An Open Letter to Farhad Manjoo

Dear Farhad,
In your No Exit

piece (6/15/2005) you linked to this summary

of my DU posts.

I should have responded to "No Exit" a year ago. But since you have continued your crusade against "fraudsters" with your thrashing of RFK's Rolling Stone article, now it's my turn. The pathetic rebuttal to RFK came exactly one year after "No Exit". in which you claim that "expert" sources debunked USCV and the "fraudsters". Quite the contrary; they're down to their last-ditch hypothesis ("false recall"). They are still trying to spread the fiction that the pre-election and early exit polls were wrong; that Bush won the popular vote by a 3 million "mandate". In so doing, they imply that the FINAL National Exit and state exit polls, which were matched to the "recorded" vote, are correct. Based on the mathematical proof below, you will see why the FINAL NEP is IMPOSSIBLE and OVERSTATES the Bush vote by AT LEAST 4 MILLION.

Here's a more complete, up-to-date, compilation of my pre-election state and national models, which forecast the Kerry 2-party vote to be 51.80% and 51.63%, respectively. Also included are over 100 pre-election and exit poll analytic threads. If you read the replies, you'll see how the arguments of responding naysayers and trolls were debunked virtually every time.
Election Model

THE MARGIN OF ERROR
You make the fallacious statement that there is no real significance to the 3% deviation from Kerry winning by 51-48% in the National Exit poll to the Bush recorded 51-48% win. That statement alone exposed your lack of understanding of the "margin of error" and mathematical probability. The 1.1% MoE (assuming a 30% cluster effect for 13,047 respondents) means that the 3% deviation is extremely significant. It's almost 6 standard deviations from the sample mean.
Let's calculate the probability of the deviation:
Prob = 1-NORMDIST(0.51,0.48,0.011/1.96,TRUE) =2.05809E-09 or 1 in 485,887,838

THE REAL EXPERTS
It's apparent to many that your analysis is full of holes. You say that an "amateur mathematician" proved Ron Baiman, Kathy Dopp, Steve Freeman, myself and other analysts wrong. Really? On DU, we have won every online debate with that tag-team naysayer duo from Bard College and Britain. Baiman and Freeman, both PHDs, have had extensive, real world analytical experience. Dopp has a masters degree in mathematics. Their expertise is reflected in the quality of their published work on the Stolen Election. I'm NOT an amateur mathematician. I have two MS degrees in applied mathematics and have been developing financial, investment and engineering models long before you were born. What are YOUR academic credentials? What makes you an expert in polling analysis?

2.5 MILLION SPOILED DEMOCRATIC VOTES
It's a FACT that over 3 million votes (the great majority Democratic) are spoiled and never counted in every election. In 2004, 3.6mm votes were lost - just check the Census. I'm sure you would agree that Bush stole the 2000 election, despite having lost by 540,000 votes, 51-50.5mm. Add 75% of the 3 million spoiled votes to Gore and it's clear that he actually won by 2 million.

13 MILLION NEW BUSH VOTERS?
In 2004, the Bush recorded win was 62-59 million. Where did he gain over 13 million new votes (1.75mm Bush 2000 voters died before 2004) to get from 49mm to 62mm? Kerry won between 54% to 61% of those who did not vote in 2000 (depending on the exit poll timeline) and over 70% of 3 million Nader voters.

CONFIRMATION: rBr IS DEAD
Manjoo, I've got news for you; rBr is dead. You called the Reluctant Bush Responder theory persuasive, but it was debunked before you wrote "No Exit". The Exit Poll Response Optimizer (July 2005) confirmed the prior USCV simulation which refuted rBr.
Response Optimizer

THE EXIT POLLS WE WERE NOT SUPPOSED TO SEE
Many voters went to sleep on election night assuming that Kerry won. Jonathan Simon stayed up long enough to spot and download the 12:22am state exit polls. Luckily for us, the Washington Post chose not to delete the corresponding National Exit Poll (13047 respondents at 12:22am) which showed Kerry as the 51-48% winner. We also have the earlier NEP timelines at 4pm (8349 respondents) and 7:30pm (11027 respondents) which established the 51-48% Kerry trend. He held the lead until the Final National Exit Poll (13660 respondents), when the numbers magically reverted to a 51-48% Bush win. The numbers (weights and/or vote shares) were revised to match the vote.

