| Updated: September 25, 2025As a highly popular casino game, blackjack is played by thousands of people, but few of them are proficient at card counting. Even fewer manage to earn a decent, sustainable profit from applying this technique at the blackjack table. Successful card counters, on the other hand, are skilled not only at keeping track of the cards that have been dealt but also at structuring their bets properly to maximize their advantage while remaining anonymous.
Once beginners learn the basic blackjack strategy, they can start learning how to count cards. Contrary to popular belief, card counting is not a skill that only exceptional individuals can master, and it does not require players to memorize every card in the deck. Most people can learn this technique, which enables them to identify the moment when the odds shift in their favor.
First Steps into Card Counting
Before blackjack players learn how to count cards like the pros, they should first gain a thorough understanding of the game. Knowing the rules and the basic strategy is a must, since even the best card counters will fail spectacularly if they do not base their tactics on mathematically proven moves and concepts. As an introduction to card counting, we present several fundamental ideas every blackjack player should focus on – what card counting is, how to determine the betting spread, and how the Kelly criterion applies to blackjack.
Fundamental Concept
Bankroll and Betting Spread The Kelly CriterionFundamental Concept
The concept of card counting in blackjack is quite simple – players keep track of the ratio of high to low cards remaining in the deck or the dealer’s shoe. They do this by assigning values to different cards, allowing them to see at a glance whether the deck is rich in high cards or in low ones.
Because blackjack is made with an Ace and a 10-value card, high cards (10, Jack, Queen, King, and Ace) are mathematically proven to work in favor of the player. Blackjacks pay the player 1.5 times the wager but only even money to the dealer; the dealer must hit on 16 and stand on 17; and the dealer is more likely to bust when more high cards remain in the deck.
Therefore, when the deck contains more high cards and blackjacks become more likely, the player stands to gain. Low cards, on the other hand, favor the house. Cards from 2 through 6 are considered low, cards valued 7 through 9 are neutral, and – as always – the greater the proportion of high cards remaining, the better the player’s odds.
Consider a single-deck blackjack game without a cut card or any burnt cards: we start with 20 high cards, 12 neutral cards, and 20 low cards. After several hands, the remaining deck might contain 17 high cards, 5 neutral cards, and 12 low cards. In this situation the player has the advantage, and to maximize profits, he should place a larger bet.
Novices must also remember this: once they identify a deck with favorable odds, they should increase their stake. If, however, the deck holds 17 low cards against 12 or fewer high cards, the house has the advantage and players should keep their bets at the minimum. By following this principle, they will win more on hands that are statistically likely to succeed and lose less on hands that are likely to fail.
Bankroll and Betting Spread
The most important aspect of card counting – and the one that actually helps players make money – is the proper adjustment of bet size based on the count. When the count is high and suggests good odds, players raise the bet; when the count drops, they reduce the wager. This variation in bet size is known in blackjack literature as the betting spread (or bet spread).
The optimal betting spread differs for each player because it depends on several factors: the table limits, the maximum amount the player is willing to wager, and the size of his or her bankroll. The bankroll is simply the money set aside exclusively for blackjack play, whether that is $100, $500, or $5,000 per session. Novices should begin with a modest bankroll and a smaller betting spread to better understand how the counting system works.
For example, suppose their betting spread is $5 – $100. Their lowest bet would then be $5, or 1 unit, while their highest wager would be $100, or 20 units. The same spread could also be expressed as 1-20. However, a 1-20 unit betting spread is aggressive and risky, so most players prefer spreads of 1-10 to 1-15. In fact, casinos become suspicious when a player suddenly increases a bet seven times after wagering the table minimum for half an hour.
Selecting the proper betting spread is therefore essential. Equally important is the expected return for a given betting spread, as well as the so-called risk of ruin, which will be discussed in detail in the following sections.
The Kelly Criterion
Several methods help players determine the optimal bet size in blackjack and the number of units to wager when the count is high or low. One method that has been mathematically proven is the Kelly criterion, a principle that balances risk and potential reward. Developed in 1956 by John L. Kelly Jr., a researcher at Bell Labs, it has since been applied successfully to both gambling and investment theory.
