9/22 ELECTION MODEL ( TIA ): Obama! 330 EV, 99% win probability -- "on the Rise!" -x

[tiptoe]

2008 ELECTION MODEL
A  Monte Carlo  Electoral  Vote  Simulation

TruthIsAll

Updated: September 22

Press REFRESH after linking to a graph to view the latest update

Chart   State Poll Aggregate + Projection Trend

Chart   National 5-Poll Moving Average Projection

Chart   State vs. National: Vote Share Projection Trends

Chart   Battleground-State Polls

Chart   Battleground-State Win Probability

Chart   Obama Electoral Vote Simulation Frequency

Chart   Electoral Vote + Win Probability Trend

Chart   Electoral Vote + Projected Vote Share Trend

Chart   Undecided Voter Allocation + Win Probability

Chart   Monte Carlo Electoral Vote Simulation Trials

 2008 Election Model Fraud Analyzer 

Uncounted  &  Switched Votes

Chart   Effect on Obama Projected Vote Share

Chart   Effect on Obama Projected Electoral Vote

 

“I’m glad that TIA raises the questions. A Top 10 list.”  — Sancho

 

This
State
National
State
National
Monte  Carlo
Simulation

Update
Poll
5-Poll
2-party
2-party
Expected

9/22/2008
Aggregate
Average
Projection
Projection
Electoral Vote

Obama
McCain
 47.01 (50.77) 
 45.58 (49.23) 
 48.67 (52.05) 
 44.83 (47.95) 
51.46
48.54
52.57
47.43
 330 
 208 

    

15-Poll

End

Sample

Poll

NATIONAL MODEL

 

Pre   Undecided Voter Allocation
5-Poll Mov Avg

 

2-Party Projection  (60% UVA)
5-Poll Mov Avg

Trend
                            
Gallup
Rasmussen
Hotline/FD
CNN
Quinnipiac

CBS/NYT
Zogby
Ipsos
Pew Resrch
Newsweek

AP/gFk
FOX News
NBC/WSJ
CBS/NYT
CNN

Registered V
vs     Likely V
Poll Averages

Date
            
9/21
9/21
9/21
9/21
9/16

9/16
9/15
9/15
9/14
9/11

9/10
9/09
9/08
9/07
9/07

Size   
              
2796 RV
3000 LV
915 RV
697 LV
987 LV

800 LV
1008 LV
1046 RV
2307 LV
1038 RV

812 RV
900 RV
1000 RV
655 RV
942 RV

RV avg
LV avg
Total

MoE
         
1.85%
1.79%
3.24%
3.71%
3.12%

3.46%
3.09%
3.03%
2.04%
3.04%

3.44%
3.27%
3.10%
3.83%
3.19%

Obama
            
48
48
47
51
49

49
47
45
46
46

44
42
46
44
48

45.6
48.3
46.7

McCain
            
44
47
42
47
45

44
45
45
46
46

48
45
45
46
48

45.4
45.7
45.5

Other
            
8
5
11
2
6

7
8
10
8
8

8
13
9
10
4

9.0
6.0
7.8

Spread
            
4
1
5
4
4

5
2
0
0
0

(4)
(3)
1
(2)
0

0.1
2.7
1.1

 

Obama
            
48.7
48.5
48.0
48.2
47.2

46.6
45.6
44.6
44.8
44.4

44.8
45.4
45.0
44.6
45.0

McCain
            
44.8
45.0
44.7
45.2
45.0

45.2
46.0
46.0
46.0
46.0

46.4
46.6
46.6
48.4
47.2

Spread
            
3.8
3.5
3.3
3.0
2.2

1.4
(0.4)
(1.4)
(1.2)
(1.6)

(1.6)
(1.2)
(1.6)
(3.8)
(2.2)

Win Prob
               
98.5
98.0
86.2
80.2
77.3

66.7
44.5
30.9
26.3
28.4

30.9
34.8
29.0
14.8
23.2

 

Obama
            
52.57
52.4
52.4
52.2
51.9

51.5
50.6
50.2
50.3
50.2

50.1
50.2
50.0
48.8
49.7

McCain
            
47.43
47.6
47.6
47.8
48.1

48.5
49.4
49.8
49.7
49.8

49.9
49.8
50.0
51.2
50.3

Spread
            
5.1
4.8
4.8
4.3
3.8

3.0
1.3
0.5
0.6
0.3

0.2
0.4
0.1
(2.4)
(0.6)

