who was essentially unelectable. In January 2013 the Republicans will have to admit the same thing, barring the following set of coincidences all happening at the exact same time:

* _all_ the polls are systematically underestimating Romney's chances
* _all_ the prediction markets are being manipulated
* _all_ the forecasters' models are wrong
* _all_ the Republicans come out to vote
* _all_ the Dems stay home

Possible? Yes. Guess where my money is staying? And this is, of course, what the prediction markets say too -- only people with nothing to lose are still holding to the "polls are wrong!' mentality, and that's why nobody's putting up action on Romney in the bettin' thread, and why shorting Romney/going long for Obama on InTrade/$PREDICTION_MARKET is the popular position to work from.

ETA: Heh. Checked 538 after I wrote this, and sure enough, look what I found.

 

[JoeTheJuggler]

You know, I'm going to have to retract that.

There are a plenty of people, including Nate Silver, who have Obama up 60, 70, 80 points... but I wouldn't consider that a landslide. If a landslide is a victory on par with 2008 (365-173), then I don't know of anybody making that call for Obama.

And to clarify, a landslide refers to large margin of victory. Silver's 70% is a probability of Obama winning (by any margin at all).

Anyone predicting a landslide for either side is not being realistic.

 

[MattusMaximus]

And to clarify, a landslide refers to large margin of victory. Silver's 70% is a probability of Obama winning (by any margin at all).

Anyone predicting a landslide for either side is not being realistic.

Actually, Silver has an Obama win at 84% probability as of this morning, and it's going up by the day.

Incidentally, here is a good analysis by Silver about why Romney's chances are so poor at this stage.

Disclaimer: the above link contains actual math and statistical analysis, so certain right-wingers may wish to avoid facing the facts by not clicking it

 

[not_so_new]

And to clarify, a landslide refers to large margin of victory. Silver's 70% is a probability of Obama winning (by any margin at all).

Anyone predicting a landslide for either side is not being realistic.

Actually, Nate has Obama's chances at 83% now......

 

[JoeTheJuggler]

Silver had a good rip on pundits. It came as a tag on the end of a discussion about a comment he's made that if Romney wins, it can only be that the state polls have a bias against Romney--something he doesn't believe to be the case.

At the end of his clarifying what he was saying is this:

If the state polls are right, then Mr. Obama will win the Electoral College. If you can’t acknowledge that after a day when Mr. Obama leads 19 out of 20 swing-state polls, then you should abandon the pretense that your goal is to inform rather than entertain the public.

Now I happen to think pundits as entertainers might be a bit too dismissive of them. First, it's overly dismissive of entertainment, and second pundits do much more than discuss the likely outcome of elections. But I'm 100% with Silver that if a pundit broaches the topic of the likely outcome of an election and ignores the facts, they lack any credibility.

I'm reminded of when Letterman used to rip on Dr. Phil talking about a "book" of his, More Advice Dr. Phil Just Pulled out of His Ass.

 

[RandFan]

Actually, Nate has Obama's chances at 83% now......

83.7% but who is counting?

 

[JoeTheJuggler]

Actually, Silver has an Obama win at 84% probability as of this morning, and it's going up by the day.

Actually, Nate has Obama's chances at 83% now......

Still, it's a probability and not any reflection of a margin. Mikedenk seemed unaware of that fact in considering whether Silver's high probability of an Obama win indicated a landslide.

ETA: In fact, Silver currently has the probability of an Obama landslide win at 0.5% (defining landslide as a double digit popular vote margin--which I take to mean double digit percentage--that is, a 10% or greater margin, and not a margin of 10 votes or more!) By comparison, he has Romney's probability at a landslide win as even nearer to nothing (less than 0.1%).