The Kelly criterion is a mathematical formula that defines what fraction of a bankroll should be wagered at any point in the game, based on several variables, including the likelihood of winning and losing. The formula is:
f = pb-q / b
Here, f is the fraction of the bankroll we should bet per hand, b is the odds offered on the wager (b to 1), p is the probability of winning, and q is the probability of losing. After we determine the variables on the right side of the equation, we can easily calculate the bet per hand. For instance, assume the odds for the next hand are exceptionally good and the probability of winning is 75% (0.75). The probability of losing will then be 1-0.75=0.25, or 25%. In this case, b is 1.5 (derived from the 3:2 or 1.5:1 payout for blackjack).
Using these variables, the bet size should be 0.75*1.5-0.25/1.5 = 0.5833, or 58% of the bankroll, which is, of course, extremely aggressive betting. We can approximate the same Kelly formula as f = a/v, where a represents the player advantage and v denotes the game’s variance. Variance is the square of the standard deviation. The standard deviation of blackjack is typically around 1.15 bets, although it may vary significantly depending on the rules, so in our example the variance is 1.3325.
If the player calculates that, on a certain count, he has an advantage of 2% (0.02, which is quite high) and the variance is 1.3325, then f will be 0.02/1.3325 or 0.015. In other words, the bet size in this case should be 1.5% of the bankroll – if the player has a dedicated bankroll of $1,000, the proper wager here is $15.
Basics of Card Counting
As mentioned above, card counters do not memorize every card that has been dealt in order to calculate exactly which cards remain in the dealer’s shoe. Doing so is impossible in a fast-paced game like blackjack, especially when 6 or 8 decks are in play. Instead, card counters roughly estimate the ratio of high to low cards in the shoe without knowing the exact cards that will be dealt next. The card-counting process can be broken down into four basic steps:
Assigning Values to Cards Keeping a Running Count Calculating the True Count Adjusting Bet Size Spotters The Big Player Avoiding DetectionAssigning Values to Cards
There are more than a dozen methods for tracking the decks and counting cards in blackjack, but we will start with the simplest and most straightforward system: the Hi-Lo. It divides cards into three groups – low (2-6), neutral (7-9), and high (A, 10). This system assigns the following values to each group:
- 2-6 – +1
- 7-9 – 0
- 10, Ace – -1
Keeping a Running Count
The sum of the point values for all cards is exactly 0, and the Hi-Lo system is very easy to learn and apply. We start the count at 0; each time a low card appears, we add 1, and each time the dealer reveals a high card, we subtract 1. Neutral cards have no effect. This ongoing tally is called the “running count,” and players should know it at all times during the game.
When the running count is high (1 or more), it indicates that more low cards than high cards have been played since the start of the shoe. Consequently, more high cards remain, giving the player the advantage. For example, three players are betting against the dealer, and their hands are 7-4, 2-Ace, and Jack-6, while the dealer’s upcard is a 5. The running count in this situation is +2, which signals favorable odds for the players.
Calculating the True Count
So far, the Hi-Lo system seems simple and straightforward, but to make counting more difficult, casinos use 6 or 8 standard decks of cards. This means there can be up to 416 cards at the start of the game. Therefore, players must consider the number of decks in play by converting the running count into a “true count”.
We can calculate the true count, or count per deck, in several ways, but the easiest is to divide the running count by the number of decks remaining in the shoe. If the game begins with 6 decks and the running count is 8, and after several hands only 4 decks remain, the true count is 8/4=2 (+2).
Sometimes the true count is not a whole number – if we estimate the remaining decks in the same example to be 5, the true count would be 8/5=1.6. Most card counters round to the nearest whole number or negative integer; the true count could be negative, as well.
Another critical skill is estimating the number of decks left in the shoe. A perfect estimate is impossible – some players look at the cards still in the shoe, while others examine the discard tray to see how many decks have been dealt. No method is absolutely precise, however.
Players can also try to count the exact number of cards seen – those on the table plus those in the discard tray – but this is quite difficult for most people, especially while trying to maintain a running count during play.
Adjusting Bet Size
Once players estimate the true count and see whether the odds are stacked against them or in their favor, they can adjust their bets accordingly. If they keep betting the same amount throughout the game, the entire process of counting and estimating the true count is practically useless. When the true count rises, it suggests that the house has lost its advantage and it is time for players to raise their bets.
This is the only way to capitalize on favorable odds and on your card-counting skills. With most counting systems, it is best to increase the bet size once the count reaches +2 or higher. The exact bet per hand will, of course, depend on the player’s bankroll and predetermined betting spread. The wager should increase proportionally with the true or running count – for instance, the betting spread is 1-10 and the basic betting unit is $5, the table minimum.
