My feeling is that a lot of polls are designed to create news and not to gage true opinions.

Polls actually tell us nothing but what they want us to think.

I've been alive for 62 years and have lived in both New York and Connecticut. I have NEVER received a request to participate in any major poll and neither has anyone I know. The education range is from High School Graduate to Ph D's in many areas including Doctorates. The work range is equally as diverse. Flawed is too nice a word for those who can choose any demographic they wish in order to obtain the right spin on a question (s) answered.

Phone interviews are as skewed as the rest, if not more so, in today's world. I am registered with the Do Not Call Registry for my home and cellular services.

Although they try to obtain a "statistically significant sample" and ensure they use a representative cross-sample of the US, I think that such a small number of people can't possibly predict what the entire country thinks. There seems to be some fundamental bias associated with these types of polls that exclude people described in the debate description.

Makes you wonder, doesn't it. If all those polls are not really giving an accurate picture, why does the media spend so-o-o-o-o-o-o-o-o much time talking about them?

The problem with polls is you have to remember who has access to even vote on the poll. If it's put online, especially on a site that isn't huge, then you're not including all the people in middle america who either don't have the internet, or don't use it for more than e-mail and movie listings, not to mention all the people who simply aren't looking for the poll in particular.

The questions are generally phrased to reflect the bias of the pollster and force a biased answer from the respondent.

I used to work at Gallup Poll in Austin Texas. One day they had one of the corporate heads (I think it might have been the main guy actually) come and discuss "Gallup's Future" with us. This was 1997 and the corp guy gleefully regaled us with the fact that Gallup had moved their headquarters to a boat three miles (the point where American jurisdiction stops) offshore. They were accountable to no one.

Now take that little fact to heart and take a look at the current polls per McCain and Obama. Add that to the fact Gallup bases their responses on landlines ONLY. Now who has cell phones? Yes, likely Obama voters. Add a ten percent margin FOR Republican voters and viola, you have a tie.

The polls are baloney.

I think the science of polling is advanced enough that they are accurate. The problem is that I believe people lie and that pollsters rarely reveal to us the exact wording they use in their polls and so we dont know what intentional biases are built deliberately into the poll or exactly what question they are designed to answer. Lastly, dont forget that most polls are ordered by an less than non interested party.

They're accurate to the percent that they say they are. If you notice, all of them have disclaimers about the percentage they might be off based on sampling error. For the first half of 2007, only 15% of households didn't have a landline. The gallup poll also tests cellphone users. If you randomize the people you call enough, that's not a very significant loss. The rest is just statistics about sample size and random sampling math. Those footnotes explain what the survey really is surveying. Just remember that if the error margin is 2%, it could be that one is 2% higher and the other is 2% lower