1 This is what I predicted. There is ONE reason for this bounce.

No, the convention hasn't had much crossover appeal, maybe a bit. Maybe a point or two. I doubt even that.

However. Many of Hillary's voters are going to come home at her and her husband's direction. A lot were out to sea.

That's why Obama's gaining ground.

2 If there is no bounce, it will be played off as backhanded racism at the sight of the first black nominee from a major party.  If there is a bounce, it will be played as undeniable surge of support for The One.  So the actual quality of the convention (which so far has been a big meh if you ask me) or the traditional bounce due to non-stop exposure that most conventions produce will be irrelevant.  That is, until next week when no one in the MSM will be surprised by any McCain bounce (since it's wholly expected), or no bounce will be forcasted as the death knell of the Republican Party forevermore.

3 It's a massage.  They massaged Barry's numbers down before, then up now.

And even if -- 6 points after the Dem convention?  Ha ha ha.

4 Rasmussen's polling only includes Mrs. Messiah's and the Pant Suit's speeches.  Gallup's reflected  up to Biden and Bill's.  Rasmussen's next poll will most likely show the bounce as well.

5 The "votemaster" at election-projection will have a rock hard boner over this.

6 Personally, I wouldn't expect much bounce from the convention until afterwards. Give it a little time to sink in, as it were.

7 A bounce is pretty typical.  It usually fades, and, if valid, this one is not a big one.

And remember, Vero Possum speaks tonight.  Even if he doesn't embarrass himself and sticks only to the teleprompter, there is still the sizeable possibility of a spectacle of absurdity that would make the Berlin speech look statesmanlike.

8

I said this over at Hot Air and I'll say it here too....

Honestly we’d have to be idiots to worry about these minor polling fluctuations right now, regardless of what direction they turn. These are itty bitty snapshots and it’s all within the margin of error. McCain will get a little bounce after the GOP convention too and we shouldn’t get too excited about that either.

We need to be focused on the long term polling trends, which are slightly better at the moment for McCain than they are for Obama. Still it’s way too early to matter much. It will start to matter a whole lot more if one of them seems to be pulling away from the other in 4-6 weeks. That’s a long time in politics. Let’s all calm down.

The big thing we need to be concerned about, and no one is really talking about it, is turnout. In the primaries the Dems had something like twice the turnout we had. If that’s the case in November we are in big trouble.

9 I'm sorry, there are not that many wishy-washy , back and forth type voter from one week to the next for that big of a swing.  It only proves polls are a farce according to 51% of statiticians.

10 "And even if -- 6 points after the Dem convention?  Ha ha ha."

The convention isn't over yet. You've no good reason to laugh.

"I'm sorry, there are not that many wishy-washy , back and forth type voter from one week to the next for that big of a swing."

No, polynikes, it proves you didn't understand what I said in my first comment or Allahpundit said in his post. It's not a bunch of wishy washy voters. It's Hillary supporters who were withholding their support until they heard it from her lips, which they did, reinforced by Bill.

"..."
Posted by: t.ferg at August 28, 2008 03:14 PM (2YVh7)

Hear, hear! Every word you said.

11 And why are we trusting these polls anyways? 

Wasn't Kerry beating Bush for all of 2004 by about 3 or 4 points? 

12

I call Bullshit, BUT remember Mike D was up EIGHT after the democrat convention, Kerry was up six.

How'd that work out for them?  Not that good.

Let them have their LAST moment of "HOPE". 

The Bad Boys are coming to town!  It's NOT going to be pretty!

As they said in Animal House,

EAT ME!

 

13 Republicans excelled at turnout in 2004, Dan. This time around the Democrats have lapped Republicans in primary turnout. Not a good sign.

14 Any poll that includes 'likely voters' is GARBAGE.  Oh and when the media says it's SO CLOSE that usually means the Repubs are ahead and are just padding the numbers.  So when I see a poll that says 'neck and neck' I don't believe it. 

15

Christoph I don't read your posts. The only reason I read your last one is because I saw my moniker.

By the way I took a poll it it showed people think Christoph is still a douche by 2 to 1. 

 

16

The convention to me has seemed weak.

Yeah, an Edwards speech would've butched it up a little.

17

Do any of the morons remember the "Bradley effect". 

He was going to be the governor of CA ahead by 6 points, but gee, I guess people lie about voting for a Black person, racist assholes.

So how much did Bradley lose by, if I remember it was 3%, which is a NINE point spread.

Now, we don't have any of that now, so NO ONE is lying about who they will vote for. Right?

Believe that? I have a bridge I want to sell you near China Town, call me.

 

18

1. Cristoph is likely right about some Dems coming home (they had been eroding pre-convo;

2. The bump also comes not from the DNCC as much as the MSM coverage of same.  Relatively few people watch the convention coverage directly.

3.  There has also been plenty of volatility in the middle for the past few months.  People who read blogs always should remember that there are many people who are just starting to pay attention to the election now.  So I would think that there will be some low info types who see Obama speak for the first time tonight -- with all of the commentary about how historic it is (and it is) -- and be greatly impressed by The One (unless the event really gives off a creepy Nuremberg vibe visually). 

4. I would not be surprised if the bounce gets bigger and does not begin to subside until a similar point in the GOP convention, when the same low info types learn more about McCain's stint in the Hanoi Hilton (we've heard it a zillion times; they have not).

