This probability analysis differs from the Electoral Vote
Simulation analysis I have posted. The EV simulation
calculates the probability of Kerry winning each STATE based
on the latest STATE polls. The total electoral votes are
summed in 100 trial runs using Monte Carlo simulation.
This NATIONAL POLL analysis calculates probabilities of Kerry
winning the POPULAR vote based on the latest National Polls if
the election were held today.
I make the following assumption:
1)Conservative:Kerry and Bush split undecided/others (50/50
allocation)
2)Most likely: Kerry gets 70%
3)The MoE is 3.5% in each poll.
For a group of 10 polls, each with a MoE of 3.5%, the
combined MoE is lower.
Two sets of probabilities are shown:
The first is based on average of the polls since Feb.
The second is based on the latest polls (June/July).
I have separated FOX, AP and CNN from the other 10 polls to
illustrate the sharp differences in them.
The key results based on the latest pollsare as follows:
FOX, CNN, AP
The average Probability of a Kerry win is conservatively
40.7% and most likely 55.5%.
For the other 10 polls:
Probability of a Kerry win is 75.1% (again, this is
CONSERVATIVE).
The Most Likely win probability is 83.3%
Feb-July Polling Data Avg Latest Polls Prob (Kerry)
Poll Kerry Bush Diff Prob Kerry Bush 50% 70%
CNN 48.00 48.71 -0.71 0.458 50 46 0.779 0.867
AP 44.25 46.50 -2.25 0.362 45 49 0.272 0.410
FOX 42.60 44.80 -2.20 0.360 42 48 0.170 0.388
AVG 44.95 46.67 -1.72 0.394 45.67 47.67 0.407 0.555
IBD 44.00 43.00 1.00 0.565 49 44 0.779 0.867
ABC 51.00 45.60 5.40 0.788 53 45 0.878 0.896
NWK 49.00 44.60 4.40 0.749 51 45 0.814 0.861
ARG 48.67 44.67 4.00 0.730 49 45 0.730 0.830
NBC 47.00 45.75 1.25 0.576 54 43 0.947 0.959
CBS 47.67 43.50 4.17 0.743 49 44 0.779 0.867
PEW 47.60 46.00 1.60 0.596 46 48 0.381 0.523
LAT 50.00 45.00 5.00 0.774 51 44 0.854 0.901
ZOGBY 46.40 43.60 2.80 0.672 46 44 0.625 0.804
TIME 49.33 47.00 2.33 0.635 49 45 0.728 0.820
AVG 48.07 44.87 3.20 0.688 49.70 44.70 0.751 0.833