One of the simplest approaches is to flat bet the single-deck game. Few casino personnel are capable of detecting a counter through play of the hands alone. However, this is only worthwhile under special conditions: you need a break-even edge off the top, deep penetration and a count system optimized for playing efficiency to gain a 1% edge.
Several advanced methods of gaining an edge have been outlined in "Blackbelt in blackjack" by Arnold Synder. Synder explains the concepts of "opposition betting" and "depth-charging". The first method involves betting randomly throughout the shoe but averaging higher amounts in positive counts. The second involves raising your bet in increments as the cards are dealt out regardless of the count, to capitalize on the increasing gains from playing strategy as more and more cards are dealtout.
Two other methods can be effective. The first involves betting three hands at table minimum and a large wager on the fourth hand. In positive counts you play only the one hand. This has two benefits: you are "eating cards" in negative decks, and your losses in negative decks are the lesser because your playing decisions on your main wager benefit from having seen the other hands played out. This technique can be improved with intelligent play of the small bet hands. Because their outcome is of little or no importance to you, you can misplay the hands in order to further eat cards. It is advisable for example to split tens so that you may play one less big hand before the shuffle. Your play seems crazy to the pit, since you are departing from basic strategy, by "engineering" the deck in your favor.. Ideally, because the play is so unusual, you would have a partner play the small hands. This would avoid them barring you because "they figured you must be doing something". I doubt there is a pit boss in Nevada or New Jersey who could detect a partnership using this technique correctly.
Another method involves using a progressive betting system, in conjunction with a count system optimized for playing efficiency such as AOII. The crucial condition is that you must reset at your bet at the shuffle. There is a weak correlation between the loss of a hand and increased advantage (about .1% or.2 high/low points), so you gain a little from betting more after a loss. There is also a mild depth-charging effect, since you always start with a minimum bet and can only increase it by following the progression as more and more hands are dealt out.
You cannot gain all that much from this: between .15% -.30% over flat-betting, depending on the penetration, the rules and the progression you are using. However, I have found this approach is also much more effective cover than flat-betting. Systems such as the Fibonacci or the Labouchere are ideal since they escalate rapidly but retreat after a series of wins. However, these may not be easily recognized by the pit as progressions. To my mind , the D'alembert (or pyramid) is the best system to try this method with. The system exploits a series of losses and does not escalate bets too rapidly. You simply add one unit for a loss and subtract one for a win.
Professionals often assume that progression betting is undesirable because it increases the volatility of your results. While there is some increase in fluctuation with this approach, the gain in EV usually makes it worthwhile. For example, flat-betting a typical Reno game yields a desirability index of 2.2. With a D'alembert progression reset at the shuffle the DI is 3.1. It would not be impossible to design a system which would provide less standard deviation than flat-betting, though strictly this would be a "regression" rather than a "progression".
