Silicon and Synapse to Blizzard
Blizzard Entertainment began as Silicon and Synapse, founded in 1991 by three UCLA graduates — Michael Morhaime, Frank Pearce, and Allen Adham — in Morhaime's apartment. The company's initial work was game porting: adapting existing titles to new platforms, a service that generated steady revenue without requiring the capital investment of original game development. Rock N' Roll Racing (1993) and The Lost Vikings (1991), both developed under the Interplay relationship, were original games that demonstrated the team's design capability, but they were modest commercial successes rather than defining works.
The company renamed itself Blizzard Entertainment in 1994, the same year Warcraft: Orcs and Humans launched. The renaming was partly about marketing — "Blizzard" communicated energy and scale that "Silicon and Synapse" didn't — and partly about identity: the team was transitioning from a studio that executed other people's designs to one with a distinctive design voice of its own. Warcraft was the demonstration of that voice: a real-time strategy game built with a visual clarity and unit responsiveness that its competitors lacked, and an online multiplayer component through Battle.net (Blizzard's online gaming service, launched with Diablo in 1996) that made the games community experiences rather than solitary ones.
Warcraft and the real-time strategy genre
Dune II (1992), designed by Brett Sperry and Louis Castle at Westwood Studios, had established the real-time strategy template: resource gathering, base building, unit production, and real-time combat on a top-down map. Command and Conquer (1995) refined it. Warcraft: Orcs and Humans (1994) did something different: it centred the player's attention on individual units rather than on economic systems. Where Command and Conquer's units were largely interchangeable expendables managed in groups, Warcraft's units could be selected individually, had distinct abilities, and accumulated experience — and were scarce enough in number per engagement that losing one was a meaningful setback rather than a statistical variance.
Warcraft II: Tides of Darkness (1995) was larger, more polished, and shipped with a map editor that allowed players to create and share custom scenarios — a decision that built a level-design community around the game and extended its commercial life far beyond what the base game alone would have sustained. The map editor established a model Blizzard would return to repeatedly: games shipped as platforms for player creativity as much as as finite experiences with defined endpoints. The StarCraft and Warcraft III editors generated map communities that produced custom game modes — Blizzard's DOTA (Defense of the Ancients) evolved from a Warcraft III custom map, and eventually became the genre template for the MOBA category — that had no direct relationship to the original game's design intent.
Diablo and the action RPG
Diablo (1996) was designed by Condor — a small San Diego studio that Blizzard acquired during Diablo's development, renaming it Blizzard North. The game's original design was turn-based, closer to NetHack or Rogue than to action games; Blizzard's intervention during development shifted it to real-time action. The final game combined the dungeon structure and loot system of roguelikes with direct mouse-click combat and online multiplayer through Battle.net, producing a game that could be played for five minutes or five hours with equivalent engagement per unit of time.
Diablo's loot system was its primary engine of engagement. Monsters dropped items whose quality and properties were randomised within defined parameters; the most powerful items were rare enough to require sustained play to encounter. The probability of any specific drop on any specific kill was low, but the cumulative probability of encountering something valuable over an extended session was high enough to sustain engagement through a play loop that could be described simply: kill monsters, collect items, become stronger, kill stronger monsters. The system's formal simplicity belied the depth of engagement it produced: players could be absorbed in Diablo's loop for hundreds of hours without the experience repeating in ways that felt mechanical. The randomisation ensured that each session's specific contents were unique even when the genre of the experience — dungeon clearing for loot — was the same.
Diablo II (2000) expanded the system to the point where the loot hunt became the primary game rather than a feature of it. Character builds optimised for specific item sets, the item economy between players on Battle.net, and the extended endgame content in Lord of Destruction's expansion created a game that players sustained for years. Blizzard had, with Diablo's design decisions, accidentally invented an engagement loop that subsequent games including Path of Exile and Borderlands built entire franchises upon.
StarCraft and the competitive ceiling
StarCraft (1998) was initially regarded as Warcraft in space — a reskinned real-time strategy game that offered moderate improvements over its predecessors. This initial assessment missed what StarCraft had actually accomplished: the design of three factions — Terran, Zerg, and Protoss — that were completely asymmetric in unit composition, building structures, and resource mechanics, but balanced against each other at the highest levels of competitive play. Achieving this balance required that each faction's advantages and disadvantages be precisely calibrated against the others', a design problem orders of magnitude more complex than the symmetric-faction balance that most strategy games attempted.
South Korea's adoption of StarCraft as a competitive spectator sport produced the evidence that Blizzard's balance was genuinely exceptional. StarCraft became the subject of professional leagues, television broadcasting, and player contracts in Korea from the late 1990s through the 2010s. Professional players trained for eight to twelve hours per day, achieving actions-per-minute rates of 300 or more — an execution standard that made professional StarCraft play look different in kind from casual play, not merely different in degree. The game's skill ceiling was high enough to sustain professional competition for over a decade, a period during which most games' competitive scenes lasted months before their player populations consolidated around a solved metagame.
Blizzard's four-year period of defining genre templates — 1994 to 1998 — was the product of specific decisions: hiring designers who understood the genres they were working in, building iteratively and removing features that didn't improve the play experience rather than shipping everything that had been planned, and investing in online infrastructure that made their games community experiences. The quality of execution was high enough that the games they produced remained competitive references for their genres for a decade after their release. That record is rare enough in the games industry to require explanation, and the explanation is mostly about discipline: knowing what to cut and what to keep.