Card counting is the only widely known method for gaining a mathematical edge at casino blackjack. The Hi-Lo system, introduced by Harvey Dubner in 1963, is the most commonly used counting strategy — and for good reason: it balances simplicity with effectiveness.
This guide walks through how Hi-Lo works, how to convert a running count into a true count, what edge you can expect at different count levels, and why deck penetration is the single most important variable most beginners overlook. For basic strategy fundamentals, start with our Blackjack Basic Strategy Calculator.
The Hi-Lo Card Values
Hi-Lo is a balanced, level-1 count. “Balanced” means the sum of all card values in a complete deck equals zero. “Level-1” means every card is assigned +1, 0, or -1 — nothing larger, which keeps it fast enough for real-time play.
| Cards | Hi-Lo Value | Count per Deck (×4) | Why |
|---|---|---|---|
| 2, 3, 4, 5, 6 | +1 | 20 cards | Low cards favor the dealer (less bust risk on stiff hands) |
| 7, 8, 9 | 0 | 12 cards | Neutral — roughly equal effect on player and dealer |
| 10, J, Q, K, A | -1 | 20 cards | High cards favor the player (more blackjacks, better doubles) |
Net count per complete deck: 0. Twenty +1 cards and twenty -1 cards cancel out exactly — this is what “balanced” means. When the running count is positive, more low cards than high cards have been dealt, leaving the remaining shoe rich in tens and aces. This benefits the player because:
- More blackjacks — which pay 3:2 (or should; avoid 6:5 games)
- Better double-down situations — a ten-rich deck makes doubling on 10 or 11 more profitable
- More dealer busts — the dealer must hit stiff hands (12-16) and is more likely to bust with high cards coming
- Insurance becomes profitable — at high true counts, the probability of a dealer ace hiding a 10 rises above the break-even threshold
Running Count vs True Count
The running count alone is not enough to make betting decisions. A running count of +6 means very different things depending on whether 5 decks remain or 1.5 decks remain. The true count normalizes per deck:
Example Walkthrough
Scenario: 6-deck shoe. Roughly 2 decks have been dealt. Running count: +6.
Decks remaining ≈ 6 − 2 = 4
True Count = +6 ÷ 4 = +1.5 (round to +1)
At TC +1, the house edge is roughly break-even. Not yet time to raise bets significantly.
Same running count, deeper into shoe: 4.5 decks dealt. Running count still +6.
Decks remaining ≈ 6 − 4.5 = 1.5
True Count = +6 ÷ 1.5 = +4
At TC +4, the player has roughly a 1.5% edge. This is when maximum bets are warranted.
Estimating remaining decks is a skill that requires practice. Most counters eyeball the discard tray — with experience, you can estimate to the nearest half-deck reliably. The Wizard of Odds Hi-Lo guide provides additional detail on rounding methods (floor vs truncate vs round).
Player Edge by True Count
The central question for any card counter is: at what true count does the edge swing from the house to the player? The following table shows approximate player edge at each true count for a standard 6-deck game (dealer stands S17, DAS allowed, no surrender, 3:2 BJ payout).
| True Count | Approx. Player Edge | Betting Action | Frequency (~) |
|---|---|---|---|
| −2 or lower | −1.5% or worse | Table minimum or Wong out | ~15% |
| −1 | −1.0% | Table minimum | ~15% |
| 0 | −0.5% | Table minimum | ~30% |
| +1 | ~0.0% (break-even) | Minimum or 2× min | ~15% |
| +2 | +0.5% | 4-6× minimum | ~10% |
| +3 | +1.0% | 8-10× minimum | ~6% |
| +4 | +1.5% | Maximum bet (10-12×) | ~4% |
| +5 or higher | +2.0% or better | Maximum bet, 2 spots | ~5% |
The ~0.5% per true count rule is a useful approximation. Starting from a baseline house edge of about −0.5% (with basic strategy), each +1 true count adds roughly +0.5% to the player’s expectation. This relationship comes from simulation data and is consistent across most standard rule sets, as documented in Don Schlesinger’s Blackjack Attack.
Notice the frequency column: you spend roughly 60% of your time at TC 0 or lower (betting minimum), and only about 25% of the time at TC +2 or higher (where you actually have an edge). This is why card counting is a grind — most of your session is spent waiting for favorable counts.
