A crisis management simulation where you govern a city with interconnected infrastructure, competing interests, and no easy answers.
Set in the fictional city of Ankh-Morpork, the simulation models how disruptions in digital and operational systems translate into real-world consequences for cities, organisations, and the people who depend on them, without relying on technical detail or domain-specific jargon.
The setting is fictional. The dynamics are not.
The simulation is designed to support conversations and workshops around:
-
The impact of digital and operational disruptions on:
- public services
- economic activity
- trust and legitimacy
- regulatory and political pressure
-
Decision-making under uncertainty, time pressure, and budget constraints
-
The cascading effects of infrastructure failures across sectors
-
The relationship between operational choices and societal outcomes
Rather than focusing on how systems are built, the simulation focuses on what happens when they fail, degrade, or are disrupted.
This project is intended for use with:
- Executive leadership
- Board members
- Policy makers
- Regulators
- Public-sector decision-makers
- Mixed technical / non-technical groups in joint exercises
No prior technical or cybersecurity knowledge is required to participate.
Players take responsibility for a city with interconnected services, including:
- Energy
- Water and sanitation
- Communications
- Transport
- Public and private facilities
- Residential areas
Each building or district represents a concentration of services and dependencies rather than a single system.
Over time, the city experiences incidents, disruptions, and stressors inspired by real-world OT, ICS, and digital infrastructure risks. These events affect:
- service availability
- economic output
- public sentiment
- institutional trust
- regulatory scrutiny
Players respond by allocating resources, prioritising interventions, and managing trade-offs between short-term stability and long-term resilience.
- It is not a technical training tool
- It does not simulate networks, protocols, or exploits
- It does not reward detailed technical knowledge
- It is not a competitive or leaderboard-based game
The emphasis is on outcomes, consequences, and choices, not on technical mechanisms.
Using a fictional city allows participants to:
- Engage seriously without referencing real-world sensitive locations
- Speak openly about failure, pressure, and trade-offs
- Recognise familiar patterns without defensive reactions
The city's exaggerated complexity and personality make interdependencies visible and memorable, while keeping discussions grounded in reality.
Requires Python 3.12 or later.
# Clone the repository
git clone git@github.com:ninabarzh/ankh-crisis-sim.git
cd ankh-crisis-sim
# Create and activate a virtual environment
python3 -m venv .venv
source .venv/bin/activate
# Install dependencies
pip install -e ".[dev]"Dependencies are minimal: customtkinter, Pillow, and pyyaml.
# Run the simulation
python -m src.main
# Run the test suite
pytest tests/ -vOn launch you see a brief introduction screen. Pressing Start enters the main game: a city map with building indicators (green/yellow/red), a metrics dashboard, a news ticker, and time controls. The simulation starts paused.
- Play / Pause: control the flow of simulated time; the active button is highlighted (green = playing, dark = paused)
- Slower / Faster: change simulation speed (1× or 10×)
- Hover over a building lamp to see its name, district, and status
- Click a building with an active event to choose a remedy
- Click a headline in the news ticker to read the full story
- Watch the news ticker for headlines as events unfold
- Monitor the dashboard for trust, budget, regulatory pressure, political stability, and legitimacy
- ⚙ (gear icon) in the dashboard title bar opens game settings
The game ends when a loss condition triggers (trust collapse, revolt, bankruptcy, infrastructure failure, assassination) or when you survive your full term. There is no victory screen, only "You served your term. Here is what happened."
The gear button opens a settings panel that lets you adjust difficulty before or during play:
| Setting | Options | Effect |
|---|---|---|
| Game duration | Standard, Short, Extended | Term length in simulated years |
| Event frequency | Normal, Rare, Frequent | Multiplier on incident probability |
| Discovery speed | Normal, Fast, Slow | How quickly hidden failures surface |
| Cascade risk | Normal, Low, High | Multiplier on failure propagation probability |
All settings are applied immediately and persist for the session.
The entire simulation is data-driven. All game content lives in YAML files under config/. No Python changes are needed for balancing, adding content, or creating scenarios.
One file per district in config/districts/. Each defines wealth, density, infrastructure quality, political influence, media attention, local trust, stressors, and narrative hooks.
