8/7 KERRY WIN PROBABILITY: 98.05% ; VOTE: 53.11%; EV: 329

[TruthIsAll]

								
ELECTION FORECAST / SIMULATION MODEL								
Created by TRUTHISALL								
								
National Polling data: Pollingreport.com								
State Polling data: Electoral-vote.com 								
								
								
LATEST NATIONAL POLL AND FORECAST VOTE %								
Assumes 70% of undecided/other to Kerry								
								
______	15Polls	Forecast Vote %						
______	Nat. %	Nat.	State	Comb.				
Kerry:	48.53	53.39	52.82	53.11				
Bush:	44.53	46.61	47.18	46.89				
Spread:	4.00	6.77	5.65	6.21		
						
						
WIN PROBABILITIES						
______	Nat.	State	Comb.	State EV		
Kerry:	 98.41 	97.70	98.05	329		
Bush:	 1.59 	2.30	1.95	209		
						
						
NATIONAL VOTE/ WIN PROBABILITY SENSITIVITY ANALYSIS						
						
Impact of Kerry undecided/other allocation						
Alloc :	50%	60%	70%	80%	90%	
Vote %:	52.00	52.69	53.39	54.08	54.77	
Prob %:	89.72	 95.60  98.41  99.52 99.88 	
						
						
ELECTORAL VOTE/ WIN PROBABILITY SENSITIVITY ANALYSIS						
One thousand  (1000) trials are run in each simulation.						
						
Simulation I: Conservative Case 						
Assume:	60%	of undecided/other to Kerry.				
						
Kerry:	51.90%	of the vote.				
Wins :	93.9%	of 1000 election trials. 				
Avg  :	313	electoral votes.				
Max  :	397	electoral votes.				
						
						
Simulation II: Most Likely Case 						
Assume:	70%	of undecided/other to Kerry.				
						
Kerry:	52.82%	of the vote.	
Wins :	97.7%	of 1000 election trials. 	
Avg  :	329	electoral votes.	
Max  :	407	electoral votes.	
			
			
Simulation III: Optimistic Case 			
Assume:	80%	of undecided/other to Kerry.	
			
Kerry:	53.75%	of the vote.	
Wins :	99.6%	of 1000 election trials. 	
Avg  :	346	electoral votes.	
Max  :	434	electoral votes.	
			
			


LAST THREE PRESIDENTIAL ELECTIONS												
(in millions of votes)												
Dem	138.78	52.61%										
Rep	125.03	47.39%										

												

BUSH JOB APPROVAL												
The data source is PollingReport.com    												
												
2004	Avg	Nwk	Fox	CNN	Pew	Harris	CBS	ABC	Time	NBC	AP	Zogby
Jan.	54.50	50	58	60	56	na	50	58	54	54	56	49
Feb.	49.67	48	48	51	48	51	50	50	54	na	47	na
Mar.	48.75	48	48	49	46	na	51	50	na	50	48	na
Apr.	48.80	49	50	52	48	48	46	51	49	na	48	47
May	45.20	42	48	47	44	na	41	47	46	47	48	42
June	47.00	na	48	49	48	50	42	47	na	45	48	46
July	48.00	48	47	47	46	na	45	50	50	48	50	49
Aug	45.86	45	44	48	na	na	44	47	na	na	49	44
												
												
												
												
NATIONAL POPULAR VOTE TREND												
Based on latest national polls from:												
IBD, ABC, NWK, ARG, CBS, PEW, LAT, ZOGBY, TIME												
Data source: PollingReport.com    												
												
												
Kerry projection is the current average of these
polls												
plus 70% of undecided/other voters.												
						
