State of the Race - From the State Level

[tanda]

The following analysis appear consistent with national polling and market-based indicators.

I track each state with a "super-poll" of all polls released with a sample date ending within the last seven days.

Based on these "super-polls", Bush leads 300-238. From the 200 results, he loses NH and wins HI, IA, NM and WI.

However, my analysis differs from that of realclearpolitics.com averages in that RCP has Bush ahead in MN. Thus, an analysis based on the RCP averages has it Bush 310-228.

Further however, Bush has a "soft" lead (less than 1% in the averages) in FL, HI and OH. Actually, FL has ticked up slightly for Bush and there are rumors that internal polling on both sides is better for Bush than the public polling. Nonetheless, I will leave FL in the "soft" lead category for now. Kerry only has a "soft" lead in MN. Ignoring these "soft" leads, Bush only leads 249-228.

This analysis assumes no surprises such as a flip of AR, CO, MI, NJ (new Quinnipiac poll has the race 46-46) or PA.

From this analysis, a win for FL for Bush gets him to 276 and he can now lose IA or NM. A win of OH gets him to 269 and the House of Representatives (as would happen with a win of FL and a loss of IA).

[El Barto]

You may be interested in this site. Stanford Electoral Probabilities

They calculate the probability of each candidate winning each state based on all the polling data, and then run all the different probability permutations.



The weakness is in states with too few polls (look at Colorado), but the figures for the hotly contested states are interesting.

[JimBob2232]

This is positively absurd. Polls show bush is up one in hawaii. Quinnipiac now has bush and kerry tied in NJ (ive been saying all along NJ is going to switch). Not saying in the end it will be wrong, but alot of those states have 100% chance, and there is no way you can be that confident.