October 9, 2012

Real Clear Politics has Romney up 0.7%.

I don't know if Romney had ever been ahead in this average of the polls. He was tied a few times in early September, between the conventions, after which he sank until the debate turned things around.

39 comments:

Mark O said...

First time in a year.

Anonymous said...

Are polls that precise, to .7% when their error is +- 3%?

Emil Blatz said...

Yowsah dat!

Anonymous said...

If you set the range on their graph to "Max", you'll see that Romney was slightly ahead a couple of times last fall. Not that it meant much of anything at that point.

Hari said...

Can anyone actually define "margin of error" as used by pollsters?

For example, how can two polls each with a margin of error of 3% differ by 7%?

Mike (MJB Wolf) said...

It's a trending topic.

buwaya said...

Its a trap.
The polls, most of them, are sheer BS no matter where they point. They are adjusted arbitrarily to make any point or serve any rhetorical strategy.

Steve Austin said...

Nate Silver! Nate Silver! Nate Silver!

Good news. We get another least day where Obama's version of Toyko Rose will need to address these bad polls for his man.

David said...

Oh good another poll.

D. said...

"Can anyone actually define "margin of error" as used by pollsters?"

see here:

http://pjmedia.com/blog/polling-theory-when-polling-is-useless/

Anonymous said...

These polls are a joke. I REFUSE to ACCEPT that The-Greatest-and-the-Best POTUS is behind the corporate raider, Romney.

Obama is our Economic-compass. He is our east, west, north, and south. He is our savior. He is the best person to be on Mt. Rushmore.

Anonymous said...
This comment has been removed by the author.
sane_voter said...

You have to look at the D/R/I split of the polls. It seems that alot of the variation is due to how the pollster views the electorate and how much they compensate to reach that split. For example, if you have one poll that is a D+9 and has Obama up by 3, and another is D+2 and has Romney up by 2 then they are actually pretty close to one another. Another thing to look at is how are independents splitting between Obama and Romney. Many national polls have had Romney winning the I's by high single to mid double digits. Thus the pollsters have to have a substantial D tilt to the poll to have Obama in the lead.

In 2008 the electorate turned out to be D+7. Most other recent presidential elections are closer to D+2. For the 2010 midterms it was R+1.

Gabriel Hanna said...

@Hari:Can anyone actually define "margin of error" as used by pollsters?

For example, how can two polls each with a margin of error of 3% differ by 7%?


It's essentially based on the number of people they polled, which is usually around 1000. The basic idea is that 2 out 3 polls would end up within 3% of the "true" number, and 9 out of 10 would end up within 6%.

So it's not a contradiction if they differ by more than the margin of error.

Gabriel Hanna said...

The accuracy of any poll depends critically on having a representative random sample of the population. And it is hard to poll more than a few thousand people. So margins of error are rarely going to be lower than 3-4%, and 3-4% represents the ideal case.

Anonymous said...

POTUS Obama Hates Romney:

"The Democrat said that Obama's inner circle was dismayed at the 'disaster' and that he believed the central problem was that the President was so disdainful of Romney that he didn't believe he needed to engage with him."

http://www.dailymail.co.uk/news/article-2215173/Obama-believed-beaten-Romney-Denver-debate-ignoring-advice-aides.html

HE IS CORRECT. Who is Romney? He is not the great POTUS.

Obama is the BEST and the GREATEST.

edutcher said...

You can take the RCP average and cram it someplace.

These guys have featured the most skewed polls around, but now are trying to salvage their rep when it looks as if their boy won't make the cut, after all.

When he didn't have anything else, shilol always pointed there.

Well, I guess now it's useful for something.

The last of his strongholds gone.

Anonymous said...

I have a theory.

In the last eight elections the better looking candidate won. During Bush v Kerry opponents were tied until Memogate. I maintain swing voters were confused by the fake accuracy of the memo and focused on the pictures every network was running which showed a narrow faced scruffy Kerry in fatigues and a young George Bush looking Hollywood handsome in a pilot uniform. When I saw those pictures I knew Bush would win.

I'm not sure who is better looking here. Obama is a handsome guy but he looks so angry and tired, and Romney is handsome as well, but not really sexy. During the debate, Romney certainly looked like the guy Hollywood would cast as president. Obama seemed to have lost his charm.

Obama v. McCain
Bush v. Kerry
Bush v. Gore
Clinton v Dole
Clinton v HW Bush
HW Bush v Dukakis
Reagan v. mondale
Reagan v Carter

Anonymous said...

I hate woodward. I hate him.

Obama is the best. DO NOT Listen to anyone.

http://www.chicagotribune.com/news/opinion/ct-oped-1009-byrne-20121009,0,4232146.column

Anonymous said...

Check out this singer, who will never be employed. If you want a job in Hollywood or Nashville, vote for POTUS Obama.

https://www.youtube.com/watch?feature=player_embedded&v=dS1N6bLriRc

McTriumph said...

