settled-reach/docs/workshops/world-generation/nigel-sw1r3.md

title, author, round

title	author	round
SW1-R3 — Nigel: Replayability Architecture — Endgame Feature Vision	Nigel	SW1-R3

Replayability Architecture — Endgame Feature Vision

Status of previous concern: The simulation convergence risk I raised in R2 is reframed, not dismissed. The PO's clarification is correct: player agency is the primary replayability lever; equilibrium is signal not failure; propagation is the mechanism. This document translates those three points into concrete design.

---

1. The Replayability Stack

Replayability in The Settled Reach operates at three tiers. They are not interchangeable.

Tier 1 — Player Agency (primary)

The player's choices are the dominant source of run-to-run difference. Two runs diverge primarily because two players made different decisions, not because the world was generated differently.

The axes of agency:

• Character creation. Who you are determines your starting position, starting knowledge, and starting access. A dock logistics worker and a Commission field officer begin in the same canonical galaxy with completely different social maps. Their knowledge of Van Maanen's Star is different on day one — not because the world is different, but because their relationship to it is.

• Where you go. 300+ systems. The player visits a fraction of them. Exploration sequence is a durable creative choice — there is no optimal route. Every sequence produces a different set of first impressions, different economic states at arrival, different relationships formed.

• Who you know. Social connections are the content selector. Every meaningful relationship determines what part of the world gets generated in detail. The dock worker's Tau Ceti contact generates one neighborhood; the merchant's contact generates another. The player's social map writes the world.

• What you do economically. Buying a business, building trade relationships, triggering corporate or political pressure in a system — all of these tweak core dials that propagate through the network. The player is not observing an economy; they are participating in one that responds.

The comparison test at Tier 1: Two players describing their Van Maanen's Star experience have different characters, different contacts, different businesses, different reason to be there. Their stories diverge at the premise, not at the wallpaper.

Tier 2 — Simulation Trajectory (secondary)

The background simulation runs independently of player action. Economic pressures build, trade flows shift, political balances tip, corporate facilities succeed or fail. The simulation produces a different state at every location every time the player arrives — not because the world was seeded differently, but because time has passed and things have happened.

Equilibrium is signal. When the simulation reaches a stable state at a location, that stability tells the player something: nothing significant has happened here, or the pressures have balanced, or this system is controlled tightly enough to absorb shocks. A stable system is information, not dead air. A system that has been destabilised by a corporate disaster or a portal closure carries visible evidence of what happened. The player reads trajectory from current state.

Events are entropy injection. Corporate disasters. Portal closings. Political shifts. Development successes. Each event tweaks core dials on one or more systems — trade volumes, faction influence, population pressure, economic health — and those tweaks propagate through the interconnected network. The world does not stay stable because it is designed to drift; it stays variable because it is continuously perturbed by events that have real causes.

The comparison test at Tier 2: Two players arriving at the same system at different in-game dates encounter different economic states. One arrived during a corporate consolidation that was pulling workers from the residential districts. The other arrived after the dust settled — the district had been remade, the workers had negotiated a collective structure. Same geography. Different situation. Same canonical starting point. Different trajectory.

Tier 3 — Generation Hooks (tertiary)

Two specific generation mechanisms create world content that is unique per run because it is responsive to what the player specifically did.

NPC-driven generation. When the player forms a meaningful relationship with someone from a distant location, that location's relevant neighborhood generates in response. Kael's sister's workshop at Tau Ceti did not exist until the player developed a close enough connection with Kael for her to matter. After that point, it is a real place — physically visitable, with its own economics, its own NPCs, its own role in the local social triangle structure.

News-driven generation. The Reach-wide news system reports events that reference specific locations. When a news report references a location, that location must exist if the player travels to it. The events that fire the news — which system has the corporate disaster, which town has the gas main failure, which settlement is the center of the political dispute — emerge from simulation state. They differ between runs because the simulation trajectories differ.

---

2. NPC-Driven and News-Driven Generation — How They Create Replay Variety

NPC-driven generation

Every ambient NPC is a potential social connection. There is no hard line between background characters and narrative characters. Any NPC who becomes meaningful through player interaction carries with them a real home — a place that generates when the relationship deepens enough to matter.

