Calculating the probability that Bush stole the 2004 election....

[TruthIsAll]

assuming Kerry was leading in the final polls.

Just check the final state polls, run the simulation and
compute the probabilities.

I just ran 100 trial simulations for Kerry winning with 50%,
51%, 52%, 53%, 54% and 55% of the vote 

If Kerry has a 51%-49% lead in the final polls, he has a 90%
probability of winning, with an average electoral vote total
of 295.
If he loses, there is a 90% probability that Bush stole it.

If Kerry has a 53%-47% lead in the final polls, he has a 99%
probability of winning, with an average electoral vote total
of 330.
If he loses, there is a 99% probability that Bush stole it.

If Kerry has a 55%-45% lead in the final polls, he has a 100%
probability of winning, with an average electoral vote total
of 361.
Thus, if Kerry loses, it would be a near CERTAINTY that Bush
stole it - again.

Here is the table of final Kerry poll numbers and
corresponding probabilities of winning a FAIR election:
Kerry  Probability
Poll%   of Win
50% -  55%
51%  - 90%
52%  - 98%
53%  - 99%
54%  - 99.5
55%  - 99.999%
..................................................................

Here is the 55% case, displaying the Electoral Votes for 100
trial simulations.

TIA Daily Kerry EV Forecast/ Simulation			
			
State Polls
         Pct	ElectVotes	
Kerry	55.03%	361	
Bush	44.97%	177	

Kerry winning %: 100%	of 100 trial simulations	
			
National Polls:			
July	Act.	Adj.	
Kerry	49.1	54.9	
Bush	45.0	45.1	
			
Avg. Bush Job Rating 48.0	
Undecided % to Kerry 99%
			
Simulation Using Latest State Polling			
SIM	Electoral Votes		
1	398	Win	
2	416	Win	
3	351	Win	
4	402	Win	
5	388	Win	
6	361	Win	
7	394	Win	
8	358	Win	
9	317	Win	
10	332	Win	
11	360	Win	
12	364	Win	
13	370	Win	
14	369	Win	
15	394	Win	
16	371	Win	
17	397	Win	
18	410	Win	
19	386	Win	
20	420	Win	
21	422	Win	
22	311	Win	
23	374	Win	
24	343	Win	
25	392	Win	
26	397	Win	
27	361	Win	
28	354	Win	
29	356	Win	
30	363	Win	
31	366	Win	
32	354	Win	
33	361	Win	
34	371	Win	
35	366	Win	
36	338	Win	
37	357	Win	
38	358	Win	
39	358	Win	
40	374	Win	
41	394	Win	
42	363	Win	
43	374	Win	
44	395	Win	
45	376	Win	
46	368	Win	
47	383	Win	
48	391	Win	
49	352	Win	
50	333	Win	
51	347	Win	
52	359	Win	
53	389	Win	
54	328	Win	
55	350	Win	
56	365	Win	
57	339	Win	
58	406	Win	
59	405	Win	
60	311	Win	
61	377	Win	
62	370	Win	
63	374	Win	
64	352	Win	
65	388	Win	
66	397	Win	
67	303	Win	
68	382	Win	
69	377	Win	
70	384	Win	
71	356	Win	
72	363	Win	
73	358	Win	
74	320	Win	
75	340	Win	
76	353	Win	
77	373	Win	
78	366	Win	
79	336	Win	
80	321	Win	
81	353	Win	
82	351	Win	
83	347	Win	
84	344	Win	
85	337	Win	
86	349	Win	
87	355	Win	
88	299	Win	
89	327	Win	
90	350	Win	
91	331	Win	
92	357	Win	
93	336	Win	
94	350	Win	
95	290	Win	
96	332	Win	
97	358	Win	
98	321	Win	
99	379	Win	
100	341	Win	

[bigwillq]

1. How long did it take you to calculate this?

Just wondering. Some pretty interesting stuff, some of it is a little confusing but it's a lways good to try and learn something new.

[TruthIsAll]

2. This is a work in progress. The calculation is instantaneous.

I created an Excel model using the random number function for the MonteCarlo simulation.

[salinen]

3. Here is my calculation

2+2 = the fucking bastards will try to steal it again.

[TruthIsAll]

4. This analysis will be extremely useful in monitoring the

election. We need to track ALL the polls, state and national.

What is key is the % SPREAD between Kerry and Bush. If it grows, and it will, Nader will soon become a non-issue.

So far, state and national polls are showing a Kerry trend which may bring him a 10% lead. That would be 55/45 - without Nader. If that happens, there is ZERO chance that Bush can win fairly. He'll need Diebold. But it would be too obvious.

In fact, even if Kerry is up by just 6% (53/47), Bush still has virtually no chance of winning, unless...Diebold comes to the rescue in selected states. Again, it would be obvious.

With Kerry up by only 51-49, Bush will probably pull it off, but even here, the odds would be 90% that he cheated.

[Wednesdays]

5. Statistics don't mean squat if the votes aren't counted.

I'm not pooh-poohing your work--you're a genius. It's just that the criminals occupying the White House are probably way ahead of you, and will do everything they can to skew the numbers their way.

They get rid of enough Dem votes and/or manufacture enough Repug votes, they win...and there won't be any way to show they cheated by using probability. If the election numbers are different from opinion polls, then all sorts of excuses will be made for that difference. Remember how the exit polls were trashed in 2000 and 2002?

[TruthIsAll]

7. Excuses won't cut it; Diebold could. We should be prepared for anything...

by realizing that BushCo will HAVE to cheat if even if Kerry has just a small lead. Because they can't assume that they will win when the odds are 9-1 that they will lose.

You know they will cheat. So do I. If they are even or behind in the hundreds of forthcoming polls, they will surely cheat.

The further behind they are in the polls, the more desperate they will be. If they are close (within 3%), think Diebold. If they are farther behind, think martial law.

Solid circumstantial, mathematical evidence is always good to have before (keep vigilant) and after the fact.

[ClassWarrior]

10. This is awesome TIA!

Be sure to post again as the election nears.

[Cookie wookie]

6. I'm math impaired

However, I'd like to share this with others who aren't. Is that okay? Maybe we could use this in Georgia to prove taht they steal the vote here (I have no doubt they will).

BTW: could you compare Max Cleland's polling numbers from the 2002 election with one of your formulas? I'll have to search around for them on the net, but the polls just days before the election gave him a good lead.

[TruthIsAll]

8. This is an old post which calculated the probability that they

stole the Seanate in 2002. And Georgia was one of the four critical states which turned around at the last minute for the Repukes, after they had been trailing by more than the margin of error in each state.


The odds that that the Repukes stole those elections is better than 43,000 to 1.

http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=104&topic_id=85732

[newyawker99]

11. Hi desert!!

Welcome to DU!! :toast:

[linazelle]

9. Latest WSJ poll has 10-point spread

http://online.wsj.com/public/resources/documents/info-b...

Bush 43.9
Kerry 53.3
Nadir 0.7
Margin of Error +/- 4.2

[TruthIsAll]

12. Lunch kick

tia