Methodology

This page describes how the baccarat.global school calculates house edges, which primary sources it uses, how the free-play simulator is built, and how the daily challenge shoe is generated.

House edge calculations

All house edge figures used in the school are derived from the standard probability calculations for an 8-deck punto banco shoe, unless a variant is explicitly specified.

The primary calculation method:

Expected return per unit = (win probability x payout) + (lose probability x -1) + (tie probability x 0)

For the Banker bet (standard 5% commission):

  • Win probability: 0.4586 (Banker wins 45.86% of hands)
  • Payout on a win: 0.95 (after 5% commission)
  • Lose probability: 0.4462 (Player wins 44.62% of hands, causing Banker to lose)
  • Tie probability: 0.0952 (ties are a push)
  • Expected return: (0.4586 x 0.95) + (0.4462 x -1) + (0.0952 x 0) = 0.43567 - 0.4462 = -0.01053
  • House edge: 1.06% (rounded to 2 decimal places)

For the Player bet:

  • Expected return: (0.4462 x 1) + (0.4586 x -1) + (0.0952 x 0) = 0.4462 - 0.4586 = -0.0124
  • House edge: 1.24%

For the Tie bet at 8 to 1:

  • Expected return: (0.0952 x 8) + (0.9048 x -1) = 0.7616 - 0.9048 = -0.1432
  • House edge: 14.36%

These figures are canonical and treated as accepted fact within the school -- in the same category as a chemistry constant. No external citation is required for standard house edge figures derived from the drawing tableau. The underlying win/loss/tie probabilities derive from combinatorial analysis of all possible card draws from an 8-deck shoe under the standard drawing tableau.

For variant-specific edges (EZ Baccarat, No Commission, Mega Baccarat), the school uses figures drawn from manufacturer game-rules documentation where available (e.g. Galaxy Gaming for EZ Baccarat variants, Evolution for their variant pages) and peer-reviewed combinatorial analysis for edges not covered by manufacturer documentation. Any edge figure that can't be cross-referenced against at least one primary or manufacturer source is flagged as an estimate rather than stated as a confirmed figure.

Primary sources

We cite primary sources only. The school's factual canon draws from the following categories:

Regulatory and legal sources

Manufacturer and operator sources

  • Galaxy Gaming: EZ Baccarat, Dragon 7 side bet, and other proprietary variant rules
  • Evolution: Live baccarat product documentation and variant rules
  • Pragmatic Play Live: Live product documentation
  • Casino operators' own websites for confirmed current information about table types, variants offered, and minimum stakes

Court and parliamentary records

Editorial sources (for factual news events only)

  • The Guardian, Financial Times, Bloomberg, Reuters, BBC, South China Morning Post, The Times, GGB Magazine, Asia Gaming Brief

Historical sources

  • Britannica for game history and definitions
  • Wikipedia for definitions only, not for facts requiring independent verification

The following sources are not cited and are removed from any lesson where they previously appeared: wizardofodds.com, askgamblers.com, casino.guru, trustpilot.com, casinomeister.com, casino.org, vegasinsider.com, gamblingsites.com, or any affiliate review aggregator.

The free-play simulator

The free-play simulator on baccarat.global uses a software implementation of the standard 8-deck Punto Banco drawing tableau. The virtual shoe is generated using a cryptographically secure pseudorandom number generator (CSPRNG) seeded at the start of each shoe. The shoe contains exactly 416 virtual cards (8 decks of 52 cards each). Cards are drawn from the virtual shoe without replacement, exactly as in a physical shoe. The shoe is reshuffled when the cut card position is reached (typically at around 300 to 330 cards dealt, replicating standard casino practice).

The simulator applies the standard drawing tableau:

  • Player draws on 0 to 5, stands on 6 to 7, natural ends hand
  • Banker draws or stands based on the full conditional tableau (Banker total 0 to 2 always draws; Banker total 3 to 6 conditional on Player's third card; Banker total 7 stands)

The simulator is for entertainment and education. It doesn't use real money. Its purpose is to allow players to observe the drawing rules in action and become familiar with the rhythm of the shoe before playing with real funds.

The house edge experienced in the simulator over a large number of hands will approximate the theoretical 1.06% on Banker and 1.24% on Player, with variance around those figures for any individual session. This is a design feature, not a limitation: it accurately represents what players will experience at a real table.

The daily challenge shoe

The daily challenge shoe is a predetermined 8-deck shoe that is the same for all players on a given calendar day. The shoe is generated at midnight UTC using a deterministic seeding process that takes the date as its input. All players playing the daily challenge on the same day play the same shoe in the same card order.

The purpose is to allow comparison between players and to create a shared daily experience. The shoe isn't accessible in advance and can't be predicted from the seed without access to the seeding algorithm.

The daily challenge scoring system records the result of each hand in the shoe and calculates a player's performance against a defined scoring rubric. The rubric is described on the daily challenge page and is based on correct bet selection (Banker and Player bets score; Tie bets don't) and session discipline (staying within defined stake limits scores; exceeding them doesn't).

The daily challenge doesn't use real money. The leaderboard is for entertainment only.

Updates to this methodology

If the school's methodology changes (for example, adopting a new primary source for variant edge calculations, or modifying the simulator's cut card position), this page will be updated with the date and nature of the change. The current methodology version is shown at the foot of this page.

Current version: June 2026.