Baccarat Odds: Banker, Player, and the Tie Trap

Q: What is the house edge in baccarat?

It depends on the bet. The Banker bet has a house edge of about 1.06%, the Player bet about 1.24%, and the Tie bet about 14.4%. Banker and Player are two of the lowest-edge bets in the casino; the Tie is one of the worst.

Q: Is the Banker or Player bet better in baccarat?

Banker. Even after the 5% commission the casino charges on Banker wins, its house edge of about 1.06% beats the Player bet's 1.24%, because the Banker hand wins slightly more than half of all decided hands.

Q: Why is the baccarat Tie bet a bad bet?

The Tie pays 8 to 1, but a tie happens only about once every 10.5 hands, so the true odds are roughly 9.5 to 1. Paying 8 when the fair price is 9.5 is what creates the roughly 14.4% house edge, more than ten times the Banker bet's.

Q: Is baccarat beatable? Can you count cards?

No, not in any practical sense. Unlike blackjack, baccarat's card composition barely shifts the edge, so card counting yields a negligible, impractical advantage. There is no skill, memory, or trend-reading system that changes the outcome. It is a low-edge game, but the house still wins over time.

By Alex Thornton · author of the MIT-licensed teachersbettextbook library · Last reviewed 24 July 2026

Baccarat takes no skill. You place one of three bets; the cards do the rest. Two of those bets carry the lowest house edge on the floor. The third, the exciting one, is a 14.4% tax dressed up as an 8-to-1 jackpot. All three, worked out.

How to read this game

Baccarat looks complicated (drawing rules, a third-card table, ritual), but you never make any of those decisions. You pick a bet. That’s it. So the only question is each bet’s price, and expected value answers it. Payoff times probability, summed:

\mathrm{EV} = \sum_i p_i \cdot (\text{payoff}_i)

The house edge is the negative of that, per dollar staked. Every number below comes from the standard 8-deck game.^[1] The three outcomes land like this:

P(\text{Banker}) \approx 45.86\%, \quad P(\text{Player}) \approx 44.62\%, \quad P(\text{Tie}) \approx 9.52\%

Ties push on Banker and Player bets: your stake comes back. So of the hands that get decided, Banker takes slightly more than half:

\frac{P(\text{Banker})}{P(\text{Banker}) + P(\text{Player})} = \frac{0.4586}{0.9048} \approx 50.7\%

That tiny lean is the whole story: why Banker is the cheapest bet on the table, and why the casino charges a fee to make it.

The Banker bet: the cheapest in the house

Banker wins a little more than half of decided hands, so the casino claws the edge back with a 5% commission on Banker wins. Win $100, keep $95.

Worked example: the Banker house edge

\text{Banker win pays } +0.95, \quad \text{Player win } -1, \quad \text{Tie } 0\ (\text{push})

\mathrm{EV} = \underbrace{0.45860}_{\text{Banker}}(+0.95) + \underbrace{0.44625}_{\text{Player}}(-1) + \underbrace{0.09516}_{\text{Tie}}(0)

= 0.43567 - 0.44625 = -0.01058 \approx -1.06\%

Even after the 5% haircut, Banker costs about a penny per dollar: roughly

-1.06\%

. One of the cheapest bets in the entire casino, in the same neighborhood as the craps pass line.

The Player bet: no commission, worse price

Player pays clean even money, no commission. It also wins less often, and that gap costs more than the commission saves.

Worked example: the Player house edge

\text{Player win pays } +1, \quad \text{Banker win } -1, \quad \text{Tie } 0\ (\text{push})

\mathrm{EV} = 0.4462(+1) + 0.4586(-1) = 0.4462 - 0.4586 = -0.0124 = -1.24\%

No commission, but a bigger edge: about

-1.24\%

. The banker vs player verdict in one line: Banker’s

1.06\%

beats Player’s

1.24\%

. The shortfall in wins outweighs the 5% fee.

The Tie trap: an 8:1 payout that isn’t

The Tie pays a juicy 8 to 1. A payout is only generous if it beats the true odds. A tie shows up about once every 10.5 hands, so the fair price, the true odds, is roughly 9.5 to 1:

\text{true odds against a Tie} = \frac{1 - 0.0952}{0.0952} = \frac{0.9048}{0.0952} \approx 9.5\text{ to }1

The casino pays 8 where fair is 9.5. Rare is not what makes the bet expensive. Underpaid is.

