5/29 Election Model (Monte Carlo Simulation): Obama 51.4%; 297 EV; 95% Win Prob

[tiptoe]

2008 Election Model
Monte Carlo Electoral Vote Simulation

TruthIsAll

Updated: 30-May
     http://www.geocities.com/electionmodel/2008ElectionModel.doc

If the election were held today, there is a  95%  probability that Obama would defeat McCain.
The analysis is based on a Monte Carlo Simulation consisting of 5000 election trials using the latest state polls.
Obama is expected to win by a  297–241  electoral vote margin with  51.4%  of the 2-party vote.

The model generates five scenarios of 5000 election trials for various undecided voter assumptions:
Obama captures  50%,  55%,  60% (the base case),  65%  and  70%  of the undecided vote.

Polling Data Source:
http://www.electoral-vote.com/evp2008/Obama/Maps/May29.html


Simulation Summary

             	EV	Vote%	Prob%	
	Obama 	297	51.4	95	
	McCain	241	48.6	 5	
 					
                        Undecided Voter Allocation
Obama	                50%	55%	60%	65%	70%  
Trial Wins              3540	4323	4735	4951	4991 
 									
Projected%										
Obama	                50.5	51.1	51.4	52.1	52.7 
McCain	                49.5	48.9	48.6	47.9	47.3 
 									
Win Prob%										
Obama	                70.8	86.5	94.7	99.0	99.8 
 									
Electoral Vote									
Average                 276	286	297	307	317  
Median  	        279	287	297	307	316  
 
Maximum	                336     352	357	369	386          
Minimum	                181     189	201	223	238          
 									
95% Conf. Interval								
Upper Limit	        311	320	329	337	346  
Lower Limit	        241	253	264	277	287  
 									
States Won	        22	23	24	24	25   

State Model Projections

Base Case Assumption:
Obama captures 60% of undecided / other voters

Polls before Undecided Voter allocation


State

AL
AK
AZ
AR
CA
 
CO
CT
DC
DE
FL
 
GA
HI
ID
IL
IN
 
IA
KS
KY
LA
ME
 
MD
MA
MI
MN
MS
 
MO
MT
NE
NV
NH
 
NJ
NM
NY
NC
ND
 
OH
OK
OR
PA
RI
 
SC
SD
TN
TX
UT
 
VT
VA
WA
WV
WI
WY
 
Total


EV

9
3
10
6
55
 
9
7
3
3
27
 
15
4
4
21
11
 
7
6
8
9
4
 
10
12
17
10
6
 
11
3
5
5
4
 
15
5
31
15
3
 
20
7
7
21
4
 
8
3
11
34
5
 
3
13
11
5
10
3
 
538
Kerry
vs.
Bush

37.1
36.8
44.7
45.1
55.0
 
47.6
55.3
90.5
53.8
47.5
 
41.6
54.4
30.7
55.2
39.6
 
49.7
37.1
40.0
42.7
54.6
 
56.6
62.7
51.7
51.8
40.5
 
46.4
39.5
33.2
48.7
50.7
 
53.4
49.6
59.3
43.8
36.1
 
48.9
34.4
52.1
51.3
60.6
 
41.4
39.1
42.8
38.5
26.7
 
60.3
45.9
53.6
43.5
50.2
29.7
 
48.8
Obama
vs.
McCain

41.4
46.4
45.6
39.0
54.8
 
54.0
59.8
90.6
55.4
46.0
 
43.8
65.8
44.4
65.4
46.6
 
56.0
40.6
38.6
45.2
57.6
 
57.2
57.0
50.2
58.4
43.2
 
49.2
47.4
37.4
48.4
53.4
 
63.2
51.2
56.8
46.6
48.8
 
55.8
39.4
58.0
52.2
58.4
 
49.2
43.0
37.6
44.4
33.6
 
67.8
54.4
56.4
42.2
49.0
41.6
 
51.4


Obama
leads

Flip to
Obama

McCain
leads


Diff

4.3
9.6
0.9
( 6.1 )
( 0.2 )
 
