Learn Card Counting

How to Use This Tool

Card Counter Pro is an educational simulator for learning blackjack card counting. Configure your table rules on the Game Variant tab, then practice counting live in the interactive Play tab. Sharpen each sub-skill on the Trainer tab, print the optimal chart for your exact rules on the Strategy Chart tab, and run Monte Carlo simulations on the Simulate tab to see your theoretical edge at each true count—this is the distribution you’re playing for. The Deviations tab shows exactly when to deviate from basic strategy, and the Profits tab projects your expected earnings based on your bet ramp and team size.

Getting Started

Start small. The biggest mistake new counters make is jumping in with real money before their count is automatic. Gather 30 hours in our simulator before you try counting live.

Use a team bankroll. A team of 3–5 players sharing one bankroll is one of the best ways to make money counting cards. Each player sits at a separate table, pooling results at the end of the session. The team’s expected profit scales linearly with headcount, but the standard deviation only grows by the square root—this is the fundamental mathematical advantage of team play. A solo counter with a $10,000 bankroll faces brutal variance. Four players with a pooled $40,000 bankroll have the same expected hourly rate per person but cut their collective risk of ruin dramatically.

Learn basic strategy cold before adding the count. If you can’t play perfect basic strategy under casino pressure—noise, distractions, cocktail waitresses—the count won’t save you. The count only adds 0.5–1.5% on top of near-perfect play.

Once you’re ready, start with Hi-Lo. It’s the standard for a reason: simple enough to maintain under pressure, powerful enough for serious profit. Move to level-2 systems only after Hi-Lo feels effortless.

The Hi-Lo Count

Hi-Lo is the most widely used system in the world and the only count we recommend. Simple level-1 tags, excellent betting correlation (0.97), and every published deviation index is calibrated to it. Master Hi-Lo before considering anything more complex.

The Tag Values

+1: 2 · 3 · 4 · 5 · 6 0: 7 · 8 · 9 −1: 10 · J · Q · K · A

Every card you see modifies your running count by its tag. Low cards (2–6) favor the dealer—they help dealers make a hand without busting. High cards (10s and Aces) favor the player—they make blackjacks and bust the dealer’s stiff hands. When low cards leave the shoe, the remaining cards skew high, and the player’s edge grows.

Step 1: The Running Count

Start at 0 on a fresh shoe. As cards are dealt—yours, the dealer’s, every other player’s—add the tag value to your running total. Practice in pairs: a 4 and a King cancel to 0. A 5 and a 6 add to +2. A 10 and a Queen are −2.

Cancel as you go. Dealer turns up a 5 and you have a Jack? Net 0—don’t bother. Look for pairs that sum to zero across the whole table to reduce mental load.

Step 2: The True Count

The running count alone is meaningless without knowing how many decks are left. A running count of +6 with one deck remaining is much stronger than +6 with six decks remaining. Convert to true count:

True Count = Running Count ÷ Decks Remaining

You don’t need exact deck precision—estimate visually from the discard tray. With practice you’ll eyeball ½, 1, 1½, 2, 3, 4, 5 decks remaining. Half-deck precision is enough.

Step 3: What the True Count Means

Each unit of true count is worth approximately 0.5% in player edge. The exact MC-simulated edges (from our 13M-hand sim) are on the Deviations tab. Your bet ramp follows these numbers:

True CountApprox. EdgeAction
≤ 0House: 0.4–2%Bet minimum
+1~0.1% houseBet minimum (still negative)
+2~0.6% player2× minimum
+3~0.9% player4× minimum; take insurance
+4~1.5% player6× minimum
+5~2.2% player8× minimum
+6 or more+3.3% or moreMax bet

Step 4: Playing Deviations

At certain true counts, the correct play deviates from basic strategy. We have a full table of MC-derived indices on the Deviations tab. The most valuable:

  • Insurance at TC ≥ +3: Take insurance when offered (single most valuable deviation)
  • 16 vs T at TC ≥ 0: Stand instead of hitting
  • 15 vs T at TC ≥ +4: Stand instead of hitting
  • T,T vs 5 at TC ≥ +5: Split instead of standing
  • 12 vs 4 at TC ≤ 0: Hit instead of standing

Practice Drills

Single Deck Speed. Take a shuffled deck. Flip cards one at a time, counting. A balanced count ends at 0. Time yourself. Pros do it in under 25 seconds.

Pair Cancellation. Deal two cards at a time. Train your eye to instantly see net values. A 5+J is 0. A 4+6 is +2. A pair of 10s is −2.

Live Practice. Use the Play tab here. Click “View Count” periodically to check your running count and true count against the app.

Strategy + Count. Once comfortable, add playing decisions. Basic strategy must be automatic before the count helps you.

Common Mistakes

  • Losing the count under pressure. Practice with deliberate distractions.
  • Forgetting to convert to true count. A +5 running count on shoe 1 of 6 is nothing. Convert every time you’re about to bet.
  • Betting too aggressively for your bankroll. See the Risk of Ruin calculator on the Profits tab. A $25–$300 spread needs roughly $28K for <5% ruin risk.
  • Playing against bad rules. 6:5 blackjack adds 1.4% to the house edge—a death sentence.
  • Wonging in too aggressively. Back-counting (only playing when count is positive) draws heat fast.

What to Expect

Card counting gives you a 0.5–1.5% edge over the house in favorable games. For every $100 wagered, you profit $0.50–$1.50 on average. The variance is enormous—you can and will have losing sessions, losing weeks, even losing months. A typical $25–$300 spread with a $10,000 bankroll produces roughly $30–$50/hour in expected value.

See the exact MC-simulated edge distribution on the Deviations tab. Use the Profits tab to project team earnings with your specific bet ramp.

