In defense of pollsters
By Steve Benen Tuesday Jan 08, 2008 7:20pmEveryone in the political world saw all the polls over the last several days, and they all pointed to a big Barack Obama win in New Hampshire. Given Hillary Clinton’s narrow victory, this has prompted many to suggest polls shouldn’t be taken seriously anymore. I think that’s a little rash under the circumstances.
Josh Marshall reminded us overnight that, by and large, polls are usually right.
[B]y and large they have a very good record. It’s silly to think that we — whether ‘we’ is reporters or political junkies or ordinary voters — are going to ignore the information that’s right in front of us. And why should we?
It’s true I guess that in an abstracted reality we could simply listen to the candidates, ignore all probabilistic data available, go to the polls with no idea of the result and learn the outcome the following morning. But that’s not the world we live in nor do I think it’s one I’d want to live in.
Agreed. Pollsters put surveys in the field, and tell us the results. The numbers offer us hints of what’s to come. When all the polls agree on a likely outcome, far more often than not, that’s what’s going to happen. Yesterday was obviously the exception, but that doesn’t necessarily mean it’s time to trash the rule.
On a related note, Matt Yglesias offers an interesting observation (and accompanying chart) from one of his commenters: “No one is talking about how the polls actually nailed Obama’s number. Obama didn’t lose this election. He stayed steady and Hillary surged ahead.”
Login or Register to post comments.
The fix is in. Demand a hand recount.
Would we not be better off if we did not hear one single poll two years before the election? I think a lot of us would vote differently.
Obama and McCain both polled well with independents. Independents were allowed to vote in either primary.
When the polls made it look like Obama had it going away, independents who had originally planned to vote in the Democratic primary for Obama switched and voted in the Republican primary for McCain.
McCain did better than the polls indicated and Obama did worse. The polls themselves caused the aberration.
A classic Heisenberg uncertainty problem ...
If you vote based on what the polls say about the candidates, your favorite candidate will probably will not win.
Oh and no more electronic voting. So shouldn't polls be done by paper/in person as well since they are subject to manipulation and fraud?
Perhaps we shouldn't take Diebold seriously anymore. Voters should be demanding a paper trail.
Agreed with Ron. Vote your heart, don't vote strategically. If voting strategically cost Obama this primary, then he should make that part of his message.
The outcome of the vote is NEVER certain until the votes have been counted. So let that be an important message to Obama's supporters about over-confidence.
But I agree about the Hillary Surge. Obama did nothing wrong, Hillary just rallied her side better.
Ron @ 4:
typo, but you get what I mean.
E-voting sucks! Proven to be flawed, yet they continue to use. Go figure. Democracy is a farce.
Mr. Marshall, why wouldn't you want to wait until after voting to know the winner? Is it an addiction to instant gratification that makes you not want to live in such a world? What is it that makes people want to know what happened before it happens? NO ONE AND NOTHING KNOWS THE FUTURE.
I don't take polls seriously. The polls are not going to inform us that Obama for example voted to continue the war each time financially. You have to check his voting record which means putting out a little extra effort.
Polls serve a purpose but they can say . . A new poll says a majority of Americans have decided to jump off a cliff. And some people would think "wow why not everyone else is doing it."
No thanks I can do fine without them.
curtilingus @ 2:
Think of how much Iowa and New Hampshire would matter if there were no published poll results going in. If you think the press overplays those results now . . .
We'd be better off with an educational system that did a better job of conferring mathematical literacy on the general population.
get rid of the polls and the talking heads.Just give the people the truth....and let them make up their own minds.......why not?
while polls maybe usefull, i think they are used WAY to much by the media. Instead of talking about the candidates and thier issues, they just focus on the top two or so candidate and talk to death aout them. Im not a John Edwards guy, but I am kind of miffed on why he gets no play in the media. But then, if just comes down to the polling. He gets what..10-15%, so in the medias eyes he is not a contender. To me, that seems rediculous
L.A. Confidential @ 11:
they have to have something to put on the idiot box.......................not to mention the influence........for the sheeple....
http://www.votefraud.org/
"The Nation's First Website Dedicated to Exposing and Combating Computerized Vote Fraud"
While, in general, I have to agree that polls are usually correct predictors, Marshall's first sentence is belied by the elections that are often cited as demonstrations of the Bradley Effect. Well, maybe not belied, since it's his own memory that he's talking about (and I don't have Andy's ability to see into other people's minds), but you get the idea.
