TIA’s 2004 Election Model


Exit Poll Analysis:  The Essential DU Threads

Exit Poll Analysis:  7/12/05 Update

2004 Election Collection




Other Links

2004 Election Results & Discussion

Election Reform, Fraud, and Related News

Tim Lohrentz on Exit Polls

Diane Perlman on Facing Election Fraud



TruthIsAll: By The Numbers

Edited on Thu Jul-14-05 07:50 AM by TruthIsAll


The Unanswered Question: Who Really Won in 2004?


The Election Model projected a Kerry victory, based on final state and national pre-election polls (two separate models). State polling data was input to a Monte Carlo Simulation model consisting of 5000 trials to determine the probability of Kerry winning the electoral vote, assuming a range of undecided vote split scenarios. The National Model calculated the probability of Kerry winning the popular vote based on an average of 18 national polls over a range of undecided vote scenarios. /

To believe that Bush won the election, a number of implausible circumstances had to occur.

To throw doubt on the exit polls, tortured hypotheticals were put forth by the naysayers who have been very prolific myth promoters. Now they are left with rBr, the Mother of All Myths, which has being thoroughly debunked by USCV and informed DUers. Here's a timeline of threads which refute the myths.

Much evidence (polling, political, anecdotal and incidental) pointed to massive fraud.

How does one explain the exit poll discrepancies?
Mitosksy states that Final Exit polls have always been weighted to match the actual vote. Now we know the reasons why. His statement, though literally true, is a canard. In fact, the vote count should be re-weighted to match the exit polls. Mitofsky would have us believe that the vote counts are correct and therefore preliminary exit poll numbers must be re-weightedto match the actual vote. That's why the final, weighted exit poll matched the votes. But this assumes the vote counts are accurate. Is this a valid assumption to make? The re-weighted exit polls are "correct" only if one defines “correct” as exactly matching bogus vote counts.

Given the recorded vote, exit poll response rate and error by precinct partisanship category, the EXIT POLL RESPONSE OPTIMIZER calculates Kerry’s true vote and refutes the rBr (reluctant Bush responder) hypothesis.

It’s the naysayer’s Hobson's Choice: Which do you believe,the Final NEP or rBr? One can argue, against all mathematical logic, that the Final National Exit poll was correct (43% is a valid weighting multiplier for Bush and 37% for Kerry). Therefore the Final Exit Poll correctly matched to the recorded vote count, as reflected by the 51-48% Bush win.

Conversely, one can hypothesize the reluctant Bush responder (rBr) theory. Of course, by doing so, one must must reject the 43/37% split, since that weighting mix implies that Bush 2000 voters outnumbered Gore 2000 voters by a whopping six percent.

Assume that 100% of Bush 2000 voters still alive (48.7 million, or 39.82% of 122.26) turned out to vote in 2004 and that the Final National Exit Poll "How Voted in 2000" weighting for Gore (37%) was correct. Also assume that the voting percentages are correct (the Final was the only poll Bush won.). Kerry still wins by 50.22% - 48.4%, a 2.23 million vote margin. In fact he wins all 120 scenarios.

This is an Exit Poll Analysis by Time Zone. The bulk of the damage was done in the East. All 22 states deviated to Bush from the exit poll to the vote. The odds: 1 in 4 million. Of the 22, 12 deviated beyond the MOE to Bush. The odds: 1 in 16 trillion.

Of 16261 respondents in the East, 40.40% said they voted for Bush. But Bush received 43.91% of the 2-party vote in the region.

According to the census, 125.7mm voted on Nov. 2 but only 122.3 million according to the vote count, so 3.4 million votes are missing. We know that millions of democratic votes are spoiled in minority precincts in every election. Naysayers would have us believe that the 2.70% census discrepancy is due to polling error, but the census margin of error is 0.30% for the gender question. The discrepancy dovetails with national and state exit poll deviations.

What happened in Ohio? The exit poll showed that Kerry was a solid winner. He won the majority of every demographic category. If you believe the Exit Poll, Kerry won OH by 160,000 votes. If you believe the count, Bush won by 119,000. If Kerry won this solid Republican state by 51-48%, he must have done better than this nationally - which means he won the election by 6-8 million votes.

Kerry held a steady 3-4% lead at each point in the timeline of the National Exit Poll, starting at 4pm (8349 respondents) to 7:33pm (11027) to 12:22am (13047). The Final Exit poll(13660) was released 1:25pm and radically changed the consistent timeline weightings and percentages in order to match the recorded vote.

State (73,000 respondents) and National (13047) exit polls were in agreement: Kerry was the winner. The fact that they matched within one tenth of a percentage point is further confirmation that they were accurate.

Analysis of state and National exit poll by region provides further confirmation and shows where the discrepancies were most dramatic.

In Florida, two voting methods were used. The party registration percentage split in counties using touch screen computers were virtually identical to counties using Optical scanners, with Democrats holding the edge. Why did Kerry do so much better in touch screen counties then he did in optiscan counties?

Even assuming 100% voter turnout for Bush 2000 voters and 88% for returning Gore voters, Kerry is a winner, using realistic weightings for the national exit poll. The Final Exit poll -- How Voted in 2000: demographic (43% Bush/ 37% Gore) is impossible, since 43% of 122.3mm voters is 52.59mm. Bush received 50.45mm votes in 2000. Since the number of Bush 2000 voters still living in 2004 was approximately 48.7mm, his maximum percentage is 8.7/122.3mm or 39.8%.

Regardless of the percentage of Election 2000 Gore voters returned to the polls his winning margin hardly changes. The percentage turnout factor is of minimal effect. However voter turnout weighting is the major factor in the wide discrepancy between the mathematically impossible Final Exit Poll at the 1:25pm timeline (13660 respondents) and the plausible 12:22am timeline (13047 respondents).

Based on the pre-election polls, 41 out of 51 states deviated to Bush. Based on the exit polls, 43 out of 51 deviated to him .

Exit Poll Margin of Error vs. Vote Deviation. Here’s a table of probabilities.

What are the odds?

Which of these facts convinced you that the election was stolen?