Political prognosticators are not nearly as scientifically grounded as you might think. The Big Three (Cook, Rothenberg, Sabato) all do considerable number crunching, information gathering, historical research etc, but even they wind up looking at available polling and taking an educated guess as to what the actual results will be.
Others like Nate Silver and Jay Cost essentially just aggregate all the educated guesses, plug in the Generic Ballot number of their choice, and then make their own estimate. This too is what 98% of the various talking heads across all the networks do. In fact I think that's about all anyone CAN do. There is no oracle that can whisper the future in your ear.
What all that means is that no one knows for sure what the heck will happen today. Group-think is quite powerful and humans hate to go out on a limb and then look silly because of it. I suspect that all the various experts base their predictions to some extent on what "everyone else is saying". Thus the fact that most experts have slowly been expanding the number of seats they either think are in play or that the Republicans will pick-up, may be in part a function of each of them feeding off of all the others. They don't want to get out in front of the field but they don't want to get left behind either.
One calculation I've been closely watching is RealClearPolitics' House numbers. Two months or so ago they showed both parties at roughly 200 seats with 35 Toss-Ups. Slowly but inexorably those numbers have moved against the Democrats. This morning their numbers are 224 R, 167 D, 44 Toss-Up. That is a massive slide and means that RCP (which is not known for making wild predictions) considers that the GOP has a statistically significant lead in enough races to pick up 45 seats with another 44 seats too close to call.
Consider the above as a casserole to which I'll now add my three secret ingredients. Those three are:
1. How a select group of states and districts are polling compared to historic norms.
2. My experience of over 30 years of observing and often times being fully immersed in our political system.
3. What my gut tells me.
My prediction is in two parts.
First, I believe that the Republicans will pick up in excess of 58 seats in the House. Pick-ups in the range of 62 to 70 will be accompanied by the pick up of 9 Senate seats.
Second, my gut tells me that the probable number of net seats the Republicans will gain is closer to 85 then it is 60. Any number of House pick-ups in excess of 75 will, I believe, translate to a pick up of 11 or 12 Senate seats.
For what its worth.