Choosing a Counting System

Every viable counting system trades simplicity against power. The two numbers that matter most are Betting Correlation (BC) — how well the count predicts your edge, which drives when to bet big — and Playing Efficiency (PE) — how well it informs strategy deviations. Insurance Correlation (IC) measures accuracy on the single most valuable deviation. Higher-level systems squeeze out a few extra points of PE, but they demand harder card tags, side counts, and flawless arithmetic under pressure — and a level-2 count played with errors is worse than a level-1 count played perfectly.

SystemLevelBalanced?BCPEICSide CountDifficultyBest For
Hi-Lo 1 Yes ~0.97 ~0.51 ~0.76 None Easy Everyone — the standard starting system
Knock-Out (KO) 1 No ~0.98 ~0.55 ~0.78 None Easy Players who want to skip true-count division
Hi-Opt I 1 Yes ~0.88 ~0.61 ~0.85 Ace (optional) Medium Players prioritizing playing efficiency
Hi-Opt II 2 Yes ~0.91 ~0.67 ~0.91 Ace (recommended) Hard Serious, disciplined players
Omega II 2 Yes ~0.92 ~0.67 ~0.85 Ace (recommended) Hard Serious players wanting top playing efficiency
Zen Count 2 Yes ~0.96 ~0.63 ~0.85 None required Hard A strong all-round level-2 balance of betting and playing
Ace-Five 1 Yes High (betting only) ~0.00 None Very Easy Absolute beginners; pure bet-ramp camouflage

BC, PE, and IC figures are approximate, illustrative values drawn from published blackjack literature (Wong, Snyder, Schlesinger, and others) and vary slightly by source. Ace-Five tags only Aces (−1) and 5s (+1) and ignores playing strategy entirely, so its PE is effectively zero. Real-world performance depends far more on table rules, deck penetration, and your bet spread than on which system you pick.

Which Should You Use?

For almost everyone, the answer is Hi-Lo. It is the most-documented, most-taught system in existence, its ~0.97 betting correlation captures the overwhelming majority of your available edge, and every strategy chart, deviation index, and training tool (including this one) is built around it. Master Hi-Lo first; do not even consider anything else until you can run a deck error-free and convert to a true count in your head without slowing down.

  • Start here — almost everyone: Hi-Lo. Balanced, level 1, easy, and the universal standard. The marginal gains from harder systems rarely survive contact with real play.
  • Hate true-count division? Use KO (Knock-Out). Its unbalanced design bakes the count adjustment into a running count, so there is no division step — at a tiny cost in flexibility across game conditions.
  • Total beginner or want maximum camouflage? Use Ace-Five. It is trivial to run and great for practicing a bet ramp, but it offers no playing-strategy advantage — treat it as a training-wheels or low-effort spread, not a serious edge system.
  • Disciplined and already fluent in Hi-Lo? Only then consider a level-2 system — Zen Count for the best all-round balance, or Hi-Opt II / Omega II (with an ace side count) for peak playing efficiency. These add a few points of PE, but a single tag error per shoe can erase that gain.

The hard truth: the system you execute perfectly beats the one you execute almost perfectly. Edge comes from penetration, rules, and a disciplined bet spread — not from chasing the last decimal of playing efficiency.

Frequently Asked Questions

What is the best card counting system for beginners?

Hi-Lo is the most widely used system and the only one we recommend to start. It uses simple level-1 tags, has excellent betting correlation (0.97), and every published deviation index is calibrated to it. Master Hi-Lo before considering anything more complex.

How much edge does card counting give you?

Counting adds roughly 0.5–1.5% on top of near-perfect basic strategy, depending on the rules, deck penetration, and your bet spread. The count only helps if you can already play perfect basic strategy under casino pressure.

Does card counting work in online blackjack?

Traditional card counting does not work at most online blackjack games because RNG games reshuffle every hand. Only live-dealer games dealt from a real shoe offer any opportunity, and even then early cut cards and flat or narrow bet limits sharply reduce the edge.

How should I practice before counting for real money?

Learn basic strategy cold first, then gather about 30 hours in a simulator until your count is automatic before playing live. A pooled team bankroll cuts variance: expected profit scales with headcount while standard deviation grows only with its square root.

Game Variant

Configure the blackjack rules for your specific table. These settings affect the base house edge and your bet ramp.

Counting & Betting

Bankroll: $10,000 $0 | Bet:
Dealer

Real Money Online

Online blackjack is a different game from live. Here’s what actually works and where you can play legally.

Reality Check: Card Counting Online

Traditional card counting does NOT work at most online blackjack games. Here’s why:

  • RNG games shuffle every hand. The deck is reset to full after each round. No count can develop. This includes every “Blackjack” game at every online casino.
  • Live dealer games cut early. Most live dealer blackjack shoes are 8 decks with the cut card placed at 50%. Even with perfect counting, this penetration is too shallow for a profitable edge after the spread vs. heat tradeoff.
  • Bet limits are flat or narrow. Live dealer tables often have low maximums relative to minimums, killing the bet ramp you need to extract edge from a positive count.

If you want a real shot at card counting, you need live in-person play. See the Where to Play tab.

What Actually Works Online

Card counting fails online, but other advantage plays can still profit:

  • Bonus abuse / advantage play. Welcome bonuses with reasonable wagering requirements + 3:2 blackjack = positive EV. Most jurisdictions allow this. You play through the bonus on the lowest-house-edge game (basic strategy blackjack), banking the bonus as profit.
  • Match play / promotional offers. Casinos run promos (deposit matches, free play, leaderboard contests) that can be EV+ if you read the terms carefully.
  • Specific live dealer rooms. A few operators run live blackjack with 65%+ penetration. Rare, but they exist. Verify before committing.
  • Side bet counting. Some online side bets (Lucky Ladies, Royal Match) can be exploitable when you count the specific cards that resolve them.

