Back to The Blackjack Page

The Blackjack Page Archive

High Stakes Blackjack

Note from the GameMaster: I recently acted as a consultant to four wealthy individuals who normally receive "red carpet" treatment from any casino they wish to patronize. Of course, the reason they're treated that way is because they lose big $$$ on every trip. Free is one thing; "free" from a casino is something else entirely. Don't get me wrong here; these guys (yes, all men) know how to play Blackjack pretty well - not 100%, but pretty good nonetheless. That said, they are (relatively) consistent losers - generally, one or two will walk away with a profit, but overall, the casinos they visit are making a nice profit from them.

They always go on these trips together and one of them asked me if I could come up with a plan to help them either make a small profit or at least not lose so much. The comps they get are great, believe me, so I first had to warn them that if they start winning on a consistent basis, the welcome mat might get pulled out from underneath them, pronto. Their reaction? They and I are originally from New Jersey, so their response sounded familiar: "We should be so lucky." Amen, brother.

Okay, how much are we talking about here? What stakes do they play, etc. First of all, 95% of all their gambling is Blackjack, with some Craps and Baccarat thrown in, as you can well imagine. Naturally, there's not much I can do for the other games, but oh, brother...can I ever help them at BJ! Especially when they told me they'll gladly risk $50,000 each on a four- or five-day trip. That's a chunk of change, any way you cut it. Generally, they told me, their average bet is in the area of $500 per hand - usually at least $300 with a maximum of $1000 or so; unless "someone takes a shot", in which case he might bet $3000 or $4000 in an attempt to get even.

They usually all play at the same table, which is reserved only for them; no outsiders. Because they're New Jersey residents, six and eight deck shoe games are what they grew up on, so very little play on their road trips has been single- or double-deck Blackjack. To get an idea of the type of revenue they're generating for the casinos, I figured 4 players, each betting $500 per hand on average. At $2000 total per hand and a playing rate of roughly 75 hands per hour, their composite bets are about $150,000 per hour. Every casino calculates the comps they'll offer a patron by first using some sort of "theoretical"; which is another way of saying casino edge. In Blackjack, it typically ranges from 1 to 3%, depending how skilled the player being rated is perceived by the game supervisors. In other words, if you know how to play the game well, your comps may be reduced, as compared to some doofus who knows nothing of Basic Strategy.

I had the "Jersey Four" complete my test of basic Blackjack knowledge and all but one scored on the high side. I estimated that the casino where they played most of the time was using about a 1% theoretical on all of them. If I'm correct, this means that the casino felt it was making about $1500 an hour from them as a group. Since most casinos are willing to give back 40% of the theoretical as comps, these guys were entitled to about $600 in comps for each hour of play. If they played 4 hours a day, that meant the group should be receiving about $2400 in comps per day. Even though they were being treated well, when I told them this, they were....well, pissed off is probably the "Jersey" way to put it. Perhaps the suites they were staying in are worth $1200 a night, but there was no way they were eating $300 worth of food, each. I pointed out that their airfare was being comped, as were limo rides, the odd female "companion" (they claim that never happens - they're all married) and a myriad of other knick-knacks, so they're probably doing okay from a theoretical point of view. However, if the casino has them down for a 2% theoretical, they're getting screwed. And if they're playing more than four hours a day each (they said they are), they're really getting screwed on a four or five day trip.

If nothing else, my analysis made them want to beat the casinos - no mercy. I needed an attitude like that because if they did what I told them to do, the fun and games were over. This wasn't going to be easy, but neither was I going to turn them into a bunch of ruthless card counters - they didn't have the desire to get that involved. All they wanted to do was keep enjoying the comps and lose less overall, on a year-to-year basis. If accomplishing that required some work, their opinion was, bring it on. The first thing I got them to agree to was learning Basic Strategy for a typical game by heart; dead cold, no doubts. By "typical", I mean a Double-deck game where the dealer hits soft 17, but you may double on any first two cards, including after splitting pairs, no surrender. This is the game I discuss in my Blackjack School, Lessons 21-23 and one that can be found fairly easily in Vegas, as well as Tunica and other casino centers - but not Atlantic City. While I realize they won't always be able to play a DD game, it's definitely the way to go whenever possible, plus using the Basic Strategy for this game in a 6- or 8-deck game won't cost them very much. The DD game I described has a house edge of 0.40%, whereas a 6-deck, S17, DA2, DAS game without surrender is 0.41%, so it's basically a "wash" from that point of view.

