Election Scenario Analysis

[Benny B]

Scenario Analysis

How often the following situations occurred during repeated simulated elections.

    Electoral College tie (269 electoral votes for each candidate) 0.2%
    Recount (one or more decisive states within 0.5 percentage points) 5.5%
    Obama wins popular vote 67.0%
    Romney wins popular vote 33.0%
    Obama wins popular vote but loses electoral college 1.7%
    Romney wins popular vote but loses electoral college 3.0%
    Obama landslide (double-digit popular vote margin) 12.0%
    Romney landslide (double-digit popular vote margin) 2.2%
    Map exactly the same as in 2008 0.9%
    Map exactly the same as in 2004 <0.1%
    Obama loses at least one state he carried in 2008 87.3%
    Obama wins at least one state he failed to carry in 2008 29.8%


Ohio
FiveThirtyEight Projections    Dem    Rep    Margin
Polling average              46.7    42.0                            Obama +4.7

Adjusted polling average    45.7    43.0    Obama +2.7

State fundamentals    45.6    45.5    Obama +0.1

Now-cast    46.0    44.0    Obama +2.0

Projected vote share ±6.9    50.4    47.8    Obama +2.6

Chance of winning              65%           35%

[Benny B]

Can you tell me where this is from, BennyB? In particular, I'd be interested to know if the models were exclusively models of economic data; those seem to be the ones with the best predictive ability.

Nate Silver, bro. The best and most scientific in the business.