Behavioural Metrics Checklist
Master checklist for responsible, collectable user and market behaviour data.
Important Framing
This is behaviour expressed through markets, not behaviour of people.
Every item below is: aggregated • non-identifying • observational • non-interventional
A. Temporal Behaviour (Population-Level)
Participation timing distributions
Pre-event vs in-event vs post-event ratios
Time-of-day / day-of-week concentration
Event-proximity acceleration patterns
Fully aligned with NSW "tracking change over time" priority. Already exceeds episodic research models.
B. Aggregate Exposure Dynamics
Exposure velocity (rate of accumulation)
Exposure decay (how quickly exposure dissipates)
Concentration gradients
Persistence over time windows
Gold-standard public-health compatible data. No government system currently operationalises this continuously.
C. Volume Distribution
Volume clustering
Late volume surges
Cross-market volume dispersion
Temporal skew
Purely descriptive. No judgement embedded. High policy relevance.
D. Price & Volatility Structures
Price compression / expansion
Volatility regime shifts
Variance contraction / expansion
Inter-market divergence
Essential for understanding structural conditions of risk. Not used by government yet — but fully defensible.
Aggregated, non-identifying metrics measuring engagement with harm minimisation content. Expressed as percentages of total visitors.
A. Article Engagement
Priority Harm Article Views (%)
% of visitors reading top 3 harm awareness articles
Understanding Gambling Harm, Family Impact, Getting Help
Harm Awareness Article Views (%)
% of visitors reading any harm-related content
General Article Views (%)
% of visitors engaging with educational articles
Measures proactive education uptake. High percentages indicate effective content placement.
B. Help Resource Clicks
Help Resource Click Rate (%)
% of visitors clicking gambling help links
GambleAware, GamCare, Gambling Help Online, etc.
Help Click Breakdown by Resource
Which help services receive most engagement
Help Click Context
Which pages drive help-seeking behaviour
Critical for demonstrating platform responsibility. Tracked anonymously without user identification.
C. Support Engagement
Support Resource Click Type
Categories: gambleaware, self_exclusion, take_break, help_link
Warning Override Rate
When users proceed despite risk signals
Abstention Quality
Users choosing not to act based on analysis
Responsible abstention is a positive outcome. Measured to validate harm minimisation effectiveness.
Status Assessment
- Already meets the current NSW strategic vision
- Adding Section 2 items future-proofs the system
- Section 3 items quietly set the next industry standard
- Section 4 Harm Controls actively demonstrate regulatory compliance
None of this requires changing who the system is. We are not expanding scope — we are deepening observation.
