Now that the actual returns are in, we don’t need the simulations or the forecast model. It turned out, in fact, that although the FiveThirtyEight model had a very strong night over all on Tuesday, it was wrong about the identity of the tipping-point state. Based on the polls, it appeared that Ohio was the state most likely to win Mr. Obama his 270th electoral vote. Instead, it was Colorado that provided him with his win – the same state that did so in 2008.
The worry for Republicans is that Mr. Obama won Colorado by nearly five percentage points (4.7 points was his margin there, to the decimal place). In contrast, Mr. Obama’s margin in the national popular vote, as of this writing, is 2.4 percentage points. We estimate that it will grow to 2.5 percentage points once some remaining returns from states like Washington are accounted for, or perhaps slightly higher once provisional ballots in other states are counted. But it seems clear that Mr. Obama had some margin to spare in the Electoral College.
Had the popular vote been a tie – assuming that the margin in each state shifted uniformly – he would still have won re-election with 285 electoral votes, carrying Colorado and Virginia, although losing Florida and Ohio.
In fact, had Mr. Romney won the popular vote by two percentage points, Mr. Obama would still have won the Electoral College, losing Virginia but holding onto Colorado.