Legal Real Money Online Blackjack (USA)

Online casino gambling is regulated state-by-state. As of 2026, these states have legal real-money online blackjack:

StateLegal OperatorsNotes
New JerseyDraftKings, BetMGM, FanDuel, Caesars, Borgata, BetRivers, Hard RockMost mature market; best variety
PennsylvaniaDraftKings, BetMGM, FanDuel, Caesars, BetRivers, PokerStarsStrong live dealer offerings
MichiganDraftKings, BetMGM, FanDuel, Caesars, BetRivers, PokerStarsMature market since 2021
West VirginiaDraftKings, BetMGM, Caesars, BetRiversSmaller market
ConnecticutDraftKings, Mohegan Sun, FanDuelLimited operators
DelawareState-run only (888 Holdings)Limited selection
Rhode IslandBally’s BetNewest market (2024)

Other states allow only retail/sports betting. Always verify your physical location is within a legal state before depositing.

UK & Europe

Most of Europe has regulated online casino markets. UKGC-licensed operators are the gold standard for player protections.

RegionNotable OperatorsNotes
UK (UKGC)Bet365, William Hill, Sky, Ladbrokes, PartyCasino, Mr GreenStrong regulation; live dealer common
SwedenCasumo, LeoVegas, Unibet, BetssonSpelinspektionen licensed; clean market
GermanyTipico, Mr Green (regulated)State licensing since 2021
Spain888, Sportium, PokerStarsDGOJ regulated
Malta (MGA)Many international operatorsCommon license for EU operators

What to Look For

3:2 blackjackNon-negotiable. Walk away from 6:5.
S17 (dealer stands soft 17)Reduces house edge by 0.22%
DAS allowedAlways preferred
Late surrenderRare online but exists in some live dealer rooms
Reasonable rolloverFor bonus play: 10× or lower wagering = EV+ possible
Game weightingBlackjack often counts 5–25% toward wagering — read fine print

This site does not run affiliate links or recommend specific operators. Verify licensing, terms, and your local laws before depositing real money.

Deviations

Index-based playing deviations and bet sizing for Hi-Lo counting. Deviate from basic strategy when the true count crosses the listed threshold.

Where these numbers come from

Edge by True Count (the bet ramp below): From our own Monte Carlo simulation — 13 million hands across 300,000 shoes on 6-deck S17 DAS LS at 75% penetration with Hi-Lo. Edge values reflect actual play with deviations applied.

Deviation indices (the matrix below): Published Hi-Lo indices from Stanford Wong’s Professional Blackjack (the Illustrious 18) and Don Schlesinger’s Blackjack Attack (the Fab 4 surrenders). These were derived via exact combinatorial analysis, which is more precise than Monte Carlo for individual cell EVs. We use them directly — the math has been peer-reviewed for 30+ years.

Bet Sizing by True Count

Optimal wager at each true count based on your spread. Adjust on the Game Variant tab.

Playing Deviation Matrix

Cell format: +3 HS = Hit below TC +3, Stand at TC ≥ +3. Colored cells deviate from basic strategy at the listed index.

S Stand D Double P Split R Surrender
Insurance: Take at TC ≥ +3 (Hi-Lo). Basic strategy: never insure.

Hard Totals

Soft Totals

Pairs

Illustrious 18 + Fab 4 (Ranked by Value)

The 22 most valuable index plays for Hi-Lo, ordered by expected gain.

Profit Calculator

Team Total (100 hrs)
$ / Per Member
Hourly EV
Hourly SD
Win %/hr

Configuration

Bet Size by True Count

Risk of Ruin

Probability of losing your entire bankroll before the edge plays out. Based on your bet ramp above.

Risk of Ruin
Bankroll / Avg Bet
Kelly Fraction
Recommended Min Bankroll

Where to Play

Casinos worldwide with potentially countable blackjack games. Look for 3:2 payouts, S17 (dealer stands on soft 17), DAS (double after split), late surrender, and deep penetration (75%+).

What Makes a Game Countable

Must Have
  • 3:2 blackjack payout — never play 6:5
  • 75%+ penetration — the single most important factor
  • Reasonable minimums — your min bet must be small enough for your bankroll
Strongly Preferred
  • S17 — reduces house edge 0.22% vs H17
  • DAS — double after split allowed
  • Late surrender — critical for Fab 4 deviations
  • Re-split aces — small but real edge

Las Vegas — Strip

The best shoe games are in high-limit salons. Most pit games hit soft 17 (H17). S17 games exist at $100+ minimums.

CasinoDecksRuleMin BetNotes
Aria6S17 DAS LS RSA$100High limit salon; 0.26% base edge
Bellagio2 / 6S17 DAS$100Double deck in high limit; strong penetration for known players
Cosmopolitan6S17 DAS LS RSA$100High limit; one of the best rule sets on the Strip
Fontainebleau2 / 6S17 DAS LS RSA$100–$300Newest major property; competitive rules
MGM Grand6S17 DAS LS RSA$100High limit; standard MGM rules
Mandalay Bay2 / 6S17 DAS LS RSA$100Double deck available; 0.19% base on 2D
Treasure Island2 / 6S17 DAS LS RSA$50Lowest S17 minimum on the Strip
Venetian / Palazzo6S17 DAS LS RSA$100Good penetration reported in high limit
Wynn / Encore6S17 DAS LS RSA$100Premium room; generally counter-friendly for lower spreads
Paris / Bally’s6S17 DAS LS RSA$100Caesars property; high limit only

Las Vegas — Downtown & Off-Strip

Better rules at lower minimums. Downtown and locals casinos often have the most beatable games.

