TruthIsAll
06-06-2008, 06:10 AM
2008 Election Model
Monte Carlo Electoral Vote Simulation
TruthIsAll
http://www.geocities.com/electionmodel/2008ElectionModel.htm
The Election Model is designed to project the most-likely outcome if it were held today.
View the current 2008 projection based on the latest state and national polls.
This overview contains a brief discussion of the following:
- Time-series regression models vs. Monte Carlo polling simulation
- Final 2004 Election Model state and national projections confirmed by the exit polls
- Analysis of 2004 registered voter (RV) and likely voter (LV) polls
- Election Model methodology
- Basic Polling Mathematics
- Overview of Monte Carlo Electoral Vote Simulation
2004 Election Model polling data and projection trend tables:
- Final state and national pre-election and exit polls
- Real Clear Politics (RCP) 102 RV and 31 LV poll trend and analysis
Election Forecasting: regression analysis vs. Monte Carlo Simulation
Academics and political scientists often forecast an election months in advance.
Virtually all the models utilize a unique set of time-series data using multiple regression analysis.
They develop a linear model which projects vote share as the dependent variable.
Typical factors are economic growth, inflation, job growth, interest rates, foreign policy, etc.
But the models fail to account for the daily events which affect voter psychology.
It’s like the stock market.
Statistical analysis of prior elections and polling trends is a better indicator of voter preference.
There are inherent problems in academic and media pundit election forecast models:
The implicit hypothesis is that the official recorded vote accurately reflects the True vote.
In other words, the assumption is that the election will be fraud-free. But what if it isn’t?
Academic forecast models never account for the fraud factor.
Are academics and pundits correct when they forecast the official “winner” in a stolen election?
In 2000, Bush had fewer votes than Gore. But Bush was the official winner.
Their models had to be wrong to project a winner who actually got fewer votes.
Were those who forecast Bush the winner right - or wrong?
It is not as well known that Kerry easily defeated Bush in 2004. But Bush was the official winner.
Were analysts who projected Kerry the True vote winner wrong because the election was stolen?
How could their models anticipate that the massive election fraud?
The evidence is clear: millions of mostly Democratic votes are uncounted in every election.
And it’s a fact that over one million Democratic voters will be disenfranchised.
In 2000, 110.8 million votes were cast, but only 105.4 million recorded,
In 2004, 125.7 million votes were cast, but only 122.3 million recorded.
Why should we expect 2008 to be any different?
Can we be confident that unverifiable DRE touch screens will reflect the intent of the voter?
Can we assume that central tabulator software will not be tampered with?
Obama’s True Vote (T) will be reduced by uncounted (U) and switched votes (S).
His recorded vote is given by the formula: R = T - U - S
(note: the formula does not account for disenfranchised voters)
The bottom line is that Obama will need a landslide to overcome multiple levels of fraud.
The 2004 Election Model:
Final pre-election projections confirmed by the exit polls
The pre-election aggregate State Poll average matched the National Poll average to within 0.7%
The aggregate state model projection matched the unadjusted aggregate state exit poll to within 0.8%
The national model projection matched the 12:22am National Exit Poll to within 0.1%.
There was a 0.87 correlation ratio between the Bush monthly approval rating and the national poll trend.
Kerry’s final projected vote shares were within 3.0% of the exit poll in 30 states, 2% in 21 and 1% in 12.
Kerry’s recorded vote shares were less than the exit poll in 45 states.
Kerry’s recorded vote shares were at least 2.5% below the projected shares in 27 states.
Kerry’s recorded vote shares were less than the projected shares in 47 states.
The Monte Carlo simulation indicated that Kerry would win 51.0% with 337 EV (51.8% of the 2-party vote).
The National projection model indicated that he would win 50.9% (51.6% of the 2-party vote).
The aggregate pre-election state polls matched the national polls.
Kerry led the aggregate final state pre-election polls by 47.7 - 47.0%.
The aggregate state polling average closely matched the national 18-poll average (47.3-46.9%).
Unadjusted state exit polls indicated that Kerry won by 51.8-47.2%.
The Election Model State projection matched the unadjusted exit poll aggregate to within 0.8%.
The 12:22am National Exit Poll update of 13047 respondents indicated that Kerry won by 50.8-48.2%.
