Also interesting is the Gallup unemployed number - 7.5%

romney now ahead.

And, we are working the Thursday and Friday bedwetting numbers out of our system. Expect it to get a little worse then get a whole lot better.

... next debate, and people will show that they do not want to get rid of him.

It's a marathon. I love that the R's are (1) running around on an adrenaline high right now ("Lordy, Lordy, the polls aren't skewed anymore!) because once the real world comes back to meet them (their candidate still sucks and Mitts numbers are going to drop) the adrenaline wears off and you crash and (2) not seeing that once the Thursday and Friday numbers are gone, then Romney's fade will be fast and probably too late for him to recover.

This is our Battle of the Bulge. Stand fast, fellow democrats!

We can't have another debate performance where Obama looks like Jason Bay out there (That's a Mets fan inside joke for those of you who don't get it).

This couldn've been a landslide for Obama.

Unfortunately it will now be a close election.

Something was going to be cited as a reason for things to tighten up. Sit tight, the skies are clearing in a couple of days.

It could change with another debate or happening.

they were deluding themselves!

Are people who annoy me, but that aside, how does the pollster figure that?

Just asking them? I'm registered, I'm for Obama, but not likely to vote. What kind of an idiot says that?

Though I agree with you re: non-voters who are eligible to vote.

Gallup uses a series of eight (iirc) questions that have historically been good predictors of whether or not someone is likely to vote. Do you know where your polling place is? Did you vote in the last election (if eligible)? They then rank people on a scale of how likely they are to vote and take a sliding percentage from each category (because in reality, even fewer people vote than those who pass the LV screen).

Last cycle, I think they modified it slightly to account for the fact that more younger voters were indicating that they intended to vote. I don't know whether they retained that adjusted model.

A couple of recent elections have had a few people questioning whether their results are correct, but they've tightened up just prior to the election, so it's hard to say.

However... they are historically the gold standard for this. Almost everyone else essentially uses their model with minor tweaking.

And again... if anything, there are more people who get through their screen than actually vote on election day.

They were when they were the only game in town.

Gallup does have a good approach in using the 7 day roll. But their screens are mainly historical - they only new blood they bring in is based on their funky enthusiasm questions.

Did you have an alternative?

There certainly aren't any LV screens that I'm aware of that do anything much beyond comparing prior answers to these decades-old questions to prior turnout.

For the record, the five point gap between RV/LV is actually two points tighter than it was in August. Note also that Silver identified at that point that the LV/RV gap for other polls was also much larger than normal.
http://fivethirtyeight.blogs.nytimes.com/2012/08/25/aug-25-an-above-average-likely-voter-gap-for-romney/

Oh... and while we're on the topic, please remember that we saw very similar comments two years ago (that the enthusiasm gap was a fictional artifact of the polling)... and Gallup turned out to be right.

LV screens change over time. Everybody sit tight and watch this play out over the next week. It'll be fascinating.

In the three days leading up to the debate, Obama led Romney among registered voters by five percent, 50-45. In the three days following the debate, they found the race tied at 47 percent and in the the two days since have found Obama retaking his 50-45% lead.

Keep in mind that's among registered voters, so the numbers will be tighter among likely voters. But it's clear Romney's bump is temporary.