- True count is negative – Do not play the game
- True count is 0 – Bet $5
- True count is 1 – Bet $5
- True count is 2 – Bet $10
- True count is 3 – Bet $15
- True count is 10 – Bet $50
Team play has long been popular among blackjack card counters, and it has been depicted – albeit imperfectly – in several books and movies. The famous story of the MIT Blackjack Team, for example, is not entirely accurate; over the years, multiple teams have recruited students from the Massachusetts Institute of Technology, and many other card-counting groups have existed as well.
While many people believe these teams are no longer active, some members still play in casinos. Moreover, it is almost certain that new and more advanced group counting tactics have been developed and that counting teams likely operate in casinos even today, despite modern counter-detection methods.
Numerous schemes allow groups of players to gain a significant edge over the casino, but the most common systems employ several “spotters” and at least one “big player”. This type of group play was invented by Al Francesco, who has trained and supported multiple teams over the years.
Spotters
Spotters play a critical role on a card-counting team – they are the people who maintain the running count at the tables. Sometimes a spotter will sit at the table, wagering only the minimum. This allows the spotter to observe the game and keep the count without arousing casino suspicion.
Alternatively, a spotter may stand behind a blackjack table (a practice known as wonging) and watch the cards until the count climbs. This can be risky because blackjack tables rarely attract spectators – unlike roulette or craps tables, where many onlookers gather without betting.
When the count rises, the spotter discreetly signals another player to begin wagering large amounts in order to capitalize on the favorable odds. When the count falls, the spotter informs the player that the table is no longer hot.
The Big Player
A typical counting team includes several spotters and one “big player.” When the spotters identify a “hot” table – a table where the count is high – they signal the main player, who is usually the most skilled member. The Big Player must know perfect strategy and feel comfortable wagering large sums.
If the count is exceptionally high, this player wagers the maximum allowed by the casino. With only a slight edge, he spreads his bets according to the running count. When the table cools down, the Big Player either leaves or reduces the wager size.
Avoiding Detection
Avoiding detection is the most challenging aspect of team card counting – because the practice is strictly forbidden in casinos, all team members should try to remain inconspicuous. Teams use various methods to avoid scrutiny, from employing discreet signals to adding extra players. Along with the spotters and the Big Player, some teams have also had back-spotters and “gorilla” players.
A gorilla player watches for signals from the spotters but, instead of keeping a low profile, is loud, extravagant, and flashy. His job is to draw attention to himself and away from the team. This player switches between hot tables frequently and places large bets.
Even with additional players and very complex tactics, card counting teams are often detected by casinos since all these strategies have already been described in books and films. Floor managers and security experts usually know how to easily spot card counters and card counting teams. This is why group counting is no longer so popular. It is not advertised publicly and if such teams still exist, they are probably extremely secretive and secluded.
- Multi Hand Blackjack
- Atlantic City Blackjack
- European Blackjack
- Blackjack: Introduction to the World’s Most Popular Casino Game
- How to Play Blackjack
- Blackjack Hands
- Blackjack: The Good and the Bad Rules
- Mathematics of Blackjack
- Blackjack Strategy: The Best Way to Win Consistently
- Blackjack Side Bets
- Blackjack Variations
- Blackjack: Tips for Beginners
- Advantage Play in Blackjack
- Money Management in Blackjack
- Edward Thorpe and the Ten Count System
- The Zen Count System
- The Advanced Omega II Card Counting System
- The Uston SS Card Counting System
- The Ace/5 Card Counting System
- The Knock Out Card Counting System
- The Red 7 Card Counting System
- The Wong Halves System for Card Counting
- Choosing the Best Blackjack Game Online
- How to Play Live Dealer Blackjack
- Microgaming Blackjack
- RTG Blackjack
- NetEnt Blackjack
- Mobile Blackjack Apps
Additional Factors to Consider
Certain card counting systems, such as the Hi-Lo, are very easy to learn, but to apply them quickly and confidently in real-money games, players must practice. They can start with a single deck to learn the basics, but anyone planning to use the system on the casino floor should be well-prepared. Preparation includes mastering the system for extended periods in multi-deck games, maintaining an adequate bankroll, and knowing how to adjust bet sizes as the true count rises and falls.
Naturally, card counters should always look for tables with favorable rules to maximize their edge. How large can that edge be? It varies significantly from one counting system to another, but in most cases skilled counters can gain up to a 2.5% advantage over the house; more commonly, the edge is around 1% to 1.5%. Before sitting at the blackjack table, players should also consider deck penetration, game variance, and the theoretical Risk of Ruin and Expected Profit.
