Open Decisions

You break a crowd with a machine you assembled, find something that changes the next expedition, and bring a person or a question home. The remaining design decisions should make that journey more compelling. A technically consistent rule is useful only if it supports something worth doing, noticing or caring about.

Start with The Game or the whole campaign. This guide carries forward the outside review of 14 September, including its concern that controlled systems could crowd out discovery, ridiculous builds and human feeling. Its assessments are criticisms to examine, not owner decisions. Earlier wording remains in this page’s revision history.

What can be judged now

The current rescue comparison is Rescue Under Pressure. Yeva chooses her moment to cross a short, cleared window in a horde fight. If the astronaut dies after she starts, protected retreat leaves her sheltered but unsent: the player must find her after the shift and prepare another attempt. Protected completion brings her aboard before the shift, removes that retry and gives the revival a human homecoming. Both suspend hostile interference during their death safeguard; neither is an ordinary unprotected escort. Completion simplifies the rescue but can reward dying to finish it. Retreat gives failure a lasting inconvenience but risks repeating preparation instead of deepening attachment. The recommendation is completion; the existing owner question remains unanswered.

Other prepared comparisons can be read without selecting them: different build controls, two core mysteries, the final station connection and 40 versus 36 rooms. These are separate choices, not a package of approvals.

The Owner Brief establishes discovery first, expressive builds second and meaningful massed combat. New expeditions renew hordes; regular fighting grows temporary power, with larger transformations, and every return home resets that power. Carried property returns even through death. Permanent work and selected connections remain; death shifts unjoined areas. The journal fully pauses, ordinary fire is mostly automatic, specials are manual, and cooperative rescues must stay short. The opening direction is one part used in both rig and station. The hospital story remains revisable. Exact mechanisms, quantities, first survivor, corruption, ending and game size remain proposals.

Horde and build play

Make the next trip worth fighting

Fighting the Horde now places renewed crowds along a familiar route, including a version whose local sources are already sealed. Permanent victories remain useful. The remaining problem is repeated travel: fighting the same warm-up for a new clue can become a toll. Those worked routes do not cover every shifted arrangement or combination of closures.

The proposed growth rhythm awards a mark when roughly half an announced crowd band falls; three marks offer an ordinary upgrade. A major transformation arrives during a larger fight. This avoids hunting the last enemy, but the unpaid remainder may feel like waste and frequent journal choices may interrupt movement. Per-body progress is a credible alternative, with cleanup and farming risks. Early sealing currently buys safety, air and quiet while forgoing the prevented bands’ growth; any additional reward remains design work.

Two Trips, Two Different Builds gives the same equipment different choices: longer electrical jumps sacrifice contacts, while a disc-and-arc hybrid combines two learned behaviors. Its four-offer pool can run out. A curated path chosen aboard is cheaper and easier to teach, but loses decisions made after seeing today’s crowd. The team still needs a broader range of worthwhile choices and different expedition goals, not just another completed example.

Let the machine become ridiculous

The twelve-part catalogue still makes all three developed climaxes store charge and release it manually. Different geometry helps, but the review’s concern survives: limited coverage and shared controls can make three machines feel like one with different effects.

Build Comparison proposes replacing Ricochet’s stored master disc with deliberately triggered disc endings, and Gravity’s battery with a movement rhythm that leaves temporary embers. Neither replacement governs the detailed equipment rules yet. They buy distinct decisions at the cost of new exceptions, feedback and balance work.

The review’s more radical alternatives also remain available: a much larger, roughly forty-part wiring toybox, including feedback loops, with physical overheating, misfires or self-harm instead of strict conservation; or twelve parts with stranger behaviors, such as a movement-only machine or a build that burns valuable loot. The first invites surprises and difficult exploits; the second remains a small, partly predesigned toybox. Replacing wiring with grid packing or fixed weapon evolutions would simplify it, but weaken the distinctive connection between combat equipment and station machinery. None is adopted here.

Within the selected opening direction, the current proposal makes Fork personally useful: replace manual Barrier timing with sensor-triggered attack and defence, then decide whether to lend that same part to the station. It adds control, not force or extra capacity, and may still feel modest. The body-lift opening remains the stranger physical alternative, with an extra landing and transport work. Later equipment acquisition separately compares finding working objects with fabricating observed patterns: tangible surprise and irregular access versus targeting, material costs and possible shopping-list exploration.

Vary pressure without erasing victory

The moving scan now has a location, rising crowd, build climax and optical discovery that opens a bridge. It costs a major mechanism and substantial geometry. The enemy guide explains nine roles, but neither creature count nor ten encounter placements establishes enough variety for the campaign.

The review proposed making ordinary travel a regional flood, keeping finite stages mainly for setpieces. That makes pressure less like a wave counter but complicates earned quiet. Alternatively, retain finite stages and greatly expand their variety—chases, oxygen sieges, moving work and enterable wards. That costs content. More particular former patients could humanise the horde, but records, recognition and repeat-visit states cost more than names alone.

Discovery and people

A hospital worth remembering

The clinical room descriptions now distinguish a refuge ward, pressure approach, treatment-comparison bay and sampling loop through care, danger and useful actions. That scoped improvement does not settle whether the station as a whole feels like a hospital rather than machinery with hospital labels.

The review’s broader option is still substantial: rebase more rooms around wards, theatre, mortuary, pharmacy and everyday staff life; combine rooms whose only purpose is a socket; give machinery a human use and include views beyond the hull. Distinct regional materials, scale and light could help recognition. Threshold wear could make shifting history visible. These require spatial and visual design; evocative prose does not make the assets or fit the mechanisms.