19 Christoph,
I don't think you're very astute regarding polling.
NTTAWWT.

The weighting of these polls (internals) has likely changed to manufacture a "bounce". Pollsters need bounces or nobody pays them any mind.

20 As said before, it's all about voter turnout.

The blacks will probably turnout more, and the kinds perhaps less.  The Republicans are anybodies guess.  GOP voters usualyl turnout much more than dems.  If they do this time, Mccain will rin, and if they don't he will lose.

So if this is really important to you, just volunteer to help get out the vote.  Knock on doors and remind peopel to vote (they have lists of likely Mccain voters).  Drive people to polls in swing states.  It makes a huge difference.  Ten volunteers can easily drive enough people to the polls to flip a state like Ohio.

If you're not volunteering for the candidate you like, then don't complain too much if the other guy wins.  Of course, that's the problem... so many conservatives are stretching just to vote for Mccain, and he could have a problem getting worker bees.  Obama might have an advantage.

21 Wasn't there actually poll manipulation in the last presidential election season though?  I seem to remember something coming up with Zogby, but can't find anything on it now.

22

Re: The Bradley Effect.

People seem to overlook that this is more likely to be a serious factor in heavily Democratic contests.  Conservatives/Republicans have plenty of reasons to tell pollsters why they won't vote for The One besides his skin color.  It's the Libs/Dems that lie, because they should have fewer policy differences.  I'm not saying there will be no Bradley Factor, but whether it will be significant -- and show up in states like MI, OH, PA, etc. -- is not knowable.

23 I am always amused by the conspiracy theories about polling.  No doubt Gallup is making up these numbers at the behest of the Bilderbergers and the Trilateral Commission.

24 and Col Saunders

25 J. Hill is likely correct about the turnout game.  Obama probably cannot jack up black turnout in enough places to make a difference.  The unknown is whether his youth vote gets re-energized, or has been put off by The One's many flip flops over the Summer.  If Obama can get voting for him to be the cool thing to do with your friends again, he can exploit his social networking ops to drive turnout the way the GOP has done since 2002 with their "72-hour projects."

26

Any poll that includes 'likely voters' is GARBAGE. 

To be fair, these guys polled likely voters who just happened to be in Denver this week.

27 kempermanx is grossly overly optimistic and came up with a conspiracy theory without evidence implemented by Gallup sometime between Wednesday and Thursday.

Now, there wouldn't be any other possible explanation for this happening during the same period, like, for example, the Democratic convention and the speeches by Hillary and Bill Clinton.

.ferg and J. Hill are probably right about turnout.

28

Christoph,

WTF, Conspiracy theory?  What you talking about?

 Pray tell where have I said or even implied that there is a conspiracy?

We have alway know there would be a bounce. READ my posts, did that work for Mikie or Kerry? NO.

Get back on your meds.

29 "I am always amused by the conspiracy theories about polling.  No doubt Gallup is making up these numbers at the behest of the Bilderbergers and the Trilateral Commission." -- Karl above

So now it's a conspiracy theory to think a company might act with a profit motive in mind?

Do you think it's odd that every Democrat has been ahead of every Republican at this point in the race since 1968 and yet 7 out of 10 elections have been won by the Republicans? Are you telling me that voter identification in 1968 was less favorable for Republicans than it is now? Do you find it odd that only Jimmy Carter has garnered 50% of the vote since 1964 and yet Democrats have all been ahead at this time in the polling?

Look, I'm sure it's all accidental that the errors have occurred in one and only one direction. But if you play 3-card Monty and lose 10 times in a row, you might begin to think it's a non-random occurrence.

And that's not conspiracy. That's simply prudent.

30   This is not the Bounce I thought I knew

31

Yes, Gallup would risk its entire business on the chance that some whistle-blower or disgruntled former employee outed manipulation of their polls.

Folks need to grow up some.  Polls are far from perfect, but no one (save perhaps Zogby) deliberately manufactures results of a tracking poll.  Other polls can be more subtly affected by the types of questions included, wording, etc.  But manipulation of a tracking poll would be tough to do, and too easily discovered to be worth risking the firm's cred.

32

Without third party candidates, democrats would have not won one race. The big earred asshole from texas comes to mind.

Oh, I will say Anderson helped Reagan, but Teddy got Wilson elected, which started the whole damn socialist shit, and WW I.

Ego's, what a bunch of shit.

Let's pray that Barr doesn't cost us FL.

33 "So now it's a conspiracy theory to think a company might act with a profit motive in mind?"
-- Nom de Blog

It is without any evidence. On Wednesday they had one result. That night, Hillary Clinton with huge numbers of outstanding voters not yet committed to Barack Obama spoke and urged her supporters to support Obama.

On Thursday, Obama was up in the daily tracking poll.

"WTF, Conspiracy theory?  What you talking about?"
-- Kempermanx

I apologize. I got your post mixed up with Nom de Blog. S/he posted the conspiracy theory earlier, and is defending it both without proof and in the face of a more obvious explanation per Occam's Razor and common sense now.

34 The Oracle from Mount Obamus tonight is going to be kickass.  I've got two PVRs recording, no lie.  It's going to be the shark-jumpingest thing I've ever seen and I can't wait.

