The good news in the battleground states per Quinnipiac is that he now leads by five in Ohio and hasn’t suffered any real damage in North Carolina and Pennsylvania despite the fact that voters in every state had Clinton winning the debate handily. In fact, check the last six polls of Pennsylvania and you’ll find that that state is now highly competitive in the four-way race, with Clinton’s lead ranging from four points in today’s Qunnipiac survey to just one point in multiple other recent polls. The most tantalizing prize for Republicans in the electoral college is within reach for Trump thanks to his strength with the white working class. If he can put together the trifecta of Pennsylvania, Ohio, and Florida, he’ll have a very real shot at 270.

But what if, for once, Pennsylvania actually isn’t the most elusive of those three states?

Florida: Clinton leads Trump 46 – 41 percent, with 5 percent for Johnson and 2 percent for Stein;

North Carolina: Clinton at 46 percent to Trump’s 43 percent, with 7 percent for Johnson. Clinton had 42 percent to Trump’s 38 percent September 8;

Ohio: Trump tops Clinton 47 – 42 percent, with 6 percent for Johnson and 1 percent for Stein;

Pennsylvania: Clinton at 45 percent to Trump’s 41 percent, with 5 percent for Johnson and 2 percent for Stein. Clinton was up 44 – 39 percent September 8.

Three of those states look similar to how they looked a month ago. Florida doesn’t. Clinton’s lead of five points isn’t huge, but five points would be larger than Obama’s winning margin in the state in 2008 and 2012. It’s a biggish lead by Florida’s standards, and Quinnipiac’s not the only pollster lately to see Clinton with that sort of cushion. Monmouth actually had her up five a week before the debate; a few days ago, Mason-Dixon had her up four. Now this. It should go without saying at this point that Trump’s Pennsylvania dream matters if and only if he also locks down Ohio and Florida too. In fact, Harry Enten looked at hypothetical electoral maps in which Clinton wins Florida and found that there are a multitude of ways she could go on to win the presidency from that point even if Trump strongly overperforms expectations everywhere else. For instance:

Let’s say, for example, that on Election Day, Clinton underperforms with white voters without college degrees even more so than she is doing now. Her support would collapse in the Midwest, and her troubles in Maine would likely bleed over into New Hampshire. But Florida is more diverse, with one of the largest groups of Latino voters of any battleground state and a sizable African-American population. So, in this scenario, the nonwhite vote in Florida might allow Clinton to hold onto the state even if she were to lose Michigan, New Hampshire, Ohio and Wisconsin

Indeed, winning Florida opens up so many electoral paths for Clinton that it’s probably a must-win for Trump. If he loses Florida, Trump wins the presidency only 5 percent of the time according to our polls-only forecast. (Clinton, on the other hand, wins the election 33 percent of the time without a victory in Florida.) Considering that Clinton has led in every post-debate survey in Florida and has a small lead there in our polls-only forecast, you can understand why Clinton is currently a favorite to win the election.

Clinton could face a total apocalypse among the white working class of the Rust Belt and still squeak through to a national victory thanks to more racially diverse battlegrounds like Virginia, (eastern) Pennsylvania, and of course Florida. Quinnipiac finds Trump winning in Florida among whites without a college degree by 34 points(!) and even among whites who do have a degree by six points. He still loses by five, though, thanks to Clinton’s huge advantage with nonwhite voters (and her 20-point advantage among women of all races). See now why Team Hillary is talking up that Trump/Cuba story that most of us are shrugging at? Cuban-Americans in Florida might not shrug at it. If that tips Florida to Clinton, that’s ballgame nationally.

Forget Florida for a second, though. What the hell happened here in Colorado? New from Monmouth:

Among Colorado voters likely to participate in November’s presidential election, 49% currently support Clinton and 38% back Trump. Another 7% intend to vote for Libertarian Gary Johnson, 3% support Green Party candidate Jill Stein, and 3% are undecided. In mid-July, the race stood at Clinton 48%, Trump 35%, Johnson 5%, and Stein 3%…

Colorado voters continue to hold a more negative view of Trump than Clinton. Fewer than 4-in-10 (37%) have a favorable opinion of Clinton while 51% hold an unfavorable view of her. However, only 29% have a favorable opinion of Trump while 63% hold an unfavorable view of him. Both of these results have changed little since July…

Clinton has an even wider advantage on the issue of presidential temperament – something that has been dominating the news since last week’s debate. Twice as many voters say Clinton has the right temperament for the job (61%) compared to those who say the same about Trump (31%).

Monmouth sees Hillary winning by 13 among women in Colorado and by seven among … men, which seems unlikely. Also unlikely is the margin: The last four pre-debate polls of Colorado had the race Clinton by one, Clinton by two, Trump by four, Trump by one. Suddenly, the first post-debate poll has her up eleven? That’s some bounce. Also of note: The last Monmouth poll of Colorado, conducted in July, had Clinton up 13 points, her biggest lead there all year. It may be that Monmouth’s model of the Colorado electorate has an unusually large outlier-ish advantage for Clinton built in. Trump had better hope so, as the tight race in Colorado the past few weeks was his brightest hope of peeling off a key blue state from Hillary in November.

There is, however, a reason to think Monmouth’s onto something in seeing a large Clinton lead in the state now. For one thing, Democrats now have a slight registration advantage in there, an ominous trend given the traditional red tilt in Colorado. Beyond that, it’s always been a long reach for Trump demographically. There are enough Latinos there and libertarian-leaning Republicans there that it’s hard to see how he’d piece together a winning coalition. His best chance would be if Gary Johnson carved a chunk out of Clinton’s support, especially among millennials — but the noteworthy thing about Monmouth’s polls of the state, both today and in July, is how poorly they have Johnson doing compared to everyone else. Most other pollsters have him in the 12-13 point range. Monmouth has him at seven points today and had him at just five this summer. Clearly some of Hillary’s larger lead in their surveys is due to the third-party guy underperforming. If voters are beginning to desert Johnson in the home stretch of the race — and there’s reason to think they are, per Quinnipiac’s data — then Colorado might be off the board after all. Trump would need the PA/OH/FL trifecta, no ifs, ands, or buts. We’ll see what the next polls there say.