What Is Card Counting?

I’m working on a series of blog posts about gambling that I call the “what is” series. These posts try to get to the essence of what gambling in general is or what the essence of a specific game is. This post breaks that pattern slightly, because it’s about a specific strategy for a gambling game:

What is card counting?

It’s an advantage play technique most commonly used in the game of blackjack. I touched on it briefly in the post about what blackjack is, but I want to get into more detail in this post.

My hope is that someone who knows nothing about counting cards but understands how to play blackjack would be able to learn to count cards using nothing more than this post. This means the post will need to be thorough and concise at the same time.

I touched on the how and why of counting cards briefly in my blackjack post, but I want to thoroughly explain the concept here. The first thing to understand is that most card counters don’t memorize anything about the order of the cards in the deck. That’s just not necessary to gain an edge.

What the average card counter does is track the ratio of high cards (10s and aces) to low cards (2s, 3s, 4s, 5s, and 6s) that have already been played and that are left in the deck. When you shuffle a deck of cards, the distribution of high cards to low cards can vary based on where you are in the deck.

When there are proportionally more 10s and aces in the deck than low cards, you have an edge over the casino. If you bet more per hand when you have that edge, you can turn the entire game into a positive expectation experience for yourself.

Why do you have an edge?

It’s as simple as this:

You have a higher probability of being dealt a blackjack (or “natural.”)

A blackjack is a 2-card hand with a total of 21. Unless the dealer also has a blackjack, the hand automatically win, and you get paid off at 3 to 2 instead of getting paid off at even money.

Think about it this way:

If you are playing in a single-deck blackjack game, and all 4 aces have already been dealt, what’s the probability of being dealt a blackjack?

It’s 0%.

You MUST have an ace and a 10 to get a blackjack.

It gives you the opportunity to do in blackjack what you do in poker—bet and raise when you have good cards, and fold when you have bad cards.

You can find plenty of schemes and systems for tracking this data. One of the simplest is the ace-five count. You only track 2 cards as they’re dealt—the aces and the 5s.

Every time you see an ace dealt, you subtract 1 from the count. Every time you see a 5 dealt, you add 1 to the count. If you’re playing in a single-deck game, that’s all you need to, besides raising and lowering the size of your bets according to the count.

When the count becomes +2 or higher, you raise the size of your bet. The higher the count, the more you bet, as follows:

• When the count is +2, you bet 2 units, or $10.
• When the count is +3, you bet 4 units, or $20.
• When the count is +4, you bet 8 units, or $40.

Your “betting spread,” in this case, is $5 to $40. A betting spread is the amount of your lowest possible bet and the highest possible bet during a session.

You make all your strategy decisions based on basic strategy. Your edge over the house using this system and this betting spread is 0.30%.

What does this mean in terms of hourly expected win amounts?

Let’s assume you’re playing 50 hands per hour. On average, let’s say you’re wagering $10 per hand. (Sometimes you’re betting more, but most of the time, you’re betting $5 per hand.) That’s $500 per hour in action.

With a 0.30% edge over the house, your theoretical projected win is $1.50 per hour.

You could theoretically use this system with a game with 8 decks, in which case, you might be able to have a larger betting spread.

The size of your bet doubles with each +1, so your bet sizes would look like this if you wound up with a favorable deck in an 8-deck game:

• +5, 16 units, $80
• +6, 32 units, $160

This would double your edge over the house, which means you’d expect to win $3/hour.

At this point, you might be thinking, gee whiz. That’s an awful lot of effort to win less than minimum wage.

But keep in mind that this is just a simple card counting system with a low minimum bet. This isn’t the system most card counters use at all.

Most card counters use, or at least start, with the hi-lo system. This is a more robust system than the ace-five count I described. In this system, you still only add 1 or subtract 1 to the count. But you’re now going to keep up with a wider range of cards, as follows:

• 2s = +1
• 3s = +1
• 4s = +1
• 5s = +1
• 6s = +1
• 10s = -1
• Aces = -1

The other cards (7s, 8s, and 9s) are counted as 0—or ignored, basically.

If you’re counting from a single deck, you’ll size your bets according to the count +1. In other words, whatever your running count is, add 1 to it to determine how many units to bet. You should decide in advance what your betting spread will be, too.

Here’s an example, using $10 per hand as your single betting unit.

• If the count is 0 or negative, you bet $10 per hand (one unit).
• If the count is +1, you bet $20 per hand. (2 units).
• If the count is +2, you bet $30 per hand. (3 units).
• If the count is +3, you bet $40 per hand. (4 units).
• And so on, up to whatever the top of your betting spread.

But here’s the thing with the hi-lo count, as well as other, similar card counting methods:

If you’re playing with multiple decks, you need to account for the extra cards in the deck.

This requires converting your count (which is “the running count”) to the “true count.”

Here’s what all that means:

You can use the count as a rough estimate of your advantage against the casino at any time.

But when you’re dealing with 8 decks, the effect of each card becomes proportionally smaller to the overall probabilities.
This might not make sense at first, but let’s look at the probabilities:

Let’s say that you’re dealing with an 8-card game. That’s 52 cards X 8 decks, or 416 cards. You have 4 aces in each deck, which is 32 aces total.

32/416 is the same as 4/52, so the probability of getting an ace is the same in either situation. That’s 7.69%.
But once you’ve seen one of those aces deal, the probability of getting an ace changes.