If he promises a Bubo for every family I'm going to lose it.

35

"Do you think it's odd that every Democrat has been ahead of every Republican at this point in the race since 1968 and yet 7 out of 10 elections have been won by the Republicans?"

No, I think that until the advent of the Internet, most pollsters stuck with polls of adults and registered voters much longer into the process and switched to likely voters later.  But I only think that based on the degree work I did in public opinion polling, so what do I know?

36 "...manipulation of a tracking poll would be tough to do, and too easily discovered to be worth risking the firm's cred."

Certainly between Wednesday and Thursday it would.

37 Donk Bounce makes my clothes smell like ass.

38

We need to focus less on polling and more on voter turnout.

It's not enough to get on down to the polls and congratulate ourselves for voting even though we didn't really like our candidate. Yay for us!

No - that won't cut it. We need to get everyone we know and a bunch of people we don't - people who wouldn't normally make the effort - get them registered and drag them to the polls in November. Otherwise we are going to lose to the democrats' ground game. We'll have Barack Obama making judicial appointments and signing every left wing bill the democrat supermajority cranks out to his desk. And you know what - we'll have ourselves to thank for it because we sat around and reminded ourselves how much we hate McCain and how we can't vote for him because we're too principled to do that and yada yada yada.

I don't care if he deserves it. I don't care if you hate McCain as much as I do. We have to go get this guy elected. If we don't, we get Obama. I'm not ok with that. Are you?

That's all for now.

 

39 I suppose my couple of advanced degrees, including one in public policy in which I took several statistical modeling and polling courses (and the work I've done as a university professor that involving analysis of surveys) comes to naught, as well. Fuck you. Don't try to big ball people on the fucking internet, dip shit.

When all the polls have been consistently wrong in the same direction, it's an indication that the fix is in. Now you might call the historical trends "without any evidence" but that doesn't make it so.

40 Nom de Blog, I'm saying you're a conspiracy-theorist jackass with situational ignorance to boot.

Maybe there's some bias in polling methadology, by innocent error or otherwise.

However, your claim is this bounce was introduced by a sudden change in methadology, by nefarious intentional design, but you offer no evidence at all.

However, what do we actually know. We know that Hillary has a ton of supporters who have been reluctant to support Obama. We heard Hillary on Wednesday say she's supporting Obama and urge her supporters to.

On Thursday, Obama's numbers are up.

Yet you don't get it. Why not?

It really isn't complicated.

41 Christoph,
You are painfully ignorant of the history of American presidential polling.
It's not my fucking fault.

The same bull shit happens every time.
We're now 11 for 11.
But of the 10 that have run to completion, 7 have ended in Republican victories.
And all the results show the lead of the Democrat shrinking right up to election day.
Now if you can offer a reasonable alternative theory as to why this has been true, every time, I'm more than willing to listen.

But when every example shows the exact same fucking thing, I'm not going to take the 11th time to be the "this time it must be for reals" faith-based explanation on its face.

42

So Christoph,

Does that mean that you don't have a fucking clue what I wrote about, or does Kempermanx look like Nom de Blog in your unmedicated state.

You an idiot. What is it about statistics you don't understand, or do you know what the word means??

Inquiring minds want to know, just how many course in Stat did you take, or don't they teach that in jr hi?

43 Nom de Blog,

You're saying pollsters have oversampled Democrats for a long time and incorrectly boosted their numbers.

Duh. What else is new? Neither I nor Karl dispute it, and Karl offers a reason for it. I offer another one, bias, and accept them as both possible.

However, your claim is rather more grandiose. You claim the methadology was suddenly and intentionally manipulated overnight to produce this Obama bounce. You have no proof for this whatsoever.

Did you miss the part about Hillary having almost as many supporters as Obama? And both her and her husband giving speeches last night throwing their supporters to Obama?

You don't think this may have caused this short term bounce?

44 This is not the kempermanx that Christoph thought he knew...
Under the bus, kempermanx...

45 Dan-O: And why are we trusting these polls anyways? 

Wasn't Kerry beating Bush for all of 2004 by about 3 or 4 points?

That doesn't mean the poll was untrustworthy.  It means Bush and the Swift Vets were able to connect with the American people between the convention and the general election, and convince enough Americans that John F. Kerry is a douche bag.

46

Nom,

Given your vast experience, the proof of your theory -- aside from the explanation I just gave -- is what, exactly?

Or did you miss the day where the "correlation-causation" fallacy was covered?

47 I guess Ramussen forgot to poll all those Hillary voters that decided to return to the fold. 

48 Yeah, kempermanx, I get your position. You believe Democrats usually receive too much weight in polls so things are better for Republicans than the polls indicate. I agree that's more than possible.

Further, you believe the "Bradley effect" may be boosting Obama's numbers artificially. Again, I think that's likely for at least a couple points, but probably less than Bradley because I think as time goes on, racism fades in importance. Fortunately.

However, I'm not addressing those issues. I'm talking about Nom de Blog's fantastic conspiracy theory that it isn't Hillary's speech and asking her 18,000,000 supporters to go to Obama that caused this swing, it's gotta be intentional manipulation. Within the last 24 hours. By Gallup.

There's a simpler explanation.

49 No, I don't.
Obama has polled within a fairly narrow range (all below 50% recently) since he secured the nomination. I'm not sure if he's cracked 50% in any of the polls.

