Build Comparison — Three Ways to Break a Horde

Storm Choir runs toward a connected knot and watches its charge meter. Ricochet Foundry circles for hard angles, watching several discs cross the room. Gravity Furnace stays just ahead of a tail, preserving a rhythm of glowing footsteps. All three can become absurd; they do not ask the player to watch, move or press for the same reason.

These are proposed directions, not settled recipes or balance. The Owner Brief selects connected reusable parts, mostly automatic ordinary fire, regular temporary growth through fighting with major transformations, fresh hordes each expedition, and loss of temporary power on every return. Physical carried parts remain. Exact growth cadence and numbers stay open.

Storm Choir — conduct the crowd

A rig is the equipment mounted on the astronaut and the links that make it work together. Here, the Threat Sensor triggers the Pulse Projector automatically. Its impacts feed the Impact Dynamo; the Fork Relay divides that charge between the Arc Emitter and a Capacitor, which stores it for later. The player reads the crowd as a circuit, crossing toward mutually visible bodies so one arc keeps finding the next. That dangerous move also fills the store.

Regular growth first lengthens or branches the local song: Echoing Arc turns one useful route into several. At a major transformation, Thunderhead keeps the current full-store rule. The player sees three chain heads and a pressure-lance corridor previewed across the crest, then chooses whether to spend now on a merely good knot or risk holding for the dense centre. At its peak, blue-white routes tear open a lane while ordinary fire later rebuilds charge.

The six parts use seven carrying units and seven mounted spaces, requiring the later eight-space rig. One two-unit canister brings carrying to nine of ten, leaving one unit for discovery. The build has no Barrier. Wide spacing, hard cover or a moving gap can strand its branches, and crossing toward density can put the astronaut inside the horde before the store is ready.

The same Fork can be seated in Cargo Turn's service board, where one pulse visibly reaches two lamps. That station discovery teaches branching and branch refusal, but not the finished combat recipe.

Ricochet Foundry — compose the room

At expedition start, the Threat Sensor triggers ordinary Pulse Projector shots. A Manual Trigger controls the Vector Nozzle, a directional shove that moves loose cover or bodies it can affect while the automatic fire holds a lane.

Regular fighting can grant Rebound Disc, changing later accepted shots into bounded banking flights. Fragment Wake then gives each normal terminal a shown fan. The player starts circling and deciding whether Nozzle should shape the crowd, move cover or preserve an escape.

The proposed major Foundry Drop changes the current rule: remove Dynamo, Capacitor and the full-store master disc from this direction. Instead, a Nozzle pulse that visibly intersects a friendly disc may force that disc to take its normal terminal immediately. At the peak, the player waits until two or three automatic discs cross different sides of the horde, faces one pressure cone through them and presses. Several existing fans open together from player-chosen points. Pressing early wastes the geometry; waiting too long lets the discs terminate naturally. Nozzle still does not steer a flight.

Sensor, Projector, Manual and the two-space Nozzle consume five carrying units and five mounted spaces, so they need the six-space rig. One canister brings the carried total to seven of ten, leaving three units for finds. That small footprint is also the largest balance risk. The build gives up stored charge, Arc and Barrier, while its one manual shove must serve setup, escape or forced terminals. Open atriums, soft surfaces and fans aimed into cover spoil it.

Seated in a decontamination air-knife mount, the same Nozzle accepts the local control and pushes a loose shield away from a hard service face. The revealed reflection mark suggests a possible bank without listing every useful angle.

Gravity Furnace — make pursuit into fuel

The Stride Switch triggers the Vector Nozzle through deliberate movement; Inward Draft later turns each qualifying traversal into a rear pull. The Threat Sensor and Pulse Projector supply modest automatic fire while the player draws a pursuing tail. The player watches spacing and cadence, not an impact-fed battery.

This direction also proposes replacing Dynamo and Capacitor. A regular Heatstep transformation makes each deliberate Stride-powered pull scorch one visible ember into the floor wake. Only the three newest remain; the oldest visibly dims and goes out after a short shown interval. Forced motion, recoil and conveyor travel leave none. These are not battery charges: they show that the astronaut has kept the hot pursuit rhythm alive.

