Breaking News!  Eric again uses the phrase Obama Clings in a headline.  Opens himself up to reader abuse!

Sure glad he can take it!

Remember - McCain MAINTAINS, but Obama CLINGS.

Fuck you, Eric.

Eric, stop pissing everyone off! (including me).

"Clings to?" Really? I don't see it. None of these polls measures ground org, but most importantly, none of them were taken on Nov 4.

Besides statistical noise what explanation do you have for Obama's massive collapse.

I just have to say: I'm disappointed in my fellow commenters. Eric put out the bait, and y'all just fell for it.

Is this just a subliminal attempt to reassert the whole "clings to guns/religion" meme? I ask you, Morton Kondracke!

Well, as Eric said the other day, the change could be due to statistical noise, or it could be due to something else.

I mean, that's some cogent analysis for you right there!

When McCain was ahead.... it was

"McCain Leaps to a 3 point Lead!!!!!!!'

When obama is ahead... is it

"Obama CLINGS to a narrow lead."

Excuse me, you puke...... but there is not a major national poll that shows McCain leading...and the polls are showing a 1 to 7 point lead. You keep figuring in that RASSMUSSEN sh*t poll that is skewed with too many Republican voters and McCain winning 18% of the African American Vote.

Your true colors nauseate me.

Somewhat off topic, but how is it that Missouri is not in that list of "battleground states" to our right? The polls here are just as close as they are in FL or CO or OR and we have more electoral votes than those last two.

i think the verb cling is accurate. three of the polls are more tight today than they were yesterday. and the lead is slim. what's the big deal?

deep inhale... now, full exhale. c'mon people

I am wondering why Eric has still this job, he has no competence. He can't analyze a poll...He is plain dumb...

Eric just loves the attention.

In the McCain Administration that Eric is gleefully helping to usher in, I hope he gets draft and shipped off to Iran.

Well played, Eric, well played.

How does Obama's lead not increase, if not stay the same, given McCain's monumental flip flops with the financial scandal? McCain changes lifelong positions on a dime and we're supposed to believe him? McCain thinks our economy's fundamentals are 'strong', his words, not mine, and we're supposed to trust anything he has to say about the economy? The gap should have doubled.

We're looking at weekend numbers. No worries!

I just wish Barack Obama would concede my pathetic state of Florida. It would be so cool if he would say, "we're not putting another dollar in a state that can't run itself."

He's not going to win here, whether he gets 100% of the votes or not. Please don't waste your money.

I'd like to see a more targeted narrative on just how extreme the Bush administration is and how they've essentially ruined the country. McCain has been Bush's right hand man through all of this.

I want to hear about 9-11, Iraq (yes, go there), the plunging economy.

Really, I think Americans need to be reminded about the disastrous 5 years that we have spent in Iraq. Nearly a trillion taxpayer dollars, thousands of soldiers killed, tens of thousands of Iraqi civilians killed. And what are the results? Better than Iraq under Saddam?

And, Obama needs to make national security more of a centerpiece to his campaign.

I think Eric does this just to get y'all riled up. And it works.

Anyway, this was a mini-bounce for Obama. And mini-bounces, like all bounces, recede. The first few days of the crisis the market was going crazy and it was all over the news. It's still the top news story but it's not as crisis-oriented as it was last week and we're starting to get into the nitty gritty of the bailout deal. People don't get fired up about details. The first few days scared the crap out of them and they ran to Obama, the guy they trust on the economy. Things are starting to settle down and some of the undecideds who switched briefly to Obama have switched back to undecided. They like Obama's initial response but now they're waiting to see a) how he reacts to the final bailout deal and b) his detailed plan for working within that deal if he's elected president. They're reserving judgement. The fact that they went to him in the crisis moment is a good thing - they can easily be won back.

The assumption is that the states, just like they lagged in reflecting the bounce, will also recede next week, like the national polls. That depends. We have a debate coming up. If Obama ties or wins that debate, especially since it's about McCain's alleged strong suit, he'll get another bounce, negating any recession of this mini-bounce. He needs to tie national security to economic security as much as possible in that debate. And he needs to keep the answers short and to the point, because lord knows McCain is just going to spew out lines from his stump speech and the standard "POW" BS.

For Economides:  Sorry for taking so long to get back to you.  Other threads have been busy :-)

Why is sampling variability newsworthy?

I'm saying the daily posting of the tracking polls is worthwhile.  Whether they show large movement, small movement, or no movement is, well, it's up to the numbers.

You are arguing that the appearance of movement when no movement has occurred is interesting.

The numbers are interesting.  In this case, it's likely that a small movement has occurred.  If no movement had occurred, that would have been useful to know, too.

I'm sure you realize you have no basis for claiming that no movement has occurred.

Or you are arguing that no movement at all is newsworthy because for some unexplained reason you *expect* movement.

I mostly don't know what to expect.  I wouldn't be surprised by a small movement in either direction or by no movement at all.  A big movement in either direction would get me to scratch my head.

Maybe you can explain why you expect movement of fractions of % back and forth from day to day.

Again, my expectations are irrelevant.  The numbers say what they say.  In this case, they say there's been a small movement down for BigO over the last three days.  And, if you do the math, you'll find it's well outside the MoE -- i.e., the expected noise from statistical variability.

What is your theory as to why preferences would change back and forth by such small amounts?

The numbers don't say why people change their minds.  But if you look at everything else in the world around you, perfect constancy is much more rare than variability.  Sampling variability does play a part, but with big enough samples, precision greater than 1% is readily possible.

Trouble is that is't not linear with sample size.  You'd have to quadruple the sample size to double the precision.  That's costly.

Random movement that happens to be in the same direction is still random, right?

Yes.  But you are presupposing it's random without any basis for doing so.

If you don't know a priori that they're random, then three movements in the same direction is mild evidence for them not being random.

Your mind is playing tricks on you.

It's happened before.  But in this case, I think you're not being rigorous in your logic and math.  You're conflating whether a small movement is real (not an artifact of the method) with whether it's important and/or worthy of a post.

As for the coin-toss, of course it's 50-50 -- providing your assumption about the tosses being truly random is correct.  But if you see 80 heads and 20 tails, you'd probably give a lot of consideration to the possibility that something is biasing the outcome.  With 8 heads and 2 tails, maybe only a little consideration.

And thanks for the advice about day trading.  I've dabbled a bit in options, and I've lost a bit of money by too often being right too early :-)

There've been times, though, when I've wished I had opened an Intrade account.  There are lots of suckers there waiting to be fleeced.  Well, not all the time, but there have been times when it's been painfully obvious.