Therefore, any changes so far have been within the margin of error. If the polls are within the margin of error then, statistically speaking, there has been no change. Now, if we accept that there's been no change we must accept that there has been no bounce. But if there's been no bounce, then there's no news story. And if there's no news story, then Gallup is a loser in the marketplace.

But that's just the sort of analysis that is "without any evidence" so what the fuck do I know.

50 polynikes, there's a couple possibilities about Rasmussen. One, the data could have been gathered at a different time of day so it may not all be reflected., even though it does show a bump. A much smaller bump.

Or Gallup's poll could indeed be an outlier. These things happen.

There's no evidence, however, that this is fraud. Karl just asked Nom de Blog what his evidence and/or reasoning is. I'm waiting.

51 Karl,
I'll put down these three cups. I'll put a ball under one of the cups. If you can choose the cup with the ball under it, then you win. If not, I win.

Now if I win 10 times in a row, you just keep on betting. After all, there's no causation there. It's just correlation.

If you need to run to the ATM, I'll wait around.

52 Christoph,
You can wait until I post #49 above.
It's just above your latest.

Whoopsie for you, and shit.

53 "Obama has polled within a fairly narrow range (all below 50% recently) since he secured the nomination. I'm not sure if he's cracked 50% in any of the polls."

He didn't crack 50% in Gallup's new poll; he's at 48%. This is a number he was at 2½ weeks ago.

You think the methadology was intentionally changed to reach this number. But you're overlooking the fact that he's in the middle of his convention with his opponent who received almost as many votes as he speaking on his behalf yesterday.

That was actually rather a big deal.

54

polynikes,

Rasmussen is polling likely voters. Gallup is tracking registered voters.

Also, if Nom is such an expert about statistics, Nom should know how reliable a sample set of n=10 is.

55 Oh, and Christoph, please feel free to my post above where I use the word "fraud".
Words have specific meanings.

56 So you'll play 3-card Monty with me Karl?
And you'll keep betting after I win 10 in a row?

What?
We have to wait another 76 years to have n=30 before you'll believe these polls are horse shit?

57

I'll put down these three cups. I'll put a ball under one of the cups. If you can choose the cup with the ball under it, then you win. If not, I win.

It's the one on the right.  What do I win?

58 Nom's #49 demonstrates that he hasn't really looked at the Gallup tracking data. It has an MoE of 2%, so there has been a bounce in its samples.  That BO hasn't cracked 50% is significant, but not very relevant to this discussion.

59 Sorry, JackStraw, you lost.
Please donate my winnings to Ace via PayPal.

60 Nom's #56 is an implicit admission that n=10 is too low for any meaningful conclusion, so he's adopting a three-card monte analogy that assumes the chicanery for which he has no evidence as to Gallup. Not impressive as a matter of stats or logic.

61 #58
It's called noise.
Statistical noise.

62 Son of  a ......

63 Ooh, I love it when Nom throws out a piece of jargon as though it's an argument.

64 Yeah, Karl, if you assume that the same thing has happened 10 times in a row due to random variance, then there's no evidence of the polls being garbage.

Assume the same thing about my 3-card Monty game.

But visit the ATM first.

65

When all the polls have been consistently wrong in the same direction, it's an indication that the fix is in.

Not necessarily. Occam's Razor suggests that this is due to factors that don't require the collusion of various polling groups over decades. Either the polls are measuring something different than what the election does, or they're measuring the same thing in slightly different populations, or (more likely IMO) the polls actually reflect a real but temporary rise in support for the Democratic candidate.

66

Karl, I would assume that the theory of Clinton's speech causing this swing would only be with likely voters as they are the one's paying attention.  Thus Rassmusen's polling should be the one reflecting any big swing based on this theory.  It did not.

67 Sorry, JackStraw.
Notice that I'm not taking the money for myself.
Therefore, you know I'm not cheating you.

68 Nom de Blog, you said this:

The weighting of these polls (internals) has likely changed to manufacture a "bounce".

Your strong suggestion is the weighting internals changed, overlooking the very real world change on Wednesday and the logical explanation of why Obama would pick up disaffected Democrats support.

You followed the above sentence with this one:

Pollsters need bounces or nobody pays them any mind.

Therefore implying it was intentional on behalf of Gallup, that they did it for their own reasons.

Taken together, this would add up to a fraud on the public. So I was paraphrasing you and pointing out how absurd your supposition is.

Likewise, you never said, "I have a conspiracy theory," but Karl interpreted your words to be one.

69 Maetenloch,

I, for one, am not suggesting collusion. With whom would Gallup collude?

I am saying their product is more valuable if they show some "news worthy" bounce.
And I'm saying Gallup knows that to be true.
As a capitalist, I don't allege anything other that people responding to the profit motive.

70 Karl lack of interpretive skills are not my problem, Christoph.
And your use of fraud, which has a particular meaning was wrong.

71

polynikes,

I haven't argued that it was "Clinton's speech" causing the big swing, per se (I'm assuming you are referring to HRC here, as Bill's would not have showed up in Gallup's data yet, afaik).  Moreover, "likely voter" are arguably less likely to be swayed by the DNCC, as they are more likely toi have been payinga ttention long before the convention.  The swing is more likely to come from less likely voters just starting to pay attention now -- which is more likely to show in the Gallup tracking poll of Registered Voters.