CasinoDecksRuleMin BetNotes
El Cortez1 / 2S17 3:2$5–$25Famous single-deck 3:2; the best low-stakes game in Vegas
Red Rock6S17 DAS LS RSA$25–$50Station Casinos; locals favorite
Green Valley Ranch6S17 DAS LS$15–$25Station Casinos; Henderson
South Point2 / 6H17 DAS$10–$25Good penetration; H17 on shoe games
The Orleans6H17 DAS LS$10–$15Boyd Gaming; low minimums, decent pen
Palace Station6H17 DAS$10Low stakes; typical locals game

Atlantic City & East Coast (US)

New Jersey law mandates S17 and prohibits mid-shoe entry bans, making AC uniquely counter-friendly. You cannot be barred from playing — only from the casino property entirely.

CasinoDecksRuleMin BetNotes
Borgata8S17 DAS LS RSA$25–$50Best AC game; NJ law = can’t be barred from play
Hard Rock AC8S17 DAS LS$15–$25Competitive rules; good pen on weeknights
Ocean Casino8S17 DAS LS$15–$25Boardwalk; varies by pit
Mohegan Sun (CT)6 / 8S17 DAS LS$15–$25Tribal casino; good 6D games in high limit
Foxwoods (CT)6 / 8H17 DAS$10–$25Huge floor; penetration varies widely by dealer

United Kingdom & Europe

European rules: no hole card (ENHC), typically S17, and often no surrender. ENHC costs ~0.11% vs American rules. London clubs have the best games.

CasinoLocationDecksRuleNotes
HippodromeLondon6 / 8S17 DAS ENHCLargest UK casino; good pen; counters tolerated longer than Vegas
Empire CasinoLondon6S17 DAS ENHCLeicester Square; reasonable minimums
Grosvenor CasinosUK-wide6S17 DAS ENHCChain with 50+ locations; S17 standard; penetration varies
Holland CasinoNetherlands6S17 DAS ENHCState-run monopoly; good rules but low pen (~65%)
Casino de Monte-CarloMonaco6S17 ENHCHigh limits; historical; no DAS on some tables
Casino Gran MadridSpain6S17 DAS ENHCSpanish regulation allows counting; no barring

Asia-Pacific

Macau is the world’s largest gambling market but penetration is generally poor (50–65%). Australia and South Korea have better conditions.

CasinoLocationDecksRuleNotes
Crown MelbourneMelbourne, AUS6 / 8S17 DAS ENHCBest Australian game; good pen; counters get backed off eventually
The StarSydney, AUS6 / 8S17 DAS ENHCCompetitive rules; varies by pit
SkyCityAuckland, NZ6S17 DASOnly NZ casino with blackjack; decent rules
Venetian MacaoMacau8S17 DAS ENHCMassive floor; 50–65% pen limits counting
Wynn MacauMacau8S17 DAS ENHCBetter pen in premium rooms; high minimums
Paradise CityIncheon, KOR6S17 DASNear Seoul; foreigners only; decent games
Marina Bay SandsSingapore6 / 8S17 DAS ENHC$25+ min; continuous shufflers on most tables; manual shoe in high limit

Canada & Caribbean

CasinoLocationDecksRuleNotes
Playground PokerMontreal, QC8S17 DAS LSStrong rules; Quebec regulation allows counting
Casino NiagaraNiagara Falls, ON8S17 DAS LSOLG-operated; S17 standard in Ontario
River Rock CasinoVancouver, BC6 / 8S17 DASBC casino; shoe games available
AtlantisBahamas6 / 8S17 DASTourist casino; rules vary by table

Key Abbreviations

S17Dealer stands on soft 17 (player-friendly)
H17Dealer hits soft 17 (adds ~0.22% house edge)
DASDouble after split allowed
LSLate surrender allowed
RSARe-split aces allowed
ENHCEuropean no hole card (costs ~0.11%)
3:2Blackjack pays 3 to 2 (always required)

Casino conditions change frequently. Always verify rules, minimums, and penetration on-site before committing bankroll. Penetration varies by shift, dealer, and pit boss. Information sourced from Wizard of Vegas blackjack survey and community reports.

Spanish 21 Card Counting Analysis

Monte Carlo-derived strategy and edge analysis from 180+ million simulated hands.

The Game

Spanish 21 uses a 48-card deck — all four 10-spot cards are removed per deck (J/Q/K remain). The missing 10s give the house ~1.89% extra edge, clawed back by player-friendly rules: player 21 always wins, double on any number of cards, double down rescue, re-split aces, and bonus payouts for multi-card 21s.

5-card 21Pays 3:2
6-card 21Pays 2:1
7+ card 21Pays 3:1
6-7-8 suitedPays 2:1 (spades 3:1)
7-7-7 suitedPays 2:1 (spades 3:1)

Best Counting System: Secret Monkey Count

The Ace is 2.5x more important than face cards in Spanish 21. Standard Hi-Lo fails because it weights them equally. The Secret Monkey Count (SMC) corrects this:

2, 3, 4, 5, 6+1
7, 8, 90
J, Q, K-1
Ace-2

Balanced (sums to 0 per deck). True Count = Running Count ÷ (remaining cards ÷ 48).

Monte Carlo Strategy (Derived from 30K trials per decision point)

Optimal play derived empirically — not copied from published charts.

Hard Totals (2-card, 6D H17)

2345678910A
8HHHHHHHHHH
9HHHDHHHHHH
10DDDDDDDHHH
11DDDDDDDDHH
12HHHHHHHHHH
13HHHHHHHHHSu
14HHHHSHHHHSu
15HSSSSHHHHSu
16SSSSSHHHSuSu
17SSSSSSSSSSu

Can You Beat Spanish 21? (10M hands per scenario)

ScenarioSMC EdgeHi-Lo Edge
6D S17 75% pen+0.21%-0.52%
6D S17 83% pen+0.26%-0.26%
6D H17 75% pen-0.31%-1.08%
6D H17 90% pen+0.10%-0.55%
8D H17 75% pen (Borgata)-0.47%-1.23%

Verdict: S17 games are beatable with SMC and a 1:12 spread. H17 requires 90%+ penetration. Standard Hi-Lo does NOT work for Spanish 21.

