a "game" about the 2004 election

[OnTheOtherHand]

Officially, Al Gore received 51.004 million votes in the 2000 presidential election; George W. Bush received 50.460 million; and other candidates received 3.953 million votes (per David Leip's Atlas of U.S. Presidential Elections).

In the run-up to the 2004 election, it was widely expected that new voters would favor John Kerry, an expectation that seems to be borne out in the exit poll results for 2000 non-voters. (The 2000 non-voters would include both literal new voters -- people who have never voted before -- and people who have voted before but did not in 2000.) Thus, for Bush to defeat Kerry in the 2004 popular vote -- never mind to defeat him by a bit over 3 million votes, as in the official returns -- he would have to overcome Kerry's presumptive lead among returning voters plus any additional Kerry edge from the 2000 non-voters.

This is a problem because the exit polls seem to say that Kerry captured about the same proportion of Bush2K votes as Bush did Gore2K votes, or even higher -- and that Kerry won the DidNotVote2Ks by a solid margin (and the Other2Ks by a huge margin). The weighted exit poll results render Bush the winner only by stipulating that Bush2K voters outnumbered Gore2K voters in the 2004 electorate by a proportion of 43%/37%. As far as I know, we all agree that this did not happen. We disagree about whether massive fraud is the only plausible or the most likely explanation of these and other data. Anaxarchos has proposed a "game" to explore alternative scenarios, of which more below.

some assumptions

TIA reasonably proposes as an analytical starting point the maximum numbers of 2000 voters who possibly could have voted in 2004. He offers a mortality estimate of 0.87%/year, or 3.48% over four years. Absent strong evidence of differential mortality (older respondents tend to skew Democratic, but Anax has some counterpoint that I have forgotten but am sure is reasonable), we apply this attrition rate to the three groups of 2000 voters. (We also discount felon or "felon" purges and the like.) Thus we have approximately

49.228 million Gore2K voters
48.704 million Bush2K voters
3.815 million Other2K voters

The official 2004 returns show 122.293 million total voters (62.040 million for Bush, 59.028 million for Kerry, 1.225 million for others).

Note that if _all_ surviving voters from 2000 turned out, we would need 20.545 million new voters to account for the total official 2004 turnout. It is ridiculous to suppose that 100% of surviving voters turned out, and I am willing to entertain a lower figure. I think we have agreed to assume that Gore2K and Bush2K voters turned out at the same percentage in 2004, whatever that percentage may be.

Please advise of any errors in or even caveats about these assumptions. (It is acceptable to explore alternative assumptions as we go, although we all agree to reject the assumption that Gore2K voters are immortal and/or undead.)

the proposed rules

1) The game is a game of "Best Fit" but the boundaries are subjective. Someone (maybe Descartes?) once said that "mathematics has a soul but it has no heart" (i.e. "conscience"), meaning any value is acceptable in calculation but not necessarily in life. We use a standard of "plausible", stretch our self-discipline to the max, and see where we end up.

2) The game gets played right here and anyone can play. Despite the "crazy" tone we sometimes adopt, there are close to 65000 experts here on what is "plausible" in politics. A few are truly nuts so we identify them early and make them the board of directors.

3) We keep it really simple.

4) This game is only indirectly about Exit Polls and Polling in general. We are actually trying to reconstruct the election of GW Bush, in a "plausible" way.

5) This game is deadly serious for anax and other players, but turning it into a game avoids what is "conceivable" for some, what is "proof", what is "acceptable", and all the rest... the only standards for this game are "reason" and "virtue".

6) First one who says "the real vote count has to be right because it is the real vote count", loses.

Note: anax proposed to "get rid of rule #7," but I'm not sure I got rid of the right one. The original rule 7 was, "I (anax) further agree not to comment on your (OTOH's) serve until we get to the 'double-dare-you' stage." Since I'm not even sure what that means, it wouldn't bind me, and I could not enforce it, I unilaterally waive it. But if the point was to eschew trash talk at least until further notice, then I endorse that.

[Melissa G]

1. there are also no Bush 2k Zombies allowed just to be fair n/t

[Time for change]

2. Just a thought, but I'm probably breaking a rule, because I don't

understand them very well.

One of your assumptions is that Gore and Bush voters turned out in approximately the same percentages. Yet, in order to fit with the official returns, the "weighted" exit polls (isn't "adjusted" a more appropriate term for this?) had to say that Bush voters outnumbered Gore voters by 43%-37%. Therefore, in order to explain a Bush victory, one of our primary assumptions was already broken.

To fix this, you must say that in 2004 we had 40% Gore 2K voters and 40% Bush 2K voters. This would raise Kerry's percent of the 2004 vote by about 5.4%, since Bush 2K voters voted about 90% for Bush, and Gore 2K voters voted about 90% for Kerry. Since Kerry lost officially in the 2004 election by 2.5%, that would mean that he actually won the 2004 election by 2.9% (I hope I'm not breaking rule # 6). Coincidentally (NOT) that is almost exactly what the unadjusted exit polls had Kerry winning by.

Well, I definitely kept to rule # 3, but I'm not so sure about the other ones.

[anaxarchos]

3. TfC, the game began, specifically from your objection.

... so it is entirely in order. Here is the link:

http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x379573#389550

starting from my response to OTOH's post #111. I argued that what you said was the only way a Bush "victory" was "plausible".

The purpose of the game is to prove otherwise.

Commentary is entirely welcome before we start.

After that, OTOH serves...

[Time for change]

5. So did I win??

[anaxarchos]

6. I thought so.... n/t

[Time for change]

7. Wait a minute

You're saying the purpose of the game is to prove a plausible Bush victory? OTOH said that the purpose of the game was to propose the best fit. If the purpose of the game is to prove a plausible Bush victory, it seems to me that there wouldn't be too much motivation in this forum.

[anaxarchos]

8. Read the thread...

The argument is that without the 37/43 split, one "plausible" way Bush wins is through fraud. Any others? We're getting there through negation. Sort of Aristolean...

BTW, I wouldn't worry about "people on this forum" (at least "temporarily" on this forum). If this gets hot enough and TIA plays, the Republicans will show up on cue. I'm waiting for the "Rocks eat Popcorn" guy....

[Time for change]

10. I read the rules thoroughly

Rule # 1, which notes the objective of the game, mentions plausible and best fit. It says nothing about not allowing fraud.

I want my trophy.

[Melissa G]

11. OK TFC..we all think you won which is why we are here in this forum

But that was way too short of a game.. let's let the folks who think they can explain the inexplicable have a shot.. Pull up a chair and have some popcorn :popcorn:

[Time for change]

12. Ok, I'll get over it.

But tell me this. When trying to come up with the best plausible fit, are we allowed to take into account all the non-exit poll related evidence of fraud? It seems to me that we should be, because that's part of the reality that we're faced with.

But that might lead us back to a fraud conclusion, which is not allowed in this game.

So let's say that someone comes up with a plausible scenario for Bush winning without fraud, but that scenario doesn't take into account the non-exit poll related evidence. What that would do would be to suggest that it is possible that Bush won the popular vote, not that he did win it, or even necessarily that it is likely that he did win it.

Do I have that right? I just wanted to make sure that I understand the purpose of the game, and how it is supposed to be interpreted.

[Melissa G]

13. As far as I understand it you have it right.. No one we know of to date

according to anax has come up with a theory of how Blivet won.
some folks around here are going to give it a shot mathematically..
i have not even seen a non mathematical scenario myself..
see anax post on TIA's thread
http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x390154#390179

[TruthIsAll]

15. Not ONE mathematical scenario, but many theories and hypotheses

based on reluctant Bush voters, the Bush Bandwagon effect, lying Gore voters, forgetful Gore voters, bad weather at the exit polls, exuberant Kerry-biased exit pollsters, inexperienced exit pollsters, exit poll not designed to catch fraud, young exit pollsters, early Democratic voters, late Republican voters, inaccurate exit polls, cluster effect, faulty assumptions, High MoEs, massive Fundamental Christian turnout, Bush the War President, conspiracy fraudsters, sore Kerry losers, rabid spreadsheet-wielding liberal bloggers, Democrats weak on defense, Kerry a lousy campaigner, Repubs united, Democrats divided, Bush strong on moral issues, Bush a religious born-again Christian...

But no mathematical scenarios based on factual data to back it all up.

[tommcintyre]

25. Hey, you forgot early women voters (who skewed for Kerry) ;)

But, just like Mitofsky didn't bother to substantiate rBr (how could he possibly substantiate a lie?); none of the other excuses have ever been substantiated either. But hell, who needs to, when ya got the new "golden rule" working for you (He who has the gold, makes the rules.)

I think I'd rather go watch a "rock eat some popcorn" right now. (I missed that one; but I think I can guess who "elucidated" it. ;) )

[anaxarchos]

35. Thank you, Tom.... I think. n/t

[Land Shark]

28. Of TIA's listed theories and hypotheses used to support election results

my favorite is "exit polls not designed to evidence fraud".

A good test of silliness sometimes is whether or not a criminal defendant could make the same argument:

"Your Honor, those fingerprints are not evidence because in no way was the placing of the hand upon the glass INTENDED to produce evidence!"

In other words, it matters not what the exit pollers INTENDED, it matters only what the evidence is and what qualities it has standing alone.

[OnTheOtherHand]

32. possibly off-topic, but let me respond

I'm not aware of anyone who has argued that the exit polls should be ignored because they weren't designed as evidence.

Some of us argue that the exit polls don't provide substantial evidence of fraud, and go on to argue that this is unsurprising _even if there was fraud_ because the polls were not designed to detect it.

We've just seen a study in San Diego that was designed to detect fraud, and still comes nowhere near proving it.

I don't advocate ignoring the exit polls (I had better not, considering all the time I have spent on them this year).

[Febble]

37. Objection m'lud!

Wrong analogy.

"Your honour, this ruler is not evidence, because it was deliberately constructed of elastic material so that it could be re-calibrated to fit as required"

Exhibit A:

FAQ on E-M site (already posted once on this thread)

How are projections made?

Projections are based on models that use votes from three (3) different sources -- exit poll interviews with voters, vote returns as reported by election officials from the sample precincts, and tabulations of votes by county. The models make estimates from all these vote reports. The models also indicate the likely error in the estimates. The best model estimate may be used to make a projection if it passes a series of tests.

http://www.exit-poll.net/faq.html#a10

(again, my bold)

If you want to run an exit poll as an auditing tool, it needs to be constructed rather differently (and rather more expensively). This one submitted as evidence was designed to enable the networks to "call" the states.

Also from the E-M FAQ:

When will we know who is elected President?
Edison/Mitofsky will add together the electoral votes won by each candidate for states that have completed their voting and closed their polls. When one candidate has a majority, that is, 270 electoral votes, a winner will be announced. The members of NEP will announce the result when their analysts affirm an electoral vote majority. This may be earlier or later than the Edison/Mitofsky announcement to the members.

For a paper addressing questions on how exit polls should be designed in order to form an effective part of an election audit see here:

http://www.electionscience.org/Members/stevenhertzberg/report.2005-05-10.0988159730/report_contents_file/

[tommcintyre]

89. LOL! My fingerprints weren't designed to prove I was there either ;)

"Your Honor, those fingerprints are not evidence because they clearly were "designed" (evolved) to facilitate grip (traction); and NOT to place me at the "scene-of-the-crime."

The argument "exit polls weren't designed to indicate fraud" is equally absurd. Like the unintended (but clear) indentifying-value of the fingerprint, this is a natural (even if unintended by M-E) by-product of ANY exit poll that is useful at all.

As far as the "weighting" (mixing in actual vote results, etc.) - this is equally untenable. This action DOES produces the end result they (networks, etc.) are looking for (exit polls matching the vote totals); but it certainly doesn't improve the real accuracy of the predictive value of the exit polls one iota, does it?

But you will hear these two logically unsupportable arguments put forth repeatedly - along with all the other equally unsupportable arguments that are listed above.

Of course, this is done because the exit poll discrepancy argument IS very important; and the "opposition" has no valid (supportable)arguments against it. So, weak "obfuscated spin" is all they got.

[Febble]

111. Re this:

"As far as the "weighting" (mixing in actual vote results, etc.) - this is equally untenable. This action DOES produces the end result they (networks, etc.) are looking for (exit polls matching the vote totals); but it certainly doesn't improve the real accuracy of the predictive value of the exit polls one iota, does it?"

You are absolutely right, indeed it doesn't, except in the narrow sense that it allows the networks to "call" the state before all the results are in. If you like, it "predicts" the official count; it doesn't predict the real votes cast, if these differ from the count.

But I think we are all agreed on that.

The germane point here, though, is that if you want to conduct an exit poll that will make a prediction that is independent of the official count, one that might be a better yardstick for fraud than the E-M poll, you need to spend a bit more money, and do a more thorough job of the actual polling. Three things would probably make a substantial difference: more training for interviewers; more interviewers per precinct (to avoid coverage gaps and misses); and larger samples. The BYU study had all these things, and got closer than E-M to the counted result in Utah (see links above).

Now, regarding "obfuscating spin": you are of course entitled to speculate why people might argue that the exit poll discrepancy was not necessarily due to fraud. I can only say that their number include people who are on record as supporting the campaign for electoral reform, and who have published work that identifies some of the gross inequities of the US electoral system. Walter Mebane springs to mind. For my own more humble part, you can google my attempts to demonstrate evidence of corruption/disenfranchisment during the 2004 election in Florida, Ohio and New Mexico. While I am not convinced of the validity by my own work on Florida, I am much more convinced by my findings on Ohio and NM, which have been borne out by other researchers.

But for the record: the reason that I investigated the exit poll story was that I wanted to know the answer as to why the poll was so discrepant from the count. I hoped the answer would unambiguously point to fraud. I no longer think it does. And I happen to think that the cause of electoral reform is not advanced by a statistically suspect arguments. If they are used as supporting evidence, fine - though I think it is worth examining the counter-arguments, as any virtual or actual cross-examiner will do exactly that. I think the case for investigating the election stands up without reference to the polls. I think, therefore, that the case should not be presented as though it rests on evidence that can be relatively easily undermined. The exit polls are the weakest strand in the argument, not the strongest, and the strongest are unassailable.

And actually, the findings of ESI on the exit polls in Ohio (that the exit polls do not indicate fraud) are of use in themselves, in that they set parameters on HOW any fraud could have worked. Some kinds of fraud wouldn't show up in the exit polls, as TfC has cogently argued

http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=203&topic_id=380878

(points 2 and 5)

and yet could have swung the election, as BB has also argued on this thread:

http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=390193&mesg_id=390335
.

[tommcintyre]

149. Disqualified: Hyperbole (exaggeration) and false appeals to authorities

Gee, I don't know where to start, so I guess I will just hit on three low points:

Fallacious argument by hyperbole (exaggeration):
"The exit polls are the weakest strand in the argument"??? I'm frankly surprised you would put forth such an obviously weak claim. First, the obvious innate "extremeness" of calling something the "weakest", rather than "weaker": for example, considering the number and diversity of "claimants" of "pro-fraud" arguments put forth, how could you possibly know this? I certainly don't know this - and I doubt anyone here (outside of your small group) would make such a claim/pronouncement.

Secondly, no doubt this "fact", that the "exit polls are the weakest strand", will come as quite a surprise to many experts who have, prior to your profound pronouncement, believed otherwise. Please be sure to inform, among many others, USCV (Dopp, Bainam, et. al.), etc. who are still laboring under the false belief that this "strand" is VERY important argument. I'm sure they all will appreciate your "enlightening" them, so they can stop wasting their time on the "weakest strand".

You use ESI as an "authority" to substantiate your claims?
Ah... there is a... credibility problem with ESI around here (an understatement):
The following post sums this up well:
http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=203&topic_id=368978#369157

"I officially question the motives of ESI and Hertzberg"

there is a pattern:


minimizing the value of exit polling

giving credit to the "reluctant Bush supporter" hypothesis with no scientific explanation or data

saying that people who question the results of the election are on a "witch hunt"

runs the election science institute, but according to his own bio he has no scientific credentials or background. he mysteriously leaves out any description of his schooling, diplomas, credentials.

out of 19 people on the "vote watch team" only two have a "Dr" in front of their name. This is "science?" We're supposed to believe their report over the USCV report that was compiled by a team of virtually all PhD statisticians?

did an "investigation" into the Ohio election and didn't think it was important to look at any of the voting machines, ballots, or investigate any of the allegations of stickers on ballots. When asked specifically about the stickers on ballots, ignored the question.

misrepresented the facts about the random hand counts in the Ohio recount. stated that the directions were not clear on the definition of "random," while Blackwell's own commission condemned the Lucas County Board of Elections for not following the clear directions on randomness and stating emphatically directions were clear. (and they were)

Refusing to even look at much evidence that was presented to him, including the entire Election Incident Reporting System (EIRS), which shows 99% of vote switching went from Kerry to Bush. Says it's not a scientific study. (personal note - I'm trying to figure out what gives him the authority to decide what is scientific and what isn't. Maybe if he told us what kind of education and/or degree(s) he has we could better understand why he can or can't judge what is scientific and what isn't. For some reason, other people who have PhD's are willing to look at the EIRS.)

Says the evidence of machine allocation and long lines at democratic precincts in Ohio is "heresay" and "innuendo" while the raw data CLEARLY shows otherwise.

IS THERE A PATTERN HERE?

BTW: This is from landsharks original thread where he challenges the claim by ESI that exit polls aren't good enough to monitor elections. (Of course you (and ESI) well-know that they have been historically used for just that purpose around the world.)
http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=203&topic_id=368978

From the OP:

snip
"Reminds me of the "exit polls were never intended to detect or prove election fraud" argument. That's just brain-dead, like going into court and saying that your client's fingerprint evidence should be suppressed since it was never INTENDED or DESIGNED to constitute evidence of a crime.

God, what have we done to deserve this level of BS?

Hmm. Come to think of it, these kinds of arguments are the sign of a

DESPERATE

FACT-FREE

CRIMINAL DEFENSE LAWYER.

Perhaps these specious arguments showing up in articles that will be cited against the case for election fraud or irregularities is a good sign?

So, is it also a "good sign" that we are getting such "specious arguments" from you and few others now? Is the facade starting to crumble?

<I have to go now; but I will finish my reply to your post in a new post below later. Part two of my reply to you will discuss your other "false appeal to authority" - to yourself.>

[Febble]

151. It happens to be

my view that the exit polls are weak evidence. It is possible that other evidence is weaker, but none that I have recently seen advanced.

And as you know, the people you mention are aware of my views. I don't suppose they would appreciate hearing them yet again, but they can log into DU if they want to.

I did not "appeal to authority" - I cited a source of evidence. If you do not regard it as good evidence, that is of course, your prerogative. I am all for regarding evidence with skepticism. Like you, I would not accept anything simply because it was said by someone in authority. I do, however, find the ESI finding persuasive, not because it comes from ESI, although its independence is in its favour (I would also regard Fritz Scheuren as a legitimate authority - see a bio here:
http://www.hrdag.org/about/fritz_scheuren.shtml - as his expertise considerable and in a relevant field). But that was not my point. I should have been more specific (I was on other threads): ESI found that Bush's change in vote share (relative to 2000) was not greater where the exit poll "red shift" was greater.

Report here:
http://www.votewatch.us/Members/stevenhertzberg/report.2005-07-19.7420722886/report_contents_file/

Figures here:
http://www.votewatch.us/Members/stevenhertzberg/report.2005-07-19.2452304843/report_contents_file/

This is an important finding, and while it is perfectly compatible with certain types of fraud (as proposed by TfC) it indicates that fraud did not account for the red shift. It may nonetheless have accounted for Bush's win in Ohio.

Disclaimer: this post contains my views only, and should not be regarded as an authoritative source.

[autorank]

152. Febble, your citing ESI (votewatch) to support ESI (votewatch)

This is not allowed. You need to cite someone other than vote watch or the independent opinion of board members listed (isn't OTOH involved with them in some way? serious question). That's like citing the Bible to support the Bible.

I read their report on New Mexico. Sounded like Sunday school just off the green with the Vicar and Dean of the Cathedral, every thing in order. Then I read the Albequerque media and all the repors of shenanigans. What a mess. Are you sure they were in New Mexico and not Six Flags over Texas. How could they miss this. More importantly, how could they not mention it in their report on NM. They don't have to agree but they do have to respond.

When I was at the 1/6/05 Rally/March, I met an activist from NM who was reallyi worked up about what happened there. Of course ESI which has received Mitofsky data was there and dismissed any problems in the area they focused on, Bernalillo County. This is contrast to the reports in the Albequerque Journal and the local CBS affiliate which had stories about "phantom votes" -- machines failing, incorrect choices (when selecting Kerry).

What was ESI/votewatch "watching" while this was going on? Didn't they have access to newspapers?

E.G. From the Albequerque Journal (all quoted on Eriposte:
http://vote2004.eriposte.com/swingstates/newmexico.htm

"A Democratic attorney and an election volunteer for the party said workers disqualified hundreds of provisional ballots cast in Bernalillo County because of names that had a missing middle initial or some other minor discrepancy. They urged commissioners to reconsider the rejection of those ballots, if they could.
The commissioners, however, said they didn't have authority to question the qualification of provisional ballots. They voted 3-0 in favor of certifying the Nov. 2 election results, which were sent to the Secretary of State's Office.

Roughly half of those ballots were disqualified in the canvassing process. The most common reason was that the person wasn't registered to vote anywhere in the county.

But Democratic attorney Jim Noel said his party's observers had seen at least 330 provisional ballots rejected due to minor discrepancies. Those were cases where voters had to show identification but the ID didn't exactly match the name on the voter registration."

http://www.abqjournal.com/elex/257849elex11-13-04.htm (requires registration)

More from the Albuquerque Journal:

"Herrera's staff had been combing through 2,000 "questionable" ballots, which led to the certification of 1,400 of them.

Those that weren't certified bothered her staff. The main reason for disqualifying them, she said, was because an affidavit testifying to the voter's identity, which is supposed to be signed by a presiding judge, was not in the outer of two envelopes that are supposed to be turned in to election workers. That rule was prescribed by New Mexico Secretary of State Rebecca Vigil-Giron.

Today Herrera said a Republican presiding judge in one particular precinct was in charge of several hundred bad ballots. The problem, Herrera said, was that the bad ballots, with affidavits inside, were largely Democrats. The good ones were for Republican voters.

"It made us kind of sick," Herrera said. "It was too obvious." "

Guess ESI/Votewatch missed this data in their final report

More problems reported to the Albuquerque Journal:

Kim Griffith voted on Thursday— over and over and over.
She's among the people in Bernalillo and Sandoval counties who say they have had trouble with early voting equipment. When they have tried to vote for a particular candidate, the touch-screen system has said they voted for somebody else.

It's a problem that can be fixed by the voters themselves— people can alter the selections on their ballots, up to the point when they indicate they are finished and officially cast the ballot.
For Griffith, it took a lot of altering.
-----------------------------
So nothing like the facts. People had problems voting in more places than would considered reasonable and, when they did in Bernilillo County NM, even the (vote)watch dogs didn't spot it.

[Febble]

158. I didn't cite ESI to support ESI

And OTOH is not, to my knowledge, involved in ESI. You'd have to ask him.

I cited ESI's finding that Bush's vote share was not greater where the E-M poll redshift was greatest. That finding is illustrated here:

http://www.votewatch.us/Members/stevenhertzberg/report.2005-07-19.2452304843/report_contents_file/

(Figure 2)

The study was done using the "blurred" dataset released to ESI by E-M and prepared by E-M, which is why Scott Dingman and Warren Mitofsky are credited in the author list (qualified by the word "with"). If further blurred datasets are released (and as I've said, I support the release of further datasets) any studies done on the data will, of course, similarly credit Mitofsky and whoever does the data preparation..

There is no way of using the E-M data and NOT crediting the source of the data. It would be unethical not to, as data itself is only as good as its source. If you distrust the source, you will, distrust the data. That is why it has to be acknowledged.

And of course you are entitled to your views on ESI. However, I was citing a specific finding (though I should have made that clearer). It is an interesting finding. TfC has raised the possibility that it may indicate that the election in Ohio may have been stolen by destruction of votes (regardless of choice) in strongly Democratic precincts, e.g. in Cuyahoga county. There seems to be good evidence for this (and I would argue also that machine rationing in Democratic precincts was another mechanism by which Kerry was deprived of votes). However, neither of these would affect the exit polls, and are thus compatible with the ESI finding.