6.4
4.5
0.1
1.6
( 1.5 )
 
2.2
11.4
13.7
10.2
7.0
 
6.3
3.5
( 1.4 )
2.5
3.0
 
0.6
( 5.7 )
( 1.5 )
6.6
2.7
 
2.8
7.9
4.2
( 0.3 )
2.7
 
9.8
1.6
( 2.5 )
2.8
12.7
 
6.9
5.0
5.9
0.9
( 2.2 )
 
7.8
3.9
( 5.2 )
5.9
6.9
 
7.5
8.5
2.8
( 1.3 )
( 1.2 )
11.9
 
2.6
 
 
 
24


5


1

Obama
EV
 
 
 
 
 
55
 
9
7
3
3
 
 
 
4
 
21
 
 
7
 
 
 
4
 
10
12
17
10
 
 
 
 
 
 
4
 
15
5
31
 
 
 
20
 
7
21
4
 
 
 
 
 
 
 
3
13
11
 
 
 
 
296
 
 
 
states


states


state

Flip
To


 
 
 
 
 
Obama
 
 
 
 
 
 
 
 
 
 
 
Obama
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Obama
 
 
 
 
Obama
 
 
 
 
 
 
 
 
 
 
 
 
Obama
 
 
McCain
 
 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 


Obama
 
36
41
39
33
47
 
48
52
90
50
40
 
39
61
39
60
43
 
47
34
32
41
51
 
53
51
37
53
39
 
45
39
29
40
48
 
56
44
49
43
38
 
48
34
52
45
53
 
45
34
31
39
27
 
63
49
54
35
43
35
 
45.1


McCain
 
55
50
50
57
40
 
42
35
9 
41
50
 
53
31
52
31
51
 
38
55
57
52
38
 
40
39
41
38
54
 
48
47
57
46
43
 
32
44
38
51
44
 
39
57
38
43
38
 
48
51
58
52
62
 
29
42
42
53
47
54
 
44.1


Diff
 
( 19 )
( 9 )
( 11 )
( 24 )
7 
 
6 
17 
81 
9 
( 10 )
 
( 14 )
30 
( 13 )
29 
( 8 )
 
9 
( 21 )
( 25 )
( 11 )
13 
 
13 
12 
( 4 )
15 
( 15 )
 
( 3 )
( 8 )
( 28 )
( 6 )
5 
 
24 
0 
11 
( 8 )
( 6 )
 
9
( 23 )
14 
2 
15 
 
( 3 )
( 17 )
( 27 )
( 13 )
( 35 )
 
34 
7 
12 
( 18 )
( 4 )
( 19 )
 
1

Obama
EV
 
 
 
 
 
55 
 
9 
7 
3 
3 
 
 
 
4 
 
21 
 
 
7 
 
 
 
4 
 
10 
12 
 
10 
 
 
 
 
 
 
4 
 
15 
 
31 
 
 
 
20 
 
7 
21 
4 
 
 
 
 
 
 
 
3 
13 
11 
 
 
 
 
274
 
_______________________________________________________________________________________________________________________________________________
  
Latest Polls - May 29	
      Obama   McCain	
AL	36	55	Capital Survey-9
AK	41	50	Rasmussen-1
AZ	39	50	Rocky Mtn Poll-9
AR	33	57	Rasmussen-1
CA	47	40	LA Times-2
 
CO	48	42	Rasmussen-1
CT	52	35	Quinnipiac U.-6
DC	90	9	Election 2004-1
DE	50	41	SurveyUSA-3
FL	40	50	Rasmussen-1
 
GA	39	53	Rasmussen-1
HI	61	31	SurveyUSA-3
ID	39	52	SurveyUSA-3
IL	60	31	SurveyUSA-3
IN	43	51	Research 2000-4
 
IA	47	38	SurveyUSA-2
KS	34	55	Rasmussen-1
KY	32	57	Rasmussen-1
LA	41	52	Rasmussen-1
ME	51	38	Rasmussen-1
 
