Online slot testing examines whether software implements its stated rules and probability model correctly. In Great Britain, the Gambling Commission sets remote technical standards and a testing strategy. Testing is a consumer-protection requirement, not a prediction that a game will produce a particular result for one person.
Reviewed: 25 October 2026. This guide concerns the Great Britain framework. It explains testing and monitoring without recommending games or betting methods.
What an RNG test examines
A random number generator, or RNG, supplies inputs used to determine game outcomes. The game’s mathematics map those inputs to outcomes and prizes. Reviewing the RNG alone is therefore different from testing the complete game.
RTS 7 requires acceptably random results, including unpredictable output and outcomes distributed according to the expected probabilities. It prohibits adaptive or compensated behaviour that changes probabilities during play.
The standard also addresses correct implementation of the rules, mapping of random inputs and misleading presentation of results. Randomness does not mean that every symbol or every prize is equally likely; the relevant probabilities are those of the game’s defined model.
Testing involves more than a simulation
The Commission’s testing procedure includes review of RNG documentation and source code, assessment of known weaknesses and statistical testing. Game testing also examines design, mathematics, artwork, rules and software implementation.
The procedure requires new games and RNGs to complete testing and have the report provided to the Commission before release. Reports identify details such as software versions, scope, platform, results and limitations. This makes the tested build important: a general laboratory logo does not describe exactly what was examined.
RTP is a mathematical property, not a session promise
The Commission’s RTP monitoring guidance distinguishes theoretical RTP—the designed return—from actual RTP measured using recorded wins and turnover.
A hypothetical theoretical RTP of 96% does not promise that someone will receive £96 from a particular £100 of spending. Observed results can differ from the model over a limited sample. The guidance explains that meaningful monitoring must account for volatility and the amount of play measured.
A short personal run is therefore not a substitute for a technical investigation. Equally, the existence of expected variation does not excuse a confirmed software error.
Rules and the displayed version matter
RTS 3 requires understandable rules and information about how outcomes and prizes are determined to be easily available before a customer commits to gamble. The applicable information concerns the actual game and version.
A familiar title, new artwork or shared series name does not establish identical mathematics. When documenting a game query, retain the title, version information, rules and date rather than relying on an unrelated review.
Changes and monitoring after release
The testing procedure distinguishes updates affecting fairness from minor changes that do not. Not every visual or technical change requires the same external retesting, but change-control records remain important.
The RTP guidance recognises that design, implementation or operational issues can escape initial testing. Ongoing monitoring and investigation are consequently part of the framework. A completed test is not a guarantee that no later fault can occur.
What a consumer complaint can establish
For an existing concern, preserve the relevant game records, displayed rules, dates and correspondence. Distinguish a disputed settlement, missing account entry or apparent rule mismatch from a disappointing sequence of random results.
Testing addresses operation according to the stated model. It does not remove gambling losses, make a game suitable for a particular person or replace account protections. No timing or pattern-based method is offered here as a way to change the outcome.