Learn casino probability without confusing averages and predictions
Casino mathematics becomes easier to follow when three questions stay separate: how likely an outcome is, what that outcome pays, and whether a recorded result follows from the published algorithm. The guides in this section connect those questions to working tools, so a definition can be checked against an example instead of being accepted as a slogan.
Start with the subject that caused the uncertainty. If you are comparing returns, learn about RTP and expected value. If a short session looks surprising, study variance, drawdown and the difference between median and mean. If a result is disputed, start with seed commitments and the inputs needed for a replay.
Probability explains frequency; RTP explains weighted return
A slot probability is not a payout rate. A common outcome can return less than the stake, while a rare outcome can contribute a substantial share of theoretical RTP. The house edge guide explains why averaging total returns differs from calculating net profit.
Use the Plinko learning section for game-specific examples, then inspect the odds tables and calculator. The formulas describe the model and the selected table. They do not make a random sequence follow its average at regular intervals.
| Start with | When you want to understand | Next step |
|---|---|---|
| House edge | Return, turnover and expected loss. | Inspect a captured table in a calculator. |
| Provably fair | Commitments, seeds and replay evidence. | Choose a supported operator in Verify. |
| Glossary | A specific unfamiliar term. | Follow its explanation back to the relevant guide or tool. |
A session is more than its expected value
Two payout tables with similar average returns can produce very different balance paths. Rare large outcomes make the mean sensitive to a small number of sessions, while the median describes the middle of the sample. Neither number alone communicates how often a bankroll runs out before the planned number of bets.
The simulator connects these concepts. Change one setting at a time and compare the outcome distribution, not just the most impressive line. A simulation is an estimate under chosen assumptions; it is not a proof of what a casino did in a particular round.
Fairness checks need evidence as well as an algorithm
The provably fair guide explains the relationship between a hidden seed, its hash, a client seed and a nonce. A check can compare a revealed seed to a saved commitment and replay the resulting outcome. The timing and provenance of the saved commitment affect the strength of the conclusion.
The glossary defines the individual terms, and Verify lists working adapters with their evidence status. Our editorial policy and methodology describe how explanations and checks are presented. Reading them together helps distinguish a supported result from a broader claim that the evidence does not establish.