The NEP margin of error was 1.12%, assuming a 30% exit poll “cluster” effect.
The Election Model projection matched the NEP to within 0.1%.
Professional pollsters allocate undecided voters to the challenger, especially if the incumbent is unpopular.
Bush had a 48% approval rating on Election Day.
The Gallup Poll allocated 90% of undecided voters to Kerry.
Harris and Zogby projected that he would get 67-80%.
The Final Election Model base case scenario was that Kerry would capture 75% of undecided voters.
Bush won the recorded vote with 50.7% and had 286 EV.
Approximately 3.4 million votes were uncounted.
A few naysayers still argue that polling analysis cannot prove that the 2004 election was stolen.
But a careful analysis of the pre-election and exit polls provides powerful evidence that it was.
They claim the early exit polls were wrong and that the Final National Exit Poll was correct.
But the Final was forced to match the recorded vote.
And it was proved to be mathematically impossible.
It assumed that 4 million more Bush 2000 voters than actually voted turned out in 2004.
They also argue that the pre-election polls favored Bush; the data is provided in the tables below.
The unweighted state poll average favored Bush, but Kerry led in the aggregate weighted average.
They claim that Bush was leading the national polls, but the data indicates a virtual tie.
To believe that Bush won, you must believe that all pre-election and unadjusted exit polls were wrong.
And that only the Final Exit Poll, which was forced to match the recorded vote, was correct.
2004 Pre-election RV vs. LV Poll Trend Analysis
The National pre-election polls, after adjusting for voter turnout, undecided voters, reflect the True Vote.
Data for 18 final Election Model polls and 133 total polls from Real Clear Politics (RCP) are shown below.
The final 15 RCP polls all sampled likely (LV) voters.
Earlier polls were a mix of likely voters and registered (RV) voters.
The Election Model matched the pre-election LV polls after adjusting for new (RV) voters and undecided voters.
Kerry led the pre-election polls in August, fell behind in September and rebounded in October.
When undecided voters were allocated, Kerry was projected to win.
Kerry did much better in RV polls than in LV polls.
There are several reasons for this:
1- RV polls include new registered voters.
2. The majority of new voters (60%) were Democrats.
3- There was a 17 million net increase in the recorded vote (from 105.4m in 2000 to 122.3m in 2004).
4- Approximately 5m who voted in 2000 died prior to 2004 and 5m did not vote (95% turnout).
5- Just 95m returned to vote in 2004.
6- There were 125.7m total votes CAST in 2004 (3.4m were uncounted).
7- Therefore, there were approximately 30m (125.7- 95.4) new voters in 2004.
8- Kerry won approximately 18m (60%) of new voters, Bush 12m.
The final 15 RCP poll national average had Bush leading by 49-47%.
The final Gallup poll was an identical 49-47 before it was adjusted for undecided voters.
Gallup assigned 90% of undecided voters to Kerry, so the final Gallup poll had the race tied at 49-49%
Assume that Kerry and Bush split the vote of the 95m returning (LV) voters.
So the race was tied at 47m with 1m to other 3rd party candidates.
Allocating the 18 of the 30m new voters to Kerry, he is the winner of the True Vote by 65-59m (51.8-46.9%).
Is it just a coincidence that Kerry had 51.8% in the unadjusted aggregate state exit poll (WPE method)?
Now let's compare the True Vote result to the pre-election national polls.
As per Gallup, the final 15 LV poll average, adjusted for undecided voters, was 49-49%.
But the LV polls did not sample newly registered voters.
The Election Model utilized 18 final polls, split equally between RV and LV.
After allocating 75% of undecided voters to Kerry, he led by 50.9-48.1%
Based on the 122.3m recorded vote, Kerry won by 62.3-58.8m
Kerry's projected share was still 0.9% below his True Vote share (51.8%) calculated above.
But the projection did not include the 3.4m uncounted votes (2.74% of total votes cast).
Assuming that Kerry won 75% of the uncounted votes, we can add 2.5m to his total and 0.9m to Bush.
Kerry's margin becomes 64.8-59.7m (51.6-47.5%).
That's within 0.2% of the True Vote and the unadjusted aggregate State Exit poll share.