# config/districts/the_shades.yml
id: the_shades
name: "The Shades"
wealth: 12
density: 300
infrastructure_quality: 3.0 # 3x more likely to fail
political_influence: 0.2
media_attention_multiplier: 0.2
local_trust: 15
discovery_time_hours: [72, 120]infrastructure_quality is a failure probability modifier: 1.0 = baseline, 3.0 = three times as likely to fail, 0.3 = 30% as likely.
To playtest with fewer districts, comment them out in config/game.yml:
active_districts:
- small_gods
# - nap_hill # disabled for testing
# - the_shadesBuilding types are defined in config/buildings/_types.yml (shared behaviour: what they consume, produce, dependency strengths, detection speed). Concrete buildings are placed in config/buildings/instances.yml:
- id: mended_drum
name: "The Mended Drum"
type: tavern
district: merchant_quarter
position: [460, 280]
status: operational
narrative_hooks:
- "Mended Drum brawl spills into street, breaks new fountain"Add a building by appending to instances.yml. Add a new type by adding a key to _types.yml.
Current building types: guild_hq, tavern, civic_amenity, slum_dwelling, middle_class_housing, civic_institution, palace, water_source, power_source, clacks_tower, food_supply, brewery, transport, communications_office, apothecary, healthcare, security, intelligence_service, tech_business, hackerspace, workshop, fire_service, media, university.
Threat categories live in config/threats/categories.yml. Systemic stressors (underinvestment, inequality, vendor monoculture, etc.) in config/threats/stressors.yml. Concrete event templates in config/threats/events.yml:
- id: pump_failure
name: "Water pump failure"
category: degradation_and_neglect
target_building_types: [slum_dwelling, middle_class_housing]
probability_base: 0.003
stressor_amplifiers:
underinvestment: 2.0
impact:
immediate:
- metric: local_trust
delta: -5
scope: district
headlines:
- "Water pump fails in {district}"All seven remedy types are defined in config/remedies.yml with base costs, downtime, recurrence risk, trust effects, and narrative framing. Adjust a number, change the game.
Discovery times per district and building type, detection mechanisms (Watch patrols, citizen complaints, Vimes intuition, media investigation), and player investments to improve detection are in config/detection.yml.
config/narratives/headlines.yml holds simple one-line templates (always visible). config/narratives/stories.yml holds richer expandable stories with tone markers. Both support {district}, {building}, {duration}, and other template variables.
Loss conditions, term completion, and post-game reflection questions are in config/end_conditions.yml.
| I want to... | Edit this file |
|---|---|
| Tweak a district's wealth or trust | config/districts/<name>.yml |
| Add a new building | config/buildings/instances.yml |
| Change remedy costs | config/remedies.yml |
| Add a new event type | config/threats/events.yml |
| Adjust how fast failures are noticed | config/detection.yml |
| Change starting budget or trust | config/game.yml |
| Write new headlines | config/narratives/headlines.yml |
| Change loss thresholds | config/end_conditions.yml |
| Playtest with fewer districts | config/game.yml → active_districts |
- Python 3.12
- CustomTkinter + Pillow (GUI)
- PyYAML (config loading)
- Python dataclasses (typed models, no Pydantic required)
- Engine fully decoupled from GUI and testable without a display
Under active development. The simulation engine, YAML config system, and GUI are functional. Current focus is on balancing, playtesting, and expanding event variety.
This project is licensed under the Polyform Noncommercial Licence.
You are welcome to use this software for:
- Learning and personal experimentation
- Academic research, coursework, and dissertations
- Classroom and curriculum use by educators
- Non-commercial workshops, civic exercises, and community events
- Game design research and serious games scholarship
Additional permissions for academic and educational research are set out in RESEARCH-EXCEPTION.md.
You may not use this software for:
- Paid workshops or training courses
- Commercial product development or integration
- Any context where money changes hands
A commercial licence is required to use this project in paid or commercial contexts. All proceeds from commercial licences, beyond the costs of maintaining this project, will be donated to Alzheimer's organisations in memory of Terry Pratchett.
This project is actively developed and maintained. The licence ensures that:
- The simulation remains freely accessible for education and research
- Commercial use contributes to something worthwhile
- Exploitation without contribution is not an option
If you are unsure whether your use case is commercial, ask. Ambiguity is solvable; silence is not.