A sample average can differ from a theoretical average
Theoretical return and an observed sample describe different objects. One is a model; the other is a particular set of outcomes.
The distinction that matters
Observed return divides recorded payouts by recorded turnover for a defined sample. The theoretical expectation requires an applicable probability model. A difference between the two does not, by itself, identify an incorrect model, a software issue or misconduct. Missing transactions and inconsistent definitions can also distort a sample.
A worked example
A fictional C$100 sample returning C$130 has observed return of 130%. That does not establish a theoretical RTP above 100%. Another small sample can return zero. Without the distribution and a valid statistical method, attaching a confidence claim to either sample would be unsupported.
Your next steps
- Define the sample boundaries and included transactions.
- Reconcile stake and payout units.
- Keep observed and theoretical values in separate fields.
- Record missing data before interpreting the ratio.
Where this answer stops
Do not gamble to generate a larger test sample. Northline does not audit random-number generators or infer a precise RTP from personal records. A documented game dispute belongs with the operator's formal process.
Sources & reading
Original Northline explanation. Worked examples use stated assumptions; they do not establish an operator’s policy or predict personal results. Read how we handle evidence.
An editorial review date is not a new policy effective date. Always check the current source for operator-specific conditions.
Suggest a correction