72 Nom de Blog, Gallup's product is more valuable if it's accurate.

73 "Your strong suggestion is the weighting internals changed, overlooking the very real world change on Wednesday and the logical explanation of why Obama would pick up disaffected Democrats support." -- Christoph @ 68

Gallup's possible internal thinking:
Obama has gained the PUMA vote. We know this to be true because Bill and Hill support Obama unambiguously.
Therefore, we know that identification with the Obama candidacy has gone up.
Therefore, we should weight our models accordingly.
Hey, look!
A bounce.

There's no "fraud" there. But Gallup suddenly has a more valuable product.

74

Maetenloch,

Don't confuse Nom witth Occam's razor; he's busy imputing dishonesty to a company which relies on its cred to make its money.  They would risk it all to manufacture a bounce for some short-term gain.

75 "And your use of fraud, which has a particular meaning was wrong."

Nom de Blog, you used the word "changed to manufacture a bounce," and then you gave a motive why pollsters need bounces.

How specifically would intentionally manufacturing a bounce to achieve Gallup's selfish motive not be fraud?
A deception deliberately practiced in order to secure unfair or unlawful gain. A piece of trickery; a trick. One that defrauds; a cheat. One who assumes a false pose; an impostor.

76 But Karl, Nom de Blog isn't saying Gallup is fraudulent, merely that Gallup intentionally changes the weightings of its polls to manufacture a more financially profitable result.

77 Karl,
Where's the dishonesty in the narrative I propose in #73?

Christoph,
Where's the fraud?

78 Christoph,
Are you saying making money is "unfair" because I know it's not "unlawful" (a word you oddly chose not to highlight)?

79

Karl I believe you have it backasswards but you'd have to believe that way in order to explain away why Ramussen's poll shows no real bounce.  The difference between the two cannot be explained away because of likely and registered voters.

80 I'm saying that when you're selling polls to major media organizations and political organizations, warranting to them you will try to make them as accurate as possible, and you then manufacture a result to make your polls more interesting, then, yes, that is fraud insofar as it's trickery and deceipt and an unfair way of making money.

It would probably even be illegal, but my goal with the highlighting was just to show that the word fraud is clearly justified considering what you're heavily implying.

Without proof.

81 You guys are all Teh Ghey for arguing like this about this.

But Christoph is more Teh Ghey.

82 Christoph,
You know nothing of the law of fraud.
Stick to other topics about which you know very little.

83

Nom,

The imputation of dishonesty comes from the suggestion that Gallup would weight the sample in the manner you suggested.  Apparently, you have no clue that Gallup is on the record against weighting samples even by Party ID.  Thus, under your scenario, Gallup is saying one thing, but doing another -- which would be dishonest.

84 "Without proof (sic)."  -- Christoph @ 80

And you have proof that Gallup always plays it straight?
Please, offer said proof.

Would you have me sue and go through the discovery process to uncover all the internal memos sent to change, ever so slightly, the methodology before I can make the case that 10 straight observations in the same direction indicate conscious decisions to increase the value of Gallup's product?
Feh!

85 And since they're selling this for financial gain, and are even a publicly traded company and that forms part of their public statements, this would be fraudulent on several levels, moral and perhaps legally.

86

polinikes,

"The difference between the two cannot be explained away because of likely and registered voters."

Really, why not -- esp. after you account for the MoE of each poll?  Gallup could have as small as a 4-pt gap, and the Rasmussen numbers could be as large as 4 pts.  And that's just the MoE factor.  Of course, there are other factors possible -- Rasmussen is a fully automated poll, question wording, etc., that may also be minor factors.

87 Nom de Blog, fuck dude, maybe you're just stupid.

You're the one making the extraordinary claim. Therefore, you're the one who should provide proof if any's to be offered.

No, I can't prove Gallup has always under every circumstances never been dishonest under any circumstances in its history. You'd be asking me to prove a negative. That would be impossible.

What I can say is when Hillary has 18,000,000 voters who voted for her in the primaries, voters who many polls say a large and growing number of whom weren't supporting Barack Obama and were instead supporting her, and when she gives a speech at the Democratic Convention supporting Obama and asking her supporters to support Obama...

... I'm not especially surprised when Obama gets a bounce the next day.

88 I don't think this has been a weak convention at all.

It has been very powerful in rallying the Democrat base of leftist extremists.
They are excited and motivated.

The problem is it has done nothing to draw in conservative democrats and independents. 

89

Nom throws out that whole presumption of innocence thing, based on that impressive sample size of n=10.

Heaven forfend someone should ask for evidence, let alone proof by some particular standard.

90 "But it is our assumption that the party identification measure also has the characteristics of short-term attitudes for a limited proportion of the population." ...
"We are willing to do weighting if we know -- based on comparisons to official statistics -- that the sample is not perfectly representative of the larger population to which we want to generalize." ...
"The distribution of party identification may vary in the total population from day to day, week to week."

All of those quotes are from your Gallup article, Karl
And they do nothing but bolster my scenario above.
Thanks for offering proof that my argument is sound.

91

Chris,

I know your meds are wearing off, read post #27 where you say I talked about a conspiracy.  So, asshole, when did I do that?

An apology is what you need to send, and when you do let us know how many Stat course you took in Jr High.

