All evidence shows this is going down to the wire in the key battleground states. You on the other hand are doing exactly what you proclaim they are. Pretending that the evidence shows anything but a nail biter.
The chance of either candidate winning this election ~= the chance they'll win Ohio.
Obama's been consistently ahead in the polls in Ohio for the past few days. Since an average of polls has a lower effective margin of error than a single poll, there's something like a 99% chance that this lead in the poll average represents an actual lead. The only reason Fivethirtyeight even gives the state only a 90.6% chance of going to Obama is because they're hedging there bets and trying to take into account the possibility that the polls are all off the mark. However, that's a pretty desperate hope to cling to.
If Obama wins Ohio, then Romney has several other courses. They are below, along with the respective Fivethirtyeight probability of Romney winning each state, and also a calculation of the overall probability of the scenario occurring, using the fact that the probability of several events occurring in a row = the probability of each multiplied together:
1) Florida (49.7%) * NC (74.4%) * Wisconsin (3.3%) * Colorado (20.5%) * Iowa (15.7%) * Nevada (6.6%) * New Hampshire (15.4%) = a 4e-6% (using scientific notation), or
chance
2) Same as 1) with Virginia (20.6%) rather than Wisconsin = a 1.6e-4%, or
chance
3) Florida (50.3%) * NC (74.4%) * Virginia (20.6%) * Wisconsin (3.3%) * Iowa (15.7%) * Nevada (6.6%) = a 2.6e-5%, or
chance
4) Florida (50.3%) * NC (74.4%) * Virginia (20.6%) * Wisconsin (3.3%) * Colorado (20.5%) * New Hampshire (15.4%) = a 1.5e-4%, or
chance
5) Same 4) but with Nevada (6.6%) instead of NH = a 6.8e-5%, or
chance
6) Same 4) but with Iowa (15.7%) instead of NH = a 1.6e-4%, or
chance
Altogether, there is about a 5.7e-4%, or
chance of any of these occurring.
I'd stick to Ohio.