Posted on 10/30/2004 11:56:36 AM PDT by PubliusEXMachina
Probability analysis by LatestPollResults.com using Real Clear Politics data Date/Time Bush Win Kerry Win 269-269 Tie 10-29-04 87.22% 10.97% 1.81% 10-28-04 76.58% 21.73% 1.69% 10-27-04 80.24% 18.12% 1.64%
thanks, nice post
Any chance of deciphering your data for those of us who don't read ASCII code?
Interesting. I've always had a gut suspicion that these "nationwide" polls were actually San Francisco/Greenwich Village polls.
That one is on my favorites now!!
Great find, thanks for the post. I'll be very interested to see if their prediction is accurate. Interesting that only PA and NJ (NJ!) are "tied."
Sorry, hit the link and the info will be there for you.
Probability analysis by LatestPollResults.com using Real Clear Politics data | |||
---|---|---|---|
Date/Time | Bush Win | Kerry Win | 269-269 Tie |
10-29-04 12:01am PDT | 87.22% | 10.97% | 1.81% |
10-28-04 12:01am PDT | 76.58% | 21.73% | 1.69% |
10-27-04 12:01am PDT | 80.24% | 18.12% | 1.64% |
Bush Electoral Votes
1 standard deviation (65% of probable outcomes) range is about 250 to 330
2 standard deviations (95% or probable outcomes) is about 230 to 350.
Median (most probable outcome) is about 291.
According to this calculation, 78.1% of the probable outcomes are Bush gets more than 270 Electoral Votes.
Everybody is forgetting about that faithless elector in West Virginia. Heard about him lately?
Median outcome (highest probability) = (291-"1 faithless elector") = 290
Just for fun, I ran a different check. I took all the states listed on realclearpolitics.com as toss ups. I wrote a short computer program that tried all 128 combinations of the outcomes in those states, and totaled EVs for Bush and Kerry.
The results? 89 cases result in a Bush win, vs. 39 in a Kerry win. That's 69.5% for Bush.
Of course, this assumes all the states are really 50-50, which I don't believe. I think Florida will likely go Bush and PA likely go Kerry. Assuming those two cuts us down to 32 cases. Bush wins 27 of them and Kerry wins only 5. That's 84%+ for Bush.
Just some idle diddling in VB...
An analysis is only as good as the data going into it. If you take garbage like the Zogby polls, Research2000 polls or ARG polls in the south, and shove it into the analysis, you come out with garbage.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.