I figure your 19, am I right?

When you are wrong, admit it.

92 kempermanx, read post 33.

93 What the hell is there so much discussion about this for?  Its simple:

For years, polls have shown results that trend left of what the elections actually show. 

Reasonable explanation: the polls are run by leftists for the most part.

This is about as controversial a statement as saying that the MSM trends left. 

So stop trolling.

94

Nom,

You are cherry-picking from a statement explaining why Gallup does not wight its samples by Party ID. 

But if we're going to do that, this quote is the thesis: "The decision to weight a national sample to a population parameter of party identification (PID) assumes that there is a reliable measure of PID to which the sample can be weighted. There are, however, no such measures at this point in time."

Your theory supposes there is some official measurement of PUMAs that Gallup looked at, which does not exist.

95

Chris,

Accepted, sorry did not see it. 

You still are barking up the wrong tree here.  It matters not.  Let up pray we are behind after their convention, we always win when we are.

Morons for MILF's

96 Maybe Gallup just had young male telemarketers flirt with some of the respondents, Karl, and make a snap judgment:
Milf?
Not Milf?

If they figure it's a milf, they might think female voter therefore more likely Democrat, and older, therefore Hillary supporter.

Ergo, that's how they swung the sample toward Obama today to make it more interesting and get capitalistic gain.

Non-fraudulently, of course!

97 No worries, kempermanx. When you read 48 you'll see I agree with most of your opinions.

98

dan-O,

"For years, polls have shown results that trend left of what the elections actually show. 

Reasonable explanation: the polls are run by leftists for the most part.

This is about as controversial a statement as saying that the MSM trends left. 

So stop trolling."

More reasonable explanation: Polls of registered voters skew more left than polls of likely voters.  As more pollsters started switching to LVs earlier in each cycle, the polls have become less bouncy overall.

However, as I note in #86, the "bounce" is probably not all that in the first place.

99 "We are willing to do weighting..." -- Gallup article Karl linked
"You are cherry-picking from a statement explaining why Gallup does not wight (sic) its samples by Party ID. " -- Karl

Uh, you might want to ask the Gallup organization why they say the exact opposite of you.

100 "However, as I note in #86, the "bounce" is probably not all that in the first place." -- Karl

I think you disagreed with me above when I said all the day-to-day changes were with the MoE.
Huh, well perhaps if I said it wasn't "all that" you'd be more likely to agree.
Now I see the error of my ways.
Fucking ridiculous.

101

Nom,

Learn to read, really.  Anyone can read at the link and figure out that you are misreading Gallup's position.

Or stick with your baseless theory and a sample set on n=10. Whatever.

102

I definitely do not want to get in the middle of this beautiful flamewar, but I would like to offer kudos to Nom for pointing out the lack of a Democrat majority since 1964.

I had to go look up 1996 and, sure enough, Clinton never got a majority. In addition, Carter only won by 50.1, or so, despite being "way ahead" early in the election.

Those numbers don't look good for Obama. McCain must be the luckiest son of a bitch I've ever seen.

103 As to the MoE argument for the slow among us (READ: Karl and Christoph)...

If Obama had been at 46% for the entire 3 week period reported in the linked article above, then we could test the null hypothesis that Obama has polled at 46% each time.

His low: 44%.
His high: 48%.

I like my chances of the last three weeks being nothing but statistical noise.

104 And McCain has polled from 46 to 42%.

Therefore, I think it's highly defensible that McCain has polled at 44% and the rest is statistical noise.

Result: Obama is ahead by 2% in the polling.
And 10 out of 10 have been wrong in the same direction.
So, eh... I'm still feeling really fucking solid about the original point.

105 "So, eh... I'm still feeling really fucking solid about the original point."

But not about insinuating Gallop manufactured a bounce today for their capitalistic financial gain?

106 Uh, yeah, Christoph. I'm still pretty solid on that point too.
The Gallup organization didn't have to do much to generate a 4% bounce and they get headline.
All they had to do was change an assumption or two.
And now they're in the news with a poll.

(I'm just bringing it down for the sake of ease of reference.)

107 "...they get headline."

I'm funny and perverted.

108 They are exactly where they were 15 days before. Exactly: 48/42.

You don't think there's a chance that Obama's just picking up a few points in his convention as the Dem party comes together a bit more than it was?

109 Christoph,
Now you're just being a fucking idiot.
They're not at 48-42.
They're at 46-42 and have been for three weeks.
Everything else is noise.
And the noise, miraculously, benefits Gallup.
10 for 10 in one direction begins to look like a trend.

110 Damn it.
I meant to say they're at 46-44.
I don't know why I typed 42.

They've been at 46-44 for three weeks.
The rest is noise.

111

Nom,

I would actually agree as to the overall trend numbers, but if it's statistical noise, that's a much more reasonable explanation than Gallup dishonestly reweighting samples.  And if you bothered to read carefully, I wrote that there is a "bounce" outside the MoE in Gallup's samples.  I wasn't referring to the longer-term trend in that entry, just observing that JM and BO were within the MoE for the past week or so, and that the gap now is outside that MoE.

112 Well then Karl, I guess you fucked up when you tried to argue MoE with me.
Since you're now admitting that you argued with something I didn't write, I accept your apology.
Fucking douche nozzle.

