Scott Rasmussen and I have emailed back and forth quite a number of times and I finally
got a straight answer on his party weighting...
I had estimated 38/35/27 Dem/Rep/Ind based on his results to date.
I was wrong...
39/35/26 Dem/Rep/Ind.. so I was pretty close...
As an aside... I have sent Scott a number of very deep technical questions, and he has always had a good answer.
I have not always agreed with his answers, but they were always, at the very minimum, defendable position that he took.
In other non news/news...
The huge Kerry Fathers day sample dropped off, and the Robo track now shows Bush 47/Kerry 45
Bush had a semi-strong sample on Tuesday night which is propping up Bush's "lead" in the poll right now.
First, remember when I suggested that Scott had really been rocked by the 2000 results and had increased the Democrat weighting in his polls?
Second, in all fairness to Scott, who has largely done a very good job, many pollsters reacted to the 2000 election in the same manner.
This is why so many polls are skewed a point or to in Kerry's favor, as the 2000 election was the best activation drive the Democrats have had in a presidential election in decades.
Given the republicans have a well run activation effort this year, and the only activation effort I have seen so far has been by the pseudo-independt committees funded by Soros et al., the electorate is more likely to resemble a normal distribution by party (or perhaps even slightly more republican than normal) in 2004.