What kind of number-crunching prognosticator would I be if I didn't look at the House & Senate elections?! These are genuinely tricky, since the polling is much lighter. In addition, there's some evidence that state's overall partisan lean can have an impact on downballot races.
The contested Senate elections this cycle were in Wisconsin, North Dakota, Montana, Nevada, Virginia, Massachusetts, and Maine. In addition, gaffes by Republican candidates made the elections in Indiana and Missouri much more competitive than they otherwise would have been. Last but not least, October polls in Nebraska and Arizona showed these races were unexpectedly close. Current polling shows that Democrats will flip Indiana, Massachusetts, and replace Maine Senator Olympia Snowe (R) with a Democratic-leaning Independent, while losing Nebraska and North Dakota, with all the remaining seats staying with in the same party. There's an outside chance that Nevada (more likely) and/or Arizona (less likely) will tip in the Democrats' favor, or that they lose an upset in Pennsylvania, but I wouldn't bet on it. So the end result will most likely be a Senate with one more Democratic-leaning vote.
It's worth pointing out that while the Senate isn't getting much more Democratic, it will be getting much more liberal. Four seats will become significantly more liberal: Wisconsin (where Tammy Baldwin replaces Herb Kohl), Virginia (where Webb was a very moderate Senator outside of criminal justice issues), Massachusetts, Indiana, and Maine. Meanwhile, only North Dakota and Nebraska will shift to the right. The end result will be that the median Senate vote moves a tick to the left, from Carper/McCaskill/Manchin to Donnelly/Bill Nelson/Landrieu. That's a small movement, but it is movement nonetheless. EDIT: How could I forget! The blogosphere's favorite whipping boy, Joe Lieberman, is retiring. As long as Chris Murphy holds on in Connecticut, the Senate gets another touch more liberal and tremendously less douchey. That would move the median Senator on economic issues to something like Michael Bennett or Max Baucus, which is not such a terrible place to be.
As for House elections, I confess that I've not paid enough attention to individual races. The consensus forecast is for Democrats to make gains, but not enough to take the House, and I have no data to support or refute that consensus.