Alan, for a 20% loss rebate example, let's use 99.17% bonus poker, which, except for the two or three major mistakes you make every 40 hands or so, you play perfectly. Thus, with no loss rebate (and without your players club card inserted), you can expect to lose at least 83 cents for every 100 dollars you play through on this game.
Now there are two outcomes for each deal-draw resolution (let's be generous and assume you never make mistakes):
1) You win or break even with probability 45.51%
2) You lose, with probability 54.49%
As stated above, with no loss rebate, this is a losing game (83 cents per $100 coin-in):
Expected_Value=ExpectedValue_of_nonlosinghands +(Probability of loss)*(amount lost)=$53.65-$54.49=-$0.83
Under the conditions of a 20% loss rebate let's see what the loss is per $100 coin-in:
Expected_Value=ExpectedValue_of_nonlosinghands +(Probability of loss)*(amount lost)*.80=$53.65-$43.59=$10.06
Wow, what a difference Alan ?! With no loss rebate, you lose 83 cents per 100 dollars wagered and with a 20% loss rebate you
win $10.06 for that same 100 dollars wagered! This assumes, of course, that you make no errors and that you play 99.17%
Bonus Poker exclusively, and that each loss is reimbursed 20%.