I'm happy Obama has done better but I still think ARG is crap and don't trust them (although i think someone pointed out they started in NH and have been accurate there, but I might be confusing them with another polling place)

Why is it a fraud . You have to have real reasons for that . Have they accurately predicted in the past.

but I do think their voter screen for the Iowa caucuses is way off. Turnout may well be below 20 percent in Iowa.

ARG polls have shown Hillary around 30 percent in Iowa all year, but I can't recall another pollster that has ever found her with 30 percent here. It doesn't seem right to me.

But hey, if the Hillary boosters want to raise her expectations by saying she's leading Edwards in Iowa, I guess that's ok by me.

I haven't seen any evidence that they are "clearly pushing Hillary", but I'll agree that these numbers are quite different from the others that I've seen. Personally, I'm a big fan of Zogby numbers, as they have an excellent historical record of accuracy. IIRC, Thom Hartmann is a fan also.

Terrible to even bring up Kerry leading in those places as an example, when it's clear that KERRY DID WIN THOSE STATES.

The exit polls proved that.

It's much more intelligent to compare ARG to all other polls in Iowa.  Why is ARG so different from EVERY OTHER POLL that comes out of Iowa?

Because it's clearly doing something to inflate Hillary Clinton's numbers there.

... if Edwards turns this around and squeaks out a narrow victory in Iowa, it will be a miracle. Given Richardson's surge, it will be an even bigger miracle if he finishes anywhere above 4th in New Hampshire.

We Edwards supporters will keep doggedly on, because we believe in the issues he represents, but we also have to be realists.

8-)#

MOE, don't forget that a poll diary without MOE is noise. Indeed, it is best to simply give the range from the central number minus MOE to the central number plus MOE, so statistical leads can be easily distinguished from statistical ties.

I have never held ARG in high regard and I will not start doing it now just because it shows favorable results for my candidate.

I like to see those numbers for Obama but I still don't trust ARG but the General trend is correct.

I know that Obama benefited from the spat but by how much, I can't say.

As I commented earlier, averaging polls like Pollster does is a joke because a poll is a measure in time. Pollster's methodology creates too many variables which are not accounted for in the numbers.

On the average good numbers for Obama.  Even in Iowa based on ARG numbers.

Which are the credible polls

I'm down with that.  Fill in the undecided's... Clinton takes third?  I'd be down with that.  

Ah the details say NH is +/-4% "if opinion is evenly divided".

The three earlier from Wikipedia,

ARG (Date): Clinton/Obama/Edwards
July 26-30: 27 to 35 / 27 to 35 / 10 to 18
June 27-30: 30 to 38 / 21 to 29 / 7 to 15
May 23-25: 30 to 38 / 11 to 19 / 12 to 22
April 27-30: 33 to 41 / 22 to 30 / 10 to 18

Oddly:

• for Hillary, 33 to 35 lies within all four ranges
  
• for Edwards, 12 to 15 lies within all four ranges
  
• For Obama, his first and second range do not overlap, his second and third do not overlap, and his third and fourth only barely overlap.

That's a 95% confidence interval, so a roughly 1 in 20 chance for any individual estimate being an outlyer ... and 12 estimates observed. One of those Obama results having a "true" value outside the confidence interval is certainly not implausible, but which one ... one of the two high ones, or one of the two low ones?

I guess its a good thing they still actually hold the elections rather than deciding it based on opinion polling.

Being progressive on foreign policy.

I think Obama had the best week of his campaign, and Bill Clinton knew it too, but let's not use ARG as a reason to jump for joy.

Those aren't trends. They are three separate races, so they have three separate trends.

ARG July 2007: IA, NH, SC
Clinton: -2%, -3%, -8%
Obama: +2%, +6%, +12%
Edwards: -8%, +3%, -4%
Richardson: +8%, +1%, +1%

In my considered opinion, Clinton is losing a bit of ground in all three because of Cleavagegate, Obama is kicking in NH and SC because of Obama Girl, and Edwards sagged in Iowa because he spent fewer than two weeks there in July and they are pining for him. Richardson zoomed in Iowa because for some reason they understood his YouTube clip when nobody outside the genuine Midwest really got it.

Obama will have a tough time winning anywhere other than Illinois when the actual votes are counted.

You all like citing history, well, how many polls say that people will vote for a minority candidate and then on election day they do not?

Fact is, I think many Democrats will be weary of Obama and Clinton for fear that the rest of the country won't elect them.  They won't hold his race against him, they'll just perform their own personal risk assessments and determine that it's too big of a risk to take, and do the same for Hillary Clinton who is so hated by 50% of the country that nominating her would be suicide.

Should Democratic voters think about whether the country would actually vote for whoever they nominate?  

As for John Edwards, he probably expects to win Iowa, and then ride the wave to a win in union-heavy Nevada.  After that, Hillary Clinton and Barack Obama don't look so "inevitable" and they immediately lose their bandwagon support in New Hampshire and SC.    

Now, Edwards will have to contend with an establishment media, which cannot stand him, and the media will begin using reverse psychology, claiming that Democrats claim to embrace diversity but they are about to nominate another "southern, white man."

Get ready John Edwards, it's coming.

Edwards plan is to win Iowa and then count on major union endorsement to carry him through NH and SC.

The problem with that is Edwards has a huge "diversity" problem when you take a look at his base of support.

Even if Edwards wins Iowa, i just cant see how in the world in wins S.C with so many AA being part of the S.C democratic electorate.

Edwards will finish 3rd among black voters even if he wins Iowa and NH...Black voters have long written out Edwards and if you listen to them, you'd think that Edwards did not exist.

Even if Edwards hangs on to Iowa by a few points, he will have to close the gap in NH now so that his Iowa-boost of 10-13 points would be enough to overtake the frontrunner in NH.

Right now,Clinton leads Edwards by an average of about 20 points,therefore,an Iowa boost may not be enough for Edwards unless it's a blow-out win.

I dont think NV will go second and my guess is NH will leapfrog them somehow but even if that doesn happen,Edwards is trailing the NV fronrunner by even larger margin then the NH frontrunner,so i dont buy it that Edwards path to the nomination will easy if he wins Iowa.

The problem with Edwards is, he has to spend money in NH and NV right now, but he also doesnt have that much and needs to save his cash for the final 2 weeks going into caucus night because Obama and Hillary are both expected to unload a large amount of cash there.

If Edwards cant move up his poll numbers in NH, then a victory in Iowa is almost meaningless.

ARG in Iowa with thier likely voter model.....I think means nothing, I don't think much of thier polling anywhere else but in Iowa it's a joke, I believe the ones that have Edwards in front with Hillary and Obama a little behind, the demmoines register should be interesting when they come out........don't disregard the crowds and that Jhonson county thing for Obama either, he's still winning every non passive way of measuring support.

It appears Richardson's voters are coming from Edwards. This poll shows a 2 pt bumb for Obama from their poll last month.

Again don't make too much of this poll. They have seemed to be the outlier since Dec. 2006 with respect to IA,

The question is their accuracy in Caucus and primary states not general election states.

My question is what is their track record for the IA caucus. Not their tracking poll.

No, those talking about ARG being an "OUTLIER" poll are exactly right.  ARG has had Hillary Clinton winning everywhere since the beginning.  Who in the world are they polling in Iowa?  Clearly not the same people that the other polling companies are polling there.

... once.