How opposite geniuses invented Doom, then tore apart

1

Two opposite kinds of genius

Id’s success ran on complementary opposites: Carmack’s monk-like engineering and Romero’s showman instinct for fun; neither trait alone would have produced Doom.

Kushner treats the founders’ childhoods as mechanism, not wallpaper. Carmack grows up isolated and obsessive, treating code as private discipline. As a teenager he is arrested for breaking into a school to steal Apple II machines he cannot otherwise afford—reckless, but single-minded about hardware.

Romero grows up in a chaotic, itinerant house, learns on early home computers, and treats game-making as performance. Arcades and comics teach him that games are spectacle first. He wants players to feel something big.

The book’s frame is that breakthrough work often pairs a technical purist with someone obsessed with audience experience. Carmack pushes engines, math, and hardware limits. Romero pushes pacing, levels, and exhilaration. The tension between those roles, not just raw talent, drives the innovation that follows—and plants the seed of later conflict when shared decisions become hard.

In practice: Carmack’s school break-in for Apple IIs and Romero’s arcade-and-comics hunger set the monk-versus-showman split that id would run on.

2

Shareware as the real hack

Id’s early wins depended less on polish alone than on shareware that gave episode one away free and let them bypass publishers for direct royalties.

Before id, PC games sat in retail boxes under publishers who took most of the money and set creative terms. Carmack, Romero, and the Softdisk circle in Shreveport find another path: ship the first episode free on bulletin boards and mail order, then charge players to unlock the rest.

A technical trick makes the business trick land. Carmack’s adaptive tile refresh copies the smooth side-scroll of Nintendo’s Super Mario Bros. on PC hardware that was not supposed to manage it. They wrap that feel into Commander Keen and release through Apogee’s Scott Miller, who pays royalties instead of a flat fee.

Kushner’s point is that the code breakthrough became a company breakthrough only because the distribution model wrapped around it. Id keeps creative and financial control a traditional deal would have stripped. Young outsiders reward speed and iteration over marketing budgets, and the give-away-then-monetize template shapes PC games and later digital software for years.

In practice: Commander Keen used Carmack’s smooth PC scrolling and Apogee’s royalty shareware path instead of a boxed publisher cut.

3

Engine and design locked together

Id’s 1991 founding and Wolfenstein 3D worked because every engine advance unlocked new play instead of sitting as background utility.

After Softdisk, Carmack, Romero, Tom Hall, and artist Adrian Carmack form id Software. The breakout is Wolfenstein 3D in 1992: shareware again, plus Carmack’s raycasting that fakes three-dimensional corridors on 2D hardware, plus fast combat against Nazi soldiers.

Other teams had tried pseudo-3D. Kushner argues what landed was Romero’s instinct for momentum—quick pacing, secret rooms, a relentless forward push that made players feel powerful. Carmack made the corridors possible; Romero made them fun to run through. Technology and design stay inseparable.

Studio culture is long nights, junk food, loud metal, and a flat rule that the best idea wins regardless of title. That informal meritocracy moves fast on raw ability. It also has no clean way to settle fights once the founders’ visions later diverge. Shareware money far beyond a boxed release validates the whole approach and funds the next leap.

In practice: Wolfenstein’s raycast corridors only became a hit when Romero’s secrets and forward rush made players feel unstoppable.

4

Doom invents a new feeling

Doom’s 1993 hit was not the first 3D shooter; it won because Carmack’s efficient engine and Romero’s dread-soaked levels created unmatched speed, agency, and deathmatch.

Doom brings texture-mapped spaces with varying height, colored lighting, and networked multiplayer, all smooth on home PCs. Carmack’s binary space partitioning lets the engine draw complex levels efficiently. Alone that would be a tech demo.

Romero’s design turns it visceral: darkness as atmosphere, sudden monster placement, escalating dread, levels built for momentum. The team play-tests deathmatch obsessively inside the office until internal rivalry becomes a feature. Competitive network combat spreads with the game.

They ship as shareware again. It moves through university networks and bulletin boards. Kushner cites internal industry estimates that it sat on more computers than Windows 95 within a year. The model scales—a bigger technical leap on the same low-cost pipe—and cultural reach explodes, along with hearings over violence and early fights about story versus pure action.

In practice: Office deathmatch sessions became product design, and shareware copies spread until industry estimates put Doom on more PCs than Windows 95 in a year.

5

Success raises the stakes

Doom’s money and fame did not heal the founders’ different values; they amplified them into open conflict and an early exit.

Profits make the founders wealthy and famous inside gaming. Kushner tracks their priorities peeling apart. Carmack doubles down on engineering discipline, wanting more engine time and less business or public noise.

Romero steps into the spotlight—magazine profiles, expensive cars, less hands-on level work, more studio image and bigger ambitions. Disputes hit shipping deadlines, whether to rebuild engines from scratch or iterate, and how much control newer hires should get.

Tom Hall leaves after clashing with Carmack over Doom’s design philosophy. The flat meritocracy that once moved fast cannot absorb deep creative disagreement without someone walking out. Kushner’s lesson is blunt: creative partnerships are load-bearing, and success is a heavier load than failure.

In practice: Tom Hall’s exit after design clashes on Doom signaled that informal meritocracy had no tool for serious founder-level disagreement.

6

Quake as the breaking point

Quake’s 1996 development made Carmack’s perfection demand and Romero’s need for freedom incompatible, ending in Romero’s ouster from the company he co-founded.

Quake aims at a true 3D engine with real-time lighting, far past Doom’s approximations. Carmack retreats into pure engineering and grows frustrated that Romero, in his view, misses finished levels and old discipline.

Romero feels micromanaged and creatively choked. He sees id turning into a joyless engineering shop that forgot fun. Factions form inside the small studio. Carmack informally polls other employees on whether Romero should stay—an act that itself ends the sense of equal partnership.

Quake ships in 1996 to major commercial success and cements id’s technical name. Shortly after, Carmack and the remaining partners force Romero out. Kushner frames it as tragedy, not villainy: two definitions of success—technical purity versus creative spectacle and recognition—can no longer share one roof once each man has the means to chase his version alone.

In practice: Carmack’s staff poll on keeping Romero marked the end of founder equality before the formal ouster after Quake shipped.

7

Solo paths prove the pair

Romero’s Ion Storm disaster and Carmack’s cooler solo run showed that design flair without engineering rigor, and rigor without showmanship, each lost the chemistry Doom required.

Romero founds Ion Storm with grand promises and loud marketing that claims to redefine gaming. Daikatana drifts through years of delays and technical trouble. Ambition outruns what the team can build without a Carmack-level engine and tight production. When it ships in 2000 it is widely panned; the gap between hype and product becomes a cautionary tale.

Carmack stays at id, pushing engines through Quake II and Quake III Arena. Technical leadership holds. Cultural dominance like Doom’s does not fully return.

Kushner uses the split paths to validate the thesis in reverse. Romero needed Carmack’s rigor and scoping to finish; Carmack alone made impressive but less explosive work. Industry shifts and bad luck play a part, yet the record still suggests the specific chemistry behind Doom and Quake was hard to reproduce with either man solo.

In practice: Daikatana’s 2000 launch, years late and widely panned after boastful ads, became the industry’s exhibit of ambition without Carmack-grade discipline.