So much for polls!

[imanamerican63]

We are looking at a positive change, but will the other states follow the trend?

[zackymilly]

1. Only the Shadow (Inc.) knows! nt

[Quixote1818]

3. Good one.

[TwilightZone]

2. Polls don't work in a direct manner for caucuses.

They only provide a snapshot of the intended first vote and can't/don't account for where supporters of non-viable candidates end up.

That's one of many reasons why focusing on them too heavily was a big misguided.

[The Velveteen Ocelot]

4. Iowa is a small, rural, almost totally white state

with caucuses that exclude anyone who can't spend a whole evening wandering around a high school gym - people who can't get babysitters, people who can't get transportation, the disabled, people who have to work, many elderly, etc. Iowa doesn't mean a damn thing, and especially not after this debacle.

[2naSalit]

5. +1

[Igel]

7. But with caucuses it's always a different matter.

You go ready to vote for A, and your neighbor makes her case for B. Maybe you stick with A, maybe you don't--but if A doesn't come out on top you get to change your mind.

If one-person/one-vote, to be sure, but with a few twists. If there were ballot boxes, people file in and don't argue about who to vote for--they vote and leave.

So not only does accessibility provide constraints, but the process itself makes polling, which assumes that everybody deciding things in the shower will implement those decisions in the ballot box, problematic.

As for "almost totally white," the proof of the pudding is in the tasting. Iowa's got a good track record for helping to predict ultimate winners of contested primaries. There's a confound--it's not like Iowa doesn't effect other states' voters' choices--but focusing on one trait as the most socially important one when arguably results are more important doesn't seem all that important.

[blueinredohio]

6. I never pay attention to polls.