	Avg Poll Trend		Projection			
2004	Kerry	Bush 	Diff	Kerry	Bush 	Diff
Jan 	44.0	49.4	-5.4	46.4	53.6	-7.1
Feb	48.0	45.4	2.6	52.6	47.4	5.2
Mar	48.6	44.0	4.6	53.8	46.2	7.5
Apr	47.4	44.9	2.5	52.8	47.2	5.6
May	47.8	44.0	3.8	53.5	46.5	7.1
June	47.1	44.7	2.4	52.8	47.2	5.7
July	48.1	45.0	3.1	52.9	47.1	5.9
Aug	49.5	43.8	5.7	54.2	45.8	8.3
						
						
						
KERRY NATIONAL POPULAR VOTE FORECAST/WIN PROBABILITIES 						
						
48.53%	 Average (mean) of most recent 15 national polls
												
4.85%	+ Kerry 70% allocation of undecided/other												
53.39%	 = Projected Kerry vote												
													
98.41%	Kerry probability of at least 50% of popular
vote												
													
______				Kerry%	Kerry% undecided/other 								
______	Date	Kerry	Bush	vs.Bush	50%	60%	70%	80%	90%				
													
TIME	722	50	45	52.6	52.5	53.0	53.5	54.0	54.5				
FOX	804	46	42	52.3	52.0	53.2	54.4	55.6	56.8				
CNN/GAL	801	48	48	50.0	50.0	50.4	50.8	51.2	51.6				
LAT	721	48	46	51.1	51.0	51.6	52.2	52.8	53.4				
PEW	718	46	44	51.1	51.0	52.0	53.0	54.0	55.0				
													
IBD	807	49	43	53.3	53.0	53.8	54.6	55.4	56.2				
CBS	730	49	43	53.3	53.0	53.8	54.6	55.4	56.2
DEMC	713	52	45	53.6	53.5	53.8	54.1	54.4	54.7
ABC/WP	802	52	45	53.6	53.5	53.8	54.1	54.4	54.7
NWK	730	52	44	54.2	54.0	54.4	54.8	55.2	55.6
									
ZOGBY	729	48	43	52.7	52.5	53.4	54.3	55.2	56.1
AP	807	48	45	51.6	51.5	52.2	52.9	53.6	54.3
NBC/WSJ	721	45	47	48.9	49.0	49.8	50.6	51.4	52.2
ARG	801	49	45	52.1	52.0	52.6	53.2	53.8	54.4
QPAC	722	46	43	51.7	51.5	52.6	53.7	54.8	55.9
									
Mean	______	48.53	44.53	52.15	52.00	52.69	53.39	54.08	54.77
Prob	______	______	______	91.31	89.72	95.60	98.41	99.52	99.88
									
Std	______	2.26	1.64	1.45	1.38	1.32	1.32	1.37	1.49
StdP	______	1.58	1.57	1.58	1.58	1.58	1.58	1.58	1.57
										
95% Confidence Interval										
Max	______	51.63	47.61	55.25	55.10	55.79	56.48	57.17	57.86	
Min	______	45.44	41.45	49.05	48.90	49.60	50.29	50.99	51.69	
x	______	47.61	47.61	50.00	50.00	50.00	50.00	50.00	50.00	
Prob>x	______	71.97	2.50	91.31	89.72	95.60	98.41	99.52	99.88	
										
99% Confidence Interval										
Max	______	52.61	48.59	56.22						
Min	______	44.46	40.48	48.07						
x	______	48.59	48.59	50.00						
Prob>x	______	48.61	0.49	91.31						
										
										
Notes:										
Std = Standard Deviation (variability) of polls										
StdP = Sqrt((1-p)*p/1000), where 1000 is average poll sample
size								
								
								
95% Confidence Limits:								
Max = Mean + 1.96 * StdP								
Min = Mean - 1.96 * StdP								
								
99% Confidence Limits:								
Max = Mean + 2.58 * StdP								
Min = Mean - 2.58 * StdP								
								
								
								
								
								
								
								
ELECTORAL VOTE SIMULATION FORECAST METHODOLOGY								

Kerry National and State Vote Projections								
State vote%: Latest state poll plus assumed share of
undecided/other voters.								
National vote%:Weighted sum of projected % state
votes.								
Probability of a state win is based on projected %
vote.								


Most Likely Case								
Assume 70% of undecided/other voters for Kerry								
Win	Prob.	Pct	EV					
Kerry	97.7%	52.8%	329					
Bush	2.3%	47.2%	209					

Historical Vote %, Kerry Projection and Win Probability
								
Dem Hist. is Democratic share of 2-party vote in last three
presidential elections.													
													
