Hold the Flood | Two concept directions

DES101 | r1 | 2026-09-06 | Lead Game and Systems Designer | In review

Recommendation - proposal: Develop A: The Borrowed Route as the working direction. Learn environmental rules to turn an approaching enemy surge into a window for exploration. Knowledge changes when and how you enter danger.

Owner requirements: The Flood is massed enemies (Brief D011); survivorlike; exploration, experimentation and mystery; assume nothing exists; Unity; Steam. Design only until explicit owner authorization after document review. No AI-generated art. Prefer existing free assets; purchases and rare, minimal custom game assets need owner approval before purchase/creation.

Provenance: The six named inspirations and their aims come from the owner, not comparative research. All mechanics, settings and numbers below are proposals; examples are illustrations, not playtests. No game or asset collection is assumed. Brief D012 records owner-confirmed staffing; no other agent activity is inferred. Both assume solo expeditions and automatic basic attacks; camera, 2D/3D, controls and operating systems remain open.

A. The Borrowed Route

Fantasy and setting assumption: A salvage surveyor enters a derelict pumping district to learn why certain rooms were deliberately isolated. The Flood is a mass of hostile organisms with forecast migration surges. Shallow water is an environmental tool, not the Flood itself; enemy identity and origins remain proposals.

Actions and pressure: Choose a route, dodge swarms, gather salvage, rearrange equipment at raised shelters, operate machinery, and return to an elevated dock. Three announced enemy surges enter through different passages, making direct routes increasingly dangerous. A final evacuation ends the expedition; missing the dock means rescue without cargo. Health depletion has the same loss. Combat protects movement; repeated kills yield no farmable progression.

Rhythm and tradeoff: Scout between surges; commit to a live experiment; survive the announced approach; reach a shelter and reconsider. The next surge and its entrance are visible. Opening the notebook pauses the whole world and permits inspecting recorded observations, without harvesting or advancing experiments. Shelters stop attacks but the surge schedule continues outside pause. Thinking is free; obtaining new evidence spends exposure and route time. An early exit banks less salvage; investigating a side room may cost the interval before enemies occupy the return corridor.

Inventory, builds and gathering: A proposed 2-by-3 equipment rack also holds salvage: each bundle occupies one square. Powered modules must share an edge with a one-square battery. A two-square coil attacks across connected wet ground; one-square insulation protects its wearer from that current. A basic cutter remains available without rack space. The coil setup trades four squares for wet-area control and machinery access; a lightly equipped cutter carrier hauls more and prefers dry paths. Repacking requires a shelter. A salvage bundle can instead power a local pump until extraction, draining one channel for a dry crossing but forfeiting that colony contribution. Thus loadout, cargo and retreat compete directly.

Discovery and signature interaction - illustration: A valve joins two shallow channels; their lamps then blink together, and a cable leads to a sealed service hatch. Infer that connected water carries power. An approaching swarm blocks the direct archive corridor: open the valve and discharge the coil to stun enemies in the channel and power the rear hatch. Cross while they recover; without insulation, the same current harms you. Next visit, bring that setup to exploit the surge, or drain the channel and take a longer dry route with more cargo space. Enemy pressure makes the discovered shortcut worth its exposure.

Persistence and colony: Notes, mapped relationships, unlocked equipment recipes and banked salvage persist. Choose the next colony repair: workshop access adds an alternative starting module; a dock repair adds an extraction point. Limited salvage delays the other. Starting equipment is reissued; found modules, cargo, enemies, pumps and surge state reset. Colony decisions change expedition options; exclude citizen schedules, hunger and passive stat grinding.

Scope cuts: One compact district, a few connected water channels, discrete dry/wet states and reusable machines. No continuous fluid physics, terrain destruction, arbitrary chemistry, open world or colony construction grid. This is a content boundary, not a feasibility claim.

Sources: Hub, Brief, Tasks, Team, read 2026-09-06. No external comparative research claimed.

B. The Listening Below

Fantasy and setting assumption: A field investigator enters a buried civic archive occupied by massed enemies proposed to behave as one collective organism. This Flood follows sustained repeating signals. You learn to redirect its attention and use the swarm to reach sealed evidence; its origin remains the central unanswered question.

