Methodology
How the course math is computed, how the simulators work, and what assumptions we use.
How the course math is computed, how the simulators work, and what assumptions we use.
House-edge figures, expected loss-per-hour calculations, and probability statements in the courses use standard published values that we cross-check against multiple analytical sources. The medium-term goal is to compute every figure in-house from first principles, publish the assumption stack, and ship the working with each lesson. Today the assumption stack below is what we use; the in-house calculator is in development.
The baseline assumptions, unless a lesson states otherwise:
Expected loss per hour uses a default of 60 hands or spins per hour for table games, 30 for live-dealer baccarat, and the operator's published spin rate for game-show formats. If you play faster or slower, scale linearly.
The planned Monte Carlo simulators will run in your browser using a Web Worker for the math loop, with the pseudo-random number generator seeded at session start from crypto.getRandomValues and the seed exposed so any run is reproducible by anyone who copies it. The simulators currently shown on /tools/ pages are previews of that design. Treat their output as illustrative, not as final.
Off-the-shelf casino simulators tend to be slow, hard to audit, or built to sell a system. The version we are building will have none of those constraints. The math will be open in the page source. The PRNG will be industry-standard. The simulators will ship the seed so you can verify any specific run.
Simulators are not strategy oracles. They illustrate the math at scale. They do not predict any specific outcome. In particular, our simulators assume independent, fair rounds with no correlation between hands (which is the casino-floor reality for shuffled-each-hand RNG games and shuffled-shoe live games before a single deck is dealt). They do not model dealer signature, biased physical wheels, or any other edge-finding tactic that requires sustained data collection.
Course content and review scores are reviewed every six months. If you spot an error, write to [email protected] and we will correct and publish the correction here.