But it leaves the red-shift in Ohio unexplained - leaving the possibility of polling bias as the explanation for the redshift in Ohio.

As BB has suggested - perhaps there was vote theft AND rBr.

[tommcintyre]

155. Your "Disclaimer" is a good start

I just finished my "part 2 post" below (as promised); and I suggested that you should put a disclaimer in all election-fraud-related posts of yours here.

I suggest you modify it as such:

"Disclaimer: this post contains my views only, and should not be regarded as an authoritative source; and, the fact that I have accepted (and continue to accept) money from Warren Mitofsky should be taken in consideration when evaluating my views.

P.S. You set yourself up as an authority on this issue when you: 1 published you paper, 2 Mitofsky used that paper to support his position at AAPOR, 3 you presented yourself as an "implied" authority here on the board. (As a note: I DO realize there is a difference between being posed as an "expert"; and presenting valid information - the former does not necessarily lead to the latter.) If you choose to relinquish your "authoritative posture" in this forum; I, for one, will accept that.

[Febble]

156. In turn:

1. I posted a paper linked from a blog. It has no more intrinsic authority than anything anyone posts on a blog including posts on DU. It needs to be taken on its merits (or lack or them).

2. What Mitofsky's response to my paper was does not give me any authority, unless you regard it as a positive endorsement (which I somehow doubt you will). His response needs to be considered on its merits (or lack) of them also. I have never claimed Mitofsky's endorsement as validation of my paper. I simply reported that Mitofsky agreed that my measure was better than the WPE. It may, nonetheless, not be the best measure there is. My paper has had no formal peer-review. An evolution of my paper (currently being written by myself, OTOH and Rick Brady) will be submitted to a peer-reviewed journal at some point. Peer-review is good.

3. I have not at any time adopted an "authoritative posture" except in the sense that I do not express opinions I do not hold. To that extent, I regard myself as "an authority" to myself. I expect you respect your own opinions too. If have inadvertently appeared to claim authority (as opposed to citing evidence and reasoned argument, I here make it absolutely clear that I have no intrinsic authority. I am simply a scientist with statistical expertise and a concern for electoral justice in the world's most powerful nation.

I am also happy to repeat the information that Mitofsky engaged me to run some analyses for him using my proposed variable. I have now completed these. You, however, transgress the site rules when you post that I "accept and continue to accept" money from Mitofsky. However, I will make it clear now: I was engaged by Mitofsky to run some analyses on the exit poll data; I quoted an hourly rate for my work that was identical to what I receive for coaching dyslexic students in statistics at my university; I have completed the work but not yet submitted my invoice; when I do, the money will be paid directly into my American niece's college fund. It was about a month's work.

It was not a bribe, nor payment for PR work. I was simply engaged to run analyses. I was engaged because Mitofsky read my paper and agreed that the measure I proposed was a better measure than the one used in the E-M report; he therefore wanted to have the data reanalysed using my measure rather than the WPE. I have no ongoing vested interest in Mitofsky's enterprise.

Yes, people should take this into account when reading my posts. But I expect my arguments to be assessed on their merits, not on any perceived "authority" they may have. I therefore take pains to link to my sources and explain my reasoning. If you have reason to doubt my sources and/or dispute my reasoning, that is fine by me.

And I would also ask people to take into account my work, hosted on the USCV website, investigating voting machine effects in Florida; undervotes in New Mexico and voting machine provision in Franklin County Ohio (links below), if they want a fuller picture of my general stance on the probity of the 2004 election.

http://www.uscountvotes.org/index.php?option=com_content&task=view&id=32&Itemid=43

http://www.uscountvotes.org/index.php?option=com_content&task=view&id=70&Itemid=43

http://www.uscountvotes.org/index.php?option=com_content&task=view&id=65&Itemid=43

I confess unashamedly to having sent these analyses to John Conyers, Cliff Arnebeck, Keith Olbermann, and Daniel Hoffheimer (Kerry-Edwards counsel in Ohio). I am also unashamed of having induced Warren Mitofsky to re-analyse his data in the light of my work.

[tommcintyre]

169. There ya go again

What I have a problem with is your regular distortion of the facts. For example, you quote me as saying: <you> "accept and continue to accept" money from Mitofsky. Well, I immediately searched the thread for the phrase in quotes, since I didn't recall saying that. (People normally add credibility to their writing by putting phrases in quotes, indicating that is PRECISELY/ACCURATELY what was said.) Not surprisingly, your instance of the phrase was the ONLY one that came up - I NEVER SAID IT. Here's what I actually said: "...that you were <and probably still are> working for Warren Mitofsky." Speculative yes, but clearly NOT as you just misrepresented it - that I was implying a statement of known fact to me. Febble, that kind of thing is just sloppy and inaccurate. Frankly, it's one of the things that is part of the pattern that makes me view your "claims" with such skepticism.

Another is (IMO) your inconsistency. You say: "I expect my arguments to be assessed on their merits, not on any perceived "authority" they may have." Yet, in the very next paragraph you say: "I would also ask people to take into account my work..." Well, if that's not an example of trying to garner increased authority for your current claims, I just don't know what is! That is so inconsistent.

"Peer-review is good." Ah, yes it is. But saying your going to try to get it, and actually getting it is two entirely different issues, isn't it? Especially, when there is only one Phd among you; and that one is in a different field than statistics? Could this be another example of attempting to garner unearned authority? Also, one of your "triumvirate" claims he was banned from here for being an "evil republican". From the posts I've seen from him, I expect the reasons were much more "extensive" than just that. The posts are over at Mystery Pollster - a site that is generally VERY unfriendly (critical) to DU. When you've posted over there, did you notice how vitriolic and mean-spirited his posts often are? How he attacked and misrepresented us here at DU (another DUer posted a couple months ago how they generally cut DU down - like at freerepublic?)? I would imagine working with someone like that would feel like "putting your hand in a blender" (an expression I saw over there referring to DU).

Finally, the fact that you didn't disclose you were currently working for Mitofsky, while you were peddling your paper around here, was less troubling (IMO) than the fact that you chose to hide that information.

[Febble]

173. Enough already

I apologise for misquoting you. Here are your exact words (copy and paste): "I have accepted (and continue to accept)". Here's the link: http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=390193&mesg_id=390777
When I did it before I forgot the brackets and changed a tense. I will make sure I use the copy and paste facility in future.

Second point: I would expect people to assess my work on its merits. You asked me to recommend that people take into account my work for Mitofsky. I merely added that they might also take into account my investigative work into the electoral irregularities of 2004. Neither gives my work authority. But, as you point out (correctly), if people want to know where someone is coming from, it can be helpful to have some information about what they've done previously. Is there a problem in presenting a rather fuller picture?

Third: yes, getting peer-review is demanding. I merely said that our paper would be submitted for peer-review. I am clearly not claiming that status for the paper at present. It is very much a work-in-progress. And yes, of its three authors, at present only one has a PhD. That PhD however is in political science. It is a paper in the field of political science. No problem there. Nor is fact that neither Rick nor I have a PhD a problem. Fortunately peer-review does not take into account qualifications - the work has to stand on its merits. Some journals even review blind. I review for several journals which conduct double-blind reviews. However, given the time it takes to get papers reviewed, it is possible that even I will have a PhD by then. Regarding Rick Brady; yes, I believe he was banned from DU. He is, after all, a Republican, so I'm not surprised he was banned from a Democratic forum (not sure that Republicans have the monopoly on vitriolic postings, however). However, I have never judged anyone's virtue, skill, or logic on the basis of their political views. Mine are probably far closer to yours, tomm, than they are to Ricks's. But that doesn't stop me disagreeing with you, nor from agreeing with Rick, as to what the exit polls are telling us. He also happens to be one of the most intelligent and hard-working people I have ever collaborated with. I also value the very fact that we have quite different political perspectives.

[tommcintyre]

182. Wow! You really whacked that one out of context, didn't you?

You wrote:
"You, however, transgress the site rules when you post that I "accept and continue to accept" money from Mitofsky."
Then modifyed it to: "I have accepted (and continue to accept)".

Here's the context you actually lifted it from (a proposed disclaimer I wrote for you to use):
"I suggest you modify it as such:

Disclaimer: this post contains my views only, and should not be regarded as an authoritative source; and, the fact that I have accepted (and continue to accept) money from Warren Mitofsky should be taken in consideration when evaluating my views."

This, of course. was written (proposed to you) ASSUMING it was true that you were still receiving money from Mitofsky. Read in context, it is pretty obvious that you would modify/correct the "suggestion" as necessary (isn't that what people do with "suggested" (proposed) ideas?).

As far a Rick Brady goes: I said, "The posts are over at Mystery Pollster - a site that is generally VERY unfriendly (critical) to DU. When you've posted over there, did you notice how vitriolic and mean-spirited his <Rick's> posts often are? How he attacked and misrepresented us here at DU..." So, it's not about him being a republican. I have said before, we will have to work with repubs to get meaningful election reform. Here's the link to the thread I referred to. Anyone can go there and judge Rick's questionable demeanor for themselves (Question: Is this a man you could ever believe, or trust?):
http://www.mysterypollster.com/main/2005/06/uscv_vs_uscv.html

[Febble]

183. Oh, drop it tomm

Your post conveyed information about me. You did not link to the source of your information. Yes, I could have corrected it, and yes I did. But I do not think that the DU rules permit personal information to be posted even if correction is invited. There are reasons for these rules. Computer security is one of them.

And what's sauce for the goose is sauce for the gander. When you "suggest" I word a "disclaimer" in the way you did (see above for context) you might like put IT in context. Put the way you did, it suggests that I am (subject to correction) accepting a retainer from Mitofsky for expressing these views on DU. Which is misleading bunkum. I ran some analyses for him at an hourly rate, and the only sense in which that ties me is that the analyses must remain confidential.

As you say, context matters.

I'm glad your objection to Rick is not that he is a Republican. I agree with you that a bipartisan approach to electoral reform is important - it's one of the reasons I am glad to work with Rick. As for vitriolic posts, as I said, I don't think Republicans, or even Rick, has a monopoly on those. TIA is maligning him for his religious views downthread, which I think is not just vitriolic but unacceptable.

Let's call a truce tomm. I've put the disclaimer in my sig. Now you stop these posts impugning my motives. If others want draw their own inferences from my disclaimer they can do so.

[tommcintyre]

186. OK, but that was NOT personal information. Dropped n/t

[Name removed]

153. Deleted message

Message removed by moderator. Click here to review the message board rules.

[Febble]

157. Done

Now, please stop posting personal information (especially speculative information) about other posters on DU. It's against the rules.

I don't have the benefit of anonymity on this forum. I don't mind that, as I have nothing to hide. But that doesn't entitle other posters to post personal information even if publicly available without their consent. From the DU rules:

Do not post personal information about any other person, even if that information is publicly available. Do not "dig up" information about any other visitor to this website. You may post the public contact information for public figures, but you may not post anyone's private information without their consent.

[tommcintyre]

161. Wow! You really like to stretch things, don't you?

These are posts you voluntarily made yourself at DU. You were VERY proud to share a lot of information when you got here. It seemed like you were "basking in the line-light", and wanted everyone to know who you are, and that you wrote that paper (don't blame me if you're not so proud of it anymore).

Anyone who has donated to DU can simply punch in your screen name, mine, or anyone else's in the search function to find what we have VOLUNTARILY posted. Febble, it's YOU who finally admitted you work for Mitofsky, etc. (remember YOUR "report and apology thread"? So, claiming it is "speculative information" is... well... as someone once said,"there ya go again").

In addition, you have voluntarily revealed a lot more info about yourself in past threads (no doubt, in an attempt to gain credibility).

If you now regret some of the information you posted in previous threads here, maybe you can search and find these threads and ask the administrators to remove them? (You can try to re-write history, so-to-speak.)

Also, you can re-sign up (start over) with a new screen name, since even if you succeed in getting the information removed that you work for Mitofsky, etc., I'm sure there are literally thousands of DUers that will remember you publicly admitting to this, and other things (which I'm not specifying now, since it's not germaine to our current conversation). (You probably can "start-over" anytime you feel like it; but I doubt you would want to, since you would no longer be able to attempt to "trade on your name" - for whatever that's worth.)

Meanwhile (until you do this), IMO, I think you are once again trying to control information (spin) to your preferences. Is THAT very ethical?

So, Febble, let me make this perfectly clear to you. ANYTHING that you have posted here (providing it is still available), is "fair game" to be referred to (obviously not for malicious purposes, of course). By posting this information, you have implied (implied consent) that it is fully available for "public consumption". But you really knew this, didn't you?

[Febble]

163. Yes, anyone can search my posts

and they are welcome to do so. But you went well beyond what I have publicly posted to speculate that I have received and continue to receive money from Mitofsky, but provided neither link nor context (i.e. you omitted to say why I might have received money from Mitofsky). If you want to cite what I have posted on DU, please include links.

You also continue to speculate as to my motives (that I have posted information "in an attempt to gain credibility"), and your speculations happen to be false. This latest is somewhat ironic seeing as you yourself were the one who pointed out to me (rightly) the onus of full disclosure. To avoid future problems the information is now in my sig, which will save people the trouble of searching for it. I will, of course, not sign up with a new screen name. I would regard that as dishonest.

However, I certainly do not wish to "trade" on my name. Frankly, seeing as it is largely mud around here, I don't think I'd do much trade. But the reason I came to DU was that my work (in my own name) was cited. I'm not going to pretend to be anyone else. Actually, I did a search for my name a while back and found quite a few favourable references on DU to my work on Ohio, NM etc, from long before I started posting. The irony.

But the rules state that you may not post personal information about other posters without permission, even if it is publicly available. And as I say, I have no problems with links to any of my posts. But selectively edited information without links is not on, IMO. Nor is digging up information about me from elsewhere on the web (my university for instance).

I am REALLY fed up with your insinuations, tomm. If I give personal information you call it trying to give myself authority. If I withhold information you call it failing to disclose. You ignore all evidence I have provided of the work I have done to further investigation into the 2004 election, and continue to allege that my motives in expressing doubts as to the "fraud" explanation for the exit polls discrepancy are because I'm a paid mouthpiece for Mitofsky. And as for accusing me of "basking in the line-light" (what's a line light?) - it's not as though my paper exactly went down a storm here, is it? Basking in the brickbats, more like.

You are entitled to your assessment of me, of course. It is quite wrong, as I would thought any dispassionate reading of my posts or googling of my name would reveal - but whatever.

Edited for typo (extra word deleted)

[tommcintyre]

175. And...

I do believe we are allowed to speculate here (not maliciously, of course). So, no febble, the "thought-police" have not taken over DU. You will not be able (IMO) to twist the well-intentioned rules to your own apparent ends to choke-off legitimate information that may be helpful to people in making evaluations. Could you please leave that kind of behavior to the MSM and the Bush administration? ;)

Are you actually saying I dug up personal information on the web about you (such as your university); or are you just trying to get away with implying I did this? That really "came out of left field". Although, as I mentioned earlier, you have provided more than enough information in your previous posts to make it easy for anyone to find more information, if it exists. Personally, I have no interest in doing this, or finding out more about you.

But again, don't try to use this gambit as a ploy to choke-off the flow of appropriate (non-personal) information. I feel that has been your intent since you started this approach several posts ago. If you have had a legitimate problem with anyone revealing ACTUAL personal information about you, it should be dealt with; but don't try to get away with manipulating this important rule to your own ends. It will NOT be appreciated, I'm sure.

I've illustrated above (Subject: "There ya go again") some of the main "problems" I perceive with what I've read from you, so I won't repeat here. The expression is actually "lime-light" (bright stage-lights - vaudeville era).

[Febble]

181. Well, it has seemed

malicious at times, tomm. And while you yourself have not, to my knowledge, posted googled information about me, other posters did so on your thread awhile back. Luckily when I discovered the full set of rules I was able to get the thread deleted. And yes, the information included details of my university, as well as funds held by colleagues in my department.

I am certainly not attempting to manipulate a rule for my own ends, but I think it is an important rule. I was seriously concerned for my computer security. But it must be apparent to you that I am not manipulating the rule to suppress information as I have now put the information you want promulgated into my sig. What I consider unethical is posting information about me, without links to the posts in which you gained the information, and without context.

No, we don't want thought-police on DU, tomm, and I'm glad to hear you say that. I hope that includes thought-detectives. Because, to be blunt, you are not an accurate thought detective when it comes to me. But only I can know that.

[tommcintyre]

184. Actually, they would not have deleted the thread for that reason

They would've only deleted the offending post(s).

Here's the facts:
I have only had one thread ever deleted, and it WAS regarding you. The thread was in response to your "Report and Apology" thread (that was described as being heavy on "report", and light on "apology". ;) ) Mine was the "Raw data" thread (since it contained all the "un-spun" details of the "outing" from PM correspondence between us ( which you had given me permission to post a thread on). Actually, the only reason that thread got deleted was because it had become an embarrassment for you, right? (I got quite an angry PM from you (regarding it's continued existence) at the time it "disappeared".

I was online monitoring it at the time of it's demise, and I can assure you, of the 38 posts NOT ONE revealed the information you describe above.

Febble, this seems like a very "convenient" claim for you to make since the thread has now disappeared.

It reminds me of when you claimed you did not post an affirmation to Kathy Dopp's post, which explained you were working for Mitofsky. What you seemed to forget then, was that your post was up for nine minutes before you self-deleted it; and, someone else posted (in another thread) what they saw when they read it. Is there a pattern here?

[Febble]

188. tomm

I do not know why that thread of yours was deleted, but it certainly was deleted shortly after I sent an alert. The post that finally pushed me over the edge was a post giving details of my university, and my colleagues' fundholding details. I subsequently exchanged PMs with the poster, and I'm sure s/he will confirm this if you PM him/her. I have PM'd you with the name.

I also dispute that I claimed I did not post an affirmation to Kathy's post. I certainly self-deleted, then subsequently PM'd you to confirm its contents. You know this.

[tommcintyre]

215. Here's the proof:

http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=203&topic_id=371175#371807
Post by you:
"I have explained in emails to you, that I did not "attempt to cover up" after I was "outed" as you call it."

You were responding to the paragraph in the post immediately above by me:
"We now know Febble's since she has admitted (after being "outed" by Kathy, and subsequently trying to cover it up) that she works for Mitofsky. And, we also know (again, by her own admission, only after being "called" on it), that she feigned no foreknowledge of information to be presented at a conference, in an attempt to advance her position, and discredit USCVs'."

Also, due to the "discrepancies" in our two stories, I offered to re-post the deleted thread (#66 - same thread) that would clarify all of this. (Plus, as I have pointed out here, there is a post that identifies what you wrote in that deleted post.)
"As far as your first six paragraphs, we have quite different interpretations of what happened. I stand by my "raw data" thread, and, with your repeated permission, would be glad to re-post it."

I this clear now?

[Febble]

219. Will you stop this, please, tomm?

Edit: full rewrite of response

Thanks at least for providing a link. Here are my words in context:

Tom, you know, because I have explained in emails to you, that I did not "attempt to cover up" after I was "outed" as you call it. And I hope I have now made it clear that I was engaged by Mitofsky to do some work 8 days before the AAPOR conference at which he presented his new analysis, and that I agree I should not have posted at DU during that time. He engaged me because it was my work, discussed on DU and elsewhere, that had led to the new analysis.

You also know that far from "attempting to cover it up" I posted a confirmation of the fact as a comment on a thread as soon as the information was public.

You also know that I then deleted my comment. You do not know why I deleted the comment, although you do know it wasn't an "attempt to cover up" the fact, as I actually emailed you personally, in response to your call for information, to confirm the its contents.

Interested readers (are there any?) are welcome to read the rest of my post, linked to by tomm.

I've really had enough of this witch-hunt.

[tommcintyre]

221. YOU ASKED FOR PROOF IN A PM TO ME-Glad to stop it

Edit: I wrote most of MY reply, to the reply you had posted PRIOR to completely re-writing it. (For the record, I understand all edits are retained on the server.) I discovered your total edit just as I was about to post the reply (I wrote it on-and-off over several hours.) I'm just going to leave my original response intact, and add any additional comments in <>. I think this is fine since both you and I know what you wrote in both messages.
---------------------
Here's part of the PM:
You said:
"...I accept the possibility of False Recall..."

That's why I ended the post you just responded to with: "I this clear now?" I assumed I had clarified and verified, with links, to the point in question.

Frankly, I didn't expect a reply from you since the statements were so definitive, and even in your own words.

I won't reply to the specifics in your description in the post above <original one was much more specific and detailed>, of what actually happened, and when - other than to say there are major inaccuracies (such as mostly switching the time-order - to your benefit). If you want me to delineate these inaccuracies for you, just ask (in a public post this time because of the previous "confusion"). I will provide links and quotes once more, to clarify the situation. Fair enough? <Because of your "dual replies", if you request this, I will just lay out, accurately (and in the right order, "what happened, when" - with links. Also, since you know I DID read your original reply here, I'm sure you now realize I DO know the reason you deleted that message (reply to sunshinekathy) - you said so very clearly in that original unedited post.>

Also, I think you may have forgotten that it is the administrators here who make the rules, and not you. There is a good reason they made that rule (personal information), and it is NOT here to be used by you to control (for your convenience and preference) the flow of information. The information you had released about yourself previously, FAR exceeded SunshineKathy giving us fair-notice of your potential conflict-of-interest. And, THAT (and other legitimate reasons) is why the mods let the info stand.

Finally, you have made major accusations against other DUers at multiple times in this thread alone. First, you falsely accused me of releasing personal information on you (now you accuse sunshineKathy <in original post> ). Next, you accused someone of making religious slurs. As I pointed out (Reply #177), that particular "slur" ("fundie") has been used in over 7,000 topics in just one DU forum, in less than 10 months (wouldn't it be wise to check your facts a little, before throwing around such serious accusations?) And, finally, you accuse me in the post above of being on a "witch-hunt"? <You left THAT in both replies. > Hopefully, such unsubstantiated innuendos do not sit well with the admins or mods.

P.S. I will be glad to terminate this conversation with you. If you do not request more clarification; or make more unsubstantiated claims (that involve me, as in the post above) consider the conversation closed. Just don't try to "spin" things to your advantage, and then try to abruptly squelch any fair chance for reply, OK? That's just NOT fair, and progressives DO try to be fair, right? (The "right of response" is pretty fundamental.)

Message to mods: If you feel it's appropriate (violates any rules I can't figure out), please feel free to delete this post (you would anyway ;) ). Admittedly, this is the most unusual post I have ever made; but I only cared about "serving the truth", and this way seemed to do that best. If you have the time, you may find it useful to retrieve her original post (prior to total editing), and compare the two. Thanks for your consideration.

[Time for change]

16. I didn't say non-mathematical -- just non-exit poll related

For example, I believe that there is lots of evidence to suggest that Kerry won Ohio, irrespective of what the exit polls show, as indicated here in Bob Fritakis's book and summarized here: http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=203&topic_id=380878.

And as indicated by apparently fraudulent voter registration numbers, as described here: http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=132&topic_id=1974460

If these accounts are accurate, that would mean that not only did the official results in Ohio lean to Bush because of fraud, but by the same token so did the "adjusted" exit polls. If there is good evidence that this is the case for Ohio (and I believe this is) then that makes it substantially more likely that the same phenomenon occurred in other states where Kerry did substantially better in the exit polls than in the official results (and therefore the adjusted exit polls).

[OnTheOtherHand]

14. ah, heck, I'll give you a trophy, but let's play on...

I suspect that the phrase "Best Fit" (anax's, not mine) covers some differences in objective, but I don't think the differences prevent us from playing the game. I admit that I am not clear on the rules of the game myself -- but I trust anax not to exploit my confusion. So we begin.

Anax seems to think that there is no plausible scenario that allows Bush to win. So, for him, perhaps the object of the game is to prove that all other scenarios are implausible.

My position (EDIT: may not be -- I may misunderstand anax's) symmetrical, because I am not convinced that with available information, we could ever resolve the "best fit" question one way or another. For me, the object of the game is either to encounter the information I have overlooked or misconstrued, or to figure out what information we could gather to resolve the issue. (I'm not sure how one could prove that the vote count was right without counting all the votes, which we have no way to do. It has not escaped my attention that this fact has implications beyond the confines of the game.)

