Rasmussen does not use that standard. His basic standard is a running poll of 15,000 respondents.
Rasmussen Reports tracks this information based on telephone interviews with approximately 15,000 adults per month since November 2002. The margin of error for the full sample is less than one percentage point, with a 95% level of confidence.http://www.rasmussenreports.com/public_content/archive/mood_of_america_archive/partisan_trends/gop_edge_in_partisan_id_slips_to_2012_low
Ok, I believe that.
That being the case, if his own party ID shows R+2, why are his polling models using D+5?
Still chicanery, just a different type.
I don't see a discrepancy between the two sets of information.
One is a poll unrelated to the presidential race, determining party affiliation of the population.
The other is a presidential race poll in which party affiliation is also asked.
It's two different data sets. They may or may not even be related; to the degree they are related, their relationship is likely not simple.
It's the second data set that Mr. Rasmussen needs to tie in to the election, because on election day, no matter how many Republicans, Democrats and Independents there are in the United States, the only ones that will matter are the ones that show up and vote, and what these voters call themselves as they leave the polling place (which may or may not be an accurate reflection of the actual population of voters).
Thus, it appears to me that Mr. Rasmussen doesn't weight his presidential race poll by his on-going party affiliation poll results. Indeed, there's nothing at your link to suggest that he does.
I'm going to rely on the York article, as it's pretty explicit in having Mr. Rasmussen say that he's going D+2 - D+4. The information at your link is interesting, but there's nothing explicit there that Mr. Rasmussen is using that data.
sitetest