Worked example: the Tie house edge

\text{Tie pays } +8, \quad \text{otherwise } -1

\mathrm{EV} = 0.09516(+8) + 0.90484(-1) = 0.76128 - 0.90484

= -0.14356 \approx -14.4\%

About

-14.4\%

: more than ten times the Banker edge, worse than almost anything on the craps or roulette table. The 8:1 sign is bait. Paying big and rarely is what disguises the price.

Why you can’t beat baccarat long-term

Is baccarat beatable? No, and unlike blackjack, it’s not close.

There’s no decision to get right. The drawing rules are fixed and automatic. You bet, the cards fall. Skill has nothing to grip.
Card counting doesn’t work here. Blackjack is countable because a rich deck genuinely swings the edge. In baccarat, card composition barely moves the Banker/Player probabilities. The theoretical counting edge is so small and rare it’s effectively worthless, a result Edward Thorp, the man who beat blackjack, published in 1966.^[2]
“Trend” and scorecard systems are theater. The gambler’s fallacy in physical form. Each hand is independent; five Bankers in a row say nothing about the sixth. The scorecards exist to keep you betting.

Banker and Player are among the fairest bets a casino offers. But “fair-ish” is not “beatable”: the edge, however small, tilts one way.

The scoreboard (baccarat house edge)

\begin{array}{ll} \text{Banker (lowest edge)} & \approx -1.06\% \\ \text{Player} & \approx -1.24\% \\ \text{Tie (sucker bet)} & \approx -14.4\% \end{array}

Two great prices and one terrible one, on the same table. Even the best of them is a slow leak, not a strategy. These are prices, not picks.

The takeaway

The headline bets are genuinely cheap: Banker about $1.06\%$ , Player about $1.24\%$ . No skill to learn, so no edge to win. The Tie’s 8:1 payout is designed to catch your eye; its price is a $14.4\%$ house edge. On this table, the best price is still the house’s.

The same expected-value math runs the sports side of the house. See it work on the free +EV tool, read how expected value finds an edge, or open it in the live app.

The slow bleed

Banker, the cheapest price baccarat offers, at −1.06%. Six players, flat $25, 1,000 hands. The lines barely wobble, yet they drift down. Low variance just means you lose more smoothly.

Want more math?the Nerd Corner

Nerd Corner

Advanced material. Nothing above depends on it.

Where these numbers come from

The three probabilities that run this page (45.86% Banker, 44.62% Player, 9.52% Tie) are exact eight-deck combinatorics, not observation.^[1] The counting question was settled in a statistics journal, by the same man who proved blackjack beatable.^[2]

Ethier, S. N. (2010). The Doctrine of Chances: Probabilistic Aspects of Gambling. Springer. A full baccarat chapter: the exact Banker/Player/Tie probabilities and the house edges on all three bets.
Thorp, E. O. & Walden, W. E. (1966). “A Favorable Side Bet in Nevada Baccarat.” Journal of the American Statistical Association 61(314), 313–328. The classic counting analysis: the main Banker/Player bets shift too little to beat in practice; the one exploitable bet they found was a side bet, not Banker or Player.

Frequently asked questions

What is the house edge in baccarat?

Depends on the bet: Banker about 1.06%, Player about 1.24%, Tie about 14.4%. Banker and Player are two of the lowest-edge bets in the casino; the Tie is one of the worst.

Is the Banker or Player bet better in baccarat?

Banker. Even after the 5% commission on wins, its edge of about 1.06% beats Player’s 1.24%, because the Banker hand wins slightly more than half of decided hands.

Why is the baccarat Tie bet a bad bet?

The Tie pays 8 to 1, but ties hit about once every 10.5 hands, so true odds are roughly 9.5 to 1. Paying 8 against a fair 9.5 creates the roughly 14.4% house edge, more than ten times the Banker bet’s.

Is baccarat beatable? Can you count cards?

No. Unlike blackjack, baccarat’s card composition barely shifts the edge, so counting yields a negligible, impractical advantage. No skill, memory, or trend-reading system changes the outcome. Low edge, but the house still wins over time.