MD	53	40	SurveyUSA-3
MA	51	39	Rasmussen-1
MI	37	41	SurveyUSA-2
MN	53	38	Rasmussen-1
MS	39	54	Research 2000-3
 
MO	45	48	SurveyUSA-3
MT	39	47	Mason-Dixon-3
NE	29	57	Research 2000-3
NV	40	46	Rasmussen-1
NH	48	43	Rasmussen-1
 
NJ	56	32	Monmouth U.-5
NM	44	44	SurveyUSA-3
NY	49	38	Siena Coll.-3
NC	43	51	SurveyUSA-3
ND	38	44	Dakota Wesleyan U.-11
 
OH	48	39	SurveyUSA-3
OK	34	57	SurveyUSA-3
OR	52	38	Rasmussen-1
PA	45	43	Rasmussen-1
RI	53	38	SurveyUSA-3
 
SC	45	48	SurveyUSA-3
SD	34	51	Dakota Wesleyan U.-11
TN	31	58	Rasmussen-1
TX	39	52	Research 2000-3
UT	27	62	Dan Jones-7
 
VT	63	29	SurveyUSA-3
VA	49	42	SurveyUSA-3
WA	54	42	SurveyUSA-1
WV	35	53	SurveyUSA-3
WI	43	47	Rasmussen-1
WY	35	54	SurveyUSA-3

[tiptoe]

1. The 2004 Election Model projected Kerry 51.8-48.2 with 337 EV

Election Models

[DemocratSinceBirth]

2. If You Believe Elections Are Fixed You Have To Be A Sucker To Vote

It would be akin to betting on wrestling matches...

[tiptoe]

3. "In order to win, we MUST have a MASSIVE turnout like in 2006 to counter the machines. It CANNOT be

a close race." --helderheid

[DemocratSinceBirth]

4. If You Can Steal A Close Race Couldn't You Steal One That Is Not Close?

~

[tiptoe]

8. not as easily undetectable. nt

[tiptoe]

10. k

[Hippo_Tron]

9. It is much more difficult to steal a race that isn't close

Look at Florida in 2000. Between the convicted felons list, the butterfly ballot, hanging chads, etc. Gore was probably cost several thousand votes. If he had won the state by a million votes, none of it would have mattered.

[GoesTo11]

5. Bad analysis

95% chance of winning? It all depends on your assumptions.

First assumption: Obama captures 60% of undecideds.
Second assumption: Polls they happen to pick are right plus or minus sampling error.
Third assumption: Turnout doesn't favor either candidates' supporters.


I tend to think that their baseline assumptions might hurt Obama -

assumption 1 is about right given the mood of the electorate
assumption 2 is misleading because Obama and Clinton have been slugging it out, so his numbers should improve.
assumption 3 - I think Obama will have better GOTV than McCain - it's his strength and Republicans are discouraged.

But all these assumptions also add noise that isn't captured in the Monte Carlo model, so all in all, it's probably too optimistic. Although I'm optimistic, I'm not that optimistic.

[DemocratSinceBirth]

6. We Have Been Down This Road Before

I'm weary about even getting back in this debate...

In 2004 TruthIsAll did the same thing...He used the same methodology and claimed or predicted that a Kerry win was a lead pipe cinch...There were several of us at DU who demurred from that assessment and said the race was essentially a pick em... We were vindicated by history...


Truth Is All's retort was that he was right and that Kerry did win but Bush* stole the election...

Then the argument becomes circular...That's why I'm weary of re-entering it...

As to the current race, every metric favors the Democrat yet McSame is running a competitive race...That's where we are now...

[tiptoe]

7. "We were vindicated by history..." More like duped by a pollster's methodology of "forcing" Final

exit polls to match Recorded Vote counts, implying elections exit-polled in this manner are themselves implicitly fraud-free...this, despite the fact that one of the pollster's own category-forcings in 2004 (How Voted in 2000) was shown to be a MATHEMATICAL IMPOSSIBILITY! (It was impossible, because the recorded vote to which the Final exit poll was being "adjusted" was a fraudulent number.) TruthIsAll did not retort, merely, that "he was right...Kerry won...the election was stolen". He provided numerous analyses: The follow-up DU-Game -- a challenge to establish a quantitative rationale for the 2004 Recorded Vote margin -- persuaded me of the implausibility of Bush's 3-million-vote "victory". As for "proof", TIA put it this way:

Some argue that exit poll discrepancies from the recorded vote can't prove that fraud occurred. But that's not a valid criticism. The question should be:

Do the exit polls, in conjunction with the pre-election and approval polls, indicate that fraud was likely?

