Unofficial public experiment. Not affiliated with Lantto Games or Hooded Horse. Frozen representative product; not a production launch.
Hard expiry: Sep 11, 2026, 12:57 PM UTC
Privacy and measurement
This experiment stores aggregate interaction counts by UTC day, one of five event types, coarse source class, and represented surface only. It uses no cookies or persistent identifiers and does not store raw IP, user-agent, full referrer/query, free text, or per-user/session history. Cloudflare may process network metadata at its infrastructure layer under its own policies.
A field guide for systems that misbehave
Find the failing flow.Change the right thing.
Start with what you can see. Trace the material, reaction, geometry, and signal behind it. Then adapt a pattern—with its limits visible.
Looking for an item or building definition?
Use the official Wiki ↗
FLOW / LIVE MODEL
2 represented systems
SYMPTOM Filter stalls material stops moving
check state
TRANSFORMATION Sand + Water becomes Wet Sand
adapt
PATTERN Reaction-safe line separate before contact
DIAGNOSE / FILTER FLOW
Material is stuck—or never falls through. Official mechanic 3 causes represented
STEP 1 · WHAT DO YOU SEE?
⇄
A pixel is sitting inside the filter It entered, then stopped
≋
The target material rides past It never reaches the filter
■
Everything backs up above it The output looks packed
Why these three causes?
They map to current official Filter behavior: transformation in transit, pile density/position, and an obstructed exit.
Official Filter reference ↗
STEP 2 · LIKELY CAUSE
Official mechanic The material changed while it was passing through.
A configured pixel can enter the Filter, touch another material, and become a type the Filter no longer allows. The changed pixel can remain trapped instead of moving on.
S + W → WS new state
SMALLEST NEXT CHECK
Pause the belt and inspect whether Water is contacting Sand or Spores at the Filter.
If the state change is happening there, move the wetting step before sorting—or keep Water away from the Filter mouth.
Evidence at the decision point
OFFICIAL MECHANIC
Filter + Water pages · checked 19 Aug 2026
Applicability: current Early Access pages
PATTERN LAB / STEAM POWER
Choose a loop by its constraint—not a magic ratio. Community tests are useful evidence, not universal truth. Geometry and game changes can alter the result.
BASELINE TEST CELL
Lava heat 2–3 pixels / tile in one test
steam rises
Condensers ≈ 4 / lava tile in one test
water returns
Snowmaker ≈ 0.5 / lava tile
Lumling ≈ 1 / lava tile
A compact starting point for comparison. The reported quantities came from one public player test and should be tuned against your geometry.
COMPARE THE CONTROL STRATEGY
Baseline
Tileable
Reserve-aware
BEST WHEN
You need one inspectable cell before scaling.
WATCH FOR
Steam backing up, Lava becoming covered, and assuming approximate counts are exact.
ADAPTATION
Start with one cell. Keep the return path visible, then add consumers or cells only after observing the water level.
Provisional values
One player reported that 2–3 Lava pixels per tile boiled faster than a full line of 4, with roughly 4 Condensers, 0.5 Snowmaker, and 1 Lumling per Lava tile.
Read the attributed test ↗
SYSTEM MODEL / WATER ↔ STEAM
The same loop, explained as states and movement. This surface explains the interaction. It does not prescribe a factory layout.
Community configuration
LIQUID STATE Water pool must stay available to the heat zone
CONTACT + HEAT Water → Steam Lava geometry changes flow
GAS STATE Steam rises clearance affects the path
RETURN PATH Condense or rain Steam reaching the sky can return as rain
01 State matters Water and Steam are not interchangeable inventory counts; they move and interact differently.
02 Geometry matters Angles, clearance, and contact area can change flow—so a screenshot is not a universal recipe.
03 Control matters A community logic pattern closes extra generation when the primary water reserve is low.
Official mechanic Water behavior and rain return Current official Wiki establishes Water interactions and Steam-to-rain behavior.
Official Water page ↗ Configuration and approximate quantities Current public player tests supply the pattern—not a Factory experiment.
Attributed test ↗ Provisional Reserve-aware control A public logic example reports gating extra power generation based on water supply.
Attributed logic example ↗
COMPATIBILITY / IMPACT
What changed—and which flow needs attention? This is an impact view, not a copy of patch notes.
PATCH 0.5.4 Changed surface
Filter editing changed. The causal model was rechecked.
The patch changed Filter editing and fixed several progression/puzzle issues. Current official Filter and Water pages still support the represented density, output, and transformation causes.
NEXT TRIGGER Filter mechanics, material density, or Water interaction changes
COMMUNITY PATTERN Recheck required
Steam quantities remain provisional.
No represented patch establishes universal throughput. Recheck the attributed test after changes to Lava, Water, Steam, Condensers, Snowmakers, or Lumlings.
SAFE ACTION Use the values as a starting observation and tune one visible cell
EVIDENCE & METHOD
Know what kind of answer you are using.
Official mechanic Direct current official support
Cross-checked Independent claim-appropriate corroboration
Attributed public configuration; not universal
Provisional Useful but materially uncorroborated
Factory Flow is an unofficial independent prototype. Canonical definitions and support belong to Hooded Horse and Lantto Games. Corrections should narrow or remove unsupported claims—not hide uncertainty.