Looking back at the election, my prediction about gerrymandering came true. That was about the strategy of dividing up voters so you have the maximum number of 5% win districts, and most of the other Party packed into a few districts. Normally a great plan, except for when the other side wins by 8.6%. Then all of those "safe" districts go to the other Party and you have a tsunami.
So it's kind of like buying options instead of the underlying stock. You can rig it so you increase your returns, if your predictions prove true. But if they don't, then you are really screwed. The interesting thing is like with options, you can rig things either way: You could pack fewer districts so they had 10% wins, and thus be assured of some presence in the new House. But maybe when your Party platform has a bunch of unpopular positions, you should go for 15%.