Defense: States that John Kerry won

OK, let's tackle that SUSA Minnesota poll. The crosstabs show Obama and McCain tied among voters aged 18-34. That is just patently ridiculous. We'll just have to go back to the 6/11 Rasmussen poll which showed Obama leading by 13. Obama will likely win Minnesota by double digits.

The Oregon poll is weird too, also showing the race far closer than any other poll has. In that one, SUSA didn't even bother to show crosstabs by age, and after the Minnesota poll, that casts this poll very heavily into suspicion as well. The 6/11 Rasmussen poll from Oregon has Obama up by 8, which is very much in line with other polls done here.

SUSA's Wisconsin poll predictably showed Obama's strength with 18-34 voters, and also won every other age group. Too good to be true? Well, it falls in line with the U. of Wisconsin poll (6/8-6/10), although I find fault with the question order there. Even so, two consecutive polls in agreement suggest that Obama is doing quite well in Wisconsin.

The polls appear to suggest that McCain is doing well in Connecticut and New Jersey. No matter what the polls say, these states are highly unlikely to flip to the GOP. New Jersey in particular is infamous for seeing Democrats outperform the polls. And Connecticut, where civil unions are legal? That poll was taken before Hillary endorsed Obama - just forget it.

ARG and Rasmussen both released New Hampshire polls last week, and both show Obama winning by double digits. And that was supposed to be McCain's best chance at turning a blue state red. Pennsylvania looks good according to Quinnipiac, but Rasmussen will release a new poll there at 5:00 PM EDT today and I expect they'll show a closer margin, just as they did in Ohio and Florida (see below).

Red-to-Blue: States that John Kerry lost where Obama is leading or tied

I include Ohio here as "tied" because Quinnipiac and PPP released polls last week showing Obama up by 48/42 and 50/39 respectively. PPP, however, is a Democratic firm. Then again, one could make a case for Rasmussen being GOP-leaning, but they are officially non-partisan whereas PPP is officially partisan. Rasmussen had a very good track record for the 2004 presidential election, in any case, though that doesn't guarantee that they are operating without bias.

The Iowa poll seems surprisingly close. One has to wonder how accurate it could be considering it was conducted in the middle of severe flooding. Today's Rasmussen poll in New Mexico, not surprisingly, shows a nice Unity bump for Obama compared to last month's tied SUSA poll. Expect people who claimed that Latinos were racist and wouldn't vote for a black candidate to end up looking very very stupid.

Rasmussen's Virginia poll is confirmed by a PPP poll showing a 2 point lead for Obama. Meanwhile, there's nothing to compare Rasmussen's Colorado polls to, since no nonpartisan firm has polled this critical swing state since February's 50-state SUSA poll. I have no idea why that is, or whether or not the race in Colorado really is getting closer the way Rasmussen shows, but SUSA will be doing another round of 50-state polling within the next week or so, I'm told.

Frontiers: States where Obama is close behind McCain

All right, Georgia's newly added to this list, but don't get the vapors; we're talking about InsiderAdvantage here. They've been amazingly inaccurate in pretty much every Southern state this cycle. Don't believe them until another poll confirms it. Likewise, I'd ignore the very outdated North Dakota poll.

Also newly added is Alaska. Rasmussen's poll shows McCain's lead being cut in half since last month. Obama is advertising there, now. In a state as small and as usually forgotten as Alaska, showing up at all can have a big impact.

The Florida polls have been just as wacky as the Ohio ones. Quinnipiac and ARG both released polls last week showing Obama leading, by 47/43 and 49/44 respectively. Of course, ARG polls are even crappier than InsiderAdvantage. So if you average Rasmussen and Quinnipiac, you get a slight lead for McCain here, which would be my expectation.

Rasmussen's Nevada poll shows a 3-point McCain lead, confirming the Mason-Dixon poll from the prior week showing a 2 point gap. And the North Carolina poll's 2-point margin is corroborated by a GOP partisan Civitas poll  showing McCain up by 4. My policy is to display on these posts the most recent nonpartisan poll.

If McCain wins Ohio, the Electoral Vote tally stands at Obama 297, McCain 241. If Obama wins Ohio, the count remains at Obama 317, McCain 221 as it was last week.