Deck Penetration Variance Expected Profit and Risk of RuinDeck Penetration
Earlier in this guide we defined “deck penetration” as the percentage of cards dealt from the shoe before the dealer reshuffles. If 39 cards have been dealt in a single-deck game, for example, penetration is 75%. For most blackjack players, even those who use basic strategy, deck penetration is not especially important.
For card counters, however, penetration is critical. Deeper penetration yields a more accurate running or true count – when more cards are dealt, the counter collects more information. In six-deck games, up to 100 or 130 cards are often cut off, dramatically affecting the player’s edge. With the Hi-Lo system and a 1-12 betting spread, the advantage in a six-deck game is about 0.64% when two and a half decks are cut off (130 cards). When the penetration is deeper (around 92%) and only 26 cards are cut off, the advantage jumps to around 1.67%.
Typically, card counters avoid double-deck games when more than one deck is cut off (penetration below 50%) and shoe games with 4 to 8 decks when more than 2 decks are cut off. A good rule of thumb is to never play on tables with penetration of less than 75% penetration.
Variance
Another crucial concept for card counters is game variance, which can be described as the “luck factor.” Even in games with deep penetration of around 80%, counters cannot predict which cards will be dealt next. The count they keep simply shows whether the shoe is rich in high-value cards or not.
The next cards will always be controlled by the randomness of the shuffle. Variance measures how much that randomness will affect the advantage we gain through good strategy, deep penetration, card counting, etc. The advantage, as expressed in percentages, is only a theoretical concept and reaching it typically takes thousands of rounds, if not more.
Expected Profit and Risk of Ruin
Expected profit – often called expected value (EV) – is one of the first concepts card counters examine, though they do not always grasp it fully. Expected value, expectation, or long-term expectation, expresses the worth of each decision in the game. It could be expressed as a percentage or as a fraction and it is used in various casino games, including blackjack, poker, and video poker. The expected value depends on multiple factors such as the penetration, the betting spread, the game rules, the strategy used, as well as the number of hands that are played per hour.
Because variance is considerable in blackjack, players would typically need hundreds of hours of play to make a good profit. Since the EV changes on every hand, it could be used to make decisions – they would affect the player’s bankroll in the long term, however. For instance, the dealer’s upcard is a 3, while the player holds a 12 – that’s a tough decision for most blackjack players and the right decision is to Hit.
The reason is straightforward – based on the probabilities, the EV of Hitting is -23.30%, while Standing offers an EV of -25.20%. Both are negative, which means that the player is more likely to lose with both decisions, but it is better to lose less, especially in the long term when losing less is nearly as equally good for the bankroll as winning.
Another concept players must understand before risking real money is the Risk of Ruin (ROR) – the possibility of losing the entire bankroll. In a negative-expectation game such as blackjack, the ROR is always present even when the player uses a good card counting system. Although it cannot be eliminated, however, it could be reduced by increasing the player’s bankroll – not as an amount of money but in terms of a number of betting units.
As a guideline, start with a bankroll that is at least 200 betting units for a ROR of around 40%. This means that four in every ten players are expected to lose their entire bankrolls. This applies to games with standard rules, a betting spread of 1-12, and a perfectly executed strategy. As you can see, this is a high-risk game and it would be safer to increase the bankroll (or divide it into more betting units) to avoid going broke. The Risk of Ruin would be 20% with a bankroll of 400 units, around 10% with 500 units, and around 1% when the player has 1,000 betting units.
Advantages and Disadvantages of Card Counting
Card counting is the only method for gaining mathematical advantage over the house.
It is difficult to count cards in a real-life situation.
Skilled card counters can make huge profits from blackjack.
Card counting does not work for virtual blackjack in online casinos.
There are many simple systems for counting cards that are easy to learn and implement.
Not all card counters manage to make a living from blackjack.
The different methods are readily available online, in various blackjack books, and guides.
The most efficient counting systems are also the most complex and difficult ones.
Card counting could be applied to all blackjack variations.
Card counting is strictly forbidden in casinos, although it is perfectly legal in most jurisdictions.
Card counting allows deviation from basic strategy when necessary.
Card counters must remain inconspicuous and pretend to be amateurs.
Card counting could be practiced solo or in teams.
Casinos take a wide range of measures to detect card counters. Even suspicion could mean lifelong ban from the casino.