Basically, the number of apportioned seats will stay the same. The idea here is to gerrymander two House maps which each max out the number of Republican/Democratic districts. Per tradition, the default factor in determining who gets each seat will be based on the 2008 election results, unless there's justification for why a seat is not safe for one party even if a candidate won it overwhelmingly (i.e. UT-4).
Remember that 2008 was a pretty Democratic year, so the baseline for pure tossup should probably be 53% Obama rather than 50-50.
I'll probably throw in some maps.
The best measure is the one-cycle PVI. Take the Dem percentage of the 2008 two-party presidential vote in the district and subtract 53.7%. Positive values favor the Dems and negative values favor the Pubs. Any result within 1.5% of even is a tossup, between 1.5% and 5.5% is lean and over 5.5% is solid.
To test how strong the gerrymander is, use my skew factor. Find the state PVI as if it were a district and multiply it times 4 and then times the number of districts. You can round that off to the nearest integer as the expected state skew for a non-partisan map. Now add the number of Dem districts minus the number of Pub districts to get the district skew. The difference between the district and expected numbers is the plan skew.
