I am not saying Rasmussen accurate or inaccurate.
However, in all seriousness, I think Rasmussen has the tendency to be biased in poll selection and result report wording.
For example, before Palin's speech, when she was faced with a lot of negative press, they did a poll asking whether people thought "the media was trying to hurt her". 51% replied yes. But after Palin's speech, which was praised in the media heavily, Rasmussen did not do the same poll to see if people thought the media were trying to help her.
Furthermore the poll headline reads:
"51% Say Reporters Are Trying To Hurt Palin; 39% Say She Has Better Experience Than Obama"
And within the body of the report the 39% number is again repeated. It is not until after this that we find out that 49% thought Biden had the better experience. So why highlight the 39% over the 49%?
The networks or Gallup do not do this as far as I can tell. Only Rasmussen does. I could mention more examples, but these two are representative.
Rasmussen is a Republican. His polls are great and not biased, but his reporting certainly is.