Any betting scheme patterned after our benchmark will produce an expectation similar to that portrayed here. You can, therefore, create a model that's easy for you to implement. To formulate a practical betting scheme that's suitable for real-world applications, we are bound by two overriding concerns. First, we need to be aware of the key count

. As we learned in Chapter 5, it makes sense to begin increasing our wagers once the running count is equal to the key count. Second, we'd like the ceiling (maximum bet) to be made at or near the pivot point, where we know our advantage is nearly +2%. In between the key count and the pivot point, wagers can be made in any manner-as long as they generally increase as the running count increases.

Your betting scheme does not have to be overly rigid, however. In the 2-deck schedule, you could go from 2 units at the key count to 5 units at the pivot point and float randomly from 3 to 4 units in-between, and give up very little. The plethora of betting schemes is limited only by your imagination, and by what you think you can get away with. That is, as long as you have the proper size bankroll to support it. It's always most important to pick a betting spread and ramp that you and your bankroll can live with.