Overview
Steppe is a compact, fully passable land map built around a simple strategic contrast: ample food, wood, and gold occupy the broader field, while its four neutral settlements draw both players through open central and cross-map routes.
The economy is rich but not uniform. Hunt volume and species mix changed markedly across the reviewed worlds, forests shifted in density and shape, and large mines extended far beyond the starting ring. Scouting decides which outer lane is actually worth holding.
Steppe rewards control more than shelter. A position that covers a settlement, an outer mine, and one strong hunt lane can be more valuable than a technically closer resource that cannot be defended in the open.
Open Steppe & Outer Grassland
All 30 reviewed worlds were dry, with no water region or water area, and every analyzed terrain cell remained passable land. Roads, rocky mine patches, and forests supplied the main local texture without creating a separate water or chokepoint layout.
Exact forest, road, and rocky connectivity changed from world to world. Those differences matter to a nearby fight, but the final ten worlds introduced no new player-relevant terrain system or alternate layout beyond the open-center, resource-bearing-perimeter pattern.
The overhead is one example, not a canonical seed. Use it to read the broad open structure, not to memorize a forest edge, settlement, mine, or relic position.
Food
Practical initial finite food ranged from 19,200 to 30,800 across the 30 unique worlds, with a 23,470 mean and 22,600 median. This total combines normal hunt, berries, chickens, and goats at generated value; predator carcasses and full goat fattening potential are tracked separately.
| Food layer | Observed range | Mean | Median |
|---|---|---|---|
| Practical initial total | 19,200–30,800 | 23,470 | 22,600 |
| Hunt · predators excluded | 15,200–27,200 | 19,600 | 18,800 |
| Berries | 1,200–1,800 | 1,440 | 1,500 |
| Chickens | 800–1,200 | 993 | 1,000 |
| Goats · generated value | 1,200–1,700 | 1,437 | 1,400 |
| Same goats · fully fattened | 7,200–10,200 | 8,620 | 8,400 |
| Predator carcasses | 1,200–3,000 | 2,247 | 2,000 |
The herdable range represents 24–34 goats. Fully fattening them adds major long-term capacity, but that future food should not be mixed into the starting-map total. Predators are likewise separated from dependable hunt because gathering their carcasses carries risk.
Observed hunt families
| Broad species family | Observed worlds |
|---|---|
| Boar + Deer + Spotted Deer | 12 of 30 |
| Boar + Deer + Gazelle + Spotted Deer | 6 of 30 |
| Deer + Spotted Deer | 4 of 30 |
| Boar + Gazelle | 3 of 30 |
| Deer + Gazelle + Spotted Deer | 3 of 30 |
| Boar only | 2 of 30 |
These broad families describe useful scouting categories, not branch probabilities. The cohort contained 29 exact hunt-count signatures, but the extra count and cluster variation did not create 29 distinct strategic map types.
Do not budget the opening around a favorite hunt family. Confirm the species mix and volume early, then compare that food lane with the settlement and gold positions it exposes.
Wood
Observed wood ranged from approximately 152.2k to 186.4k, averaging 167.6k with a 167.2k median. The 30 worlds contained 761–932 trees, while forest-classified ground covered 5,924–7,212 square meters.
World 29 established the high endpoint without revealing a new forest system. In play, the more important variation is how several forest patches line up with open raid routes, settlement roads, and outer-resource positions.
Steppe is not short on total wood, but an exposed lumber line can still be tactically poor. Plan the next camp before pressure arrives instead of judging safety by map-wide abundance.
Gold
Every reviewed world contained 54,000 gold in ten mines: two 3,000-gold medium mines and eight 6,000-gold large mines. All 300 mine placements across the cohort occupied rocky cells.
| Reserve layer | Observed count | Nearest-player distance | Gold on map |
|---|---|---|---|
| Medium mines | 2 in every world | 21.5–24.7 m | 6,000 |
| Large mines | 8 in every world | 53.9–162.1 m | 48,000 |
The total was stable, but access was not. Large mines reached from the first expansion layer to far outer ground; 144 of 240 large-mine placements lay within 35 meters of an edge, and 28 of 30 worlds placed at least one within 20 meters.
The strategic question is where the fixed observed gold landed. Securing two compatible large mines along one route is usually stronger than stretching across the open map for isolated reserves.
Settlements & Map Control
All 30 worlds contained four neutral settlements. Their nearest-player distances ranged from 63.2 to 120.1 meters, and every one of the 120 observed placements sat on road terrain.
The cohort repeatedly showed a closer and farther opportunity on both player sides. That is a useful reading pattern, not guaranteed ownership: a nearest-player measurement does not lock a settlement to that player or make the route equally defensible.
The first settlement contest often decides more than population capacity. Road positions can connect the same open corridor to outer hunt, forests, and large gold mines.
Relics
Every reviewed world contained four relics. Nearest-player distance ranged from 71.0 to 174.2 meters, while pairwise relic spacing ranged from 80.6 to 318.2 meters.
Relics showed a strong outer-map tendency: 81 of 120 were within 35 meters of an edge, and 25 of 30 worlds had at least one within 20 meters. Those frequencies do not guarantee an edge relic, fixed symmetry, or a two-per-player allocation in a future world.
Scout the perimeter after the first economic routes are known. Steppe's relics can sit far outside the settlement fight, so central vision alone does not finish the search.
Steppe's stock script is protected, so this overview does not invent script counts, placement rules, or probabilities. It analyzes 30 unique Standard-size 1v1 worlds from build 606844: steppe_001 through steppe_029 plus a canary holdout. An unchanged parser verified identity, build, complete entity resolution, meter coordinates, and normalized world-fingerprint uniqueness before aggregation. Observed minima and maxima are cohort endpoints, not script limits; frequencies are not probabilities; and generated worlds cannot expose failed placement attempts or inaccessible branches. Broad hunt families and cluster distances are documented analytical heuristics. The approved overhead is one representative world only.
- Official game stringQuoted game copy
- Generated-world observationIndividual generated-world facts
- Multi-world empirical evidenceRanges and frequencies across unique generated worlds
- Documented heuristicReproducible grouping or distance interpretation
- Scenario Editor observationScenario Editor totals and layout
- Strategic analysisPlayer-facing interpretation