And since the errors always go in one direction, I'm sure it's just too small a sample to matter.
I'll see you in 2084 and we can discuss how you were really fucking wrong back in 2008.
Please keep the bookmark.
Shit heel.

113

To sum up:

I pointed out that Nom's claim that Gallup was manipulating its samples was lacking in proof or a foundation in statistics.  And that his claim of statistical noise, if true, would explain things better than his crackpot fantasy.  And he's hoping that people won't notice that Gallup has had JM and BO within Moe for more than a week, but do not today.

Which makes me a douche nozzle and a shit heel.

If I was Nom, I guess I would feel the same way.

114 "If I was (sic) Nom..." Karl
Now there is a young man with ambition...

Seriously, Karl, you conceded every point I've made so far after making all sorts of bull shit claims about what you thought I was saying. And now you're stuck wishing you were me? I'm truly fucking honored. No, really and for true, I'm honored.

Oddly, Gallup shows a McCain gain just before the Democrat convention. And then it shows the "expected" bounce for Obama. And that bounce happened like fucking clockwork. And that clockwork has happened each of the last 11 times (10 have been completed.) while the Republican outperformed the poll every time and won 7 out of 10 elections.

But this time it's for reals.
Have you had a chance to stop by the fucking ATM yet?

115 The Bradley Effect.

I call it, I Don't Want My Car Keyed Effect.

116

Chris.

Morons for MILF's

That is code for nominate Sarah Palin.  Read all of Ace's shit, I've only written this a hundred times!

Kemp

117 wedding dresses wedding gowns bridal gowns lace front wigs wedding invitations bridal shower invitations baby shower invitations custom wedding invitation diablo 2 cd key

118 YouTube Video Converter helps you download YouTube video to your computer and convert YouTube Video to MP4, 3GP and other videos, such as convert YouTube to avi, DivX, XviD, rm, rmvb, MOV, MPEG, MPEG4 AVC, H.264 and WMV, VOB; VCD, SVCD and DVD Format .mpg; And the independent player of the YouTube Converter will amuse you by playing your favorite videos.M2TS Converter is a powerful m2ts conversion software that can convert M2TS files to other video and audio formats. Convert m2ts, ts, trp hd video files is so fast and perfect.For e.g. With the M2TS Converter you can convert m2ts to AVI, convert M2ts to MPG, convert m2ts to iPod, convert m2ts to WAV, etc. The software support a series of output formats like, WMV, AVI, ASF, MPEG, MP4, MOV, 3GP, VOB, FLV, WMA, MP3, AAC, WAV, etc. It can aslo supprt convert m2ts files to iPod, iPhone, Apple TV, Zune, PSP, PS3, Creative Zen, Archos, MP3, and other Video and audio players. The Transfer iPod to Computer is premium iPod transfer software for iPod music, video and photo transferring. It can transfer iPod to computer and backup iPod to computer quickly and smoothly. It's also a computer to iPod transfer tool which can transfer iPod music, videos and photos from computer to iPod.iPod to Mac Transferis perfect and powerful iPod transfer for Mac software. It offers iPod to Mac transfer of music, photo, video, movie, podcast and TV program. Moreover, iPod to iTunes, and Mac to iPod transfer are also supported.DVD to iPod for Mac is specially designed for Mac Intel and Mac Power PC users to convert DVD movies to iPod video format on Mac OS X (including Mac OS X 10.5 Leopard). Supports playback on iPod touch, iPod classic, iPod nano and other portable video players.

119 DRM Converter, DRM Removal, Protected Music Converter, DRM Video Converter, DRM WMA converter, DRM WMV converter, DRM ASF converter, M4A converter, M4B converter, M4P converter, M4V converter

120 iTunes Converter, Protected Music Converter,Protected Video Converter,Protected M4A Converter, Protected M4B Converter, Protected M4P Converter, Protected M4V Converter, Protected MP4 Converter, Napster Converter, Zune Music Converter, Yahoo Music Converter, Protected WMA Converter, Protected WMV Converter, Protected AAC Converter, RAX Converter

121 AVI to DVD ,Mp4 to DVD,FLV to DVD,MOV to DVD,VOB to DVD,MPG to DVD,MKV to DVD

122 MKV Converter For Mac ,MKV to VOB For Mac ,MKV to MP4 For Mac ,MKV to WMV For Mac ,MKV to MPEG For Mac ,MKV to DivX For Mac ,MKV to FLV For Mac ,MKV to iPod For Mac ,MKV to iPhone For Mac ,MKV to Zune For Mac ,MKV to AVI For Mac ,MKV to PSP For Mac ,MKV to MP3 For Mac ,MKV to 3GP For Mac,MKV to ASF For Mac

123

www.ircask.com ircask.com sohbet sitesi üzerinden sohbet ve aşk odalarında sohbet edilebilir. ircask sohbet siteleri blog irc forum sohbet ekle islami sohbet galeri msn mp3 güzel sözler chat directory

124 Air force 1 shoes
Nike air force one
Nike air force
Air force 1 mens
Air force one
UGG Boots
UGG Australia
UGG Shoes
UGG Sale
Lebron Shoes
Cheap UGG
UGG Boots
UGG Shoes
Discount UGG
UGG Australia
UGG Sale
UGG Online

125 discounthair straighteners
ghd hair
chi hair straightener
ghd straighteners
chi straighteners
chi hair
chi hair iron
hair straighteners ghd
ghd hair straighteners
free shipping!

