I've been tracking how the state polls are relative to each other, and came up with this:
Zogby 7.0
Rasmussen 3.5
ARG 3.4
Research 2000 2.9
LA Times 0.9
Mason-Dixon 0.9
SUSA 0.0
Sorry I don't have time to format this better.
SUSA is the most accurate using a methodology of using 2000 as a base line. (that is, take the 2000 results and add or subtract the national results to guess where a state is).
I have read the 3 most accurate polls last time were Mason-Dixon, SUSA and Research 2000. I have just a few Research 2000 polls in my sample so far, so their skewedness may easily be just random fluctuation. The ARG, Rasmussen and Zogby polls aren't, however, as there is a lot of data for them.
The Zogby poll results are just the first batch. I quit following them as it's methodology and results disqualify it from being considered, but the skewedness of 7 towards Kerry gives some idea of how unreliable it is.
At any rate, any methodology which relies on the Zogby, Rasmussen, and ARG state polls will present a look at the election picture which is rose-colored from a Kerry perspective.
Did you notice on the Rasmussen thread that Scott has now admitted that he gives Democrats a 4 point advantage in his polls!!!!
I'm not sure I understand what you are analyzing in your post #16.