/magi

Business Wargaming Simulation Reference — Magi

Purpose: Use this file when Magi must stress-test strategies through competitive wargaming simulation, build scenario trees from competitor responses, or integrate wargaming findings into strategic roadmaps.

Contents

• Business wargaming in strategic context
• Magi's wargaming role vs Compete's role
• Wargaming simulation patterns
• Scenario tree integration
• Financial impact modeling
• Strategic stress-test protocol
• Templates

Business Wargaming in Strategic Context

Why Strategic Wargaming Matters

Traditional scenario planning asks "What if the market changes?" Wargaming asks "What if we act, and competitors react?" This transforms passive forecasting into active strategy stress-testing.

Key insight from military-to-business adaptation: AI-supported wargaming outperforms traditional methods by an average of 12.8% in decision-making accuracy, with the most significant improvements in complex multi-actor scenarios.

2025-2026 Generative Wargaming State-of-the-Art

• Johns Hopkins APL — GenWar / GenWar Sim (announced 2025-03-03, dedicated GenWar Lab established 2025-11): integrates LLMs into the Advanced Framework for Simulation, Integration and Modeling (AFSIM) so new scenarios stand up in under two weeks vs months previously. Designed to "build wargames in days, analyze dozens of alternative futures at scale, and focus human attention on the scenarios that most demand thoughtful deliberation." (JHU APL 2025-03-03, Military Times 2025-11-24)
• CSIS Futures Lab — "It Is Time to Democratize Wargaming Using Generative AI": argues GenAI shifts the cost of running wargames from honoraria + travel + facilitator time to data curation, enabling many more analytical games per quarter; companion executive-education course Building Better Strategies with AI, Red Teaming, and Gaming extends the methodology to enterprise risk management, corporate governance, and strategic planning.
• RAND continues to publish on AI for wargaming and modeling (Artificial Intelligence for Wargaming and Modeling, EP68860), with a 2025 focus on guardrails for AI red-team agents to avoid escalation bias.
• Open research line — Scaling Intelligent Agents in Combat Simulations for Wargaming (arXiv:2402.06694) — provides the LLM-agent-as-Blue/Red-player baseline most enterprise adaptations build on.

For Magi, the practical takeaway: when Compete supplies competitor profiles, an LLM-agent loop can now generate the 3-5 most plausible response sequences per move at near-zero marginal cost. Use the financial scenario tree in WG-1/WG-2/WG-3 to translate those branches into $ impact; do not trust the LLM-generated probabilities — assign them yourself or via Magi-style multi-perspective scoring.

Magi's Role vs Compete's Role

Magi consumes Compete's wargaming outputs and translates them into financial scenarios and strategic roadmap adjustments.

Wargaming Simulation Patterns

Pattern WG-1: Response-Adjusted Scenario Simulation

Extend Magi's standard 3-scenario model with competitor response probabilities.

For each strategic move under consideration:

Scenario A: No effective competitor response (P = Compete estimate)
  Revenue_A = Base forecast × (1 + move_uplift)
  Cost_A = Base cost + move_investment

Scenario B: Expected competitor response (P = Compete estimate)
  Revenue_B = Base forecast × (1 + move_uplift - response_impact)
  Cost_B = Base cost + move_investment + counter_response_cost

Scenario C: Aggressive competitor escalation (P = Compete estimate)
  Revenue_C = Base forecast × (1 + move_uplift - escalation_impact)
  Cost_C = Base cost + move_investment + escalation_defense_cost

Expected value = Σ (P_i × (Revenue_i - Cost_i))

Pattern WG-2: Multi-Move Strategy Simulation

For strategies requiring sequential moves:

## Multi-Move Simulation

### Move Sequence
| Move # | Our action | Investment | Expected competitor response | Response probability |
|---|---|---|---|---|
| 1 | [action] | $[X] | [from Compete wargame] | [X]% |
| 2 | [action, conditional on Move 1 outcome] | $[X] | [from Compete wargame] | [X]% |
| 3 | [action, conditional on Move 2 outcome] | $[X] | [from Compete wargame] | [X]% |

### Cumulative Financial Projection
| Path | Probability | Cumulative investment | Cumulative revenue impact | Net value |
|---|---|---|---|---|
| Best path (all moves succeed) | [X]% | $[X] | $[X] | $[X] |
| Expected path | [X]% | $[X] | $[X] | $[X] |
| Worst path (all moves face escalation) | [X]% | $[X] | $[X] | $[X] |

