The traditional media loves to create sweeping narratives about elections, and they worship the winners and detest the losers. The winners are geniuses whose every plan played out to perfection, while the losers are insufferable idiots who were terrible campaigners with worse campaign teams (this is true, by the way, even when the "loser" wins the popular vote, as Gore did in 2000). Well before elections are over they start their sweeping analysis of why the expected winner created a cresting wave that could not be held back.
Here's the dirty little secret, though: little things like who actually controls Congress are decided by narrower margins than the sweeping-narrative writers like to admit. While there are of course tides that go for and against both parties, the fact is that even in tide years, who actually controls Congress will be settled by a few votes in a modest number of races.
Take 1994, for example. It certainly felt like a tidal wave to those of us sitting in the White House that year. The Republicans picked up a devastating 52 House seats, 12 more than needed to win control. But they also won 16 races by four percentage points or less, meaning a stronger field operation in those 16 races might have saved us control of the House. The losses still would have been devastating, but Newt Gingrich would not have been Speaker. In 1996, Republicans won even more close races, 21, as we fell 10 seats short of retaking the House.
Over the next four elections, the number of closely contested races dropped, as the Republicans averaged only five close victories a cycle in years where not too much changed in the House in general. But 2006 was another wave year, and the number of close elections jumped again. Democrats won enough, 14, to make sure we won back control of Congress, but the Republicans won enough not to be slaughtered even worse (17). In 2008, part of the reason Democrats were able to pick up another 21 seats on top of the wave in 2006 is because the Obama GOTV organization helped carry Democrats to victory in 13 of 20 (65 percent) of the closest House races.
What does all this mean? Regardless of all the mathematical formulas on likelihood of Republicans winning the House, ultimately this will all come down to GOTV operations in probably 30-35 races (since wave elections produce more close elections). Going into the last weekend before the election, if you and ten of your friends join with other Democratic and labor and netroots and other progressives in turning out the vote, we can still win this thing. The kind of races I have been talking about generally come down to one, two, three votes a precinct, and that can be made up by knocking on extra doors, making extra calls, bringing in those last straggling absentee ballots this weekend, working people outside of polling places one more time as they come to vote on Tuesday. What you do matters, so keep fighting until the last poll is closed.