POLLS??? Kerry won Ohio in 2004.
Voting based on what the polls say is almost the same as betting on the super bowl based on what the spread is. The bookies almost never lose.
What? Are you crazy? We all know how notoriously unreliable polls are. We're tired of being told by the media (blogs included) who the winners are. That's as bad as picking movies to see based on the opinions of idiot movie critics.
ineedanemail @ 14:
You've raised a good question.......why did Edwards come in at 17 percent....and stay there the whole time.....didn't budge all night....from beginning to end...17%
Yes, believe in the polls.
Just imagine what might happen if people actually voted their conscience, or for the good of the country, instead of "sticking with a winner."
I ALWAYS lie my ass off when it comes to polls. Polls are used by big media/govt to manipulate their customer/victim base.
Everyone should lie to all pollsters and marketers.
mudshark @ 21:
I have been asking that same question all day. Diebold maybe?
http://www.youtube.com/watch?v=PiiaBqwqkXs
Seriously... there is a reason we do polls. They are statistically predicable.
Something is happening AGAIN!
WE NEED TO STOP IT NOW!!!
WE NEED TO HANG FOR TREASON THOSE WHO ARE MANIPULATING OUR SYSTEM!!!
Why isn't there a post with all the voter fraud news?
http://www.youtube.com/watch?v=MBfmksI_wYY
http://ronrox.com/paulstats.php
Ron @ 24:
Thank You for the U Tube link....but I couldn't watch the whole thing............I might go berserk..........know what I mean......thanks again.
The polls are unreliable for a bunch of reasons. The main problem is that you reach very few voters, so you have a built-in error right from the beginning. Polling has to assume that you are drawing a random sample from the population that you want to test or else none of the math makes any sense at all. The issue with cell phones is one of the sorts of things that cause problems -- the poll results are from a "random" sample of people without cell phones, who happen to be home, who answer the phone, and are willing to take part in the survey. By the time you meet all of those conditions the assumption of randomness that all of the statistics are based on is violated.
Another serious problem is that you are trying to answer two or three questions at once, and the simplistic margin of error does not capture the true error. In NH you are trying to find "likely voters" who are going to vote in a particular primary. In theory you have to start multiplying the margin of error at that point, but the assumptions of randomness are so far gone by then that the math does not match reality.
The margin of error that is reported is always based upon a binomial outcome -- a survey that answers two questions. The math is based upon detecting whether or not a coin is "fair." When you have more than two outcomes the simple form of margin of error is meaningless. The margin of error by the way is a very simple caulculation based upon the estimated population size and the sample size. There is no complex math behind it.
Note that the way that most people think about margin of error is backwards. The margin of error is the amount that the results must differ by for you to be able to reject the hypothesis (usually at a %95 confidence) that the two outcomes are equally likely and your results are meaningless. That is all, "close to the margin of error" does not count -- if Candidate A is at %35 and B is at %37 with a %5 margin of error, the study has proven nothing. It does not mean that the candidates are tied -- perhaps you just didn't poll enough people to detect the difference.
Not making a point here, just feeling pendantic.
Voter fraud needs to be brought to the forefront...and kept there.
It’s true I guess that in an abstracted reality we could simply listen to the candidates, ignore all probabilistic data available, go to the polls with no idea of the result and learn the outcome the following morning. But that’s not the world we live in nor do I think it’s one I’d want to live in.
I believe the world you would not like to live in is perhaps the world before television. These countless polls, touted almost daily for a full year or more before the election, are an abstraction. It's like watching paint dry. I simply want to see voting records, stances on present and possible future issues. Voting for someone because you've been told they have a better chance of winning seems risky and lazy.
mudshark @ 21:
The variance of a proportion is at a maximum when the expected value is 0.50. The farther from 0.50, the less it should vary as the sample grows over time.
Nothing to see here.
Keith O. just said on his show that Obama won in the areas that had paper ballots and Clinton won in the areas that used Diebold machines.
I wrote down the returns last night as they were coming in.
