National polling data sites analyzed

Because I am a news junkee in some respects, but one who tries to find the middle ground in reporting through a series of balancing acts, I am always on the lookout for what could be correct or indicative polling data. For the past few weeks, I have focused much of my attention on the Real Clear Politics poll and Rasmussen's data. Apparently, my reliance on Real Clear Politics is a little misplaced.

Granted, this site has its own biases also, but I thought the post was worth reading. See FiveThirtyEight: Real Credibility Problems (discussing the inclusion and exclusion of certain polling data from the RCP poll) and his follow-up post, RCP Follow-Up. He also pointed me to this Pollster.com polling site, which seems to say what every other poll is saying: the Republican ship is sinking, and this election is not going to turn on one state.

Whether or not any of these polls have any prescient accuracy remains to be seen, but the point, as always, is to look at these things with a grain of salt and a skeptic eye. My prediction remains rather general: Voter turnout will be through the roof, and none of these polls are going to get it exactly right.

This election is going to be over with in a couple of weeks. Whether poll A or B or C puts Obama ahead or behind, or McCain ahead or behind in other states is irrelevant. All of this voter fraud nonsense is the same swan song that is played during every election. See also An Analysis of Voter Fraud in the United States. The system is not perfect. But it's the system we've got and that's the way it is.

The bottom line is that all of these polls have significant margins of error and none purport to give what is required to be scientifically statistically accurate. This has more to do with cost efficiencies more than anything else, and in the end it's just a conversation piece that adds fuel to the fire of 24/7 media, a recurring problem I have discussed many times over the past few months.

I will be curious to see whether my own prediction is going to come true - so far it seems on track, although I certainly miscalculated the breadth and success of Obama's 50-state strategy. We'll see how exactly it plays out in two weeks, and obviously it is going to impact future elections regardless of who wins.

No real point to this entry, just thought the article regarding the RCP polling information was interesting. Why can't all of these polling sites simply disclose their information? I'm more interested in how much money these sort of businesses actually make, but I can't seem to find that anywhere.

Tags: election polls, accuracy of election polling data, rcp poll, real clear politics poll only semi-accurate