Nadi’s encounter now connects one ordinary Presser’s paired marks to a patient record and Ivo’s memory. A powerful build can defeat that body normally. The cheaper record-only version keeps her history but loses recognition inside the moving crowd. Neither proves consciousness remains or makes enjoying combat a wrong choice.

People beyond their useful projects

The cast now gives Mara an argument about preserving dangerous material and Ivo control over his own movement. Yet five people arranged as rescue, project and changed belief can still feel predictable. The review’s smaller alternative folds roles into three or four people, varies how many projects each has, and gives them wants no machine can satisfy, stubborn beliefs and disagreements that can cost goodwill. Keeping five preserves more rescues but needs more distinctive writing and relationships.

Yeva and Aster remain alternative first rescues: Yeva brings the memory of people lost; Aster brings distrust of the arrival and a costly request to close a route. Neither is selected. Home still needs a settled management scope and post-rescue mortality policy. No management economy and permanent safety aboard remain credible simpler options, with fewer later pressures or losses.

Short transfer time alone does not make rescue satisfying. The review’s simpler preparation lets the survivor judge the canister and retreat to their own door, saving the refuge checklist but accepting less mechanical certainty. A refused or already failed rescue would cost a person aboard. Allowing death would change the survivor stakes further. These remain alternatives, not hidden additions to the current two-rule comparison. The survivor’s judgment and the relief at home must carry the scene.

Give the astronaut and the ending a personal stake

The astronaut now speaks in Nadi’s exchange, but a pre-docking wish, history and personal identity remain thin. Defining those could shape existing conversations without adding a dialogue system. Resurrection and Corruption compares safe ordinary revival, voluntary removable treatment and accumulating marks. The review also proposed permanent voluntary treatment. Involuntary marks make the threat personal but spend bodily agency; a cap or slow rate does not by itself prevent deliberate-death farming. Removable treatment preserves choice but can keep the horror external.

The core alternatives present living tissue behind isolation or an empty template engine. The team prefers the captive; the colder engine loses that image and risks a settings-menu reveal. A reversible isolation control can still feel emotionally cheap even when its evidence is sound.

The review proposed a priced intervention—ending revival for the walk home or a survivor leaving over the choice—or stronger visible and conversational consequences while retaining reversibility. Ending revival would change the confirmed return rule; a survivor’s departure also needs an unselected policy. Both need an owner decision. Reversible consequences cost meaningful reaction states. The story map already distinguishes lost live observations from retained captures, but that is not necessarily a climax the player cares about. Voluntary irreversible undocking is also still a proposed apparent ending. The team must make each competing history coherent before asking its clues to carry that weight.

Readable information

Language and Journal now explains capture, comparison and testing through two actual language uses. The language map groups four registers into two languages; technical stamps do not count as another language. Survivors can explain workplace use without translating alien text.

The remaining criticism is depth: practical pattern recognition may never feel like understanding meaningful speech. The addressed reply is an interaction, but does not establish its speaker or precise meaning. Distinct visual writing systems and one consequential, intelligible sentence in each language remain credible developments. A slower physical alternative could preserve progress for a non-reader. These need actual language and evidence authoring; replacing the second script with spatial patterns would still need to satisfy the two-language requirement. A smaller two-chain version saves registers but loses wall-access interpretation and the act of replying.

The information hierarchy puts urgent threats in play and detailed diagnostics in inspection. It can still overwhelm the spectacle. The review’s call to remove repetitive caveats survives; hiding uncertainty must not turn a guess into a guarantee. More physical tank and pack interactions could reduce arithmetic, but instant canister swaps would change the current live-transfer rule. Scaled weak shots or overheating consequences instead of refusals would likewise change equipment behavior. The journal’s alternative—personal notes without structured test outcomes—reduces clerical presentation but makes failed experiments harder to diagnose.

Practical scope

Joining the station should turn disorientation into ownership. The rigid-group proposal has not demonstrated a valid changed arrangement from every reachable state, or a complete forty-room network. Its fallback to unchanged geography conflicts with the every-death-shift requirement. A room-deck network could simplify arrangement while losing a consistent physical floor plan. A requested living shift remains unselected; repeatedly dying for one useful neighbour is still a failure risk.

Completing the Station compares securing a dependable home with taking the whole station away. Stability concludes the geographical puzzle but can leave errands; travel adds scale and an ending at significant new production and story cost. Neither makes full assembly a prerequisite for the core investigation.

Campaign Scope Choices compares forty rooms with thirty-six denser rooms. The smaller version saves layouts but retains the headline systems and loses separate observation and recovery spaces. Neither size establishes the desired 30/90/100–200 hours. The review’s lower-duration ambition remains an alternative; replay and a larger build pool can extend play but may pull effort away from authored mystery. Later measured experience would be stronger evidence than dividing hours by objectives.

Production Scope also retains more aggressive room or staging cuts with unallocated losses. Its historical 350–590 role-hour documentation allowance predates the fourteenth mechanism and is no current production budget. Development and validation require a separately authorised scope; this guide starts no prototype, playtest or asset work.

Open questions

The existing rescue question, build-priority choice and optional experience reactions remain unanswered. Prepared story, corruption, cast, acquisition, completion and scope alternatives remain available; they do not require simultaneous answers before useful team work continues.

The team still owes varied expeditions, readable supporting pages, coherent evidence and consequences, and credible geometry and scope. A few reference repairs remain: the captive story gives the builders an abandonment motive where the story map leaves it unknown; the regional language note still calls a register relationship unsettled despite the dedicated map’s classification; and the catalogue’s warning about a self-starting Projector is stale against the current controller-led progression examples. The historical shift-options link also needs its original revision identified. These are alignment tasks, not reasons to invent owner answers.

The proposals remain untested. Written examples and completed editorial passes do not establish fun, balance, accessibility, feasibility or campaign duration.