WEIGHTED RANDOMNESS
Random outcomes with tuned probability distributions — loot tables where rarity tiers carry weights, drop chances biased by context, gacha rates published to the decimal, the dice that are loaded on purpose and often for the player's benefit.
- SOLO-DEV COST LOW
- TIMING REAL-TIME / TURN
- COMMON IN GACHA · ARPG · ROGUELIKE · CARD-GAME
Every loot game runs weighted tables under the hood (Diablo's rarity weights, smart-loot biasing drops toward the player's class), gacha games are weighted randomness as regulated commerce (published rates, pity-adjusted weights — the 0.6% five-star as an economic constant), roguelike generation weights everything from room types to item pools by depth and context (the early-floor tables that won't spawn the run-enders), card games weight pack contents by slot, and pseudo-random distribution systems (Dota 2's PRD) reweight each failure to tame streaks. Designers weight their dice because uniform randomness is almost never the experience they want: pure uniform distribution produces streaks, absurdities, and pacing accidents (five legendaries in an hour, then nothing for a month — statistically normal, experientially broken), while weights let randomness be directed — rarity encoded as probability (the weight IS the value hierarchy), context-sensitivity (drops biased toward usability, encounters weighted by area theme — the dice as invisible dungeon master, curating while appearing to be chance), and pacing control (weights shifted by drought duration are pity systems, the mechanic's most player-visible form). Weighting is thus the craft of designed luck: the goal isn't fairness in the statistical sense but felt fairness — distributions whose outputs match human expectations of variance, which are themselves biased (players experience true randomness as rigged and gently-rigged randomness as true — the gambler's fallacy isn't a bug in players, it's the spec the dice must be built to). The transparency spectrum is the modern battleground: hidden weights maximize designer freedom and community distrust (the datamined drop table as betrayal event), while published rates (now legally mandated for gacha in several jurisdictions) convert the weights into contract. Key decisions: weight visibility policy (published, discoverable, or hidden — and what the community will do when they find out), dynamic reweighting rules (pity escalation, anti-streak damping, smart-loot context — each is a deviation from independence that must never be exposed as fake randomness mid-experience), distribution shape beyond simple weights (guaranteed slots, shuffled-bag draws that promise exact counts over a cycle), and tuning instrumentation — weighted systems need telemetry, because the difference between the intended distribution and the experienced one is where the design actually lives. Pitfall: weights fighting the fantasy — when players discover the dice were loaded in ways that contradict the presented fiction (the 'random' drop that was actually gated, the streak-breaker that denied an earned crit), the revelation retroactively poisons every past outcome; weighted randomness survives disclosure only when the weighting serves the player's experience legibly, because loaded dice in the house's favor is a scandal, and loaded dice in the player's favor is just good game design wearing a poker face.