Why is anyone believing the poll numbers?

I just looked up Dem numbers for California and different polls all seemed to have different results. One had Clinton 9-points ahead, another had them tied. And yes, same dates.





Why the variability?Why is anyone believing the poll numbers?
poll numbers are not acurate until TuesdayWhy is anyone believing the poll numbers?
Not really. Recent RCP polls show Hillary at 41% in California and Obama at 39%... which means she has not gained but he was at 36% yesterday. No one really knows until Tuesday, and even then we may not know much more about the Dems. RCP has McCain at 35 % and Mitt at 30%, with Mitt gaining 2 %.
That's why I never look at just one poll. And which ever poll I'm looking at, I want to see the actual questions and how they were phrased. The two I find the most reliable are Zogby and Rasmussen because they have a good track record of calling things right..
They are different because they are asking different people. Why worry about polls anyway. Tuesday will tell all.
ill explain the polling process as i understand it.





basically they call up about 10000 people in the state and they take the responces and count them... that is the poll number you see. they include a margin of error and there is also a category for undecided voters (which they do not show you on tv).





the polls are almost always within their margin of error after you include the undecided voters... they have like 20% to play with.





basically you should consider it a tie if the poll is within 10%. its kind of a toss up with one person having a slightly better chance.
You can take all of the polling results and add them together to get more accurate information. Also, you can eliminate some of the polls that have results far away from the norm and then average the remaining polls.





The polling has to be done correctly and usually it isn't. With phone polls only one in twenty people pick up their phones. Polls taken during the day may exclude more workers. If the poll taker has the official party list of registered voters, the poll will turn out differently than if the poll taker relies on the participant to state his or her party. A poll taken by Clinton workers turns out differently than a poll taken by Obama workers.





Voting results differ by community, so to be accurate, a poll must proportionally survey from the entire state population. Some polls just take a percentage of all of the voters within a state. I know the districts around me are going for Hillary because they have lots of senior citizens, but 1 hour north of me, the voters are going for Barack.
I don't think the polls are accurate at all. It all depends on the source of the poll, the people that are answering , and several different factors. For example, CNN polls in my opinion are biased, if you watch their coverage on the democratic side the air time they give Obama is tremendous over Hillary. Also, people that answer internet polls may or may not vote . Random votes do not reflect the majority. So, I take the polls with a grain of salt particularly in this election.