Is an 8-deck game much different than a 6-deck game? Not really, but the penetration you get at an 8-deck game is very important, as you'll see when I discuss the simulations I did for this. The 8-deck game is most commonly found in my old stomping grounds, Atlantic City, New Jersey due to the fact that Blackjack was determined to be a game of skill by the N.J. State Supreme Court way back when. Because the casinos could no longer bar players from the game for demonstrating their skill, the AC casinos - in their infinite greed for profits - first went from 4 decks to 6, then from 6 to 8. From a card-counting point of view, an eight-deck game is almost the mathematical equivalent of an "infinite" deck, which essentially means there is little additional advantage that accrues to the casino if more decks were to be added to the shoe, so it’s not likely you will ever see a 10- or 12-deck game.

Just as I did in the lesson on Double-deck games, I'll begin with some comparisons based only on penetration. In other words, for the simulations I did to develop the chart below, the only thing that changed was the penetration. The rules, the bet spread (which I'll cover later) and all other aspects of the game remained the same. If you've read my previous lessons, you'll recall that I say you're wasting your time if you play at any 6-deck game that has penetration of less than 65%. For an 8-deck game, the minimum penetration needs to be 75%. Here’s some data to consider:

Pretty sad, isn't it? Even with a 1-12 bet spread (I used the spread I teach for 6-deck games) and good penetration, the long-term edge one can get is only about 0.30%. But don't worry, I'll show you how to get a bigger edge, because we really need an absolute minimum of 1% to make it all worthwhile.

No long explanation needed here; use the same variations discussed for the 6-deck game because they apply to any game that uses more than one deck and has the same rules, like the dealer hits soft 17 (or not), you may double on any first two cards and so forth. Make the insurance bet at a True Count of 3.1 or more at an 8-deck game.

As we’ve discussed previously, it’s usually necessary to "ramp" your bets, which is another way of saying gradually increase them. If your minimum bet is $5, then a 1-12 spread will make your top bet $60, no matter how high the count gets. Depending upon when you'd like to get your top bet on the table, that is, at which True Count, it's then a simple matter to calculate just what size your total bankroll should be. Let's say you wanted to bet $60 at a TC of 5 or more. The optimum bet for that count is 1.57% of your total bankroll, so if you divide $60 by 0.0157, you get $3821 as the proper bankroll. Now remember, you won't be making every $60 bet at that count because it's your "top" bet and some will be made at a higher advantage, but $4000 is a good number and one that I'll recommend if you’re playing 8-deck games.

Important! That $4000 represents the total amount you should be willing to commit to this adventure, but it's not what you'll carry with you on a trip to the casino. For most trips, a "session" bankroll of 20 top bets or $1200 should suffice.

With a $4000 bankroll, the betting schedule could look like this:

You can see that I topped the bet out at $60 when the True Count is at 5 or more. That's just an artificial limit, based upon what kind of a spread one can usually expect to get away with in actual casino play at a $5 minimum bet table. Remember, even though they can't bar you for counting in A.C., you'll still want to disguise your skill so they don't "half-shoe" you or take other steps to neutralize your skill. However, if your "act" will allow you to get away with using a bigger spread , the optimum bets shown will apply to the same $4000 bankroll which will give you a 13.5% risk of ruin factor. By capping the top bet at $60, your risk of ruin is below 10%. For those who want to play at a $10 minimum bet table and can spread $10-$120, the problem is that you're seriously over-betting when the count is at 1 or lower so the risk of ruin factor will be higher; in the area of 25% for the recommended $4000 bankroll. Earlier I alluded to the fact that we really need a long-term edge of 1%, as an absolute minimum to make all this effort worthwhile. The primary way to accomplish that is to leave the table when the True Count drops to -1 or lower. With the A.C. casinos being such big places, it's really not a problem to leave a table when the count drops. As I teach in lesson 4, you should leave only after losing a hand, because "gamblers" seldom leave after winning a hand and we want to look like gamblers. The next simulation will show you the impact of leaving:

Simulation #1: 8-decks, standard A.C. rules, 80% penetration; $5-$60 bet spread shown above but leave when count drops to -1 or lower.

