Normalized map (50/50 PV split)

[cinyc]

This is the Nevada/NH swing from 2000 that produces the tie many thought could happen this year.

This kinda proves how possible the tie really was.  Had the election been 49.6 -49.4 McCain, it probably would have been a tie, according to this.

I was surprised that Iowa was at the center.

Based solely on the last 3 elections, Iowa looks like the current bellweather.

Oregon is still a swing state too, even though everyone's forgotten.  It's one half northern California, one half Idaho.

...and 26% not in yet. Oregon might shift slightly when all the votes are tallied.  Smith could still lose.

[cinyc]

This is the Nevada/NH swing from 2000 that produces the tie many thought could happen this year.

This kinda proves how possible the tie really was.  Had the election been 49.6 -49.4 McCain, it probably would have been a tie, according to this.

I was surprised that Iowa was at the center.

Based solely on the last 3 elections, Iowa looks like the current bellweather.

I guess McCain spending so much time in Iowa makes some sense after all.

[cinyc]

The interesting thing to me was the following:

National swing needed for McCain to pull ahead in the popular vote: about 6.2%

National swing needed for McCain to tie in the electoral vote: about 6.8% (Obama's lead in Colorado)
National swing needed for McCain to pull ahead in the electoral vote: about 9.3% (Obama's lead in Iowa)

In other words (assuming uniform swings), Obama could have lost the popular vote by 3% and still pulled off a tie in the electoral vote.  Now that would have been a nightmarish scenario in the House. 

What was the actual result in ME-2?  The exits had it at 54%-43% Obama.   Would ME-2 have fallen before Iowa?  Looks like they were fairly close.