THE IMPOSSIBLE FINAL EXIT POLL
According to the Final National Exit poll, 43% of the 122.3 million who voted in 2004 were Bush 2000 voters and 37% were Gore voters. The weightings in and of themselves debunk rBr. Now 43% of 122.3 is 52.57 million. And since Bush only got 50.5mm votes in 2000, of whom about 1.75mm died, only 48.7mm (39.8%) could have returned to vote in 2004. Therefore the 43% Final NEP weighting was mathematically impossible. Since a bogus Final NEP was forced to match the vote, the total Bush recorded vote was also bogus and had to be inflated by at least 4 million. I have just shown that the Final Poll at 1:25pm on Nov.3 was bogus and that the earlier 12:22am numbers are at least close to the truth.
Here's proof that the Final NEP weights were impossible.
Bogus Weights

Check out the National Exit Poll Timeline. Notice the smooth trend in the first three timelines (4pm,7:38pm,12:22am). Compare them to the discontinuous jump in ALL the Final NEP Bush demographic weights and/or vote shares.
National Exit Poll Timeline

REINCARNATED BUSH VOTERS?
We have proved that the Final NEP weights are mathematically impossible - unless you believe that 4 million Bush voters were reincarnated. Simple logic tells us that if the Final Exit poll was matched to the Bush recorded vote and was overstated by at least 4 million, then the Bush recorded vote must also have been overstated by at least 4 million.
If you read my related posts, you would know all this. But since you never mentioned them in your article, one can assume that a) you never read the posts, or b) you read them but chose not to discuss them. If it's (a), why didn't you? If it's (b), why didn't you? An impartial journalist seeking the truth would have.

NEP or rBr: NAYSAYER HOBSON'S CHOICE
Which do you believe: the Final NEP or rBr? You can't have both. They are MUTUALLY EXCLUSIVE.
Hobson's Choice

THE CLINCHER
You MUST read the "CLINCHER" thread. Kerry wins all plausible scenarios of 2000 voter turnout in 2004, assuming 100% Bush voter turnout and declining Gore voter turnout (from 100% to 70%) over a range of NEP Kerry vote shares (51-60%) of those who did not vote in 2000.
The Clincher

THE GAME THREAD
As far as the myth of "false voter recall" is concerned, I suggest you read the "GAME" thread. A Bard College professor's implausible explanation for the 3 million Bush "victory" margin (that almost one in 6 Gore voters defected to Bush in 2004 while just one of 16 Bush voters defected to Kerry) requires a leap of faith beyond the pale; the odds against it are astronomical. It's the naysayer's last stand. Their case against the "pristine" exit polls is in its "final throes", having been reduced to a totally implausible hypothesis. Farhad, ask yourself who you voted for in 2004 and 2000. Or did you forget?
The Game

PRE-ELECTION AND EXIT POLL SIMULATIONS
You say you want a simulation? Well, you know...
We're doing the best we can.
Pre election poll vs. Exit Poll vs.Simulation vs. Final

THE NEP TIMELINE - A TRAGEDY IN FOUR ACTS
11/2/04, 3:59pm 8349 respondents: Kerry 51-Bush 48
11/2/04, 7:33pm 11027 respondents: Kerry 51-Bush 48
11/3/04, 12:22am 13047 respondents: Kerry 51-Bush 48
11/3/04, 1:25pm 13660 respondents: Kerry 48-Bush 51

THE VILLAINS
Oh, sure, you will probably say: they matched the final exit poll to the vote. They do it all the time. Yes, Farhad, they do, even when the vote is corrupted. How come you never entertained the possibility that the votes were scammed? Or do you trust Bush/Rove to do the right thing and not try to steal the election? Remember Katherine Harris/ Jeb Bush in Florida 2000? Remember Ken Blackwell in Ohio 2004? Remember WMDs, mushroom clouds, Bush reading "My Pet Goat" on 911, domestic spying, torture, stem cells, Diebold, black disenfranchisement, spoiled votes, hacked voting machines, Hava bribes, Plame, the Aug.6, 2001 PDB, clearing Crawford brush.

