Hi-Lo Card Counting Tutorial: From Tag Values to Real-Table Play

A player is at his kitchen table with a shuffled deck face-down. He flips cards one at a time, saying nothing aloud, just tracking a single integer in his head. He flips the 52nd card and the integer reads zero. He's done it correctly. He looks at the timer: 22 seconds. He shuffles and starts again. After three weeks of doing this most mornings, the count is no longer a calculation — it's something his brain does in the background while he ties his shoes or watches the cards turn. He hasn't sat at a blackjack table once in those three weeks. But the work is most of what counting actually is.

Hi-Lo is the most-used card-counting system in blackjack, and not because it's the most powerful one. Wong Halves and Omega II are technically more accurate. Hi-Lo wins because it's the right balance — strong enough to extract most of the available edge, simple enough that a focused amateur can drill it to muscle memory in roughly a month and execute it under casino pressure without breaking concentration. This tutorial walks the full path: tag values, the running count drill, the true count conversion, the bet ramp, the deviations, and the practice ladder from kitchen table to live shoe.

Why Hi-Lo and not something else

Hi-Lo's stats from the TableSharp counting reference:

• Tags: 2-6 = +1, 7-9 = 0, 10/J/Q/K/A = -1
• Balanced: yes (the full deck sums to zero)
• Level: 1 (single-unit tag values, no fractions)
• Betting correlation: 0.97
• Playing efficiency: 0.51
• Insurance correlation: 0.76

Betting correlation is how well the count predicts when bets should go up. Playing efficiency is how well it tells you when to deviate from basic strategy. Insurance correlation is how well it tells you when to take insurance. At 0.97 betting correlation, Hi-Lo captures nearly all of the available bet-sizing edge. The 0.51 playing efficiency is the trade-off — higher-level systems like Omega II reach 0.67, but they're harder to learn and run under fatigue. For 95% of working counters, Hi-Lo's simplicity is worth the marginal accuracy loss.

Step 1: The tag values

Every card gets a tag value when it comes out of the shoe. Add the value to a running integer in your head. That integer is the running count.

• 2, 3, 4, 5, 6 → +1 (low cards leaving is good for the player — the shoe gets richer in 10s)
• 7, 8, 9 → 0 (neutral cards, no count change)
• 10, J, Q, K, A → -1 (high cards leaving is bad for the player — the shoe gets thinner in 10s and Aces)

The intuition: high cards (10s and Aces) help the player. They make blackjacks more likely, double-downs more profitable, and dealer busts more frequent. Low cards help the dealer — they let her draw out of stiff totals like 12-16 without busting. When the cards leaving the shoe are mostly low (RC going up), the shoe gets relatively richer in high cards. When the cards leaving are mostly high (RC going down), the shoe is getting starved of the cards that help you.

Step 2: The single-deck drill

Practice doesn't start in a casino. It starts with a deck in your hands. The opening drill:

1. Shuffle a standard 52-card deck.
2. Flip cards one at a time, face-up, into a discard pile.
3. Track the running count silently in your head.
4. At the last card, your count should be exactly 0 (Hi-Lo is balanced — a full deck sums to zero).
5. Time the run. Target: count down a deck in 25 seconds with a final RC of 0.

First attempts will be 90+ seconds and frequently off by 1 or 2 at the end. Don't be discouraged. The goal of week 1 is accuracy, not speed. Hit a clean 0 ten times in a row at any pace; then start pressing the clock. By week 3-4, the target is 25-30 seconds with reliable accuracy.

Two upgrades to the drill once basic accuracy is locked in:

• Flip cards in pairs (matches the way they come off the deal — two cards per player per round). Add their combined values. A 5 and a J cancel out to 0; two 6s sum to +2.
• Watch a televised blackjack game (or a YouTube playthrough) at casino speed and count the visible cards. This is the closest you'll get to live-shoe rhythm without leaving home.

TableSharp's counting trainer at /train/counting includes a Hi-Lo drill mode that times you, randomizes pace, and tracks your last-30 accuracy so you can see whether you're plateauing or improving.

Step 3: The running count vs true count

Hi-Lo is a balanced system. A running count of +6 means very different things in different shoe states:

• RC +6 with 6 decks remaining: ~1 extra high card per deck. Mild edge.
• RC +6 with 1 deck remaining: ~6 extra high cards in the last 52. Massive edge.

True count corrects for this. The formula:

TC = RC / decks remaining

Decks remaining is estimated from the discard tray. In a 6-deck shoe with 3 decks dealt, 3 decks remain. RC of +6 / 3 decks = TC of +2. The same RC of +6 with only 1 deck remaining is TC +6 — three times the edge per dollar bet.

Estimating decks remaining is its own skill. Most counters use the discard tray as the visual cue: half a tray of cards means 3 decks gone in a 6-deck game; a tray that looks 2/3 full means 4 decks gone; and so on. With practice you can eyeball it to the nearest half-deck without staring. TableSharp's decksRemaining helper rounds to the nearest half-deck, which is the precision live play actually needs.

Common rounding rule: round decks remaining down (toward fewer decks remaining), which rounds the true count up. That's the conservative bet-sizing direction; you'd rather be slightly aggressive than slightly under-bet on a hot shoe.