At 2% counted - Obama 39% Clinton 37%
At 3% counted - Obama 37 % Clinton 3% Edwards 17% (his didn't change all night)
At 3 % again at 6:33pm
It was Obama 37% Clinton dropped to 33%
She stayed at 33% up to 6% when she jumped to: Obama 36% Clinton 38% at 7% counted.
I have more... but not sure what it means. Funny how Edwards hit one number and stayed there all night untill the very end.
I'm for Edwards.... would rather have Hillary than Obama... but most of all I'd like a gd democracy w/out fixed elections again.
Clinton signed on to the Neocon prowar agenda, so the least they could do is get Diebold working for her. I don't think the war party thinks a republican can win, so they better
get a democrat that does their bidding. And that's what she is.
The fix... is in.
Nice work FriendsofSamAdams! :)
Polling is an excellent way to counter voting fraud. Multiple independent polling does usually predict the outcome of elections within a few percentage points. I think the problem here comes from the sudden drop-out of both Biden and Dodd a few days before leaving many voters who were undecided heading in to cast their ballots. When the polls and actual voting results differ significantly, that is when people should be calling for a hand recount.
Is there any evidence that Republicans were ordered to the New Hampshire ballot boxes to keep Hillary in the contest? Emails?
JTM @ 31:
I think not.We need to scrutinize everything now..if anything looks irregular...we need to thoroughly check it out.
Leave nothing to chance.
"When the polls and actual voting results differ significantly, that is when people should be calling for a hand recount."
Moose, Clinton got a plus 5% improvement on Diebold counting versus hand count where as Obama did 2% worse on Diebold machines than handcounting... that is a 7% difference... a SIGNIFICANT difference.
Count me as one who would prefer to do away with pre-election polls altogether.
Need to fix this: at 3% , 4% & 5% counted Clinton was at 33
then jumped to 36% and 38% when 6 percent of the vote counted.
friendofSamAdams @ 32:
I wrote down the returns last night as they were coming in.
At 2% counted - Obama 39% Clinton 37%
At 3% counted - Obama 37 % Clinton 38% Edwards 17% (his didn't change all night)
At 3 % again at 6:33pm
It was Obama 37% Clinton dropped to 33%
She stayed at 33% up to 6% when she jumped to: Obama 36% Clinton 38% at 7% counted.
I have more... but not sure what it means. Funny how Edwards hit one number and stayed there all night untill the very end.
I'm for Edwards.... would rather have Hillary than Obama... but most of all I'd like a gd democracy w/out fixed elections again.
obviously technology should not be trusted when it comes down electing the President of the US................I'm sorry...but that's a no brainer......how stupid can people be?
I think election polls should be outlawed.
The polls are usually correct, however, we witnessed the corporate media trying to end her campaign.
They have CDS and should be fired.
friendofSamAdams @ 32:
More technically (than post 31): the variance of a proportion equals p times (1-p) divided by N. This is why it is at a maximum when p = .50. Plugging in some values, when p is around .37, as it was for Obama and Clinton, variance is .37 times .63 divided by N or .23 over N. When p is down around .17, as it was for Edwards, variance is .14 over N. Farther from .50, more variance.
Now, if you have something like a series of percents and the associated Ns, and find that a shift in percent is impossible for the associated shift in N, then you'd have something. For an example of such a smoking gun, search around for Ohio precincts with more votes for a candidate than living population. For an example of something more subtle, google Gregor Mendel.
I think that pollsters must get nervous when they see Diebold machines... and it goofs up their polling data.
Or maybe the islamofascists created fake polling data in districts with Diebold machines to subvert our democracy.
There has to be some simple, rational explanation that accounts for the discrepancies between the hand counted districts and machine counted ones. And as a patriot I cannot accept the idea that there is anything wrong with the machines themselves. Machines don't make mistakes. Obama supporters do.
I think Obama fans are suffering from over active imaginations. Just because Obama lost New Hampshire doesn't mean there is vote fraud.
Regardless, now we have Dems fighting each other and charging Diebold! and Fraud! in what was just probably an Obama loss.
Weird
pollsters can't account for Diebold!
