Ahead by a single point, I should note, whereas Trump was ahead by a single point last week, making the movement here statistically insignificant. But it’s still interesting, for two reasons. One: Rasmussen has been one of the most reliably pro-Trump polls out there. Until recently, in fact, it was the only pro-Trump poll. From June 1st until six days ago, no one else had Trump with a lead nationally. Ras, however, had him up four at the end of June, two a week later, seven(!) a week after that, and last week had him ahead by one. At a moment when various other national polls are finally catching up to them and finding Trump passing Clinton, Rasmussen suddenly sees the opposite. Hmmm.

The other interesting element here, of course, is the timing. This poll was conducted on Monday through Wednesday of this week, with virtually the entire sample polled before Obama’s speech last night. Is Ras picking up the beginnings of a convention bounce for Hillary?

A new Rasmussen Reports telephone and online survey finds Democratic nominee Hillary Clinton with 43% support to Republican nominee Donald Trump’s 42%. Ten percent (10%) like some other candidate, while four percent (4%) remain undecided…

The latest survey was taken Tuesday and Wednesday nights, so we won’t know until next week whether Clinton generates any bounce from her acceptance of the Democratic nomination in a speech before the party’s national convention tonight. Last week’s survey was taken before Trump’s acceptance speech, so if he got any bounce from his convention, it has been blunted by the opening days of the Democratic convention.

Trump still has stronger support among voters in his own party (86%) than Clinton does among Democrats (79%). But this week Clinton has taken a five-point lead among voters not affiliated with either major party.

The last three national polls before this had Trump up three points, one point, and seven points; either Ras is off here in not seeing a bounce for Trump or they’re ahead of the pack in detecting a counter-bounce for Hillary. The poll that had Trump up seven, by the way, is the LA Times daily tracking poll that I blogged yesterday. His lead actually expanded further overnight in that one, although again by a statistically insignificant margin. He went from a lead of 6.6. points yesterday to a lead of 7.3 today, even though we’re already most of the way through the Democratic convention. Even so, don’t be too quick to draw conclusions. Click the last link and you’ll see that Trump’s own numbers were flat throughout the Republican convention. Only afterward did he start to take off. Hillary’s bounce might be coming but is similarly delayed right now.

One more new poll for you, this time from Suffolk. Guess which state these numbers come from.

Massachusetts? Nope, Pennsylvania. That state is … not supposed to be a gimme for Democrats this time; if it is, Trump is all but done. This poll, like Rasmussen’s, was conducted over the past three days, when you’d expect to see some evidence of a post-convention bounce for Trump. Instead Suffolk has Hillary’s margin exactly the same as it was in the last poll of Pennsylvania, conducted by the WSJ/NBC in mid-July. It was nine points then, it’s nine points now. Huh. Her average lead in the state is now up to a comfortable-ish 4.4 points despite the fact that she led by no more than four points in any poll in June and that Trump led for the first time in the state just 17 days ago. How do we square all of that? Is Pennsylvania reverting to bluish form at a moment when the rest of the country seems to be taking a hard look at Trump, perhaps because they’re hosting the Democratic convention this week, or is today’s Suffolk poll a little off? I lean towards the latter: If you scroll down through the crosstabs you’ll find that they also have Democrat Katie McGinty with a solid lead over Pat Toomey for his Senate seat, 42.8 percent for her versus a measly 35.6 for him. No other poll of that race over the last two months has Toomey below 40; less than three weeks ago, Quinnipiac had him at 49. Even with today’s terrible Suffolk number baked in, he still leads McGinty in the poll average (by three-tenths of a point). What probably happened here is that Suffolk inadvertently oversampled Democrats, driving down the numbers for Trump and Toomey. Still, it’s worth flagging this for future reference. If we get another poll of PA showing Hillary leading easily, the sirens are going to go off at Trump HQ.