Balance and Simulations
All numbers are Tentative prototype baselines unless marked otherwise. They came from hand-run simulations of the Tier 0 opening during brainstorming.
Proposals for every Tier 0 number this doc doesn’t set yet (tick length, vision, Scout, crew, noise, combat, HP, hives) are in tier0-proposals.md, awaiting review.
Current baseline (latest)
Twice as fast since the second playtest Tentative (Nov 2026). On screen 1× felt far too slow and 2× felt right, so the prototype doubled every speed and rate and halved every time. The rows below that the prototype has built give both numbers; the runs and sims further down are at the old speed, so their times are double what the prototype now plays. Whether the pacing targets in minutes (5 buildings at 12 minutes, Tier 2 at 36–42) halve too is open question #52.
| Parameter | Value | Notes |
|---|---|---|
| Simulation tick | 20 ticks/s (50 ms) | Decided. The server owns the tick; see player-and-platform.md |
| Harvester drive speed | 2 tiles/s (was 1) | Lowered from 2 to 1, then doubled with the rest of the game. “Slow to drive, quick to collect,” so distance hurts and players want logistics. |
| Harvester cargo | 50 | One stack of ore |
| Harvester collection | 2/s at 4 productive cells (was 1/s) | Linear: 0.5/s per productive cell of its 2×2 footprint (was 0.25) Decided |
| Harvester unload | 2.5 s (was 5) | |
| Harvester build/respawn | 15 s (was 30) | Lowered from 90 s Decided, then halved |
| Truck speed | 3 tiles/s | Must be much faster than Harvesters |
| Truck cargo | 2 slots × 1 stack | |
| Truck build/respawn | 60–90 s | Suggested; “trucks can be higher” |
| Building construction time | 15 s (was 30) | Per building, one crew: 0.15 s per unit of material. Foundation is 0.5 s a tile |
| Depot slots | 3 | Possibly 4; capacity per unit size still to be set |
| Depot storage | about 6 stacks | Matches a full load for its 3 trucks (2 slots each) |
| Crashed Ship storage | about 6 stacks | Holds the first Harvester loads |
| Storage Silo storage | about 24 stacks | 6×4 grid in a 2×2 building |
| Depot | 100 Basalite | |
| Storage Silo | 67 Basalite | |
| Barracks | 100 Basalite | |
| Foundation | 10 Basalite / tile | Raised from 1–2 on purpose [Decided direction] |
| Stone Wall | 3 Basalite / tile | |
| Ore stack size | 50 | |
| First Evolite sample | 10 units | |
| Evolite study | 3 min | |
| Dropped Evolite lifetime | about 5 min | |
| Basic alien detour limit | about 20 tiles (80 m) | Decided |
| Spawn protection radius | about 150 tiles | |
| Starting Scout range | about 500 tiles | Now 150 tiles of flight, reaching 75 out (tier0-proposals.md) |
| Building vision radius | about 20 tiles | Not additive |
| Nearest Basalite | 15–20 tiles | Generation guarantee |
| Colite drill (placeholder) | burns 1/min, produces 50/min | Must be seeded with a first batch |
| Vehicle cost | Free | Decided |
Starting stock: none in the latest sims. Giving the ship a little salvaged Basalite (25–50) was floated. The designer prefers relying on nearby Basalite, because starting stock is odd story-wise: Basalite is a material native to this planet (only similar to stone from home), a ship wouldn’t carry stone, and letting the survivors build right away undercuts the survivor opening. The cost is that the Engineering Crew idles until about 3:20 (see open-questions.md).
Harvester throughput by distance (1 tile/s, 50 cargo, full 2×2)
| One-way distance | Round trip | Resources/min |
|---|---|---|
| 10 tiles | ~1:15 | ~40 |
| 20 tiles | 1:35 | 32 |
| 40 tiles | 2:15 | 22 |
| 80 tiles | 3:35 | 14 |
| 120 tiles | 4:55 | 10 |
A Depot placed 10 tiles from a distant deposit raises each Harvester back to about 40/min. Trucks then carry the stock home. That’s the motivation for Tier 1 logistics.
