FL/OH/PA Quinnipiac: Clinton in big trouble

[muon2]

Cook PVI, by state

90% saturation for D+15 or R+15 or greater
70% saturation for 12 to 14, inclusively  
50% saturation for 10 or 11, inclusively
40% saturation for 6 to 9, inclusively
30% saturation for 3 to 5, inclusively
20% saturation for 1 or 2, inclusively  
white -- zero PVI

That looks like a pretty swingy map, with nothing ensured for either side. In a tie election for the PV one would expect 272D, 253R and VA(13) up for grabs. A two point swing either way would shift the map from D332-206 to R332-206. Other than a slight bias for Dems in a perfectly even election, any small swing in PV will easily give the WH to either party.

[muon2]

The electoral college vote actually correlates well to the popular vote. This is a graph I posted in 2009 from the 20 most recent presidential contests through 2008. It compares the Pub EVs to the fractional difference in PV between the Pub and the Dem (percent divided by 100). If there is more than two percent difference between the candidates the debate about swing states probably won't matter much.