Methodology:
5-poll moving average projection trend
75% of undecided voters allocated to Kerry
Other 3rd party projection: 1%
Election Model Methodology
It actually contains two independent models:
a) Monte Carlo Electoral Vote Simulation - based on the pre-election state polls.
b) National average model - based on the latest national polls.
In the state model, the average weighted poll share is calculated. The vote shares are projected by adjusting the polls for the allocation of the undecided voters. In the Monte Carlo simulation, 5000 election trials are executed to calculate the expected electoral vote and win probability.
A powerful feature is the built-in sensitivity analysis. Five scenarios of undecided voter allocation are executed to project state and national vote shares, electoral votes and win probability.
Political scientists generally use one of three methods to project elections.
The first method analyzes historical economic data: growth, jobs, inflation, etc. using regression analysis models to predict the popular vote. Unfortunately, these models lack precision; they often use limited, outdated time series data. The data doesn’t reflect current news and information which affect voter psychology and preference.
The second method tracks national polls to project the winner of the popular vote.
The third method tracks the latest state polls in order to project the popular and electoral vote.
The Election Model uses current state and national polls. The only projection assumption is in the allocation of undecided/other voters. Historically, undecided voters have split at least 2-1 for the challenger. So if a poll has the race tied at 45-45, then allocating a 60-40% split of the undecided 10% derives a 51-49% projected vote share.
The winner of the popular vote will almost certainly win the electoral vote. But that might not be the case if the winning margin is less than 0.5%.
Polling Mathematics
An advantage of national polling is its relative simplicity. If the polling spread exceeds the margin of error (3% for a typical 1000 sample) then the leader has at least a 97.5% chance of winning the election assuming a) it was held that day and b) the poll is an unbiased sample. But that’s just the probability for a single poll.
If three independent national polls are done on the same day, that is approximately the equivalent a single poll of 3000 sample size, and the theoretical MoE is 1.8%. Assuming that the average split is 52-48%, there is a 95% probability that the leader will receive between 50.2% and 53.8%. But since there is a 2.5% probability that his vote share will exceed 53.8%, there is a 97.5% probability of winning at least 50.2%.
The MoE is 1.96 times the standard deviation, which is a statistical measure of the variability of polling observations. The standard deviation and projected vote share is input to the normal distribution function in order to determine the probability of winning a vote majority.
To calculate the expected EV from state polling data, the final vote is projected. State polls typically sample 600 voters, so the state MoE is 4% compared to 3% in the National polls. The probability of winning a state is based on the projected vote: the 2-party state polls adjusted for undecided voter allocation.
For example, in the case of a 50-50 projection, each candidate has a 50% probability of winning the state. For a 51-49 split, the leader has a 69% chance of winning; for 52-48, the probability is 83%; for 53-47, 93%; for 54-46, it’s 97.5%.
Monte Carlo Simulation
In the simulation, 5000 election trials are run to determine the probability of winning 270 Electoral Votes. The probability is the number of election trial wins divided by 5000.
The projected vote share for each state is first calculated by applying the undecided voter allocation to the most recent 2-party poll. A win probability is then calculated based on the projected 2-party vote. A random number (RND) between 0 and 1 is generated for each state and compared to the probability to determine the winner. For example, assume the latest poll indicates that Obama has a 90% probability of winning Oregon. If the RND generated for Oregon is less than 0.90, Obama wins Oregon’s 7 electoral votes; otherwise, McCain does.
The procedure is repeated for all 50 states and DC to determine the election trial winner: the candidate who has won least 270 EV. A total of 5000 election trials are executed. The EV winner is determined for each trial. The total number of trial wins is calculated for each candidate. The probability of winning the electoral vote is the total number of trial wins divided by 5000. The expected (mean) electoral vote is just the average EV.
An advantage of running a simulation is that minor shifts in individual state polls have minimal impact since the expected EV is the average of 5000 simulations and not just a single snapshot. Independent national and state polling models provide a mathematical confirmation of each method.
In summary, the Election Model projects the latest national and state polls after adjusting for the allocation of undecided voters. The probability of winning each state is calculated. A Monte Carlo simulation of 5000 election trials is then executed (using the individual state probabilities) to determine the expected final electoral vote and win probability.