The Constitution @ 38:
I agree. I'm not saying for sure there is any type of foul play going on here (Clinton had a sizable lead in every poll a couple of weeks before the primary). But, I am saying, a call for a hand recount is very reasonable given the discrepancy between the polling data and the actual outcome.
As a Dodd supporter who doesn't give a shit whether it is Obama or Clinton, I can sit back and observe this process with some aloofness. The ludicrous speculation about how the polls could have possibly been wrong is quite entertaining. Josh Marshall's sniveling essay was among the worst--he has been one of the chief cheerleaders for the Obama machine and he was just dead wrong about the inevitability of it all (sound familiar?). If he hadn't selectively chosen polls that showed Obama way up for the last few weeks, he might not be so shocked.
L.A. Confidential @ 46:
It's not vote fraud, it's election fraud. Diebold was used in 80% of the precincts. The results come in from different precincts at different times. The percentages stayed the same. Are you saying that all the people in NH think on the same level throughout the whole state? The percentages should have fluctuated.
http://www.youtube.com/watch?v=PiiaBqwqkXs
L.A. Confidential @ 46:
Not just obama...Diebold sucks...bad!http://digg.com/2008_us_elections/New_Hampshire_District_Admits_Ron_Paul_Vote_Skew
L.A. Confidential @ 46:
What I find weird is that Edwards came in at 17% and stayed there....all night long................that to me is peculiar
the media drives the polls. media played and replayed hillary's tearing moment where she came across as sincere and the next day all the polls turn out wrong because of that media coverage. the media showing hillary's human moment won her new hampshire.
Polls do not serve any purpose that is useful to the citizens of the US.
They are a tool of the media, to give them something to talk about, to sound more authoritative when foretelling the future. The are also a tool which can be manipulated by whichever 'Powers that be' control them.
Neither of these functions are useful to citizens, in fact, they have a negative influence, or at least the potential of a negative influence, with no positive, or 'up' side, in favor of them.
Unless you consider that the gossip and speculation they generate is positive, in which case you may find satisfaction in a lifetime subscription to a supermarket tabloid, with features of deformed interstellar mutants hacking into voting machines to allow their 'pets', the multi-legged dogs of Sirius, to become the elected officials of the US.
Stephen @ 49:
Which poll are you talking about?
2008 New Hampshire Democratic Presidential Primary Polls
New Hampshire has an estimated population of 1,309,940
96.97% White
1.29% Black
I mean come on lets be real. Obama didn't do to bad when your really think about it. There's not that many young people there either.
LOL
vote fraud: the ugly elephant in the room
Statistical probability is a science.
Bush and the rethugs hate science.
If you rig elections, as was done in 2000 and 2004, it makes science look silly. Everybody points at the pollsters and laughs.
And this puts people like Bush in office.
Polls are like a Svengali looking into a crystal ball......foretelling the future.......BS.
L.A. Confidential @ 56:
Got any specs on Iowa?
prior to the 2000 election, the media didn't provide info on bush's shady business dealings or much of his past history at all. how many people were aware that laura bush was involved in a fatal car accident. had hillary been involved in a fatal car accident, it would be talked about all the time.
Polls, particularly exit polls, are a good test of vote tabulation equipment. The majority of the New Hampshire primary votes were tabulated by Diebold on equipment that has been proven defective. Results are showing that the towns and counties that hand counted votes has Obama ahead by four percent.
No one in the media is willing to examine this angle ... See
News Updates from Citizens for Legitimate Government
09 Jan 2008
http://www.legitgov.org/
http://www.legitgov.org/index.html#breaking_news
and, of course ...
http://www.bradblog.org
Ohh, warm soup for my soul. Please, let's get over it, the fact is that although I'm a Hillary supporter, I believe that Obama can be an excellent president too. In fact, anybody on the democratic side is better than anyone on the other side. We have a bigger enymy here and it's name is conservatives.
mudshark @ 52:
Edwards' constant 17% is the result of the Pre-Vote Determination Meeting held by BigOil, Goldman Sachs, AIPAC, and the US Chamber of Commerce, at which Edwards drew the short straw.
L.A. Confidential @ 56:
excellant point......but how do you explain Edwards?...I'm not trying to pick a fight with you LA....but this really bothers me.