At the major Furnace Burst, temporary behavior pauses the Sensor's automatic path and lets Manual→Thermal Lens→Projector draw from those three live embers. The press consumes them into one lance and bloom, then ordinary Sensor fire restores without a journal edit. Three is a proposed readability target: the first mark teaches the response, the second makes a pattern, and the third creates a recognizable “turn now or lose it” moment without a long count. It is not a tested optimum; the number and lifetime can change.

At the peak, three hot seams glow behind a gathered tail. The player must turn and fire before the oldest fades, but turning early may put the bloom behind the mass. Hard cover divides it, a missed stride breaks the rhythm, and Bellback's pressure can scatter the wake. Replacing Dynamo and Capacitor with Manual makes six parts use seven carrying and mounted spaces, requiring the later eight-space rig. One canister brings carrying to nine of ten. It has no Barrier and cannot bank a burst while standing still.

A station gait rail fitted with Stride ignores a conveyor-carried pass but answers deliberate steps; a sterilization exchanger shows Lens turning a valid input into heat. Those observations reveal ingredients, not the Heatstep sequence or best combat route.

Why the difference matters

Storm asks which crowd is connected and when to spend stored power. Ricochet asks where moving paths will end and when to collapse them together. Gravity asks whether the player can maintain movement cadence, gather pursuit and turn before the window closes. Fresh hordes let those skills face new crowd shapes each expedition; regular fighting supplies incremental transformations, while major milestones supply Thunderhead, Foundry Drop or Furnace Burst without this page choosing thresholds or pickups.

The accepted cost is three clearer fantasies but two new temporary exceptions and more preview work. Current shared Mechanical rules continue to govern until these replacements are separately reconciled; this comparison does not silently change their power, heat, flight-slot or finite-route safeguards. Written scenes cannot prove that any direction is fun, balanced or sufficiently discoverable.

Can a shot carry lightning?

The second mapped trip in Upgrade Choices creates a fair test inside Storm Choir rather than a fourth fantasy. The astronaut carries the same Sensor, Projector, Dynamo, Fork, Arc and Capacitor through the same sealed pressure route. They already earned Thunderhead and chose Rebound Disc at the first Tune. At the sixth mark, one 40–75 hall band has broken for Echoing Arc while surviving bodies and the other band still contest Receiver Vault Approach. Neither treatment below is selected.

Proposed alternative — assemble a bank-delayed carrier

The hint appears after Rebound Disc is chosen, before the player owns Echoing Arc. In the paused journal, selecting the Arc Emitter's physical-effect collar and the Pulse Projector together reveals a faint matching outline at the Projector's payload rail. It says that the two physical interfaces can be linked; it does not award or name a finished special. The player decides to test that relationship.

They disconnect Fork's ordinary Arc branch, so Projector impacts now charge only the Capacitor, connect the Capacitor's chosen release to Arc, and clip Arc's physical collar to Projector's payload rail. Sensor still owns automatic Projector targeting. The new carrier link changes the Rebound preview: its electrical trace is hollow through the initial segment and any bodies before the first valid hard-face rebound, then becomes bright. The first body shown after that rebound carries a release ring. That honest preview teaches the configured rule without promising where a moving crowd will be when the disc arrives.

The first useful experiment can happen at Balance Gallery. From the open side of the fixed rib, the nearest visible Presser is a legal automatic target. It stands in front of a hard rib face whose shown reflection crosses the rib's far side, where a second Presser occupies the outgoing segment. The astronaut arms the chosen release. The disc hits the selected Presser without spending its electrical payload, reaches the rib, rebounds, and releases base Arc when it contacts the second Presser. The player has made two familiar behaviours answer each other: a bank arms the carried action, and the next body releases it. A route ending before a bank produces a visible fizzle instead.

At Receiver Vault Approach, Echoing Arc turns that learned trick into the later horde payoff. The astronaut stands on the clear near side of the existing shelter, just outside Siphon's line and Shutterback's fixed fan. One exposed Presser in front of a pillar is the Sensor's nearest visible legal target. The incoming segment passes through that Presser and reaches the pillar's hard angled face; the shown reflection runs through the shelter's far opening. The first body on that outgoing segment is another Presser at the near edge of the connected far-side group. The shelter blocks a clear electrical line between the two Pressers and blocks direct Arc selection of the far-side one from the rig.

On the successful press, the first Presser takes the physical contact but not the Arc. The pillar rebound arms the payload. Contact with the far-side Presser releases Echoing Arc inside that connected group while the disc continues along its remaining shown route. If the displayed branches find their current clear hops, the far opening loses enough pressure for the astronaut to cross the shelter's end instead of entering the exposed centre. That is the proposed advantage: delivery establishes a useful electrical origin beyond a hard-cover break while automatic initial targeting still obeys the nearer visible body.