______	Dem	Kerry	Kerry	EV (trials 1-10 of 1000)									
______	Hist.	Proj.	Prob.	1	2	3	4	5	6	7	8	9	10
State	52.60%	52.82%	97.70%	319	309	260	360	368	325	319	260	302	322
													
AL	44.8	41.0	0.0										
AK	37.6	40.7	0.0										
AZ	48.8	49.9	49.0				10	10		10			10
AR	55.2	50.2	52.0	6	6	6	6	6	6	6	6	6	6
CA	57.4	57.3	96.6	55	55		55	55	55	55		55	55
													
CO	48.8	49.3	43.1	9		9	9	9	9		9		
CT	57.7	62.6	99.9	7	7	7	7	7	7	7	7	7	7
DE	56.8	57.1	96.2	3	3	3	3	3	3	3	3	3	3
DC	90.3	89.5	100.0	3	3	3	3	3	3	3	3	3	3
FL	50.7	54.9	89.0	27	27	27	27	27	27	27	27	27	27
													
GA	47.6	45.9	15.3		15			15					15
HI	59.0	60.9	99.7	4	4	4	4	4	4	4	4	4	4
ID	35.7	39.0	0.3										
IL	57.9	58.9	98.7	21	21	21	21	21	21	21	21	21	21
IN	44.8	45.9	15.3										
													
IA	51.8	52.2	70.9	7	7			7	7			7	7
KS	42.4	41.6	1.8										
KY	46.7	46.2	17.1						8		8		
LA	49.2	44.4	8.1										
ME	57.1	52.3	71.7	4	4	4		4	4	4	4	4	4
													
MD	57.8	58.6	98.4	10	10	10	10	10	10	10	10	10	10
MA	65.3	65.8	100.0	12	12	12	12	12	12	12	12	12	12
MI	54.7	56.9	95.8		17	17	17	17	17	17	17	17	17
MN	55.7	52.6	74.2			10	10	10	10	10			10
MS	44.3	36.3	0.0										
													
MO	52.5	51.1	60.8	11		11	11	11	11	11	11	11	11
MT	44.9	42.8	3.6										
NE	37.5	36.5	0.0										
NV	49.9	51.3	62.7	5	5	5		5		5		5	5
NH	51.7	55.3	90.7	4		4	4	4	4	4		4	4
													
NJ	56.5	59.5	99.1	15	15	15	15	15	15	15	15	15	15
NM	53.0	51.1	60.8	5	5		5	5		5	5		5
NY	62.6	65.0	100.0	31	31	31	31	31	31	31	31	31	31
NC	46.6	47.5	26.6		15		15			15			
ND	40.8	37.2	0.1										
													
OH	50.8	48.9	39.2									20	
OK	42.8	39.2	0.3										
OR	53.6	52.8	75.8	7		7	7	7	7		7	7	7
PA	54.2	55.2	90.3	21		21	21	21	21	21	21	21	21
RI	65.6	67.2	100.0	4	4	4	4	4	4	4	4	4	4
													
SC	44.4	47.5	26.6	8				8			8		
SD	44.5	44.8	9.7		3								
TN	50.5	50.2	52.0	11	11		11						
TX	44.3	42.6	3.2										
UT	33.6	29.7	0.0										
													
VT	59.4	60.1	99.4	3	3	3	3	3	3	3	3	3	3
VA	47.3	49.9	49.0				13	13			13		
WA	55.9	55.2	90.3	11	11	11	11	11	11	11	11		
WV	54.0	53.6	81.6	5	5	5	5		5	5		5	5
WI	52.7	51.9	68.3	10	10	10	10	10	10				
WY	38.3	30.8	0.0										
													
Avg	52.60%	52.82%	97.70%	319	309	260	360	368	325	319	260	302	322

[gristy]

1. Very good, TIA

It would seem that one could use your data to help identify states where the Dems should put some extra resources and plan some extra campaign visits.