126 Discount Boots
Womens UGG
UGG Shoes
UGG Australia
UGG UK
UGG USA
discount ugg boots
cheap ugg boots

127 buy Discount Boots
Womens UGG
UGG Shoes
UGG Australia
UGG UK
UGG USA
discount ugg boots
cheap ugg boots
air force 1 shoes
nike air force
air force 1 mens
air force 1 boots
air force onefrom usa

128 discount Discount Boots
Womens UGG
UGG Shoes
UGG Australia
UGG UK
UGG USA
discount ugg boots
cheap ugg boots
air force 1 shoes
Nike air force
air force 1 mens
air force 1 boots
air force onefree shipping

129 Come to our website for shopping with high quality and best service! There have new products,MBT M. Walk
and MBT Lami,Welcome everybody to buy it.

130 Batiencela, a Rolex stevedore at Rolex Watches the Navotas fish port, made Replica Watches the tag heuer revelation during Tuesday¡¯s Omega Watches inquest before Makati Assistant Prosecutor Cartier Watches Lody Tanchoco.Batiencela and two other arrested gang Breitling Watches mates, Dennis Serquina and Jose Rodriquito, were charged with Chanel Watches robbery with homicide for the brazen midday assault on the audemars piguet watch store. The robbers took 64 Tudor armani watches watches worth P6.17 million when they fled after a three-minute fire fight patek philippe with two Taguig policemen who happened to be in the iwc watches area. One of the robbers, who was carrying the more expensive Rolex Replica Rolex watches, died mont blanc watches in the firefight.
j7

131 Buy discount ugg shoes
cheap ugg shoes
cheap ugg
ugg rainier
buy ugg
ugg usa
discount ugg boots
ugg 5825
ugg shoes sale
ugg sundance
ugg shoes
cheap ugg boots
ugg 5815
ugg sale
ugg uk
ugg knightsbridge
womens ugg boots
ugg boots
ugg australia
ugg 5819free shipping

132 buy [url=http://www.reviewups.com/]discount ugg shoes[/url]
[url=http://www.reviewups.com/]cheap ugg shoes[/url]
[url=http://www.reviewups.com/]cheap ugg[/url]
[url=http://www.reviewups.com/ugg-coquette-casual-shoes-c-22]ugg rainier[/url]
[url=http://www.reviewups.com/]buy ugg[/url]
[url=http://www.reviewups.com/]ugg usa[/url]
[url=http://www.reviewups.com/]discount ugg boots[/url]
[url=http://www.reviewups.com/ugg-classic-short-c-1]ugg 5825[/url]
[url=http://www.reviewups.com/]ugg shoes sale[/url]
[url=http://www.reviewups.com/ugg-sundance-c-5]ugg sundance[/url]
[url=http://www.reviewups.com/]ugg shoes[/url]
[url=http://www.reviewups.com/]cheap ugg boots[/url]
[url=http://www.reviewups.com/ugg-classic-tall-c-3]ugg 5815[/url]
[url=http://www.reviewups.com/]ugg sale[/url]
[url=http://www.reviewups.com/]ugg uk[/url]
[url=http://www.reviewups.com/ugg-knightsbridge-c-27]ugg knightsbridge[/url]
[url=http://www.reviewups.com/]womens ugg boots[/url]
[url=http://www.reviewups.com/]ugg boots[/url]
[url=http://www.reviewups.com/]ugg australia[/url]
[url=http://www.reviewups.com/ugg-classic-cardy-c-2]ugg 5819[/url]from usa

133 Replica Breitling Watches | soccer jersey supplier | Replica Breitling Watches | Replica Omega | Replica Breitling Watches | Longines Watches Replica | Roger Dubuis Watch | Breguet Replica Watches | Replica Bvlgari | Cartier Watch Replica | Maurice Lacroix Watch Replica | Hermes Watch Replica | Bulova Replica Watches | Fake Gucci Watches | Louis Vuitton Watch

134 Fake Watches | Replica Watches | Fake Watches | replica soccer jersey | Replica Watches | Fake Raymond Weil | Ebel Watch | Zenith Fake Watches | Bvlgari Replica Watches | Baume & Mercier Watches | Roger Dubuis Fake Watches | Oris Replica Watch | Oris Watches Replica | Gerald Genta Watches | Ulysse Nardin Replica Watches

135 Fake Cheap Rolex | replica soccer jersey | Replica Rolex | Replica Watches | Replica Watches | Replica Audemars Piguet Watches | Replica Jaeger LeCoultre | Glashutte Replica Watches | Chopard Replica Watches | Replica Alain Silberstein | Panerai Replica Watch | Piaget Watches Replica | Maurice Lacroix Replica Watch | Ulysse Nardin Fake Watches | Wholesale Raymond Watch

136 Replica Rolex | Replica Cartier | Replica watches | Replica Rolex | Fake Watches | Tag Heuer Wristwatch | Titoni Watches Replica | Replica Maurice Lacroix | Christian Dior Replica | Frank Muller Replica Watch | Titoni Replica Watches | Oris Watch Replica | Panerai Watch | U-boat Replica Watches | watches brand