I don't see a rule that says one can't introduce fraud. Anax made a procedural suggestion that I would start by laying out some assumptions, then he would try to lay out the scenario where Rove won this, and then we would all tear it down... something like that. Probably it is useful to try to work on one scenario at a time, trying to get each one as lucid and perhaps "maximally plausible" as we can. I think the game will work best if all of us try playing on all sides at various times.

[anaxarchos]

17. You convinced me that wasn't needed...

I am happy to start with your scenario as you initially proposed. As far as "best fit" goes... is just science guys...

Don't they teach "thesis, antithesis, synthesis" anymore?

Sheesh....

(on edit - below)

You said: I'm not sure I understand the relationship between the "initial assumptions" and the "first test scenario."

I said: OK, If you don't think we ought to try it out first, then let's go straight to your red shift case. We can refine initial assumptions in that context just as easily. We'll do fraud later.

Don't tell me we are sitting and waiting for each other to serve... Hah.

Go ahead... It's your serve.

[Melissa G]

9. there are at least couple of three folks on this forum who believe

the Shrub actually won... they do some math and I hear they are gonna show us how it could be done.. I have popcorn :popcorn:

edit to say that the rules say anyone can play and so I have several bags of popcorn in reserve..

[Febble]

4. Well the attractive thing

about the unadjusted table (apart from the fact that it shows Kerry winning, which is really attractive) is that the proportions of new voters, 2K Bush voters, 2K Kerry voters and 2K Independents looks about right.

So TIA gets the first break of serve, no doubt. The question is: once the polls have been "adjusted" with reference to the returns, can the proportions of 2K voters be accounted for without zombies?

I'll break one rule (but this is a time out) by saying that an alternative game plan for those who would claim plausibility for the adjusted table is to argue that when the reweighting was done (and weights were used - we know that because they are in the "raw" data sets that are available - in the column marked "weight") the weights could simply have been applied to the wrong demographic groups. The algorithm was set up to match the vote the count, and there would have been an infinite numbers of way to reweight the demographic features to get the match. There is no a priori reason to suppose that the one the system settled on was the right one. It could simply have been the most parsimonious one. At the time the reweighting was done, no-one knew whether the problem was in the precinct selection or in voter selection (and we know the algorithm was based on the assumption that the count trumps the poll). Only later was the problem pinpointed to precinct level (i.e. voter selection - or vote counting).

OK, back to the rules: I assume what we are trying to do is to produce a scenario whereby the second table can make sense without zombies. I've got a few on my desktop, but I'd like to check it out a bit more thorough before putting it out there. (I've also got to drive to a distant airport early tomorrow so unless there's a rain check I might miss the fun). My self imposed rules are to try and produce a scenario where the assumptions about false recall are withing spitting distance of the figures OTOH posted on the other thread.

http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=379573&mesg_id=390057

[Bill Bored]

18. So nice to see you all playing so well in the sandbox for a change!

I'd stick around, but I gotta go decertify some DREs now.

Later! :)

[Bill Bored]

19. OK I'm back.

OTOH wrote:

"The weighted exit poll results render Bush the winner only by stipulating that Bush2K voters outnumbered Gore2K voters in the 2004 electorate by a proportion of 43%/37%."

Not so, OTOH. The poll as of 11/2 7:33 PM showed Bush 41/Gore 38. Based on this, Bush still could have won, but by a smaller margin. Yet Kerry was ahead 51/48 in this poll.

Also, since 3/4 of the respondents didn't answer (or weren't asked) this Y2K question, how can we compare these responses to those of the entire ~13,000 sample?

I don't think I like this game. I never did.

I like the Party ID question because almost everyone was asked and answered that one. But TFC and others have pointed out that even Party ID can be affected by poll bias, although it's not exactly the same as "are you a Bush responder?"

The Party ID shift was greatest in the West though and this suggests a time-dependent effect whereby the bias or the fraud intensified after the early exit poll results were released and either the Dems stopped voting in the West, or more of the Bush responders stopped responding there. This makes sense if you consider that Kerry had been reported as being ahead, and CA isn't exactly a swing state. A good portion of the popular vote swing could have happened there if the Dems stayed home after hearing the "good news" from the East.

Sorry if I'm cheating or perhaps just raiding the game.

[TruthIsAll]

20. Bill, good point . I have always said that if ONE demographic was bogus..

then they ALL had to be - by simple logic:

Assume Proposition A <> B.
Assume B = C = D = E = F = G = H
If A = TRUE and B = FALSE then B-H are all FALSE.

The Time Line
http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x363938


But we KNOW how to disprove the How Voted in 2000 demographic. The reason we use THIS one to play the game is that the FACTS are directly verifiable based on 2000 stats. We KNOW exactly HOW the votes were cast in 2000 and we KNOW that 122.27mm votes were cast (officially) in 2004. There were probably 3 million more votes.

According to the 2004 census (0.30% MoE), 125.7 million votes were cast. Even though they were lost/spoiled forever, it makes the case for fraud that much more likely since they were democratic votes. But we must work with what we've got. It's more than enough.

Party ID: another obvious mystery. Not. At 12:22 am the split was 38/35/27, which agreed with the prior THREE elections. At 1:25pm (the Final) it was 37/37/26.

Gender: Kerry was winning the female vote by 54/46 at 12:22am. This agreed with the Gore numbers. He ended up with 51-48.

I'm sure that the other categories (race, military, region, religion, education, ideology, income, when decided, etc.) were ALL manipulated to match the vote - last minute adjustments to weights, voting percentages or both.

As Deep Throat said to Woodward: "You are missing the overall. It was a Haldeman operation".

In this version of Votegate, it was a Rove operation.

[Bill Bored]

26. Not quite the timeline I had in mind.

Come on TIA, the idea that the last 600 respondents were all Bushies is just your way of interpreting the weighting change. I think we all understand that you don't mean that quite literally, right?

Here's what I think:

Let's say that OH and FL were in the bag because of Blackwell's tactics and paperless DREs that can't be verified, respectively. That about sewed up the Electoral College for Shrub, even if PA went for Kerry. The backup plan could have been IA, NM and NV, which would have tied the EC if Bush lost OH. Then they still had HI to play with for the win, with that Republican Gov. and several hours left. And Cheney was out in HI too, remember?

So they had the EC all locked up.

Now, the popular vote could have been manipulated by getting CA and other Western Dems and Independent Kerry voters to stay home. How do you do that? You tell them Kerry is going to WIN, based on the early exit polls!

At 7:33 PM on the East Coast, Kerry had his 3% lead in the popular vote. The Warren County lockdown hadn't happened yet and the rest of the crap in Ohio was probably not well reported outside of Air America and local media there. In any case, OH was still in play and Kerry led in the state exit poll there and in FL too.

Since CA was a blue state, I can see how under those circumstances, some Democrats, or some of their large number of Independents, could have stayed home because they still felt Kerry would win as of 4:33 PM Pacific time. Let's remember this is the blue state who elected what's his name -- the robot guy -- so they can be quite fickle.

So you have Bush winning the Electoral College fraudulently, but the popular vote through suppression of the Western Kerry voters by leaking the early exit polls that showed Kerry ahead. This is why the Party ID was so far off in the exit polls out west. From 4:30 PM on, the Kerry voters stayed home. Mitofsky doesn't figure this out until the next day (server crashes and all on election night -- amateurs!) when he makes a huge adjustment in the Western region Party ID weightings, practically reversing them. Other regions change too but not as much as the West because in the West, Kerry voters really did stay home because they "knew" he was going to win without them.

A late surge in Bush voters gave him more votes in Red and Blue states out West because they wanted to deny Kerry his mandate, even though they thought Kerry would win the EC based on the exit polls.

This is just my theory, but it reconciles a few things:

1. Shrub really didn't win hearts and minds. Some voters who were on the fence just stayed home. There is no mandate.

2. The popular vote may have been a toss up, like 2000. You had 9/11 and gay marriage and the hapless Democrats refusing to use 9/11 to their advantage in the campaign even though it was clear that at the very least, Bush was asleep at the wheel.

3. The fraud theory remains intact in FL, OH, IA, NM and NV where a few hundred thousand bogus votes gave Shrub mucho Electoral Votes. Additional fraud was unnecessary.

4. Exit polls were right in Freeman's original 3 swing states (FL, OH and PA???).

5. Ohio is so fucked up that comparisons between 2004 and 2000 indicating consistency are meaningless. (Refutes ESI's and DNC's analysis which are based on the false assumption that 2000 was somehow on the up and up.)

6. FL was stolen using DREs (Berkeley report comes close to confirming this statistically) and other miscellaneous tactics to cage Black voters and so on.

I'm tired now. Nighty night everyone.

[Time for change]

29. A problem with your electoral scenario

You say:
"The backup plan could have been IA, NM and NV, which would have tied the EC if Bush lost OH."

That can't be right. Bush DID win (at least officially) IA, NM, and NV, and yet if he had lost OH (officially) the elctoral college would not have been a tie, Bush would have lost.

[Bill Bored]

39. You are right.

They would have tied the EC if Kerry had won them and lost Ohio.
So rather than a backup plan, it was part of the primary plan for Shrub to win those states by any means necessary, and as it turns out, by an aggregate of only 20,000 votes!

At that point, even with HI going for Kerry, Congress would decide in favor of Bush if they could show he won the popular vote, and also because they're Republicans.

[TruthIsAll]

31. I said nothing here about the 613; "last minute" refers to after 12:22am

YOU
"Come on TIA, the idea that the last 600 respondents were all Bushies is just your way of interpreting the weighting change. I think we all understand that you don't mean that quite literally, right?"


ME
Of course.

Bill, of all people, why do you raise that favorite canard of the naysayers: That I said all of the final 613 had to be Bush voters.

I have always been consistent. The weights AND percentages were changed to match the vote across all demographics-period.

The fact that the final 613 were added to the pristine 13047 just buttresses the case. It was all a big scam to fool the public. They could not base the final on the 13047 respondents, could they? They had to make it appear that more were interviewed to better match the vote count.

You
Now, the popular vote could have been manipulated by getting CA and other Western Dems and Independent Kerry voters to stay home. How do you do that? You tell them Kerry is going to WIN, based on the early exit polls!

At 7:33 PM on the East Coast, Kerry had his 3% lead in the popular vote. The Warren County lockdown hadn't happened yet and the rest of the crap in Ohio was probably not well reported outside of Air America and local media there. In any case, OH was still in play and Kerry led in the state exit poll there and in FL too.

ME
You forget, no one was supposed to view the early exit poll timelines and Bush was winning the popular vote from the very first returns shown in full view on the TV screens.

The state exit polls would have stayed a mystery had Simon not downloaded them at 12:22am (coincident to the 13047 NEP).

In fact, I recall that YOU were the one who, just before the New Year, made all of us at DU aware of the 11027 (7:33pm) National Exit Poll. In fact, I used YOUR post as basis for my WTF!- that Kerry won 59% of NEW voters and therefore MUST be the winner.

In any case, we are diverting from the purpose of the thread, which is to see how Bush won (or stole) the election.

[Febble]

33. That's interesting, TIA

I never quite got your point about the 613 before.

But looky here:

You argue that

"The fact that the final 613 were added to the pristine 13047 just buttresses the case. It was all a big scam to fool the public. They could not base the final on the 13047 respondents, could they? They had to make it appear that more were interviewed to better match the vote count."

This makes no sense to me. No-one with an ounce of numerical nouse is going to believe that 613 respondents could have swung the numbers by 5+ points. As Bill says, they would virtually all have had to be Bush voters, and if so, the MoE would be as big as the Grand Canyon. But more to the point, no scam was required. It said quite clearly in the FAQ on the website:

How are projections made?

Projections are based on models that use votes from three (3) different sources -- exit poll interviews with voters, vote returns as reported by election officials from the sample precincts, and tabulations of votes by county. The models make estimates from all these vote reports. The models also indicate the likely error in the estimates. The best model estimate may be used to make a projection if it passes a series of tests.

(my bold)

http://www.exit-poll.net/faq.html#a10

So the obvious answer to any question regarding the final numbers is to refer to the FAQ - it changed because the precint vote returns and county tabulations had started to come in.

Obviously, if these were fraudulent, the thing will swing wildly at this point. As it will also do if the poll was strongly biased.

But no scam is required to account for the 613.

[Time for change]

36. Technically you are right

The information was there on the web site.

However, IMO, the good majority of people in this country would not get it, either because they didn't go to the web site, or because they wouldn't know how to interpret it if they did go to it. So, the good majority of people in the country would think, based on the final "adjusted" exit polls, that Bush won in the exit poll.

Of course, since that time there has been so much clamor about exit polls that quite a few more people than on election night now realize that Kerry in fact won the exit poll, both in the popular vote and in the electoral college. But I presume that the MSM didn't anticipate the outcry that would ensue. And anyhow, I don't doubt that still over half of the country may not realize that Kerry won the exit poll.

I'm not saying anything about Mitofsky. But I think that it is quite fair to say that the MSM would rather that there be no controversy about this election -- and that is an understatement. So, I am not accusing them of outright fraud in this regard. But I certainly don't doubt that they would think it was just fine if people didn't know that Kerry won the exit poll, and they certainly wouldn't go out of their way to enlighten us on this issue.

So, I see this as part of a much larger issue. The bigger issue, of course, is that the MSM has been irresponsibly silent on the whole issue of election "irregularities" -- unless you consider Michael Jackson's personal and legal problems more important than the state of our democracy.

Consequently, I think that it's quite understandable that so many people would see the early morning exit poll adjustment as just one more aspect of the MSM's desire to withhold information from us.

[Febble]

38. Fair enough

I'm not saying it's not understandable that people would think it was a scam. Clearly they did.

But while I am willing to entertain the possiblity of vote fraud as an explanation for the exit poll discrepancy, I am not willing to entertain the notion that the reweighting itself was a fraudulent.

Fair enough if people thought it was. But if you are going to run a scam, you don't advertise it on your website first. On the other hand, if Rove had read the FAQ he'd know that if his plan succeeded, the exit polls would automatically follow suit.

The hypothesis that the exit pollsters were knowingly involved in the alleged fraud makes no sense to me, and if anything, undermines the fraud case (why involve the pollsters when it clearly wouldn't be necessary?)

[Bill Bored]

40. Two ways they could have been involved:

1. By releasing the results early, they caused suppression of Dem votes on the west coast. (Certain Repukes have claimed this would have the opposite effect, but I don't buy that.) I refer here not just to E-M, but also their subscribers who released the polls to the public.

2. The pollsters could have leaked their precinct locations in advance so that these could remain non-fraudulent so as not to arouse suspicion. Fraud could then have been limited to non-exit-polled precincts, which were of course the vast majority. This would make it appear as if DREs had lower WPEs, for example, even though there could have been massive fraud in unpolled precincts.

[Febble]

41. Ha!

Re two:

Let me see, you are saying: that there was no fraud in the polled precincts, because the wicked Mitofsky leaked the locations to the Republicans, so that the exit polls wouldn't arouse suspicion, but then Mitofsky got the exit polls wrong anyway, and had to change them to fit the count, in case people got suspicious....

So there was rBr AND fraud. Okaay....

Re leaks: accident, design, it shouldn't happen. It's illegal in the UK.

[Bill Bored]

46. Not saying this happened everywhere.

I'm saying that if, for example, you wanted to show that paperless DREs had a lower WPE to make that technology look favorable, you could do so by letting the exit polls take their course without corrupting DRE exit polled precincts or entire paperless DRE states that were not crucial to the outcome of the election.

It might even be possible to do this without a leak, once the interviewers were spotted at these polling places, since the count could be altered even after the election on the central tabulators.

In these polls, we see that DREs and Op Scans have similar WPEs which contradicts studies that show DREs to be inferior at recording voter intent. I'm suggesting that this could be by design.

Let me ask everyone something:

Suppose a massive vote count discrepancy could be uncovered in a state that closely matched the exit polls, through actual auditing of the election. What would be your reaction to that? Feel free to start another thread about this if you like.

[Febble]

51. A lot of this depends on where you start.

If you start from: Bush couldn't have won because that many Americans couldn't have been so daft, I'm with you.

In which case, the exercise becomes, how do we square a stolen election with some actually fairly good evidence that the exit poll discrepancy did not correlate with an increase in Bush's share of the vote (ESI, Ohio) or was discrepant with an apparently clean recount (NH).

There could still have been fraud of state-swinging magnitude in Ohio if it was a) in unpolled precincts (your theory, which I accept is valid, just funny, in the circs) b) was carefully calibrated to give a linear relationship with the 2000 vote (TfC's theory) c) was mainly accomplished by voter suppression (which wouldn't have shown up on the exit polls) and targetted provisional vote/ballot rotations scams in Democratic counties, which the coarse grain of the exit polls missed (my favoured candidate).

And NH discrepancy could still have arisen from fraud if the fraud was at tabulator level and got missed by precinct recounts, or if the discrepancy arose from lost, rather than switched, votes (Land-Shark).

But it's all getting a bit elaborate - looks piecemeal, rather than a single throw of a master switch. That's what puts me on the skeptical side of popular-vote-stealing-scale fraud.

To take your last point: if an actual audit showed a massive vote count discrepancy in a state that closely matched the polls - I'd say that the audit was correct. Audit is what you need. The polls, as I've said too often already, were a kind of cut-price affair, designed to be calibrated by the early returns.

However, polls on the lines of the ones proposed here:

http://www.votewatch.us/Members/stevenhertzberg/report.2005-05-10.0988159730/view?searchterm=None

or executed here

http://exitpoll.byu.edu/

might answer questions better. The BYU methodology was very much more meticulous than the E-M poll (major difference: pollsters worked in teams, not singly), and got the Utah answer pretty well right (IIRC, E-M was four points off).

[Time for change]

113. Most definitely -- There are at least 4 kinds of fraud that wouldn't show

up as within precinct error in the exit polls.

This thread describes 3 of them:http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=203&topic_id=380878:

Item # 2: Electronic deletion of votes (via central tabulator I presume) in highly Democratic precincts, where the deleted votes are characterized by approximately the same vote share as the undeleted votes.

Item #5: Electronic addition of votes (via central tabulator) in highly Republican precincts, where the added votes are characterized by approximately the same vote share as the unadded ones.

Item #3: Voter suppression.

And lastly, and perhaps most importantly, this thread describes how it could have occurred via voter registration fraud: http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=132&topic_id=1974460


Note: The first two mechanisms would not show up as within precinct error, but they would show up in the state exit poll. The second two mechanisms would not show up in the exit polls at all.

[Time for change]

44. I'm saying a little bit more than that

And remember, my comments apply to the MSM, who own the exit polls, not the pollsters.

What I'm saying is:

1. The MSM would like the fact that Kerry won the exit poll (and all other 2004 election related matters) to be buried.

2. They did an extremely poor job of explaining their late night adjustment to the US public, even the fact that it was done at all.

3. I strongly believe that #1 and #2 above are related.

[Febble]

55. That I will buy.

I absolutely agree that the magnitude of the reweighting (as opposed to the fact of the reweighting) should have raised alarm bells in the media as it did in my living room. Expecially after the long queues in the rain (those were what convinced me Kerry was going to win, and later convinced me that the election had been corrupt - what kind of democracy requires people to lose a days wages in order to vote?)

I couldn't believe how little the debate over the legitimacy of the results made the mainstream press. No-one in the UK had any idea that the electoral vote was going to be challenged (thank God it was, and I did my best to do my bit).

So we are on the same page here.

But in the interests of fairness (because even on this thread, Mitofsky is accused of being behind the "scam" - why does no-one throw brickbats at Joe Lenski?) it is worth pointing out that the actual poll methodology, including the re-calibration with results, was up-front.

[Time for change]

60. It's good to know we're on the same page Febble

With regard to the fairness issue, I don't know enough of the details to accuse any specific individuals or to defend any specific individuals.

What I do strongly believe is that our news media is in the process of being hijacked by our Facist government, and that that is almost as great a threat to our democracy as election fraud.

[Febble]

62. Greater, I would say

and that's not to underestimate the importance of election fraud.

The thing that has frightened me for a long time about US politics is the lack of an independent media, which, TG, we still have, sort of, in the UK, although the way certain tabloids go can still swing an election. But the Guardian and the BEEB are great bulwarks against election-by-television, and our comparatively tiny election budgets and snap elections (as little as three weeks notice) means that our choices have sporting chance of being informed by the issues, not by the way the isses have been packaged.

Also our parliamentary procedures (PM's question time for instance) do mean that our politicians are regularly exposed to difficult questions, and can be seen to squirm when they do.

Not that all's well this side of the pond!

[Time for change]

64. Well, if things get much worse here I'd have to consider transporting

myself to your side of the pond.

If only it wasn't so damn cold there.

[Febble]

66. And Blair is the leader of the

progressive party.

(sigh).

Wait till we've got rid of Blair and installed Brown. It'll happen eventually.

[Bill Bored]

106. Well that's the guy you actually "voted" for right? nt

[Febble]

109. Precisely not

except in the sense that around 10 years I voted for him as leader of the Labour Party.

In our UK electoral system we vote for members of parliament, and the leader of the party with the most members gets to form a government. But the leader may not even win his own seat. And there is no guarantee his party will keep him on as leader.

I voted for my Labour Party candidate, an excellent man who voted against the Iraq war, as did more Labour Party MPs than MPs from any other party. It was a difficult election, and a lot of people voted tactically. No-one in progressive politics wanted Howard's Tories in.

But if Tony Blair had been my candidate I would have voted for one of his opponents. And I hope that my party will eventually succeed in forcing his resignation.

For more information see here:

http://neweuropeantimes.com/forum20/3.html

[Bill Bored]

122. I was referring to Brown! You voted for him by voting for the party.

It just happened to be Tony at the helm.
One advantage of our system, as corrupt as it is, is that we get to vote for PEOPLE and not just parties.

[Febble]

124. huh?

Your's is possibly an advantage. Although we stand a greater chance of getting rid of Blair before the end of his term than you do of Bush.

But this is definitely off topic. It was Hobson's Choice for "liberals" (quotes because Liberals are something else) in the UK at the last election, and I opted for a tactical vote for a good guy (Alan Simpson). Honestly, you wouldn't have wanted Howard as PM, Kennedy didn't stand a chance, and this way we should end up with Brown, who is at least as good as Kerry.

At least we cut Blair's majority down to size. People like Simpson have more clout now.

[Bill Bored]

42. TIA, it was well established by MSM and Air America

that Kerry was ahead. The fact that we didn't have Simon's screen captures is only relevant to quants. As far as potential voters, there was PLENTY of media hype about Kerry winning the Presidency as early as the afternoon on the east coast on 11/2. Keith Olbermann commented on this too, so it's not my imagination or my choice of Air America for my election day coverage that leads me to this conclusion. Victory was in the air and this is one reason why we were so disappointed.

But it's not unreasonable to suggest that there was vote count corruption in selected counties in a few key states to deliver the Electoral Votes to Shrub (as Wally O'Dell said he would do with Ohio's), combined with a very efficient method of suppressing the popular vote elsewhere by telling us that Kerry was going to win!

[TruthIsAll]

50. The reason people decided whether or not to vote had nothing to

do with who was leading. That is an old wives tail, promulgated by Repukes in 2000 FL. Remember the Panhandle?

They may stay home in a landslide, but not in 2004. BUSH was leading (on TV), not Kerry. And he was leading in the East, of all places.

Democrats were angry, pissed, motivated. That's why there was a massive turnout. Kerry was leading the exits, but very few knew about it. They were glued to their TV screens, not computers like you and I.

BILL, THE BEAST WAS IN THE EAST, THE PEST WAS IN THE WEST.

[Bill Bored]

53. Well, perhaps you will be more comfortable with my response to

Peace Patriot's post which explains the Boxer/Kerry discrepancy.
Now this would also explain the Connolly/Kerry stuff in Ohio, but since judges are non-partisan and OH was mostly punch cards, it's much more plausible for this to have happened in CA. DREs are the Devil's playground!

[tommcintyre]

24. Or, calling it for what it is:

"framing the debate".

When one is allowed to pre-set the parameters of the debate, it inherently limits (controls) the possible outcomes.

For example, one could ask the limited question, "why DID Bush voters stop responding late in the day? Assuming I
buy in" to the limited framing that would only choose an rBr answer, my answer may be: :because of their inherent sour dispositions, they get cranky when they are tired". ;)

Of course, if ALL possibilities are allowed (no framing), the most probable truth comes out (Occam's razor): Mitofsky made rBr up (did it in '92 also, for "papa" Bush).

