9/19 ELECTION MODEL: KERRY 248 EV, 49.47% VOTE, 14% WIN PROB

[TruthIsAll]

DON'T PANIC!

Independent pollsters (Zogby, ARG, IBD, etc.) show the race TIED. However, the model uses ALL the polls, including the latest state polls from electoral-vote.com, which show Kerry trailing in a few key states which he must win.

Polls change daily. Kerry's National trend is up. The states will follow shortly.

Disregard the corporate pollsters whose agenda to defeat Kerry is as clear as day.





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

[molly]

1. There were 2 pollsters on C-Span

this morning. They both agreed that the real picture comes from averaging all of the polls. The misleading numbers come from the people polled. They don't tell you that they have polled mostly a pug base.

[DeepModem Mom]

2. Keep telling it like it is, TIA -- thanks -- (nt)

[unblock]

3. nice commentary :)

no doubt, many of your fans aren't quants like us, and therfore may not dig in past the headline numbers.

so a few words are very helpful.

[the Kelly Gang]

4. kick

[coloradodem2005]

5. What I don't get is....

...how we go from having a win probability in the high 90's to the low 10's. The thing is, it is the real picture of the election model. But it may not be the real picture of reality. Especially when you consider how they weight things and who they don't poll.


The other thing is: How do you follow a national trend? How does it show that Kerry is up? I would like to figure that out. What is Bush's national trend? Down like it always is?

[TruthIsAll]

7. The Kerry trend is up with independent pollsters. Link.

The corporate pollsters (Gallup, Newsweek, Time, etc.) have Bush ahead by 8-12%. The Indies have the race tied. Believe them.

See my post from yesterday:
http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=132x837238


To answer your questions:
The 14% probability is based on the latest state polls, from which I derive Kerry's expected EV using 5000 Monte Carlo simulations.

Look at the decline in Kerry's EV. from an expected 330 to his current 247. He needs 270.

With 330 EV, his prob is near 100%.

With 270 EV, his prob is slightly higher than 50% (since he has the advantage of the high EV states).

With 247 EV his prob drops to 14%.

Makes sense, no?

Look at the graphs. They tell the whole story.

[Owlet]

6. Aside from Election Day, the only polling number that counts...

...is George Bush's approval rating. This election, without a strong economic argument either way that's gonna have folks vote their pocketbooks, is a referendum on the incumbent. Right now his overall approval rating is 51%. If that holds up, he wins. If not, he loses. Some things really are as simple as they look. ;)

[TruthIsAll]

9. His average approval is 49.5%. See the chart.

tia

[Owlet]

11. Er.....what chart?

and are you looking at "job" or "favorability ratings"? You've been at this a lot longer than I have, and I'm not questioning your numbers. Just want to uderstand the guts of all this so I can be a better person.

BTB, do you agree with the notion that job approval is the only poll number worth watching?

[TruthIsAll]

13. I use Job Approval (see graph). Do NOT rely on it.

Job approval numbers can be biased as well from the corporate media pollsters who tend to inflate for Bush.

First look at the independents (Zogby, ARG, PEW, etc.)

So Bush's current 49.5% average may actually be lower by a couple of points.

tia

[TruthIsAll]

15. Bush average approval is 49.5%, NOT 51%

See the numbers and chart in the model.

[TexasSissy]

8. I can't help but get concerned when I see much models & polls.

I mean, if I get excited when Kerry's ahead in the polls, it stands to reason that I'd get concerned when I see him lagging in those polls.

I hope Kerry's campaign can pull the proverbial rabbit out of the hat....and he just might.

[Toots]

10. Your trend has been steadily down

I don't really mean your trend but only of what you have posted. I remember only a few weeks ago you had Kerry winning at 99% and almost every day the percentage gets less. Now it is at 14%. I hope it starts to reverse itself and get back to higher percentages. Possibly after a couple of the Debates if there are any. Anyway I believe Kerry is the better man for the job and hope most Americans will see that as well.

[TruthIsAll]

12. Small gains in these states will boost Kerry:247EV/14% PROB to 298/90%

It's NOT that bad, folks, IF Kerry can just improve his current state poll numbers a little bit in these states.

CO 44/45 to 46/45
FL 45/51 to 50/48
IA 46/48 to 49/48
NJ 45/49 to 48/44
OH 45/48 to 49/48
WI 43-49 to 47/44

[Gyre]

14. Aha! So that's the purpose of skewed polls.

To skew the estimates of the EV results! (demoralizing dems from voting). Then, that would make the EV estimates bullshit as well!

Stake through the hearts of the EV people who allow bogus polls into their equation!!!

Gyre

[TruthIsAll]

16. Good point, but the model shows us HOW much the EV results can be skewed

by slight "adjustments" to the state polls.

For THAT reason alone, it's useful.