[wtmusic]

2. unsound and speculative

60%-80% undecided will go to Kerry? Based on what?

The fact that we're still 3 months away from the election is not even accounted for in your methodology. By this logic we could predict 2008 results as well.

Not only are these predictions faulty but they encourage overconfidence.
Counterproductive.

[TruthIsAll]

3. On the contrary, they show the strength of Kerry's trend.

Your comments are typical of those who fail to understand the importance of probability theory. I do not predict anything. The only assumptions I make are in the undecided/other voter allocation.

Saying Kerry has a 97% probability of winning is a totally accurate statement - if you believe the polls and the 70% allocation assumption.

And if you believe Kerry will split the undecided vote with Bush, the odds still favor him (90%), because he is leading right now.

The 60-80% goes to Kerry because that is the historical average split for the challenger.

Basic logic should tell you that a voter who is undecided after four years of Bush is probably going to vote for Kerry. Did you even consider that?

By your logic, polling is useless. Tell that to the Kerry/Bush campaigns.

To say I can't forecast major events which may impact the election stating the obvious. But I can give you a better analysis of the polling trends then you will ever find in the whoremedia.

By the way, you can check out blogs who run election forecast sites (See the list on the right side of democrats.com home page).

They mostly agree with my numbers.

[wtmusic]

4. Again, you don't address the time factor

your probability calc would be well-served by including the disclaimer "if the election were held today".

Do you have anything to back up that 60-80% split going to the challenger? Seems to me we should only be looking at wartime presidencies, when the results might be considerably different.

[TruthIsAll]

5. The Incumbent Rule.

http://www.mydd.com/story/2004/8/2/111725/5894

As for the time factor, we are not dealing with option expiration dates here. Just take a look at the 9-poll trend analysis and extrapolate out to November.

The trend has Kerry breaking away from Bush, after holding a steady 3% lead since Feb.


Kerry projection is the average of these polls plus 70% of undecided/other voters.

Avg Poll Trend Projection
2004 Kerry Bush Diff Kerry Bush Diff
Jan 44.0 49.4 -5.4 46.4 53.6 -7.1
Feb 48.0 45.4 2.6 52.6 47.4 5.2
Mar 48.6 44.0 4.6 53.8 46.2 7.5
Apr 47.4 44.9 2.5 52.8 47.2 5.6
May 47.8 44.0 3.8 53.5 46.5 7.1
June 47.1 44.7 2.4 52.8 47.2 5.7
July 48.1 45.0 3.1 52.9 47.1 5.9
Aug 49.5 43.8 5.7 54.2 45.8 8.3


Assuming NO MAJOR EARTH-SHATTERING EVENT, LIKE BUSH CANCELLING THE ELECTIONS, I BELIEVE KERRY WILL GET 53-55% OF THE VOTE.

[wtmusic]

6. ??

"As for the time facor, we are not dealing with option expiration dates here."

So you're saying that the amount of time until the election is not a factor in the probability?

That's ridiculous. You need to hit the books again. :crazy:

[TruthIsAll]

7. I told you to extrapolate the trend. Do you know regression?

The trend is a factor. The trend is based on obvservation changes over time. I use the latest data to forecast probabilities.

You can look at the polling trend and extrapolate yourself, if you know the meaning of the word EXTRAPOLATE.

You will find, after you do this, that the chances are very high Kerry that will have a solid margin come November.

Will you do this? Or will you just continue to throw a strawman at my work?

I seem to recall your attempt to debunk some of my prior threads.

Let's see your election analysis. Or are you too busy for that?

[wtmusic]

8. So your contention is that

extrapolation based on the trends you present is equally valid for November, July 2012, and February 2060?

Your strawman charge is a strawman, as is your insinuation that I'm ignorant of what "extrapolation" means. I also think it's interesting your 97% probability has been reduced to "chances are very high Kerry that will have a solid margin come November", with which I would agree.

I have no election analysis to present. Unfortunately that doesn't preclude me from looking at other analyses critically.

[TruthIsAll]

9. You did not read my edited post.

The chances are 97% assuming Kerry gets 70% of the undecided/other.

