Because there were almost new state polls during the two weeks of the conventions, it was not really possible to see separate bounces from the two conventions. Right before the conventions, Obama managed to reverse some of the momentum McCain had in late July and early August. Now that the conventions are over, we are getting bunches of state polls again, now looking at the state of the race after both VP choices and both conventions.
Today the news is all bad for Obama. In order of electoral college votes, here are the states that change category in today's updates:
Ohio (20 ev): I am half tempted to not even report when "leaning" states flip from one side of the fence to the other. Either way, the real way to look at it is that the state is too close to call. The moving back and forth on who is slightly in the lead is probably just random. In this case, we go from Obama being ahead by 0.2% in my five poll average, to McCain being ahead by 0.2% in the five poll average. Either way, it is a tie race. But it gives McCain bragging rights to put Ohio in the "if everybody gets their leans" category.
Michigan (17 ev): Of the changes today, this is actually probably the most significant. Obama's lead in Michigan drops (once again) to less than 5%, thus moving the state back from "Weak Obama" to "Lean Obama" and bringing the state into the inventory of swing states that could easily go either way. With McCain having a shot in Michigan, it opens up the range of possible ways for him to win significantly. Instead of having to sweep seven swing states to find a path to victory, if he can bring Michigan into play, he gets a situation where the only "Must Win" state is Florida.
Virginia (13 ev): This is another one where there really isn't a significant change other than bragging rights to the current lead. Virginia goes from Obama being ahead by 0.2%, to McCain being ahead by 0.2%. In reality, this is just too close to call. As I keep saying, all of the swing states should really be considered too close to call... even the ones where one candidate is ahead by 4.9%. But in the cases of Ohio and Virginia, we really are just bouncing around the "completely dead even" line.
Washington (11 ev): As happened with several states in August, the five poll average in Washington now drops below a 10% lead for Obama. So the state moves from "Strong Obama" to "Weak Obama". Still quite a way from becoming a swing state, but it shows overall weakening, even in states that should be completely secure for Obama.
Given all of the above, the overall summary of the race changes a decent bit:
McCain Best Case - McCain 298, Obama 240
Obama Best Case - Obama 381, McCain 157
If everybody gets their leans (and Obama gets DC) - Obama 273, McCain 265
That last set of numbers, with everybody getting every state they are even slightly ahead in, is once again looking like an incredibly close race. One that could easily come down to very close results in one or two states.
I also note that although no category changes happened, new polls showed Obama weakening in Pennsylvania and Colorado as well. This was not a good poll day for Obama. The only bright spot for Obama was a very slight improvement in Florida.
Over all though, it looks like in the immediate aftermath of the conventions, McCain has undone Obama's "pre-bounce bounce" and we're in approximately the same place we were around August 22nd... which is an Obama advantage, with Obama only a couple of states away from the win when you give him all the states where he is ahead by more than 5%... but with McCain with the apparent edge in most of the swing states... enough to bring him right up to the edge of winning (but not quite) if he got all the states he is ahead in.