Actions and pressure: Scout cleared corridors, fight mobile swarms, inspect signal traces, operate resonators, extract specimens and return to the entrance. Powered investigations and harvesting add visible expedition attention; crossing marked thresholds releases a new swarm at a telegraphed entrance. Basic movement and short-impact combat do not add attention. Existing enemies keep acting; attention never decays through waiting or kills. No rewards from repeated kills. Extraction ends the expedition; health depletion loses carried finds.

Rhythm and tradeoff: Clear an observation position, read its traces without a clock, choose an intervention, then survive and reposition after its consequences. A sealed refuge permits unhurried notebook reading and equipment changes, but offers no harvesting or remote operation. Notebook pause freezes everything. Each operation previews its attention cost and whether it triggers a swarm. You choose how much danger to incur, but a threshold cannot be erased by resting. A specimen may fund the colony while using the remaining margin for a mystery experiment would teach a new route. Cleared spaces provide real observation time.

Builds, inventory and gathering: Use two equipment sockets, one combat and one investigation, plus four cargo spaces. A narrow cutter with a muffled sampler keeps extraction attention low, but cannot send the long-range signal needed to redirect distant Flood. A broad sweeper with a resonator handles crowds and reaches signal mechanisms, but each powered signal costs more attention. Bulky specimens take two cargo spaces; scrap takes one and can be consumed to place a remote signal beacon. Carrying a beacon supply sacrifices specimen capacity. Exclude a packing grid and adjacency bonuses: destination, signal reach and cargo commitment already express this concept's choices.

Discovery and signature interaction - illustration: Repeatedly pulsing panels draw tendrils only on the third beat; tracks converge around a pressure plate mechanically linked to a locked gate. Infer a sequence and a use: place a beacon on the plate and repeat three pulses. The pursuing mass moves onto it, opening the gate while occupied. Cross to inspect the chamber, knowing that the swarm stands on the return route and the operation may trigger another arrival. Later, bring beacon scrap and open the shortcut deliberately, instead of treating that mass solely as enemies to kill.

Persistence and colony: Notes, learned signal rules, unlocked equipment and banked finds persist; attention, local mechanisms, enemies and carried finds reset. Reissued starting tools prevent a loss spiral. Spend recovered specimens and scrap to restore either a listening post revealing the next expedition's arrival entrances or a workshop unlocking the muffled sampler. Both remain attainable, but the order changes preparation. Exclude colony needs, social simulation and production chains: the settlement supplies investigative options.

Scope cuts: One archive wing; a few authored pulse rules and marked arrival sites. No simulated acoustics, adaptive AI learning, procedural room shuffling or required sound-only clues. Pulse timing also has visible cues. No asset availability or production estimate is assumed.

Recommendation, risks and PM handoff

Reasoning: A makes one environmental rule serve combat, access and route timing; its cargo powers either a safer return or colony repairs. Its external enemy schedule preserves urgency even after a mystery is solved. B provides more deliberate observation and player-controlled escalation, but may become slow attention-budget accounting; its provocation-and-recovery loop also risks losing survivorlike momentum. Recommend A, retaining its full pause for thought and requiring live exposure for evidence.

Unresolved: Neither rhythm, loadout balance nor repeated-expedition appeal is validated. A risks deadline anxiety and unclear water/current feedback; B risks repetitive lure puzzles and over-safe clearing. Both need a fair missed-clue/recovery path and later learning-driven expedition examples in DES102. Presentation must fit verified existing free assets; none have been shortlisted or licensed here. No game assets, implementation, simulation or playtesting were produced.

Acknowledgment and PM report: Lead Game and Systems Designer; DES101 / Design r1 / In review; own work page: Design. Read Brief D011/D012 at 2026-09-06 13:08:30 UTC and Team revision 3; acknowledge the design-only gate and asset restrictions above. No accepted decisions changed. The unpublished draft was corrected to make the Flood massed enemies. Reasoning uses those records and written causal analysis, not player evidence. Old build work is explicitly superseded; no active phase mismatch found. All four work pages returned 404 at intake; World and Mystery and Design Review still returned 404 on recheck. No contribution or critique is invented.

Next handoff: PM reviews this packet, obtains owner steering under OWN101, and records the working direction before DES102. Exact request: record A, B or the required premise change, and update DES101's canonical status. PM can route this first submission to the recorded critic and coordinate WLD101; their outputs remain pending in this packet. No written feasibility review is requested at this concept stage. Concept acceptance does not authorize development.