CoIntelPro @ 57:
I thought Rudy Giuliani was the ugly elephant in the room...
Lollimom @ 60:
Iowa has an estimated population of 2,982,085
91.0% white
2.5% black
Damn! Obama! DIEBOLD!
LOL!
I read somewhere that there was a discrepancy between the hand counted ballots and the Diebold (computer) counted ballots... and that Obama actually won. Anyone know if this is true or not???
mudshark @ 65:
I don't know.
Interesting... so the Diebold machines were used to keep Hillary in the race?
Well, it's possible, I suppose. Some geek should borrow one of the machines and check the BIOS or the firmware or whatever the hell...
L.A. Confidential @ 56:
been using your brain long?
For that matter how the Hell did McCain rise from the dead in New Hampshire?
He wasn't even on the radar.
CoIntelPro @ 71:
Alright now. Don't start up with me.
L.A. Confidential @ 72:
McCain bypassed Iowa...and focused on New Hampshire...that's what he did in 2000...and it payed off.again.
L.A. Confidential @ 67:
You need to up your percentage of whites by 3 percentage points.
I'm keeping an eye on the Brad Blog. I see vote irregularities.
L.A. Confidential @ 72:
Not true. The polls predicted the outcome for the Republican Primary fairly well for each candidate.
After the 2000 election and the 2004 election......if we didn't question any of the results...that would be foolish.We all bitched and complained over the previous election results....who's to say anything has changed...These Diebold machines and the like are the most serious threat to our Democracy...and should be thrown out.
mudshark
Not only did Edwards stay at 17% all night, but when only 10% of the districts had reported in, CNN announced that he would come in third but didn't announce the Clinton Obama ranking until about 95% had reported in. Given his 2nd place win in Iowa, how did they know he'd rank third in NH? The state's voting population is 45% independent and many had not made up their minds until the last minute. Given Hillary's 3rd place in Iowa, how did they know, long before the votes were counted, she'd rank above Edwards?
I REALLY dont' like what news agencies do to elections and voting. As someone else said, no ranking or winners should be announced until all the votes are counted and recounted if needed.
L.A. Confidential @ 73:
there was no racial angle to the Obama numbers. were you making a joke?
LanceThruster @ 76:
I haven't see Obamas official camp protesting the vote count. Have you?
I didn't see Hill's camp screaming Diebold! when she lost Iowa.
Maybe they are both part of some vast left wing conspiracy secretly working for the Neocons.
There has to be something right?
Jeff @ 68:
According to this link there was a discrepancy in which the machines favored Hillary +5.4% and disfavored Obama 3.0%. (On the repub side Romney got a +7.5% bump. This compared to Huckabee, Paul and McCain who all seem to have taken a -2/-2,5% hit):
http://ronrox.com/paulstats.php?party=DEMOCRATS
anney @ 79:
Thank You for this......being in Calif...most times it's a foregone conclusion...and Calif doesn't really matter.....but apparently..Calif is a player this time......we'll see what happens.
L.A. Confidential @ 81:
Iowa was a caucus, NH was a primary. Voting machines or ballotts are not used in a caucus.
CoIntelPro @ 80:
Yes I was but I'm not surprised Hillary went over the top there with all them Old NH Hillary Fans, especially the Women out in force. I suppose Kissinger could have ordered the Diebold buttons pushed, to keep his boy toy Kristal from being embarrassed by that stupid NY Times debut article he wrote.
Ron @ 84:
Right. Regardless what would we Americans do if we weren't paralyzed by some sinister plot about everything that goes on in this country?
That was the Neocon goal wasn't it? Spread fear and paranoia?
Off-topic here and FWIW:
The Woman Who made Hillary Cry Voted for Obama
The woman whose empathetic question — "how do you do it?" — sparked uncharacteristic emotion Monday from Sen. Hillary Clinton ended up voting for Sen. Barack Obama in the New Hampshire primary.
Marianne Pernold Young, 64, a freelance photographer from Portsmouth, N.H., told ABC News that while she was moved by Clinton's emotional moment, she was turned off by how quickly the New York senator regained her "political posture."
"I went to see Hillary. I was undecided and I was moved by her response to me," Pernold Young said in a telephone interview with ABC News. "We saw 10 seconds of Hillary, the caring woman."
