Baccarat Advantage Play

Why Baccarat Is Hard to Beat

The structure of baccarat creates a uniquely difficult advantage-play target. Three properties combine to neutralise the techniques that work elsewhere. First, the game is symmetric: both Player and Banker hands follow fixed drawing rules from our rules guide, and no player decision interacts with deck composition to produce edges. Second, the probabilities are near-symmetric (Banker 45.84%, Player 44.62%, Tie 9.54%), and card-removal effects on these probabilities are tiny, an order of magnitude smaller than the comparable effects in blackjack. Third, casinos use 6-deck or 8-deck shoes with cut cards that prevent deep penetration; the last 30 to 60 cards of every shoe are never dealt, so late-shoe counting advantages cannot be exploited.

The combination produces a game where advantage play is mathematically near-impossible through legitimate techniques. The exceptions, covered below, require either microscopic edges that variance overwhelms or techniques that have been legally classified as cheating.

Card Counting: The Honest 0.01%

Edward Thorp (the mathematician who beat blackjack in the 1960s) and Peter Griffin (the academic behind The Theory of Blackjack) both analysed baccarat for advantage-play potential in the 1960s and 70s. Their conclusion: card counting in baccarat is mathematically possible but practically useless.

The mechanics: certain cards favour the Banker bet (removing 4s, 5s, 6s, 7s and 8s shifts probabilities marginally toward Banker), others favour the Player bet (removing 1s, 2s, 3s shifts toward Player). A perfect counter could identify the rare shoe configurations where the count crossed the threshold for a positive expected bet.

The numerical result: under perfect play, the theoretical edge gained by card counting baccarat is approximately 0.01% per hand. Compare to blackjack, where good counters achieve 0.5% to 1.5% edges. The baccarat figure is fifty to a hundred times smaller.

Practical implications:

• To realise a meaningful profit from 0.01% edge, you would need to wager enormous amounts. RM1,000,000 of wagering at 0.01% edge produces RM100 expected profit, swamped by normal variance.
• The positive-edge configurations are rare. Most of the time the count favours the house, requiring the counter to sit out or bet minimum.
• Casino staff notice players who watch cards and adjust bet sizes, regardless of the actual technique. The attention attracted is disproportionate to the realisable edge.

The popular advice “do not count cards in baccarat” is correct, but for an unusual reason. It is not that counting is impossible; it is that the edge gained is so small that it is not worth the effort. Card counting in baccarat is intellectually interesting but practically a waste of attention.

Edge Sorting Baccarat: The One Technique That Worked

Edge sorting is the one advantage-play technique that has produced documented multi-million-dollar wins in baccarat. The technique exploits manufacturing imperfections in playing cards: subtle asymmetries in the patterns printed on the card backs, where one long edge has slightly different markings than the other.

The technique is the focus of the most famous baccarat case of the 21st century, the Phil Ivey edge sorting baccarat lawsuits at Crockfords (London) and Borgata (Atlantic City). The mechanics involved:

1. Identify the asymmetric pattern. The Gemaco purple Borgata-branded playing cards used in both casinos had geometric patterns where one long edge was slightly different from the other. The difference was visually imperceptible at normal viewing distance but visible to a trained observer focused on the pattern.
2. Request favourable conditions. Ivey and his partner Cheung Yin “Kelly” Sun requested specific table conditions: purple Gemaco cards (with the exploitable pattern), an automatic shuffler (which preserves card orientation through shuffles), the same 8-deck shoe used throughout multiple sessions, a Mandarin-speaking dealer, and the right to play multiple sessions.
3. Manipulate card orientation during play. Sun, posing as a superstitious player, asked the dealer to rotate certain cards 180 degrees before dealing. The “lucky” cards were the high-value cards (7s, 8s, 9s); the rotated orientation placed their identifying edge marking in a different position from the low cards.
4. Exploit the sorted shoe in subsequent sessions. The automatic shuffler preserved the orientation of cards through the shuffle. After multiple sessions of sorting, the deck was substantially organised by value: high cards faced one direction, low cards faced the other.
5. Bet aggressively on favourable hands. Once the deck was sorted, Ivey could identify the value (high or low) of cards before they were dealt by looking at their orientation in the shoe. This gave him approximately a 6-7% edge against the house, enormous by gambling standards.

Across multiple sessions at Crockfords in August 2012, Ivey won £7.7 million. Across four sessions at Borgata in 2012, Ivey and Sun won approximately $9.6 million. Total winnings across both casinos: approximately $20 million.

The Phil Ivey Court Cases

Both casinos refused to honour the winnings once they identified the technique. The subsequent legal battles established the modern legal status of edge sorting.

Crockfords (UK). Crockfords (owned by Genting Casinos UK) refused to pay Ivey his £7.7 million winnings. Ivey sued in 2014. The initial court ruled against Ivey, finding that his actions constituted cheating under the UK Gambling Act 2005. Ivey appealed, ultimately reaching the UK Supreme Court.

The Supreme Court delivered its judgment on 25 October 2017, in the landmark case Ivey v Genting Casinos UK Ltd. All five judges rejected Ivey’s appeal. Lord Hughes wrote that Ivey’s actions constituted “a carefully planned and executed sting” and that the edge sorters had “taken positive steps to fix the deck.” The decision established UK legal precedent that edge sorting, regardless of intent or whether the player physically touched the cards, constitutes cheating. The £7.7 million stayed with Crockfords.

