I’d put more credence in this if the first poll didn’t come from PPP, which has played a lot of games in sampling the last few weeks, and the second from McCall/Muhlenberg, which hasn’t disclosed even a portion of their sample or crosstab data.  Their new poll apparently has Pat Toomey losing a seven point lead in the space of two weeks to fall behind by three, with both candidates only scoring in the low 40s with 13 days left to go before the election.  Can this get any more questionable?  Via Jim Geraghty … yes:

Democrat Sestak now leads Republican Toomey 44 percent to 41 percent with 15 percent undecided, a Morning Call/Muhlenberg College Tracker poll shows.

. . . It’s hard to explain Sestak’s latest apparent comeback. There’s a sense that faithful Democrats are paying closer attention in the closing weeks as President Barack Obama crisscrosses the country making the case for keeping Democrats in charge.And Sestak’s campaign appears adept at using an opponent’s own words against him in political TV ads. On air now is a commercial showing Toomey, a former Lehigh Valley congressman, saying his voting record is “hard to distinguish from Rick Santorum‘s.”

That caused a ten-point swing in the gap in two weeks?  Bear in mind that Toomey isn’t exactly an unknown quantity in Pennsylvania.  He served in the House for three terms during both the Clinton and Bush administrations, and has made two runs for the US Senate, narrowly missing against Arlen Specter in the 2004 Republican primary.  Pointing out that Toomey has a conservative voting record may excite the Democratic base a little, but won’t move the needle ten points.

Speaking of which, in the middle of this supposed ten-point swing, Rasmussen also polled in Pennsylvania and found Toomey had a 10-point lead over Sestak.  In their sample, Democrats had an eight-point advantage, 45/37, so the lead didn’t come from a lack of Democratic turnout.  That was actually one point higher than Democratic turnout in the 2008 presidential election, according to CNN’s exit polling (44/37).  If anything, Rasmussen’s likely-voter model from last week was probably a little too generous for Democrats, and therefore the idea that Democrats are suddenly thundering to the polls seems an unlikely explanation for these results.

I’d wait to see what Quinnipiac, Survey USA, and potentially another Rasmussen poll show in terms of movement rather than rely on these two surveys.