The really interesting content is the explanation of what goes into creating the election models. Basically, polls (with all their problems) and some very basic economic stats. Can HN denizens whip up a better model using public data?
Misleading headline. 61.8% chance of winning. Expected 50.5% of popular vote.