Win Prob
               
99.7
99.6
92.7
87.3
88.1

80.5
65.8
56.2
62.1
54.1

51.8
54.8
51.0
27.0
42.2

 

 

Obama’s EV and Popular Vote Win Probability is on the Rise

Assuming that the election is held today (and is fraud-free), the Election Model projects that there is a better than 99% probability that Obama would win the election. His expected electoral vote margin is 330–208 with a 51.46% two-party vote share. He also leads the National model (based on the latest 5 national polls) with 52.57% of the vote — giving him a 99.7% popular vote win probability.

The EV win probability is a simple calculation: Obama won 4926 of 5000 simulated election trials. His win probability is therefore 98.5% (4926/5000). View the Election Model Electoral Vote Simulation Frequency chart. The model is a snapshot in time. Ideally, the weighted average vote shares would be identical in both the national and state models. But if they are, it’s just a coincidence; state polls lag the national polls. View the State vs. National Vote Share Projection Trend.

The base case scenario assumes that Obama will win 60% of the undecided vote. And this is conservative; he is presumed to be the challenger, since McSame is running for the third Bush term). Note that the national polls lead the state polls, so that we can expect a rise in Obama’s expected EV and win probability. The national model also assumes that he will win 60% of the undecided vote. The probability that he will win the popular vote is over 99% — again assuming zero fraud, fairly accurate polls with the election held today.

If Obama captures just 50% of the undecided vote, he is expected to win by 311–227 EV with a 50.6% vote share and a 96% win probability. The probability of winning the electoral vote is very close to the probability of winning the popular vote. In other words, if the projected vote share is nearly tied, the probability of winning the electoral vote will also be close to 50%.

Democrats are strongest in high EV urban states, and Republicans are strong in low EV rural states. That’s why Obama can win the Electoral vote with slightly less than 50% of the total popular vote. The sensitivity analysis (see below) shows that if Obama wins 40% of the undecided vote, he will have 50.0% of the 2-party vote, 290 electoral votes and an 80% win probability.

The fivethirtyeight.com site projects Obama with a 311.5–226.5 EV lead and a 74.4% win probability. Obama would surely win more than 74% of the trials in a Monte Carlo simulation with that expected EV split, unless they are factoring in a fraud component without saying so.

As of Sept. 22, electoral-vote.com has Obama leading by 273–265; realclearpolitics has Obama by 219–189 (130 tossup). But the 2008 Election Model (EM) has Obama leading by 330–208. Why the difference?

THE 2008 ELECTION MODEL

Last
Aggregate
5-poll
2-party
2-party
Monte  Carlo
Simulation

Update
State
National
State
National
Expected

9/22/2008
Average
Average
Projection
Projection
Electoral Vote

 
 
 
60% UVA
 
 

Obama
McCain
47.01
45.58
48.67
44.83
51.46
48.54
52.57
47.43
 330 
 208 

 

Nov 1, 2004 Final Election Model
75% UVA
 
 

Kerry
Bush
47.88
46.89
47.80
46.60
51.80
48.20
51.77
48.23
 337 
 201 

 


Sensitivity Analysis — Impact of Uncounted and Switched Votes on Obama

Uncounted
1%
 
2%
 
3%
 
Switched
2%
4%
6%
Vote%
50.4
49.4
48.4
EV
303
272
240
Vote%
50.2
49.2
48.1
EV
289
257
227
Vote%
49.9
48.9
47.9
EV
274
244
213

 

Sensitivity Analysis — Impact of Aggregate State Projected Vote Share

Undecided Voter Allocation Scenario
Base Case

Obama
40%
50%
60%
75%
80%

 

Projected 2-Party Vote Share

Obama
McCain
50.0
50.0
50.7
49.3
51.46
48.54
52.6
47.4
52.9
47.1

 

MoE
Obama Popular Vote Win Probability (Normdist)

1.0 %
2.0 %
3.0 %
47.9
49.0
49.3
91.9
75.8
68.0
99.78
92.3
82.9
100.0
99.4
95.3
100.0
99.8
97.3

 

Obama Electoral Vote (Monte Carlo - 5000 election trials)

Mean
Median
289.9
289.0
311.2
311.0
330.4
332.0
355.2
358.0
362.2
364.0

Maximum
Minimum
385
210
382
224
400
244
417
268
420
283

 