Pacing targets
| Metric | Target | Status |
|---|---|---|
| Meaningful buildings at 12 min | ~5 first playthrough, 10+ experienced | Decided |
| Time to first Depot | 4–6 min | Tentative |
| First Depot payback (typical) | 3–5 min | Tentative |
| Time between affordable useful builds (early) | ~1–2 min | Tentative |
| Time between major new capabilities | ~5–10 min | Tentative |
| Time with no meaningful decision | ≈ 0 | Decided in spirit |
| First ship repair (electricity) | 30–60 min | Tentative |
| Prototype length | 60–120 min | Tentative |
Metrics to track
- Time to afford, time to first delivery, time to break even (payback) for each economic building, especially the Depot. The designer asked to track Depot payback specifically.
- The designer’s caveat: a short payback isn’t bad in itself. Players are spending on many things at once, and a fast-growing economy is the fun. The bigger risk is stockpiling hundreds of Basalite with nothing interesting to buy.
Depot payback (all 3 Harvesters on Basalite, computed with the older 90 s deploy)
| Distance from Depot | Payback after completion |
|---|---|
| 10 tiles | 2:45 |
| 20 tiles | 3:05 |
| 40 tiles | 3:45 |
| 80 tiles | 5:05 |
| 120 tiles | 6:25 |
With fewer Harvesters on Basalite, payback is slower (for example, 1 Harvester at 40 tiles ≈ 6:00). With 30 s deploys, subtract about 60 s.
Foundation cost check
Foundations must cost less per tile than any building (which includes its own foundation): Depot 100/9 ≈ 11.1, Silo 67/4 ≈ 16.75, Barracks ≈ 11.1, Foundation = 10. Recheck whenever footprints or prices change. A 40-tile corridor = 400 Basalite = four Depots.
Simulation history
Each run changed the numbers. Kept here so the reasoning behind them is visible.
| # | Setup | Result | Takeaway |
|---|---|---|---|
| 1 | 2 tiles/s, Basalite 40 tiles away, Depot 150, 90 s deploy, 8 min | First Depot 5:45; acceleration only at ~8:30–9:00 | Too slow; mostly waiting |
| 2 | Depot 100 and other buildings cut by a third, 12 min, all Basalite | First Depot 4:10; at 12 min: 2 Depots, Silo, Barracks, 10 foundation tiles, 7 Harvesters, 723 Basalite unspent | Economy snowballs; Basalite could become effectively unlimited, needs realistic multi-resource sim |
| 3 | Harvester speed lowered to 1 tile/s | First Depot ~5:30 | Distance now hurts; logistics is attractive |
| 4 | Foundation raised to 10/tile | Corridors become a big investment | Trucks preferred for distance |
| 5 | 1 tile/s, mixed resources (Basalite 40, Colite 30, Ferrite 35, Cuprite 45, Evolite 60 tiles), 12 min in 30 s steps | Only 2 buildings (Depot, Barracks); 300 of 500 gathered resources unusable until refining | Too slow. Led to the 5 / 10+ target |
| 6 | Basalite 15 tiles away (small patches), 90 s deploy, 10 min | 4 buildings (2 Depots, Silo, Barracks) + 5 foundation tiles + 10 wall tiles; Evolite study started 9:20 | Close; deploy delay holds back growth |
| 7 | Same, 30 s Harvester deploy | 5 buildings at 10:00 (3 Depots, Silo, Barracks) + 5 foundation + 10 wall, 7 Harvesters + 3 deploying, study started ~9:00 | Current baseline. On track for the 12-minute target |
| 8 | Run 7 continued from 12:00 to the electrical repair, with placeholder Tier 1 numbers, three scripted players (see Run 8) | Tier 2 at 36–42 min; the Harvester-only rush is fastest | Tier 1 timing works; Mining doesn’t pay off before Tier 2 |
Run 7 timeline (building-focused player; not optimal, since the Silo and foundations at 5:55–6:10 delay the second Depot by about 2 minutes):
| Time | Event | Basalite |
|---|---|---|