The chances are 90% if he splits the vote with Bush. Still quite high.

We are forecasting THREE MONTHS out. Not TEN years. Get REAL.

You are becoming shrill in attempting to justify your arguments.

I have said all I will say on the subject. I am not a seer. I an sure my probabilities are mathematically valid, based on latest available polling data, both state and national. I believe the trend is solid for Kerry. I do believe he will get a solid majority of other/undecideds (including Nader).

And I believe you are full of gas.

[DemocratSinceBirth]

11. Here's Democrat Since Birth's Analysis

and for the record I have done post grad work in Government...


If Bush doesn't change the dynamic he will lose... One way the dynamic can be changed is for a terrorist attack to occur on U S oil....


Anybody who knows if a terrorist attack will help or hurt Chimpy would be speculating....



As for the seminal poster's model it is only valid if the election was held on the day of the polls he cited... To infer more from that would get you a failing grade in any research methods of social science course....

[goclark]

10. Question? Why are they sasying it is so close?

Why is the Media continually saying how close it i/to close to call?

[TruthIsAll]

13. Because the media is pro-Bush. They are also incompetent.

For instance, if Kerry has a 49-45 lead in one poll, they say its too close to call and within the MoE. That is true.

But when 15 polls show an average Kerry lead of 4 points, that is something quite different.

But they won't tell you that.

Its what they say (misleading, incomplete) and what they don't say (the full truth) that has this nation all fucked up.

[Selwynn]

19. you can't just "extrapolate out to November," which you admit

in your last sentence.

Assuming "major earth shattering event." What that should really says is, "assuming nothing unforeseen happens." Which should really say "assuming the election were held today, in a context and environment that we have knowledge of, this would be the likely result."

So yeah, assuming nothing changes between now and November, great. But that's a fry cry from saying there is a 98% probability that Kerry will win in November. That has all kinds of qualifications

1. Assuming my blanket assumptions about swing voters are correct
2. Assuming all the major variables in our nation stay the same between now and then
3. Assuming there is no huge debilitating scandal in the Kerry campaign
4. Assuming there is no October surprise
5. Assuming there is no major event like a terrorist strike, etc.


What we can say about this data, is that contrary to the media reports, if the election were held today, all indicators are that Kerry would win and win easily. This is a great place to be going in to November, but has nothing to do with certainty or even a great statistical likelihood, because the farther you speculate out, the more variables come into play, that must either be scientifically taken into account or undermine the accuracy of data.

[TruthIsAll]

20. No one is talking CERTAINTY, we are talking PROBABILITY.

There is a vast difference.

We CAN say, based on a variety of current polls, that Kerry has an extremely HIGH probability of a win.

If you assume he splits the undecided/other vote with Bush, he has a 90% CHANCE of winning.

If you assume he gets 70% of the undecided/other vote, he has a 97+% CHANCE of winning. End of story.

I have made NO predictions. I have just calculated probabilities.

I hope that finally sinks in. Just accept my mathematical analysis. Also accept that I am not predicting anything.

And I bet you have probably learned something from my post.

[Selwynn]

18. "The only assumption I make"

"The only assumptions I make are in the undecided/other voter allocation."

But that's a pretty huge and ungrounded assumption.

"Basic logic should tell you that a voter who is undecided after four years of Bush is probably going to vote for Kerry. Did you even consider that?"

Unfortunately that's a gaff in logic, not logic at all. Logic can refute that assumption by direct counter-example. Undecided voters may still not vote for Kerry but vote third party, or decide after four years of bush that they like Kerry less and vote for Bush - you just don't know and there's any argument for any one of those options is guesstimating - its not "basic logic." It's basic hypothetical inference, which could or could not be right.

"By your logic, polling is useless. Tell that to the Kerry/Bush campaigns."

By logic in general, not "my" logic, polls and probability data are often severely misused by people not trained enough to appropriately understand their limits. Polling is so often mis-characterized and misused as "factual" or "proving" this or that, that it is in many ways worthless. And more often that not, people see polling information as "logically" proving an absolute when they don't have a real good understanding of principle logic in the first place. It's sad, but true.

