Game
reading the chain…
Game
reading the chain…
3 sealed trees · set #1
The inverse map with memory — a persistent board steered to the exact settled award.
All-time
first round 33d ago · last 1d ago
How it's built
A falling-block chain-reaction slot whose board survives between spins — the first presentation on the floor with persistent state. That breaks the pre-mined seed library every stateless game leans on (which seeds realize an award now depends on the standing board), so the inversion runs constructively: the chain settles the award first, and an inverse planner synthesizes the piece flow that makes your standing board pay exactly that number under the published physics. Awards are integers in printed units (20 units = 1× stake), so every ladder value maps to an exact committed multiplier. The load-bearing math is V-universality — a proved invariant that from any reachable board, every award in the ladder stays exactly realizable forever — enforced at runtime by the planner's backwards-looking admissibility filter (any move whose successor board leaves the invariant is refused) and certified by an exhaustive fuzz artifact in the self-test. Three sealed stake rungs (25× / 100× / 400× tops — the stake picks the deepest board the pool seats), every root committed at EXACTLY 985000ppm.
Build one yourself — the SDK seals any distribution the edge band admits.
Rungs
Recent rounds
Each round's roll is re-derived from the revealed seed on its channel's proof page — nothing here is read from a private database of outcomes.