Considering the way they are conducted, by calling "likely" party voters (people who voted in the last primaries) on a landline telephone from the previous election cycle primaries, why should they be considered fair, accurate, and tantamount in determining a candidate's support? When someone who crosses party lines runs for president, and they appeal to more than just the hardcore base of either of the 2 main parties, that person, no matter how much support they SEEM to be drumming up nationwide, will always be marginalized by the outcome of the landline polls. This disenfranchises Independents, people who were too young in the previous election cycle to vote, people switching parties, people who haven't voted in decades due to apathy, people who don't operate their communicative lives via landline phone, etc. To me, this would seem about as UN-scientific as it gets. Shouldn't these polls be improved to include cell phones, and shouldn't scientific polls reporting in the media also include the results of official state and local straw polls? I mean, straw polls are the closest thing we have to a foresight into how an election would have gone. Why do we continue to allow the media to influence us with these so-called scientific polls, when they are obviously so flawed and inaccurate?