"But then when she turned away from me, I noticed that she stiffened up and took on that political posture again," she said. "And the woman that I noticed for 10 seconds was gone."
But in the end, she said it was Obama's message of hope and change that won her vote.
"I went to see Obama on Friday and he moved me to tears, I was in awe," she said in a telephone interview with ABC News. "I'm 64 years old and nobody does that to me."
L.A. Confidential @ 85:
then I'm sorry for the question. It dawned on me after a re-read.
but nothing as yet has been able to embarass KKKristol. or Kissinger, for that matter.
L.A. Confidential @ 86:
Without a doubt. That's why Rudy keeps repeating, 9/11, 9/11. Be afraid, be afraid.
Is it inappropriate for me to find Rachel Maddow attractive?
CoIntelPro @ 88:
I'm with on you on that.
Ron @ 89:
Check out this McCain commercial.
http://www.youtube.com/watch?v=bce2rev5mq8&feature=user
anney @ 87:
Wow...There's an endorsement for ya. Obama has the post-menstrual, Oprah book clubbers getting all wet over him.
Cue the Porno Music.
CoIntelPro @ 80:
"...no racial angle..."?
Are you making a joke??
Johnny2Bad @ 94:
Trying to maintain some sense of humor. I have lots of great black friends and always have my entire life. I even got beat up one time when I was kid because I had a black friend I was buddys with.
I'm no Racist.
Wait a minute...hold up.
If Hillary was ahead in New Hampshire for most of 2007...then suddenly a few polls in the same 2 day timeframe show Hillary down, is it really Hillary that "surged"?
The POINT is that everyone was ready to call Clinton's campaign dead...and eager to do so.
Disgusting.
I hope she sticks around, and wins.
Johnny2Bad @ 93:
HAHAHAHA, seriously man...."oooo, he's so handsome!"
Josh @ 96:
The media was running story's with headlines like Clinton Campaign Collapses! Clinton Backers Flee! Hillary Finished!
Hillarious. The media must really think people are stupid.
Hillary's the safest candidate out there...every bit of her life has been examined; there's a reason why republican nutters don't want her to win.
Not electable? Is that why millions were invested before she even announced her run to "stop Hillary"?
BS!
Obama's voting record is identical...don't care for cocky, and his performance in Chicago while I lived there was not exceptional.
What did he do? WHAT DID HE ACCOMPLISH.
L.A. Confidential @ 98:
Can't imagine how that might be.
I want all 50 states to have their primaries on the same day
I don't suspect vote fraud. I suspect hell hath no fury like a losing woman scorned by other women.
L.A. Confidential @ 81:
You think Hillary is "left wing"??? LOL back at you, fool.
I just can't believe all the conspiracy theories all over this blog, don't you dumb asses ever think that maybe she just kicked his ass and yours. Now its the polls. The polls were right about obama, look it up. What they didn't know was that the women came out for clinton. Everything she does is questioned and she did not cry she got teary eyed. She was around women in a small setting, she was vulnerable and tired. I don't care what you people that suffer with the HILLARY-HATRED-SYNDROME say she is real,but she has had to get a thick skin to put up with all the vile and mean things the right wing and the media has put her through. You morons don't even remember the nineties or maybe you were in pampers. And unlike what christ jr. says the past matters it forms the future. By the way this change mantra is getting old, change to what. He would do the world better and become a bonified preacher.
I'm gonna keep an eye on www.bradblog.com for sure from now on.
This is scary. All the people suddenly saying that we don't need any polls are scary. Who puts ideas into the heads of these people.
It's getting pretty obvious that NeoCon thinking is seeping into the Democractic base (that, or that partisan Democrats have always been insane and stupid). Anti-democratic talking points going around in the Democratic base? Plenty of it on C&L?
Still moving towards becoming a bana republic, I see. The Democratic base is split between reason and irrational ignorant people who will repeat what Bush voters did in 2000 and 2004. Don't people get the serious situation the country is in?
Will the fact that Bush came to power through a coup go away on its' own?
Hillary is not my first choice, but I like the fact that she won New Hampshire. This competitive race is at least energizing people to get off their butts and go to the polls. I want the democratic nominee batte tested for the GOP scum machine in the general election.
