What would be really interesting would be for someone to calculate the trip results distribution–including frequency of winning trips and approximately breakeven trips–for various levels of house edge. Make reasonable assumptions as to game variance and total number of spins per trip. If raising the house edge reduces the frequency of winning trips by 75%, which is a real possibility, that’s going to have a powerful effect on the customer experience and willingness to return. People like to be able to brag to friends and themselves about those winning trips, even though long term they’re big losers.