Seeded Procedural Universe
A seeded procedural universe stores no world at all, deriving planets, names, economies and positions from one starting number, so the same seed always reproduces exactly the same galaxy.
Instead of data, the game holds an algorithm and a seed. Running that seed through a deterministic sequence produces coordinates, planet names, government types, technology levels and trade prices, and running it again reproduces them precisely. Nothing needs saving because nothing was ever stored. Scale becomes a function of arithmetic rather than of disk space, and a seed shared between two machines yields identical worlds on both.
David Braben and Ian Bell wrote Elite in 1984 for home computers with kilobytes of memory, where storing thousands of star systems was flatly impossible. Their generator produced far more galaxies than the game shipped with, and the pair cut the number deliberately on the reasoning that eight was already more than anyone would visit. Constraint dictated the technique, and the technique then dictated what kind of game could exist.
Vast spaces became cheap, which permanently raised the ambitions of space and survival games and turned the seed itself into something players trade. Uniformity is the cost. A million systems from one formula are a million variations on a single idea, and the real labour shifts to making that variety mean something. Later games mix generated bulk with authored landmarks for exactly this reason.
Lineage — traced back to a root
Every mechanic in this dataset resolves to an ancestral chain. This one is 2 steps deep, beginning with Procedurally Generated Levels in 1980.
Descends directly from
Forked into
No Man's Sky computed planets, creatures and languages at galactic scale, Minecraft made the seed a social object, and FTL: Faster Than Light generated a fresh sector map each run.
Carried forward by 10 later games
Look this mechanic up elsewhere
Also called: seeded generation, deterministic generation, world seed