Strong cheaper baseline — keep two useful effects

The current hybrid keeps disc and Arc independent. Automatic discs follow their shown bank paths; the local Dynamo gives at most one contribution per Projector wave, and Fork feeds both a rig-origin Arc action and Capacitor storage. A bad disc can still leave useful local lightning, and the full store still offers Thunderhead's reliable three nearby chain heads and straight lance. All of this uses current links, sensing, previews, heat and reservations.

At the same shelter position, that strength does not solve the far-side opportunity. Ordinary Arc begins on the exposed Presser and its clear routes stop at the shelter. Thunderhead's extra seeds remain near that initial target, while its straight lance reaches the Presser and then the pillar rather than following the reflected lane. The baseline can break more of the visible front and is safer when that front is the real problem; to reach the far-side group, however, the astronaut must move around the shelter and accept Siphon's or Shutterback's angle. This is why the carrier proposal is more than matched effects, while the baseline remains the stronger general-purpose choice.

What the proposed relationship changes and costs

The carrier link is a new direct physical compatibility, not a general signal port, new part, upgrade card or automatically granted mode. It works when Arc and Projector are deliberately connected; Rebound Disc supplies the bank that arms delayed release. Echoing Arc improves the carried action later but is not a prerequisite, and Thunderhead remains a separate full-store arrangement. Returning to ordinary local Arc or Thunderhead requires another paused-journal reconfiguration, so the carrier cannot supply all three benefits at once.

The proposed command uses Capacitor's normal 2 U chosen release rather than the full store. A brief hold reserves the next natural-ready Projector instant, but commitment still requires Arc and Projector readiness, a current legal Sensor target, a shown valid bank path, and room for the normal Projector launch plus one Arc action. Refusal or lapse before launch spends nothing. Acceptance removes 2 U; the parent disc keeps its normal flight slot and 1 P, while the carried Arc keeps 2 P and its current heat until it releases and its routes finish, or until the disc ends and the payload dissipates. The disc retains its current contact, distance, time and rebound limits and continues after release.

A soft terminal before the bank, a changed reflection, no body after the bank or blocked Arc routes gives no refund. Pre-bank contacts never seed lightning. The carried wave cannot charge Dynamo, refresh readiness or create another discharge, and one body remains one Arc body even when it also takes the disc contact. These rules and the faster single-branch charging cadence are proposals, not hidden changes to the current catalogue.

The cost is loss of ordinary local Arc while configured, more commitment to a bad trajectory, and substantial interface work: a side-link state, dormant/armed/release preview, held readiness, contact and fizzle feedback, sound and accessibility equivalents. Adoption would also require the Equipment Catalogue and Mechanical Interfaces to reconcile heat, automatic-launch timing, Fragment Wake, Long Arc and overlapping flights. The cheaper baseline needs only clearer causal feedback between its existing disc impact and rig-centred electrical response.

After either treatment clears or separates the surviving pressure, the reward stays quiet: the astronaut captures the same seal strip and compares it with the hospital commissioning card.

For the common equipment rules, see Build Rules and the Equipment Catalogue; Mechanical Interfaces holds the detailed quantities.

See these ideas in Three expeditions into the observation ward, or return to Hordes and Builds. Back to The Game.

Open questions

• Are Ricochet's forced terminal and Gravity's three-ember heat rule understandable without turning the journal into a recipe list?
• Does Ricochet's smaller permanent footprint need another physical part or higher operating cost, or is its dependence on room geometry enough?
• What regular-growth choices lead toward each major transformation, and what milestone triggers it?
• How should renewed hordes occupy permanently opened and joined rooms without making return travel repetitive?
• Which exact power, heat, duration and coverage values preserve overpowering success without making one build universal?
• Does the direct physical carrier link and its dormant/armed preview teach assembly without feeling like a hidden recipe or a supplied special?
• Is access to a group beyond a hard-cover break worth giving up ordinary local Arc, Thunderhead access while configured and 2 U on a failed bank?
• Should the carrier keep the proposed normal 2 U release and hold, and how should it interact with Fragment Wake, Long Arc and overlapping flights?
• The rules are untested and all quantities remain proposals.