You're right... I don't like playing "framing the debate" either. ;)

[OnTheOtherHand]

27. yes, I overstated that

and I didn't mean to imply that Bush would need every point of that 43/37 retrospective margin in order to win the popular vote.

It might have been better not even to mention the exit polls, and just to frame the problem as: if Bush2K and Gore2K voters turn out in roughly equal proportions, and they "defect" to the other party at roughly the same small rates, and Other2K and DidNotVote2K voters break for Kerry, then Kerry wins the popular vote, period. Which is apparently what most people on DU believe happened. (This view does not appear to be widely held among political scientists, and I find this difference in perspective interesting in itself -- but that would be a different thread.)

If I were going to sort out your last point about party id shift, I would look closely at two things. First, I would expect more Dems to vote for Bush in the south than elsewhere, and I wonder whether that affects your analysis. Second, there is a literature on whether exit poll calls affect West Coast turnout, and I think it is pretty dismissive. Even if that literature holds up, that wouldn't necessarily refute your idea in 2004.

I don't think I will comment separately on your other post, but I think it proceeds in a reasonable direction. It's sort of a hybrid scenario (maybe Bush won the popular vote -- or didn't, hard to say -- but stole the election anyway in one or more battleground states). I don't think we are ready for hybrid scenarios on this thread. At this point I will be amazed if we can make it through one scenario. But the investigations in battleground states should continue. I'm not convinced that Bush stole any states (which I mean in the sense of actually stealing or inventing votes), but I'm certainly not in a position to whistle the all-clear.

[anaxarchos]

34. Couldn't agree more on your first paragraph...

The rest are artifacts of previous conversations.

We'll get there...

[Bill Bored]

47. I like the hybrid scenario

but I will say that we also have to take the media (useless for the most part), 9/11 (fear and "war on terror") and the wedge issues (God/guns/gays) into consideration to eplain the popular vote.

Given the numbers, new voters, etc. I'd say that the answer to the puzzle, other than fraud, is that more Dems voted for Bush than Repubs for Kerry and this offset the new voters who voted for Kerry. The exit polls do show more Dem defectors than Repubs, even at 7:33 PM.

[TruthIsAll]

54. Bill, that's not true. 12:22am 8% Gore to Bush; 10% Bush to Kerry

__________________________________________________________


National Exit Poll: 12:22am, 13047 respondents				

Kerry wins by 4.5 million votes

Voted	Mix	Bush	Kerry	Nader	Votes
No	17%	41%	57%	2%	20.78
Gore	39%	8%	91%	1%	47.68
Bush	41%	90%	10%	0%	50.13
Nader	3%	21%	71%	8%	3.67
98%	100%	47.62%	51.41%	0.97%	122.26
	122.26	58.22	62.85	1.19	
_________________________________________________________

[Bill Bored]

80. TIA, I didn't quote this one so how can you say it wasn't true?

I quoted 7:33 PM and the final adjusted version.
I use them because they were relased officially by Mitofsky, via AlTheCat. The midnight one is from the Washington post, right?
I don't have it handy.

On edit, see this post:

<http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=390193&mesg_id=390431>

We may be talking apples and oranges.

[TruthIsAll]

90. Not so. Take a look at the 7:33 poll of 11027

Voted Pct Kerry Bush Other
No 17 59 39 1
Gore 38 91 8 0
Bush 9 90 0
Other 65 13 16

And why are you even citing the the Final 1:25pm numbers?
They were matched to the vote, remember?

http://www.exitpollz.org/mitof4zone/US2004G_3798_PRES04_NONE_H_Data.pdf

[Bill Bored]

96. That's what I said: 90% - 91% Y2K party loyalty. Final version, the same.

I cited the final to show that for those who consider the vote count to be accurate, the final still showed that Y2K voters had 90% Y2K party loyalty, which was what OTOH was disputing.

You guys really spend too much time on this stuff though! :)

[Melissa G]

98. some of us just refuse to let the IMO Lie that Blivet won continue

to be spread unchallenged but I know I and others also do things like work on de certifying DRE's as instructed by our good friends Roj and BB as well..we even link to their Greatest thread...
http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=390084&mesg_id=390084

[TruthIsAll]

105. Except that Bush went from 8% of Gore votes to 10% in the final.

Why the 25% (8>10) loss in Gore voter loyalty?

Or was it something else?

A percent or two here, a percent or two there, and before you
know it, you have a completely different result.


Voted Mix	Bush	Kerry	Other	Votes
No      17%	45%	54%	1%	20.78
Gore 37%	10%	90%	0%	45.24
Bush 43%	91%	9%	0%	52.57
Nader 3%	21%	71%	8%	3.67
Total 100%	51.11%	48.48%	0.41%	122.26
          122.26	62.49	59.27	0.50	

[Bill Bored]

107. You're right and there were probably some rounding errors too

but the point is that the Y2K result was substantially similar in both the pre- and post- adjustment polls and this would seem to refute OTOH's claim that the Y2K Dem voters defected at a higher rate in 2004, which is the serve we're supposed to be returning at the moment, albeit further down the thread.

[Dynasty_At_Passes]

21. Here's how to play this: Stick to the facts.

Lets go with the official reports of http://www.votersunite.org and lets completely throw out one million votes for Bush, that were stolen directly from John Kerry.

We have enough data to show that one million votes were stolen from John Kerry.

Lets say those votes are retained in Kerry's column, and there is a voting poll bias of 54/50 biased towards Kerry, but biased towards Bush in the democratic strongholds due to supression.

Now, Bush still wins the pot by 2 million votes and still gets 10 million new voters.

He still has won the popular vote, despite 1 million fraudulent votes thrown out.

The challenge is to explain where those ten million votes came from, that went directly to Bush's column

This is assuming the maximum possible level of spoilage: He STILL wins, because of these votes, so where do they come from?

Did fundamentalists, all 10 million of them from the 122 million pool go out at 12 midnight and vote for Bush?

Is it POSSIBLE for all the vote returns of these 10 million votes to come in late, adding an extra 2 million votes to Bush's final lead over Kerry?

How would this happen? How do you come to the results? And finally what affiliation do these voters have, exactly....?

[TruthIsAll]

22. Bush won by 3 million votes, NOT 3.5. n / t

?

[Dynasty_At_Passes]

23. Thanks, edited for clarity. n|t

[Time for change]

30. I'm not sure what you're saying

There was no late surge for Bush.

The sudden exit poll change at 1:30 a.m. on Wednesday morning was due to an "adjustment" of the raw exit polls by the pollsters to bring them into line with the official vote. I believe that that fact is accepted on both sides of this issue.

[Dynasty_At_Passes]

77. Do we know that for sure?

It seemed to me that while Kerry was ahead at 12 midnight, he was no longer ahead by 2AM in the morning and lots of Bush votes just appeared "out of nowhere" which indicates either a surge or unreported results.

Mysteriously Bush ended up ahead by the same amount he was behind.

[Time for change]

166. That's what I'm saying

Kerry was ahead in the exit polls at midnight, and even after 1:00 a.m. He's still ahead with respect to the raw exit polls. The unadjusted final exit polls have Kerry ahead by 3.0 nationally and by 4.2% in ohio. But at about 1:30 the exit polls were "adjusted" to match the official vote count. There was no change in the numbers, they just changed the percents to match the vote count. My own son (and probably hundreds or thousands of others) has a CNN screen shot that shows the abrupt change in Ohio from one minute to the next a little after 1:00 a.m.

But there's nothing terribly mysterious about it. It's even explained on some website (Mitofsky's?), but the explanation is so poor that you wouldn't expect many people to understand it.

See posts 33, 36, and 38 of this thread for a more complete discussion of this.

[Past_World_Doubt]

178. So then why is this called an invisible surge?

Were the voters who appeared for Bush invisible? Were all the polls flawed? Are the machines doing something naughty?

[Time for change]

187. I believe that massive voter fraud was the reason for the Bush "victory"

And I have posted numerous threads explaining why I believe that.

I can't explain why this is called an invisible surge. I guess because very few people (including me) expected Bush to "win".

The polls may or may not have had some flaws. But there are massive amounts of evidence for fraud, especially in Ohio:

http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=132&topic_id=1974460

http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=132&topic_id=2020121

http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=203&topic_id=380878

As far as the "surge", it seems to me that these fraud factors had been operating all day long, and even in the weeks and months prior to the election, not to mention during the Ohio recount. But I don't believe that there is evidence of a late in the day "surge" for Bush.

[Time for change]

192. Whoops -- I meant to say election fraud, not "voter fraud" n/t

[Peace Patriot]

43. Bill, I don't think California voters are/were fickle. Here's why...

You cite our possible fickleness because of the vote for the Robot, but I suspect that that election, too, was fraudulent.

2003: Kevin Shelley in office only about six months when that Recall campaign began (Sec of State who sued Diebold and decertified their touchscreens in 2004). New electronic voting systems just coming on line, many of them Sequoia, authorized by former Repub Sec of State Bill Jones, who then (along with his chief aide Alfie Charles) went to work for Sequoia. Very smelly.

Shelly barely had time to grok these new systems and to be the "new sheriff in town" as he vowed upon taking office (to miscreant, corrupt county election officials, who immediately started fighting him). Backed by rightwing $$$ billions, the petition for Recall qualified, and was designed as a rushed, 6-week campaign, with no primary (to vet the candidates), and included a two-part ballot (1. yes or no on Davis. 2. Choose one.) with ONE HUNDRED AND TWENTY-FIVE CANDIDATES on the ballot, to replace Davis.

Due to corporate news monopoly malfeasance, the general population had not been informed about Schwarzenegger's meeting with Kenneth Lay/Enron in May 2001, to plot all this (Enron stole $9 billion from Calif., thus getting Davis in trouble with the voters due to budget shortfall--and Lay wanted this debt to be forgiven), and there was not enough time for alternative news sources to get the word out that Schwarz was a Bush Cartel shill. Time magazine put Schwarz on its cover during this very short campaign, and Larry King and others gave Schwarz (already famous from his movies) millions and millions of dollars in additional free publicity. They almost didn't need electronic vote-stealing, but they had it all set up, if needed--and had ONE HUNDRED AND TWENTY-FIVE minor candidates to shave votes onto (away from the main opposition, Lt. Gov. Cruz Bustamante).

I was simply horrified by this Recall election--by its design, by its speed, by its unfairness (Davis had just won reelection the year before), and by its sheer nutsiness. It was a frigging coup. And I am totally convinced that Schwarz could not have won a fair election here--against Bustamante or any other half-way decent Democrat. I think it was a combo of Time/Larry King and vote rigging in the new electronic systems.

After Shelly had some time to assess these systems, he immediately started trying to protect the rights of voters--in all sorts of ways, including suing Diebold and providing Californians with a paper ballot option for 2004--with corrupt county officials, like DEMOCRAT Connie McCormack in Los Angeles, fighting him tooth and nail all the way, and ultimately driving him from office, when a Bush Cartel black op succeeded in floating substanceless charges of corruption against Shelley.

I know that your "fickle" remark was a toss-off, for the sake of the game. I just wanted to set the record straight about Californians and the Robot. (His polls have tanked just like *'s, once Californians had a chance to see who/what he really is.)

(Note: Calif voter registration, as of March 2004, was Dem 43%, Repub 34.7%, other parties and indies 32.28%. The Calif legislature is 2 to 1 Dem/Repub.)

------------

THE GAME: We need a calculation for which I don't have all the stats.

To assess whether or not Californians stopped voting late in the day on Nov. 2, 2004, upon news that Kerry was winning, we need to do a calculation of California voter turnout in 2004 vs. the rest of the nation. On the face of it, California voter turnout kept pace. In 2000, there were about 10 million votes cast for president in Calif. In 2004, there were about 12 million--an increase of 2 million votes. Is that comparable to the overall increase in voter turnout nationwide in 2004?

Does someone have the nationwide vote totals for president, 2000 vs. 2004? I think 2004 was 120 million, wasn't it? What was 2000?

And to refine the answer, I guess we need some party affiliation stats, too. Is there some demographic of Dem/Repub turnout, or Gore/Bush vs. Kerry/Bush votes, that would tell us whether or not Calif Dem voters, or Calif Kerry voters, fell down on the job in 2004?

California:

Kerry 04: 6,745,485 votes (54.4%).
Bush 04: 5,509,826 votes (44.4%).

Gore 00: 5,861,203 votes (53.5%)
Bush 00: 4,567,429 votes (41.7%).
Nader 00: 418,707 votes (3.9%)


The really interesting statistic is this: While John Kerry beat Bush in Calif. by a 10% margin, senate incumbent Barbara Boxer beat her opponent (Bill Jones) by 20%. Some of this difference between Kerry and Boxer might be attributable to incumbency, except that ALL of Boxer's edge over Kerry, as to votes and margin of victory, comes in the most rightwing counties in the state. Go figure. (People voting for Boxer...and Bush?!)

If it were Boxer's incumbency--or some other factor (her being a woman?)--the impact would likely be across the board in all counties (Dem or Repub). The Boxer edge over Kerry ONLY APPEARS in Repub counties (and, in Calif, some of our Repub counties are real rightwing hotbeds.)

I can't find a correlation between this odd Kerry deficit (vis a vis Boxer) with voting technology. It seems to be related only to Republicans having political control in that county. But, Boxer being to the left of Kerry, that doesn't make sense.

Boxer 04: 6,955,728 votes (57.8%).
Jones 04: 4,555,922 votes (37.8%).

(Boxer got 2,399,806 more votes than Kerry, overall; got 3.4% more of the total vote; beat Jones by 20%--while Kerry beat Bush by 10%--and all of that difference is found in the most rightwing areas of the state.)

What it looks like is that Repub county election officials stole votes from Kerry, using whatever technology they had to hand. (Note: From what I've seen, the corruption on the expensive new electronic voting systems is bipartisan. We have a big problem of Dem corruption on this matter--not just Repub.)

Shelly was still the Sec of State during the 2004 election, but he hadn't yet been able to do anything about the secret, proprietary programming code, which Bill Jones (former Repub Sec of State; Boxer opponent; now working for Sequoia) had locked into contracts containing "trade secret" provisions, setting the precedent. Shelley had demanded to see Diebold's code, in the course of his lawsuit against them, but that demand was left in limbo when they got rid of him. I suspect that THAT is one of the reasons they got rid of him--he was the first and only Sec of State who had demanded to review source code--in addition to suing Diebold and decertifying their TS's. Review of any of Diebold's source code could possibly have blown the 2004 election fraud scandal wide open.

We now have a Schwarz APPOINTEE as Sec of State, who has packed his advisory board with Bushites and Diebold proponents. This does not bode well for future Calif elections--although citizens here are waking up, and a big protest in Sacto, and outpouring of letters, prevented the worst of Diebold's election theft machines from getting certified this year (--in addition to a 10% failure rate in tests of the machines).

I've heard a lot of explanations of the Boxer/Kerry discrepancy (from argumentative friends), but none of them holds up very well. For instance, that the Boxer incumbency factor was at work in Repub counties, but NOT in Dem counties. (Kerry and Boxer got approx. the SAME number of votes in Dem counties.) To me, incumbency at work in Repub counties only can't really explain a TEN PERCENT difference in Boxer's and Kerry's margins of victory. From Repubs' point of view, they had a perfectly respectable Repub choice for Senate in Bill Jones (no skin off their noses if he'd totally corrupted Calif's voting system--and they likely didn't have a clue about that anyway). SOME Repub women voting for a woman? Possible, yes--but if women's rights were a concern, how could they vote for Bush? --just schizophrenia maybe? okay. Boxer wasn't all that far to the left of Kerry in 2004, or wasn't perceived as such (? --but she did vote against the Iraq war!). okay, maybe.

But still...it's such a big and dramatic difference--10%--and ONLY in Repub counties, that I think a more relevant question is, if they were stealing votes from Kerry, why didn't they also steal them from Boxer? (Possible answers: vote stealing comes from the top, from a directive, and the Repub powers-that-be possibly didn't want Jones to become a rival to Schwarz--which he would have become, if he'd done well against Boxer; and, Boxer was a shoo-in, no way to beat her, why double the risk of detection, if you can't change the outcome? Also, votes for Boxer were NOT needed to pad Bush's popular nationwide majority!)

CONCLUSION FOR THE GAME: If it's true that California had a comparable turnout with the rest of the nation, and if no one can provide a fact-based explanation for the Boxer/Kerry discrepancy in Repub Calif counties, then any impact that Calif Kerry votes (a lack thereof) had on the national popular vote was not likely due to "fickleness" (Calif Kerry voters staying home), and was more likely due to fraud, specifically, votes stolen from Kerry in Repub counties (especially given what we know about Repub election officials' behavior in other states).

[Bill Bored]

48. As usual, a lot to consider in your post PP.

I agree with almost all of it though.

One reason for the Boxer-Kerry discrepancy could be this:

A certain percentage of the machines were set up so that anyone who voted for BB (not me -- the other BB!) also had their vote switched from Kerry to Bush. This might be shown by an audit of ballot definition files, not necessarily the source code, because alterations to the source code may not be required to set this up. The issue is whether this would show up on the summary screens, how many voters would be savvy enough to see and correct it, etc. Or, this could have been done to Republicans for Kerry who voted against Boxer. They would be less likely to complain about it than Dems.

As far as the SoS stuff, it's time for you guys to get one of your famous ballot initiatives to take the power to run elections away from this office and move it to a bi-partisan state BoE. And also, get those voting machine rules, VVPATs, auditing, hand recounts, etc. into state law, instead of SoS or BoE regs if they're not there already.

What do you think?

Also, CA turnout was about the same as 1992. Should it have been higher? I'd say yes if it weren't for Perot!

[anaxarchos]

45. Ladies and Gentlemen...

If someone (and you know who you are...) doesn't post a scenario soon, this game is going to collapse like a flan in a cupboard.

(on edit - Below)

Though I do dig the fact that we get 45 comments BEFORE the first serve...

[OnTheOtherHand]

49. is this a "serve"?

OK, here are some numbers on my laptop. Nothing sacred about them.

Currently working with 94.0% turnout among Gore2K and Bush2K voters, 97% among Other2K voters -- gives a decent number of DidNotVote2K voters under my other assumptions.

So: I've got

Gore 2000 46.275M 84.83% Kerry, 14.63% Bush
Bush 2000 45.782M 7.20% Kerry, 92.31% Bush
Other 2K 3.701M 65.90% Kerry, 18.10% Bush
DNV 2K 26.535M 52.90% Kerry, 46.50% Bush

which yields, within a few hundred, the official result of Bush winning by about 3 million votes.

Then downstream I have a pro-Kerry "response bias" on the order of 1.12 to give Kerry a 3% margin in the national exit poll.

I further have some estimates, inspired by the 2000-04 NES panel, of false recall. For instance, I observe that among panel participants who said in 2000 that they had voted for Gore, 7.3% said in 2004 that they had voted for Bush in 2000. In this group, 26% said they voted for Kerry, and 65% said they voted for Bush. The sample is small, but hey, it's a starting place, so I use those proportions. (I similarly assume, also based on the panel, that 1.6% of Bush2K voters false-recall themselves as Gore2K voters, and this group breaks 95/5 for Kerry. In the actual panel, there were only 5 such people, and they all voted for Kerry.)

I end up with the final results for the unweighted exit poll:

reported Gore2K voters, 38.8%, of whom 90.3% Kerry, 9.2% Bush
reported Bush2K voters, 40.2%, of whom 9.3% Kerry, 90.3% Bush
reported Oth2K voters, 3.3%, of whom 69.2% Kerry, 17.0% Bush
reported DNV2K voters, 17.6%, of whom 56.1% Kerry, 43.2% Bush

all reasonably close to the actual unweighted exit poll results -- I am low on reported DNV2K voters (18.4% in the actual unweighted results). The actual unweighted results show Bush2K voters outnumbering Gore2K voters by only about 1.1%. (But, by assumption, Kerry voters -- many of whom were Gore2K voters -- are overrepresented in the exit poll. Downweighting the Kerry voters to compensate should yield a reported Bush2K/Gore2K split close to the 43/37 margin in the official results.)

[anaxarchos]

52. That, most definitely, is a serve....

More like two, actually (both an actual vote hypothesis on how the 2004 elections played out and a second hypothesis on how the exit polls could reflect that scenario).

Before anyone attempts to "return" (Is either hypothesis "plausible"?, etc.), do you want to explain (or defend) what you've done? As far as the actual vote scenario goes, I see a good deal of point shaving (at least against the adjusted NEP - BTW, perfectly legit if it can be explained) but the main premise still remains: "Bush won because of a dramatic red-shift of Gore voters from 2000, at approximately twice the rate of blue-shift in the other direction".

But I don't want to put words in your mouth...

Would you prefer that you list your basic premises or that we ask questions?

[OnTheOtherHand]

56. I think that is the basic premise

Honestly, I'm not sure how I ended up with these exact numbers -- I am sorry to confess that I haven't performed a complete sensitivity analysis yet. I think you could nudge both those steal percentages down, and twiddle with turnout, perhaps obtaining more realistic results.

However, I (necessarily) embrace your main premise:

"Bush won because of a dramatic red-shift of Gore voters from 2000, at approximately twice the rate of blue-shift in the other direction".

And my warrant for this premise -- beyond the fact that I think I was given the assignment of coming up with a scenario that would confirm the official result -- is that in the 2000-04 National Election Study panel survey, in fact 6.3% of people who reported in 2000 that they had voted for Bush, reported in 2004 that they had voted for Kerry. And in fact 12.4% of those who reported in 2000 that they had voted for Gore, reported in 2004 that they had voted for Bush.

(Those are unweighted results, FWIW. When I apply the panel weights -- which should be, if I understand rightly, purely demographic -- the figures are 6.6% and 14.0%. Let me underscore that the NES results, whatever else their merits and failings, are not weighted to official election returns.)

I don't think that crosstab result should be accepted as gospel. But it should at least be considered.

[anaxarchos]

57. OK then,... the game has begun n/t

[anaxarchos]

67. Before I respond, a question....

(Before we get to even that: Could you post a link whenever you present "evidence", please? I'm guessing you are referring to the 2000-2002-2004 NES panel study but it helps to be specific {what and where exactly}, so we don't have a lot of digging to do.).

I really want to start with the actual election (the first part of your serve) but I have a problem. The only evidence you have presented for the plausibility of your split is the research you have presented in the second part of your serve. Before I go digging out NES panel studies, isn't that an unverified panel (i.e. a panel of "eligible voters")? If so and it is not a panel of "verified" Gore voters, then you can not apply that conclusion to a REAL population of Gore voters. You might (and I have some doubts) be able to apply it to a sample of people who SAID that they voted for Gore in 2000 and actually voted in 2004 but, even here, we would be on thin ice.

I know it can't be that simple but, I thought I would ask before I dig.

[OnTheOtherHand]

84. hmm

I've tried to specify numbers that are based on the 2000-04 NES panel, yes.

You are absolutely right that I cannot authoritatively argue from, well, any survey whatsoever to any particular finding. We have a secret ballot, after all. So, we could and can speculate about the actual proportion of (true) Gore voters in the panel who voted for Bush in 2004. We can speculate about everything.

I do hope we will be sort of even-handed about the speculation. It's certainly not obvious to me why the NES should have some sort of intrinsic bias in favor of exaggerating Gore defections. Actually, I sort of think it might tend to understate them, but I won't press that point.

[anaxarchos]

99. 'Snot the point....

I asked the question correctly. Are you sure you want me to respond to this in the way you have proposed it?

If you like, we can call a fault and go to second service but the issue seems glaring to me (i.e. Using a yardstick to recalibrate a micrometer).

But, then again, we are on your "turf".... not mine.

...still, I am itching to answer this as it stands.

[OnTheOtherHand]

137. whoops, didn't mean to leave you itching

(I have to chuckle at the statement that we are on my "turf," but whatever.)

I don't claim to have precise estimates of false recall, although I think I have pretty good evidence of false recall.

Sure, answer away.

[Melissa G]

76. Question here OTOH.. 'You were not really given the assignment'..

You actually previously said that you believed the Shrub actually won..I had my popcorn out to see how you would explain this... So in this scenario you give are you 'playing the game' or saying what you really think happened? Just curious....:popcorn:

[OnTheOtherHand]

85. we play the game to find out or refine what we think

I do believe, with fairly high confidence, that Bush won the popular vote. I'm quite certain that my scenario here isn't exactly right. But I don't see a knock-down rebuttal that proves it is impossible.