Monte Carlo Electoral Vote Simulation
TruthIsAll
http://www.geocities.com/electionmodel/2008ElectionModel.htm
The Election Model is designed to project the most-likely outcome if it were held today.
View the current 2008 projection based on the latest state and national polls.
This overview contains a brief discussion of the following:
- Time-series regression models vs. Monte Carlo polling simulation
- Final 2004 Election Model state and national projections confirmed by the exit polls
- Analysis of 2004 registered voter (RV) and likely voter (LV) polls
- Election Model methodology
- Basic Polling Mathematics
- Overview of Monte Carlo Electoral Vote Simulation
2004 Election Model polling data and projection trend tables:
- Final state and national pre-election and exit polls
- Real Clear Politics (RCP) 102 RV and 31 LV poll trend and analysis
Election Forecasting: regression analysis vs. Monte Carlo Simulation
Academics and political scientists often forecast an election months in advance.
Virtually all the models utilize a unique set of time-series data using multiple regression analysis.
They develop a linear model which projects vote share as the dependent variable.
Typical factors are economic growth, inflation, job growth, interest rates, foreign policy, etc.
But the models fail to account for the daily events which affect voter psychology.
It’s like the stock market.
Statistical analysis of prior elections and polling trends is a better indicator of voter preference.
There are inherent problems in academic and media pundit election forecast models:
The implicit hypothesis is that the official recorded vote accurately reflects the True vote.
In other words, the assumption is that the election will be fraud-free. But what if it isn’t?
Academic forecast models never account for the fraud factor.
Are academics and pundits correct when they forecast the official “winner” in a stolen election?
In 2000, Bush had fewer votes than Gore. But Bush was the official winner.
Their models had to be wrong to project a winner who actually got fewer votes.
Were those who forecast Bush the winner right - or wrong?
It is not as well known that Kerry easily defeated Bush in 2004. But Bush was the official winner.
Were analysts who projected Kerry the True vote winner wrong because the election was stolen?
How could their models anticipate that the massive election fraud?
The evidence is clear: millions of mostly Democratic votes are uncounted in every election.
And it’s a fact that over one million Democratic voters will be disenfranchised.
In 2000, 110.8 million votes were cast, but only 105.4 million recorded,
In 2004, 125.7 million votes were cast, but only 122.3 million recorded.
Why should we expect 2008 to be any different?
Can we be confident that unverifiable DRE touch screens will reflect the intent of the voter?
Can we assume that central tabulator software will not be tampered with?
Obama’s True Vote (T) will be reduced by uncounted (U) and switched votes (S).
His recorded vote is given by the formula: R = T - U - S
(note: the formula does not account for disenfranchised voters)
The bottom line is that Obama will need a landslide to overcome multiple levels of fraud.
The 2004 Election Model:
Final pre-election projections confirmed by the exit polls
The pre-election aggregate State Poll average matched the National Poll average to within 0.7%
The aggregate state model projection matched the unadjusted aggregate state exit poll to within 0.8%
The national model projection matched the 12:22am National Exit Poll to within 0.1%.
There was a 0.87 correlation ratio between the Bush monthly approval rating and the national poll trend.
Kerry’s final projected vote shares were within 3.0% of the exit poll in 30 states, 2% in 21 and 1% in 12.
Kerry’s recorded vote shares were less than the exit poll in 45 states.
Kerry’s recorded vote shares were at least 2.5% below the projected shares in 27 states.
Kerry’s recorded vote shares were less than the projected shares in 47 states.
The Monte Carlo simulation indicated that Kerry would win 51.0% with 337 EV (51.8% of the 2-party vote).
The National projection model indicated that he would win 50.9% (51.6% of the 2-party vote).
The aggregate pre-election state polls matched the national polls.
Kerry led the aggregate final state pre-election polls by 47.7 - 47.0%.
The aggregate state polling average closely matched the national 18-poll average (47.3-46.9%).
Unadjusted state exit polls indicated that Kerry won by 51.8-47.2%.
The Election Model State projection matched the unadjusted exit poll aggregate to within 0.8%.
The 12:22am National Exit Poll update of 13047 respondents indicated that Kerry won by 50.8-48.2%.
