The Only Three Bets That Matter
Baccarat looks complicated because of the drawing rules. It isn't. You are not making decisions โ the third-card rules are fixed and automatic, applied identically at every table on earth. Your only decision is which of three bets to place, and then how much.
Two hands are dealt: Player and Banker. Neither belongs to you. You are betting on which one lands closer to 9. The Banker hand wins slightly more often because it draws its third card after seeing what the Player did โ a small informational advantage baked into the rules.
| Outcome | Probability | Pays | House edge |
|---|---|---|---|
| Banker | 45.86% | 1:1 minus 5% | 1.06% |
| Player | 44.62% | 1:1 | 1.24% |
| Tie | 9.52% | 8:1 | 14.36% |
Why the 5% Banker Commission Still Wins
New players see the 5% commission on Banker wins and assume Player must be the better bet. It isn't, and the gap is easy to prove.
The commission is real, but Banker wins often enough to absorb it and still come out 0.18 percentage points ahead. Over 1,000 hands at $50, that difference is about $90 โ small per hand, meaningful over a year.
House Edge on Every Baccarat Bet
Casinos have added side bets to baccarat precisely because the main game is so cheap. Every one of them is worse than the base bets, most by a factor of five or more.
| Bet | Typical payout | House edge | Verdict |
|---|---|---|---|
| Banker | 0.95:1 | 1.06% | Play this |
| Player | 1:1 | 1.24% | Acceptable |
| Tie (8:1) | 8:1 | 14.36% | Never |
| Tie (9:1) | 9:1 | 4.84% | Still no |
| Player Pair | 11:1 | 10.36% | Never |
| Banker Pair | 11:1 | 10.36% | Never |
| Perfect Pair | 25:1 | 17.07% | Never |
| Either Pair | 5:1 | 14.54% | Never |
| Big / Small | 0.54:1 / 1.5:1 | 4.35% / 5.27% | Avoid |
The Tie bet is a trap by design. 8:1 sounds generous until you notice ties happen once every 10.5 hands, not once every 9. That gap is the casino's 14.36% โ worse than most slot machines and roughly 13ร the cost of simply betting Banker.
The Roadmap Myth: Patterns Don't Exist
Every baccarat table displays a "roadmap" โ big road, bead plate, big eye boy, cockroach pig. Players study these grids looking for streaks, chops and dragon tails, then bet accordingly.
The roadmaps are real records of what happened. They contain zero information about what happens next. Each hand is dealt from a freshly shuffled 8-deck shoe with essentially fixed probabilities; the cards have no memory of the last twelve results. Banker is 45.86% to win the next hand whether the last six results were BBBBBB or PPPPPP.
Why Card Counting Fails in Baccarat
Card counting works in blackjack because removed cards meaningfully shift your payoff โ a 10-rich deck means more blackjacks at 3:2 for you. Baccarat has no equivalent lever.
The published research is blunt about it. Even under perfect play with a full 8-deck count, the advantage that occasionally appears on the Player bet is roughly 0.7% of hands, worth a fraction of a percent, and requires enormous bet spreads to exploit. The Tie bet becomes theoretically countable near the end of a shoe โ but casinos cut the shoe long before that point.
- No player decisions โ you cannot act on information even if you have it
- No bonus payout equivalent to a 3:2 blackjack to swing EV
- Deep cut cards remove the end-of-shoe windows entirely
- Continuous shufflers on most modern tables end the discussion
Punto Banco, EZ, Speed and No-Commission
Hands per hour is the real variable. A 1.06% edge at 40 hands/hour on $50 costs about $21/hour. The same edge at 130 hands/hour costs $69/hour. Table speed matters more than almost any other choice you make.
Bankroll and Bet Sizing for Baccarat
Baccarat's low edge is paired with low volatility โ Banker/Player is close to a coin flip. That combination means your bankroll depletes slowly and predictably, which is exactly why flat betting works well here and progressions do not.
| Bankroll | Flat bet | Expected session length |
|---|---|---|
| $200 | $5 | ~3 hours |
| $500 | $10 | ~4 hours |
| $1,000 | $25 | ~3 hours |
| $2,500 | $50 | ~4 hours |
The rule of thumb: bet 1โ2% of your session bankroll per hand. At 40 units of bankroll you have enough cushion to survive normal swings; below 20 units, a routine losing streak ends your session.
Martingale is especially dangerous in baccarat because the near-coin-flip odds make it feel safe. It isn't. Eight consecutive Banker losses isn't rare โ it happens roughly once every 400 hands โ and by then a $5 start has grown to a $1,280 bet, most likely above the table maximum.
The 30-Second Strategy
Played this way, baccarat is genuinely one of the fairest games in any casino โ you'll pay about $1 per $100 wagered. Played the way most people play it, chasing ties and patterns, the effective cost is closer to $8. Same table, same dealer, eight times the price.
Play with the math on your side
SlotDrop's free tools show you the real house edge, EV and variance behind every bet โ before you place it.
Open the Black Box โReality check: no strategy on this page turns a negative-expectation game into a positive one. Strategy reduces how much the house takes and how fast your bankroll disappears โ it does not guarantee profit. Only ever bet money you can afford to lose entirely.