[OnTheOtherHand]

13. and the answer is: no, they don't

The pre-election polls on average showed Bush slightly ahead -- so TIA massaged them.

The OLS best fit between approval ratings and incumbent vote shares shows that Bush's approval rating was good enough to win -- so TIA pretends(?) not to know what OLS is. (Of course he could suggest a more sophisticated method.)

The exit polls in conjunction with the pre-election polls show no (or negative) relationship between exit poll "red shifts" and "surprise" compared to the pre-election polls. That is, Bush didn't do better (compared to pre-election polls) in states with big red shifts than in states with no red shifts. No one has explained why that would be true if red shift reflected vote miscount.

At the precinct level, the exit polls showed no relationship between red shift and "swing" compared to 2000. That is, Bush didn't do better (compared to 2000) in precincts with big red shifts than in precincts with no red shifts. Again, no one has explained why that would be true if red shift reflected vote miscount.

You claim to have been persuaded by the DU Game, but you have never said how or why. The heart of the DU Game is that people often misreport having voted for the incumbent; there is a great deal of evidence for that. Maybe you are convinced that it can't have been true in 2004, but that is no reason for anyone else to be convinced.

:shrug:

[DemocratSinceBirth]

15. "The pre-election polls on average showed Bush slightly ahead ..."

Res ipsa loquitur:



http://www.pollingreport.com/2004.htm#Pollster


I am weary of these debates...I spent yesterday morning disabusing a gentleman of the notion that African Americans will comprise 80% of the Mississippi electorate in the Fall...

[OnTheOtherHand]

17. exhaustion is TIA's modus operandi

He simply wears out anyone stubborn enough to correct him.

Interesting about Mississippi. I think a lot of people have in the back of their minds that black turnout in the South has hardly changed since the 1960s. I'm glad that isn't so, although a corollary is that the 80% scenario is wildly unrealistic.

[DemocratSinceBirth]

18. I Have Researched It (A Bit)

African Americans comprised 34% of the 04 Mississippi electorate...That's only a bit less than their total percentage of the population which is 37.1%... Kerry lost because he got 14% of the white vote, not because he didn't get enough African American votes or the African American vote was supressed or depressed... The numbers all across the south are more or less uniform... That's why I remain skeptical that outside of VA and FL which are atypical southern states the South is in play...And as a Floridian I doubt Florida is in play...

[OnTheOtherHand]

19. I pretty much agree with all that

Right now I wouldn't expect Florida to be in play unless it doesn't matter anyway. This isn't 2000 any more. Of course, on June 1 it is hard to be sure of anything.

[tiptoe]

11. 5/31 Election Model (Monte Carlo Simulation): Obama 52.1%; 304 EV; 98% Win Prob

2008 Election Model
Monte Carlo Electoral Vote Simulation

TruthIsAll

Updated: 31-May
     http://www.geocities.com/electionmodel/2008ElectionModel.htm

If a fraud-free election were held today, there is a   98%  probability that Obama would defeat McCain.
The probability is based on a Monte Carlo Simulation consisting of 5000 election trials using the latest state polls.

Obama leads the state poll weighted average by  45.8–43.5%  and the national polling average by  46.0–44.2%

The State Model projects that Obama will win a  304–234  electoral vote margin with  52.1%  of the 2-party vote.
The National Polling Model projects that he will win  51.9%  of the vote.
There is just a  0.2%  difference between the models.

The Monte Carlo EV Simulation Model generates five scenarios of 5000 election trials each for the following undecided voter assumptions:
Obama captures  50%,  55%,  60% (the base case),  65%  and  70%  of the undecided vote.