| 0:30 | Basalite discovered; Harvester sent | 0 |
| 1:55 | First 50 delivered | 50 |
| 3:20 | Second 50; first Depot ordered | 0 |
| 3:50 | Depot done; 3 Harvesters start building | 0 |
| 4:20 | 3 new Harvesters active (2 on Basalite, 1 on Colite) | 0 |
| 4:45 | Original Harvester delivers 50 | 50 |
| 5:55 | New Harvesters deliver 100; Silo ordered (67) | 83 |
| 6:10 | Original delivers 50; 5 foundation tiles (50); aliens spotted → Barracks revealed | 83 |
| 7:30 | +100; Barracks ordered (100) | 83 |
| 8:00 | Original +50; Evolite found and original Harvester sent there; second Depot ordered (100) | 33 |
| 8:30 | Barracks and second Depot done; 10 wall tiles (30) | 3 |
| 9:00 | Evolite sample at ship; study begins | 3 |
| 9:05 | +100; third Depot ordered (100) | 3 |
| 10:00 | Third Depot done; 3 more Harvesters at 10:05 | 3 |
Total delivered 550: Depots 300, Barracks + Silo 167, foundations + walls 80.
Note: one simulation wrongly treated patches as depleting. The designer corrected it: patches saturate (every 2×2 footprint occupied), they don’t deplete.
Run 8: Tier 1 pacing
Tentative throughout. Script: scripts/sim/tier1.py. It continues Run 7 from 12:00, when Tier 1 opens, until the electrical repair is delivered. Every Tier 1 number was picked for this run. None is design.
The model is coarse. There’s no map, only distances. Noise is one number for the whole base. A fight is a strength comparison. Players follow fixed build orders, and they also build a refinery whenever 150 ore piles up and keep 60 of each material back from the repair.
Placeholders
| Parameter | Value |
|---|---|
| Calcified Evolite collection | 0.4/s on a full footprint (40% of the ore rate). Deposit 60 tiles from home, 2 footprints. |
| Research (Evolite) | Logistics 60, Basic Fabrication 80, Defense 100, Stationary Mining 150 |
| Primitive Refinery | 150 Basalite. 30 ore + 3 Colite → 24 metal per minute (5:4) |
| Primitive Fabricator | 100 Basalite + 40 Iron + 20 Copper. 3 Iron + 2 Copper → 1 Basic Component every 6 s; 1 Iron → 5 rounds every 2 s |
| Outpost | 60 Basalite + 20 Iron; holds 4 stacks |
| Mining Drill | 80 Basalite + 50 Iron + 20 Copper. 120 ore/min (15/min on Calcified Evolite); burns 2 Colite/min; takes a Harvester footprint |
| Gun Turret | 50 Basalite + 30 Iron + 10 Copper; uses 20 rounds per fight |
| Truck | 1 Depot slot, 75 s respawn, 100 cargo, 3 tiles/s |
| Electrical repair | 2000 Iron, 1300 Copper, 50 Basic Components, 300 Colite, 100 Evolite |
| Noise | Harvester 1, truck 1, fabricator 2, refinery 3, drill 8. A group arrives 3 min after base noise reaches 40, then every 6 min while it stays there. Strength 2, 3, 4…; a Sword Bot counts 1, an armed turret 3. |
| Map | Basalite at 15, 20, 50 tiles (3, 3, 6 footprints); Colite 30 (4); Ferrite 35 (4) and 80 (8); Cuprite 45 (4) and 90 (8) |
Results
| Player | Plan | Mining | First fight | Tier 2 |
|---|---|---|---|---|
| Designed path | Refining → Logistics (remote Depot on Evolite) → Fabrication → Mining (2 Outposts, 4 drills) → Defense → far patches | 27:25 | 25:30 | 41:49 |
| Doctor rush | Fabrication only; Harvesters on every patch; no trucks, drills or turrets | — | 24:30 | 36:16 |
| Industrial | Logistics → Mining as early as possible → Fabrication → Defense | 24:55 | 26:35 | 40:28 |
Designed-path beats: Tier 1 opens 12:00 → refinery 13:55 → Logistics 17:35 → Fabrication 22:25 → first fight 25:30 → Mining 27:25 (first drill 30:05) → Defense 31:20 → Tier 2 41:49. That’s a new capability about every 5 minutes, at the fast end of the 5–10 minute target.
