Willy's op brought this to mind ; as someone who can be charitably described as, "ahem", obsessive
LOVE polls, particularly political polls. However, their accuracy has been called into question. Let me cite an egregious historical example : '48 POTUS race, Dewey favored to win, but, Truman pulls out upset of century & wins instead. How did this happen ? Fucked up polling methodology : pollsters heavily polled on phones,
which were considered something of a luxury item in '48, that largely belonged to folks who were Republicans. Those without phones ( largely Truman supporters ) were not polled, hence the POTUS "upset".

Flash forward to today : vast majority of polling done on landline phones; here's the rub; large slice of electorate only has cells & no landlines. No polling is done on cells. The trend for folks eliminating landlines is sure to increase exponentially by '12 POTUS race.

IMHO, the ' 12 POTUS race will be a close one. With the accuracy of polls called into question, I ask you this : can we have a repeat of '48 Truman upset in '12 ?

I scoff at polls.
Most polls done months ahead are partisan, or they contact insufficient realistic sampling, to generate a true analysis of voters intentions.
Months in the future polls are as accurate as flipping a coin.
As you can see below,I am not wavering on my view of polls.

http://www.theeroticreview.com/discussion_boards/viewmsg.asp?MessageID=117423&boardID=39&page=

Obama will easily win the election in 2012.
There is no Republican available that has a chance in the 2012 Presidential election, unless Reagan or *"JFK"* is reincarnated.
America might have millions of ignorant voters, however they are vastly outnumbered by voters, who won't vote for monopolies.
The Republicans have a good chance of taking over the Senate and keeping the House in 2012.
No way I am taking a chance and voting on a Huckabee or Romney dictatorship.
Why I didn't include Sarah Palin. She quit her job as Governor of Alaska.Standing alone ,that fact disqualifies her.



-- Modified on 12/14/2010 7:59:34 AM

Opinions and votes.  
The exit polls are a liberal farce and meant to deceive the stupid among us. Last minute, undecided voters can be swayed to vote for the "winner". No one really wants to be a looser.

The long range polls are useless unless the source and those polled can be identified as a block. That's another liberal trick. Poll where you already know what the consensus is and try to pass that off to everyone.

I don't have a clue whether the 2012 will be a close one or a landslide. It would stand to reason that a reinvigored Obama could win by a landlide against a poorly regarded Republican, like a Palin or a Huckabee. It would also seem likely that if the GOP gets their act together and nominates that "Ronald Reagan" type candidate, he/she would still only win in a close race.

It's way too early to make any type of informed prediction about 2012. I doubt that any of the GOP candidates that are household names today have a snowballs chance unless Obama and the Dem completely implode, but we have well over a year for someone to emerge. Remember 2008 and 1992, noone had even heard of either Bill Clinton or Obama two years before either of those elections.

Why is Rasmussen poll generally thought to be conservatively biased ?! And why was "GOP tsunami" accurately predicted by all the polls in this election ?

Btw it's spelled "loser", a word you should be familiar with---------

hence why many of them are paid for by network and paper news companies. Regarding your specific issue about phones, some polling organizations, I can remember which, poll both land-line and cell phones, so they will not be as susceptible to the issue you are addressing.

at least not in the literal sense. lol

Don't you know that all bias is liberal bias? or to listen to liberals, all bias is conservative.

Sounds kind of stupid regardless of which side is making the claim, doesn't it?

and not to encourage you to continue acting as the TER branch of the spelling police, but why do so many losers not know how to spell the word? It's mind boggling how many people spell it "looser".

Polls have their purpose, and are often times right on the money, but polling an election two years out is a complete excercise in futility.

There are always statistical margins of error. And as land lines become less popular, it tends to shift polling more towards older more conservative voters who still have them.

However, prior to the election, I was looking closely at the national polls over at electionprojection.com.

The polling they conducted, and the election results were very close.

One poll may be on or off, but when you average out all the polls, it'll give you pretty accurate results.

Long range polls are less accurate because you don't know what will happen in two years that may change people's minds and priorities.  In 2006, Iraq was a huge issue. By 2008, it had dropped in importance.  In 2006, no one thought the economy would be the primary issue, almost to the exclusion of all others.    

Terrorism is a concern of a certain percent of the population now, but if there were another major attack that does not flop, it may increase in importance.

Short range polls are the best (actually, only) measure that anyone has.  The fact that every major party relies on them shows that they think there may be something to them, and there is no substitute.

All methods of anything - including polls - can be improved.  

All sociology and anthropology is to a large extent a question of polling, unless you have a pure "particpant observer" study, and even then the anthropologist is asking, "Do you ....."

Political polling is as accurate as the means. If you only call people at home at noon on week days you will get a different sample than if you call at 6 in the evening.  If you call people at both times, you sample increases in accuracy. Add a few more, and it gets better.

If you try to find out where your polling is off and adjust, it gets better.

unfortunately there is no way to procure a "cell-phone" list, however pollsters can purchase consumer data (know how they ask you for your phone number at certain stores?).  until pollsters can open up polls to anyone with a cell phone, the polls will become even more unaccurate.

i wonder how pollsters will make their money in '12-- should be interesting.  maybe the switch will be to focus groups, but those are incredibly expensive..

Every poll has a "margin of error" that will give you an idea how accurate it is.  
But all polls are a measure of peoples opinions at that point in time.  It is not a prediction of an election outcome.  People can and do change their minds for any number of reasons.  
Also you need to look to see how it was conducted.  Some polls only poll likely voters.  Others ask questions of whoever they reach.  The latter is less accurate because some responders won't take the time to do their civic duty and vote.