Borgata Atlantic City

8-deck, S17 (mandated by NJ regulation), no redoubling, $10 minimum, ~75% penetration. Base house edge: 0.40%. At a 5:1 spread the game is NOT beatable. At 10:1 you barely break even (+0.03%). You need 20:1+ spreads for meaningful profit, which requires a $50K+ bankroll.

Baccarat Advantage Play

Deep dive from 21+ million simulated hands. The main game is unbeatable — but the Dragon 7 side bet is provably exploitable.

Main Bets: Unbeatable

Confirmed across 35 million hands (500K shoes). No counting system flips the main bets positive at any count level.

Player bet-1.24% house edgeNever flips
Banker bet (5% commission)-1.06% house edgeNever flips
Tie bet (8:1)-14.4% house edgeNever flips

Even computer-perfect counting with $1,000 bets yields ~$3/hour at 98% penetration. Dead end.

Dragon 7 Side Bet: BEATABLE

In EZ Baccarat (commission-free), Dragon 7 pays 40:1 when Banker wins with a 3-card total of 7. Base probability: 2.25%. Base house edge: 7.41%.

Why counting works: When 8s and 9s leave the shoe, fewer naturals occur, forcing more 3-card draws. More draws = higher Dragon 7 probability.

The Count: Dead Simple

4, 5, 6, 7 dealt-1
8, 9 dealt+2
Everything else (A, 2, 3, 10, J, Q, K)0

True Count = Running Count ÷ decks remaining. Bet Dragon 7 when TC ≥ 3.

You can track this openly with pen and paper — casinos provide scorecards at baccarat tables.

Edge by True Count (300K shoes, 21M+ hands)

True CountPlayer EdgeBets/ShoeProfit/Shoe ($100)
TC ≤ 1-1.61%20.1-$32
TC ≥ 2+0.74%11.8$8.73
TC ≥ 3+3.80%7.0$26.47
TC ≥ 4+5.64%4.2$23.77
TC ≥ 5+8.78%2.5$22.00
TC ≥ 8+11.32%0.5$5.86
TC ≥ 10+21.06%0.2$3.47

The Money

StrategyBet at TC ≥ 3 (optimal)
Edge per bet+3.80%
Bets per shoe~7
At $100 side bets~$27/hour
At $25 side bets (typical max)~$7/hour
Bankroll needed ($100 bets)$10,000+ (100 units)

The constraint is bet maximums, not math. Most casinos cap Dragon 7 at $25-$100. At high-limit tables with $500 side bets: ~$135/hour.

Practical Advantages

  • Open tracking allowed — casinos provide scorecards and pens. Your count sheet looks like normal baccarat pattern tracking.
  • Simple count — only 6 card ranks matter (4-9). Ignore everything else.
  • Low heat — pit bosses focus on main bet whales, not side bet players.
  • Bet the main game on every hand (Banker $25) to look normal. Add the Dragon 7 side bet only when the count is favorable.

Trainer

Isolated drills that build the individual sub-skills counting requires — far more efficient than only playing full hands. Every drill tracks your accuracy. Drills use the rules and counting system you set on the Game Variant tab.

Correct 0 Wrong 0 Accuracy Streak 0

Basic Strategy Chart

The mathematically optimal play for every hand, generated live from the rules you set on the Game Variant tab. Change a rule there and this chart follows. Learn this cold before adding the count.

H Hit S Stand D Double P Split R Surrender

D = double if allowed, otherwise hit (or stand on soft hands). R = surrender if allowed, otherwise hit. Columns are the dealer’s upcard.

Hard Totals

Soft Totals (hand with an Ace counted as 11)

Pairs

Monte Carlo Simulator

Deal and play thousands of shoes using the rules, counting system and bet spread from the Game Variant tab. Every hand is played with correct basic strategy plus the Illustrious 18 / Fab 4 deviations, then aggregated into your real edge by true count. This is the distribution you’re actually playing for.

Blackjack Side Bets

The felt is covered in optional wagers with names like 21+3, Lucky Ladies, and Buster. Almost all of them exist for one reason: their house edge dwarfs the main game. Below is an honest breakdown of the major side bets, their typical edge ranges, and whether a counter can actually beat them. One critical caveat up front: side-bet house edges swing enormously with the number of decks and the specific pay table posted at the table. Treat every figure here as a published range (drawn from widely cited analysis such as the Wizard of Odds), not a fixed number for the game in front of you.

Insurance

Offered whenever the dealer shows an Ace. You bet up to half your original wager that the dealer has a ten in the hole for a natural. It pays 2:1.

The Flat-Play Verdict

For a player with no count information, insurance is a sucker bet. There are only 16 tens per 52 cards, so the dealer makes the blackjack less than a third of the time. The typical house edge is roughly ~7% off the top of a fresh shoe, varying with deck count.

Pays2:1
Basic strategyNever take it
Off-the-top edge~7% house (varies by decks)

The Exception That Matters

Insurance is the single most valuable card-counting deviation in the game. It is purely a bet on the ratio of tens to non-tens remaining — exactly what a Hi-Lo count tracks. When the deck is rich in tens, the dealer hits blackjack often enough to make insurance a positive-expectation wager.

Hi-Lo True CountInsurance Decision
Below +3Decline — still a losing bet
≥ +3Take insurance

This is the “Illustrious 18” index play number one, and it ties directly to the running/true count you track on the main game. See the Deviations tab for the full index list.

