Pollsters are diddling with "Likely Voter" models.

[longship]

The recent changes in the polls are not real. In the eleventh hour, the pollsters are tweaking their analysis models. Specifically, they are upping the number of likely Repuke voters. That's what's causing the uptick, not anything in the data.

They are basing this change on historic data of tightening races and of the recent negative ads by the Repukes. What is important to remember about this is that it is not based on any current measurement but on the rather arbitrary criteria that this election must be close, just like all the recent previous ones. Of course, this sells many more newspapers and dishwashing detergent ads on TV. That's what this is all about, not about any real effect.

Ignore all the candidate polls now

[C_U_L8R]

1. don't they know 911 changed everything !?

ack.. hairball

[DaveJ]

2. Hasn't 'likely voters' always been factored into the results?

Also, I did not notice the uptick on http://www.electoral-vote.com.

[Gormy Cuss]

3. If republicans are more likely to vote, it's not diddling with the numbers

It's a necessary tweak. The press reporting that tweaking as an upswing for Republicans is what's wrong in that case.

It is true that with tight races a certain number of disillusioned voters-- those who earlier claimed they would vote for the candidate from the opposing party -- end up voting along party lines. The historic effect of negative ads, I don't know how well documented their effect is late in the campaign.

So, the pollsters may simply be taking the right steps to adjust the estimations.

Or you could be right --- it's about selling more stuff on TV.

[David__77]

4. I don't think Pew did. But it doesn't matter either way.

Polls are by nature "all over the place." Make a quick reasoned analysis of available polls and you'll see Democrats are in a very good position, the best since at least 1992.

[longship]

6. The Pew poll was why I posted this.

This morning, one of the prognosticators (looking for citation, but not finding it) cited the internals on this poll and decried that Pew somewhat arbitrarily upped the ante for the Repukes strictly based on historic data, not on any actual information about this year.

Let's get this straight right now. There is no reason to expect that this election cycle is going to be in any way like previous ones. If a pollster changes their analysis model based on such an arbitrary criterion as "Republicans always close strong", that pollster is not doing their job.

This election cycle one cannot ignore the global indicators which are dramatically and monolithically favoring Democrats. It very well may be that the Democrats and those very pissed-off Republicans and Independents are going to be the ones who will finish strong.

The question is, What criteria do the pollsters have in changing their models in the eleventh hour? If it is merely "Repukes finish strongly." I must reject that. If there is actual data to indicate that this time that would be the *only* valid reason to change the model. What I am hearing is that there are no such indications this election.

The only conclusion I can reach is that the pollsters are diddling with their models for the sole purpose of showing a tightening race.

[David__77]

7. What I meant to point out...

...is that Pew applied the same methodology to the poll previous to its latest one. That previous poll showed a larger Democratic lead. But I say: so what? There are polls moving in opposite directions. All of them show Democrats on top. It's a good, good position to be in.

[longship]

8. Agreed.

Nothing has changed, but the poll models.

If Ford was ahead of Corker this last weekend, he'll be still ahead of him tomorrow. We're not going to know until tomorrow night.

You are correct:

Ignore the polls and vote

[kid a]

5. just vote