[Melissa G]

92. Sorry OTOH, can't resist...Tried, but can't...

You write.."I do believe, with fairly high confidence, that Bush won the popular vote. I'm quite certain that my scenario here isn't exactly right. But I don't see a knock-down rebuttal that proves it is impossible.'

If you are Believing with fairly high confidence that the blivet won, yet you are not confident in a scenario you just came up with... Where does your high confidence that you seem to have had for a while come from??? OK... ( :dilemma: and here is the part of the question I tried to resist and couldn't) Is your belief Faith Based? :evilgrin:



edit clarity..

[OnTheOtherHand]

114. of course you could have resisted

and I really think that the "Faith Based" bit should be summarily banned from the board. I can't think of any case in which it clarified the issues.

(EDIT: But I do understand why you might have been confused about what I was saying about confidence and models, so let me try to explain that again.)

I am pretty confident that Bush took a lot more of the Gore2K vote than Kerry took of the Bush2K vote. That fits with the general incumbency advantage, the fact that the pre-election polls were close and leaned toward Bush, and the results from the 2000-2004 NES panel cross-tabs. Probably other things, too.

But my specific numbers were based on something like 29 distinct parameters, and I doubt that any one of them -- much less every one of them -- is exactly right. That isn't the point of the model. (Of course, the 29 parms aren't equally important. The really crucial ones, I think, are Bush2K->Kerry, Gore->Bush'04, and one or two of the false recall numbers. All the values I am using for those are within spitting distance of figures derived from the NES.)

I could do a "sensitivity analysis" on the model, and -- as long as you let me set the range of the parameters -- I could give you 120 or 12 million scenarios, in each of which Bush wins the popular vote. But that would be a cheap rhetorical trick.

[Melissa G]

139. okay, thanks for the clarification and no, if you give me a hole

big enough to drive a Mac truck through in the game i am not gonna resist...:evilgrin: ( FYI.. from reading your posts and pm's I think you might need to know that you expect folks to mind-read pretty often..)

So I now hear you saying that the reason for your previous confidence
in a Blivet win is a pile of (IMHO stinky) 29 parameters that resulted in an inordinately difficult to imagine blue to red shift, the biggest of which was about a 15% Gore 2K defection.
This really stretches my imagination ..
I would refute this but Anax does it so well in post 134 that it seems unnecessary. I would just add that I live in the South....
Not only in the South, but in Texas as in Austin where the Shrub used to live and i can personally report MASSIVE Bush shift to Kerry
.. The first time around folks were saying that they were going to be voting for Blivet not because the liked him or because they thought he was smart but because they thought it would be good for Texas to have him in DC. The same folks the second time around were very clear about the error of their ways...barely willing to admit or want to be reminded they had voted for Blivet the first time.
I find your 15% evidence very unconvincing in the face of my southern,Texas experience.

[OnTheOtherHand]

142. that also seems like pretty circular reasoning to me

The country is pretty sharply polarized right now, so it is pretty easy to go around having conversations with people who already agree with us.

But the polls pretty much agreed that the race was at least close, so I think people who are sure that Kerry won it -- absent clear and direct proof of fraud -- are sort of blowing smoke.

If you don't have any specific criticisms of any of my 29 parameters, then, yes, you too are blowing smoke. But it doesn't mean that I don't like you.

[Melissa G]

146. Sure it is easy to have conversations with people who agree

with us.. But these are the SAME people who were disagreeing in 2000...
In Austin in 2000, there were plenty of folks supporting shrub. You bumped into them all the time. They were your friends . It was a shocking experience how many intelligent folks who you thought ought to know better were going to vote for the Blivet, even after his sorry Texas record.
In 2000, very liberal Travis county where Austin resides, Horror of Horror, went red to the Shrub. This was because of the 10% or so Nader defection which was boosted by the huge amount of Nader Trades (Remember those motivated Dems who found Nader voters in other states who promised to vote for Gore in your calculations when you are thinking of that ridiculous 15% shift) added back in we still had more votes against blivet than for him but it was small consolation. My world was collapsing and I thought about moving...Canada maybe..

In 2004 there was none of this feeling. These former Bush supporters had come to their senses.. There were massive voter registrations and Dems retook travis county by about 50,000 votes..and increased the Dem vote by 71,000 This is NOT anecdotal. This was in spite of the straight Ticket democratic voting irregularities that were reported BTW... I also believe there was likely vote padding in texas but that remains in my clearly labeled belief category.

http://www.usatoday.com/news/politicselections/vote2004/countymap.htm
http://www.austinchronicle.com/issues/dispatch/2000-11-24/pols_feature4.html
We also sent folks from Texas all over the country to swing states doing voter registration and GOTV efforts. I was among the speakers at the first spontaneous Dean rally in March, there were 3000 people there. This was unprecedented this early in the campaign season for this kind of candidate. Folks were energized to defeat Bush. These people were the base of Deaniacs that Glen Maxey organized. We also sent folks from Texas all over the country to swing states doing voter registration and GOTV efforts. These were real people doing real work not that illusion of a grassroots campaign that Rove alleged got the shrub ohio. 15% forgetful shift?? I don't think so..Maybe about voting for Bush perhaps...

[OnTheOtherHand]

168. I'm not denying that those things happened

My county (Ulster in mid-Hudson NY) shifted away from Bush, too. Mostly demographic, I think, Democrats moving in from the south. But also some moderate Republicans who see Bush as totally irresponsible. There are 300 million people in the country, and some of them are Republicans who wince when they think about That Man being president.

If we are trying to assess fraud, I don't really see where these comments can lead. According to a plot of CBS/NYT and Gallup polls I am looking at here, right after 9/11 Bush had something like a 97% approval rating among Republicans and 80% among Democrats. Around election time, he had about a 15% approval rating among Democrats (in one October poll, it was 8%), which apparently is the lowest out-party approval rating in the history of the Gallup poll. But among Republicans, it was over 90%.

So, it's perfectly possible to have the biggest grass-roots movement against an incumbent in at least the last 50 years (I'm not saying I know whether it was), and still have the incumbent win. (I probably should have pointed that out to Chi.)

I don't think either of us is really in a position to assess the intensity of support for W., among people who do support him. I think we are predisposed to discount it or rationalize it away, and we need to be careful about that. As Febble said, tens of millions of people voted for Bush -- we are 'only' arguing about how many tens of millions. (Of course, I also think that a lot of Bush's votes came from people who aren't intense supporters at all.)

[Melissa G]

216. OTOH, you are not remotely comparing apples to apples..

I live in Texas, Shrub's home state..I am less than two hours from his ranch. (handy for protesting) I LIVE in Bush country.. You are talking about upstate NEW YORK! This is not remotely comparable information. Talk about blowing smoke!!!
Shifts in Bush support in Texas are a LOT more significant than shifts in NY. Bushies were noticeably turning on our homeboy here. Not implying all of them by any means but a significant shift over the previous election. Totally different political climate. You could check but I have no belief it was due to a demographic shift in our area. Our big pop shift was before either election. Voter reg efforts by dems were also a big factor here.

[OnTheOtherHand]

220. I think you are refuting a point I didn't make

Actually, I'm aware that Austin, Texas, is not a whole lot like Kingston or Woodstock or even New Paltz, New York.

Meanwhile, back in Brooklyn ("Kings County"), Bush got 90,355 votes or about 14.1% of the vote in 2000, and 159,200 votes or 22.9% of the vote in 2004. He lost Staten Island in 2000 with 41% of the vote, and won it in 2004 with 53% of the vote. He improved his vote percentage in every borough. (Kerry's overall vote margin in the five boroughs in 2004 was just a bit smaller than Gore's in 2000, around 1.2 million votes. But his percentage share was over six points lower.)

Could those numbers have been jiggered? Maybe, but I doubt it. Kerry officially won the state by 17.3%. The last four polls showed him winning by an average of 16.5%.

So I don't think we can safely reason from the climate of any one place to the climate of the whole country -- any more than we can infer from the fact that Gore lost his own home state in 2000 that he must have gotten destroyed in the popular vote.

[Bill Bored]

78. This premise is refuted by the exit polls, sort of.

Even the adjusted final version showed:

2000 Gore voters voted 90% for Kerry;
2000 Bush voters voted 91% for Bush.

Same in the 7:33 PM 11/2 version.

However, when you look at Party IDs, it's:

Final:
Dems 89% Kerry;
Repubs 93% Bush.
Or 11% Dem defectors and only 6% Repub defectors.

11/2 7:33 PM:
Dems 90% Kerry;
Repubs 92% Bush.
Or 9% Dem defectors and 7% Repub defectors.

Fewer Dem defectors before the final adjustments were made, but no change in the 2000 Gore voter defection rate between these two versions of the poll.

[anaxarchos]

102. You know how to play this game, BB... n/t

[Time for change]

88. How do you reconcile these NES numbers with 2004 raw exit polls

which results were much more in favor of Kerry?

[anaxarchos]

103. You too, TfC... n/t

[Time for change]

112. Seriously, I don't understand your comment to BB or me

I think this is a critical point. I think that the proposed scenario for OTOH's first serve, regarding proposed Bush/Kerry 2004 vote share for the Gore and Bush 2000 voters is not plausible, and I believe that that can be shown. But I need to understand his rationale for favoring the reference he cites over the exit polls.

[OnTheOtherHand]

116. Tfc, I'm not sure I understand your question

But maybe this is the answer:

My reading of the General Social Survey results is that false recall favoring the incumbent pretty reliably increases over time.

So I regard NES panel respondents' reported 2000 votes _after the 2000 election_ as a more (although not perfectly) reliable measure than exit poll respondents' reported 2000 votes four years later.

[Time for change]

135. But I don't think we're talking about "4 years later"

We alreday made the assumption, according to the rules you postulated in your initial thread. That Gore and Bush 2000 voters turned out in equal numbers. To quote you,

"I think we have agreed to assume that Gore2K and Bush2K voters turned out at the same percentage in 2004, whatever that percentage may be."

Now, given that, the exit poll question is, who did you just vote for. I believe that the raw exit poll numbers said that 7 or 8 percent of these voters said they voted for Bush. You're postulating twice that number.

I think I know what you might say to this, but I'll leave it at this for now, since I'm trying to save my energy for the next serve.

[OnTheOtherHand]

140. I swear, there is nothing up my sleeve here

What I am saying is that some of the Gore2K voters say they voted for Bush2K -- and, in fact, vice versa. In fact, some of the Gore2K voters say they voted for Bush in 2000 and Kerry in 2004. But more of them say that they voted for Bush both times.

And, therefore, the raw exit poll crosstabs are thrown off. Folks who actually voted for Gore in 2000 (but who presumably weren't paying much attention, most of them -- I'm not positing some weird Gore Shame Complex, just the same phenomenon that occurs to loser after loser*), and then voted for Bush in 2004, don't even remember voting for Gore in 2000, which depresses the defection rate.

*I've found two exceptions in the GSS. McGovern holds his own, because Nixon is _so_ unpopular. And Carter holds his own, not because Reagan is so unpopular (although at times he is), but most likely because he is a remarkably fine ex-president. But Ford, Mondale, Dukakis, Dole, Gore? All toast, of various degrees of crispiness. Probably someone has written a brilliant article about all this -- I just haven't looked.

The proportions are unknown, but the basic idea has a solid factual basis. You can download the NES 2000-04 panel data for yourself, and see the people who said in 2000 that they had voted for Gore, and said in 2004 that they had voted for Bush 2K (or vice versa) -- and see who they said they voted for in 2004. Obviously those people are Not Like Us.

[Time for change]

145. Then I think you should specify your scenario in more detail

In other words, for all categories of voters that you specified in your scenario, what percent said they voted for Gore, etc. Then we'll have a better idea of how plausible this sounds.

Sorry to put you through all this, but this was your serve.

[OnTheOtherHand]

170. let's see

I'm not sure what information you are looking for, but I
certainly didn't provide the DNV splits, so let's review --
and then, if you have further questions, I will tackle them.
Also, BB pointed out that it would be better if I used a
non-proportional font (ouch, doesn't that bring back painful
memories?).

Hypothesized votes by "actual" 2000 vote (assuming
94.0% Gore and Bush turnout and 97.0% "other"
turnout; fiddling with preference percentages ad lib) --

Gore 2000  46.275M    84.83% Kerry, 14.63% Bush
Bush 2000  45.782M     7.20% Kerry, 92.31% Bush
Other 2K    3.701M    65.90% Kerry, 18.10% Bush
DNV 2K     26.535M    52.90% Kerry, 46.50% Bush

(interim result: Bush wins by three million votes)

Kerry voters assumed to participate in the exit poll at a rate
12% higher than Bush and other voters. Probably this
propensity should vary across "type" of Kerry or
Bush voter, but I already have way too many parameters -- and
this is only a thought experiment, not a time machine.
(Actually, the parameter is 1.1164 -- this just happens to be
the number that yields a 3.00-point Kerry margin in the exit
poll, if the votes are constrained to the official totals.)

Now, we have the following retrospective error rates, and
hypothesized Kerry/Bush splits:

2000 actual/recalled   rate    Kerry    Bush
Gore -> "Bush"          7.3%   26.0%   65.0%
Bush -> "Gore"          1.6%   79.0%   21.0%
DNV -> "Gore"           7.0%   95.0%    5.0%
DNV -> "Bush"          11.0%   25.0%   75.0%
DNV -> "Other"          1.5%   60.0%   15.0%

And, as I reported before, I end up with the following
"results" for the unweighted exit poll, assuming
zero sampling error:

                       % of sample   Kerry     Bush
reported Gore2K voters    38.8%       90.3%    9.2%
reported Bush2K voters    40.2%        9.3%   90.3%
reported Oth2K voters      3.3%       69.2%   17.0%
reported DNV2K voters     17.6%       56.1%   43.2% 

What else?

[Time for change]

198. One major problem with this scenario is that it requires

actual Bush 2K voters who recalled voting for Bush in 2000 to give Bush a 107% share of their vote in 2004.

I don't doubt that this actually happened. However, I thought we were supposed to develop a scenario that doesn't include fraud.

[OnTheOtherHand]

206. smile (but confused smile)

I see 45.78 million actual Bush2K voters splitting 92.3/7.2 for Bush, and I see 1.6% of Bush2K voters forgetting they had voted for Bush2K, so I don't see how I am getting pushed up to 107%.

There could be a gaping hole in this scenario, but I don't see that one, yet.

[tommcintyre]

207. Ah... you two must be SO pleased your game is going so well

Maybe you ought to go celebrate?

[anaxarchos]

147. Please forgive me TfC and my rudeness in answering so late...

I don't know what to say because I think you are precisely correct in this post and those that follow (which in itself is a positive experience for me... pleasure to meet ya).

The "game" was an "creation". It came from many crying fraud and OTOH saying that there were other plausible explanations... The "game" also came from a rather primitive attempt to put some rules around the discussion so that one person could propose a scenario, others could attempt to refute and still others would defend. The idea, I think, was to prevent 15 scenarios from boiling around at once. OTOH may have confused us both by putting up a premise without a defense (i.e. what makes this "plausible"). I'm guessing that part is coming....

[TruthIsAll]

59. The rubber has it the road, OTOH.

So you have 26mm who did not vote in 2000?

Gore 2000 46.275M 84.83% Kerry, 14.63% Bush
Bush 2000 45.782M 7.20% Kerry, 92.31% Bush
Other 2K 3.701M 65.90% Kerry, 18.10% Bush
DNV 2K 26.535M 52.90% Kerry, 46.50% Bush

*************************************
You say this matched the Kerry margin.
OK, but... 
Why do you give Bush 40.20%?
His maximum voter turnout was 39.82%.
I thought we agreed on this. 

You are overshooting for Bush from the get-go.

Voted	Mix	Bush	Kerry	Other	Votes
DNV	17.70%	43.20%	56.10%	0.70%	21.64
Gore	38.80%	9.20%	90.30%	0.50%	47.44
Bush	40.20%	90.30%	9.30%	0.40%	49.15
Nader	3.30%	17.00%	69.20%	13.80%	4.03
	100%	48.08%	50.99%	0.93%	122.26
	122.26	58.78	62.34	1.14	
	Margin		3.56		
**********************************************************


Explain why you increase DNV in favor of Bush, and decrease
Kerry's share on top of that - even below that of the Final
Exit Poll.

****************************************************
Voted	Mix	Bush	Kerry	Other	Votes
DNV	21.90%	46.50%	52.90%	0.60%	26.77
Gore	37.50%	14.63%	84.83%	0.54%	45.85
Bush	37.50%	92.31%	7.20%	0.49%	45.85
Nader	3.10%	18.10%	65.90%	16.00%	3.79
	100%	50.85%	48.14%	1.01%	122.26
	122.26	62.17	58.86	1.24	
	Margin	3.31			

You increase DNV to 21.90% from 17% (21.5mm to 26.77mm).
And you claim that Kerry only got 52.90% of DNV.
So you're low-balling the Final Exit Poll, which already
low-balled Kerry's DNV share. The Final Exit Poll said 17%
were DNV, who went to Kerry by 54%-45%.

Finally, please explain how Bush won 14.83% of Gore voters,
while Kerry won only 7.20% of Bush voters.

What made the Gore voters (1 in 7) desert Kerry?
Did they forget it was stolen in 2000 from Gore?
Did they approve of the Bush job performance (he was at 48.5%
in the final approval polls on Nov. 1)

How come only 1 in 14 Bush voters deserted him in 2004?
Didn't they see Michael Moore's "Fahrenheit 9/11"? 

The rubber has it the road, OTOH.

[OnTheOtherHand]

70. let's see...

I don't give Bush 2000 voters an actual share of 40.2% of the electorate -- 45.782M / 122.293M is 37.4%, no? Reported Bush 2000 voters end up being a larger share than 37.4%, because of false recall.

Similarly (but in reverse), my actual share of DNVs is 21.7% (not 21.9%, I think), but reported DNV-2000s end up as only 17.6% of the unweighted exit poll sample, which is actually lower than the observed value of 18.4%. (This attrition is again due to false recall -- the recalled turnout rate goes up, so the reported DNVs go down.) That will take some tweaking. Part of the riddle is that I'm not at all sure how to allocate the DNV-2000s between new young voters (who wouldn't be represented in my panel study) and other people who were old enough to vote in 2000, but didn't.

Similarly, I don't give more (actual) DNVs to Kerry than the exit poll does. Rather, more DNVs falsely recall having voted for Bush than falsely recall having voted for Gore (etc. etc.), so Kerry ends up with a higher proportion of recalled DNVs than of actual DNVs. (But Kerry's margin in this group is in fact 0.7 points lower in my simulation than in the unweighted exit poll results, so that could use a further tweak -- although my results are easily within sampling error of an actual exit poll, and anyway, 0.7% of the DNVs are obviously not going to swing the popular vote.)

"How" did Bush win twice as many Gore voters as Kerry did Bush2K voters? How could I know? The general result is consistent with the 2000-04 panel survey, although we can try goosing down the specific percentages and seeing what happens -- I certainly don't see a Kerry victory in the cards. (If I use the NES figures of 12.4% and 6.3%, Bush wins by about 2 million votes instead of 3 million votes. Do bear in mind that my assignment was to come up with a scenario that fits the official totals. My figures don't seem to be 'impossible,' but I'm not wedded to them.)

However, Bush had the advantage of incumbency, lots of people had doubts about Kerry, and -- the point you seem to miss most often -- many Americans just aren't paying that much attention.

James Campbell points out in the latest PSQ that if one looks at the relationship between presidential approval in July and final incumbent two-party share, the break-even point (from a linear regression) is about 40.5% approval (Bush's, according to Campbell, was 47%). In 1980, Carter was at 21% approval in July, and still won 44.7% of the two-party vote. In 1992, Bush was at 31% in July, and still got 46.5%. Incumbents do a heck of a lot better than the approval-rate rule indicates. (The approval-rate rule isn't much of a rule, because we don't have all that many cases to look at.)

[TruthIsAll]

73. FALSE RECALL. FALSE RECALL. FALSE RECALL.

So that is your argument.
False recall.
Over and over and over again.

False hypothesis.

Calling Land Shark.

[OnTheOtherHand]

79. facts are stubborn things n/t

[Febble]

110. Are you disputing OTOH's facts?

Because they are facts in the sense that all survey data are facts, including the raw exit poll responses.

Or does "FALSE RECALL FALSE RECALL FALSE RECALL" translate as

"I'm not listening I'm not listening I'm not listening"?

[TruthIsAll]

117. NEWS FLASH! Census 60,000 sample (0.30% MoE): 125.7mm voted in 2004

That's right Febble, 3.5 million votes were never counted.

But, before you and OTOH naysay, let me say:

I realize that the 60,000 sample-size does not compare to the NES 600 sample.

I realize that the MoE as given by the Census Bureau is a ridiculously low 0.30%.

I realize that although most of these lost votes were Kerry votes (because only Gore voters lie or forget), it may very well be that a few, but only a few, (because Bush voters don't lie or forget) were Bush votes.

I realize that many of the 60,000 probably forgot that they didn't vote in 2004.

I realize that many of the 60,000 probably lied and didn't vote in 2004.

I realize that this is not compelling evidence.

I realize that the NES study is scientific.

I realize that the Census study is not scientific.

I realize that the Census 3.5mm differential from the recorded vote exceeds the MoE by a factor of 10. That is not compelling evidence.

I realize that this in and of itself is not indicative of fraud.

I realize that this is probably irrelevant to the discussion.

But it should still be placed in evidence.

*************************************************************

2-3 MILLION LOST KERRY VOTES? CENSUS: 125.7MM; RECORDED:122.3MM
http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x372831

CENSUS DATA MATCHED NATIONAL EXIT POLL WEIGHTS(12:22AM, 13047 SAMPLE)
http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x373187

Census/Exit Poll: Gender Vote Probability Analysis
http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x373414

[Febble]

121. Nice one, TIA. However....

Regarding MoE for the CPS numbers, this is from the Source and Accuracy Statement for the CPS survey on Voting and Registration in the November 2004 Election:

Accuracy of the estimates

A sample survey estimate has two types of error: sampling and nonsampling. The accuracy of an estimate depends on both types of error. The nature of the sampling error is known given the survey design; however, the full extent of the nonsampling error is unknown.

(my italics)

http://www.census.gov/population/socdemo/voting/cps2004/sa2004.pdf

From their report of the survey on the 2000 Election, in which the CPS estimate of turnout was greater than the count (as it was also in 2002):

Over reporting of voting in the CPS.

Some of the error in estimating turnout in the CPS is the result
of misreports, population controls, or survey coverage. Previous
analyses based on reinterviews of respondents showed that respondents
and proxy respondents are consistent with their reported
answers, and thus misunderstanding of the questions do not
account for the overestimate of voting in the CPS. However, other
studies that matched survey responses with voting records indicate
that a significant part of the discrepancy between survey estimates
and official counts is the result of respondent misreporting.
Incorrect reports of having voted from survey participants may be
due to a reluctance to admit being outside the mainstream of
American culture or the desire to exhibit a civic responsibility.

http://www.census.gov/prod/2002pubs/p20-542.pdf

Now, if the CPS's estimate of the vote is regularly larger than the counted vote you may interpret this as evidence of regular vote-embezzlement, to quote Kathy Dopp's apt phrase, rather than as, to quote CPS, evidence of a "desire to exhibit a civic responsibility". But the discrepancy certainly appears not to be unique to 2004. And bear in mind that the exit poll estimates in 2000 were fairly accurate (the WPE was small anyway).

[TruthIsAll]

119. Those are not facts. 122.26mm votes, 48.69 max Bush 2000 votes are FACTS

Stick to the facts.

You guys are grasping.
Taking a 600 sample survey (4.0% MoE) as a FACT.

Will you also accept as FACT the 60,000 U.S. census survey (0.30% MoE) finding that 125.7 mm actually voted in 2004?

Will you also accept as FACT that the Dems lose millions of votes (3%) in every election due to spoilage (innocent or otherwise)?

[Febble]

123. Survey data

are facts. However, inferential statistics based on survey data are not facts; they are in fact, inferences. Your 125.7 million is not a fact, it is an inference, just as OTOH's extrapolation is also an inference. Inferences based on FACTS.

I agree with you that one interpretation that can be made of the CPS data is that its apparent routine overstatement of the vote may atleast in part due to vote spoilage, which, I also agree, appears to differentially disenfranchise Democratic voters. I also agree with your implied point that this a disgrace to democracy, and have cited sources that make this point on many occasions both on DU and elsewhere.