21+3

Combines your two cards with the dealer’s upcard to form a three-card poker hand. Payouts trigger on a flush, straight, three of a kind, or straight flush.

House Edge

Typically around ~3%, but this is one of the most pay-table-sensitive side bets in the casino — the edge can range from low single digits to well into double digits depending on the posted schedule and number of decks. Always read the paytable before assuming the “~3%” version is in front of you.

Countability

Largely not practically countable. The bet depends on suit and rank distributions across three specific cards, which requires a multi-parameter side count that delivers little edge for enormous effort. Skip it.

Typical house edge~3% (varies widely by pay table)
Practically countable?No

Perfect Pairs

Wins when your first two cards form a pair, with escalating payouts for a mixed pair, a colored pair (same color, different suit), and a perfect pair (identical suit).

House Edge

Typically in the ~2–11% range depending on the number of decks and the pay table — fewer decks generally mean a lower edge. As always, the exact figure depends on the specific schedule posted.

Countability

Generally not countable for practical play. Pairing probability shifts only slightly as cards deplete, and isolating it requires tracking exact rank/suit residuals. Not worth it.

Typical house edge~2–11% (decks & pay table)
Practically countable?No

Lucky Ladies

Pays when your first two cards total 20, with large bonuses for a matched 20, a suited 20, and the marquee hand — a pair of Queens of Hearts (often boosted further if the dealer also has a blackjack).

House Edge

One of the worst bets on the table for flat play. The typical house edge runs roughly ~17–25%, again varying by pay table and deck count. The eye-catching top jackpot masks a brutal long-run expectation.

Countability

It has been formally studied for counting — ten-rich decks raise the frequency of twenties — and dedicated multi-level side counts can theoretically flip it. In practice it is impractical for most players: the required count is complex, the bet caps are low, and the variance is savage. A specialist’s curiosity, not a working play.

Typical house edge~17–25%
Practically countable?Studied, but impractical

Buster Blackjack

Wins when the dealer busts, with the payout scaling up by how many cards the dealer used to do it — a bust on more cards pays more.

House Edge

Typically around ~6–8%, varying by deck count and pay table.

Countability

Countable in theory. When the remaining shoe is rich in high cards, the dealer breaks more often on stiff totals, and multi-card busts become more likely. The catch is that the edge swing is marginal — the bet only turns favorable at fairly extreme counts, and it overlaps awkwardly with your main-game betting decisions. Most counters leave it alone unless they have specifically modeled the local pay table.

Typical house edge~6–8%
Practically countable?In theory — marginal in practice

Royal Match

Pays when your first two cards are suited, with a much larger “Royal Match” payout for a suited King and Queen.

House Edge

Typically in the ~3–7% range, depending on the number of decks and the posted pay table.

Countability

Not practically countable. The outcome hinges on suit composition rather than the high/low ratio your main count tracks, so it offers no useful hook for a standard counting system.

Typical house edge~3–7%
Practically countable?No

The Bottom Line on Side Bets

For flat play, treat nearly every blackjack side bet as a sucker bet. Their house edges — ranging from a few percent to well over 20% — exist precisely because the flashy bonus payouts feel exciting while quietly draining your bankroll faster than the main game ever could.

The rare exceptions are count-dependent, not universal:

  • Insurance is the standout — the most valuable counting deviation in blackjack, profitable at a Hi-Lo true count of +3 or higher.
  • Specialty bets in adjacent games can also be beaten, most famously EZ Baccarat’s Dragon 7, which flips positive with a dead-simple count. See the Baccarat tab for the full breakdown.

One practical upside: side-bet counting tends to draw less heat than spreading your bets on the main game. A player who flat-bets the base game and only fires a side bet when the count is right looks far less like a card counter than someone ramping their main wager from one unit to twelve. The edges are thinner and the bet caps are lower — but the cover is better.

Guides & Strategy

Is Card Counting Legal?

Counting cards in your head is legal in the United States and the United Kingdom. You are using only mental skill and memory to track which cards have been dealt — no law prohibits thinking. The legal lines are crossed when a device enters the picture or when you ignore a private property owner’s instructions.

Devices are a crime

Using any device or technological aid to count or predict outcomes is illegal. In Nevada, this is codified at NRS 465.075, which makes it a felony to use a device to aid in projecting the outcome of a game, keeping track of cards played, or analyzing the probabilities of an event. Phones, hidden computers, and counting apps at a live table all fall under this. Counting unaided does not.

Casinos are private property

In most US states, a casino may refuse service to anyone for skilled play and may trespass you from the premises. The skill itself is not a crime; being asked to leave and refusing is.

  • New Jersey is the notable exception. In Uston v. Resorts International Hotel (1982), the New Jersey Supreme Court held that Atlantic City casinos cannot bar a patron simply for being a skilled counter. They may instead adjust the rules of play (e.g., shuffling more frequently).
  • United Kingdom: counting is legal, but casinos remain free to bar players they suspect of advantage play.

Not legal advice. This is general educational information. Laws vary by jurisdiction and change over time. If you have a specific legal concern, consult a qualified attorney in the relevant state or country.

Heat, Back-offs & Getting Barred

“Heat” is casino attention directed at a suspected advantage player. Surveillance and pit staff watch for bet spreads that track the count, deviation plays, and players who only sit down when the shoe is rich. The response escalates in predictable stages.

Signs you are being watched

  • A pit boss lingers and openly records your bets.
  • Shuffle-up: the dealer shuffles early the moment your bet jumps.
  • A sudden mid-shoe shuffle after a big wager.
  • Staff ask for your player’s card or photo ID without an obvious reason.

