8/21: ELECTION SIMULATION MODEL: Kerry: 307 EV

[TruthIsAll]

Kerry's state polling numbers have dropped the last few days, resulting in his simulation electoral vote average falling from 337 electoral votes to 307. His EV win probability fell accordingly from 99% to 86%.

Except for CBS, no national polls were released, so his National Win probabilities are still high.

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

August 21, 2004

STATE POLL EV SIMULATION MODEL
Proj. Kerry EV 307

Current Wtd. Mean:
Kerry 45.76
Bush 45.96
Projected Vote:
Kerry 50.73
Bush 49.27

Kerry Win Prob 86.60

National Polling data:

NATIONAL 11 POLL MODEL
Current Avg. Mean:
Kerry 49.67
Bush 44.11
Projected Vote:
Kerry 53.40
Bush 46.60

Kerry Win Prob 97.68

NATIONAL 15 POLL MODEL
Current Avg. Mean:
Kerry 48.60
Bush 44.60
Projected Vote:
Kerry 52.68
Bush 47.32

Kerry Win Prob 94.21

[iconoclastic cat]

1. Ah, TIA---I was wondering where you'd been.

Where do you stand on the K/E04 trajectory?

[TruthIsAll]

2. Kerry's current dip will be arrested as AWOL Bush is exposed..

This smear is backfiring on Bush, as all others have. He has nothing left and when the evidence of his desertion comes out, he will be toast.

Kerry will get 53-55%, depending on turnout. With 380 EV.

My 2c

[iconoclastic cat]

4. You strike me as someone who has some exp with statistical models.

In general, what is your background? You put an amazing amount of work into the info and graphics on your webpage, so I assume that you must have more than a passing fancy with this whole thing.

[TruthIsAll]

5. FYI - my background

BS Math
MS Applied Math
MS Operations Research

5 years in defense/aerospace as manufacturing engineer/Fortran programmer (IBM 370).

10 Years on Wall Street as Investment Banking Manager/software developer (Fortran/Lotus 1-2-3).

15 years as Financial software consultant (I use Excel/VBA for this model).

Lifelong JFK Democrat.

I am maintaining this to inform and because no one else is using latest state polling data for a Monte Carlo simulation to determine win probabilities.

My goal: present ALL relevant data, models and graphs on ONE page.

[Rowdyboy]

6. BTW-Thanks for your updates

I check them religiously

:thumbsup:

[Dogmudgeon]

3. I just looked at your site

This is good stuff.

With a more "fine-grained" model, you could probably hire yourself out as a statistical consultant. I know that we could use it in Pennsylvania, where Specter v Hoeffel is the big contest, and is being wildly spun by every pollster, pundit and paper in the state.

--bkl

[oscar111]

7. question

the first tally is bush ahead, .... is this voters in only battleground states? If it were all states, the diff ought to be more like 5, as in the natl tally at the end of your post.

[TruthIsAll]

8. Bush has attained a slim national lead based on the STATE

polls in ALL states (link to electoral-vote.com). The national vote% is based on the state polls weighted by voter population. As you can see in the graph, this reflects the 2% Kerry decline (mostly due to the latest CA poll) in the last few days.

My base case projection assumes Kerry will win 60% of the undecided/Nader vote. He has 50.73% after these are added to his total.

The National polls show Kerry with a 5 point lead. The main reason for the discrepancy: only one national poll (CBS) has been released in the last few days. The state polls are updated daily at electoral-vote.com.

[TruthIsAll]

9. Charts have been updated for more clarity.

tia

[TruthIsAll]

10. Kick...No change in state polls today (Sunday)

tia

[TruthIsAll]

11. Kick for Wolf Blitzer

tia

[mconvente]

12. At least we still got a lead

Great work TIA. I checked out your site and then went to this link:

http://www.pollingreport.com/k.htm

This stuff is interesting here. In almost every poll, Kerry's favorable rating has actually gone up in early August (when the SBVFL ads started). Only one poll has Kerry's favorability rating going down (and it was only by 1% - not bad at all). Hell, even the Faux News poll has Kerry's favorability rising 8% from Late July to Early August. Those bullsh*t ads are backfiring!