Note: The model does not take into account the fraud factor. In 2004, the Election Model projected Kerry would win 337 electoral votes and 51.8% of the 2-party vote. The pre-election model closely matched the unadjusted exit polls. But Bush won 51.2% of the 2-party vote.


State Polling Data Source:
http://www.electoral-vote.com/

National Polling Data Source:
http://www.realclearpolitics.com/epolls/2008/president/us/general_election_mccain_vs_obama-225.html


State Model - Simulation Summary

	Projection	Elec. 	Pop.	Win 
	                Vote    Vote%   Prob%
	Obama 		304	52.1	98.3	
	McCain		234	47.9	 1.7	
 
                             Undecided Voter Allocation
                                     Base Case
Obama	                50%	55%	60%	65%	70%  
Trial Wins              4146	4677	4915	4986	4999 

 									
Projected 2-party %										
Obama	                51.0	51.5	52.1	52.6	53.1
McCain	                49.0	48.5	47.9	47.4	46.9

 									
Win Probability %										
Obama	                83	94	98	100	100 
McCain			17	6	2	0	0   

 
Electoral Vote									
Average                 286	295	304	313	322  
Median  	        287	295	305	314	320  
 
Maximum	                334	359	383	375	390           
Minimum	                222	244	249	248	267           

 									
95% Conf. Interval								
Upper Limit	        315	324	333	342	349  
Lower Limit	        257	265	275	285	294  

 
States Won	        23	24	25	25	26  

National Model

 	
								  Base Case
		        National Model          		  Projection
			Date	    Sample  MoE	  Obama  McCain	  Obama	 McCain
Average			5/15-29	    9491    1.0%    46   44.2     51.9	 48.1

				    	    
Pew Research		5/21-25	    1242    2.8%    47   44	  52.4	 47.6
Gallup Tracking		5/24-29	    4368    1.5%    46   45	  51.4	 48.6
Rasmussen Tracking	5/26-29	    1600    2.5%    43   46	  49.6	 50.4
Newsweek		5/21-22	    1205    2.8%    46   46	  50.8	 49.2
Reuters/Zogby		5/15-18	    1076    3.0%    48   40	  55.2	 44.8

Latest State Polling

State Polls before Undecided Voter allocation


State

Total
 
AL
AK
AZ
AR
CA
 
CO
CT
DC
DE
FL
 
GA
HI
ID
IL
IN
 
IA
KS
KY
LA
ME
 
MD
MA
MI
MN
MS
 
MO
MT
NE
NV
NH
 
NJ
NM
NY
NC
ND
 
OH
OK
OR
PA
RI
 
SC
SD
TN
TX
UT
 
VT
VA
WA
WV
WI
WY
 


EV

538
 
9
3
10
6
55
 
9
7
3
3
27
 
15
4
4
21
11
 
7
6
8
9
4
 
10
12
17
10
6
 
11
3
5
5
4
 
15
5
31
15
3
 
20
7
7
21
4
 
8
3
11
34
5
 
3
13
11
5
10
3
 
Kerry
vs.
Bush

48.8
 
37.1
36.8
44.7
45.1
55.0
 
47.6
55.3
90.5
53.8
47.5
 
41.6
54.4
30.7
55.2
39.6
 
49.7
37.1
40.0
42.7
54.6
 
56.6
62.7
51.7
51.8
40.5
 
46.4
39.5
33.2
48.7
50.7
 
53.4
49.6
59.3
43.8
36.1
 
48.9
34.4
52.1
51.3
60.6
 
41.4
39.1
42.8
38.5
26.7
 
60.3
45.9
53.6
43.5
50.2
29.7
 
Obama
vs.
McCain

52.0
 
41.4
46.4
45.6
39.0
59.8
	   
54.0
59.8
90.6
55.4
46.0
	     
43.8
65.8
44.4
65.4
46.6
	     
56.0
40.6
38.6
45.2
57.6
  
57.2
57.0
50.2
58.4
43.2
  
49.2
47.4
37.4
48.4
53.4
  
63.2
51.2
56.8
46.6
48.8
  
55.8
39.4
58.0
52.2
58.4
  
49.2
43.0
37.6
44.4
33.6
  
67.8
54.4
56.4
42.2
54.0
44.2
  


Obama
leads

Flip to
Obama

McCain
leads


Diff

3.2
 
4.3
9.6
0.9 
(6.1)
4.8 
 	
6.4 
4.5 
0.1 
1.6 
(1.5)
 	