The escalation ladder

StageWhat happens
Shuffle-upEarly or mid-shoe shuffles to erase your edge
Back-off“We’d prefer you not play blackjack” — you can stay, just not at the tables
Barring / trespassFormally told to leave; future entry is trespassing
Back roomRare; detention for questioning, typically when a device or fraud is suspected

Facial recognition & shared databases

Larger properties run surveillance imagery against facial-recognition systems and may share identifications across casino networks and third-party databases. At a high level, a flag at one property can follow you to affiliated properties, which is why back-offs sometimes happen before you have placed a bet.

Practical conduct

  • When the heat comes, leave. The math is gone for that session.
  • Do not argue with the pit or surveillance — it changes nothing and gets you remembered.
  • Do not volunteer ID you are not legally required to show.

The MIT Blackjack Team & the History of Counting

Modern card counting begins with mathematician Edward O. Thorp. His 1962 book Beat the Dealer proved that tracking the ratio of high to low cards gives the player a measurable edge, and it forced casinos to change how blackjack was dealt.

The lineage

  • Edward Thorp — founded the field with the first practical counting systems.
  • Ken Uston — popularized team play and pushed the legal fight that produced the New Jersey ruling bearing his name.
  • Stanford Wong — refined counting strategy and lent his name to “Wonging,” entering a shoe only when the count is favorable.
  • Al Francesco — credited with developing the Big Player team concept.

The Big Player technique

In the Big Player approach, spotters sit at separate tables flat-betting the minimum while keeping the count. When a table turns hot, a spotter signals the Big Player, who arrives, bets large only during the favorable count, and leaves. To the pit, the Big Player looks like a high-rolling tourist on a streak rather than a counter, because his bets never correlate with a count he was visibly keeping.

The MIT team

From the 1980s into the 1990s, teams run by students and associates connected to MIT in Cambridge, Massachusetts used this method to win millions across US casinos. Their story was popularized by Ben Mezrich’s book Bringing Down the House and the 2008 film 21, both of which dramatize events and should not be read as documentary fact.

Comps, Tipping & Casino Etiquette

“Comps” are complimentary goods and services — rooms, meals, show tickets — the casino gives back as a fraction of your theoretical loss. To get them you must be rated: a pit clerk logs your average bet and hours played, and a casino host manages the relationship for bigger players.

The counter’s dilemma

Rating is a double-edged sword. It earns comps, but it also creates a detailed record of your play — bet size, spread, and session length — which is precisely the data surveillance uses to identify advantage players. More attention is the trade-off for a free buffet.

Tipping (toking)

Tipping the dealer can serve as cover, making you look like a recreational player. It is also a direct cost to your edge: every chip toked is expected value handed back. Treat tips as a deliberate cover expense, not an obligation, and budget for them.

Table etiquette

  • Use the standard hand signals — tap for a hit, wave for a stand — so the cameras have a clear record.
  • In a shoe game, never touch the cards; they are dealt face up and stay on the table.
  • Buy in with cash placed on the felt (not handed to the dealer); bet with chips.
  • Keep your hands away from your bet once the deal begins.

Bankroll for the trip

Bring enough that normal variance cannot wipe out your session, and segregate gambling money from travel and living costs. Counting only yields a small long-run edge, so short trips are dominated by swings — plan for losing sessions as a routine outcome, not a failure.

Glossary

Plain-English definitions of the blackjack and card-counting terms used throughout Card Counter Pro. Entries are grouped alphabetically; bolded words point to related terms you can look up in the same list.

A

Ace-Five Count
The simplest viable counting system: add 1 for every five seen and subtract 1 for every ace. Easy to learn but with low power compared to a full Balanced count like Hi-Lo.
Advantage Play
Any legal technique — card counting, Shuffle tracking, hole-carding — that shifts the long-run Edge from the house to the player using skill rather than cheating.

B

Back-counting (Wonging)
Watching a table without playing and entering only when the True Count turns favorable, then leaving when it drops. Named after author Stanford Wong; see Wonging.
Backoff
When casino staff ask a suspected counter to stop playing blackjack (or to leave), usually politely. A common form of Heat short of a formal barring.
Balanced count
A counting system whose Tag values sum to zero across a full deck, so the Running Count returns to 0 after all cards are dealt. Hi-Lo is the classic example; conversion to a True Count is required.
Bankroll
The total money set aside exclusively for play. Its size relative to your Bet Spread determines your Risk of Ruin and the Unit you can safely bet.
Basic Strategy
The mathematically optimal way to play every hand against each dealer up-card, assuming no knowledge of unseen cards. It minimizes the house Edge before any counting is applied.
Bet Spread / Bet Ramp
The ratio between your smallest and largest bets, and the schedule mapping each True Count to a bet size. A wider spread raises both EV and variance — and Heat.
Big Player
In Team play, the member who is signaled into a hot shoe by back-counting teammates and makes the large bets, appearing to outsiders as a lucky high-roller.
Burn card
The card the dealer removes from play, typically face-down to the Discard tray, immediately after the shuffle and Cut card placement.
Bust
A hand whose total exceeds 21, which loses immediately regardless of the dealer’s hand.

C

Camouflage
Deliberate cover behavior — chatting, varied bets, occasional non-optimal plays, drinking — used to disguise Advantage Play and reduce Heat.
Cut card
The colored plastic card the player inserts to cut the deck after the shuffle; its placement sets the Penetration by marking where the dealer reshuffles.

D

DAS (Double After Split)
A rule allowing a Double down on a hand created by splitting a pair. It favors the player and slightly lowers the house Edge.
Deviation / Index play
A departure from Basic Strategy triggered when the True Count reaches a specific index number. The most valuable deviations are collected in the Illustrious 18 and Fab 4.
Discard tray
The holder where played cards accumulate. Its fill level lets a counter estimate decks remaining for the Running Count-to-True Count conversion.
Double down
Doubling your bet after the first two cards in exchange for exactly one more card.

