What Card Counting Actually Is
Card counting is a mathematical strategy for tracking the ratio of high cards to low cards remaining in the deck. It's not memorizing every card — it's maintaining a simple running tally that tells you when the remaining deck favors the player versus the dealer.
When the deck is rich in tens and aces (high cards), the player has an advantage: blackjacks are more likely, the dealer busts more often on stiff hands, and doubling down is more profitable. When the deck is rich in low cards, the dealer has the advantage.
Card counting doesn't guarantee wins on any single hand. It shifts the long-run edge from the house to the player by approximately 0.5–1.5%, depending on the count, rules, and penetration. That's a thin edge that only materializes over thousands of hands.
The Hi-Lo System
Hi-Lo is the most widely used counting system. It's a balanced, level-1 system, meaning it's relatively easy to learn while being effective enough for practical use.
| Cards | Count Value | Why |
|---|---|---|
| 2, 3, 4, 5, 6 | +1 | Low cards leaving the deck helps the player |
| 7, 8, 9 | 0 | Neutral cards — ignore them |
| 10, J, Q, K, A | −1 | High cards leaving the deck hurts the player |
As each card is dealt, you add its value to your running count. A positive count means more low cards have been dealt, so the remaining deck is rich in high cards — favorable for the player.
Count: −1, 0, +1, 0, 0, +1, 0, −1, +1, +2
Running count: +2 (deck favors the player)
Practice counting with an interactive trainer that deals cards and checks your count.
Open Card Counting Trainer →Running Count vs True Count
The running count alone doesn't tell the full story. A running count of +6 with five decks remaining is very different from +6 with one deck remaining. The true count adjusts for the number of decks left:
True Count = Running Count ÷ Decks Remaining
For example: running count of +6 with 2 decks remaining = true count of +3. The true count is what you use for betting decisions.
Practice converting running count to true count under pressure.
Open True Count Drill →How to Bet: Spreading Your Bets
The betting strategy is where the edge actually materializes. You bet more when the count is positive (deck favors you) and minimum when it's negative (deck favors the house). A typical bet spread:
| True Count | Bet Size | Edge (approx.) |
|---|---|---|
| ≤ 0 | Minimum ($10) | House has edge |
| +1 | 2x minimum ($20) | Roughly even |
| +2 | 4x ($40) | Player +0.5% |
| +3 | 6x ($60) | Player +1.0% |
| +4 | 8x ($80) | Player +1.5% |
| +5 or higher | 10–12x ($100–$120) | Player +2.0%+ |
Calculate optimal bet sizing using the Kelly Criterion.
Open Kelly Criterion Calculator →Bankroll Requirements
Even with a mathematical edge, you'll experience significant variance. Long losing streaks are normal and expected. The standard recommendation is a bankroll of 200–400 maximum bets to withstand variance.
With a 1–12 bet spread and $10 minimum bets, your max bet is $120 and you need a bankroll of $24,000–$48,000 to have a less than 5% chance of going broke. This is why card counting is not a get-rich-quick scheme — it requires substantial capital and patience.
Simulate thousands of sessions to see realistic bankroll swings.
Open Bankroll Simulator →Realistic Expectations
A skilled card counter playing 100 hands per hour with a 1–12 spread at a $10 minimum table can expect to earn roughly $15–$25 per hour on average. That's the mathematical expectation — but in practice, you might win $500 one session and lose $800 the next. The edge only emerges over tens of thousands of hands.
Other realities to consider:
- Casino countermeasures: Casinos use continuous shuffling machines, shallow penetration (shuffling early), and surveillance to deter counters. Being asked to leave (backed off) is common.
- Mental stamina: Counting for hours while maintaining a conversation, acting natural, and making correct playing decisions is exhausting.
- It's legal: Card counting using your brain is legal everywhere. Using a device is not.
Other Counting Systems
| System | Level | Accuracy | Difficulty |
|---|---|---|---|
| Hi-Lo | 1 | 97% | Easy |
| KO (Knockout) | 1 | 96% | Easiest (unbalanced, no true count needed) |
| Omega II | 2 | 99% | Hard |
| Hi-Opt II | 2 | 99% | Hard |
| Wong Halves | 3 | 99.5% | Very hard |
Hi-Lo captures 97% of the theoretical edge with far less mental effort than Level 2+ systems. For most people, mastering Hi-Lo is more valuable than learning a harder system imperfectly.