One thing I asked them to agree to was splitting their wins and losses evenly; in other words, to act as a team. If four players who are using a common bankroll play at four different tables, each player can safely bet as though the team's total bankroll was theirs alone. For example, if each were to put up $50,000, then each person could play as though he had a bankroll of $200,000. Instead of, say, a top bet of $800, each team member can play with a top bet of $3200 - that's the power of team play at Blackjack. These are just examples, but you get the idea.

However, if all four are playing at the same table, their individual hands are closely related - should the dealer get a 'blackjack', most likely all four will lose the hand; conversely, if the dealer busts, all four might win the hand. You can easily see that the situation here is more like one person playing four hands at the same table. As I point out in my series, "Playing Multiple Hands" (which you can access by clicking on the Blackjack navigation tab up at the top of this page), it's usually not worthwhile - at least from a mathematical viewpoint - to play more than one hand when alone with the dealer. Yes, the "Jersey Four" - 4 players - would be at the table, but because it's a team, they're actually the equivalent of one person playing four hands and that calls for some creativity when it comes to how much to bet.

How much to bet was the primary question they wanted me to answer, but the bad news is that no system of betting, by itself, can overcome the casino's edge in the long run. Of course, I could show them how to cut the casino's edge to a minimum by playing perfect Basic Strategy on each and every hand, but they'd still lose, unless I could figure in the comps as part of their return. The way I explained it was this: If they bet a total of $150,000 per hour, their long-term "expected value" would be to lose 0.40% of that, or approximately $600, which just happens to equal their comps, more or less. The downside here is that the casino may change their "theoretical" because, (A) they're now playing a game - the Double Deck version - that the casino knows has a lower house edge - and (B) they'll be making fewer Basic Strategy errors, which the casino game supervisors may notice.

A perfect example is playing the hand A-7 vs 9, 10 or Ace by hitting, which is a Basic Strategy move. That's a sure sign you know what you're doing and, because I naturally felt they should play A-7 this way, it was eventually going to be noticed. After looking up some numbers, I came to the conclusion that just standing with A-7 vs. 9, 10 or Ace (which most were doing anyway) didn't cost much at all. In a multi-deck game where the dealer hits soft 17, standing with A-7 vs. 9 costs you 0.038% or 19 cents if your bet is $500. Against a 10, it's almost an identical amount and against an Ace, it's about 42 cents. That totals to approximately 80 cents over the course of three hands and $1500 in average bets. A player can expect to get a hand of A-7 about 1130 times each 100,000 hands. Of those, a total of 491 will be when the dealer is showing 9, 10 or Ace. That's less than one-half of one percent of all the hands they'll play, so I told them they could misplay A-7, but that was as far as they should go. They'd still have to hit 12 vs. 2 and 3, double A-8 vs. 6 and all of the other correct Double-deck Basic Strategy moves most players don't make, like it or not.

Obviously, if the Jersey Four would just play "perfect" Basic Strategy, in the long run their losses would be covered by the comps they get and, with just a 0.40% negative expectation, there would be times they'd come home with a profit, just from variance (what most people call "luck") alone. But if I were to leave it at that, they wouldn't be getting their dollars' worth from me, plus we already know they won't be playing perfect Basic Strategy. I needed to give them more weapons for this battle.

The best weapon I could recommend in this situation is the Ace-5 Count. I've discussed this count before and it's really simple. This is basically an abbreviated Hi/Lo Count, where the card most important to the player - the Ace - and the card most important to the dealer - the 5 - are the only cards tracked. Aces are assigned a "point" value of minus 1 and 5s are assigned a "point" value of plus 1, so when there are more Aces left in the remaining decks than 5s, the count will be positive and you can bet more. When the number of 5s remaining is greater than the number of Aces, the count will be negative and you should bet less. Well, the Jersey Four wanted to know how much to bet and when, so this count provides that answer. Only one of them needs to learn this count because he can tell the others how much to bet by non-verbal signals (which I'll keep secret here so as not to tip off the competition - the "pit critters.")

