My take: if these numbers are true Hillary is in control (I note my prediction of last week holds). Most Richardson/Biden voters will go for her when they can't muster 15%.

I have read that predicted in a couple of places recently and will link to one here:

http://www.huffingtonpost.com/chase-mart yn/bush-fatigue-unlikely-t_b_75578.html

"While some 125,000 Democrats braved the cold Iowa weather to caucus in 2004 (a sharp increase from the 2000 Democratic caucuses), David Redlawsk, University of Iowa Political Science professor and the Acting Johnson County Democratic Chairman in 2004, says he does not expect that number to increase this year and believes that Iowa may have already reached its turnout peak."

Much more likely to be in 120-130,000 range.

Political Wire indicates a new SV poll out tomorrow will show Obama at 30% with Clinton and Edwards tied at 27%

http://politicalwire.com/archives/2007/1 2/20/another_iowa_poll_shows_deadlocked_ race.html

Two weeks to go and it's tied, not its tied.    

that if there is a clear cut winner for the Dems in Iowa -We will never examine an Iowa poll again.

So who gets "Other's" second choices?

I don't see why to would expect a surge in turnout. 2004 was a massive increase over 2000 but in 2000, there was not a competitive election in a lot of parts of the state, since Bradley didn't contest a lot of counties. Its not as if there is a trend of increasing turnout over the past few cycles.

That said, what does seem to me different from the past -- and this is from a distance -- is the density of field staff for the top 3 campaigns. That suggests to me, again speculating from 1000 miles away, that there are going to be many fewer "unorganized" voters who are not on any campaign's list of identified supporters -- ie, very few actual undecided voters or voters supporting third tier candidates or voters who show up of their own accord. Based on what Mark Mellman has written about Kerry 04, it was "unorganized" voters who put him over the top -- who had not been identified as supporters by the campaign.

I think therefore its less likely there will be anyone breaking out of the pack -- again, thats based on looking at the math, not based on any feel of whats going on on the ground.

To me, this means that breadth, not depth, will carry the day. In other words, I think its less important to guess overall turnout than to guess at whose got a broad bases support across the state -- and where those bases overlap.

Finally, in terms of 2nd-choice candidates of third-tier candidates, someone suggests upthread that Richardson supporters would go to Clinton. I'm sure Richardson would like that, but from wher e I sit, watching Richardson call for an immediate withdrawl, I'm not sure how his support gravitates to Clinton.

are right, because Obama has around 40,000 committed caucus goers right now... Surely, he would pick up a few more on caucus night.

its is right, cause nobody likes to hold down the shift key

Even if the final results are very close (and I doubt they will be, for the record), it makes a big difference to Edwards whether he finishes first or third.

If Edwards finishes third, even if he's very close behind the other two, the media will probably continue to ignore him.  They've been pushing the narrative of a two-candidate race for so long that it's really going to take a major development to force them to switch.

On the other hand, if Edwards finishes first, the media really has little choice but to talk about him - even if it's a close victory.  The $64,000 question is how many voters in other states are open to considering Edwards, once they realize he's a viable choice.  But I really think he has to crash the gates for that to happen.

Not sure I buy that turnout scheme. It's going to come down to the most passionate supporters. Now it may very well be true that lower turnout helps Edwards, higher turnout helps Obama, and for Clinton that could go either way, but whether it's 120,000 or 170,000 anyone can still win it. Iowa voters can change their minds late in the game.

Right now I think there's a kind of momentum behind Edwards. His second choice numbers time and again have been better than anyone else in these Iowa polls. Right now I think it goes Edwards, Clinton Obama, with maybe no more than a 10% difference between first and third. It's going to be a horserace.

after being massively outspent, massively out news-mediaed and writtn off by the MSM, then he wins.

Huge.

McCain is the sleeper on the Republican side.  Not to be confused with Thompson, who is the snoozer.

I'm predicting a strong third or even second-place showing for McCain.

Jerome good post. the numbers are less relevant than the type of increased turnout, if Hillary suceeds in getting older women out much more than Obama does getting highschool students and indies  or vice versa it decides who wins.

New numbers from ARG, noticibly different:

http://politicalwire.com/archives/2007/1 2/20/new_iowa_new_hampshire_polls_show_d ifferent_results.html

no they are trying to create mo for McCain.....no way in hell an honest or competent pollster could find more support among Iowa caucus goers for McCain than Romney.