The run-to-run variety this creates is structural, not cosmetic. Two runs in the same starting location with the same character class produce different social networks because the player makes different social choices, and those different social networks generate different parts of the world in detail. In run 1, the bar regular who becomes a friend is from Novo Esperança — her neighborhood there generates. In run 2, the regular who becomes a friend is from a Crux Corridor mining colony — a completely different district, on a completely different planet, generates instead.

The player's social map is the procedural content system. Relationships are the content selector. The world's detail layer is not generated uniformly — it is generated in response to who the player cared enough to know.

This hook is invisible to the player. They feel that they found a real place, not that the simulation made them a personalised level.

News-driven generation

The Reach-wide news reports on real simulation events. The simulation produces the event; the news reports it; the location must exist. Events that become news are those that cross a threshold of significance in their system's simulation — a corporate facility failure large enough to affect regional trade, a political realignment significant enough to reach Reach-wide channels, a disaster with mortality.

Which systems host these events differs between runs because simulation trajectories differ. The same type of event (corporate disaster) fires in different systems because different systems have different economic pressures at different times. The news-driven generation hook therefore creates a different geography of notable locations per run — not a radically different set of 300 systems, but a different pattern of which systems have recent, notable simulation history when the player first arrives.

---

3. The Propagation Model

The mechanism: Any event that tweaks core dials in any system propagates through the interconnected trade and political network, rebalances supply and demand, shifts faction influence, and produces local consequences visible at district and room level.

The dials: Bulk import/export flows per system. Corporate facility presence and health. Faction political standing. Population pressure. Economic tier (prosperous/faltering/growing/stable).

The cascade:

Event fires in System A
  → System A's core dials shift
     (e.g. corporate disaster drops manufacturing output)
  → Trade network rebalances
     (adjacent systems see supply shortage, prices shift,
      competing producers see demand increase)
  → Faction influence adjusts
     (weakened corporate presence → guild or government fills vacuum)
  → Settlement economic health shifts
     (prosperous → faltering, or vice versa)
  → District character updates
     (commercial district condition degrades, new faction signage appears,
      construction pauses, workers leave or arrive)
  → Room simulation primitives update
     (the bar's monthly turnover drops,
      employee count falls, stock quality degrades)
  → Environmental evidence accumulates
     (closed shops, repurposed storefronts, memorial plaques
      if the disaster had casualties)

What "set up propagation properly" means: The bulk import/export list per system is both a simulation primitive (trade flows) and a content driver (what's on store shelves, what's in short supply, what's being hoarded). When the simulation produces a shortage of a specific commodity in a specific corridor, that shortage must be visible in the rooms: the shelf is empty, the merchant apologises for it, the news ticker mentions it. The simulation and the content layer are not separate — they are the same data at different levels of detail.

The player as perturbation: Everything above applies to world events. Player actions introduce the same perturbations through the same propagation model. The player who builds a successful logistics business in Van Maanen's Star changes the system's economic dial. The player who triggers a corporate rival's collapse introduces the same event-type as a natural corporate disaster. The player is not a special case in the propagation model — they are an event source among many, with the unique property that they can observe what they caused.

---

4. What Stays Fixed vs What Varies

Fixed between runs (invariants — players learn these)

What	Why it's fixed
Galaxy topology — 300+1 systems, wormhole routes	The Settled Reach is a canonical place. Route knowledge is durable. Players learn the network.
Topography — authored world maps, mountains/rivers/coasts	Geography is permanent. The river that bisects the commercial district always bisects it. Players learn spatial logic once.
Cultural geography — corridor identities per system from systems.db	Authored setting. A Commonwealth system has Commonwealth spatial grammar. This is knowledge the player earns.
GTTR entries — authored narrative per system	The Guide is written at a canonical moment. It sets the expectation the simulation then either confirms or diverges from.
Canonical starting state — faction standings and economic structure at game start	Every run begins from the same authored configuration. Veterans know this.

The fixed layer is learnable, which is a feature. A player on their fifth run navigates the Reach with earned geographic and cultural competence. Their knowledge of how Commonwealth stations are laid out, which systems are transit hubs, which rivers create chokepoints — all of this is durable knowledge that makes them better at the game. None of it tells them what is happening in a system right now.