As I mentioned earlier, a significant problem here is that the Jersey Four are really the mathematical equivalent of one player betting on four spots and playing four hands at a time. To bet the same amount on four hands as would normally be bet on only one hand could be a financial disaster because the four hands are inter-related, which we call "co-variance". If one were to play four hands at the same table, only 35% of the "regular" bet should be placed on each spot. For example, if proper money management called for betting, say, $200 at a count of 4 on one hand, a bet of $70 should be placed on four hands at the same count. You can see that the total amount bet is $280, which is 4 times 35% = 140% of the bet for one hand. In this situation, I used a $200 unit as the starting point for a bet schedule. That being the case, the minimum bet for the Jersey Four would be $200 and, since I recommended a 10-unit maximum bet, there would be times when they had $8000 riding on one hand. Because of the co-variance, this is roughly equivalent to $5600, which is about a $1400 bet by each of them, something they do quite often. Sure, there will be times when they lose all four bets (and they won't be very happy with me when they do), but there will be more times when they win 2, lose 2 or win a double, push one and lose 2, etc. No, none of these guys are singers - no sopranos among them, if you catch my drift. But I digress.

In my simulations of the Ace-5 Count at a double-deck game where 65 cards are dealt before the shuffle (63% penetration), a running count of 4 or higher - which is the count I recommended they put out their maximum bet - happens only about 2 percent of the time, or roughly one hand in fifty, so about 65% of the time the count will be between -1 and +1. That's a bit of a problem from an average bet point of view, because the "pit critters" will see them betting $200 most of the time and it'll undoubtedly cause their average bet calculation to drop from the current $500, even though it'll ultimately work out to be $450 on the betting schedule I recommended to them, which I'll show you in a bit. That being the case, there's another interesting aspect to playing more than one hand, which is the fact that you can bet different amounts on two hands, which are mathematically equivalent to making identical bets on two hands. For example, if you would normally bet $300 per hand on two hands, a bet of $400 on one hand and $100 on a second hand is its mathematical equivalent. Of course, you might lose the $400 hand and win the $100 hand, resulting in a net loss, where it would have been break-even if the bets were equal but over time it'll work out to be the same as though you had bet $300 and $300. I don't want to detail the betting pattern for plays like this, because it's one way for the casino personnel to "get wise" to what the Jersey Four would be doing, so let it suffice to say that I taught them a relatively simple way to have varied bet sizes that ultimately amount to equal-sizes bets on each hand. The beauty of this method - and here's where I earned my fee - is that this type of betting works great if one or two of the four goes to the bathroom or takes a break, etc. It's really cool then and no, I'm not going to tell you more.

In case you're wondering, the bet schedule I recommended they use with the Ace-5 Count is to bet 1 unit at a running count of 0 or lower, 2 at a running count of 1, 4 at a running count of 2, 8 at a running count of 3 and 10 units at a running count of 4 or higher. I say "running count" because this technique does not require a True Count (TC) adjustment, which many beginning counters find hard to do. While it would be better for them to use the TC, in a double-deck game it's pretty simple to implement "on the fly", so to speak. For example, if it's getting close to the shuffle; that is, about 1 deck of 2 has been played, the running count when converted to a True Count is obviously near double that of the same count on the first hand. I didn't give them any betting adjustments for that - I wanted them to use the same count all the way through - but I did point out that they had a bigger edge at that point and whether or not they wanted to take advantage of it was up to them. For example, like most gamblers, these guys like to "parlay" their bets, which basically means, if they win the current hand, they'll let it all ride on the next hand. As you can see, the bet "ramp" I recommended allows for that, but the "on the fly" count adjustment would make it okay to parlay a 10-unit bet to a 20-unit bet every now and then.

The 1-10 unit bet schedule (remember, a "unit" is $200) used in conjunction with the Ace-5 count in a decent double-deck game produces a long-term advantage of only 0.20%, which is barely above break-even but that, combined with the comps these guys are getting, should produce what they wanted in the first place: less losses and maybe even a profit from time to time, yet still retain their preferred customer status. I honestly don't know if they have the discipline to pull this off, but they did do some practice together at one of the casinos in Connecticut, where they've never played before, using $50 as the betting unit (A $50 - $500 spread). They happily told me the 6 hours they played resulted in a $2200 profit, but as I told them, it may have been just "luck" and one trip does not a Blackjack team make.

Can they do this? I don't know, but it's a relatively simple plan, which I hope inspires them to becomed full-fledged counters, so they can really play some high-stakes Blackjack. Of course, these guys are playing with bigger bankrolls, but this will work for 5 dollar bettors as well, so if you're still undecided about whether or not you want to put more effort into improving your game, give this a shot. It's easy, fun and it works.

I'll see you here next time.