Chris, do you have a link to ARG's polling results in 2004?

looking at the Republican results which show them completely backwords from every other polling firm in the race, either they are incompetents, frauds or visionaries or they are getting paid off from Hillary and mcCain. my gut feeling is Mark Penn's inevetability team has something to do with these numbers but I can't prove it. It is interesting how quickly Hillary's spin team is out to paint these obviosly wacky results as correct intead of the polls that showed Hillary doing bad in Iowa which led I presume to the imfamous skip Iowa memo.,

in New Hampshire over the weekend than ARG has voting for him in the entire state, I guess according to them donating money or seeing the candidate in person doesn't make for a likely voter.

I posted this in georgep's original diary but I thought it would be good to be discussed here.

This is from ARG's site.

Question Wording:

Would you say that you definitely plan to participate in the 2008 Democratic presidential caucus, that you might participate in the 2008 Democratic presidential caucus, or that you will probably not participate in the 2008 Democratic presidential caucus?

Is the likely voter only those that said they "definitely plan to participate" or the combination of "definitely plan" and "might participate"?  Because if it is both, that seems to be much too loose a likely voter screen for a caucus.  To attend a caucus it takes a lot more effort, time and courage to go through with it than simply going to a voting booth and voting as in a primary.

Unfortunately, ARG doesn't have a breakdown on the 600 "likely voters" and how many were "definitely plan" and how many were "might participate".  A caucus, and especially the Iowa caucus since it's the first, is notoriously hard to poll.  I think the best thing for all the candidates to do is to not pay too much attention to each of the polls this far out and keep plugging away.

I'd love to find out what that breakdown of "definitely plan" and "might participate" were.

They had Bush by 2 points the day before the primary in NH 2000.  They are a NH polling firm and effed that up that badly?  They assume 20% independent voting in the caucus, when the actual non-D vote is more like 5% (and will likely be less this year with fewer Rs crossing over).

OK, not only are their Dem numbers out of whack, but their GOP numbers have McCain surging into the lead in Iowa, NH, and SC.  WTF???

And I still can't forget the CT-Sen general election last year, where ARG had two polls showing Lamont only 2% behind Lieberman when everyone else had him at 10% or further back.  We know who turned out to be right in that instance.