The Sixth Year Curse?

You may have heard conservative pundits argue that the elections this month (gloating...) are nothing special, that it's quite common for the midterm elections of the second year of a two-term president's second term to result in losses for his party. You may have also heard that the actual numbers of seats changing party was on par for this historical trend.

Media Matters for America notes that the sixth year curse did not affect Bill Clinton and that Ronald Reagan got off lightly, too. True. But hasn't the avid gerrymandering by the Republicans in so many parts of the country caused the number of possible losses to have gone down in recent years? Think about this: If the number of seats which are truly up for grabs has shrunk, what is the actual percentage of such seats which the Democrats gained this month?

I'm trying to find a way to express the seats won as a percentage of seats actually in play, because I think that this last election was quite different in that respect. But I could be wrong and right now I'm too lazy to study this topic in detail.