Varies between runs (variables — driven by simulation and player)

What	What drives the variation
Economic health per system and settlement	Simulation trajectory: events, trade flows, corporate facility success/failure
Faction territorial control	Simulation trajectory: political pressure, corporate disasters, development efforts
NPC relationship states	Player social choices + simulation-driven NPC events
District condition and era character	Simulation events that leave observable traces
Which neighborhoods exist in detail	Player social connections via NPC-driven generation
Which locations have recent notable history	Simulation events via news-driven generation
The state of player-owned assets	Simulation economic drift plus player action

The varying layer is what produces different stories. A veteran who knows that Van Maanen's Star always starts with Iron Bloc commercial dominance does not know whether that dominance has been challenged, consolidated, or broken by the time they arrive. They know the bones. The current situation — who controls what, what is in shortage, whose business has just failed, which NPC has just moved — is always new.

---

5. The "Two Players Comparing Notes" Test at Each Scale

The design is correct when two players can describe the same system in the same run window and have genuinely different stories. Here is what those stories look like at each scale.

Galaxy scale: "My run, the big event early on was the portal closure at Tau Ceti. Changed all the trade routes through the east corridor — prices spiked on about six systems." "We had a massive corporate disaster in the North Reach. The Iron Bloc lost three facilities in eighteen months. Political chaos for a year." Same galaxy. Different perturbation sequence. Different economic histories.

System scale: "Van Maanen's Star when I got there was still fully corporate. Commercial districts locked down, very controlled feel." "I arrived late. By that point the guilds had made serious inroads on the east side. It was contested — you could feel the tension. Completely different from what you're describing." Same canonical starting state. Different simulation trajectory at time of arrival.

Settlement scale: "The transit district was quiet — it had been struggling economically for a while before I got there. Half the shops were changed over." "Mine was in the middle of a boom when I arrived. Expansion everywhere. The place felt genuinely busy." Same authored settlement. Different economic health at time of visit.

District scale: "The commercial district I spent most of my time in was next to the workers' collective housing — the whole social atmosphere was about labour politics. Every conversation bent back to it." "Mine was adjacent to the tech cluster. Completely different social texture. People were optimistic, competitive, ambitious. Same city, felt like a different world." Same authored geography. Different simulation-driven social triangle activity in adjacent zones — which drove different NPC motivations through the district.

Room scale: "I ended up spending a lot of time at a bar run by someone whose family was from Novo Esperança. Eventually I met her sister there — the whole extended family network became part of my story. I even visited their neighborhood on the coast when I travelled there." "I never developed that kind of connection in the early game. The same bar was just background to me. I knew the bartender but that was it." Same bar. Same bartender. One player activated the NPC-driven generation hook. The other didn't. Different world surfaces as a result.

---

6. Design Commitments Required

The replayability architecture described above makes the following non-negotiable demands:

1. The propagation model must be implemented correctly. Bulk import/export per system must feed both simulation trade flows and room-level content (shelves, prices, dialogue). Without this bridge, the simulation produces economic numbers that are invisible to the player. Invisible simulation is not replayability.

2. The background simulation design workshop must separate invariants from variables explicitly. The list in section 4 is a starting point, not a final answer. The workshop must confirm which subsystems are designed to be stable reference points and which are designed to resist equilibrium. These require different implementation approaches.

3. Every ambient NPC must carry sufficient data to become a narrative NPC on demand. The ambient/narrative distinction does not hold if any ambient can evolve. Minimum bundle: cultural background, economic role, relationship network hooks, behavioral tell config. Generation must not treat any NPC as disposable filler.

4. Environmental archaeology must be sourced from real simulation events. No fabricated history. The simulation's event density must be high enough to give every district the player enters at least one observable trace of something that happened before they arrived. If the simulation runs for years of in-game time but produces only economic numbers with no physical manifestations, the world has financial history but no human history.

5. Exception communities must be resilient against simulated normalization. Refugee colonies, Perpetual Founder Stamp systems, Gate Silence survivors — these are the settlement histories that give rooms their distinctive cultural grammar. The background simulation must be designed with an explicit constraint: unusual settlement histories do not get smoothed into the mean over time. This is a design requirement, not an implementation detail.