A change from six decks to eight decks sounds substantial: the shoe contains 312 cards rather than 416. But the effect on standard punto banco main-bet probabilities is quite small. A useful comparison must hold the drawing rules, five-percent Banker commission and 8-to-1 Tie payout constant.
The published mathematical comparison
Michael Shackleford’s exhaustive baccarat analysis at Wizard of Odds produces the following house edges. These are long-run expected losses per unit wagered, not forecasts for a particular session:
- Banker, with 5% commission: six decks about 1.0558%; eight decks about 1.0579%.
- Player: six decks about 1.2374%; eight decks about 1.2351%.
- Tie, paying 8 to 1: six decks about 14.4382%; eight decks about 14.3596%.
Rounded to two decimal places, the Banker and Player values look essentially identical between the common shoe sizes. The Tie result moves by around 0.08 percentage points, although the Tie wager’s house edge remains far larger than either main bet’s in this example.
Why the cards in the shoe matter
Six 52-card decks contain six copies of each individual rank-and-suit card, while eight decks contain eight copies. Without replacement, removing one card changes the mix remaining. Baccarat’s fixed drawing tableau evaluates all possible card sequences. Different starting deck compositions therefore produce subtly different relative frequencies of Banker wins, Player wins and ties.
For example, an eight-deck baseline gives about 45.8597% Banker wins, 44.6247% Player wins and 9.5156% ties. Six decks gives about 45.8653%, 44.6279% and 9.5069%, respectively. These are rounded population values for the specified models, not a promise about the next hundred hands.
An easy numerical interpretation
Suppose someone examines 1,000 hypothetical main wagers of $10 each, for $10,000 wagered in total. Multiplying that turnover by the 8-deck Banker edge of approximately 1.0579% gives an expected loss of about $105.79; six decks gives about $105.58. That difference is only about 21 cents across the hypothetical $10,000 turnover. Individual results may be far above or below either expected amount.
The example assumes unchanged stake size, commission and settlement rules. It does not justify treating gambling as a savings plan or a source of income. Our RTP and turnover explainer separates expectation from realised results.
What changes the comparison much more
A Tie paying 9 to 1 rather than 8 to 1 changes its house edge markedly, even using the same shoe. Likewise, special Banker-six half payments or three-card-seven push rules require an entirely new table of returns. Read our Tie paytable comparison before comparing a headline number from another format.
The responsible conclusion is narrow: deck count is a legitimate mathematical variable, but it does not create a reliable predictive system from previous winners, and shoe-size changes should never be confused with a change in payout rules.
Sources and further reading
Editorial note: This article is for adult readers and explains rules and mathematics, not a method of predicting outcomes. Rules and paytables may vary; gambling can cause financial loss.