Rasmussen - President George W. Bush with 47.9% of the vote and Senator John Kerry with 47.1%.

http://www.rasmussenreports.com/Presidential_Tracking_Poll.htm ^

[politicsfan]

Hmm... Rasumussen latest poll has the race tied at 48.3 - 48.3 with leaners. Why has race tightened in this poll?

[Archie Bunker on steroids]

I'd like to discredit it too, but I'm convinced their are more human parasites in this country then you think. Factor in the criminally biased media & this race is too close for comfort.

I so desperately want to hear the concession speech that we didn't get in 2000.....and I want it early.

[rocklobster11]

No race could be so close that it stays within one or two points for week after week.

Rasmussen's poll is designed to not show big swings. He adjusts his numbers based on the party identification breakdown from 2000 exit polls (this is a change from what he did in 2000, where he ended up with the worst record of accuracy of any pollster). Thus, it would take a major shift in one party's support to/away from their candidate or a major shift in the independents/undecideds to cause the poll to swing. Zogby uses a similar method, but he applies his own special sauce to the recipe.

Newsweek, Gallup, and Battleground do not adjust their polls based on party identification, which is why they have shown wild swings in the race. Sometimes they oversample Republicans, sometimes the oversample Democrats.

The real answer is that in all of the polls, Bush is doing a better job of holding onto his base than Kerry is. This is because it's easier to be excited to vote for someone you like, than to be excited to vote against someone you hate. I believe the independents are split somewhat evenly, but they do break to Bush. At this point, the undecideds are as brain dead as Arafat.

I think Rasmussen's technique is the correct technique. However, he doesn't account for 9/11, which has made some Independents and discerning Democrats switch to Republican. Assuming there is no major unexpected historical event in the next 4 years, his method will probably be pretty good for the 2008 election (except for the fact that there will not be an incumbent running in 2008).

The real value of the Rasmussen poll is to track long-term trends. And the longterm trend is that Bush has been consistently up between 1-3 points in his poll.

[My2Cents]

Rassmussen's worthless. His polls are never outside the margin of error. Why bother publishing them?

[Illinois Rep]

Raw Data for Rasmussen

Wednesday
Bush 48.6 Kerry 47.2
Thursday
Bush 50.8 Kerry 46.9
Friday
Bush 44.3 Kerry 47.2

So what this says is on Friday only 91.5% of people polled selected Bush or Kerry. That is very odd.

On Tomorrow's rasmussen--a very slight lead for Bush falls off the track...Then Monday morning polls a VERY good Bush day falls off the poll. This means Kerry will most certainly be leading Bush in Rasmussen's poll on Monday morning unless Bush has a very good result tonight and/or tomorrow.

On Tuesday (election morning) Bush's worst day falls off and a usually good night (Mondays) goes ON the poll so going into Election Day Rasmussen no doubt will show the race 48 Kerry 48 Bush or very very close to that.