However, the NES survey data are also FACTS and a legitimate inference from those facts is the inference that people forget, on a small, but nonetheless substantial (and statistically highly significant - did you get my probability estimate?) scale who they voted for at previous elections.

I'm all for facts. But let's not cherry-pick - and let's not ignore non-sampling error.

[TruthIsAll]

125. My "FACTS" are more believable than your "FACTS"

And, please, Febble, don't lecture about "cherry-picking".

So far, your case is based on one big bowl of cherries.

[Febble]

126. And my cherries are bigger than your cherries....

Come on, TIA, we are supposed to be playing nicely in the sandbox.

I just don't think you can ignore OTOH's evidence that people forget who they voted for. In fact I don't think you can ignore anything.

What you can do is to try to see whether the FACTS can be interpreted in a way that makes sense, and which is supported by precedent.

I think the the exit poll data can be accounted for, without invoking massive fraud. And, TfC and BB have convinced me that, in principle at least, massive fraud could still have occurred, but not shown up in the exit polls. That might account for the ESI finding in Ohio (that Bush's increase in voteshare in 2000 was not greater where the redshift was greater). Those findings are also inferences based on FACTS.

But in that case, the exit poll evidence itself becomes irrelevant. Which I think it is.

(edit of typo)

[TruthIsAll]

127. You would LOVE the exit poll evidence to be irrelevant, wouldn't you?

You and OTOH have been forced you to realize just how relevant the exit polls are. They are relevant enough to force you to come up with outrageous assumptions to fit the FACTS of the maximum possible 2000 voter turnout. And it was the Exit Poll "How Voted in 2000" question which has become the starting point in our search for the truth. Yet, there has been no appropriate recognition of these FACTS on the part of naysayers until now - although, in fairness you alone have in the past agreed that these FACTS make sense.

Now, OTOH and you have been forced kicking and screaming to finally accept these irrefutable FACTS in your analysis. You are the FIRST naysayers to finally, after 9 months, come up with a spreadsheet analysis. Let's face it Febble, you are now a confirmed naysayer, since you have just called the exit polls "irrelevant". If they were truly irelevant, you would not have to go through what must surely be a very difficult exercise in spin.

This is it. This is for all the marbles.

It's too bad that to fit the FACTS, you must resort to outrageous contortions that would make an Olympic gymnast proud.

Welcome to the world of spreadsheet bloggers.

[Febble]

128. No, you are wrong

regarding what I would "love".

What I would "love" is a reasonable debate as to what we now know about the exit poll evidence can (or can not) further the case for electoral reform.

[mgr]

138. There is a problem with your assumptions

Sorry for the late epiphany, I've been busy all day here at work. This may be obvious to some, but I have not seen it touched on. I guess the lesson from Florida has not fully sunk in.

122.26 million are not the total votes cast, but the total counted. This is not the same population the exit poll sampled in 2004; and the same applies to the population sampled in 2000. Assume a 5% undercount, and you have a pool of 132.48 million.

Now we can assume that those that performed their civic duty but did not have their vote counted can be irrelevant to the discussion if the numbers of votes to each party's presidential candidate were the same percentages over the four year period. However, HAVA was implemented in the interim, and given that new improvements to any public service begin in the wealthier places first, if HAVA compliant technology reduces the undervote, then a smaller percentage of republicans would see their votes as undervotes.

I made a rough calculation assuming all voters were returning with a two D to one R undervote for 2000 (which is liberal, my intuition is that it should be closer to 3:1), and assumed implementation of HAVA compliant in republican areas primarily, that halved the number of undervotes. The rough effect was to make the ratio 4:1. This would negate the assumption above.

What we also know about the undervote effect is that it is highly localized, and could be exacerbated by increases in voter turnout. I'm thinking Cuyahoga with a possible 180-300,000 increase in new voters, possibly losing more votes (percentage-wise) to this than in the past, and partly explaining why new voters did not go to Kerry as expected.

If we consider this, the zombies stay well within the ground, as the numerical pool of Bush 2000 and Gore 2000 voters increases, and the pool of Bush 2000 voters being counted in 2004 can also be greater.

Mike

[Bill Bored]

217. I like this one.

I'm not sure I believe it, but it shows thinking beyond the survey data which is what we could use more of!

It's certainly not inconsistent with fraud since the same HAVA compliant technology can be used to switch votes or generate undervotes.

[mgr]

223. Absolutely,

All HAVA assures us of is greater precision in the vote count, not greater accuracy. I'm not sure that I agree with it either.

The thing I am having difficulty with, and may be its that I need a few days and many beers, is how does one get the exit poll to popular vote stolen on the ground with specific hypotheses that can be tested? Does the proposal beast was in the east help? Or is it a red herring? Where specifically do we look?

I sounds from OTOH downstream that the story may not be in New York, so where?

Another observation, and its consistent with an article I read in Mad Cow Prod, is that what if the republicans are not in on this, and its only the corporations? Does this alter the expected patterns, and is this testable (the short answer is, yes--if we look at Florida and the gambling initiative, and see if there are patterns to both outcomes.)

Mike

[TruthIsAll]

75. CONGRATULATIONS OTOH. YOU HAVE JUST SHOWED US HOW TO MATCH THE BUSH VOTE.

Its a faith-based argument.
Simply, Gore voters voted for Bush.
Big time.
They lied to the exit pollsters.
Or they forgot that they voted for Gore.
Faulty recall.

I kneel before the god of faith-based analyses.
The god of implausibility.

We truly live in a Bizarro world when Gore voters, 1 out of 7 of them, voted for the idiot who stole it from them the last time.

Truly persuasive argument.
Not.

So, OTOH, we must except your numbers on faith.

I guess I shouldn't even bother to repost the probabilities of a 5 million discrepancy in New voters from all the exit poll timelines.

Or of Bush getting close to 15% of Gore voters.
Or of Kerry getting only 7% of Bush voters.
Or Kerry getting less than 53% of New voters.
Why bother?

Probabilities mean nothing.
Margin of error means nothing.
Scientific exit polling means nothing.
False recall is your answer to everything.

You have shot your wad.
You have nothing left.
You are bereft of any feasible, plausible argument.

You have finally come up with your analysis.
Its about time.
That's good.
I give you credit.
You tried your best.

You have put forth your best case.
And you must know in your heart of hearts that it's a joke.

Now your argument is out there for all to see.
Why proceed with this madness?

Game. Set. Match.
It's all over.
Why prolong the agony of defeat?

The numbers have always been your enemy.
And now you are found wanting.

You have nuthin'.
Your "false recall" model proves it.

I'm done with this charade.
Let others perform the burial.

[TruthIsAll]

81. One more thing. This is what Kerry needed to just TIE Bush

A piece of cake.
For those who live in the real world.

http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x379573#390129

[OnTheOtherHand]

82. I am astonished

TIA, you or anyone else on DU can download the NES panel data yourself, and see the basis of my figures.

I have no way to prevent you from claiming that the evidence doesn't exist, but it is out there for anyone to see. No one has to accept my numbers on faith. We don't even have to apply them mechanically to the entire election. But for heaven's sake, let's admit that there are data, shall we?

I'm astonished that you would argue that 26 out of 111 (the latter being the number of panel respondents who said in 2000 that they hadn't voted, and the former being the number of those who said in 2004 that they had voted for Bush in 2000) is not significantly different from 0.

I'm astonished that you would argue that 39 out of 302 (the latter being the number of panel respondents who said in 2000 that they had voted for Gore and who said in 2004 that they had voted in 2004, and the former being the number of those who said (in 2004) that they had just voted for Bush) -- a proportion that I recently understated (I forgot to leave out the 2004 non-voters), but which is about 12.9% -- clearly equals 20 out of 304 (the proportion of panel respondents who similarly defected from Bush2K to Kerry). I will leave the CHITEST as an exercise.

And I'm astonished that you repeatedly call my positions "faith-based."

[Time for change]

63. One thing I don't understand about each of your serves

The first serve appears to be against the rules of the game because under the "assumptions" of the game you say, "I think we have agreed to assume that Gore2K and Bush2K voters turned out at the same percentage in 2004, whatever that percentage may be." It seems to me that your serve breaks this assumption.

With regard to the second serve, it appears to be closer to being within the rules, but it also has Kerry winning. Which is like a double fault, right?

Or maybe I just don't understand the rules very well.

[Febble]

65. His serves are good

First serve: Gore2K and Bush2K have identical turnout, although "other" is higher (plausible - wouldn't you expect independent voters to be more politically aware?)

Second serve: Kerry is winning the poll because he has assumed "bias" in the "poll" of his assumed voters. So his table is a sim of the "unweighted" NEP table. If he can figure a way of weighting his table back to the actual vote, and reproducing the right kind of apparent Gore/Bush proportions, he gets the point.

I would say.

[Time for change]

68. I don't know what I was thinking about that first serve

What I MEANT to say was "Isn't it highly implausible that Bush gets 14.6% of Gore voters?"

As for the second serve, you attached a big "IF" to his getting the point. Seems to me that as it stands now, he loses.

[Febble]

71. No, it's still in play

he's allocated his actual votes, he's "polled" them with a biased poll, and he's produced something like the early "unweighted" table. A nice lob for the weighted table should get him the point.

[anaxarchos]

93. Don't count on it... n/t

[Chi]

120. One question...

Which do you think is more likely....
A) 15% Gore voters voted for Bush
B) Fraud was the reason Bush won

[Chi]

180. Chi lifts his arms and sniffs his armpits....couldnt be that. n/t

[Febble]

201. Just realised this was addressed to me

Don't have a view actually. It seems incomprehensible to me that Gore voters should vote for Bush, and on election night it looked as though Bush thought he was winning. So on those hunches alone, I'd say fraud was more likely.

It was my own a priori hypothesis, the reason I got into this swamp.

The trouble is that I think there really is good evidence that the exit poll discrepancy was primarily due to widespread fraud, although widespread fraud could still have occurred in a way that was invisible to the exit poll (TfC has suggested a mechanism, and voter suppression wouldn't show up either). I think the San Diego experiment will be worth watching - it's looking to me like evidence that even a parallel election is prone to a red shift. That recount is starting to look convincing.

So although, a priori, I would have said fraud was more likely, I am prepared to accept the possibility that a sizeable chunk of Gore voters voted for Bush. I don't know about the US but people here hate uncertainty, hate elections, and hate hung parliaments. Is it possible that a tranch of Gore voters thought, oh, have it your way, I'm not going through that Florida recount business again -?

I dunno.

[Bill Bored]

218. I don't think it's possible for Gore voters to have said"have it your way"

in 04, precisely because of FL 2000 and also because Gore himself, if anyone was paying attention, was a much better campaigner against Bush in 2004 than Kerry was! He was saying stuff like the voters are the only ones left in the Constitution who could hold Shrub accountable. That's powerful stuff! And he mentioned the Aug 6, 2001 PDB, which Kerry was too much of gentleman (or something?) to even bring up in polite company, or in Shrub's company either for that matter.

The exception might be Gore's home state of TN but this would be offset by Kerry's performance in MA.

One more thing: In the reweighting of the exit polls, there was no substantial difference in the Y2K how voted response. So when OTOH says he's replicated the unweighted exit poll, he's also replicated the weighted one, which Mitofsky barely found it necessary to adjust at all.

In other words, if Gore2K defectors is what gave Bush the victory, why weren't the exit polls reweighted to reflect this? Instead, they were reweighted to reflect more 2004 Dem defectors. Not the same adjustment.

[OnTheOtherHand]

227. a few things

As I think I said elsewhere on the thread, we have to be careful when we think about "Gore voters." Committed Gore voters probably wouldn't even have had to hear what Gore was saying in 2004 in order to figure out that they wanted to vote for Kerry in overwhelming numbers.

But the sort of voter who would tell someone in 2000 that s/he just voted for Gore, and tell someone in 2004 that actually s/he had voted for Bush, is not all that committed. And there are probably some others who weren't very committed, either.

"In the reweighting of the exit polls, there was no substantial difference in the Y2K how voted response."

I don't agree with that (if I understand it), but let me take the opportunity to clarify a few points. First of all, the unweighted results I was trying to match were the percentages from the national data file, available on line, and quite literally unweighted. Those percentages won't match the "12:22 13047" numbers (although they should be pretty close), because all along the national numbers were being weighted for geography and demographic non-response adjustments. Apparently the official totals weren't factored into the national numbers until the "1:25 13660" numbers.

The 13047 numbers apparently showed 39% Gore, 41% Bush. The totally unweighted numbers I was using showed 38.4% Gore, 39.5% Bush. My scenario has 38.8% Gore, 40.2% Bush, which is somewhere between. Maybe a one-point difference, maybe more like a two-point difference.

The weighted results have Gore 36.7%, Bush 43.0% -- over a six-point difference. That shift alone is probably enough to swing the exit poll from Kerry to Bush, barely.

Bear in mind that there are always two kinds of "Gore2K defectors" (i.e., Gore2K voters who voted for Bush in 2004) -- people who remember having voted for Gore, and people who don't. According to the scenario, there are about half as many of the second kind as of the first kind. Given our puzzlement that any Gore voter could vote for Bush, maybe it shouldn't seem so strange that about a third of them wouldn't remember having voted for Gore in the first place.

[anaxarchos]

74. When I said two serves, it was a figure of speech...

The serve was "Bush gets 14.6% of Gore voters, while Kerry gets 7.2% of Bush2k voters". The NES panel study was the evidence to make the 14.6% "plausible". The reconciliation of the exit poll was a "flourish" (a way to say "this scenario ALSO explains the exit poll discrepancy") that adds to the plausibility story ( even though explaining that is not necessarily an objective of the "game"). The point is just starting (not at all clear who won or lost yet).

BTW, when you are comfortable, pick up a racket and play from either side. If you don't buy "plausibility" (or do) take a WHACK at it....

[Chi]

97. Let me get this straight...

First time voters broke for Kerry
Didn't vote in 2k broke for Kerry
Independents ... broke for Kerry
Vote regstrations broke for Kerry

But 15% of Gore voters broke for Bush?

All with the Pres. job approval rating sub 50%.....
How is that likely, or credible.

I would think a study, or at least a previous election with similar results
would be required to make that possible, never mind likely or credible.
But maybe that's just me.(shrug)

[OnTheOtherHand]

130. well, the game is a little bit stacked right now

The NES panel cross-tab seems to indicate that all the things you say are true (not literally "15%," but twice as many Gore00->Bush04 as in the other direction). Mostly, I seem to be hearing over and over again that That Just Can't Be Right. Shrug.

We know there is, on average, a big incumbency advantage, and it seems sensible that much of it would be accomplished through "defections" rather than new voters. It also seems sensible that, since folks here have presented plenty of evidence of false recall, that should be taken into account in interpreting crosstabs. But I'm not sure what would count as a convincing case for you.

Actually, I thought the idea of the game was for us to start by working on a scenario together -- not for me to present the scenario and for lots of people to hoot at it (grin).

I think people are thrown off by the category "Gore voters." I think presidential approval ratings were more polarized along party lines in 2004 than they have ever been, so it makes sense that the "typical" Gore voter dislikes Bush. But in every election, a bunch of people are voting who don't really care all that much. That's why the outcomes vary so much. Counterintuitively to most Democrats, the economic indicators in 2004 favored Bush's reelection chances.

I've just posted about the "job approval" issue. Counterintuitive as it may be, the data don't really support the argument. See via http://inside.bard.edu/~lindeman or in my other post.

[Chi]

133. OK...let's see if I can be more constructive....

If you theorize that 15% of Gore voters went for Bush, I would figure..

that somewhere in the vicinity of 24% of Dems, who voted in 2k, voted for Bush in 2004.
(This is giving the benefit of doubt to your scenario, and splitting the 15% evenly across the board for Gore voters)
This comes from 14% of Dems who voted for Bush2k, and an additional 10% by percentage according to your scenario.
(If you don't like my math...try it yourself and see what you get.)

Does this seem likely or reasonable to you?

I would think everyone would agree this is not likely, which means you/we need to lower that number.

(And BTW, Hey there 8) and welcome back)

[OnTheOtherHand]

136. fair question, and thanks

for taking my snitlet in good humor.

But I don't quite grasp the premise of your argument, and I don't agree with the conclusion, and I'm not sure how to clear things up concisely. Part of the problem is that party ID is even fuzzier than recalled vote choice. People's party IDs slosh around (mostly people moving in and out of the "independent" category -- and I imagine that there is slosh depending on the context, as well as the question wording). From the NES panel, it appears that the Democrats lost some ground to the Republicans mostly as a result of former "I"s becoming "R"s -- but the evidence is weak, and the concept is fuzzy, and besides, this conversation is too depressing already.

That said, I'm not sure where you come up with 14% of Democrats who voted for Bush in 2000 -- and those Democrats aren't compelled to vote for Bush again (we should expect Bush2K defectors to be disproportionately Democratic, especially since Bush has an strikingly low out-party approval rating) -- and one wouldn't expect proportionately many of the Gore2K->Bush04 defectors to be Democrats (and even those who were, might just leave the Dems, too).

Overall, the Democrats-for-Bush percentage may be higher, it may not, but it wouldn't be anywhere near 24%. The weighted exits put it at 11% in both years. The NES, I think I'll get different results depending on which weights I use, but I don't think I'll get higher than that.

[TruthIsAll]

141. To you, every statistic "sloshes" around except your 600-sample NES.

"But I don't quite grasp the premise of your argument, and I don't agree with the conclusion, and I'm not sure how to clear things up concisely. Part of the problem is that party ID is even fuzzier than recalled vote choice. People's party IDs slosh around (mostly people moving in and out of the "independent" category -- and I imagine that there is slosh depending on the context, as well as the question wording). From the NES panel, it appears that the Democrats lost some ground to the Republicans mostly as a result of former "I"s becoming "R"s -- but the evidence is weak, and the concept is fuzzy, and besides, this conversation is too depressing already".

This is a classic case of logic thrown on its head to support an untenable argument. This is truly Bush/Rovespeak.

Kerry won the female vote big time - 54-46, just as Clinton and Gore did.

The Party-Id split stayed constant at 38/35/27. It was 39/35/26 in the prior three elections.

There were 21 million new voters - overwhelming for Kerry.

There were 3 million Nader voters - overwhelming for Kerry.

There was widespread disenchantment with Bush from moderate Republicans.

There was widespread hatred of Bush on the part of Gore voters who saw the election stolen from them in broad daylight.

There was widespread hatred of Bush on the part of minorities and others who saw 170,000 punch cards spoiled (double-punched and under-punched) in FL 2000.

AND TO TOP IT OFF, BUSH LOST THE LAST TIME BY 540,000 VOTES (OFFICIAL). BUT UNOFFICIALLY, IT WAS BY MUCH MORE THAN THAT. OVER 3 MILLION VOTES WERE SPOILED IN HEAVILY DEMOCRATIC PRECINCTS.

You reject all of this and you low-ball the final exit poll which matched the Bush vote and blame it all on false recall, based on an obscure 600 sample. ANAXARCHOS got you on that one.

From now on, TOTAL RECALL on the part of DUers will remember the sheer arrogance of your argument.

You are reduced to reiterating your relentless mind-boggling rationale.

Unfortunately, the 43/37 Bush/Gore split was your Waterloo.

[OnTheOtherHand]

172. actually, I think every survey statistic is somewhat sloshy

whereas you simply cite as Proven Fact that "Kerry won the female vote big time - 54-46" (and so on and so on). But it is not proven fact.

You are cherry-picking, relentlessly, seemingly without even noticing it.

Whatever else one says about the NES, it is hardly "obscure" among survey researchers. Anax did quite a bit of impressive handwaving there, but the panel still offers explicit evidence of actual people actually changing their answers about whom they had voted for. If you are not willing to admit that, then you aren't really engaging the debate.

Then there are all those examples from the GSS, which I don't think you have ever commented on. I guess I will post them again.

Your latest post again implies that you are oblivious to incumbency advantage.

[Chi]

176. I can see how the party affiliation numbers can be muddy...

so let's just back up, and forget I said that, it will go nowhere.

My contention has to do with the excessive numbers of Gore voters who voted for Bush04.

What did you base that number on?

If it's just a convenient fit, thats not a reasonable method to build a likely scenario.

[OnTheOtherHand]

185. OK, Gore for Bush04

First of all, I think we have very little basis for insisting on what the number should be. I accept TIA's point that Bush could only have beaten Kerry if this %age is substantially larger than the %age of Bush2K voters for Kerry. (There is wiggle room based on turnout and other parameters, but I don't see how it would affect the basic result.) I also accept his conclusion that neither value is uniquely determined.

I think we can infer, from the closeness of pre-election polls -- and upon the assumption that "other" and "new" voters favored Kerry -- that in fact Bush was likely attracting more Gore2K voters than Kerry was attracting Bush2K voters. Either that, or the pre-election polls were biased in favor of Bush, which I guess is more or less what TIA ends up arguing. And many people agreed with him at the time, although not to the extent of estimating a 99% probability of a Kerry win. Still, it begins to seem to me that some folks are insisting upon taking the exit poll very literally, and taking the pre-election polls rather figuratively, and junking other inconvenient survey research entirely. This is not a mainstream approach among survey analysts of my acquaintance, and it does not compel assent.

The NES crosstab indicates, in fact, that the rate of Gore->Bush defections was about twice the rate of Bush->Kerry defections (the former rate is 12.9%, the latter is 6.6%). I do not take this figure literally, but I am not at all impressed by the insistence that it is obviously much too high. My first thought is that it could well be too low, because people who participate in the NES are probably stimulated to think harder about political issues than people who don't, and therefore might be disposed to express more stable views than others. But I'm not at all sure about that, and there could be a counterargument -- and besides, this is a small sample. It does seem to me that people who insist that the Bush->Kerry defection rate must be at least as high as the Gore->Bush defection rate are, well, somewhat lacking in an authoritative basis for that insistence -- beyond John Kerry's own visceral intuition, "I can't believe that I am losing to this idiot."

The false recall and conditional vote decision rates in my model are inferred from NES data as thoughtfully as I could manage. (Bear in mind also that in the 2002 GSS, Bush retrospectively beats Gore by 5.9 points. It's not as if this false recall business is clearly rebutted by available evidence.) I am sure these specific rates are wrong, but they are not just a "convenient fit." My specific use of a 14.6% defection rate is indeed a convenient fit, although I should note that (1) it could be counterbalanced by reducing the Bush->Kerry defection rate, which I think would also be supported from data, and (2) we will have to drop the Gore->Bush defection rate quite substantially before Kerry wins. (Of course I am not saying that we will have to drop it to an "impossible" figure.)

I haven't thought hard about your "historical precedent" q in part because, in the survey biz, we don't really have all that much history to fall back on. I'm open to ideas about that -- what races might be relevant? However, I believe that the extent of in-party/out-party polarization about this president may be unprecedented at least in the survey era (as I pointed out to Melissa G, apparently no president has ever had a lower out-party approval rating than W does), and this probably throws off our intuitions. In fact, someone is about to say, "Then how on earth could all those Gore voters have voted for Bush?" Well, some were never Dems, some have become Reps or Inds, and some are in that unusually small yet substantial Dem minority that say they approve of Bush's performance. Even amidst the polarization, TIA had an argument that hinged on allocating 5% undecideds -- something like six million folks who, at least by hypothesis, just couldn't decide who they liked best. Go figure.

[Chi]

191. Best info still says 14.6% too high, IMHO.

"First of all, I think we have very little basis for insisting on what the number should be."

I agree

To me, there seems to be two main sources of info on this catagory...
switched- to Bush, to Kerry
NES panel .. 12.9 .to 6.6
04 Exit poll... 8 ....to 10

But you chose to exceed the highest figure, instead of falling in between.
Even though you appear to agree, the exit polls would likely be more accurate.
I still fail to see how that achieves a likely scenario.


A question.
Is there a way to see the NES panel data without having SAS, SPSS, or Stata?
I'd like to see if there are any correlations between those who misremembered, and their attitudes and/or monetary incentives.

(edit to change exit poll numbers to 13047 results)

[OnTheOtherHand]

209. it's open to interpretation

I think the exit poll number is quite misleading, if I am right about the possible magnitude of false recall -- but there is plenty of room for further analysis on that. (I would like to look at some older exit polls, at some point. Maybe the bug will catch you?)

