You, too, can handicap the U.S. Presidental election from the comfort of your own home!
Fill in your best guess for state by state win probabilities -- using polling data, prediction markets, pundit opinions, or your own wild-ass guesses, and see how frequently your candidate-of-choice comes out on top!
Note: these numbers indicate the percentage probability of the Democratic incumbent (Obama) winning the state, not poll numbers. Generally, if a candidate leads in multiple polls beyond the margin of error -- and there's nothing wrong with how the polls are constructed -- it's a safe bet the leading candidate would win an election there all of the time (100%), no matter how large or small the actual poll numbers are, at least if the election were held at the time the polls were taken. If polls show leads inside the margin of error, the win probability decreases: although if multiple polls show a lead for one candidate, even if it's still inside the MOE, the probability that the leading candidate is actually ahead may still be close to 100%. You can accordingly aggregate polls to produce a statistical guess at how likely one candidate is to win a state, although this is complicated by the fact that poll data is taken at different times, and so each poll may be sampling from populations with actually different averages.