Deck Penetration: The Hidden Variable
Deck penetration — the percentage of cards dealt before shuffling — is arguably more important than the rules of the game. Deep penetration creates more opportunities for extreme true counts, which is where counters earn their profit.
| Penetration | Decks Cut Off | Relative Win Rate | Assessment |
|---|---|---|---|
| 67% (4 of 6) | 2.0 decks | ~40% | Barely worth counting |
| 75% (4.5 of 6) | 1.5 decks | ~65% | Playable but marginal |
| 83% (5 of 6) | 1.0 deck | 100% (baseline) | Good game |
| 87% (5.25 of 6) | 0.75 decks | ~120% | Excellent — increasingly rare |
The “Relative Win Rate” column uses 83% penetration (1 deck cut) as 100% baseline. At 67% penetration, your hourly expectation drops to roughly 40% of the baseline — meaning the same game with poor penetration may not be worth the effort and risk of detection. These penetration effects are well-documented by Stanford Wong in Professional Blackjack and confirmed by the simulation data at Blackjack Apprenticeship.
The Illustrious 18: Strategy Deviations by True Count
Basic strategy assumes you know nothing about the remaining deck composition. But card counters do know — the true count tells them. Don Schlesinger identified the 18 most valuable plays that deviate from basic strategy depending on the count, and published them in Blackjack Attack. These 18 plays capture approximately 80% of the total value from all possible index plays.
| # | Your Hand | Dealer Up | Index | Action if TC ≥ Index |
|---|---|---|---|---|
| 1 | Insurance | Ace | +3 | Take insurance |
| 2 | 16 | 10 | 0 | Stand (basic says Hit) |
| 3 | 15 | 10 | +4 | Stand (basic says Hit) |
| 4 | 10,10 | 5 | +5 | Split (basic says Stand) |
| 5 | 10,10 | 6 | +4 | Split (basic says Stand) |
| 6 | 10 | 10 | +4 | Double (basic says Hit) |
| 7 | 12 | 3 | +2 | Stand (basic says Hit) |
| 8 | 12 | 2 | +3 | Stand (basic says Hit) |
| 9 | 11 | Ace | +1 | Double (basic says Hit) |
| 10 | 9 | 2 | +1 | Double (basic says Hit) |
| 11 | 10 | Ace | +4 | Double (basic says Hit) |
| 12 | 9 | 7 | +3 | Double (basic says Hit) |
| 13 | 16 | 9 | +5 | Stand (basic says Hit) |
| 14 | 13 | 2 | −1 | Hit (basic says Stand) |
| 15 | 12 | 4 | 0 | Hit when TC < 0 (basic says Stand) |
| 16 | 12 | 5 | −2 | Hit when TC < −2 (basic says Stand) |
| 17 | 12 | 6 | −1 | Hit when TC < −1 (basic says Stand) |
| 18 | 13 | 3 | −2 | Hit when TC < −2 (basic says Stand) |
How to read this table: For play #2 (16 vs 10), the index is 0. This means: if the true count is 0 or higher, stand; if below 0, hit. Basic strategy says to always hit 16 vs 10, but once the deck is neutral or ten-rich, standing becomes the superior play because the dealer is more likely to bust. The Wizard of Odds publishes the Illustrious 18 with permission from Schlesinger and provides simulation data confirming these index numbers.
The first play — insurance at TC ≥ +3 — is by far the most valuable single deviation. At TC +3, the probability of a dealer ten in the hole exceeds 1/3, making the 2:1 insurance payout profitable. Below TC +3, insurance remains a sucker bet.
Hourly EV: How Much Can a Counter Actually Earn?
The romanticism of card counting movies aside, the math produces modest hourly expectations with enormous variance. Here is a realistic estimate for a $25 minimum bet, 1:12 spread on a standard 6-deck S17 game with 83% penetration and 80 hands per hour.
| Metric | Value |
|---|---|
| Average bet (weighted by TC) | ~$55 |
| Effective player edge | ~0.5-1.0% |
| Expected hourly EV | $22-$44/hr |
| Hourly standard deviation | ~$500/hr |
| Required bankroll (1% RoR) | $15,000-$25,000 |
The standard deviation is roughly 10-20× the hourly EV. This means individual sessions are dominated by luck, and you need hundreds of hours of play before your edge manifests reliably. For precise bankroll calculations, use our Risk of Ruin Calculator.
Related Blackjack Tools
GamblingCalc has a full suite of blackjack calculators covering every stage from learning basic strategy to professional card counting:
- True Count Calculator — Convert running count to true count interactively
- Deviations Calculator — Look up strategy deviations (Illustrious 18) by true count
- Penetration Calculator — Calculate how deck penetration affects your hourly EV
- Basic Strategy Calculator — Master perfect basic strategy before counting
- Blackjack EV Calculator — House edge by rule set and hourly cost analysis
- Blackjack Bankroll Calculator — Bankroll requirements for card counters
- Risk of Ruin Calculator — Universal risk of ruin formula