Yglesias highlights Alan Abramowitz's analysis of the midterms:
The current political environment only appears unfavorable for Democrats compared with the extraordinarily favorable environment that the Party enjoyed in both 2006 and 2008. The two structural variables in the modelprevious Republican seats and the midterm dummy variablepredict a Republican gain of 38 seats, half due to the small number of Republican seats prior to the election and half due to the fact that 2010 is a Democratic midterm year.
According to this model, the main reasons that Democrats are likely to experience significant losses in 2010 are the normal tendency of voters to turn against the president’s party in midterm elections regardless of the national political environment and the fact that after gaining more than 50 seats in the past two elections, they are defending a large number of seats, many in Republican-leaning districts.
This isn't to say that the political environment doesn't matter at all:
Even under what might be considered a best-case scenario for Democrats, if President Obama’s net approval rating were to improve from a +5 to a +20, and Democrats were to regain a 10 point lead on the generic ballot, Democrats would still be expected to lose about 20 seats in the House. On the other hand, under what might be considered a worst case scenario for Democrats, if President Obama’s net approval rating was to fall from a +5 to a -20 and Republicans were to gain a 10 point lead on the generic ballot, Democratic losses would be expected to reach 54 seats in the House.