Baccarat Odds and the True Cost of Every Bet

How often do Banker, Player and Tie win in baccarat?

Everything on this page falls out of one set of numbers. Deal an eight-deck shoe of baccarat under the standard drawing rules and the three possible results land at these rates.

The Banker advantage is structural rather than accidental. The Player hand acts first and follows a rule that ignores everything else at the table. The Banker hand acts second, and its drawing rule is keyed to the exact rank of the Player third card, so it is effectively playing with more information every round. That is the whole source of the gap, and it is set out card by card on the baccarat rules page.

Strip ties out of the calculation and the Banker side takes about 50.68% of decided hands. That is the number a casino would be facing if it paid Banker wins at full even money, and it explains everything that follows.

The 5% commission is a price correction, not a service charge

Pay Banker at 1 to 1 with no deduction and the bet returns more than it takes. So the operator deducts 5% from winning Banker bets. Run the expected value with that deduction applied and the Banker bet settles at a baccarat house edge of 1.06%.

The Player bet needs no correction. It wins less often, it pays full even money, and it costs 1.24%. Notice how small the difference between the two is. Ties returning both main stakes is what compresses them: without the push rule the Player bet would be a far worse proposition than it is.

Read the edge column as a price per ringgit wagered over a long run, not as a prediction for your session. A 1.06% edge means that across an enormous number of hands the operator expects to keep just over one ringgit per hundred staked. Across a hundred hands it is buried entirely by ordinary variance.

Where the numbers come from, and why you can trust them

Baccarat probabilities are not estimates from observed play. They are computed by enumeration. Because the drawing rules are completely fixed, every possible arrangement of an eight-deck shoe can be worked through and the outcomes counted, which is how figures like 45.86% are stated to two decimal places rather than approximated.

That is a real advantage of a game with no decisions in it. In blackjack the house edge depends on how well the player plays, so any published number carries an assumption about strategy. In baccarat there is no assumption to make. The 1.06% on Banker is the price of that bet for everyone at the table, expert and beginner alike, and it does not move because of anything that happens during the round.

It also means the numbers are checkable rather than promotional. Any published baccarat table using standard punto banco rules over eight decks produces the same three probabilities, and any figure that differs is either using a different deck count, a different tie payout or a different commission arrangement. If a source quotes an edge without stating those three settings, it has left out the part that matters.

Variance is far larger than the edge over any session you will actually play

A 1.06% house edge sounds like a slow leak, and over hundreds of thousands of hands that is exactly what it is. Over the two hundred hands you might play in an evening, it is almost invisible next to the swings.

Put two hundred RM10 bets on Banker and the expected cost is a little over RM21. The realistic range of outcomes around that is in the hundreds of Ringgit in either direction, because the bet resolves close to a coin flip and streaks of eight or ten in a row are ordinary rather than remarkable. This is why sessions feel nothing like the mathematics: the mathematics describes the destination, the session describes the road.

Two consequences follow. First, a winning night proves nothing about a method, and neither does a losing one. Any result you can achieve in an evening is comfortably inside normal variance for a game this close to even. Second, the number that actually determines your long-run cost is total turnover, so bet size multiplied by hands played is the figure worth watching rather than the edge itself.

It also explains why progressions feel like they work. Doubling after a loss produces a long series of small wins punctuated by rare disasters, so a short sample will usually show a profit. The disaster arrives eventually, and it is proportionally larger than everything the system collected before it. The baccarat strategy page takes the popular systems apart one by one.

Commission-free tables move the cost rather than removing it

Some tables advertise Banker at a flat 1 to 1 with no commission, and settle a winning Banker hand totalling exactly six at half stake instead. That single carve-out has to do all the work the 5% was doing, and a Banker six wins often enough that it does slightly more.

The practical result is that a commission-free table costs a little more per hand than the standard version, in exchange for simpler bookkeeping and a cleaner-looking payout. If you are choosing purely on price, the commission table wins. If you are choosing on how little you want to think about, the difference is small enough to ignore.

The tie pays 8 to 1 and costs about 14.4%, the worst price on the layout

Ties land about 9.52% of the time, which is roughly one round in ten and a half. To break even, the bet would need to pay a little over 9.5 to 1. Most tables pay 8 to 1, and that shortfall is where the roughly 14.4% house edge on the baccarat tie comes from.

A minority of tables pay 9 to 1 instead. That is a genuine improvement and it takes the cost of the tie down to about 4.8%. It is still several times the price of either main bet, so a better tie is a reason to prefer one table over another, not a reason to start betting the tie.