E

Edge
The percentage advantage one side holds over the long run. A skilled counter aims to convert the house’s edge into a small player edge, often around half a percent.
ENHC (European No Hole Card)
A rule variant in which the dealer takes no Hole card until players finish acting, so splits and doubles can be lost to a dealer blackjack. It slightly raises the house Edge.
EV (Expected Value)
The average result of a bet or decision over the long run, the core measure used to evaluate every counting and Deviation choice.

F

Fab 4
The four most profitable Surrender Deviation indices, identified by Don Schlesinger as the surrender companions to the Illustrious 18.
Flat betting
Wagering the same amount every hand regardless of the count. Used as Camouflage or to play near break-even without drawing Heat; see Spread.

H

Hard hand
A hand with no ace, or one in which the ace must count as 1 to avoid a Bust, so its total is fixed. Contrast with a Soft hand.
Heat
Scrutiny from casino staff toward a suspected Advantage Player, ranging from closer watching to a Backoff or barring.
Hi-Lo
The most popular Balanced, Level-1 count: 2–6 count as +1, 7–9 as 0, and 10–A as −1. It offers a strong balance of simplicity and power.
Hole card
The dealer’s face-down card. Glimpsing it (hole-carding) is a powerful but situational form of Advantage Play.
H17 / S17
Whether the dealer Hits soft 17 (H17) or Stands on soft 17 (S17). S17 is better for the player; H17 adds roughly 0.2% to the house Edge.

I

Illustrious 18
Don Schlesinger’s set of the 18 most valuable Deviation plays, capturing most of the gain available from departing from Basic Strategy.
Insurance
A side bet, offered when the dealer shows an ace, that the dealer has blackjack. Generally a poor bet under Basic Strategy but profitable for counters at a sufficiently high True Count.

K

KO (Knock-Out)
A popular Unbalanced count whose Tag values do not sum to zero, removing the need to convert the Running Count into a True Count.
Kelly Criterion
A bet-sizing formula that wagers a fraction of Bankroll proportional to your Edge, maximizing long-run growth. Many counters use a fractional (e.g. half-) Kelly to lower Risk of Ruin.

L

Level (of a count)
The largest absolute Tag value a system assigns. Hi-Lo is Level 1 (±1); higher-level systems use values up to ±2 or ±3 for more power at the cost of difficulty.

N

N0 (N-zero)
The number of hands (or rounds) of play, on average, needed for accumulated EV to equal one Standard Deviation — a measure of how long until your edge reliably overcomes variance.

P

Penetration
The fraction of the Shoe or deck dealt before the reshuffle, set by the Cut card. Deeper penetration is the single biggest factor in a counter’s profitability.
Pitch game vs shoe game
A pitch game uses one or two hand-held decks dealt face-down; a shoe game deals four to eight decks face-up from a Shoe. Pitch games allow deeper estimation but offer fewer hands per hour.
Playing Efficiency (PE)
How well a count’s Tag values predict correct Deviation decisions. It matters most in deeply dealt single- and double-deck games.
Betting Correlation (BC)
How well a count’s tags predict the player’s changing Edge, and thus how accurately it guides your Bet Spread. Hi-Lo has a near-perfect BC.
Insurance Correlation (IC)
How well a count predicts the profitability of the Insurance bet specifically. High IC means the count is a reliable insurance signal.
Push
A tie between player and dealer, in which the original bet is returned with no win or loss.

R

Risk of Ruin (RoR)
The probability of losing your entire Bankroll before reaching a target, given your Edge, Bet Spread, and bankroll size. Sound Bankroll management keeps it low.
Running Count (RC)
The raw running total obtained by adding each card’s Tag as it appears. In a Balanced count it must be divided by decks remaining to get the True Count.

S

Shoe
The box that holds and dispenses multiple decks (typically four to eight) in a shoe game.
Shuffle tracking
An advanced technique of following groups of cards through the shuffle to predict where high-value clumps land, letting a player bet and cut accordingly.
Soft hand
A hand containing an ace counted as 11 without busting, so it can be hit safely. Contrast with a Hard hand.
Spread
Shorthand for the Bet Spread — the range between minimum and maximum bets, often quoted as a ratio such as 1–8.
Standard Deviation (SD)
A measure of how widely actual results swing around their EV. It quantifies the short-term variance a counter must endure to realize a long-run Edge.
Surrender (late/early)
Forfeiting half your bet to abandon a poor hand. Late surrender is allowed only after the dealer checks for blackjack; the rarer early surrender, allowed before the check, is more valuable to the player.

T

Tag (count value)
The point value a system assigns to a card rank (e.g. +1, 0, or −1 in Hi-Lo), summed into the Running Count.
Team play
Coordinated Advantage Play by multiple people who pool a Bankroll, often using back-counting spotters who signal a Big Player into hot shoes.
Toke (tip)
A tip given to the dealer, sometimes as a bet placed on the dealer’s behalf. Counters factor tokes into EV and use them as Camouflage.
True Count (TC)
The Running Count divided by the number of decks remaining, giving the per-deck concentration that drives bet sizing and Deviation decisions in a Balanced count.

U

Unbalanced count
A system whose Tag values do not sum to zero over a full deck, allowing play directly off the Running Count with no True Count conversion. KO is the best-known example.
Unit
The base betting amount from which all wagers are scaled. Bets and Bet Spreads are expressed in units to keep sizing tied to Bankroll.

W

Wonging
Synonym for Back-counting: observing a table and only sitting down when the True Count is favorable. Named for blackjack author Stanford Wong.

Session Report

Enter your email to generate and receive your full session report with play analysis.