The NEP margin of error was 1.12%, assuming a 30% exit poll “cluster” effect.
The Election Model projection matched the NEP to within 0.1%.
Professional pollsters allocate undecided voters to the challenger, especially if the incumbent is unpopular.
Bush had a 48% approval rating on Election Day.
The Gallup Poll allocated 90% of undecided voters to Kerry.
Harris and Zogby projected that he would get 67-80%.
The Final Election Model base case scenario was that Kerry would capture 75% of undecided voters.
Bush won the recorded vote with 50.7% and had 286 EV.
Approximately 3.4 million votes were uncounted.
A few naysayers still argue that polling analysis cannot prove that the 2004 election was stolen.
But a careful analysis of the pre-election and exit polls provides powerful evidence that it was.
They claim the early exit polls were wrong and that the Final National Exit Poll was correct.
But the Final was forced to match the recorded vote.
And it was proved to be mathematically impossible.
It assumed that 4 million more Bush 2000 voters than actually voted turned out in 2004.
They also argue that the pre-election polls favored Bush; the data is provided in the tables below.
The unweighted state poll average favored Bush, but Kerry led in the aggregate weighted average.
They claim that Bush was leading the national polls, but the data indicates a virtual tie.
To believe that Bush won, you must believe that all pre-election and unadjusted exit polls were wrong.
And that only the Final Exit Poll, which was forced to match the recorded vote, was correct.
2004 Pre-election RV vs. LV Poll Trend Analysis
The National pre-election polls, after adjusting for voter turnout, undecided voters, reflect the True Vote.
Data for 18 final Election Model polls and 133 total polls from Real Clear Politics (RCP) are shown below.
The final 15 RCP polls all sampled likely (LV) voters.
Earlier polls were a mix of likely voters and registered (RV) voters.
The Election Model matched the pre-election LV polls after adjusting for new (RV) voters and undecided voters.
Kerry led the pre-election polls in August, fell behind in September and rebounded in October.
When undecided voters were allocated, Kerry was projected to win.
Kerry did much better in RV polls than in LV polls.
There are several reasons for this:
1- RV polls include new registered voters.
2. The majority of new voters (60%) were Democrats.
3- There was a 17 million net increase in the recorded vote (from 105.4m in 2000 to 122.3m in 2004).
4- Approximately 5m who voted in 2000 died prior to 2004 and 5m did not vote (95% turnout).
5- Just 95m returned to vote in 2004.
6- There were 125.7m total votes CAST in 2004 (3.4m were uncounted).
7- Therefore, there were approximately 30m (125.7- 95.4) new voters in 2004.
8- Kerry won approximately 18m (60%) of new voters, Bush 12m.
The final 15 RCP poll national average had Bush leading by 49-47%.
The final Gallup poll was an identical 49-47 before it was adjusted for undecided voters.
Gallup assigned 90% of undecided voters to Kerry, so the final Gallup poll had the race tied at 49-49%
Assume that Kerry and Bush split the vote of the 95m returning (LV) voters.
So the race was tied at 47m with 1m to other 3rd party candidates.
Allocating the 18 of the 30m new voters to Kerry, he is the winner of the True Vote by 65-59m (51.8-46.9%).
Is it just a coincidence that Kerry had 51.8% in the unadjusted aggregate state exit poll (WPE method)?
Now let's compare the True Vote result to the pre-election national polls.
As per Gallup, the final 15 LV poll average, adjusted for undecided voters, was 49-49%.
But the LV polls did not sample newly registered voters.
The Election Model utilized 18 final polls, split equally between RV and LV.
After allocating 75% of undecided voters to Kerry, he led by 50.9-48.1%
Based on the 122.3m recorded vote, Kerry won by 62.3-58.8m
Kerry's projected share was still 0.9% below his True Vote share (51.8%) calculated above.
But the projection did not include the 3.4m uncounted votes (2.74% of total votes cast).
Assuming that Kerry won 75% of the uncounted votes, we can add 2.5m to his total and 0.9m to Bush.
Kerry's margin becomes 64.8-59.7m (51.6-47.5%).
That's within 0.2% of the True Vote and the unadjusted aggregate State Exit poll share.