Obama Electoral Vote Win Probability (Monte Carlo)

Trial Wins
Probability
4009
80.2
4808
96.2
4978
99.56
4999
99.98
5000
100.0

 

95% EV Confidence Interval
Upper
Lower
337
243
358
265
375
286
392
319
396
329

 

States Won
Obama
 
25
 
25
 
29
 
29
 
29
 

 

 

 

Complete Update w State Model: http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=132&topic_id=7159446&mesg_id=7159446

 

[tiptoe]

1. Post-convention reality for McPalin is setting in... n/t

[Roland99]

2. Where are the models for adding third parties in? Nader's polling 6% in Mich.

[tiptoe]

4. Nader taking from McCain. Maybe he'll become a factor. nt

[MrsCorleone]

3. Assuming the election is "fraud-free"

[Rosa Luxemburg]

8. well this is the problem

[tiptoe]

9. especially when no other election models factor election fraud into analyses. "Ostrich Modelers" nt

[FreeState]

5. This will be really good with the new ABC poll - Obama +9 :)

Thats likely voters, he's +10 with registered voters

[MiniMe]

6. I can't help but notice the 2004 final pre-election poll in there

Kerry 337, Bush 201

Even without the fraud in 2004, this isn't close to what happened. It makes me doubt this model as a predictor.

[tiptoe]

10. TIA's 2004 Election Model's Final Pre-Election Projection of 337 EV for Kerry matched EXACTLY the

aggregate state electoral vote count (column "EV") of the first-of-three Preliminary "Unadjusted Exit Polls" on election night, identified as "WPE Method" in the chart available here:

http://www.geocities.com/electionmodel/MatchingFinalExitPollToTheVote.htm

Please reconsider your notion "this isn't close to what happened" after understanding what did happen on Election night and following day, 2004.

Note the chart lists three Preliminary "Unadjusted" Exit Polls: WPE Method, Geo, and Composite.

Pollster Edison-Mitofsky provided four aggregate state exit poll measures: the three "unadjusted" Preliminary Exit Polls and a Final "adjusted" poll (i.e. a "poll" 'forced' to match the Recorded vote).

The "WPE Method" Preliminary Exit Poll is the "pristine" poll, unmodified, purely voter-response-based for 1250 representative sample precincts around the states. ("...the relevant data has not been made available for independent analysis"). The other two, later preliminary exit polls are modifications of the pristine poll results by degrees of incorporation of incoming vote counts and pre-election poll data.

Kerry won the first three exit polls; Bush won the Final "poll":

Exit Poll

    Vote Share

EV based on
states WPE

1)    WPE

    51.8 – 47.2%

377 – 161

   (unadjusted)

2)    GEO

    51.0 – 48.5%

 

   (modified with incoming recorded votes)

3)    Composite

    50.3 – 49.1%

 

   (12:22am – modified with pre-election polls)    (13,047 random sample, 1% Margin of Error)

4)    Final

    48.5 – 51.1%

 

   ('forced' to match the official recorded vote count)

 

TIA's 2004 Election Model* final pre-election projection of Kerry 377 Expected EV exactly matched the unadjusted 'WPE Method' exit poll's calculated aggregate EV based on state WPE data provided by pollster Mitofsky (see chart).

The official 286 EV awarded to Bush -- based on a recorded vote that is arguably fraudulent -- is what "isn't close to what happened" with either 1) the calculated aggregate EV based on Mitofsky unadjusted exit poll data or 2) TIA's Election Model projection of 377 expected EV based on pre-election poll data. The latter two independent measures matched exactly; the official 286 EV is what is "out of whack" because based on election fraudulent recorded vote count).

The chart's first column is the official recorded vote count, to which the Final Exit Poll was 'forced' to match, with adjustments resulting in mathematical impossibilities and implausibilities for explaining Bush' 3Million official-vote margin mandate, implying the official recorded vote was fraudulent.


*page 3-5

[tiptoe]

11. 9/24 2008 Election Model update provided in above post: click "2004 Election Model" link. n/t

[tiptoe]

12. typo correction: "377" should be "337" and "161"-->"201" (sorry) n/t

[tiptoe]

7. correction: "...Obama won 4978 of 5000 Monte Carlo simulation election trials. His win probability

is therefore 99.56% (4978/5000)."