9/7 KERRY: 284 EV, 50.34% POPULAR VOTE, 71% WIN PROBABILITY

[TruthIsAll]

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

[kevin881]

1. yesterday you said he was at "rock bottom"

Does it still go up from here? His probability was better yesterday!

[TruthIsAll]

2. I guess my timing was wrong. I'm not a seer. But he WILL rebound.

It's only a question of how soon.

I suspect we'll be smiling two weeks from now.

[nodictators]

3. TIA, drop the "win probability", keep the EV

The margin of error used by the pollsters is based on only their sampling error. There are many other errors in polls, such as over-weighting the Repubs, as we have seen in the now-infamous, recent Time poll.

[TruthIsAll]

5. The probability analysis is a key result of the model/simulation.

By utilizing all state and 15 national polls, the model overcomes many of the problems associated with outliers. Indeed, that is what makes the model unique.

[Poiuyt]

4. Hey TIA, could you put this in English for the non-mathematicians

I've always relied on http://www.electoral-vote.com/ to get my sense of how the election is shaping up. I don't understand how Kerry can have such a large win probability when he's behind in the electoral polling.

Thanks

[TruthIsAll]

6. Kerry is ahead. His average EV is 284 (for 5000 simulation trials).

This assumes the base case in which he gets 60% of the undecided vote.

Kerry can win with less than 50% of the popular vote since his strong states ((NY, CA, IL, NJ, etc) are richer in electoral votes. On the other hand, it's extremely unlikely that Bush can win with less than 50%.

I use latest state polling data from electoral-vote.com and employ Monte Carlo simulation using state probabilities (calculated from the polls)to generate thousands of potential electoral combinations.

The problem is to calculate the probability of Kerry getting over 270 EV. This is a much stronger result than just adding up the latest state polls to determine who is leading. It's more complicated than that.

[cthrumatrix]

7. great news..talking to people...Kerry is winning hands down

[TruthIsAll]

8. National and state polls agree: less than one percent

difference bewteen Kerry and Bush.

Bush has shot his load; he had a free ride in August. No ads from Kerry, swift boat garbage, the Repuke convention.

And there's only a ONE percent difference between them.

Kerry is in a great position to close the deal.

[seabeyond]

9. you said yestersday that 80% was the low

wink

[AndyTiedye]

10. With 50% of the Undecideds We Probably Lose, by your model

in all your previous runs, we could still win with that.

"Voters" who are undecided at this point are more likely to stay home than to
vote for Kerry. 60% is really optimistic, even though it may have been common
for challengers to do that well in the past, when we had a free press. It is no
longer possible for opposition candidates to get any significant airtime.

We need to find more Democratic voters, somewhere.

[TruthIsAll]

11. You are assuming that Kerry stays where he is now; he won't.

When he gets back into the lead (2-3%) he can win with less than 50% of the undecided.