How APWebsite Evaluates Advantage Play
APWebsite focuses on observable game state, how that state carries forward, and whether it can reasonably change the value of continued play.
APWebsite's research and editorial process is maintained by the APWebsite Field Research. The standards below explain how public guides distinguish observable facts, estimates and points that still require verification.
1. Start with observable state
We begin with information a player can actually see or verify on the machine: meters, collectors, jackpot values, reel states, multipliers, Wilds, feature progress, and other persistent variables.
2. Separate true persistence from appearance
A visual that grows, fills, or looks close to triggering is not automatically valuable. We ask whether the state survives between spins and whether that state changes future probability, feature value, or the amount of progress still required.
3. Identify the mechanism before estimating value
APWebsite distinguishes between a confirmed or observable mechanism and an estimated entry point. Knowing that a state persists is not the same as knowing the exact meter, count, or machine setup that becomes positive expected value.
The feature or persistent behavior can be directly observed, documented, or consistently verified.
The value depends on assumptions, modeling, incomplete probability information, or for the exact machine version inputs.
The state appears relevant, but the exact positive-EV threshold is not established with enough confidence.
4. Evaluate the complete state
Many games cannot be reduced to one number. A strong evaluation may require multiple meters, bet level, denomination, progressive values, reel position, remaining feature time, or several persistent states at once.
5. Treat calculators as models, not guarantees
APWebsite calculators are decision-support tools. Unless explicitly stated otherwise, model assumptions should not be interpreted as manufacturer-disclosed probability tables. Different casino machine setups, bet structures, and game versions can change the result.
6. Account for variance and bankroll
A positive-EV estimate does not guarantee a profitable session. Slot outcomes remain highly variable, and a favorable state can still produce a substantial loss. Risk, volatility, and the possible cost of reaching a feature matter alongside expected value.
7. Update when better information becomes available
APWebsite guides are intended to improve over time. When better data, clearer features, or stronger verification becomes available, summaries, calculators, confidence language, and member guides may be updated.
Public review register
Guide-level confidence, claim scope and review dates are recorded in the Source & Review Register. It also explains which evidence types carry the most weight and how to report a machine difference.
Review cadence and correction standard
APWebsite reviews machine guides in dated batches, prioritizing short pages, machine setup-sensitive claims and pages with weaker evidence. A review does not certify profitability; it verifies that the public claim is narrow, observable and paired with the correct uncertainty.
A correction should identify the exact title, machine, casino or jurisdiction, denomination, bet and conflicting rule or reset behavior. Material corrections update the page, review date and register entry together.
Approximate entry-point confidence
Member guides separate stronger numerical research from provisional screening estimates. High confidence means the number is supported by a well-defined feature and consistent numerical evidence. Moderate means the number is useful but still sensitive to machine setup. Experimental means APWebsite is giving a practical scouting estimate while the exact break-even point remains uncertain.
Every entry point is approximate. RTP, bet level, denomination, feature value, progressive values, machine machine setup and volatility can change the true threshold. A qualifying state can still lose.