Game-Theoretic Simulation Checklist (Batch 6): Priority Behaviours & Build Order¶
Source: User provided text (December 2025) Context: Prioritisation of behaviors for implementation, focusing on policy relevance and "NHRA-shaped" incentives. Includes a phased build order.
Priority behaviours¶
| Priority | Behaviour to reproduce | Why it’s top-tier | Minimal game lens | What to measure (signature outputs) |
|---|---|---|---|---|
| 1 | Threshold & timing behaviour (EOFY surges, smoothing, “race-to-cap”, claim timing) | It’s the cleanest fingerprint of caps + reconciliation + lags; drives big $ and real operational decisions | Kinked payoff + repeated game | Month-by-month NWAU spikes; cap consumption trajectory; reconciliation volatility; pre/post-threshold behaviour |
| 2 | Coding intensity / complexity drift (bounded upcoding, documentation effort, “creep”) | Explains systematic growth in measured complexity independent of true acuity; central to ABF incentives | Principal–agent (hidden effort) + audit risk | Coding intensity index; complexity inflation vs clinical proxies; audit yield/repayments |
| 3 | ABF↔block (and boundary) shifting (classification/venue shifting) | Eligibility boundaries create substitution incentives that look like “efficiency” but are often reclassification | Mechanism design / boundary game | Share of activity by stream; sudden shifts around rule changes; service-line boundary switches |
| 4 | Capacity-constrained access dynamics (queues, cancellations, ED crowding) | Converts dollars/NWAU into the outcomes everyone argues about; prevents unrealistic “free growth” | Congestion/queueing game | ED LOS distribution; elective breaches; occupancy; cancellation rates; ramping proxy |
| 5 | Internal contracting / delegation (State↔LHN↔hospital target games) | Most real decisions are internal: budget envelopes, targets, and blame-shifting; generates heterogeneity within states | Nested principal–agent | LHN-level variance; target chasing; internal transfer rules; localised surges/cuts |
| 6 | Cost shifting & substitution across settings (public↔private, hospital↔community, downstream spillovers) | Captures externalities and “moving the problem” rather than solving it | Common-pool + routing/substitution | Displacement indicators: ED avoidance, short-stay/inpatient substitution, private share shifts, readmission bounce-backs |
| 7 | Audit / integrity “arms race” (targeting adapts to gaming) | Turns gaming into a measurable equilibrium with policy levers (audit intensity, penalties) | Dynamic enforcement game | Detection rate vs intensity; deterrence curves; false-positive cost vs gaming reduction |
| 8 | Renegotiation & side-payment dynamics (extensions, settlement transfers, brinkmanship) | Important for long-run realism, but less necessary for first-pass operational accuracy | Bargaining / coalition game | Probability of renegotiation breakdown; size/frequency of side-payments; rule-regime switches |
A pragmatic “build order”¶
Core v1 (Must-Have)¶
Focus: Timing/Threshold (#1), Coding Drift (#2), Capacity/Queues (#4).
- State: Time, Budget caps, Claim timing, NWAU, Coding intensity, Capacity utilization, Queue lengths.
- Actions: Submit claim, Allocate capacity, Manage queues.
v2 (High Value)¶
Focus: Boundary Shifting (#3), Internal Contracting (#5).
- State: v1 State + Eligibility rules, Internal contract targets.
- Actions: v1 Actions + Reclassify activity, Adjust internal targets.
v3 (Policy Breadth)¶
Focus: Cost Shifting (#6), Audit Arms Race (#7).
- State: v1/v2 State + Audit intensity, Penalty structure.
- Actions: v1/v2 Actions + Adjust audit params, Gaming responses.
v4 (Long-Horizon Realism)¶
Focus: Renegotiation & Side-Payments (#8).
- State: v1/v2/v3 State + Renegotiation params, Side-payment rules.
- Actions: v1/v2/v3 Actions + Initiate renegotiation, Offer/accept side-payments.