Raytri's rules of polls and commentators

There's one reason that Romney went to both states; polling shows there's at least a chance. It's also why Obama went into Wisconsin; because he knows it. IF he doesn't go into Penn now, they'll just think he's taking them for granted and that might backfire too. So I wouldn't be surprised that he doesn't show up there too.

 

http://elections.nytimes.com/2008/results/president/map.html

That is a link to actual results from 2008. Just hover over a state and you can see the points.

 

Excuse me? Partisans don't spin? Since when?

I don't believe *anything* either campaign says about the horse-race aspect of the election.

You usually won't see a heavy push in a state unless a campaign feels like they have a chance there. But there are other factors at play.

1. The Romney campaign has been working hard to project the idea of "momentum" in the last couple of weeks. They've engaged in a bunch of smoke-and-mirrors stuff, like loudly announcing they were pulling out of NC because they had it in the bag, when they transferred exactly one staffer. This feels like more of the same -- or else an effort to find a different path to 270 because they feel Ohio is slipping away. The fact that nothing in the polls suggests Pennsylvania is seriously in play (Obama is still up there by several points) makes this explanation even more likely.

2. Pennsylvania has proven a false lure for Republicans before. McCain made a late play for the state in 2008 --but iit was a sign of desperation, not good internal polling.

3. With both campaigns awash in money, you can spend a few million on a flier -- or defending against a flier -- without harming your other prospects.

4. The Romney effort in Pennsylvania is relatively small -- a single candidate visit. If they were serious, it would be sustained. The campaign and GOP Super PACs are tossing in $11 million in ad money, but Republicans admit they're only spending money there because it has become physically impossible to buy air time in swing states:
http://www.huffingtonpost.com/huff-wires/20121104/us-romney-pennsylvania/


Republicans acknowledge that decisions to buy ads in Pennsylvania were made, in part, because millions of dollars were still available – and airtime in other swing states had simply run out, already filled with political ads.


5. The Obama response in Pennsylvania makes it clear they don't think Romney is that serious, either. They're spending about $3.7 million -- a fraction of the GOP total -- and sending in a surrogate (Clinton), not Obama.

 

It's proving your point for one thing and adding a new rule,

Anyone who whines about polling methods is losing.

 

I think both sides have way too much money. So Mitt is trying to put as many states as possible into play, and the Obama campaign is being careful to counter him.

I do think it reflects an underlying sense in the Romney camp that they are behind, for several reasons:

1. The Obama campaign is going into these areas only after Romney commits major resources. This suggests they're not really worried, otherwise they would have been proactively shoring up those areas before this.

2. These states (Penn, Wisconsin) aren't true toss-up states: they definitely lean toward Obama. If Romney's campaign had 270 EV they were mostly confident in, they wouldn't be trying so hard to expand the map in states where they've trailed since the beginning, since the cost-benefit is against you: a dollar spent in Wisconsin is far less likely to result in an EV victory than a dollar spent in, say, Ohio or Colorado. So there's not a real good reason to go after Wisconsin unless you're really worried about Ohio and Colorado.

Or it could just be, as I noted, that both sides have more money than they know what to do with. If you're doing everything you possibly can in Ohio and still have money in the bank, why not throw it at Wisconsin?

But nothing I've seen suggests that the moves in Wisconsin and Pennsylvania are because the Romney campaign sees a major opportunity there. The polling just doesn't support that idea.

 

rasmussion is well known to be the worst of the pollsters.
You will note they are far right until the last few days then they correct their bias and show true numbers. That is the ONLY way they get close.

 

Thank you for proving the point.
rasmussen is the suckers poll.

 

With all the new republican voter suppression laws around the country it's no wonder your daughter had trouble, that was the right wings goal.
Make voting so hard most won't bother.
They are anti-Americans and they know the only way for the republicans to win is to steal the election.
It's not working thank god.

 

Today, Nov 5th; the day before the voting.....Rasmuessen has Romney up by 1pt: 49% to 48%
He has 1% still undecided and usually undecideds break for the challenger....especially when the incumbent has had an Approval Rating of under 50% for most of the last 3 years.

BTW, today Gallup has the race at 48-48. CNN has it also tied, but with a D+11
Pew has it Obama 48, Romney 45, with a D+6

If you take away the skewing/padding in favor of Democrats, all but one show Romeny winning.



http://washingtonexaminer.com/novem...-coal-regulation/article/2512538#.UJfTBMWHK8B

 

JP5, you're buying into several pure myths:

1. There is no "skewing/padding" to favor Democrats in the polls. That's just flat false. Most pollsters are simply reporting what people are telling them, not imposing some specific partisan breakdown on their polls.

