Donald Trump (Carlo Allegri/Reuters)

With much fanfare (and no shortage of media attention), Donald Trump revamped his campaign earlier this month with an eye toward turning things around. A new campaign manager, a new chief executive, a new habit of reading off the teleprompter and a new target: black voters.

A national poll from Monmouth University released on Monday showed that the revamp hadn't necessarily moved Trump's national numbers. But what about Pennsylvania, a state that Trump has insisted that he could win and which has a substantial population of black voters?

Monmouth polled there, too. Trump trails by eight points, in line with the average of polls from RealClearPolitics. The pollster didn't break out black voters separately, but among nonwhite voters, Trump gets 5 percent of the vote, 85 percentage points fewer than Hillary Clinton gets. That's all nonwhite voters, mind you, including any Hispanic respondents. (About 7 percent of the state is Hispanic.) Since Trump does quite a bit better with Hispanic voters than black voters, it suggests that Trump's numbers among black voters likely haven't improved substantially. (Well, that's not true: He got zero percent from black voters in Pennsylvania in July.)

To some extent, though, this is deeper in the weeds than we need to be. The simplest calculus to determine who will win in November is comparing how Trump is doing to how Mitt Romney did four years ago. Romney lost, so Trump needs to do better than him to win. The bad news from the Monmouth poll, then, is that Pennsylvania continues to be out of reach for Trump. It almost certainly won't be the state Trump picks up, if he picks up any at all.

In fact, of the 10 states that were closest in 2012, Trump trails in all 10, according to RealClearPolitics' polling average. Obama won nine of those states, mind you, but these should still be the places where Trump is most likely to make gains. What's worse is that the average of the averages in those states shows Trump trailing by more than six and a half points.

Polls change, of course, but it's also important to note that in states where we can compare RCP polling averages now to how Romney was faring in 2012, Trump does worse almost across the board. He's underperforming where Romney was on Sept. 1, 2012, in Virginia, Colorado, Wisconsin, New Hampshire, Florida, Ohio, Pennsylvania, North Carolina, Michigan and Iowa. He's doing a bit better in Nevada (which we looked at on Monday), but he's still trailing there.

Worse still, Trump is at risk of losing states Romney won. North Carolina falls into that category, as does Georgia. It's still likely that Trump can prevail in either state, but that doesn't give him the gain that he needs. What's more, it allows Clinton to play offense in those states and forces Trump to spend time and money playing defense when he really needs to be trying to gain ground in states Romney lost.

Notice that the trends in each state in 2012 often continued: Where Romney trailed at this point, he often ended up losing by more. In two of the three states on that chart where Romney led, he ended up extending that lead. On Aug. 19, we looked at the state of play as Trump's campaign revamp was announced. At that point, he was doing better than Romney in Ohio and Iowa (relative to Election Day that year). That's no longer the case.

With 10 weeks to go — and early voting beginning in less than a month in a number of states — Trump's campaign overhaul hasn't yet reshaped the state of play in the swing states. If he can improve his national numbers, that's fine. But if he continues to lose the 10 closest states from 2012 by about six points, that's it.