THE RATIONALISTS
These are just a few of those who believe that the election was stolen: RFK Jr., John Conyers, Mark Miller, Robert Koehler, Michael Keefer, Thom Hartmann, Jim Lampley, Greg Palast, Bob Herbert, Steve Freeman, Kathy Dopp, Wayne Madsen, Ron Baiman, pollsters Harris and Zogby. Over 50% of the public believe that the election was stolen - and the percentage is growing daily. In fact, all the cable networks ran polls (except for Fox) which showed that a majority of viewers believe Bush stole it.
Network Polls

THE WIMPS
Of course, we still have spineless Democratic politicians who dare not call it "fraud". And that includes wannabee Al Franken. Well, at least Mike Malloy and Randi Rhodes are not afraid to talk about the stolen election. Unlike Al, they have no plans to run for office.

SOME QUESTIONS
Just curious. What is your math background? Specifically, statistical voting trends, polling mathematics, probability theory, Monte Carlo simulation? If you can't follow the math or logic in my posts, don't feel bad - the Mystery Pollster didn't either. You will learn the basics if you download and run the Interactive Election and Monte Carlo polling simulation models. Do you have Excel? Here's an intro to polling simulation and statistics:
Intro to Polling Mathematics

A link to all of my posts and models:
TruthIsAll

View my recent posts here:
Recent posts

THE SMOKING GUN
Finally, I suggest you consider the following in your next piece:Considering that Gore won in 2000 by over 540,000 votes and that
1) Kerry won a solid majority of new voters (57-41%) and
2) He won a large majority of Nader voters (71-21%) and
3) at most 49.2mm Gore and 48.7mm Bush voters (the ones still alive) could have turned out to vote in 2004,
then the discussion should end right here. Assuming a 95% turnout of Gore and Bush voters, Bush needed 19.5% OF GORE VOTERS TO WIN BY 3 MILLION VOTES.

ARE WE EXPECTED TO BELIEVE THAT ALMOST ONE OUT OF FIVE GORE VOTERS DECIDED TO FORGIVE BUSH FOR STEALING ELECTION 2000? OK, FORGIVE. BUT VOTE FOR HIM?

AT THE 12:28AM NEP TIMELINE, BUSH WON 8% OF GORE VOTERS. IN THE FINAL, JUST 10%.
__________________________________________________ USING MATHEMATICALLY FEASIBLE WEIGHTS AND A 95% TURNOUT OF 2000 VOTERS, BUSH NEEDED 19.5% OF GORE 2000 VOTERS TO WIN BY 3.1mm, ASSUMING NO OTHER VOTE SHARE CHANGES.

VOTED IN 2000
12:22am (13047 respondents)
Assumption: 95% turnout of Gore and Bush voters
.............Vote shares.......... Votes in Millions
Weight Votes Kerry Bush Other Kerry Bush Other
No 21.59% 26.37 57% 41% 2% 15.0 10.8 0.5
Gore 38.27% 46.75 80% 19.5% 0.5% 37.4 9.1 0.2
Bush 37.86% 46.26 10% 90% 0% 4.6 41.6 0.0
Nader 2.28% 2.79 71% 21% 8% 2.0 0.6 0.2
Total 100% 122.2 48.3% 50.9% 0.8% 59.0 62.1 1.0


Assuming a 2% MoE, the chances are LESS THAN ONE IN 450 TRILLION THAT BUSH WOULD GET AT LEAST 16% OF GORE VOTERS!
THE EXCEL PROBABILITY CALCULATION IS:
PROB = 1- NORMDIST(0.16,0.08,0.02/1.96,TRUE)
PROB = 2.220E-15
or 1 in 450,359,962,737,050

THE NORMDIST FUNCTION COULD NOT CALCULATE THE PROBABILITY THAT BUSH WOULD GET 19% OF GORE 2000 VOTERS; IT GENERATED A DIVISION BY ZERO ERROR. THE PROBABILITY IS AT THE VANISHING POINT!

_________________________________________________________ Using the actual 12:22am exit poll vote shares, Kerry won by over 7 MILLION VOTES!

VOTED IN 2000
.............Vote shares.......... Votes in Millions
Weight Votes Kerry Bush Other Kerry Bush Other
No 21.59% 26.37 57% 41% 2% 15.0 10.8 0.5
Gore 38.27% 46.75 91% 8% 1% 42.5 3.7 0.5
Bush 37.86% 46.26 10% 90% 0% 4.6 41.6 0.0
Nader 2.28% 2.79 71% 21% 8% 2.0 0.6 0.2
Total 100% 122.2 52.4% 46.6% 1.0% 64.2 56.8 1.2
Kerry margin: 7.4mm

Farhad, I look forward to your next piece. Just be sure that you discuss the analysis this time.

Best,
TruthIsAll