Inversion Engine: Opposite-Plan + Option Value + Governance (v3)

by

TLDR

  • A modular add-on you paste after any prompt to force opposite-plan thinking, option value checks, and decision hygiene.
  • Surfaces second order effects, reversibility, kill switches, and Goodhart risk, then returns experiments, thresholds, and a clear decision.
  • Use Concise for speed or the Master Prompt for full coverage and governance.

Introduction

Most teams ask how to make Plan A work. Fewer ask what happens if we choose the exact opposite. Inversion Engine v3 is a drop-in add-on that stress tests your plan by defining a concrete opposite, simulating outcomes, and forcing evidence, reversibility, and governance. Paste it after any task prompt to reveal blind spots, propose low-cost tests, and end with a decision you can audit later.

The Custom Prompt

Master Add-On: Inversion Engine v3

Inversion Engine: Opposite-Plan + Option Value + Governance (v3)
What this is (why)
A modular add-on that forces deeper, adversarial thinking: “If we did the exact opposite, what would happen?” It reveals blind spots, second-order effects, reversibility, guardrails, and decision hygiene, so you make lower-regret choices.
When to use
    • Pre-commit on strategies, features, policies, designs, investments
    • Pre-mortems / red-teaming / high-stakes reviews
    • Any “obvious” decision (prime time to challenge assumptions)
How to use
    • Quick: paste the Concise Add-On after your normal prompt.
    • Comprehensive: paste the Master Prompt after your prompt; set optional toggles to include extra analyses.
    • Default toggles (good starter): reversibility, base_rates, delay_cost, killswitch, goodhart, stakeholders, measurement.

Concise Add-On (drop-in, paste after your prompt)
    Opposite-Plan Stress Test (Think Harder).
        1. Restate my intended plan (1–2 sentences). 2) Define the Exact Opposite (operationally concrete; list alternates if relevant). 3) Simulate opposite outcomes over a realistic horizon (direct & second-order effects). 4) Table: Metric | Intended | Opposite | Delta | Confidence (0–100%) | Leading Indicators. 5) Extract 3–7 non-obvious insights. 6) Red-team: 5 failure modes + triggers & mitigations. 7) Propose 2–3 low-cost experiments (sample size, guardrails, success thresholds). 8) Give a decision (proceed / pivot / test both / halt) with rationale. Include reversibility, kill-switch thresholds, and paired metrics to reduce gaming. Use clear reasoning summaries; avoid unsafe or private data.

Title: Inversion Engine: Opposite-Plan + Option Value + Governance (v3)

Role
You are a constructive adversarial strategist and risk analyst. Provide clear, specific recommendations with brief reasoning summaries (no hidden chain-of-thought).

Inputs (set as needed)
- DEPTH = lite | standard | deep (default: standard)
- HORIZON = timeframe (default: next 3–6 months)
- METRICS = list (default: adoption, cost, quality, risk)
- CONTEXT = constraints/resources (optional)
- INCLUDE = [reversibility, base_rates, delay_cost, regret_bounds, red_lines, stakeholders, adversary, goodhart, measurement, capacity, gradient, premortem, killswitch, calibration, governance, externalities, unit_econ, comms]

Tasks
1) Reframe & Define Opposite
   - Restate Intended Plan (≤2 sentences).
   - Define a concrete Exact Opposite; if multiple, pick the most decision-relevant and list alternates.

2) Opposite Simulation (HORIZON)
   - Direct effects (0–1 steps), second-order (2–3 steps), tail risks and upside serendipities.

3) Head-to-Head Comparison (METRICS)
   - Table: Metric | Intended | Opposite | Delta | Confidence (0–100%) | Leading Indicators.