With a single deck game, you now have a 3/51 probability of getting an ace. That’s roughly 5.88%, which is significantly different from your original 7.69%.

With an 8-deck game, you now have a 31/415 probability of getting an ace. That’s roughly 7.47%, which isn’t as high as 7.69%, but the difference is nowhere nearly as profound.

For example, if you have a count of +4, and you have 4 decks left in the shoe, you’d bet as if the count were +1.

+4 is the running count in that situation, but +1 is the true count.

The hi-lo system is far from the only card counting system out there. It’s one of the most basic, and it has some characteristics that can be useful to examine when thinking about other card counting systems.

One of those characteristics is that you’re only ever adjusting the count by +1 or -1. This makes it a single-level system.

Other card counting systems might count some cards differently from others. One system might count some cards as +1, +2, +3, -3, -2, and -1. That would be an example of a 3-level counting system. Generally, the fewer levels the system has, the easier the system is to implement. But the more levels it has, the more accurate it becomes.

Some systems are intentionally unbalanced. Usually, the designers of such system are hoping to eliminate the need to convert the running count into a true count. The red 7 count is one example. The KO system is another example.

Card counting systems that are more complicated are harder to implement but more accurate, which means that you play with a greater edge against the casino when using the harder systems. For the average card counter, though, using a more complicated system probably isn’t worth it.

You can also measure different aspects of a card counting system’s effectiveness.

Some card counters only get an edge by changing the size of their bets. In fact, that’s the only use for card counting I’ve discussed so far. You can measure how well a system correlates your bet size with your advantage using a number called “betting correlation.”

80% of a card counter’s edge comes from changing the size of his bets. The other 20% comes from deviating from basic strategy based on the count. Many card counters don’t even bother with the 2nd technique.

Let’s talk a little bit about basic strategy changes based on the count.

Most basic strategy decisions are straightforward regardless of the count.

But in some cases, your basic strategy changes will change based on the count. These changes are, again, based on the ratio of high cards to low cards in the deck.

The most obvious of these decision changes is whether to take insurance. I mentioned previously that insurance is a sucker bet with a huge house edge.

That’s no longer true in a deck with a higher ratio of 10s in it than normal. If the count is high enough, insurance becomes a positive expectation bet.

Most card counters focus on 18 situations where the count affects your strategy decisions. It’s beyond the scope of this post to outline all the strategy variations for these 18 situation, but you can easily find them by searching for “illustrious 18 card counting” or something similar.

Card counting isn’t cheating, and it isn’t illegal. The casinos consider it cheating, but they have no legal grounds for treating it as such. Still, they reserve the right to ask players to stay away from their blackjack games and stick with their other games.

In some cases, they’ll even ask a card counter to leave and never play any games at their casino again.

It’s not cheating, though. Think about it. All you’re doing is thinking about the conditions of the game that you’re playing as you’re playing it. It’s not much different from basic strategy when you think about it. The casino doesn’t expect you to ignore which cards you’re holding and/or which card the dealer has face-up.

Those are the same reasons that counting cards isn’t illegal. Who in their right mind would make it illegal to think hard about a game you’re playing? That would be like making it illegal to check-raise in a game of poker. It would be an arbitrary and unfair rule.

Now that I’ve cleared all that up, here are the practical considerations you need to account for. You don’t want the casino to think you’re counting cards or getting an edge over them. You need camouflage.

Some blackjack players wear disguises, as they know they’ve been identified previously. In fact, there used to be a private detective agency in Vegas that specialized in thwarting card counters—the Griffin Agency.

They published a book of headshots called “The Griffin Books,” and casinos used this book to help them identify card counters and run them off before they could make any money at the casino.

Other players wear disguises as a more general acting and persona building strategy. For example, some players might want to project an image of being a wild, superstitious gambler. This would give such a player a reason for deviating from basic strategy and for ranging his bets.

Of course, there are also the basic considerations. You don’t want to look like you’re thinking too hard at the table. You don’t want to seem like you’re paying too much attention to which cards the other players are getting. You don’t want to seem too smart or too serious.

I think most casinos are misguided in their enthusiasm for thwarting card counters, but I also don’t think that most of them plan to change their approach or ideas based on my opinions. Most would-be card counters aren’t good enough at counting to get an edge anyway.

But it would sure be a drag to get backed off by a casino, or barred altogether. Take reasonable precautions, and this never has to happen to you at all.

Having read this post, you might think that counting cards is easy. Conceptually, it’s not at all hard to understand.

But from a practical perspective, it’s tough to actually count cards in a casino and get an edge.

You need to practice at home first, because you need to be able to count cards perfectly and effortlessly. The only way to pull that off is to practice.

Get a stopwatch and a few decks of cards. Start by timing yourself as you count through a single deck of cards one card at a time. Use a balanced system like the hi-lo so that you’ll know if you got the count right. (If you wind up with a result other than 0, you lost count.)

You should practice with noise going on, too. Turn on the TV and maybe the radio, too. Let the kids play and argue in the same room where you’re practicing.

When you count cards in a real casino, you’ll face all kinds of distractions, and you’ll need to be able to maintain a perfect count and perfect concentration throughout.

Card counting is a reliable way to get an edge over the casino when playing blackjack, but it’s not for everyone. Most people can handle the math, but being able to concentrate that intently without drawing attention to yourself is probably the trickiest part.

Do you think counting cards is something you might try? Why or why not?