Ras has recently said he thinks the actual turnout will be D+2-4%. That explains his switch to a D+3 sample.
I still think his party affiliation makes more sense, so his reasoning confounded me.
Like you, I think we're about to witness a blowout of epic proportions. In that scenario the only way to explain a D+3 turnout is to realize that alot of Dems will vote for Romney.
It's the premise of this thread.
Sly one, that Rasmussen. :)
Rasmussen has to consider his own business, as well. He’s been producing the most Republican numbers of national polls. Most other polls are using D+6 or greater. So, with a D+3, Rasmussen will be closest as long as turnout is more Republican than D+4.5... which would be the second biggest D wave since Reagan.
Second, we have yet to hold an election where Ras showed more Rs than Ds the month before the election. November 2010 was the very first time Ras ever showed Rs outnumbering Ds. So, he doesn’t really have any data to guide him on what will happen on Election Day under that situation. He DOES know, though, that since Reagan got elected, turnout has been between even and D+4 every single election except 2008, with an average of D+2.5. That gives him a nice, safe model that is easy to defend. I don’t blame Ras one bit for that.
That's a far more powerful source of victory than planning on a bunch of Democrats coming down the street to pull the levers repeatedly for a Republican (which is pretty much like imagining that you could have an undecided 5 to 10% independent vote somewhere in the middle in a voting public that has no middle).