### Decision Rule
- Proceed if expected path NPV > 0 AND worst path is survivable
- Pause if expected path NPV > 0 BUT worst path threatens viability
- Abandon if expected path NPV < 0

Pattern WG-3: Competitive Equilibrium Simulation

For markets approaching equilibrium or price wars:

## Competitive Equilibrium Simulation

### Market State
- Total market: $[X]
- Number of active competitors: [X]
- Current price level: $[X]
- Margin structure: [X]%

### Price War Simulation
| Round | Our price | Competitor A | Competitor B | Our margin | Market share |
|---|---|---|---|---|---|
| Current | $[X] | $[X] | $[X] | [X]% | [X]% |
| Round 1 (our move) | $[X] | $[X] (predicted) | $[X] (predicted) | [X]% | [X]% |
| Round 2 (responses) | $[X] | $[X] | $[X] | [X]% | [X]% |
| Equilibrium | $[X] | $[X] | $[X] | [X]% | [X]% |

### Equilibrium Assessment
- New equilibrium price: $[X] (vs current $[X])
- Margin erosion: [X]pp
- Market share change: [X]pp
- Is this equilibrium sustainable? [Yes/No — why]
- Who benefits most from the new equilibrium? [Company]

Scenario Tree Integration

Converting Compete Wargame to Magi Scenarios

When receiving a scenario tree from Compete's wargaming, Magi assigns financial values to each branch:

## Financial Scenario Tree

Our Move: [Action] — Investment: $[X]

### Branch A: No response (P=[X]%)
- Revenue impact: +$[X] / year
- Duration: [X] years
- NPV: $[X]
- Risk-adjusted value: $[X] (= NPV × P)

### Branch B: Price match (P=[X]%)
- Revenue impact: +$[X] / year (reduced from A due to price pressure)
- Margin impact: -[X]pp
- Counter-cost: $[X]
- NPV: $[X]
- Risk-adjusted value: $[X]

### Branch C: Escalation (P=[X]%)
- Revenue impact: -$[X] / year (net negative)
- Defense cost: $[X]
- NPV: $[X]
- Risk-adjusted value: $[X]

### Portfolio Expected Value
Σ Risk-adjusted values = $[X]

### Decision
- [PROCEED / PAUSE / ABANDON] based on portfolio expected value and worst-case survivability

Financial Impact Modeling

Wargame Impact on KPIs

## Wargame Financial Impact Summary

### KPI Impact Matrix
| KPI | No response | Expected response | Escalation | Weighted average |
|---|---|---|---|---|
| Revenue growth | +[X]% | +[X]% | -[X]% | [X]% |
| Gross margin | [X]% | [X]% | [X]% | [X]% |
| CAC | $[X] | $[X] | $[X] | $[X] |
| Churn rate | [X]% | [X]% | [X]% | [X]% |
| Rule of 40 | [X]% | [X]% | [X]% | [X]% |
| Burn Multiple | [X]x | [X]x | [X]x | [X]x |

### Sensitivity Analysis
| Variable | ±10% impact on expected value | Most sensitive to |
|---|---|---|
| Competitor response probability | $[X] | [which probability assumption matters most] |
| Price elasticity | $[X] | [how price changes affect volume] |
| Response timing | $[X] | [fast vs slow competitor response] |
| Market growth rate | $[X] | [background market expansion] |

Strategic Stress-Test Protocol

WARGAME Mode Execution

When Magi is invoked in WARGAME mode:

1. RECEIVE: Compete wargaming output (competitor response predictions)
2. QUANTIFY: Assign financial values to each scenario branch
3. SIMULATE: Run 3-scenario model for each branch
4. STRESS-TEST: Apply sensitivity analysis to key assumptions
5. SYNTHESIZE: Produce decision-ready recommendation
6. MONITOR: Define early warning signals and deviation thresholds

Integration with FORESIGHT

After wargame-informed strategies are deployed:

• Track which scenario branch is materializing
• Compare predicted competitor response vs actual
• Recalibrate response probabilities for future wargames
• Feed accuracy data back to Compete for calibration (MAGI_TO_COMPETE_CALIBRATION)

Escalation Rules

Integration with Magi Workflow

Wargaming simulation integrates into the VERIFY phase:

• After standard scenario simulation, overlay competitor response scenarios
• Use wargaming to stress-test the strategic roadmap before PRESENT
• Include wargame findings in the Execution Roadmap section
• Set monitoring thresholds based on wargame scenario branches
• Feed results into FORESIGHT for post-deployment tracking