The House and Senate Expectations Game
At present, the CW about November 2 is that Republicans will win the House while Democrats retain the Senate, and if GOPers complain about the spin being a split decision, they have no one to blame but themselves for their year-long orgy of triumphalist rhetoric.
Perhaps anticipating a sense of disappointment among Republicans, RCP's Sean Trende tries to cheer them up today with some analysis about the nature of Senate contests and what should be reasonably expected. His main point, which should come as no surprise to those who remember this started out as a cycle in which Democrats thought they might actually pick up Senate seats, is that the Senate seats that happen to be up this year are disproportionately held by Republicans already. Thus, the Senate landscape is tougher than the House landscape, as Trende explains by looking at the states and districts from the perspective of PVI (Presidential Voting Index, the generally accepted measurement of the partisan character of any given jurisdiction):
To take the ten seats they need to win the Senate, Republicans have to either run the table in every state that is D+5 or better, or make up for any misses in even bluer states. To put this in perspective, for House Republicans to pull off the same feat, they would have to pick up about 123 seats! The landscape is also more difficult for Republicans given that four of the GOP's five most Democratic seats are open this cycle.
This is all true, but the same PVI data that makes the likely Senate results look better than they might first appear also creates some needed "perspective" for contextualizing likely House gains:
[B]ecause of Democratic gains in Republican-leaning districts over the last two cycles, House Republicans have been waging this election on relatively favorable turf. If Republicans defeat every Democrat in an R+4 district or better, they'll pick up 47 seats. Almost 100 seats are held by Democrats in districts that were D+3 or better, i.e., in swing districts or districts leaning toward Republicans.
Turn that insight around, and you can say that Republicans can pick up 47 seats without beating a single Democrat in a district where the PVI is better for Democrats than R+4. That would be an impressive GOP win, but hardly the stuff of any big pro-Republican long-term trend, particularly if you take into account (1) the especially large age-related Republican midterm turnout advantage; (2) the economy; and (3) the near-universal phenomenon of midterm losses by the party controlling the White House.
As Sean says, it's all a matter of perspective.