From my current POV, the panel number is sort of the only number, and I didn't mess with it _too_ much -- but I was pledged to match the exact result without allowing any difference in turnout between Bush 2K and Gore 2K voters (and I chose not to tinker with a lot of other parameters, because I knew I would be in enough trouble presenting this scenario as it was). I do think it would be hard to argue that the 14.6% figure is downright implausible because the panel number is 12.9 -- unless there is a reason why the panel number is biased upwards, which could be the case.

Umm, there ought to be some way to import the ASCII data into R -- or maybe you can even use an Excel filter, gack. Many variations on that theme. Depends on what you like to use. Meanwhile, feel free to ask any specifics I should run. (Please PM to be sure -- I may be intermittent on the board.) Come to think of it, I may even be able to supply you with your data format of choice. You would still need to mess with the codebook -- filter out missing values so you don't get garbage.

[tommcintyre]

211. And some things aren't open to interpretation.

They are pretty clear, aren't they? Hey... real nice "game" OTOH.

[autorank]

58. We interrupt our normal programming for this SPECIAL ANNOUNCEMENT

We interrupt our normal programming for this SPECIAL ANNOUNCEMENT

8/22/05

Bush

Approve 36%
Disapprove 58%
Undecided 6%

http://americanresearchgroup.com/

You know what this means, the "f" word as in election FRAUD is now accepteable for use by CM (Ccorporate Media)

[TruthIsAll]

61. AND AN MOE CALCULATOR, TO BOOT!

http://americanresearchgroup.com/sams.html

ENTER 100000000 AS THE POPULATION SIZE (100 million)

SEE IF THE RESULTS AGREE WITH THIS FORMULA:
MoE =1.96*SQRT(0.5*0.5/n)

for n:

n MoE
9603 1.00%
2401 2.00%
1067 3.00%
600 4.00%

[anaxarchos]

69. Wow... n/t

[Bill Bored]

72. I wonder which way those undecideds will break. nt

[autorank]

83. ==>ATTENTION -- RULING FROM THE JUDGES--IT'S OFFICIAL

"False memory" isn't a fact,it's a discredited theory put forward in
entirely different field. Take my word on just this one thing, you DO NOT want to associated with that theory or those people.

You're pushing awfully hard to come up with a Bush victory. That's the way this game has played out. Especially the DNV's -- up around 5 million over the exit's even, lord. Wuz up, you trying to get an ambassadorship? spot in the administration? date with Anne Coulter?

I'm not close to being persuaded so I think the game is still on. You don't have a credible scenario + you are using the "false memory" theory which automatically disqualifies you from winning any scenario in which you use it.

RULING FROM THE JUDGES: Begin Again OTOH!

This ruling is final. There no appeals. Begin Again OTOH.

[Dynasty_At_Passes]

86. I think you can rule false memory out.

It would have to take place on a large scale, and it doesn't seem to me false memory has been a factor on such a scale in any previous election.

I still see it happened, because there is always a certain degree of false memory with older voters. However, it did not happen everywhere. So what did happen everywhere? We need that analyzed and examined.

[autorank]

87. Right, and it's just memory from the voting booth to the door.

It's not like asking me next year who I voted for in the Supervisor's race last year, it's who did you just cast your ballot for. Memory degreades over time, but the difference between T1 (vote) and T2 (exit poll intervies) is, what, minutes.

Silly point OTOH, not even a good bluff. PENALTY, points substracted, BEGIN AGAIN.

[TruthIsAll]

95. Are we done yet?

Relentless acronyms.

eKv Early Kerry Voter.
lBv Late Bush Voter.
oFv Overpolled Female Voter.
dEp Democratic Exit Pollsters.
iEp Inexperienced Exit Pollsters.
yEp Young Exit Pollsters.
hMe High Margin of Error.
kCe Kerry Cluster Effect.
kDe Kerry Design Effect
bBe Bush Bandwagon effect.
mFt Massive Fundie Turnout.
rBr Reluctant Bush Responder.
eKr Exuberant Kerry Responder.
lGv Lying Gore Voter.

fGr Faulty Gore Recall.

What's next?

[Febble]

108. Do pay attention at the back, there

No-one (in this game, anyway) is arguing that people forgot who they'd voted for five minutes back. The argument is that a proportion of voters may have forgotten who they voted for four years back.

So no, it's not like asking you who you voted for last year. It's four times as long, and as you rightly say, auto, memory degrades over time. Using actual data from actual longitidudinal studies, OTOH has extrapolated data-backed numbers for this phenomena.

And it appears that, using these numbers, plus the postulated exit poll bias, he can reproduce the unweighted exit poll table, assuming, as per the rules, no zombie voters, and equal turnout of Bush and Gore voters.

No penalty.

Play on!

[autorank]

154. Cite some evidence for MAJOR DECISIONS.

This is a single event...not trivial, happens on a regular basis, you remember who you voted for and change it only when you're with inlaws or friends whose tedious fulminations will be touched off with if you say you voted for someone sane.

This is not like the stuff of memory experiments tracking material from college courses or events that are not important.

Who did marry the first time you were married?

Who did you vote for the first time you voted?

Who did you ... well, we'll leave that one out...

But you get the drift.

The presumption is that a single event that doesn't happen that frequently but one in which many participate is easy to remember. I don't have to prove a negative, I'm asserting the obvious.

Give us some data on major events that people forget at a !significant! frequency.

How silly, duh, I don't remember who I voted for...

In the absence of real evidence to support the specific claim that people forget significant events, not that frequent, that are shared by milions OTOH will retire from the field and colate his list of "400 political scientists/Sonambulists."

;)

[Febble]

160. Define "real" evidence

OTOH has cited evidence. Perhaps it can be refuted. But I haven't seen it refuted yet, on this thread.

Re: "significant": to you, to me, to people on DU voting is deadly serious, deadly significant (literally). To many it is a chore, one they will even try to duck out of. Politics turns a lot of people off. In Australia, where voting is (was?) compulsorily, there used to be something called the "donkey vote" whereby people, confronted with a list of candidates, would vote for the one at the top of the list. They had to rotate the ballot to stop that happening. Some people just don't care that much.


But also re "significant": yes, people do, amazingly, misremember significant events. Dammit, if they didn't, most women wouldn't have a second child.

And then there's this:

We met at nine. We met at eight.
I was on time. No, you were late.
Ah yes! I remember it well.

We dined with friends. We dined alone.
A tenor sang. A baritone.
Ah yes! I remember it well.

That dazzling April moon! There was none that night, And the month was June.
That's right! That's right!
It warms my heart to know that
you remember still the way you do.
Ah yes! I remember it well.

How often I've thought of that Friday, Monday night,
when we had our last rendez-vous.
And somehow I've foolishly wondered
if you might by some chance be
thinking of it too?

That carriage ride. You walked me home.
You lost a glove. I lost a comb.
Ah yes! I remember it well.

That brilliant sky. We had some rain.
Those Russian songs. From sunny Spain.
Ah yes! I remember it well.

You wore a gown of gold. I was all in blue.
Am I getting old? Oh no! Not you!
How strong you were,
how young and gay;
A prince of love in every way.
Ah yes! I remember it well.

http://er.neoxer.com/lyrics/gigi.html

Remember?

[autorank]

190. Evidence

Evidence that people forget who they married, who the first worked for, who they voted for in major elections. "Forget" -- not misreport. I'm alone, there's a piece of paper in front of me, I know that my answer will be confidential, the question is: Who didyou vote for in 1972...no problem for me or anybody else, except a few. Why? It's a major event. No evidence has shown up to indicate that for major events, not frequent , and shared by many (e.g., a presidential election), no evidence that a significant number of people forget.

As for exit polls which are the subject around here, that's easy and you won't find any evidence to contradict this: I vote, I walk outside, I'm approached and asked to participate, I way yes. How hard is it for me to remember if I volunteer. Not hard at all. Case closed.

[Febble]

195. OK, then what is your problem

with the NES survey as evidence, as cited by OTOH here?:

http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=379573&mesg_id=390057

In the 2000-04 NES panel study, among the folks who reported a 2000 vote in both the 2000 postelection and the 2004 postelection survey:

Among the 308 who had voted for Gore as of 2000, 29 of them said in 2004 that they had voted for someone else in 2000 (23 Bush, 3 Nader, 2 other, 1 DK -- you can skip that one).

Among the 309 who had voted for Bush as of 2000, 6 of them said in 2004 that they had voted for someone else in 2000 (5 Gore, 1 Nader).

I make it that 6% of participants in the survey misrecalled (we do not know whether they forgot or deliberately misreported) their 2000 vote when asked about it in 2004. 9.4% of Gore voters misrecalled their vote, 1.9% of Bush voters.

i.e. not zero for either group. So the next question is: is the proportion of Gore voters who misrecalled their vote significantly higher than the proportion of Bush2000 voters? The two relevant statistical tests for this question are chi square, and Fisher's Exact test. Chi square is easy, and is almost certainly accurate for a sample this large, but I happen to have a program that will give me Fisher's Exact Test, so I did that.

Fisher's Exact Test gives a probability of 1 in 27,740 that the difference could have occurred by chance.

So we can probably conclude two things: yes people can forget who they voted for four years after the event (because out 6% of the survey participants did so - no inferential statistics required) and that Gore voters were significantly more likely than Bush voters to forget who they voted for in 2000 when asked in 2004.

This looks like evidence to me. There may be reasons why it is not valid evidence (but I have yet to see such a thing). It is certainly not invalidated by the sample size which provides plenty of statistical power, as is evident from the prrobability figure.

[StefanX]

91. I'm an amateur mathematician

But I'm too tired to play this game in all its detail tonight.

I would just simplify it as follows:

(1) Bush lost the popular vote in 2000 - but won in a "squeaker" with the aid of disenfranchisement, SOTUS, and Nader.

(2) In 2004, Dems registered in droves, most Naderites became more pragmatic and realized they didn't want to throw their vote away, and exit polls showed Kerry winning.

So Bush did WORSE in 2004 than he did in 2000. And yet he won. Ergo, he cheated.

I assume this is the same result you get when you crunch the actual numbers.

[TruthIsAll]

94. Touche. It really IS that simple. Simple math. Pure logic. Common sense.

As a fellow applied mathematician, who HAS crunched the numbers, I welcome your succinct input.

It's been a long haul at DU since Nov 2004. This is the FIRST time that the other side has presented a statistical analysis to show how Bush managed to defy the odds. You are witness to the result of their efforts.

We forced them to provide their analysis after we had posted literally hundreds of analytical models which overwhelmingly made the case for fraud. Finally, after interminable verbiage, they agreed to walk the walk - with real numbers.

The essence of their argument requires us to believe that Gore voters not only deserted Kerry for Bush, but that many of them forgot that they voted for Gore or lied about it- and that's why the exit polls were off. Prior to this latest hypothesis, they said the discrepancy was due to reluctance on the part of Bush voters to be exit polled. Now they are left with this.

To believe their case requires a leap of faith that would betray everything we have learned and experienced in a lifetime devoted to developing mathematical models. And the models were designed to solve problems much more complicated than this.

This problem can be solved on the back of an envelope. You just did it.

[StefanX]

100. As an amateur mathematician and a computer programmer

I don't think we should be using computers for elections.

All we need is the following:

(1) Large lucite ballot boxes at each polling station.

(2) Paper ballots and indelible ink markers.

(3) Video cameras.

(4) A live audience - members of the public hanging around observing.


This low-tech voting system would work as follows:

(1) You mark your ballot.

(2) You fold it up and put it in the box.

(3) Anyone can hang around the polling station and "observe" until polling ends.

(4) At the end of polling, ballots are pulled one by one out of the lucite boxes, held up to the camera, and duly tallied. (Think Lotto drawings on TV. A little longer and slower but still could be done in a few hours.)

(5) At the end of this process, you have a LOCAL total for that polling station, which is duly certified and announced.

(6) Each local precinct total is totalled nationally.


See why this works? Because:

(1) Observers watch the vote casting and the totalling, so there's no opportunity for tampering.

(2) The polling station has to certify and announce its total LOCALLY - WITHOUT getting in cahoots with any of the other polling stations.
Yes, this would prevent any more "razor-thin" last-minute victories by Republicans (which usually happen when all the small districts have closed and they save the biggest Republican-leaning districts for last so they have a larger pool of votes to hide their fudging)... but sorry, George, that's the way democracy works.

(3) There's no invisible bits and bytes and modems and memory cards and touch-screens and chips and RAM and hard drives and machine code and motherboards and trapdoors and easter eggs and ActiveX controls and computer bugs and Visual Basic code and virtual memory and duplicate database tables and OLEDB access and memory modules and proprietary operating systems and coding errors and compilers and p-code interpreters and whatever other bullshit to mystify the process and turn it into incomprehensible mumbo-jumbo for 99% of the population. Remember, elections are guilty until proven innocent.


It would only take a few hours to get a total for a given polling station. From that point, it would only take a few MINUTES to get a national total.

===

As a computer programmer, I think computers should be banned from vote casting and particularly from vote totalling.

===

Just think about how Lotto balls are picked - they have an attractive woman on TV in front of a live audience, very publicly pulling balls out of a tumbler. Of course, Lotto is important, so we make sure there can't be any cheating.

[autorank]

104. Get the beautiful girls to count the votes. Lots of witnesses then.

[autorank]

101. Please return and return often.

Thanks, we needed that!

[OnTheOtherHand]

115. it depends on the value of "you"

For one thing, Bush was the incumbent in 2004. Historically, that is typically worth 3 to 4 extra points in two-party share (up to 8 points in margin) right there.

http://fairmodel.econ.yale.edu/RAYFAIR/PDF/2002DHTM.HTM

See coefficients for "PERSON."

[TruthIsAll]

118. Bush was an incumbent with a 48.5% job rating. Envelope, please?

This will fit on the back of an envelope.

FACT: Challenger's traditionally win 2/3 of the undecided vote.

Assume:
They were tied in the final polls: 47-47%-1%.
Five percent were undecided.

Allocate:
Let's allocate the undecided:

Kerry: 47 +2/3*5 = 50.33 (very conservative)
Bush: 47 +1/3*5 = 48.67

OTOH, you are grasping.

[OnTheOtherHand]

129. considering some discrepant information

(1) I've heard a lot about the horrors of having a 48.5% job rating, although I haven't seen much serious analysis of the point. Inspired by a recent article by James E. Campbell, I've taken the liberty of scatterplotting incumbents' final pre-election Gallup approval ratings against their ultimate vote margins (positive or negative). There is a lot of noise in the analysis, because the dates of those polls varies a lot -- but actually, based on this variable alone, Bush would be predicted to win by over five points.

http://inside.bard.edu/~lindeman/ (index page)


(2) The assertion that challengers "traditionally" win 2/3 of the undecided vote is so back-of-the-envelope that one wonders what to say. What is this, Fiddler on the Roof?

But, since Mystery Pollster seems to be momentarily again in your good graces, we might try the numbers he posts:
http://www.mysterypollster.com/main/2004/11/lessons_likely_.html

10/27-11/1: Bush 48.5%, Kerry 47.0%, Nader 0.9% (undecided not reported, but apparently 3.6% or less).

So, even if we accept your "fact" and give Kerry 2/3 of that 3.6%, Bush still beats Kerry. We will have to stipulate at least one more "fact" before these polls somehow point unerringly to a Kerry win.

(3) Just for fun, people might be interested in comparing some pre-election models by other experts:

http://www.apsanet.org/imgtest/Into-2004Forecasts-Campbell.pdf (seven models)

http://fairmodel.econ.yale.edu/vote2004/vot1104.htm

[TruthIsAll]

131. I suggest you look at this..and get back to espousing "false recall"

I have already stated why I used registered voters where available and likely voters if not.

You chose to cite all "likely voter" polls.
Bottom line: I was right. You were wrong.

Now about those 15% Gore voters who voted for Bush...
Any more magic tricks up your sleeve?

[OnTheOtherHand]

132. circular reasoning 101

Whether you were right is the subject of the game, not the assumption of the game. Doh.

You still lifted a finger to refute my evidence -- you just sputter about how it's obviously wrong. Or, in this case, you mostly just ignore it entirely.

[anaxarchos]

134. Return of Service....

You have asserted that "Bush won because of a dramatic red-shift of Gore voters from 2000, at approximately twice the rate of blue-shift in the other direction".

Your premise was "Bush gets 14.6% of Gore voters, while Kerry gets 7.2% of Bush2k voters".

Your only supporting evidence was the 2004 NES panel study.

I believe your “supporting evidence” in no way supports your premise:

1) There is inherently no reason to believe that a poll of 1100 “eligible voters” conducted after the election is more reliable than a poll of 13000 actual voters conducted on the day of the election. If I understand the claim, this is the reason that exit polls are “an advance”, i.e. that they are able to reliably eliminate response error inherent in the questions “Did you vote?” and “Who did you vote for?”. I am not talking about the retrospective question (yet). As ONE supporting piece of evidence, I cite:

Robert H. Prisuta, A Post-Election Bandwagon Effect? - Comparing National Exit Poll Data with a General Population Survey, SRMS/ASA 1993

available here:

http://www.amstat.org/sections/srms/Proceedings/papers/...

This is as exact a match to your argument as you are ever likely to find in research and Prisuta’s findings imply the exact opposite of your claim. I tried to bring this up with Febble (I guessed where you were going) but she chose to blow me off:

Available here (post #126):

http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x379573#389726

2) There are both extrinsic and intrinsic reasons to regard the NES as inherently unreliable on this issue. This response error is the oldest, most persistent, and largest in the NES, from its inception.

I cite as my authority on this...................... You.

(and, luckily, one hundred assorted other, fucking boring, NES Technical Reports, Research Summaries, Voter Verification Reports and other odds and ends of research that no one in prison should be forced to read).

You have been trying to tell everyone who will listen (and many who won’t) about “false memory”, etc.. By golly, you’re right. And how exactly did it go away when you decided to use it to reset the vote and the exit polls?

We can go down this very deep rat hole if you want to but I think it will be the end of the game and of my ability to stay awake. Why would we believe that you are not simply re-sampling the response error that is already stipulated?

I don’t believe anyone at the NES would support using the 2004 Panel Study in the way you have just done. But, hey, what do I know? Prove me wrong….

3) I believe you are misusing the research (an outrageous comment, I know, from an interloper like me). At the least, you are being very free with what part of your claims is actually “supported by a preponderance of research”, what part seems to you to be supported by any research at all, and what part is only according to you. In fact, I don’t remember you ever citing ANY basic research. You cite numbers, you say what they mean, and then you move on. Febble is actually worse. I ask her for a reference and she cites a source that says “people sometimes have false memories” … No shit! This is quite a catch-all. I can hardly quibble if you give me diddle. No wonder TIA says “False Memory, False Memory, False Memory”.

For starters, I think that “false memory” is largely a misnomer in this case or perhaps an artifact of previous research. I know that some researchers think that it is. If so, they claim a highly selective, “respondent-specific”, and difficult to quantify (unlike “Did you vote”) “false memory”, which applies under certain conditions. And as far as the NES goes, I know that efforts to fix this response error as “false memory” have largely failed. I cite as evidence:

Belli, Robert, Santa Traugott, and Steven J. Rosenstone. 1994. Reducing Over-Reporting of Voter Turnout: An Experiment using a ‘Source Monitoring’ Framework, NES Technical Report Series, No. nes010153.

Available here:

ftp://ftp.nes.isr.umich.edu/ftp/nes/bibliography/documents/nes010153.pdf

On the other hand (couldn’t resist) there is, what appears to me, to be the preponderance of thinking. That view attributes the response error to “the social pressure to vote”, particularly among those who were predisposed to vote but didn’t. This is indirectly implied even in the UK poll citations that Febble sent me. In those, the terms “misremembrance” and “lie” are used interchangeably, a dead giveaway as to whether this response error is a sin of “omission” or a sin of “commission” . More germane is the following citation which claims to have reduced the response error in the NES by nearly half by essentially letting people off the hook on the social pressure:

Duff, Brian, Michael J. Hanmer, Won-ho Park, and Ismail K. White. 2004. How Good is This Excuse?: Correction of Overreporting of Voter Turnout Bias in the 2002 American National Election Study, NES Technical Report Series, No. nes010872.

Available here:

ftp://ftp.nes.isr.umich.edu/ftp/nes/bibliography/documents/nes010872.pdf

If I understand the implications of this last, then how exactly does “false memory” even apply in exit poll “retrospective” questions? Exactly the same? Not at all? Cite your research.

I think the era of “false memory” has to end. I think you have to cite specific research, its implication, etc. The current path is guaranteed to insure disbelief (coming directly after rBr, etc.) as well as to confuse us as to when you are speaking for yourself and when you speak for “the dead” (i.e. “authority”).

Believe it or not, I say all of this point #3 without rancor. This is the web and not a journal…. But, you get my point.

One last issue: You just told TfC : “My reading of the General Social Survey results is that false recall favoring the incumbent pretty reliably increases over time.”

That disagrees directly with Presser who claims the exact opposite:

Stanley Presser, "Can Context Changes Reduce Vote Over-reporting?," Public Opinion Quarterly, Winter, 1990.

You might mention when you are being “controversial”.

4) I believe there is a logical fallacy in your presentation. If your premise were true, THEN one explanation of the exit poll retrospective question could be “false memory”. But this “explanation” actually drives the initial premise. You are pulling from the ass end of the donkey (sorry, couldn’t resist). The initial premise could have been cherry picked to produce this result or it could have come from your astrologer. Ultimately, this is a circular argument, with each end ultimately relying on the other. The donkey has its head up its ass.

To prevent this, you have to present independent, empirical evidence

I am not protesting but don’t blame me if the earth swallows you up.

5) I think many of the responses that you received, immediately went to the heart of the matter. BillBored went straight to the exit polls… “look at the party split”, etc. Chi asked the obvious historical question… “when has a red-shift like that ever happened when everything else went blue”. MelissaG said: “Wait a second. I thought you said you had evidence”. TfC was, in some ways, the best of all: “But, but… that’s not “plausible”…”. All of this was premature because you hadn’t made a case yet. But it IS the reason I hang around this board.


I claim first point by right of “reason” and “virtue” (unless you protest)……

Luv-15…

What else ya got?

[OnTheOtherHand]

143. whoa there

(1) I never claimed that the NES was more accurate as an election measure than the exit poll is. It is clearly subject to the bandwagon effect; it is probably less subject to non-response bias, since they work so darn hard at getting the response rates up. But either way, it wasn't part of my argument.

(2) I have almost no idea what you are saying here. At the end of the day, the NES shows -- very blatantly in the panel study -- people retrospectively changing their minds about who they voted for, and so does the GSS, and so do.... It's one thing to say that the NES numbers can't be relied upon, but it's ridiculous to say that therefore we should believe that 2004 exit poll respondents are accurately reporting their 2000 votes. And if you aren't saying that, then I'm not sure what you are contributing.

We know an awful lot about problems with the NES, because it is good enough to reward careful research into its problems. The exits -- well, the exits have the virtue of immediacy, but they have problems, too.

(3) I have to admit that I'm not really all that interested in the theory of false memory. I am interested in the phenomenon, but not even very interested in that except as it helps us to interpret the 2004 exit poll results. I am also happy to stipulate that "pretty reliably increases" is hyperbole -- I was overly astonished by the Dukakis and Dole cases. Nixon did not do especially well in the vote recall sweepstakes, and Carter did pretty well against Reagan.

Nevertheless, I have no idea why you would believe that exit polls are immune from false retrospective vote reporting, while other surveys are not. Is that obvious to everyone but me? or what part of the argument am I missing? Have you actually taken a position on whether retrospective vote reporting in the exits is likely to be accurate, or are you just busy raising doubts about the NES?

I think it's fairly silly to say that if my arguments here aren't up to APSR standards, then you win. Dammit, this is not what I am supposed to be doing with my time.

(4) The basis of the false reporting premise is studies other than the exit polls. Silly me, I thought that would make the premise more plausible.

(5) I am happy to respond to individual arguments individually. I haven't found any of the responses persuasive at all. Many people here seem to Know that it was somehow impossible for Bush to win this election, but strangely they haven't managed to convince a whole lot of political scientists. Well, here I am. Convince me.

[TruthIsAll]

144. Just who the hell are these mysterious "political scientists" you keep

referring to?

The only ones I know of (on the Net and MSM) were IMMEDIATELY quite dismissive of the notion, even the possibility, that the election could have been stolen.

Their core argument was simply this: The exit polls were obviously bogus, because they deviated from the vote count.