The tie is also the clearest illustration of a distinction worth carrying to every casino game: how often something hits and what it is worth are separate questions. One round in ten sounds frequent. At 8 to 1 it is still the most expensive way to put money on a baccarat table.

Deck count moves the numbers in the third decimal place

Published baccarat odds almost always assume eight decks, because that is what live tables deal. Fewer decks change the composition effects slightly, which shifts each figure a little.

The direction is consistent: fewer decks are marginally kinder to Banker and marginally harsher to Player and to the tie. The magnitude is in hundredths of a percentage point for the six-deck case, which is far too small to base a table choice on. What never changes is the ordering. Banker is the cheapest bet in every shoe size, the Player bet is next, and the tie is a long way behind.

This is also why card counting has no practical application here. The theoretical edges that appear deep in a depleted shoe are tiny, they arrive rarely, and live tables shuffle long before they could be acted on. Nothing about a multi-deck baccarat shoe rewards tracking.

What the edge actually costs you per hour depends on pace, not on skill

House edge is expressed per unit staked, so the money that matters is total turnover: bet size multiplied by hands played. Baccarat deals quickly, which makes pace the largest lever a player controls.

Those are long-run expectations rather than forecasts for any single hour, and the swings around them are much larger than the numbers themselves. The point is the shape: at the same stake, a speed table costs roughly six times what a squeeze table costs, purely because it deals six times as many hands. Choosing a slower table is a bigger financial decision than choosing between Banker and Player.

Side bets sell frequency and charge heavily for it

Most live baccarat tables carry a menu of optional wagers around the edge of the layout. They resolve on the cards rather than on the outcome, and their paytables are printed on the table information panel.

• Player Pair and Banker Pair. Wins if the first two cards of that hand are the same rank. Typically pays 11 to 1. The pair does not have to be the same suit.
• Perfect Pair. Wins only if the first two cards of a hand match in rank and in suit, and pays considerably more, commonly 25 to 1. Rarity is the entire product.
• Dragon Bonus. Pays on the margin of victory. A natural win pays a small fixed amount, while a non-natural win by a wide margin pays a scaling ladder that peaks at a nine-point win.
• Either Pair and Big or Small. Variations on the same idea: either hand pairing, or the total number of cards dealt in the round being four, five or six.

The honest summary is that every one of these costs several times more per ringgit than Banker or Player, and some cost more than ten times as much. Exact figures vary by paytable and by deck count, so read the panel at the table you are actually sitting at rather than trusting a general number. The broader picture across live games sits on the live casino side bets page.

None of that makes them illegitimate. They are entertainment priced as entertainment, and a small side bet alongside a main bet is a reasonable way to spend money if you know what the trade is. The mistake is treating them as an alternative to the main bets rather than a surcharge on top of them.

Rebates bite harder in a low-edge game than in a high-edge one

A cash rebate paid on turnover interacts with house edge in a way that is worth understanding, because the effect is much larger in baccarat than in most formats.

House edge is charged on turnover. A rebate is paid on turnover. So a rebate is a direct offset against the edge, expressed in the same units. On the banker bet at 1.06%, a rebate of 0.80% of turnover cancels most of the theoretical cost. On the tie at 14.4%, the same rebate barely registers.

U88 runs a daily cash rebate on live casino turnover with no maximum, paying 0.20% at Normal tier and rising by member tier to 0.80% at Diamond. Baccarat suits that structure, because the game generates high turnover at the lowest edge on the floor. It does not make the game favourable, and no promotion does, because rebate terms carry their own qualifying conditions and the edge continues to apply to every hand. What it does mean is that the value of a rebate depends on which bet you were making and which tier you sit in. Chasing turnover on the tie to earn rebate is the clearest possible example of paying fourteen to collect a fraction of one.

Reading a paytable is a skill worth ten minutes of your time

Every live table publishes its rules and paytable in an information panel, and three lines of it decide the price of your session.

None of these settings changes the drawing rules, which are identical everywhere and set out on the baccarat rules page. They change the price. Two tables side by side in the same lobby can differ meaningfully in cost while looking indistinguishable, and the only way to tell is to open the panel.

Where baccarat sits against the rest of the live floor

At 1.06% on the Banker bet, baccarat is among the cheapest wagers a live casino offers. That is a real advantage over most of the floor, and it is the main reason the game holds high-limit tables across Asia. It is also the ceiling: the number is fixed by the rules, nothing you do at the table lowers it, and no sequence of results makes it temporarily favourable.

If you want the comparison laid out across every table game, the house edge comparison for live casino games ranks them on the same basis. And if you have arrived here looking for a way to turn these numbers around, the baccarat strategy page explains plainly why no betting pattern moves them.