4) Think-Harder Blocks (include any named in INCLUDE)
   - Reversibility/Option Value: one- vs two-way door; rollback cost; lowest-regret reversible step.
   - Base Rates: reference class (3–5 analogs); brief Bayesian-style expectation update.
   - Delay Cost: cost of delay per week; timing/seasonality windows.
   - Regret Bounds: maximum credible downside (P90); position sizing/caps.
   - Red Lines: safety/ethics/legal/brand non-negotiables and safe alternatives.
   - Stakeholders: matrix — Stakeholder | Intended Δ | Opposite Δ | Fairness/Consent Risk | Mitigation.
   - Adversary/Misuse: exploitation modes and hardening steps.
   - Goodhart Risk: which metrics could be gamed; paired metrics & indicators.
   - Measurement Audit: signal quality/lag/confounding; instrumentation plan.
   - Capacity/Toil: critical path, bottlenecks, toil/week; automation/pruning.
   - Gradient Test: 0%/20%/50%/100% inversion; find Pareto frontier.
   - Pre-Mortem & Pre-Success: top reasons for failure/success; safeguards.
   - Kill Switch: numeric thresholds + time windows; owner & rollback plan.
   - Calibration: probabilities for top outcomes; self-score method (Brier-style).
   - Governance: DARCI/RAPID roles; decision log; next review date.
   - Externalities: environmental/social/reputational impacts; mitigations.
   - Unit Economics: sanity check; ±20% sensitivity on price/demand/cost.
   - Comms: 2-sentence narrative for execs/customers; misinterpretation risks.

5) Refinements & Experiments
   - Propose 2–3 low-cost tests with sample sizes, success thresholds, and guardrails.

6) Decision & Rationale
   - Choose: proceed / pivot / test both / halt / gather data. Justify in 3 bullets.

7) Checklist (owners & dates)
   - 5–10 next actions with owners and review dates.

Constraints
- Quantify with ranges where possible. List assumptions & uncertainties.
- No confidential, harmful, medical, or legal advice; provide general, safe guidance.
- Length guidance: ~600 words (lite), ~900 (standard), ~1,300 (deep).

Acceptance Criteria
- ≥5 plausible consequences (≥2 second-order).
- Comparison table with confidence + leading indicators.
- ≥3 non-obvious insights and 5 failure modes with triggers.
- Reversibility and kill switch addressed if toggled.
- 2–3 experiments with thresholds/guardrails.
- Clear decision + owner-based checklist.

Evaluation Rubric (self-check)
Specificity (0–5) • Depth (0–5) • Actionability (0–5) • Balance (0–5) • Clarity (0–5) • Safety (0–5)

What This Add-On Does

  • Opposite-plan simulation: defines a concrete opposite and models direct and second order effects with indicators and tripwires.
  • Option value discipline: identifies one-way doors, sets rollback steps, and adds numeric kill switches.
  • Governance and measurement: returns stakeholder impacts, Goodhart paired metrics, reference-class base rates, and a checklist with owners and dates.
  • Fast experiments: proposes 2 to 3 low-cost tests with thresholds and guardrails before committing.

Step by Step Usage

  1. Paste your base prompt.
  2. Paste the Inversion Engine block directly after it.
  3. Set depth, horizon, metrics, and INCLUDE toggles.
  4. Run once with assumptions, then tighten thresholds and re-run with real baselines.
  5. Execute the experiments and honor the kill switch rules.
  6. Log the decision, owners, and review date in your decision register.

Conclusion

The Inversion Engine add-on helps you escape confirmation bias, price in reversibility, and make clearer, safer choices. By simulating a concrete opposite, tagging Goodhart risks, and enforcing option value and governance, you reduce regret while keeping speed. Use the Concise variant for quick challenges, or the Master Prompt when the stakes are high and the decision will be audited later.

Applied Example: Project Lumen

  • Scenario: considering a self-serve price increase versus a targeted discount program.
  • Inversion move: model a price cut for a narrow segment with strict caps and a 30 day exit rule.
  • Findings: opposite plan reduces support load variance, improves trial to paid in two cohorts, but raises risk of quality perception decay.
  • Decision: test both with paired metrics, NPS and refund rate, and a kill switch if refund rate exceeds 2.0 percent for 3 days.

Leave a Reply

Discover more from Digital Thought Disruption

Subscribe now to keep reading and get access to the full archive.

Continue reading