I agree!  Now, I have something that I want to show you over here...in my basement.

Please come along...

There are not overrated if they have recent data .

That said the recent state by state polls that have come out recently have shown closer races , so I really don't doubt their is both national and state level movement .

Over confidence and hubris was a negative for the Clinton campaign , voters don't like it .

I think the same thing is at play here .

I agree with you, except I like the electoral maps better when they look worse for Obama.  Obama is BAD! BAD! BAD!

Electoral maps may be overrated, but national polls mean nothing at all, so they're the best we have.

That's funny, I could've sworn electoral college polls meant EVERYTHING! Right, Senator Bayh? I seem to remember you and your compatriots prattling on about battleground states and electoral college polls back when Hillary was down in the national polls.

Why don't you just grow a pair and state that you're rooting for Obama to lose. Instead, you try to be all coy about it, posting a diary insisting that the polls showing Obama winning are "overrated" and that tracking polls are all that matters (the day that he's down in one such poll).

What a sad coward you are.

But the problem with electoral college maps is that the most recent polls used to represent each state have been conducted before the most relevant breaking news that as we all know, can change the electoral snapshot in a matter of days.

It can...if there's a trend. Which daily/weekly state by state or even national polls don't take into account whatsoever.

Like it or not, the electoral map is where its at, not what Rasmussen's saying about a sample size of 1655 likely voters from Arizona during the first week of August.

Your premise is that national polls are a more current version of state polls, when nothing could be farther from the truth.

Lets say state polls were done as frequently as national polls. McCain is up +1 in the national poll. By your logic:

He would be up 1 in California, and
Up 1 in Texas

Obviously that's silly. The two candidates will lead each other based on the widely varying demographics of the individual states. Furthermore, a lead in each state is not equivalent to a lead in any other state. +1 in Rhode Island is NOT the same as +1 in Florida.

So the actual situation, the probability of each candidate winning a particular share of EV's, is not represented in the national poll. IF we ran the general election based solely on popular vote, your position might have merit.

I will agree that electoral maps lag, and we have to be careful relying on them too much. But tracking polls are not a substitute.

State polls are idiocy. I don't know anyone who successfully plays the political market sites who is married to state polls.

You'd have to be a dunce. State polls are notorious to err in specific directions depending on the state. For example, if anyone wants to believe Alaska polls at face value they don't have two functioning brain cells. That state is hellish to poll and always tilts heavily Democratic.

National poll averaging is by far the most reliable measure, combined with logical application of how individual states are likely to fall in line with that national average, based on their typical partisanship adjusted with relevant current variables. I used that method to win a 16 man political betting pool in '96 and '98. Only my runaway stupidity prevented a win in 2000. I had bragged about my method post '98 and others in the pool successfully copied it.

Previously those clowns were slaves to state polls, believing everything they spit out and swallowing whole the media/internet pronouncement that state polls were more relevant than national polls. That is true of primaries. In the general election you need to throw away everything from the spring. Deny that if you must. Eventually you'll learn.

And yes, electoral maps are overrated. Virtually every one relies on state polls so they are flawed by definition.

I will say this about 2008: The partisan index is fuzzy with a variable like first black. Same thing if we had nominated a woman.

You're going to have state outcomes that vary differently than typical from the national margin. In that regard, it's going to require more subjectivity going forward. Did that state tilt several points red from its typical partisan index based on a foundational shift, or was it reluctant to vote for an African-American?

But I'm very confident the state polls generally overstate the variance from typical partisan index realities. For example, during the primaries we had many state polls asserting Hillary and Obama might fare 8-15 points differently vs. McCain. That's laughable. Everything tends to drift back to the beginning. No chance a state will veer that dramatically from its raw partisanship, based on which one of two Democrats is the nominee vs. a specific Republican. Truth is a few points tops, other than a home state where the gap obviously will be decisive.

If anyone wants verification that state polls are less reliable than national polls, check the betting limits. That's always an indication of how confident oddsmakers are in the numbers. For example, NFL wagering limits are sky high compared to obscure events that the oddsmakers concede they don't have a good feel for. Right now many offshore outfits will gamble huge on the presidential outcome. The odds of Obama as roughly a 63-37 favorite have been established and you can plunk down virtually anything you want. Now try to get down heavy on state outcomes. Best of luck. Not many places are offering them and the limits are peanuts when you can find them. I'm talking about wagering against the house, not man vs. man similar to Intrade and Iowa Futures Market.

When they don't show the result you want.