[W. Deranged]

I find situations like this to be pretty cool theoretical exercises.

If you think about this game-theoretically, what you should basically look to do is play in such a way that your expectation is never less than the size of the rack that you are going to win. If you are playing 20/40, you should probably play the hand as you normally would, with the exception that you will probably not be folding at any point considering that the rack will pay for your flop, turn, and river bets combined. Considering you do stand to win more by winning the whole pot than you do by losing and getting the rack, you should still play to win the hand.

At the smaller limits, the key is going to be to win a giant pot. The worst thing that can happen is that you win a pot that is smaller than the size of the rack. Hence, play as is necessary to get a ton of money in the pot; make giant pot-building moves like limp-reraising pre-flop (given it doesn't shut people out) and value-check-raising the flop. Notice that you'd like to keep a lot of people around to assure the pot gets big, but I would avoid slow-playing too long, because if you do a good job of building a monster pot on early streets there will be a point at which you actually want to shut people out and win the hand instead of taking the rack. At a limit like 4/8 this point will almost certainly come up by about the turn. At 3/6 it will be marginal, and you may find yourself in some big, multiway rivers if you play correctly at 3/6.

Anything about 4/8 you probably shouldn't worry too much about changing your play on early streets because you'll probably hit the $200 mark anyway. Obviously, again, though, you are never folding at any point ever.