14 Sandustry Tips That Actually Move the Needle

A curated tip sheet — every entry source-tagged so you know what's official (VERIFIED), what's community-tested (COMMUNITY), and what's a single observation (single-source).

The First Hour

1. Spend your first 20 gold on the Shaker. It's the first research node and the whole gold loop starts there VERIFIED.
2. Build near water, not around it. 1 sand + 1 water = 2 wet sand VERIFIED; short carry distance beats a pretty base.
3. Don't dig "down and hope." Reserve vertical shafts for the press later COMMUNITY.

The Gold Line

4. Plan for ~4 wet sand per 1 gold VERIFIED — size the wet-sand stage 4× your target gold rate, not 1×.
5. Diagonal shakers beat flat rows. The gap lets belts pass between machines COMMUNITY; see Shaker Setup.
6. Route gold into a Collector, never onto the floor VERIFIED — loose gold falls into water and blocks lanes.
7. Chase the chain before adding machines. If gold stalls, trace the input side first; a bottleneck upstream makes extra Shakers pointless (full method in How to Get Gold).

Logistics & Automation

8. Learn the Shift-skip launcher trick. Throwing items with momentum and skipping precisely turns a launcher into a routing tool COMMUNITY.
9. One-way lanes, always. A belt that doubles back on itself recirculates items and clogs filters COMMUNITY.
10. Research in the community-mapped order: Refinement 1 → Logistics 1 → Filter → Pipes → Tools → Refinement 2 (Slag Press) → Drones COMMUNITY.

The Slag Press & Heat

11. EA heat = Burner Belt + Thermal Buffer. The Demo's "lava self-ignite" trick is dead; don't follow old videos COMMUNITY (see EA vs Demo).
12. The press needs a 25-block drop. 1 Burnt Slag dropped ≥25 blocks → 1 Gold + 1 Spore VERIFIED; wrong height or wrong input = clog.
13. Combustion is probabilistic. You get less Burnt Slag out than slag in — over-produce the burn stage VERIFIED.

Keeping the Save Healthy

14. Update before blaming your PC. v0.5.3 fixed a steam/rain memory leak and added a GPU-overlay launch option COMMUNITY; see Performance and v0.5.3 Notes.