Nate Silver's Senate model
I just learned something interesting about 538's Senate model. I originally assumed that the listed probability of a Dem Senate was the probability of Democrats gaining four seats. But of course, four seats isn't enough in a hypothetical world where Trump wins tomorrow. In that case, five seats would be required (since VP Pence would break the tie).
So Nate's model uses its forecast in the presidential race to determine what happens if Democrats gain four seats. That means that in 30% of the simulations where Democrats get four seats, Nate's Senate model tracks that as Republicans controlling the Senate. So if you just want to see the probability that Democrats get four seats, that currently comes out to 54.5% in polls plus (an increase of about 5%). It is interesting that as Clinton's probability in the model has decreased, that had a slight direct affect on the Senate model (above and beyond any affect measured in polls for Senate races).
Of course, a 5% increase isn't huge; it looks like it will still be neck and neck. But it's a welcome surprise to see the odds of a four seat gain higher than I thought, in his model.
= new reply since forum marked as read
But, I don't think it'll be that close. Like George Bush and Stephen Colbert, I go with my gut! 
And, my gut is telling me that there's going to be a whole lot of Republicans staying home.
