I was thinking about some of these state polls and how the one to really keep an eye on are the national head-to-head polls. In almost every election, the popular vote winner has also won the electoral college vote. I would think that these national polls are a more important indicatior of who will win the election.
We saw in 2000 how Gore won the popular vote by .5% but lost the electoral college by the slimmest of margins. Going by that, would it be safe to assume that any popular vote victory over .5% would guarantee an electoral victory? Basically that means that if one candidate won by .5% or more, a majority of the electoral votes from the close states like NM,IA,FL,OH would end up voting for the winner.
I'm not a statistician, but I think the odds of a candidate winning the popular vote by let say 1.5%, yet losing the electoral college would be pretty slim. Therefore I think I'm going to focus on the national polls, especially those by reputable organizations like Zogby.
Focus on job approval ratings. They always work in presidential re-election contests. 50% or more and it's a win. Less than that and it's iffy.