How's that for an argument? Is this the way to encourage independent, scientific analysis of the election?

Are those the political "scientists" you are talking about? Well, they don't know a damn thing about the scientific method.

[autorank]

148. "There are 400 of them, I have a list of their names in my brief case.

I don't need to show them to you, but I can assure you that they are real card carrying political scientists, all 400 hundred." (para.)

Where's OTOH, taking a "standing 8 count" after losing the graph war and the Anaxarchos left-right-head butt combination.

And OTOH, what about this, which you ignored: "Febble* is actually worse. I ask her for a reference and she cites a source that says “people sometimes have false memories” … No shit! This is quite a catch-all. I can hardly quibble if you give me diddle. No wonder TIA says “False Memory, False Memory, False Memory”." Are you going to let him talk that way about Febble? I'm shocked!

Right now, I'd prefer a boxing metaphor but tennis will do.

It's Anaxarchos in the in-his-prime, Jimmy Conners role and, you, OTOH as Ricardo Mantalban (who can barely play tennis but speaks well and is "nice").

OTOH, I tried to warn you off the "false memory" business, it's a total loser. That arises out of ... well, we won't go there ... it's a loser. The term you may be looking for is confabulation:

"Confabulation is a memory disorder that may occur in patients who have sustained damage to both the basal forebrain and the frontal lobes, as after an aneurysm of the anterior communicating artery. Confabulation is defined as the spontaneous production of false memories: either memories for events which never occurred, or memories of actual events which are displaced in space or time. These memories may be elaborate and detailed. Some may be obviously bizarre, as a memory of a ride in an alien spaceship; others are quite mundane, as a memory of having eggs for breakfast, so that only a close family member can confirm that the memory is in fact false."

This may resurrect the voting Zombie model because they certainly have brain damage.

How about a Sonambulist model? A Dr. Calagari figure (fill in appropriate political villain) hypnotizes scads of Gore voters who obey his twin commands: vote for Bush, act really stupid when you're asked about it so we can denigrate the polls. These are my two commandments!

Maybe the 400 political scientists are Sonambulists!

Somebody throw a racket. Oh, wait, they threw an election in 2004 and that's their racket.

-----
*Febble, do not take this personally. It's "trash talk" which is allowed in this game.

[Febble]

162. No probs, auto

(I'm even growing quite fond of Harvey).

But see my response to anax below and to you elsewhere.

[OnTheOtherHand]

164. actually...

There are quite a lot of political scientists who never gave the issue a moment's thought, so it is fine to dismiss them. (I have no idea of the proportions.)

A number of us -- again, I have no idea what the number is -- started poking around to see what evidence of fraud might hold up. It is ill-considered to generalize that political scientists just assumed that the vote count must be right.

One problem has been that academics who work with survey data -- probably even more than professional pollsters -- spend a great deal of time examining all the things that can and do go wrong with surveys. (Perhaps ironically, anax drew on this fact in his criticisms of the NES, just as Febble did in her response to your point about the CPS.) For this reason, these academics were not predisposed to believe that the exit polls wre strong evidence of fraud. To the extent that advocates of the fraud hypothesis have emphasized the exit polls, while spending less time developing other lines of argument, they have tended to convince academics that "the arguments" for fraud aren't very interesting.

I so far haven't found any of the arguments for massive fraud very compelling, but I can still hope.

[TruthIsAll]

150. OTOH, how does your model deal with these categories?

What about all these other groupings?
What did they forget or lie about?

National Exit Poll

Time Respondents
7:33pm 11027 
12:22am 13047 ""
1:25pm  13660 ""

        Bush	Kerry	Nader

We discussed this one...

Voted in 2000			
11027	46.96%	50.25%	1.03%
13047	47.62%	51.41%	0.97%
13660	51.11%	48.48%	0.41%

VOTED 2000

	Mix	Bush	Kerry	Nader
No	17%	41%	57%	2%
Gore	39%	8%	91%	1%
Bush	41%	90%	10%	0%
Other	3%	21%	71%	8%
	100%	47.62%	51.41%	0.97%
	122.26	58.22	62.85	1.19
______________________________________

Now would you recall how you analyzed these?

Gender			
11027	47.76%	50.78%	1.00%
13047	48.22%	50.78%	1.00%
13660	51.22%	47.78%	0.00%
			
Education			
11027	48.18%	50.34%	1.48%
13047	48.05%	50.21%	1.17%
13660	51.24%	47.82%	0.42%
			
Race/Gender			
11027	47.96%	51.04%	1.00%
13047	47.86%	50.94%	1.00%
13660	50.99%	47.81%	0.23%
			
Age			
11027	48.17%	50.53%	1.30%
13047	48.17%	50.53%	1.00%
13660	51.28%	47.96%	0.29%
			
Income			
11027	47.54%	51.01%	1.45%
13047	47.54%	51.01%	0.94%
13660	51.02%	48.13%	0.11%
			
			
Ideology			
11027	48.60%	50.07%	1.00%
13047	48.15%	49.85%	1.00%
13660	50.83%	47.95%	0.22%
			
			
Religion			
11027	47.90%	50.85%	1.25%
13047	47.90%	50.85%	1.18%
13660	50.77%	47.99%	0.17%
			
Military			
11027	47.62%	51.20%	1.18%
13047	47.62%	51.20%	1.00%
13660	50.44%	48.38%	1.00%
			
When Decided			
11027	48.04%	51.42%	0.54%
13047	47.95%	51.23%	0.54%
13660	51.22%	47.50%	0.19%
			
Region			
11027	47.73%	50.84%	1.00%
13047	47.95%	50.84%	1.00%
13660	51.08%	48.24%	0.42%
			
			
Party ID			
11027	48.24%	50.24%	1.26%
13047	47.77%	50.69%	0.92%
13660	49.69%	49.15%	0.27%
			
			
Average			
11027	47.89%	50.71%	1.12%
13047	47.90%	50.80%	0.98%
13660	50.91%	48.10%	0.31%

______________________________________

Nov. 3, 12:22am, 13047 respondents				

GENDER	Mix	Bush	Kerry	Nader
Male 	46%	52%	47%	1%
Fem 	54%	45%	54%	1%
	100%	48.22%	50.78%	1.00%
					
______________________________________
EDUCATION				
	Mix	Bush	Kerry	Nader
NoHS	4%	47%	52%	1%
HSGr 	22%	48%	51%	1%
Col 	31%	51%	47%	1%
ColGr	26%	50%	48%	1%
Postg	17%	40%	58%	2%
	100%	48.05%	50.21%	1.17%
					
				
______________________________________
RACE AND GENDER
	Mix	Bush	Kerry	Nader
WM 	36%	58%	41%	1%
WF 	41%	52%	47%	1%
NWM 	10%	28%	69%	1%
NWF 	13%	22%	77%	1%
	100%	47.86%	50.94%	1.00%
					
______________________________________
AGE	Mix	Bush	Kerry	Nader
18-29	17%	43%	56%	1%
30-44	27%	50%	49%	1%
45-59	30%	47%	51%	1%
60+ 	26%	51%	48%	1%
	100%	48.17%	50.53%	1.00%
	
______________________________________
INCOME	Mix	Bush	Kerry	Nader
0-15K 	9%	33%	66%	1%
15-30 	15%	39%	59%	1%
30-50 	22%	47%	52%	1%
50-75 	23%	53%	45%	1%
75-100  13%	50%	49%	0%
100-150 11%	53%	45%	2%
150-200 4%	53%	47%	0%
 200+ 	3%	58%	41%	1%
      	100%	47.54%	51.01%	0.94%
					
______________________________________
IDEOL	Mix	Bush	Kerry	Nader
Lib 	22%	12%	86%	1%
Mod 	45%	41%	57%	1%
Cons 	33%	82%	16%	1%
	100%	48.15%	49.85%	1.00%
	
______________________________________
RELIG	Mix	Bush	Kerry	Nader
Prot	53%	56%	43%	1%
Cath 	27%	49%	50%	1%
Jew	3%	23%	77%	0%
Other	7%	20%	75%	4%
None 	10%	29%	70%	1%
	100%	47.90%	50.85%	1.18%
					
______________________________________
MILIT	Mix	Bush	Kerry	Nader
Yes 	18%	55%	43%	1%
No 	82%	46%	53%	1%
	100%	47.62%	51.20%	1.00%
	
______________________________________
DECIDED	
         Mix	Bush	Kerry	Nader
Today 	6%	40%	53%	5%
3Days 	3%	41%	53%	4%
Week   2%	51%	48%	1%
Month 	10%	38%	60%	1%
Before79%	50%	50%	0%
	100%	47.95%	51.23%	0.54%
				
				
______________________________________
REG  Mix	Bush	Kerry	Nader
East	22%	41%	58%	1%
Midw	26%	49%	50%	1%
South	31%	54%	45%	1%
West	21%	45%	53%	1%
	100%	47.95%	50.84%	1.00%
				
______________________________________
PARTY	Mix	Bush	Kerry	Nader
Dem 	38%	9%	90%	1%
Rep 	35%	92%	7%	0%
Ind	27%	45%	52%	2%
	100%	47.77%	50.69%	0.92%

					

[OnTheOtherHand]

165. sorry, is there a question here?

Basic line of analysis: the exit polls diverged from the official count; the exit polls were weighted to match the official count; we are looking to see whether anything in the weighted results is screamingly unlikely -- and, if so, whether that evinces fraud, or just an artifact of weighting.

The 43/37 recalled-vote split seemed screamingly unlikely, but I don't think it is, because of the extensive evidence from multiple sources of misremembered votes retrospectively favoring incumbents in general (although not in all cases -- e.g. Nixon) and W. in particular. It was interesting, I learned something from it, and I hope that someday you will too.

Is there something else in here that you consider screamingly unlikely?

[TruthIsAll]

174. I'm glad you learned something from 43/37, professor.

It took quite a while for it to sink in, though. It must have been my relentless pounding until you finally got it.

And yes, there are 10 new questions here.

Now that we have your theory concerning the forgetfulness of Gore voters, what are your THEORIES about Kerry voters - did they also forget that they were women, or that they were black, or a college grad, or a liberal, or that they made up their mind at the very end, or that they live in the East?


You can proceed one at a time to provide an analogous argument to the category: "How Gore Voters Forgot Who They Voted For in 2000".

[OnTheOtherHand]

179. seems disingenuous to me

Since you give absolutely no credence to my interpretation of 43/37 -- and still don't have any substantive response to my evidence for my interpretation -- it seems odd to take credit for it.

As for the rest, it should be pretty obvious to readers that there are many unarticulated assumptions. On what basis do you infer that I must believe that Kerry voters forgot that they were women? What does that even mean?

[anaxarchos]

200. I will answer this directly (soon)... Real life inervenes.... n/t

[anaxarchos]

204. I'm going to have to suspend my participation...

...temporarily or longer. Sorry to leave you hanging but you weren't winning anyway ;-)

Two problems:

1) There is a tropical storm which might become a hurricane headed for Florida. Ordinarily, this is not much concern for us (it's a little one), but since this one is headed straight for the chair I am sitting in, I have to take a few precautions. Hope to be back in 24 hours.

2) It seems that TIA got banned. If that is the case, then I'll be moving along as well. In such an eventuality, feel free to declare yourself "winner" and then you can talk to mgr...

[Febble]

159. Below the belt, anax....

I conceded in my response to you that it wasn't a great cite, but you'd asked for some indication of how the term "consistency bias" was used in the field of psychology, and rather than cite my (long packed) undergrad psychology texts, I did a quick Google.

A full literature search would take rather longer, but as you seem more expert than I assumed, I'll leave you to do it yourself. However, the concept that memories are constructed and reconstructed rather than laid down as permanent but degradable layers is fairly widely accepted in the field of cognitive psychology as I am sure you are aware. My husband in fact is a schizophrenia researcher, and has investigated the role of the hippocampus in gating memory - and in distinguishing between semantic and episodic memory. His hypothesis is that aberrant hippocampal firing in schizophrenia leads to confusion between the two, and for internally generated thoughts to escape validation and thus be construed as external reality.

However, I am NOT suggesting that UK (or US) voters are schizophrenic (although that might account for 1%). I am suggesting that it is perfectly plausible that people construct their memories as to how they voted in line with their current narrative. In fact I believed for a while that my first vote ever was for Labour as I have been a Labour Party member most of my life. I was reminded recently that it was actually for David Steel, one-time leader of the Liberal Party (and responsible for the UK equivalent of Roe v.Wade, the 1967 abortion act). OTOH's data seems to confirm that people misremember their vote; similarly the CPS study I linked to in a response to TIA notes that a proportion of people recall voting when examination of voting records indicates that they did not.

In other words, while "reported vote" responses are facts in themselves, they are only facts in the sense that they are probably accurate records of who people said they had voted for. There may be considerable slop in the degree to which they map on to actual votes in 2000. Or there may not be. But we are talking plausible here. None of these things are actually knowable. However, the NES data has the virtue of being from a within-subjects longitidinal study - it is likely therefore that the evidence it presents that a proportion of people misreport their vote when asked some years later is good evidence.

I know there was a rule that said something about the first person to say "the vote count is right because it's the official count" loses. I'd like to suggest a complementary rule that says:

The first person to say "Bush couldn't have won because no Gore voter in their right mind would have voted for Bush" loses.

I think one intrinsic problem here is that no-one who posts on DU can conceive of a) anyone voting for Bush and b) anyone not caring much about who is president. But, like it or not, tens of millions of Americans voted for Bush (we are only arguing about how many tens) - and millions of Americans can't be bothered to vote at all.

Weird, but true.

[TruthIsAll]

171. An observation from a very biased poster on this thread..

Posters on this thread apparently fall into one of two groups:

1) Fraud conspiracists
2) Naysayers

Group I: 12 individuals
DU members who have worked diligently since Nov. 2, using publicly available but media-ignored data related to Election 2004. Each member has analyzed the data using his/her particular area of expertise.

Group II: 2 individuals
Both worked at USCV and left early last year, immediately proceeding to bash the statistical-based study they helped write which debunked the rBr hypothesis. At least one individual immediately engaged in a "consulting" assignment for Mitofsky, the originator of rBr. Both are currently collaborating with a well-known fundie, banned from DU, who throws stones at any argument pertaining to fraud. Each of them often refers to MP as an polling expert, although his "expertise" appears to derive from his claim of being an unbiased Democrat. Each has apparently deserted the discredited rBr hypothesis and cite as the basis of their case an NES study in order to promote a new theory to explain the exit poll discrepancies - the Forgetful Gore Responder (fGr). To justify the numbers, they have made assumptions regarding Gore/Bush 2000 voter percentages in 2004. The percentage deviations from the Final Exit Poll percentages are nearly triple the margin of error for this voting characteristic, considering that it is a 10/90% split.

Both groups are obviously biased.

[Febble]

177. Two small corrections:

First:

Re Mystery Pollster, aka Mark Blumenthal, is a partner in the polling firm Bennett, Petts and Blumenthal.

http://mysterypollster.typepad.com/about.html

A person who is a partner in a polling firm is generally considered to have polling expertise.

Second: I don't think either OTOH or I have "deserted rBr", although rBr is not my favored description of the hypothesis. My own view is that some form of sampling/response bias was probably a substantial contributor to the exit poll discrepancy, although I do not rule out a contribution from fraud.

The theory put forward in this thread by OTOH (and by myself, earlier) is simply a hypothesis regarding a mechanism by which the "final table" with its apparently "impossible" Gore/Bush proportions could be reconciled with a red shift that was largely due to sampling/response bias.

Otherwise, for myself, I would confirm your timeline, although not the implied causal sequence. And because I dispute the causal sequence I dispute the allegation of bias. I changed my views because I was unbiased (i.e. not committed to a fraud hypothesis come-what-may) not because I was biased. And certainly not because I did work for Mitofsky. The timeline was, as you correctly report, the other way round. My paper resulted in my leaving USCV; my paper subsequently induced Mitofsky firstly to reanalyse the data, and later to commission further analyses from me.

So yes, in your words (elsewhere), I am a NAYSAYER to the case that the exit poll evidence is unambiguous evidence of massive electoral fraud. However I am not by any means a naysayer to efforts to investigate the inequities and illegalities of the 2004 election or to bring about fair, secure, auditable elections in the US. It is why I got involved in the first place.

On edit: just noticed your reference to the religious beliefs of one of my colleagues. This is quite irrelevant and quite unacceptable. I suggest you edit this if there is time.

[tommcintyre]

214. The term "fundie" has been used 7048 topics (GD) since 11-02-04

Now, you may not like the term (and I don't use it personally); but "7048 topics match your criteria" when I just did an advanced search from 11-02-04, to now in the General Discussion forum for "fundie".

It is commonly used here to describe a "right-wing religious zealot" who is biased (and intolerant) against all, but a very narrow set of beliefs. Of course, there are some christian fundamentalists who don't fit this description; but the aforementioned attitude is so prevalent among them - hence the term.

Now I don't know Rick Brady's personal religious beliefs; but if MY term accurately describes him, his obvious right-wing bias would make him a very poor choice for participation in ANYTHING where non-partisanship would be necessary (such as co-authoring a paper on the 2004 election exit polls).

And, if this is an accurate description of him, it would explain my observations of his vitriolic and mean-spirited attitude - especially toward DU, and it's members.
http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x390193#390813

Finally, again, assuming this is an accurate description of him, how could you expect us here (or anyone else who is reasonably fair-minded) to even consider anything he has participated in (such as the paper you plan to co-author with him)?

[OnTheOtherHand]

189. whoops, we forgot the rule against ad hominems

Folks, please be cautioned -- although I suppose this is already obvious:

Several DU regulars have expressed vigorous dissent from some of TIA's arguments (and gotten hammered for it). And Febble and I are a lot less predictable than the "naysayer" epithet implies. The attempt to divide us into 12 versus 2 is misguided.

The scope of TIA's attacks is at least wide-ranging ("they don't know a damn thing about the scientific method" -- #144), so one does not feel singled out.

[mgr]

199. It's a 2X2 square, not a syllogism

Bishop Ussher(TIA)'s being a little less than honest here: the two categories are not linked by a syllogism, and it does not follow that there are 12 ayesayers to 2 naysayers by the criteria offered. It is the intent of how it was put to confuse, and for others to make the intended connection, but to not say it; and to deny it later. Very Rovian.

There are four logical divisions here, not two: DU ayesayer, DU naysayer, non-DU ayesayer, non-DU nay sayer. Where Febble and OTOH fall within these four categories, only criteria they match is non-DU.

I obviously fall into DU, and in how Bishop Ussher measures the universe, am in the naysayer category.

But I would like to know the 12 he considers the DU quants, or was I among that 12?

Mike

edited to include appropriate logical connection not

edit2 apparently brackets make edit hidden text

[Calamity Jayne]

167. "A" "game" about the 2004 Election

THE DEMOCRATS HAVE NO CANDIDATE

Well let's hope the Swedes are watching us closely. Could
someone please persuade them to invade our country...starting
with Los Alamos? It's in New Mexico. Appreciate. We need help.

Hillary 4 Prez: Even one more year of the Bill and Hill show
would be cause to actually move to Canada. His kingdom for a
b-job. They need to stop subjecting us to the hick version of
the War of the Roses.

John Kerry: Too bad. What does time do to a courageous Winter
Soldier. Basically he comes across now as a pontificating bore
with the delivery of a Greek Oracle bowing and scraping to
wealth of all types even if he has to suck on a ketchup
bottle. I still think he's cute though.

Al Bore

Joe Biden: Yadda yadda yadda yadda yadda yadda, grin, "I
like you" yadda yadda yadda, big grin, "You're the
real deal" big grin, kneel, bow, scrape and so on and so
on. By the way did you know that all the censored dead kids in
flag draped caskets come back through Dover Delaware-his
state. )

Start looking for the real thing. A taller Howard Dean with
the ideas of a Kucinich. The "super star" idea of an
Arnie wasn't a bad idea but they picked such an ubber-aas. It
worked for awhile though. He IS gov of Kaliforneeia.

How about running Chomsky or Lew Lapham? Amy Goodman for Prez.
Ha ha ha ha ha ha. Can you see Chomsky or Lapham hamming it up
in Iowa with all those pig farmers?

Problem is we need an objective brain in there cause our
survival depends on it.

All we have to choose from are screaming political bores
who've never demonstrated the expertise in world affairs nor
the courage to deserve it... That's why the coup succeeded in
the first place. But keep hope alive. There has to be someone
in a country of 300 million.

Richard Clarke for President? Would he subject himself to the
Dicksville Dorks and the Iowa jello and macaroni salad events
in intellectual Iowa?  

Threaten your Senator with expulsion for the reason of your
choosing.

 

	

[autorank]

193. RIP Calamity

[Chi]

194. And she was such a promising star of DU.

The Clinton venom is always a dead giveaway.
BUH-BYE
:nuke:

[anaxarchos]

196. Good... they are showing up.

Someone is bringing lampshades. Now the real party can begin.

[Chi]

197. I think we need a 2 syllable limit to attract more 8) n/t

[Bill Bored]

202. HALF TIME SHOW! Take a break and take this poll!

<http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x390853>

[Chi]

203. UMMM,.. Anyone know what TIA did to get Tomb stoned?

I see a couple of his posts deleted in Land Shark's thread, but nothing else.

[autorank]

205. As far as I'm concerned, this game is over period as is other Bull Shit

I'm tired of the antics and the silliness.

An election was stolen. If people want to come around and play 20 questins, or ask the professor, or "lets be nice" that's fine but I'm not participating in any more garbage.

[mgr]

208. Then why this post? n/t

[tommcintyre]

210. You are truly "amazing" n/t

[Chi]

212. Very classy comment

[mgr]

226. Not sure where you are going with this

But from someone that gloated over getting Qw... tombstoned, I would not throw stones, or start looking for someone being responsible as Tomm seems to do upslope. There is more important work, and if you believe that someone is bigger than the rules because of their contribution; then someone needs to consider getting the rules changed, or their behavior excepted. Then the pigs can run the farm. All can look to themselves (and I am not addressing just yourself in this), that supported his arguments as having failed to advise him to tone it down. It was that obvious it was coming.

If you value his contributions, then help to move them forward. If what you are starting is a witch hunt, leave it at home.

Mike

[Chi]

229. Hmmm, I don't recall kicking someone in the face...

who was upset by Q getting tombstoned.

It's your comment Mike, wear it with pride.

Looks good on you mate! :thumbsup:

(edit- for those who wish to see the extent of what Mike refers to as my 'gloating', here it is)
http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=355131&mesg_id=355238

[autorank]

213. .

So you know how I really feel.

[mgr]

222. Your expressions of contempt aside,

This exercise was proposed by those who argue that exit polls are indicative of widespread fraud that stole the popular vote. If you have nothing of substance to say, or think that action needs doing now, focus your effort on that. Many of us see the value in this discussion--it is an attempt to provide a plausible explanation, within the framework of the exit poll, for pattern and behavior within the electorate allowed Bush to win.

What I don't think you get is that the potential this discussion has to allow the development of specific and testable hypotheses as to how the popular vote may have been hijacked. Unless you can develop something from the 'beast is in the east', but I have yet to see anything applicable from that thesis.

By the way, that is a good article on the Hackett election.

Mike

[mgr]

224. 2. I had not realized that you posted upstream

That makes your comment more contemptable. Unless you think this thread was some kind of flame fest concocted by naysayers, which it was not (I don't see any deleted messages, but I could have overlooked one or two).

Honestly, if you think debate is over, and action needs taking, then do so, and stop the cheerleading when you think your side is winning a debate you have so little respect for. Unfortunately, you cannot take the ball and go home.

Mike

[Name removed]

225. Deleted message

Message removed by moderator. Click here to review the message board rules.

[Chi]

228. I sent an email to the Administrators last night about TIA...

(EDITED TO COMPLY WITH THE RULES!)
I simply asked, what was the reason TIA got banned.
It was very short and without any attitude.

I have yet to receive a response.

I (and everyone else I heard from) can find no posts that were deleted from TIA, the reason for his banning is a mystery.

If I receive a response, I will let everyone know.
If I don't receive a response by this evening, I will let everyone know that as well.

Chi

[EarlG]

230. I'm locking this

since it's turned into a flamewar.

As for TruthIsAll, we don't normally discuss why members are banned, so you will have to contact him directly. I should point out that he knows exactly why he was banned, and he was given plenty of opportunities to correct his behavior beforehand.