Polls tell you only one thing, trends.

Seeing Red said...

Tax Prof Blog has an interesting post that the worst states to do biz in are blue.

Wisconsin is on the list @ #43

The Tax Foundation today released the 2013 State Business Tax Climate Index (10th ed.) which ranks the fifty states according to five indices: corporate tax, individual income tax, sales tax, unemployment insurance tax, and property tax. Here are the ten states with the best and worst business tax climates:



Interestingly, all ten of the states with the worst business tax climates voted for Barack Obama in the 2008 presidential election, and five of the ten states with the best business tax climates voted for John McCain (and eight of the ten voted for George Bush in 2004).

campy said...
This comment has been removed by the author.
campy said...

How do they poll the trunks of democrats' cars? 'Cause there are lots of votes to be found there.

Brennan said...

You will hear this talking point over the next 30 days so get used to it.

Black voters are late to decide if they will vote.

Latino voters break for a candidate in the final week before ballots are cast.

This will be cited to explain Obama’a decline in polling. It will be hyped as his only chance top win.

What the media will not report is that Chicago cannot find 1 of every 3 Obama voters from 2008.

Nate Silver’s modeling does not factor any of this. Get used to him talking about sports.

Writ Small said...

I look forward to the day when someone says, "While it was no Obama-vs-Romney, candidate X really had his ass handed to him in last night's debate."

@Gabriel Hanna - nice summary.

edutcher said...

Brennan said...

What the media will not report is that Chicago cannot find 1 of every 3 Obama voters from 2008.

that's a problem in a lot of big cities.

Cleveland, f'rinstance.

somefeller said...

edutcher mumbles:You can take the RCP average and cram it someplace. These guys have featured the most skewed polls around, but now are trying to salvage their rep when it looks as if their boy won't make the cut, after all.

Real Clear Politics isn't known for its left-wing bias and is a go-to stop for people who are knowledgeable about politics, regardless of their political stripe. They certainly don't have to worry about their rep, especially from the unemployed loser portion of the electorate.

The takeaway here is simple: Obama had a lead. And now he doesn't. Maybe he'll regain his lead, but the momentum as of today is with Romney. No skewed or unskewed polling issues, just the movement back and forth of that part of the electorate that hasn't made a final decision on who to support.

Hagar said...

This years election is not like the past several elections, and the polls are floundering.

Ignorance is Bliss said...

Blogger Hari said...

Can anyone actually define "margin of error" as used by pollsters?

The margin of error is a standard statistical concept. It can be determined by comparing the sample size compared to the total population from which the sample is drawn. They typically specify a ~95% confidence interval, meaning that if you conducted 20 polls, you would expect 19 to be within the margin of error, and one to be outside.

However, the most important thing to note is that the margin of error assumes that there is no systematic bias introduced by the sampling method. Any error introduced by the sampling method ( and there is always some ) is in addition to the stated margin of error.

NotWhoIUsedtoBe said...

Keep in mind that McCain was briefly up, too. Not this late, but it didn't last.

Joe Biden will be declared the debate winner. It doesn't matter what actually happens- he'll win.

Don't be surprised, or depressed, when this happens. Biden must win or Obama might lose, after all.

steve uhr said...

Actually it is not 0.7% but rather 0.7 percentage points. Amazing how many commentators (and law profs) don't understand the critical difference.

steve uhr said...
This comment has been removed by the author.
gadfly said...

There is some sort of weighting going on, because arithmetically averaging the polls that include 9/30/12 do not come to +.7 Romney.

Ender's Ghost said...

Now all of a sudden Althouse is looking at polls other than Rasmussen and Gallup, the two polling places that heavily lean Republican, because there are now other polls that favor her candidate. Funny how that works!

Seeing Red said...

Via Vodkapundit:

We have a couple more strange polls to look at this morning. The first one is CNN’s headline poll of Ohio voters, showing Obama with a four-point lead. What’s the catch? Almost one in five self-described conservatives say they’ll vote for Obama. And to that, the next question should be, “Are you freaking kidding me?” There isn’t another poll in the nation with a cross tab anything like that, and it make the entire poll suspect. Suspect? No — useless. Nolte has the full breakdown.

Matt Sablan said...

Man. And I'm not even able to be on Twitter these nights to watch the epic meltdowns.

Seeing Red said...

Socialism kills, free markets feed, via Rantburg:

A talented head cook at a school in central Sweden has been told to stop baking fresh bread and to cut back on her wide-ranging veggie buffets because it was unfair that students at other schools didn't have access to the unusually tasty offerings.

Annika Eriksson, a lunch lady at a school in Falun, was told that her cooking is just too good.

Jason said...

So now that polls favor Romney, the liberal reporters are starting to come out in public saying it should be illegal to publish poll numbers.

http://www.newsbusters.org./blogs/noel-sheppard/2012/10/10/rolling-stones-taibbi-it-should-be-illegal-publish-poll-numbers