Close races are nothing new to Dino Rossi, and perhaps the third time will be the charm.  Democrats thought that Patty Murray had begun to put a little daylight between herself and Rossi, but a new poll from Marist shows the Democratic incumbent in Washington falling back into a virtual dead heat with her Republican challenger.  McClatchy calls this a “cliffhanger”:

With two weeks to go, the Washington state U.S. Senate race is a virtual dead heat, with Democratic Sen. Patty Murray holding a 1-point lead, 48-47 percent, over Republican challenger Dino Rossi among likely voters, according to a McClatchy-Marist poll released Tuesday.

The outcome could determine whether Republicans pick up the 10 seats they need to regain control of the Senate.

“This is indeed a cliffhanger, any way you carve up the numbers,” said Lee Miringoff, the director of the Marist Institute for Public Opinion in Poughkeepsie, N.Y., which conducted the survey. “The road to a Republican majority in the Senate could go through Washington state.”

Other recent polls have shown a volatile race, with Rossi having a slim lead in some and Murray up by 6 to 8 points in others.

Marist found 589 likely voters in a sample of 834 registered voters in Washington, including both land lines and cell phones.  The sample has a nine-point advantage for Democrats, 32/23, with 45% independents.  That’s a significantly different turnout model than exit polling showed for the 2008 presidential election, which had a 36/26/39 split.  The Marist poll shows about the same advantage for Democrats and a somewhat smaller Republican turnout for these midterms, both of which seem questionable choices given the political environment.

Also, the numbers don’t really seem to add up in the crosstabs.  Murray gets 96% of Democrats, and Rossi 94% of Republicans, which is not much of a surprise.  However, Rossi gets a 21-point lead among independents, 57/36.  Given that independents make up the largest part of the sample, one might expect the 21-point gap to more than make up for the natural nine-  to ten-point lead Murray gets among just the partisans, especially since independents are somewhat overrepresented in this model.  Yet Rossi trails by a point, 48/47.

The poll shows another dynamic with which Rossi is very familiar: regional swings.  Rossi loses the King County/Seattle region, where Christine Gregoire bested him in two recount-dependent gubernatorial elections.  However, he only trails 40/57 in this most liberal of all Washington regions, not a bad showing at all for a Republican.  Rossi wins all three of the remaining regions, two by significant majorities.

The big question here will be whether the turnout model actually works, and when any momentum came in this race.  Most Washington voters send their ballots in by mail; only in one county and a few other precincts will voters go to the polls on Election Day.   Will Rossi have closed the deal before this poll showed him closing the gap, and will the turnout model wind up being something a little more Republican than 2008’s model?  I’d bet the answer to both is yes.