Most everyone thought Ohio was the deciding factor. It is where the candidates spent the most time and money, and where everyone thought the election would break. It turns out they were wrong; while Ohio was the state that allowed the news media to call the election for Obama, it was a more narrow victory than Colorado.
The map I posted, from FiveThirtyEight, weighs the states as a spread of how well Obama did. While the final numbers are still being figured out, it looks like Obama won the national "popular vote" by 2.5 percent. A much smaller victory than his 2008 numbers, but a victory nonetheless. Where the polls get interesting is when you break down how well or poor Obama did in relation to that number.
The swing states of Michigan, Pennsylvania, Colorado, Virginia, Wisconsin, Iowa, New Hampshire, and Nevada went to Obama by more than 2.5 points, while Ohio and Florida went for Obama by less than 2.5 percentage points. Now, generally, a national swing can be seen in a state-by-state basis. That means Obama lost red states by a bigger margin than last time and Romney did better in solid blue states. What does that mean?
If the popular vote was split 50-50, Barack Obama likely would have won 285 electoral votes.
Of course, a 50-50 split is rare. A big Republican win would change things. Virginia, which went to Obama by 3, would be the first to go and leave Obama holding on to only 272 and unable to lose another. However, Colorado is the only other one under five percent, at 4.7 percent. So that means, assuming an average distribution of a Republican spread of 2 points, Colorado would still be .2 percent in the Democratic column.
Of course, this isn't a perfect science. Not all states follow this model exactly; in some states Obama did worse in 2008 than John Kerry in 2004 despite the huge swing, and in some states he increased the Democratic margin by even bigger points. But just because it could be more favorable to Republicans doesn't mean it would; it could actually prove to be that these so-called swing states hold Democrat even in a bigger move nationally.
Other than Florida and Ohio (which were lower than Obama's national numbers), Virginia (which he did not need to win), and Colorado, Obama won every other swing state by more than five points.
New Hampshire: 5.8
How does the Republican Party reconcile these numbers? Paul Ryan was supposed to make Wisconsin competitive; Romney was supposed to have heavy Michigan roots. Nevada has some of the highest unemployment numbers in the nation and the incumbent still won. New Mexico went for Bush less than ten years ago and it isn't even considered a swing state anymore. Virginia, despite only going to Obama by three, is predicted to become more Democratic by 2016 due to the growing Washington D.C. suburbs. People keep saying that Wisconsin, Michigan, and Pennsylvania are swing states despite how reliably Democratic they have been, but it does not happen.
If you give Obama his safe states, including New Mexico, a future safe state, Nevada, and likely safe states like Michigan, Wisconsin, and Pennsylvania, the Democratic Party has 253 electoral votes to the 191 that are considered safe for Republicans. This isn't even giving the Democrats Iowa and New Hampshire, despite their healthy margins, and not giving them the future of Virginia.
We could be potentially looking at a 253-191 future, with North Carolina, New Hampshire, Iowa, Florida, Ohio, Colorado, and Virginia as swing states. Florida or Ohio alone give Democrats victory, as would Virginia plus either New Hampshire or Iowa. Things could change demographically, even though people think it will be more favorable to Democrats in four years, but it is an uphill battle for Republicans with this map. They are going to do well in 2014 with an incumbent party and the Obama coalition not being the type to turnout in midterm elections, but Republicans should not believe the hype, just like they should not have believed the hype in 2010.