[TruthIsAll]

21. My 70% inference is a best guess, based on historical

allocations. That is MY assumption and I am entitled to make it.
I also provide a SENSITIVITY ANALYSIS to look at alternative assumptions (from 50-90%) so anyone can see how the probability is effected by applying THEIR own own allocation assumption.

I agree that polls are misused; that's why I provide a detailed probability analysis, using the latest 15 national polls and state polling data. I just calculate the means and deviations.

I also believe the media has a general Bush bias, and that's why I analyze two national poll groups.

The 9-poll analysis looks at polls which I believe to be least biased to Bush. I exclude NBC, CNN, AP, FOX. The 15-poll analysis makes no distinction and includes these polls.

I believe my model is more analytical than similar predictors on the Net. But maybe I'm a little biased there.

[Selwynn]

23. Right -and theres nothing wrong with best, intelligently made guesses..

..but its not "basic logic" that leads you there, is educated guessing.

I don't want this to be mistaken as confrontational. I think your analysis is VALUABLE, especially for confronting the lying media and the lying pundits that lie for it. I do try to think of the variables, limits, assumptions and contextual complexities associated with probability, but I am not slamming your work.

[Selwynn]

12. This data promotes one wrong assumption, though I'm still glad to have it

and that is that polls - even an average of a lot of them - are any kind of honest indicator of what will actually happen.

Polls are only accurate when there is a wide margin. The less the margin, the more innaccurate they become.

Of course, these infomation is awesome - I'm not just berating it. It is extremely cool to have the math that says, "if we just go honestly by the polls, then these are the statistics." But in terms of reality, in an election like this, with things so close and so many variables, polling data has little or no necessary relationship with reality, only possibly a coincidental one.

[DemocratSinceBirth]

14. I Disagree

If you have a dozen polls and they show a small lead for your candidate he is probably ahead as the cumulative effect of the redundancy of the polling data raises the confidence level...

[TruthIsAll]

15. The consistency of the polls says that there is a trend to Kerry.

One cannot deny this obvious fact. Whether or not it will continue, is another question. The only way to answer that is to track the polls frequently, checking the mean, std deviation and applying the normal distribution to determine the probability of a win.

Each poll has a margin of error. The MoE for a large number of polls is less than the individual MoE.

To disregard the polls, both approval and election, is silly.

All I am trying to do is to undo the Bushit that is thrown out by the know-nothing media, when they say the election is "too close to call".

I suggest you look at the historical relationship between Aug. job approval and election polls to the final result. There IS a relationship.

[DemocratSinceBirth]

16. I Agree With Your Analysis Of The Trend....

The only thing that can change the dynamic in this race is a terrorist attack or a mega Kerry scandal...

[Selwynn]

17. Undoing "election too close to call"

I know. I'm glad. But at the same time, the only time we'll know for certain who's going to win in November is once the voting booths close on election day.

...well theoretically, assuming the votes get honestly counted.

What your data rightfully does is show that all mainstream media is lying. Polls do not indicate a race too close to call. Polls indicate a no brainier win for Kerry if the election were held today. Naturally, things can change, and the American public is fickle. But you're right, you help show that the media is full of massive amounts of shit, and that's a good thing.

[TruthIsAll]

22. Let me correct one statement of yours..

You say:
Polls are only accurate when there is a wide margin. The less the margin, the more inaccurate they become.

That is totally false. It's just the opposite.

For instance if A leads B 80-20, there obviously a greater chance of A winning. But that is NOT the measure of accuracy. If the final result is A 75-25, yes we have correctly predicted the winner, but the error is 5%. That is beyond the typical 3% MoE.

OTOH, if A leads B 51-49, but B wins 51-49, that is a better forecast, because there is only a 2% error, within the MoE.

In fact, it is much more likely that a tight, hard-fought election will be within the MoE than a one-sided election. And that is why final deviations far exceeding the MoE, like Georgia in 2002, should be VERY suspect.

[Selwynn]

24. Interesting... oh and be sure to see #23

Thanks

[TruthIsAll]

25. Kick for the night crew

tia