Tell-me-how-to...

How to Count Cards at Blackjack

The principle behind card counting is that a deck of cards rich in tens and aces is favorable to the player, a deck rich in small cards is favorable to the dealer. A deck rich in tens and aces, is likely to bust the dealer more often.

To gauge the richness of the deck in high cards or lack of them, the player needs to keep track of the cards that are already played and assign a point value to each card. The calculation is quite complex, but basically the card counter will give a plus point each time the deck of cards becomes more favorable and a minus point each time it becomes less favorable. Basically, the high cards have a -1 point and the low cards a +1 point and the in between 0 point.

The counter then counts by adding and subtracting points according to the cards played and keeps a running total of the count called 'running count'. Then he also needs to divide the running count by the proportion of the size of the deck of cards left to get the 'true count'. Now he knows the relative richness of high cards in the remaining deck.

A positive count is good and a high positive count is best and the card counter will assess his hand, the dealer's up card, weigh up his options and intensify his betting accordingly.

To be a successful card counter you need to have a powerful memory and fast reaction while amassing information as you play. And for what? If everything works out well, you will be looking at a slow and tedious 1% average profit. If you wager large sums of money to make the 1% worthwhile, you are likely to be noticed by the pit boss and prompt frequent shuffling of the cards. Card counting is hard, not liked by the casinos and is not as rewarding as it may seem.

More Card Counting Techniques

Unlike casino games such as roulette and craps, where the outcome of one play has no effect on any future play, a hand of blackjack depletes the deck of the cards used in that hand, and this can alter the probability of certain events occurring on the next deal. In proper statistics terms, this is known as the Law of Independent Trials - past event have no effect on future probabilities. Specifically, if the remaining cards have a higher proportion of 10-count cards and Aces than normal, it is more likely that a player will be dealt a natural, which is to the player's advantage (yes, it's also more likely for the dealer to get a natural—but the dealer wins only even money, while the player is paid 3:2). When the deck has more small cards such as 4s, 5s, and 6s, it is more likely that the player will be dealt a bad hand and bust, favoring the dealer (likewise, it increases the chance of a dealer busting as well, but when the player busts, the dealer wins even if he later busts himself).

Because the house advantage in blackjack is so small to begin with, it is quite common for a deck that happens to be "rich" in remaining 10 count cards and Aces to offer a positive expectation to the player on the next hand. By keeping track of the cards played, a player can take advantage of these situations by betting larger amounts when the deck is in his favor and smaller amounts when it is not. In the long run, the deck will be unfavorable to the player more often than it is favorable, but it is the amount bet under each condition that counts. The player can also use information about the deck's composition to alter strategy. For example, basic strategy calls for hitting a 16 when the dealer's upcard is a 10, but this is a very close play; one loses less by hitting than standing, but not by much. If it is known, however, that the deck is depleted of small cards such as 4s and 5s, and rich in 10s, that may alter the odds in favor of standing.

It is difficult for most people to remember what cards have already been dealt, particularly from a multiple deck shoe. Therefore, most card counting schemes assign a positive, negative, or zero point value to each card in the deck. Normally, low-value cards, such as a 2 or 3, are given a positive value, and 10s are given a negative value. The exact number assigned to the cards depends on the specific card-counting method. The card counter mentally keeps a running tally of the point values as they are dealt. To make the count an accurate representation of the percentage of "good" cards left in the deck, this running tally must normally be divided by a factor based on the counter's estimate of the number of undealt cards that are left (so-called unbalanced counts do not require this additional adjustment, because that is factored into the count). Highly skilled counters have an expectation of 1 to 1.6% gain; professional counters do exist, and number perhaps a few hundred. It is probably safe to say that the best counters earn in the very low six figures of US dollars, comparable with those in other mentally exacting fields.

If the tally is sufficiently high, the counter can increase his or her bet, and also may make modifications to basic strategy. All of these calculations must be accurate, at the same time that the dealer and other players may be talking to him, and it must be done in such a way that the casino does not notice that any counting is taking place, to avoid facing casino countermeasures. In practice, the vast majority of people who attempt to count cards lose money through errors; casinos who notice a counter will often check to see if the counter is good enough to have a positive expectation, and ignore them otherwise. This detection process is mistake-prone.

In addition, a card counter can play the Insurance bet if the count of faces is sufficiently high with potentially an advantage over the house; this bet is in general almost always disadvantageous.

Counting schemes that assign point values of –1, 0, or +1 are called level one counts and are considered the easiest to perform. Slightly greater accuracy, at the cost of increased difficulty and likelihood of making mistakes, involves the use of multi-level counts, which assign point values of –2, +2, or greater to the various cards. This greater range of point values adds to the complication of keeping an accurate tally in one's head.

A final complication in card counting involves the issue of how to treat aces. While playing out hands, Aces are slightly disadvantageous for the player, which implies that they should have a positive point count; but for purposes of getting a blackjack, they are extremely valuable when they remain in the deck. Most counting schemes give aces a negative count, recognizing that there is a compromise involved in this process. Some schemes actually assign a zero value to aces, and require the counter to keep a separate side count of aces.

The theory of card-counting, and the first counting scheme, was published in 1962 by American mathematician Edward O. Thorp in his book Beat the Dealer, which is now regarded as a classic in the gambling literature genre. Much of the specific detail in the work, however, is no longer up-to-date— end play, for example, has practically disappeared because the casinos no longer deal to the last card, in a (somewhat panicked, some say) response to the book. Also, the counting system described (10-count) is harder to use and less profitable than the point-count systems that have been developed afterwards.

The most commonly used system by most professionals (both players and surveillance) is Hi-Lo. It assigns -1 to 10's and Aces, +1 to 2 through 6. Higher level counts theoretically generate higher profits, but for most players, decreased playing speed and increased fatigue and error rates argue against their use. K-O, an unbalanced count (7's are also +1) developed by Ken Fuchs and Olaf Vancura (Knock Out Blackjack), is only modestly less effective than Hi-Lo, but is substantially less error-prone.

In the early days of card-counting, it is undoubted that a few players were hugely successful. Ken Uston recounts his early successes—and court battles with the casinos—in his book Ken Uston on Blackjack. In reality, Ken Uston, though perhaps the most famous card counter through his 60 Minutes television appearance and his books, was overall only a small winner. The most financially successful card counters have made their fortunes in other businesses. Ed Thorp, for example, runs a successful fund.

There have been several MIT Blackjack Teams, made up of MIT students who team up to use a combination of card counting and group play to attempt to beat the house.