Number of decks
The number of decks is how many standard 52-card packs are shuffled together into the shoe a table game is dealt from, usually between one and eight in blackjack and normally eight in baccarat. It counts as a rule of the game rather than a detail of presentation, because it moves the house edge. In blackjack the shift is measurable; in baccarat it is small on the main bets and much larger on some side bets.
How deck count changes the odds
The mechanism is card removal. In a single deck, an ace on the table leaves three aces among the remaining 51 cards, while in an eight-deck shoe of 416 cards the same ace leaves 31 among 415, so each dealt card matters far less.
For blackjack, the commonly cited figures under one fixed rule set run from about 0.16% with a single deck to 0.46% with two, 0.64% with six and 0.66% with eight. Most of the rise comes from the first extra deck, and the step from six to eight is tiny. Fewer decks raise the chance of a natural blackjack, from roughly 4.83% to 4.75% as decks go from one to eight, and they make doubling slightly stronger.
Baccarat barely notices.
Where players see the deck count
On a live casino table the deck count normally sits in the game rules or help screen, often alongside the dealer’s soft 17 rule and the blackjack payout. RNG versions list it in the same place, though how often the virtual shoe is reshuffled varies by studio and is not always stated.
For baccarat, published figures put the Banker edge at about 1.06% on both six and eight decks, while the pair bet’s edge climbs from around 10.36% on eight decks to 29.41% on one.
Common confusions about decks
The deck count is often treated as a verdict on its own, and it is not one. A single-deck blackjack table that pays 6:5 on a natural can carry a worse edge than a six-deck table paying 3:2, so the payout and dealer rules have to be read together with the shoe size.
It is also confused with penetration, which is how far into the shoe the dealer goes before reshuffling. That is a separate setting, and operators handle it differently from one table to the next, so a figure for one game says nothing about another.
Finally, a short run of good hands at a one-deck table proves nothing, since the house edge describes the long-run average over thousands of rounds rather than any single session.
