I did news yesterday, so today is back to polling. Put simply, Mitt Romney's debate bounce is now showing signs of burning off virtually everywhere you look. My poll aggregator shows Barack Obama leading 23 of 25 national polls (including tracking at various times) between September 22nd and October 2nd. Then between October 2nd and October 10th, Mitt Romney lead out of 10 of 13 polls. Since then, Obama has lead 5 of 6 national polls with only PPP showing a lead for Mitt Romney, and PPP's three-day tracker showed the race tied yesterday, and Obama leading by a single point today.

Just visually, you can clearly see the three trends. Obama dominated September from his convention, Romney took the lead nationally but not as strongly as Obama had had it, but for only about a week, and now Obama is coming back to where Romney just was, up by an average of maybe three points.

The trackers are moving in that direction as well. Rasmussen had been showing a 1-2 point lead for Romney between the 12th and 18th, but today is showing a tied race. PPP's new tracker showed it tied yesterday, and Obama up by one today. Ipsos had Romney up by three on the 11th, by one on the 12th, Obama up by two on the 15th, and up by an average of three between the 16th and 19th. IBD/TIPP, for which I have little faith and no history, was showing a tied race yesterday, but now a 1.7 point lead for Obama today.

Of six trackers, the only two not showing and Obama gain is RAND, which has Obama down from yesterday (+5.7 to +3.1), and Gallup, which has Romney up by six points, down a point from yesterday.

To illustrate how stupid Gallup has gotten:

RAND: Obama +3.1

Rasmussen: Tie

PPP: Obama +1

Ipsos/Reuters: Obama +3

IBD/TIPP: Obama +1.7

Gallup: Romney +6

Read Nate Silver about Gallup and what you should make of them.

Speaking of Nate, he explains today why his models gave Obama a five point boost in his chances to win the election (65.7% to 70.4%) within the last day. Spoiler: state polling.

There were 13 battleground state polls yesterday and Obama gained in five of them, and Romney in four. Sounds unfair, right? They both gained in a roughly equal number of states. But it's not fair in a meaningful way because not all gains were equally meaningful. Romney gained in North Carolina, but he was already leading there and Nate's model already projects a win for Romney there. And it's not always about the change. Holding a sizable lead in a very important state, even if your lead in shrinking a little, can be quite meaningful.

If you take a quick peek at the battleground states, you'll understand why Obama is a 70% chance favorite to win, despite national public polls showing a tied race (for the moment). I say for the moment because with Obama leading in 5/6 recent national polls, Romney is about two national polls from having a 2-3 point deficit in an average of the most recent ten polls, instead of 0.5 deficit.

Last five poll average, and latest poll:

Nevada (6 electoral votes): Obama +5.2 [Obama +8, Mellman (D)]

Colorado (9): Obama +2 [Obama +3, PPP]

Iowa (6): Obama +3.4 [Obama +8, NBC/WSJ/Marist]

Wisconsin (10): Obama +3 [Obama +2, Rasmussen]

Michigan (16): Obama +6.6 [Obama +6, EPIC-MRA]

I included Michigan because Republicans are pretending that it's in play. Um, not so much.

Ohio (18): Obama +2.8 [Obama +1, Rasmussen]

Pennsylvania (20): Obama +3.6 [Obama +7, PPP]

Would be Obama +5.5 without the increasingly laughable Susquehanna, who gave Romney +4 in a poll on behalf of the Republican State Committee of Pennsylvania. Yeah, because that sounds on the level.

New Hampshire (4): Obama +0.6 [Obama +1, Rasmussen]

Virginia (13): Romney +1.6 [Obama +1, PPP]

A Romney +7 from McLaughlin from a poll paid for by George Allen's Senate campaign is skewing this. This race is tied. And remember, Suffolk hilariously pulled out of VA on the grounds that there was no chance for Obama to win there. Here are the last four polls of Virginia: Obama +1, Romney +2, Romney +1, Obama +1.

Florida (29): Obama +0.2 [Obama +4, JZ Analytics/Newsmax]

Last five polls: Romney +4, Romney +1, Romney +1, Obama +3, Obama +4

If the election were held today, Obama would win 118 electoral votes, and Romney would win 13. If you do these averages based on the last ten polls (I did five to illustrate recent movement), Florida and Virginia go to Romney, but that's it. Obama has broken a stalemate in Colorado, has come back to basically tie Virginia, and has retaken a miniscule lead in Florida. The thing about Florida isn't Obama's lead, it's the trend. Romney was up by four from his debate bounce, then down to up by just one point, and now Obama is up by 3-4. One more Obama lead in Florida and that +0.2 lead could look more like +2.

Since Real Clear Politics' map is the most favorable to Mitt Romney, and the only map cited by Republicans, I'll use it as a baseline:

Obama-Biden: 201
Tossup: 131
Romney-Ryan: 206

If the election were held today, based on the above state polling, it looks more like this:

Obama-Biden: 319
Tossup: 0
Romney-Ryan: 219 

If you disagree, feel free to come up with your own numbers. Even if you give Florida to Romney, he still easily loses:

Obama-Biden: 290
Romney-Ryan: 248

It takes 270 to win, and while it's possible to tie, it's not likely.

In fewer words, Mitt Romney has already lost almost everything he gained from the first debate, and he was already losing even with those gains. With one debate left, Monday the 22nd, and Obama's debate win on the 16th, it seems unlikely that Romney can rely on even another debate blowout to carry him over the line. At this rate, you'd have to see most national polls showing Romney up by 7+ points to drag enough states with him to win the electoral college. Probably more like +8-10.