Arizona's Left Leaning Community
I saw this elsewhere about exit polling:
I replied to this slideshow: http://www.slideserve.com/prosper/forecasting-presidential-elections with this link to the 2004 pre-election and exit poll simulation model: Introduction to the 2004 Election Monte Carlo Simulation Model.
This is an introduction to election forecasting and exit poll simulation using the 2004 Election Simulation Model (ESM). It utilizes pre-election state and national polls and unadjusted (WPE), preliminary (GEO) and final ("Composite")state exit polls. The ESM provides strong circumstantial evidence that the election was stolen.
The model produces popular and electoral vote win probabilities based on 200 Monte Carlo election trials. https://docs.google.com/spreadsheet/ccc?key=0AjAk1JUWDMyRdDU5VERHay1mZExaT0lMRVhOXzg2aHc#gid=10
In the pre-election model, state and national polls are adjusted for the allocation of undecided voters. In the post-election model, the user can analyze unadjusted and adjusted state exit polls. Monte Carlo simulation calculates state and national vote shares, median and mean electoral votes. The popular and electoral win vote probabilities are a function of projected vote shares, user-entered undecided voter allocation and a "cluster effect" factor adjustment to the exit poll margin of error.
The National Election Pool is a consortium of six media giants which funds the exit polls.
In 2000, 110.8 million votes were cast but only 105.4 million were recorded. Gore won officially by 540,000 votes nationwide. But approximately 6 million uncounted ballots indicate that he won the True Vote by at least 3.0 million. In Florida 75,000 under-punched and 110,000 over-punched cards were uncounted. It is very likely that Gore won Florida by at least 50,000 votes since he was the clear choice on a solid majority of 110,000 punch cards which were double or triple-punched.
In every presidential election, millions of voters are disenfranchised and millions of votes are uncounted. Forecasting models should have the following disclaimer: The forecast will surely deviate from the official recorded vote. If they are nearly equal, then there must have been a) input data errors, b) incorrect assumptions, c) faulty model logic and/or methodology.
The True Vote always differs from the official recorded vote due to uncounted, switched and stuffed ballots.
Were the pollsters who forecast the recorded Bush win correct? Or were pollsters such as Zogby and Harris correct in projecting a Kerry the winner? Kerry won the preliminary state and national exit polls.
In 2004, Kerry had a slight lead in 18 pre-election national polls: 47.5-47.3% and a nearly identical lead in the state polls. After allocating undecided voters, he was projected to win by 51-48%. Kerry also won the unadjusted and preliminary state and national exit polls.
The Final National Exit Poll is always forced to match the recorded vote (Bush won by 50.7-48.3%). Either the pre-election and unadjusted exit polls were wrong or there was massive election fraud.