2. Rasmussen has had a pro-GOP lean of 2-3 points this cycle. A big reason for that is its polling methodology: robocalls that only call landlines, and poll the first person to answer the phone. Another reason is that Rasmussen is one of the few pollsters who *does* impose a specific partisan breakdown on their polls.

3. Rasmussen's number has suddenly moved toward the consensus in the last few days. So now it's not quite as much of an outlier. But what would you say about a pollster who is consistently 2-3 points off from the consensus throughout the campaign, then suddenly shifts toward the consensus right before the election -- knowing that that final number is what will be used to assess their accuracy?

4. Historically, undecideds don't break toward the challenger in massive numbers. An analysis of September polls compared to the final results showed that the incumbent typically gained 3.5 percentage points in the final results, while the challenger gained 3.9 percentage points. That's just not a very big gap.
http://fivethirtyeight.blogs.nytimes.com/2012/07/22/do-presidential-polls-break-toward-challengers/

 

That is absolutely not true Raytri. When they are giving Dems a +11, that means they are saying that Dems have a huge leg up on enthusiasm and turnout. IF they are basing that on early voting, it doesn't give the whole picture.....and they'll be sadly diasppointed once that get that whole picture. I don't know of any poltical analyst....Republican OR Democrat....who thinks the Democrats are coming out in the same numbers as they did back in 2008. Just look at the crowds and their enthusian this time....compared to 2008. It's all there.

Well, regardless, in 2008 they were #1 on being the closest to the final vote.....as they predicted an Obama win. Bet ya'll weren't complaining he was "right-wing" back then, eh?

Unlike the other polling firms, Rasmussen stops the padding for what they think turnout will be a few weeks before voting day. And if you think their final numbers are accurate...good. Because today Nov 5th, he's showing R 49, O 48. He's also still showing a 1pt undecideds....which will usually break for the challenger this late in the game.

I didn't say "massive numbers." Any difference at all....like the one you point out above, can certainly make the difference in a race that is tied to start with.

 

Cherrypicking, anecdotes and your own feelings that the polls can't be right are a poor substitute for actual data.

 

That's not what it means at all. All They're saying is that they when they conducted their survey, respondents who said they were Dems outnumbered respondents who said they were Reps by 11 percentage points. *They're simply reporting what the people they polled told them.* It says nothing about enthusiasm or turnout.

They're not.

Agreed. But that has nothing at all to do with whether the Party ID numbers the polls are finding is "skewed."

Why do the Dems continue to show a party ID advantage? The reasons, IMO, are pretty simple:

1. Dems have *always* had a party ID advantage.

2. There are not *more* people IDing as Dems compared to 2008; there are fewer. It' just that the number of people identifying as GOP has fallen even more sharply. Both parties have lost people to the "independent" label.

3. Why has the GOP lost more? Two obvious reasons. Many Tea Partiers identify as independent instead of GOP, and they came straight out of the GOP ranks; and many moderate Republicans have been turned off by the rightward swing of the GOP.

4. This also helps explain why Romney has a lead among independents: many of those new independents are former GOP voters, and still lean that way.

Rasmussen has had a pro-GOP bias since the beginning. It's not necessarily because Rasmussen is skewing the results: a lot of the blame lies with their methodology -- robocalls to landlines that ask questions of whomever answers the phone first.

Your listing refers to a preliminary analysis conducted right after the 2008 election. A more complete 2009 analysis by the same author showed Rasmussen was the 9th most accurate pollster:
http://www.fordham.edu/images/acade...campaign_/2008 poll accuracy panagopoulos.pdf

And in 2010, they were *really* bad:
http://fivethirtyeight.blogs.nytime...rate-quinnipiac-surveyusa-performed-strongly/

Um.... what?

I said Rasmussen moved *toward* the consensus. I did not say they suddenly became highly accurate.

Point is, in most of the swing states Obama leads by way more than 0.4%. Romney can't overcome that gap with a minor advantage among undecideds.

 

I guess that is why they tied with Pew as the most accurate in the 2008 prediction

 

according to the .poll which used a 7 - 8% larger sample of Democrats to Republicans In Virginia, Romney leads Obama by a massive 21 points among Independents. In OH and FL, he leads by 5-6 points. Current early voting shows the Democrats down proportionally to Republicans compared to 2008. The GOP's enthusiasm is higher. All of that means bad news in Obama land

Now seriously why do you think that the electorate is more like 2008 than 2010. I am serious here, what facts can you point to that shows that the electorate shifted back from red to blue in the last two years?

 

As I pointed out in the post two above yours, "independents" are not the same thing as "undecideds". Try to keep this apples to apples.