These players play almost perfectly. Run 7’s player was slower than optimal, so a first playthrough should take longer. A rough guess is 50–60 minutes, inside the 30–60 minute target.
Findings
- The Harvester-only rush reaches Tier 2 first in every variant tried. The sweep covered about 40 combinations of repair size, far-patch distance (80–130 tiles) and size, Evolite rate, Mining cost, drill rate (75–120/min) and Depot price (100–200). Drills arrive at 25–36 minutes and get only about 10 minutes to pay off before the repair is done. Harvesters are free, and a Depot turns 100 Basalite into 3 of them, so the only limit is footprints. This is the “Harvesters versus drills” risk below, showing up inside Tier 1.
- The repair bill decides what Tier 1 is about. With a component-heavy bill (200 components), the finish time is just when Fabrication is researched plus how many fabricators run. Mining doesn’t matter at all. Only a bill of bulk iron and copper lets drills affect the time.
- Harvesters alone cover a small bill early. A 600 Iron / 400 Copper bill was in stock by about 25 minutes on every path.
- Iron is the long pole. Components use 3 Iron to 2 Copper, and every Tier 1 building, ammo round and drill needs iron. Copper piles up. That’s the ratio puzzle the design wants.
- Everyone fights. Even the quiet player crosses the noise limit at 21–24 minutes from Harvesters and refineries alone. Three Sword Bots hold the first two groups. By the third, turrets are needed. This matches “the first encounter is inevitable,” but only because Harvester noise adds up.
- Evolite runs ahead after Mining. At 0.25/s collection, Mining came at 36 minutes; 0.4/s brings it to 27. By Tier 2, players hold 300–500 unspent Evolite, so research stops pacing the game in the second half of Tier 1.
- Basalite piles up on the drill paths (880–1330 unspent at Tier 2), while the Doctor rush spends it on Depots and 21 refineries. Refinery spam is the main Basalite sink.
- Fuel can jam a small Outpost. In an early version, two trucks delivered Colite at once and filled a 2-stack Outpost. The drill then had no room for ore, so it never burned the Colite, and it stopped for good. In the game that’s either a bug to prevent (fuel doesn’t count against output room) or a “Why is my Ferrite production unstable?” lesson.
Known balance risks
- Too much compounding: free vehicles plus about 3 slots per 100-Basalite Depot. Limited by 2×2 saturation, small Basalite patches and travel time Open.
- Harvesters versus drills: a dozen Harvester Depots must not permanently beat stationary drills. Drills should be several times more efficient; Harvesters win only on low noise and quick setup. Run 8: within Tier 1, Harvesters win the race to the electrical repair under every setting tried Open.
- Unusable stock: ores gathered in Tier 0 can’t be built with, so the reveal of Tier 1 refining has to arrive before that feels pointless.
- Not enough Tier 0 variety: reaching 10 buildings means building duplicate Depots and Silos. Fine if each is clearly useful; watch it in testing.
- Silos and walls have little early job. With about 6 stacks per Depot, storage rarely runs short in Tier 0, and walls protect against nothing yet. The designer is fine with this Decided: new players don’t know that yet and will build them anyway, and those who check buildings’ storage will learn they don’t need a Silo yet. Run 7’s Silo and foundations reflect that kind of player, not an optimal one.
- Cheap long foundation corridors undermining trucks, now addressed by the 10/tile price.