SP
Standards Protocol
Decision Support Architecture
Discuss an Application
Public observability layer

Comparative Observatory

Descriptive statistical comparison framework. Evidence View primary · Dashboard View secondary.
Comparative Observatory · mandatory disclaimer

Descriptive statistical comparison only. No venue is characterised as unsafe, harmful, or non-compliant. Not a regulatory determination.

REGISTRY GAPNot yet registered as verified public evidence

The values on the Comparative Observatory are derived analytical products produced by deterministic comparative outputs derived from the locked NSW observation set. They are not yet registered as active Evidence Registry records and are presented for reference only. They should not be treated as verified public evidence. Once registered, each value will render with its status, confidence, period, jurisdiction, source, method, limitations, and provenance envelope hash.

evidence registry methodology →registered NSW state evidence →

Methodology

Stage 9 — Comparative Observatory methodology. Descriptive statistical framework. Deterministic computation, no inference.

Computation parameters
  • Latest-anchor snapshot: venue_quarter_observations from stage 4.
  • Long-term window: 6 publication anchors.
  • Sustained growth: 3 consecutive periods of strictly-positive period-over-period EGM movement on the LGA stable-coverage series.
  • Volatility: stdev of period-over-period percentage change on the stable-coverage anchor window ≥ 0.50 (50%).
  • High-density EGM: latest total EGM in the top quintile (p80+) of all LGAs.
  • Standard percentile bands: 10, 25, 50, 75, 90.
  • Dedupe: per (canonical_venue_id, anchor) with net_profit (Type B) priority over per-machine (Type C).
  • Metro/regional classification: registry-based mapping (no inferred socio-economic implication).
Baseline registry
  • BL_STATE_TOTAL_EGM — State total EGM count (latest published anchor)COMPUTABLE
  • BL_STATE_TOTAL_VENUES — State total venues observed (latest anchor)COMPUTABLE
  • BL_STATE_MEDIAN_VENUE_EGM — State median EGM count per venue (latest anchor)COMPUTABLE
  • BL_STATE_YOY_EGM — State year-over-year EGM movementCOMPUTABLE
  • BL_STATE_LT_AVG_EGM — State long-term average EGM count (6-quarter)COMPUTABLE
  • BL_PROFIT_PER_ADULT — Profit per adult (LGA grain)UNCOMPUTABLE
  • BL_PROFIT_PER_MACHINE — Profit per machine (venue/LGA)UNCOMPUTABLE
  • BL_MACHINE_DENSITY — Machine density (EGMs per adult, LGA)UNCOMPUTABLE
  • BL_VENUE_DENSITY — Venue density (venues per adult, LGA)UNCOMPUTABLE
Cluster registry
  • HIGH_DENSITY_EGM — High-density EGM areastop quintile by total EGM countCOMPUTABLE
  • EXPENDITURE_CONC — Expenditure concentrationrequires NSW-published net profit dollars at LGA grainUNCOMPUTABLE
  • SUSTAINED_GROWTH — Sustained-growth regionsLGAs with ≥ 4 consecutive periods of positive YoY EGM movementCOMPUTABLE
  • VOLATILITY — Volatility clustersperiod-over-period EGM movement coefficient of variation ≥ 5%COMPUTABLE
Stage 9 discipline
  • comparisons_are_descriptive
  • comparisons_are_statistical_not_accusatory
  • no_venue_safety_or_compliance_labelling
  • no_predictive_intervention_recommendation
  • no_causation_claim
  • methodology_and_limitations_on_every_output
  • uncomputable_metrics_remain_uncomputable
  • clustering_is_deterministic_quantile_bucketing
  • no_k_means_no_learned_models_no_ai
  • percentile_positioning_uses_published_rank_where_available
  • evidence_view_is_primary_dashboard_view_is_secondary
Descriptive statistical comparison only. No venue is characterised as unsafe, harmful, or non-compliant. Not a regulatory determination.