RTP is a long-run model
What a percentage can and cannot say about one session.
24 SEP 2026
RTP describes modelled return over a very large sample under stated rules; it is not a session promise.
What the claim leaves out
Commercial language compresses several separate questions into one attractive phrase. Cleobetra restores the missing boundaries: which entity makes the statement, which product version it covers, which event begins and ends the measurement, and which source can reproduce the result.
That distinction matters because a polished interface can make a condition feel self-evident. Evidence works in the opposite direction: it names the condition, records the date and leaves uncertainty visible.
A four-step check
- Capture the exact wording.
Save the page, date, product name and version before interpreting it.
- Define the boundary.
State what starts the event, what ends it and what is actually being counted.
- Find a primary record.
Prefer rules, regulator registers and dated reports over affiliate summaries.
- Record the unresolved part.
An honest “not established” is stronger than a confident guess.
Use RTP to compare documented versions, never to chase a shortfall.
Read the Cleobetra method →Primary-source shelf
Australian legal claims should be checked against the Australian Communications and Media Authority. Product mathematics should be checked against the exact rules and version supplied by the developer or regulated venue.