Step 4: When the true count matters

True count drives two decisions in real play: how much to bet, and when to deviate from basic strategy.

Bet sizing

The bet ramp is the table of bet sizes you make at each true-count level. A typical 1-12 spread example:

• TC ≤ 1: bet 1 unit (your table minimum)
• TC = 2: bet 2 units
• TC = 3: bet 4 units
• TC = 4: bet 8 units
• TC ≥ 5: bet 12 units (your max)

This is the betSpreadUnits ramp built into TableSharp. The exact widths vary by spread budget, table conditions, and how much heat you can absorb — a 1-8 spread is gentler, a 1-20 is sharper but draws attention faster. The shape is always the same: flat through neutral counts, ramped sharply once the count goes positive.

Deviations

At specific true count thresholds, basic strategy stops being optimal. The Illustrious 18 is the highest-EV set of these deviations. A few examples:

• Insurance at TC ≥ +3 — the single most profitable deviation. Below +3, never take insurance; at or above +3, always take it.
• Hard 16 vs 10 at TC ≥ 0 — stand instead of hit. Basic strategy says hit, but at any neutral-or-positive count the shoe is rich enough in 10s that standing wins more.
• Hard 12 vs 3 at TC ≥ +2 — stand instead of hit. The dealer's 3 is bust-prone enough at +2 that standing pat beats taking a risk card.

There are 18 of these total — see the dedicated Illustrious 18 article for the full table with all thresholds. Getting the deviations right adds roughly 0.10-0.15% in expected value on top of the bet-ramp gain. They're not optional once you're counting seriously.

Step 5: The practice ladder

Counting is a stack of three layers, and each one needs its own drill:

1. Layer 1 — tag values. Drill until you can call a card's value in under 0.5 seconds without thinking. Single-card flashcards, then deck-flip drills.
2. Layer 2 — running count under pace. Drill a full deck in 25 seconds with a clean 0. Then drill at casino speed (about 50-70 cards per minute) for 15 minutes at a stretch without errors.
3. Layer 3 — RC + TC + bet sizing + basic strategy + deviations under simultaneous load. This is where almost all aspiring counters fail. The skill is doing all of it at once, while a dealer asks 'cash or chips?', while the pit boss walks behind you, while the cocktail waitress takes your drink order.

A reasonable practice schedule: 30 minutes a day, 6 days a week, for 8-12 weeks before your first live session. Half the time on RC drills (decks and pace), a quarter on TC conversion (RC + decks-remaining calculations), and a quarter on basic-strategy refreshers and deviation drills. By week 8, the count should be running in the background while you play the hand in front.

Step 6: The cover act

Casino counter-detection is sophisticated. A player who only big-bets at high counts, only takes insurance at high counts, and only deviates from basic strategy at the chart-defined indices is wearing a sign. Real working counters wrap a layer of normal-player camouflage around the optimal play:

• Vary bet timing slightly — don't lock in the max bet the second the TC crosses your threshold.
• Order drinks, chat with the dealer, tip occasionally. A silent player who never breaks concentration is a tell.
• Occasionally take insurance at TC +2 or skip it at TC +3.5 (small EV cost, large heat reduction).
• Cap your spread. A 1-20 spread is mathematically optimal at most tables but invites a backoff in 3-4 sessions. A 1-8 or 1-10 spread is sustainable.

The cover act is its own skill. Most counters under-rate it until their first backoff.

Common Hi-Lo errors

Five mistakes that kill most aspiring counters:

1. Forgetting to convert RC to TC. RC +6 in a fresh 6-deck shoe is TC +1, not TC +6. Betting based on RC alone is the most common beginner error.
2. Estimating decks remaining wrong. If you call 4 decks remaining when it's actually 2, your TC is doubled and your bets are wildly wrong. Practice the discard-tray estimate as carefully as the count itself.
3. Counting only your own hand. Every visible card counts — your hands, the dealer's hands, other players' hands, the burn card. Missing 4 out of every 10 cards dealt halves your edge.
4. Adjusting the bet visibly with the count. Pit bosses notice. Always adjust on a new shoe, between hands, or after a clear win/loss event so the bet change looks like emotion, not math.
5. Quitting in a downswing and never returning. The variance in counting is large. Standard deviation per round at a 1-12 spread is around 1.3 units. A 1,000-hand losing run is well within normal. Bankroll management — Kelly fractions, RoR calculation — is half the job.

When to first sit at a real table

Hi-Lo is a long apprenticeship. A realistic first-session benchmark:

• Single-deck speed drill: 25 seconds, clean 0, ten times in a row at three separate sessions.
• Pace drill: 15 minutes of continuous counting at casino speed with no more than one error.
• Basic strategy accuracy at TableSharp's trainer: 99% or higher for two consecutive weeks.
• Illustrious 18 deviations memorized and tested under the counting trainer's drill mode.
• A defined bankroll, bet ramp, and exit rule before sitting down.

Most players who try counting quit during the drill phase because the work is unglamorous and the variance is real. The few who push through come out with one of the only positive-expectation games available in any casino anywhere.