The “father” of card counting was Professor Edward O. Thorp whose book, “Beat The Dealer” was initially published in 1962. In the revised edition of 1966, he presented not only his original “ten-count” system, but added his “simple point count” system and a “complete point count” system which evolved into the High-Low or Plus-Minus Count. But most people forget that he had developed an “ultimate” point count system in the first edition of his book. Thorp knew his stuff and he was aware of how powerful a counting system could be, if one could add and subtract like a computer. That wasn’t a very practical idea back in 1962 and he certainly didn’t envision the Internet or Web casinos, but here we are.
You are probably reading this at a computer which is so powerful that, if you had it back in 1945, you would have probably been burned as a witch. Well, maybe not, but when you consider that the Internet was initially based upon computers with 25K (“K”, not megs) of memory, what we have at our fingertips today is astounding. But, Las Vegas has never looked kindly upon the use of computers at the gaming tables; in fact, it’s now illegal to use one. That wasn’t always the case, though. Back in 1960, a man named Robert Bamford developed a Blackjack-playing “black box” which he used, openly and at some profit, in the casinos until they began shuffling the deck after every hand. Concealed computers followed in the ’70s, the most famous being “George” which was used by a team formed by Ken Uston. With George, that team made over $150,000 in December of 1977.
So, what’s the name of your computer? It doesn’t have to be “George” for you to use it to play very accurate, if not perfect, Blackjack. Of course, you’re not going to carry it into your favorite casino, but if you play Blackjack on the Web, you can do so with great accuracy. If you’ve read the previous installments of this series, you’ll recall that counting systems are typically measured in three ‘efficiencies’: betting correlation, playing correlation and insurance correlation. You also need to remember that no one system excels at all three. Even the best system is a compromise; so, what if you had a way to keep track of three different counts? Then, you could use one for sizing your bet, one for the proper play of the hand and one for making insurance decisions. And, for good measure, you could keep track of the number of cards played so that any conversions to the “True Count” would be perfect! Such programs do exist and I am in the process of evaluating one of them. This idea of using your computer is relatively simple; it’s really just a spread-sheet function. But, to hark back to the ‘good old days’ of the computer biz, one truism remains: Garbage In, Garbage Out. I am not a computer programmer, hell, I’m not even a skilled user of Excel but I can show you some good counts to consider. Unlike my previous lessons, there’s not a lot of information available on these counts, so when it comes to indexes (indices?) like we use in ‘regular’ counting systems (you know; “stand with 16 vs. 8 at a true of 7”, etc.), you’re on your own here. But, I’ll give you the basic figures which is a good start.
The Thorp Ultimate Point Count
Optimal System for Variation of Bets
Point Values; A -9; 2 +5; 3 +6; 4 +8; 5 +11; 6 +6; 7 +4; 8 = 0; 9 -3; 10 -7
Betting Efficiency: 100%
Optimal System for Variations of Strategy
Point Values: A +51; 2 +60; 3 +85; 4 +125; 5 +169; 6 +122; 7 +117; 8 +43; 9 -52; 10 -180
Playing Efficiency: 70.3%
The Insurance Count
Point Values: A-9 +4; 10 -9
Insurance Efficiency: 100%
Note: This was covered in the “single-level” lesson.
Although I called the “playing” count an ‘optimal’ system, it’s really only the optimum for a single-parameter count. Don’t confuse ‘parameter’ with ‘level’, however. A single-level count uses point values of 1, 0 or -1 but a single-parameter count just uses a total point value for all your decisions. We’ll cover multi-parameter counts, which use a side-count of the number of remaining cards (like Aces) by type, in the next lesson. It’s certainly possible to achieve a playing efficiency of almost 100% by keeping track of each type of card separately. Now understand that 100% efficiency doesn’t mean you’ll win every hand, because you still don’t know the order in which the cards will appear, but in ‘the long run’ order doesn’t matter; it will eventually work out to be random.
You now have most of the information needed to create a program on your computer which can play very accurate Blackjack. There are two approaches one can take; the first being a program which keeps three different counts and you use the appropriate one as you play. The other is one which doesn’t use any of these counts; it just calculates the bet and proper play based upon the composition of the remaining deck(s). That’s more in the style of a multi-parameter count and I’ll cover those next time. I’ll leave you with a teaser: I have, on my hard drive, copies of programs just like the two I’ve described above.
See you here next time.