A bit off topic but:
Does anyone remember Bill Clinton's C-SPAN broadcast seminar for college students in Arkansas a little more than four years ago?
His prescription for winning a presidential race contained three items:
A) show people that you have the qualities that people like in your opponent.
B) show people something about your opponent that if people knew it, they wouldn't like them so much.
C) show people something that your opponent can't possibly offer that you can.
First Hillary had to be for change, second they trotted out the (false) Iraq war and abortion claims and third Hillary was a woman..something Obama can't be.
Exactly Bill Clinton's prescription for winning, by the book.
This is so blatantly voter fraud. Examine the numbers by the candidates from those that were hand counted and those that were machine counted. The polls weren't wrong, the machines are fraudulent and nobody's talking about it... at least even on C&L, Commondreams, Huffington... well huffington is shit now but anyway... privatizing vote counting is the most overt affront to democracy right now.
I'm sorry, but I think Matt Yglesias' premise is wrong.
Polls look at groups of people like "likely voters", those who will PROBABLY be voting. That's what lets some of them be reported as "Undecided", etc.
However, actual vote totals count only those who ACTUALLY voted. This is not the same group as the former, unless every single "likely voter" actually made a decision and voted for someone.
Thus, claims such as "the polls got Sen. Obama's percentage correct" can't be right...they can be close, depending on how many "likely voters" stayed home, but they can't be right.
Also, it's a matter of scientific fact that polls have not just a "Margin of Error" (the idea that the specific percentages are within X points of what we'll actually see), but also a "Confidence Interval" (usually 95%). That means that the Margin of Error will be seen in 19 of 20 cases. That 20th doesn't necessarily indicate a flawed poll, however, and people need to remember that.
Sure, polls with results that differ from reality, as seemed to have happened here, may bear further scrutiny, but they're not necessarily "wrong" or "biased".
L.A. Confidential @ 100:
Followed by a runoff, or perhaps with instant runoff voting. Mark your first three choices, in order, folks.
The only defense I can think of for pollsters is that they're psychotic, take themselves too seriously, attach messianic significance to their every slightest whimsical impulse, and are remiss about taking their meds on a regular basis.
L.A. Confidential @ 98:
Correction: The media KNOWS people are stupid, and can prove it:
They put up with the media, don't they?
really they both surged and Edwards tanked. Obama did better than he was expected to pre-Iowa, Clinton did slightly worse than expected pre-Iowa, and the Iowa results made things that much closer.
Look at the psychology of this. The people who are explaining the discrepancy in the polls and the outcome are searching for any reason to continue to believe in the election system.
It was: Obama couldn't win because he is black.
The independents switched to McCain because Obama was a sure thing.
Women were angry at Chris Mathew's sexism.
Liberals were angry at the sexism of the media's attack on Hillary's crying.
The Polls are fraudulent and being manipulated by the media.
Polls are never correct.
You are grasping at straws because you want to believe that we still have a legitimate democracy. It makes you feel all warm and fuzzy like evangelicals believing in Noah's Ark.
If a bank safe (voting machine) full of money is open (easy to hack) to greedy criminals (ambitious politicians) than don't we have a right to question whether the money was stolen.
Polls are used throughout the world to spot election fraud. If Clinton continues to win in the Diebold Districts you need to ask yourself at what point will you question your belief in the system. If it happens again in Michigan? Super Tuesday? For many people on this blog the clear answer is NEVER.
L.A. Confidential @ 95:
Of course it was a joke. Just juxtapose the black and white and you got my story, we had a few run-ins, too. who knew?
neoconned @ 114:
worth repeating
Bram Weiser @ 109:
Exit Polls are supposed to reflect the actual votes. Until Diebold, TSS and Sequoia, Exit Polls actually DID generally reflect the actual voting results
CoIntelPro @ 117:
Thanks, but I beg your pardon...Exit Polls are an ATTEMPT to reflect the actual voting results. A scientific attempt, it's true, but an attempt.
Not only that, but, as I understand it, it's not that Exit Polls were "wrong" in this case, but that the polls leading up to Primary Day were "wrong". As you note, that's a far different kind of survey.
Login or Register to post comments.