Log in

View Full Version : 1988-2008: The Statistical Evidence



TruthIsAll
03-29-2009, 06:09 PM
The Evidence

1988-2008
- Democrats do 3% better in RV (registered voter)polls which include new voters) than in LV (likely voter) polls
- Undecided voters break for challenger:70-80% according to Zogby, Harris; Gallup 90%
- Democratic late vote share (provisional, absentee PAPER ballots) was 7% higher than their Election Day share
- Uncounted votes (ranging from 3.5-10m) were 70-80% Democratic
- Average 14% Democratic margin in new voters from 1988-2004; 44% in 2008

- Millions of phantom Bush voters based on 2004/2008 Final NEP
- Largest exit poll discrepancies in 1988,1992, 2004(all Bush) and 2008
- Final NEP forced to match recorded vote using impossible return voter mix and implausible vote shares
- Democrats won True Vote by 52-42% but recorded vote by just 48-46%

2000
Gore won by 540,000 recorded votes; 5.4 million uncounted (4m for Gore)

2004
- Bush average monthly approval: high (0.87) correlation to average pre-election poll
- Pre-election poll and exit polls red-shift to Bush in 40-43 states
- Pre-elect polls (with UVA) matched unadjusted exit polls to within 1% (Kerry 52%)
- 3.45m uncounted votes (2.74% of total cast)

- NEP time line: steady Kerry 51-48% at 4pm, 730pm, 1220 am
- Uandjusted MoE exceeded in 29 states (all Bush)
- All 22 EST states deviated to Bush
- Implausible Bush Final NEP approval, undecided, Gender, Party ID
- Exit poll red-shift to Bush in Battleground (OH, FL) states, strong Democratic states (CA, NY) and large Bush states (TX, NC)
- No red-shift in small Bush states: ND, MT, SD, etc.

- WPE highest in Lever and punched card precincts, electronic (touch screens and optical scanners), lowest in paper ballots.
- Bush Urban Legend: increased vote share in urban areas, no gain in rural/small towns
- NY and CA provided 2.3m of Bush 3.0m margin.
- Final NEP forced to match recorded vote: Bush 51-48%

2004 Probability Analysis
7.4% WPE
1 in 17 trillion (1.32% MoE)
= 5.71765E-14
= 1- NORMDIST(0.525,0.475,0.0132/1.96,TRUE)

MoE exceeded in 16 states for Bush (12:22am Composite):
1 in 19 trillion
= 1- BINOMDIST(15,50,0.025,TRUE)
(probability is 2.5% that MoE would be exceeded in a given state)

MoE exceeded in 29 states for Bush
Absolute ZERO
= 1- BINOMDIST(28,50,0.025,TRUE)

EIRS:
National: 88 of 92 touch screen vote switching incidents from Kerry to Bush
Prob: 1 in 79,000,000,000,000,000,000,000
NY: 19 Lever voting machines stuck on Bush, none on Kerry

2006
- Pre-election 120 Generic poll trend (Dem 56.4%) matched unadjusted 7pm exit poll
- Final NEP forced to match recorded vote 52-46%

2008
- Pre-election RV polls indicated a 56-42% Obama win;
- NEP: forced to match recorded vote (52.9-45.6%)
- NEP: 12m more returning Bush voters than Kerry (Bush 2004 recorded margin: 3m)
- NEP: 5m returning 3rd party voters, but only 1.2m voted in 2004
- Not released: Unadjusted state and preliminary national exit polls. Why?