Methodology:
5-poll moving average projection trend
75% of undecided voters allocated to Kerry
Other 3rd party projection: 1%
Election Model Methodology
It actually contains two independent models:
a) Monte Carlo Electoral Vote Simulation - based on the pre-election state polls.
b) National average model - based on the latest national polls.
In the state model, the average weighted poll share is calculated. The vote shares are projected by adjusting the polls for the allocation of the undecided voters. In the Monte Carlo simulation, 5000 election trials are executed to calculate the expected electoral vote and win probability.
A powerful feature is the built-in sensitivity analysis. Five scenarios of undecided voter allocation are executed to project state and national vote shares, electoral votes and win probability.
Political scientists generally use one of three methods to project elections.
The first method analyzes historical economic data: growth, jobs, inflation, etc. using regression analysis models to predict the popular vote. Unfortunately, these models lack precision; they often use limited, outdated time series data. The data doesn’t reflect current news and information which affect voter psychology and preference.
The second method tracks national polls to project the winner of the popular vote.
The third method tracks the latest state polls in order to project the popular and electoral vote.
The Election Model uses current state and national polls. The only projection assumption is in the allocation of undecided/other voters. Historically, undecided voters have split at least 2-1 for the challenger. So if a poll has the race tied at 45-45, then allocating a 60-40% split of the undecided 10% derives a 51-49% projected vote share.
The winner of the popular vote will almost certainly win the electoral vote. But that might not be the case if the winning margin is less than 0.5%.
Polling Mathematics
An advantage of national polling is its relative simplicity. If the polling spread exceeds the margin of error (3% for a typical 1000 sample) then the leader has at least a 97.5% chance of winning the election assuming a) it was held that day and b) the poll is an unbiased sample. But that’s just the probability for a single poll.
If three independent national polls are done on the same day, that is approximately the equivalent a single poll of 3000 sample size, and the theoretical MoE is 1.8%. Assuming that the average split is 52-48%, there is a 95% probability that the leader will receive between 50.2% and 53.8%. But since there is a 2.5% probability that his vote share will exceed 53.8%, there is a 97.5% probability of winning at least 50.2%.
The MoE is 1.96 times the standard deviation, which is a statistical measure of the variability of polling observations. The standard deviation and projected vote share is input to the normal distribution function in order to determine the probability of winning a vote majority.
To calculate the expected EV from state polling data, the final vote is projected. State polls typically sample 600 voters, so the state MoE is 4% compared to 3% in the National polls. The probability of winning a state is based on the projected vote: the 2-party state polls adjusted for undecided voter allocation.
For example, in the case of a 50-50 projection, each candidate has a 50% probability of winning the state. For a 51-49 split, the leader has a 69% chance of winning; for 52-48, the probability is 83%; for 53-47, 93%; for 54-46, it’s 97.5%.
Monte Carlo Simulation
In the simulation, 5000 election trials are run to determine the probability of winning 270 Electoral Votes. The probability is the number of election trial wins divided by 5000.
The projected vote share for each state is first calculated by applying the undecided voter allocation to the most recent 2-party poll. A win probability is then calculated based on the projected 2-party vote. A random number (RND) between 0 and 1 is generated for each state and compared to the probability to determine the winner. For example, assume the latest poll indicates that Obama has a 90% probability of winning Oregon. If the RND generated for Oregon is less than 0.90, Obama wins Oregon’s 7 electoral votes; otherwise, McCain does.
The procedure is repeated for all 50 states and DC to determine the election trial winner: the candidate who has won least 270 EV. A total of 5000 election trials are executed. The EV winner is determined for each trial. The total number of trial wins is calculated for each candidate. The probability of winning the electoral vote is the total number of trial wins divided by 5000. The expected (mean) electoral vote is just the average EV.
An advantage of running a simulation is that minor shifts in individual state polls have minimal impact since the expected EV is the average of 5000 simulations and not just a single snapshot. Independent national and state polling models provide a mathematical confirmation of each method.
In summary, the Election Model projects the latest national and state polls after adjusting for the allocation of undecided voters. The probability of winning each state is calculated. A Monte Carlo simulation of 5000 election trials is then executed (using the individual state probabilities) to determine the expected final electoral vote and win probability.