As Jim Geraghty notes, conservatives usually express the same kind of skepticism towards PPP surveys as progressives do with Rasmussen.  Usually, while the methodology is similar, the sampling isn’t analogous; like most pollsters, PPP waits until closer to the election to use a likely-voter model, where Rasmussen rarely uses anything else.  Now that PPP has switched to likely voters, it turns out that they’re more similar to Rasmussen than first thought, at least in terms of results in Pennsylvania:

In PPP’s previous survey of the Pennsylvania Senate race in June, Republican Pat Toomey and Democrat Joe Sestak were dead even at 41% among occasional voters. In the first survey using its likely-voter model, however, PPP now finds Toomey jumping out to a 9-point lead, 45-36, with 20% still undecided. …

In an electorate with a 50-42 Democratic identification edge, Sestak maintained parity with Toomey by getting almost as much of his party’s support as Toomey pulled of his, despite a 41-21 Toomey advantage with the small set of independents. But with only a two-point Democratic advantage, Sestak gets only 64% of Democrats to Toomey’s 74% of Republicans, and Toomey’s support among unaffiliateds is now a majority, 50-23. Sestak also now has a net-negative favorability rating by 10 points, after breaking even in June, while Toomey keeps a barely positive mark.

Jim pronounces this “Toomsday,” and it’s difficult to argue with that assessment.  PPP notes that the likely-voter electorate backed John McCain over Barack Obama by one point, despite Obama’s win in Pennsylvania by 10.  The people who turned out in 2008 aren’t going to the polls in 2010, while independents and Republicans angered by the Democratic agenda are on a mission to change direction in Washington.

It’s not hard to see why from the poll results.  Obama gets a 40/55 job approval rating from likely voters, and his ObamaCare gets the same treatment at 39/54.  Independent voters give both a 32/63 approval rating.  Sestak, who voted for ObamaCare, only manages a 28% approval rating despite his relatively high profile in Washington — and despite a two-point advantage in the survey sample for Democrats, 46/44.  In fact, Sestak only gets a 47% approval rating from Democrats.

Assuming this model correctly shows the support levels in Pennsylvania, Sestak would have to convince 80% of the undecided voters over the next ten weeks to prevail in the general election.  Possible?  Yes, but not likely at all.

Update: The news isn’t quite as good in PPP’s poll of Illinois likely voters in the Senate race.  Alexi Giannoulias edges Mark Kirk 37/35, with 28% of voters wondering how they wound up with these choices.

Update II: Rasmussen has a new poll coming out this afternoon showing Toomey with a 9-point edge as well, 46/37.  Toomey wins independents 45/24, gets 19% of Democrats, wins both men (52/37) and women (41/37), and every age demographic except 18-29YOs.

Update III: Yeah, I jumbled up the Rasmussen numbers, which commenters rightly point out are 46/37, not 47/36.  Is it Obamath or wishful thinking?  Is there a difference?  Anyway, my apologies for the error.