Hmmm ... they say the results could change if the economy changes, but when they give their past results they don't talk about multiple applications of the model in each election cycle. Here's what they predicted last year for this years election:
Did they apply the model at different points during previous election cycles and get different results? That might make one more skeptical about their record of success in the past and about their current prediction.
Still, if they give Trump Florida, Ohio, and Colorado (all states he might not win) and that's still not enough to win the election, you might wonder how Republicans can win this year.
It's more likely the GOP would lose Ohio, Florida, Colorado, or Nevada, than that the Democrats would lose Iowa or New Hampshire or another blue state.
Unless you want to turn the model on its head and argue that the model's awarding the most hotly contested of the traditional swing states to the GOP indicates just how bad things are for the Democrats?
It's more likely the GOP would lose Ohio, Florida, Colorado, or Nevada, than that the Democrats would lose Iowa or New Hampshire or another blue state. Trump will win Florida, Ohio, Nevada and is already competitive in Pennsylvania and Virginia.
Voter enthusiasm is overwhelmingly on the side of the GOP. Barring something unexpected, this will be a Trump landslide...
Vote Trump!
Thee most pivotal state: Pennsylvania!!!
I also think a strong running mate could be Iowa Senator Joni Ernst; add Iowa. I truly think Trump could swing his home state of New York and a recent New Jersey poll only had Trump down 45 to 38 (Christie factor??)
I pray that all conventional wisdom is thrown out the window this November. After all, we’ve never had a candidate shake things up the way Mr. Trump has this time.