How it works — 1 of 7

World & ticks

Observable lifecycle

Inputs → computed by → persisted as

Authorized world view and actor proposals.

Procedures, seeded roll, effects and guardrails.

Trace, snapshot, summary and optional next tick.

The Global Simulator advances one tick at a time. Every tick, each participating institution's decision-maker looks at what has happened so far, may act, and the platform turns those actions into a small, consistent shift in the state of the world. The completed tick is observable and the run can be paused or stopped; while it remains active, the platform may queue the next tick automatically.

A tick opens with the actors. Every decision-maker taking part in the run — typically the heads of a handful of key institutions per country, such as foreign affairs, defense, finance or the executive — reads the parts of the world it is allowed to see and can take up to a couple of actions, each carrying a stated rationale. Nothing is decided yet: an action is a proposal, not an outcome.

What happens to a proposed action is the job of adjudication, covered on its own page. In short: the model argues the case, and a transparent rule — not the model — turns that argument into a result and a fixed set of deltas.

Those deltas do not land on the world state directly. They pass through guardrails that clamp anything that looks like an outlier before it is recorded, then into the substrate and network layer: a set of difference equations that carry trade, tariff, currency and energy stocks forward tick by tick. One honest caveat lives here: trade rerouted by sanctions is drawn from a calibrated table of plausible substitution, not computed by re-solving the trade network from scratch — a deliberate simplification, not a hidden one.

The same tick also steps the population layer: a set of cohort archetypes per country carry grievance and loyalty forward, and if enough of them cross a threshold together the cascade feeds an institutional gate that can constrain what a government can still do. Two indicators — stability and military readiness — are pulled gently back toward their baseline every tick by a dedicated decay step. The rest are left out of that step on purpose, not by oversight, each for its own reason: the economy indicator is rebuilt fresh from the underlying stocks every tick, and public pressure already carries a decay term of its own inside the population layer above — running the shared step on either would correct the same drift twice.

The tick closes with a summary: the numbers are turned into narrative news for the people following the run, the whole state is snapshotted, and — as long as the run is still active and its tick budget is not exhausted — the next tick is queued. The operator does not have to click through every tick, but retains control over the run itself.

flowchart TD
    accTitle: One simulation tick
    accDescr: Actors propose actions, adjudication resolves them, guardrails and numeric layers update the world, and the platform stores a summary and snapshot before optionally queueing the next tick.
    A[Tick begins] --> B[Actors act: each participating decision-maker may take 0 to 2 actions]
    B --> C[Adjudication: the platform resolves each action]
    C --> D[Guardrails clamp any outlier deltas]
    D --> E[Substrate and network: trade, tariff, currency and energy stocks evolve]
    E --> F[Population: cohort grievance and loyalty shift, feeding an institutional gate]
    F --> G[Decay: two accumulating indicators pull back toward baseline]
    G --> H[Tick summary: narrative news, snapshot, memory update]
    H -. If the run remains active, queue the next tick .-> A