Theoretical player edge: 0.89%
Expected return for 100,000 hands of play: 1780 units
Standard Deviation for 100,000 hands of play: 1160 units

Comment: When I say "Theoretical Edge", that's the overall advantage as determined by the software I use for my sims, which is Statistical Blackjack Analyzer" (SBA), a program I've written about many times before. The problem here is that SBA plays each hand perfectly, plus it "leaves" the table religiously when the count drops. You and I probably won't play every hand perfectly and if we want to look like gamblers, we won't always leave the table immediately when the count drops. So to put a realistic spin on these numbers, my advice is to reduce them by 10% or so. That being the case, an overall edge of about 0.75% can be expected with this method of play. Just so you know, SBA calculates your expected profit for 100,000 hands, which in this case would be 1780 units, or $8900 if your unit is $5. The Standard Deviation figure is 1 SD, which will cover about two-thirds of all your playing results. What this means is, if you play a series of 100,000 "trials", two-thirds of them will show a result of 1780 units won, plus or minus 1160 units. In other words, two-thirds of the time your result will fall somewhere between a profit of 1780 + 1160 = 2940 units and a profit of 1780 - 1160 = 620 units. Remember, that's for a series of 100,000 hand trials or maybe 1-2 million hands of play. Well, darn few of us are ever going to play a million hands in our lifetime, so does that make these figures meaningless? Not really. The value of these figures is that they show you whether or not a game is worth playing in the long-term. If the SD were, say, 3000 units instead of 1160, that would mean you could play 100,000 hands and still have a high probability of being in a losing position. A game with a high SD is a high-variance game, which requires either a large bankroll or a very big player edge to make it viable. We obviously do not have a big edge in this game, but it is an edge and, with the relatively conservative betting schedule I used, the SD is kept fairly low. But we can squeeze out a slightly bigger edge by including the basic strategy variations you saw in the lessons on six-deck games:

Simulation #2: 8-decks, standard A.C. rules, 80% penetration; leave when count drops to -1 or lower, $5-60 bet spread but add basic strategy variations.

Theoretical player edge: 1.09%
Expected return for 100,000 hands of play: 2166 units
Standard Deviation for 100,000 hands of play: 1164 units

Comment: Adding the basic strategy variations gets us to where we need to be, but just barely, and you need to remember that we are faithfully leaving the table when the count drops to -1 or lower. Eight-deck games are a grind at best, but if you insist on playing them, here’s my best recommendation.

Although it's difficult to do in Atlantic City, the only sure way to get an edge of more than 1% at an 8-deck game is to backcount ("wong") the tables and not place a bet until the True Count is 2 or more. The reason it's difficult to do is that many A.C. casinos use the "no mid-shoe entry" rule, which basically prevents you from placing a bet once the first hand has been dealt from the shoe if you weren't betting from the beginning. But, because I'm also addressing 8-deck games in areas other than A.C., perhaps this technique will work for you. Here's a simulation where the amount bet was $0 whenever the true count was 2 or lower. At true counts above 2, I used a flat bet of $50, which will definitely lower any suspicions about your play if you use this strategy.

Simulation #3: 8-decks, standard A.C. rules, 80% penetration, $50 bet at TC2 or more; $0 otherwise; basic strategy variations.

Theoretical player edge: 1.46%
Expected return for 100,000 hands of play: 14,647 units
Standard Deviation for 100,000 hands of play: 3617 units

Comment: Now we're getting somewhere! But, while this method of play looks good on paper in reality it's a tough way to go. First of all, you'll spend a lot of time just watching the game, so your hands per hour rate will be low. Secondly, the casinos aren't stupid; they know this technique works very well, so you'll undoubtedly attract more than your fair share of attention. On the plus side, it gets the $$$, so if you can limit your playing sessions to one hour per shift per day in very large casinos, it's a viable strategy.

Some casinos have devices on the Blackjack tables where the dealer puts the cards from the most recent hand into the machine, rather than a discard tray. These "CSMs" render counting useless but they have no effect on basic strategy, so just use the strategy that's appropriate for the rules the game offers if you must play at all. Better yet, tell anyone who will listen that you hate CSMs and will not play at tables using these devices.