Anti-addiction effectiveness depends on identification accuracy

What it is

Anti-addiction rules (playtime caps, time-window limits) target "minors," but systems actually operate on "accounts and devices." Identification accuracy therefore sets the boundary of the rules' real effect: only when all of a minor's accounts and devices are identified and linked to that one person do the rules fully bind; every unidentified account or device is an unrestricted channel. Account and device identification is the technical infrastructure layer between anti-addiction policy text and actual constraint.

Why it happens

The identification stack has three layers covering different circumvention surfaces. Account layer: real-name-verified account identity, covering single-account use but defeated by multiple registrations. Device layer: device fingerprints (hardware identifiers, system characteristics) identify multiple accounts on one device, aggregating limits from accounts to devices—combined playtime across all accounts on one device counts together, closing the "switch accounts and keep playing" hole. Behaviour layer: cross-device, cross-account behavioural pattern recognition (play-style features, time patterns, social-graph overlap) infers identity links, covering combined multi-device multi-account evasion with the lowest accuracy but the widest coverage. The three layers' combination determines total identification rate, and total identification rate determines anti-addiction's real effect—regulators evaluating anti-addiction systems look precisely at this end-to-end identification and enforcement capability. Each layer also has a false-positive surface: device fingerprints penalise adults on shared family devices (the whole household shares one time pool), and behavioural matching mislinks adults with similar play styles.

Where it stops holding

Identification's privacy cost must be weighed against protection benefit: device fingerprints and behaviour profiles are highly intrusive data practices, justified in the minor-protection context, but collection scope, retention period, and usage boundaries still need minimisation and compliance review. The "sufficient" accuracy standard is context-dependent: spending limits demand higher identification accuracy than playtime reminders (falsely capping an adult's spending is a property impact), and time-window limits tolerate more error than spending limits. Household complexity is the identification system's inherent difficulty: shared devices, parent-managed accounts, and multi-child families (each child needing an independent cap) all require identity-device decoupling structures (sub-accounts / family groups); pure device aggregation distorts these scenarios. The system's adversarial evolution (circumvention, next card) means accuracy is a dynamic metric needing continuous evaluation, not one-time acceptance.

Applying it

• Deploy layers with matching policies: account real-name as the base tier, device aggregation for combined counting (one device's accounts share a time pool), behavioural signals for auxiliary review, with per-layer accuracy requirements set by policy sensitivity.
• Provide sub-account / family-group structures for shared devices so "device aggregation" can be corrected by "identity distinction," avoiding one shared pool for the whole household.
• Verification: test end-to-end identification and enforcement accuracy with controlled combinations (multi-account one device, one account many devices, simulated minor behaviour patterns), quantifying false positives (adults restricted) and false negatives (minors unrestricted) as separate improvement metrics.

Related

• Same group: W10.07.1 Playtime reminders must be proactive at key thresholds · W10.07.2 Forced logouts need advance warning, not mid-match interruption · W10.07.4 Crude time-management tools breed circumvention rather than less play
• Nearby: W10.06 Minor protection and real-name systems · O2.01 Privacy and data minimisation · K1.02 Devices and account systems
• Search terms: device fingerprinting · anti-addiction enforcement · identity resolution · account linkage