Honestly, why bother parsing polls?

[cali]

It's clearly tight as can be in Iowa, and the results there will have an effect on NH, so why bother with that? 2004 polling should have taught us the futility of depending on polls. Maybe in the two or three days before Iowa polls will show us a clear leader, but I doubt it. The crowing over a 2 point rise here, and the jeering over a two point loss there, is about as silly as it gets. I don't get the obsessive reading of polls. It strikes me as about as pointless as reading tea leaves or chicken entrails.

[SteppingRazor]

1. It's doubly asinine because it takes the focus off what's important...

namely, policy.

Who gives a rat's ass whether one person is ahead or behind by a point or two? What's important is what each candidate intends to do as president. The horse race has taken over everything to the point where the average, cable-news-watching voter can't even tell you what the actual policy differences are between any of the major candidates.

[flashl]

2. Because it makes as much sense as the media creation of the Blue vs Red states craze?

[Tejanocrat]

3. 2 points. First, why say "2004 polling should have taught us the futility of depending on polls"?

The 2004 Iowa polls shortly before the caucus generally show Kerry winning, Edwards rising and either in 2nd place or within the margin of error from 2nd place, Dean falling and either in 3rd place of within the margin of error from 3rd place, and Gephardt in 4th place.

Second, I agree that national polling is all but useless at this time. It is useful only for the purpose of seeing that Hillary has a large lead outside the early caucus/primary states and that fact confirms that if she wins those early states she'll likely remain strong nationally, but if Obama or Edwards win the early states they still need to overcome Hillary's large national lead.

With that said, head-to-head polling of Democrats vs. Republicans is relevant because we need to know if one of the candidates we are considering will get slaughtered in the general election or whether one of our candidates slaughters the likely Republican nominee. This is information I want to know when making my choice. It's very relevant.

Finally, the in-state polling in Iowa is relevant to those who might be selecting a candidate for the caucus. Iowa caucus participants need to make choices and those who are supporting candidates who will be non-viable in their precinct will need to make further choices. It is true that polling is only a rough indicator of how the caucus will conclude, but polling is not without some validity. No one who's polling in the mid-20s is going to caucus in the single digits, and no one who's polling in the single digits isn't going to caucus in the mid-20s. Knowing that Richardson or Biden are unlikely to caucus in the mid-20s (unless the polls change radically in the next two weeks) may very well affect whether a caucus participant wants to support them as a first or second choice. As someone who may vote for Kucinich in my primary, I'm not saying that voting for a viable is the only factor to consider, but it's certainly one factor. the Iowa polls in the weeks leading up to the caucus provide some real (even if somewhat cloudy) insight into that matter.