Borgata (US). Borgata sued Ivey and Sun in 2014. US District Court Judge Noel Hillman ruled in 2016. The ruling drew a distinction that mattered: Ivey was found NOT guilty of fraud (he had not deceived the casino about who he was or what game he was playing) but WAS guilty of breach of contract (the implicit terms of legal casino gambling under New Jersey’s Casino Control Act). Ivey was ordered to return $10.1 million.

The two rulings established a clear consensus: regardless of fraud findings, edge sorting violates the implicit terms of casino gambling. Casinos may legitimately refuse to pay winnings obtained through the technique, and may recover such winnings if already paid. The cases also resulted in litigation against Gemaco (the card manufacturer), which Borgata sued for producing cards with the exploitable manufacturing flaw.

Why Edge Sorting Is Now Effectively Impossible

The Ivey cases triggered industry-wide responses that have closed the edge sorting loophole. Card manufacturers (Gemaco, Cartamundi, USPC) implemented tighter quality control to eliminate the asymmetric back patterns. Most casinos now rotate cards 180 degrees randomly during shuffling, undoing any sorting work by the next shuffle cycle. Casinos refuse the kinds of player requests Ivey made (specific card brands, multi-session continuity, Mandarin-speaking dealers paired with specific decks) and flag suspicious request patterns automatically. High-stakes baccarat tables are continuously monitored for behaviours suggesting sorting attempts. And the legal precedent from the Ivey cases means that even if a player successfully edge-sorted in 2026, casinos can refuse to pay and recover winnings already paid.

The Ivey case was the high-water mark of advantage-play baccarat. It is unlikely to be repeated; the manufacturing flaws have been corrected, the operational countermeasures are in place, and the legal precedent is settled in casinos’ favour.

Dragon 7 Advantage Play

One other advantage-play technique deserves mention: card counting the Dragon 7 side bet in EZ Baccarat.

The Dragon 7 bet pays 40:1 when Banker wins with a three-card total of 7. The bet has a standard house edge of 7.61% but, unlike the main bets, is sensitive to deck composition: hands rich in low cards make a Banker 3-card 7 more likely.

Gambling mathematician Eliot Jacobson published an analysis demonstrating that the Dragon 7 bet can be card-counted, with theoretical edges of approximately 7-8% over the house when the count is favourable. Unlike main-bet baccarat counting (where 0.01% is mathematically insufficient), the Dragon 7 edge is large enough to be meaningful in principle.

Practical limitations:

• The favourable counts are rare. Most of the time the count is unfavourable and the player must sit out or place minimum bets.
• The 6-8 deck shoe with cut cards means deep penetration is impossible.
• Casinos that offer EZ Baccarat are aware of the technique and monitor for it. Counters are often back-rolled to flat-bet tables or asked to leave.
• The variance is enormous. The 40:1 payout structure means individual hands swing wildly; realising the theoretical edge requires thousands of hands of disciplined play.

The Dragon 7 advantage play is the closest thing to a legitimate baccarat advantage technique in 2026. It is also impractical for casual players, narrowly available (only at EZ Baccarat tables), and subject to the same casino countermeasures applied to advantage players generally. See our side bets guide and variants guide for the Dragon 7 mechanics.

Hole Carding and Shuffle Tracking

Two other advantage-play techniques deserve brief mention for completeness, with the caveat that neither is meaningfully applicable to modern baccarat.

Hole carding. The technique of glimpsing the dealer’s face-down card before its scheduled reveal. Effective in blackjack where the dealer’s hole card matters for strategy. Almost entirely inapplicable to baccarat, where all cards are dealt face up after the initial deal and there is no decision-bearing hole card to exploit.

Shuffle tracking. Tracking groups of cards through the shuffle to predict their position in the upcoming shoe. Theoretically applicable to baccarat, but baccarat’s symmetric game structure (no player decisions, fixed drawing rules) means even perfect shuffle tracking provides edges similar to general counting: approximately 0.01% per hand at best. The technique requires extensive practice for no meaningful return. Modern automatic shufflers also eliminate the tracking opportunity by randomising card positions completely.

Neither technique is worth pursuing for baccarat. The combination of small theoretical edges, casino countermeasures, and practical difficulty makes them strictly inferior to alternative uses of the player’s time.

The Honest Verdict

For Malaysian players considering edge sorting baccarat or any other advantage-play approach in 2026, the honest answer: do not. The math is not there.

Card counting gains 0.01% per hand, mathematically insufficient. Edge sorting has been closed by manufacturing improvements, operational countermeasures, and legal precedent. Dragon 7 counting has a meaningful theoretical edge but practical execution challenges that make it inaccessible to casual players. Hole carding and shuffle tracking do not work in baccarat. The James Bond strategy and other progressive systems (covered in our betting systems guide) do not change expected value.

Baccarat is a -1.06% expected value game on the Banker bet, and that figure is not going to improve through any technique available to recreational players. The honest approach is to accept the house edge as the cost of playing, manage your bankroll as a fixed fraction (see our strategy guide), and walk away on plan. This is realism, not pessimism, and knowing it protects you from the system-sellers, pattern-readers, and wishful-thinking traps that consume more player money than the house edge itself.

Frequently Asked Questions