Of course the proportion is down from 2008 -- that was a wave election where Democrats pushed early voting and Republicans didn't. Plus early voting is far more widely available than it was in 2008, GOP efforts to restrict it notwithstanding. When you have millions more people using early-voting, you're likely to see the partisan split even out a bit.

But you ignore the *meaning* of early voting to each party. Democrats use it to turn out marginal voters; since Republicans generally have higher turnout anyway, Republicans are more likely to simply be cannabilizing their election-day voters than turning out marginal voters. So it's hard to read much meaning into early-voting data this cycle. But for what it's worth, Obama is generally leading in the early voting. So Romney is hoping that they'll win the election-day voting by a large enough margin to overcome Obama's early-voting lead.

I don't. I think the electorate is more like 2012. Both 2008 and 2010 were wave elections -- one built around anti-Bush sentiment, one built around economic fears. In 2012, the economy is recovering rather than tanking, and Bush is long gone.

Further, after every wave election you have a certain "reversion to the mean." Incumbents who are marginal fits for their district, but were carried into office by the wave, tend to lose. So we should see the GOP underperform in 2012 if for no other reason than that. There are other factors that will mitigate that effect -- notably, the GOP win in 2012 left them in charge of a lot of legislatures, with the power to draw GOP-friendly districts based on the 2010 census -- but the trend will still apply.

Then there are the specific candidates. Obama is a known quantity; he's the safe choice, if you're truly undecided. Mitt Romney is the "take a chance" candidate. He has been painted as an out-of-touch rich guy -- a job made easier by the fact that he *is* an out-of-touch rich guy. He has flip-flopped in the most open and cynical way I've ever seen -- running hard right in the primaries and then disavowing most of those positions in the general. He has refused to provide any meaningful details about his economic plans, or how the numbers add up. He says "trust me". His positions on immigration have alienated Hispanics, the fastest-growing voting bloc in the country.

If Romney loses, a certain amount of blame will attach to Romney being Romney.

 

OMG that is so pathetic. You actually believe that Nash Equilibrium is only at works in theoretical models. I've got new for you son. Nash Equilibrium is in the real world at work every moment of every day. People are playing the game of life and adjusting their strategies based on what they expect their opponents do.

Let me simplify it for your sonny. I never said that you build a polling model based on Nash Equilibrium. I said that Nash Equilibrium effects the polls.

Anytime you build a model you have to predict how the model itself will effect the equilibrium.

 

and if there is a 6 point net shift from Democratic to Republican that would cancel out the 6 points Obama won by in 2008

mind showing me where you get this millions more number form , only pace that I can find that shows actual numbers copied from the states databases is here

http://elections.gmu.edu/early_vote_2012.html

fact or opinion on the marginal voters, facts require numbers, opinion just means you type it in. Link to the hard data if this is a fact

on the link above , Democrats had a several point advantage on early voting in 2008. This year looks like in most area the GOP is up a point or so percentage wise compared to 2008. I can find some exceptions but in general.

well the only 2 numbers I can find on the electorate for 2012 are Gallup and Pew. Gallup shows the GOP ahead by a significant margin while pew had this to say

[/quote]

being a known quantity is not all positive for Obama, look at how people view his ability to turn the economy around on all these polls

also if Obama loses a certain amount of blame goes to Obama for being Obama. Even ignoring everything but the last few months he made some major mistakes in my opinion. Picking a fight with the Catholic Church and his flip flop on gay marriage being 2 notable ones.

 

Guess you have a lot of explaining to do to John Kerry.

 

Mikezila only has one polling rule-only one poll counts. they will give you a clue, but they are often wrong.

 

Yep. Kerry engaged in all of the above. And lost.

So did McCain.

This cycle, it's Romney doing it. Not saying that proves he's going to lose, just that he and his supporters know the data isn't in his favor.

 

Did I leave that one out? "The only poll that matters is on Election Day" is another line trotted out by losing campaigns.

 

I am still waiting on you to prrovide data to show me how a net shift of 7+ points in the electorate, a decline in Democratic enthusiasm and a decrease in early voting in heavily democratic areas is a net positive for Obama. We won't even go into how the economy is by far the #1 issue for most voters and how Romney wins in every poll in the last month or so that I can remember on the question who can handle the economy better

 

or that he rarely, if ever, comes up with over 50%. undecideds this late in the game won't vote or vote for the challenger.

 

Did I say it was a net positive?

Obama won by 8.5 million votes in 2008. He could lose 8,499,000 of them and still win.

You guys keep trying to compare this to 2008. I'm looking at 2012 data. And so far, the 2012 data shows Obama winning.

Could the data be wrong? Sure. But at the moment Obama has the edge.