Baccarat Strategy — The Only Advice That Actually Works

Baccarat is one of the few casino table games where the best strategy can be stated in a single sentence: bet Banker every hand, set a session budget, and leave when you have reached it. Everything else — the betting systems, the scoreboard patterns, the instinct plays — either makes no difference or makes things worse. This guide explains why, with numbers.

The house edge on each bet

The three baccarat bets do not carry the same house edge. The differences are significant enough to matter over a session.

Bet	Payout	House edge
Banker	0.95:1 (5% commission)	1.06%
Player	1:1	1.24%
Tie	8:1	14.36%

These figures assume a standard 8-deck shoe and a 5% Banker commission. They represent the percentage of each bet you expect to lose in the long run. On a £100 Banker bet, the expected loss per hand is £1.06. On a £100 Tie bet, the expected loss is £14.36.

The Banker bet has the lowest house edge of any bet in baccarat, and at 1.06%, it compares very favourably to many other casino games. For context, a single-zero roulette outside bet carries a 2.70% house edge — more than twice the Banker edge. See Roulette Strategy for a full comparison of roulette bets.

Why the Banker bet wins more often

The Banker hand wins approximately 45.86% of the time, the Player hand wins approximately 44.62% of the time, and the remaining 9.52% of hands are Ties. Strip out the Ties (since neither bet wins on a Tie unless you bet Tie), and the Banker wins roughly 50.68% of non-Tie hands versus the Player’s 49.32%.

The Banker’s slight statistical advantage arises from the draw rules. The Banker’s draw decision is informed by the Player’s third card, while the Player must draw or stand blind. This informational asymmetry gives the Banker hand a structural edge. To offset this, casinos charge a 5% commission on winning Banker bets. Even after paying that commission, the Banker bet remains the best wager on the table.

Why the Tie bet is a trap

At 8:1, the Tie bet looks attractive. An 8:1 payout on a bet that wins nearly 10% of the time sounds like it should be profitable, or at least close. The numbers say otherwise.

For the Tie bet to be break-even at 8:1 payouts, it would need to win approximately 11.1% of the time. It wins approximately 9.52% of the time. The gap between required and actual win rate produces that 14.36% house edge. Some tables pay 9:1 on Ties, which reduces the edge to around 4.84% — still far worse than the Banker or Player bets, but less catastrophic than the 8:1 version.

The Tie bet should be avoided by any player who is trying to make their bankroll last. It is not a strategic option; it is the game’s most costly bet dressed up with a large number.

Does card counting work in baccarat?

Card counting works in blackjack because drawing certain cards changes the remaining composition of the shoe in ways that shift the house edge from the player to the dealer — or reverse it. Blackjack card counters track tens and Aces because a ten-rich shoe dramatically increases the player’s advantage when doubling down, splitting, and drawing to naturals.

In baccarat, card counting produces negligible advantage even in theory, and is completely impractical in practice.

Academic analysis has shown that baccarat shoes do contain states where the Player or Banker bet briefly turns player-favourable — but those states are rare, the edge is tiny (often fractions of a percent), and they cannot be identified without tracking dozens of card categories simultaneously. Researchers who have studied this conclude that the gain from perfect baccarat card counting is so small that even a player tracking the count perfectly would need to play for years to achieve a measurable return.

In practice, casinos shuffle frequently, use 6–8 deck shoes, and cut cards deep into the shoe — all of which prevent meaningful count-based edge identification. Card counting in baccarat is, for practical purposes, a dead end.

Betting systems: why they don’t change the odds

The casino floor is littered with baccarat players using betting systems. The most common:

Martingale: Double your bet after every loss, reset to base after a win. The logic is that a win will eventually recover all losses plus a unit profit. The problem is that a losing streak long enough to exhaust your bankroll — or hit the table maximum — is not just possible, it is guaranteed over enough sessions. Martingale does not change the expected value of each bet; it trades frequent small wins for occasional catastrophic losses.

Paroli (Reverse Martingale): Double your bet after each win for up to three wins in a row, then reset. This system is psychologically appealing because you are pressing winning runs rather than chasing losses. It still cannot change the house edge. Each hand is an independent event, and winning three in a row does not make a fourth win more likely.

Fibonacci: Bet sizes follow the Fibonacci sequence (1, 1, 2, 3, 5, 8, 13…). Move up after a loss, back two steps after a win. Slightly less aggressive than Martingale but subject to the same fundamental problem: it cannot change the mathematical outcome over a large sample.

Flat betting — wagering the same amount on every hand — is actually the safest approach in a negative expected value game. It minimises variance and stretches your bankroll furthest. If you must use a system, flat betting is the one that does the least damage.

Pattern tracking and “roads”

Baccarat tables, both live and online, display scoreboards showing the recent sequence of Player and Banker wins. Some casinos provide multiple derived “road” displays — the Big Road, the Bead Plate, the Big Eye Boy, the Small Road, the Cockroach Pig. Players spend considerable effort looking for patterns in these displays: streaks, chops, “the shoe is running Banker.”

This is pure superstition. Each baccarat hand is an independent trial. The probability of the next hand being a Banker win is approximately the same regardless of whether the last ten hands were all Banker wins or alternating perfectly. Cards have no memory. The shoe does not have a “character.” Past results tell you nothing about the next hand.

The scoreboards exist because players enjoy them and they increase engagement — not because they contain predictive information. A player who ignores the scoreboard and bets Banker every hand has exactly the same expected return as a player who studies the scoreboard for twenty minutes before each bet.

Practical strategy summary

1. Always bet Banker. The 1.06% house edge is as low as baccarat gets.
2. Avoid the Tie bet. 14.36% edge is among the worst available in any casino game.
3. Ignore the scoreboards for decision-making. They are entertainment, not information.
4. Set a session budget before you start. Decide the maximum you are prepared to lose and stop when you reach it.
5. Flat bet. Consistent bet sizing makes your money last longer than any progression system.

Responsible play

Baccarat moves fast — particularly mini-baccarat, which can deal 150 or more hands per hour. Even a 1.06% house edge accumulates quickly at that pace. A player betting £20 per hand at 150 hands per hour faces an expected hourly loss of £31.80 (£20 × 150 × 0.0106). Knowing this number before you sit down sets realistic expectations.

If gambling is causing financial stress, taking time away from other activities you value, or beginning to feel obligatory rather than entertaining, most casinos offer deposit limits, time limits, and self-exclusion options. Use them proactively, not reactively.

---

For a full explanation of the rules, including the Banker draw table and how naturals work, see Baccarat Rules. For background on where the game came from, see Baccarat History. Play baccarat online for free to apply these principles in practice.