Hold The Horde — Death and Persistence

Status. A provisional comparison for review, not owner-approved. The Owner Brief requires that death returns the astronaut to the ship for a reason hidden in the station; that cause is not addressed here. Damage and oxygen depletion both kill — that is fixed; oxygen quantities are not. Map structure, rescue logistics and colony systems remain undecided. Every claim below about how these rules would shape play is a prediction, not an observation.

Recommended rules: recoverable cache, retained record

• Equipment and salvage carried at the moment of death remain as a cache at the place of death, contents unchanged.
• A cache persists until recovered. A later death does not destroy an earlier one; caches coexist, and entity resets do not disturb them, a cache being the player's property rather than station state.
• Journal records and discovered-location references persist through death.
• Colony members already sent to the ship are unaffected by the astronaut's death.
• A person rescued but not yet sent to the ship is not retroactively rescued by revival and remains an unresolved situation in the station. Whether such a person can be permanently lost is a separate rescue and colony decision, not settled here in either direction.
• World changes are limited to location continuity. Enemies, traps and mechanisms may reset on their own terms.

What death threatens, stated plainly. Nothing destroys a cache. Dying costs the use of everything carried until a journey retrieves it, leaves new salvage unbanked, and may demand capabilities no longer to hand. That is the whole stake; naming it is better than implying one the rules do not create. Decay or fragility would turn a recovery choice into a deadline and push toward the compulsory errand this arrangement exists to avoid, so none is proposed.

State table

The next choice after each outcome

Returning alive asks what to do with what was banked: install it, hold it, or re-outfit for a deeper objective. Dying asks whether to retrieve the cache now, defer it, or write it off and spend the retained knowledge elsewhere. Where death interrupted a rescue, revisiting that person competes with retrieving the cache and with any other objective; which pull is stronger depends on the situation and is not fixed by these rules.

The alternative

Expedition rollback. Equipment held before departure returns with the astronaut; equipment and salvage acquired during the failed expedition are lost; records persist; rollback is limited to the objective's state rather than the whole station.

Rollback has real strengths. It protects an assembled build from confiscation, which serves expressive play directly, and it produces no gear-recovery errands and no compounding, because the player resumes with the capability they left home with.

Why the cache is still recommended. Under rollback, whatever was carried from the ship comes back intact, so the loadout decision carries no consequence at death and bringing more equipment is free. The brief's required tradeoff is equipment against salvage in one pack; rollback protects one side of it outright, while the cache makes that same decision the thing death acts on.

Deliberate death is not a point of difference. Under either arrangement, a player who has recorded what they came for, whose pack holds little they need soon, and who is far from home on thin oxygen, may find dying preferable to walking back. Only the aftermath differs: rollback returns the gear and destroys new material, the cache keeps everything and charges a journey. Whether that behaviour is acceptable is untested and not settled here.

A third arrangement, returning everything carried including newly acquired salvage, is excluded: it banks new material for free, making deliberate death strictly better than returning.

Accepted costs

• Failure can compound. The expedition that killed the astronaut may be the one that must be repeated with less.
• Confiscation risk may suppress experimentation. An expressive build is a large investment of carried parts, and a rule that withholds parts discourages that investment. A predicted effect.
• Retained records do not guarantee discovery. Retention has two distinct benefits and neither ranks above the other: it avoids recapturing evidence already understood, and it preserves evidence not yet read for a later inference. It resolves a dependency left open by Oxygen and Observation without making discovery riskless.

Recovery, and when it is not available

Recovery needs more than a place name: contents must persist as they were and the site stay locatable. Four situations are distinct and should not be collapsed.

• Available — the cache can be reached, carried and returned with the capability now aboard. Retrieval is a genuine choice against other objectives.
• Deferred — not reachable yet, but perhaps later as reach or capability grows. The cache waits; setting it aside costs nothing.
• Abandoned — the player judges the contents not worth the trip. This must remain credible.
• Compulsory — retrieval is possible but no reasonable alternative exists. This is the failure state, and the condition that would reopen the preference.

Dying of depletion does not by itself put a cache beyond reach. Depletion often reflects time spent searching, fighting or detouring rather than distance, and a direct informed route may comfortably reach what a wandering one could not; a lighter loadout can also carry more canisters. Where a cache genuinely cannot be reached, carried and returned within supply, recovery is deferred rather than forced.

Equipment-system dependency. The ship must be able to outfit a minimum useful expedition even when everything carried is cached, including a means of carrying oxygen. That is a requirement on the equipment system, not something established elsewhere: Build Comparison's basic-defence rule concerns two example builds holding the compressor head and does not establish an equipment-free floor.

What would reopen this preference

Recovery becoming compulsory rather than a genuine choice; replacements scarce enough that abandoning a cache is never credible; or later owner-authorised observation showing experimentation collapsing under confiscation risk. Rollback would then be the more honest arrangement.

Unresolved

Oxygen quantities; map generation and respawn rules; the full rescue and colony systems; whether an authored survivor can ever be permanently lost; whether deliberate death is an acceptable tactic; and the cause of revival, untouched.