Six connected modules. One loop. Evidence anchors every claim. Intelligence surfaces the signal. Assessment makes reasoning visible. Decision produces a recorded artefact. Outcome measures what actually happened. Learning returns to Evidence. The next decision is measurably better than the last.
Hover, focus or tap any module. The loop is deliberate. Learning always returns to Evidence.
Where the loop begins.
Nothing enters intelligence without a registered evidence record.
Every piece of source material is captured with provenance, publication period, jurisdiction, retrieval record and reliability classification. Evidence is registered — not narrated, not paraphrased, not inferred.
The registered evidence set is examined for signals, patterns, contradictions, relationships and unknowns. Nothing is surfaced that cannot be traced back to at least one evidence record. Contradictions are surfaced, not resolved by hand.
Scenarios and alternatives are compared against declared assumptions and identified risks. The reasoning path is inspectable — every assumption applied, every alternative considered, every risk weighted.
A Standards Protocol decision carries its evidence with it. Every decision record exposes a Decision Score, a Decision Confidence, a recommendation, a rationale, a review trigger and an outcome state. The record is the artefact.
Outcomes are captured against the recorded decision: result, variance, unexpected consequences and performance. Outcomes are not lost to a report and are not re-narrated to fit expectation.
Learning returns to Evidence: the evidence set is updated, assumptions are refined, the framework itself is adjusted. The next decision is measurably better than the last — because the record of what actually happened is available to it.
Four commitments that hold across every Standards Protocol module. If any of them fails, the record fails.
No claim enters the loop without a registered evidence record, a source, a retrieval date and a reliability class.
Assumptions are declared. Estimates are labelled. Values that cannot be observed are declared non-computable rather than fabricated.
Every output resolves back to its source publication, its parser version, its methodology version, and the deterministic calculation that produced it.
A Standards Protocol decision is the record. It carries its evidence, its confidence and its review trigger with it — always.
The Gambling Design Observatory is the first public application of Standards Protocol. Each of the six modules below is enacted against real, registry-backed NSW records.
NSW quarterly LGA gaming-machine records are ingested. Each row is registered with its source authority, retrieval date, method, reliability and limitations. 1,959 records to date.
Signals are surfaced — including changes in observed profit-per-machine, LGA-level tax variance, and reporting-period continuity. Contradictions between publication windows are recorded, not resolved silently.
Scenarios are compared against a declared 2019 post-reform baseline. Assumptions are stated explicitly. Where an assumption cannot be defended, the scenario is excluded rather than weakened.
A recommendation is recorded with a Decision Score, a Decision Confidence, a rationale, and a review trigger set against the next NSW quarterly release. The record is inspectable.
When the next publication window closes, the outcome is captured against the recorded decision. Variance, unexpected consequences and performance are logged.
The outcome updates the evidence set and refines the assumptions applied at Assessment. The next decision on the same domain begins with a better evidence base than the last.
Bring us the decision, not a predetermined solution.
Discuss an Application