2.2 
11.4 
13.7 
10.2 
7.0 
 	
6.3 
3.5 
(1.4)
2.5 
3.0 
 	
0.6 
(5.7)
(1.5)
6.6 
2.7 
 	
2.8 
7.9 
4.2 
(0.3)
2.7 
 	
9.8 
1.6 
(2.5)
2.8 
12.7 
 	
6.9 
5.0 
5.9 
0.9 
(2.2)
 	
7.8 
3.9 
(5.2)
5.9 
6.9 
 	
7.5 
8.5 
2.8 
(1.3)
3.8 
14.5 
 
 
 
 
25


5


0

Obama
EV
 
306
      
      
    
    
    
55
    
9
7
3
3
    
    
    
4
    
21
    
    
7
    
    
    
4
    
10
12
17
10
    
    
    
    
    
    
4
    
15
5
31
    
    
    
20
    
7
21
4
    
    
    
    
    
    
    
3
13
11
    
10
    
    
 
 
 
states


states


state

Flip
To


 
 
 
 
 
 
 
Obama
 
 
 
 
 
 
 
 
 
 
 
Obama
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Obama
 
 
 
 
Obama
 
 
 
 
 
 
 
 
 
 
 
 
Obama
 
 
 
 
 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 


State

Total
 
AL
AK
AZ
AR
CA
 
CO
CT
DC
DE
FL
 
GA
HI
ID
IL
IN
 
IA
KS
KY
LA
ME
 
MD
MA
MI
MN
MS
 
MO
MT
NE
NV
NH
 
NJ
NM
NY
NC
ND
 
OH
OK
OR
PA
RI
 
SC
SD
TN
TX
UT
 
VT
VA
WA
WV
WI
WY


Obama
 
45.8
 
36
41
39
33
52
 
48
52
90
50
40
 
39
61
39
60
43
 
47
34
32
41
51
 
53
51
37
53
39
 
45
39
29
40
48
 
56
44
49
43
38
 
48
34
52
45
53
 
45
34
31
39
27
 
63
49
54
35
48
30
 


McCain
 
43.5
 
55
50
50
57
35
 
42
35
9 
41
50
 
53
31
52
31
51
 
38
55
57
52
38
 
40
39
41
38
54
 
48
47
57
46
43
 
32
44
38
51
44
 
39
57
38
43
38
 
48
51
58
52
62
 
29
42
42
53
42
53
 
 
Obama
leads


Diff
 
2
 
(19)
(9)
(11)
(24)
17
 
6
17
81
9
(10)
 
(14)
30
(13)
29
(8)
 
9
(21)
(25)
(11)
13
 
13
12
(4)
15
(15)
 
(3)
(8)
(28)
(6)
5
 
24
0
11
(8)
(6)
 
9
(23)
14
2
15
 
(3)
(17)
(27)
(13)
(35)
 
34
7
12
(18)
6
(13)
 
 
 
23

Obama
EV
 
284
 
 
 
 
 
55
 
9
7
3
3
 
 
 
4
 
21
 
 
7
 
 
 
4
 
10
12
 
10
 
 
 
 
 
 
4
 
15
 
31
 
 
 
20
 
7
21
4
 
 
 
 
 
 
 
3
13
11
 
10
 
 
 
 
states
 

[gristy]

12. 538 has it a lot closer - Obama 271EV McCain 267EV

http://www.fivethirtyeight.com/

[OnTheOtherHand]

14. 538 doesn't put a thumb on the numbers n/t

[VotesForWomen]

16. heh. bookmarking. nt

[tiptoe]

20. hey. kick. nt