[HannibalHamlinJr]

Yesterday morning I believe, in a thread, at least twice it was posted that Rasmussen has moved from a D+3 to D+6 partisan model. I took it that the second post was from a Rasmussen paid subscriber. Do we have confirmation or denial of this? ... Wouldn't normally interrupt into main threads, but this is a pretty important factoid, question about veracity came up on another thread tonight. (I searched ... but without searching body of threads, can't locate it to see if it was confirmed later in the thread.)

[Senator Goldwater]

Depends on where he’s polling. I think he uses 2008 and 2010 turnout averages, and then leverages it against enthusiasm factors to get a proper sample figure.

[nhwingut]

Rasmussen uses a D+3 (39D/36R/25I).

[InterceptPoint]

I can’t give you a link but we did have a Raz subscriber post a few days back that it was definitely D+3. He sounded pretty authoritative so I took him at his word.

Also, a 3 point jump from D+3 to D+6 would cost Romney nearly 3 points in the tracking poll over a 3 day period. We saw no such jump. So I’m sticking to the D+3 theory.

[RoseofTexas]

This is my prediction...IF Romney is still in the 50% range tomorrow by a few pollster...HE WINS!! This is it..do or die, this is the last week and YOU AIN'T gonna change too many minds with nine days left TIL the big DAY! PLEASE LORD bring victory to Romney's camp....DON'T FORGET PRAYER AND FASTING TOMORROW!! This week is going to be the WEEK FROM HELL, this is the time that Obumbler and his evil demonic hordes are going to go insane and start a Racial war..or something more sinister! BE ON YOUR KNEES!!

[Political Junkie Too]

His national tracking poll weekly summary is 3.500 likely voters 36R/39D/25O.

His Florida state poll from Friday was 750 likely voters 43R/39D/18O.

His Virginia poll from Thursday was 750 likely voters 38R/37D/24O.

His Ohio poll from Wednesday was 750 likely voters 39R/38D/23O.

-PJ

[InterceptPoint]

Thanks for posting. This needed to be flushed out.

[Jet Jaguar]

Poll ping discussion.

[outofstyle]

Scott was on the tube the other day. He said: he is not manipulating the D-R number one way or the other. He says they simply make the contacts and, based on the answers to questions about past voting history, intention to vote and enthusiasm, he classifies them as “likely voters” or not. Then he asks about party ID. Specifically, he asks “as of today, do you consider yourself R, D or I.” If his contacts reveal a party advantage in the likely voter model, it’s simply the fact that that is what he is seeing according to his model.