Why the Original Grand Theft Auto Was Made
Updated
The first Grand Theft Auto makes more sense when it is treated not as a primitive version of GTA III, but as the outcome of a particular mid-1990s design problem. DMA Design had an overhead city engine, an interest in cars and police pursuit, and a team willing to discover the game by testing how systems collided. The project began under the working title Race’n’Chase. A surviving March 1995 design document describes a fast multiplayer racing-and-crashing game, while later recollections from Dave Jones and Mike Dailly show how the city simulation and the pleasure of simply driving through it gradually became more important than the original pitch.
Race’n’Chase began with movement, not a crime epic
The 1995 design material is valuable because it prevents hindsight from rewriting the project. It envisaged races, crashes, police pursuit and several modes, with PC, PlayStation and other contemporary machines under consideration. The point was not yet to build a cinematic gangster saga. It was to make movement through a city entertaining and unpredictable. That distinction explains much of the finished game. Grand Theft Auto would eventually acquire mission-givers, criminal progression and three fictional American cities, but its basic appeal remained systemic: a vehicle, a road network, traffic, police and the possibility that several rules might interact at once.
Dave Jones later recalled that DMA had an overhead city engine whose parallax buildings created a strong impression of speed and height. Attaching the camera to a moving car suggested a chase-helicopter view and, more ambitiously, the possibility of a “living” city. The important leap was therefore not simply from racing to crime. It was from a fixed course to an urban system in which the course could be almost anywhere.
The police changed the meaning of the city
A racing game normally tells the player where to go and how success is measured. Police pursuit changes that relationship. Streets become escape routes, traffic becomes an obstacle or opportunity, and the player starts reading the map tactically. Dailly’s recollection of early development describes a game in which players could originally take police or criminal roles. The final product concentrated its identity elsewhere, but the pursuit logic survived as one of the design forces that made free movement interesting.
This is also why the game’s wanted-response idea mattered beyond spectacle. A city that reacts to what the player does feels less like scenery. Even with the severe limits of 1990s hardware, a response loop could imply an institution beyond the player: commit an offence, attract attention, escape, and watch the situation escalate. Later GTA games would make that illusion vastly more detailed, but the conceptual seed is visible here.
DMA discovered that freedom was itself a feature
Dailly has described Jones as being fascinated by giving players tools and allowing them to decide what to do. That philosophy was unusually consequential. The team did not need every entertaining moment to be authored as a mission. Driving quickly, stealing another vehicle, finding a shortcut or provoking a pursuit could generate a small story through play. In design terms, this was a shift from content that is consumed in sequence towards systems that can be recombined.
The finished 1997 game still has progression. Players work through jobs in Liberty City, San Andreas and Vice City, earning enough points to move onwards. Yet the route between objectives is full of opportunities to ignore the immediate instruction. That tension — authored job on one side, self-directed city play on the other — became one of GTA’s most durable structural ideas.
A difficult production exposed the cost of interacting systems
Freedom is expensive to build because one feature can affect many others. Jones later estimated that development took about 30 months rather than the roughly 18 originally expected, and described the difficulty of compound interactions. A new mechanic could not be tested in isolation if it changed traffic, pursuit, collision or mission behaviour elsewhere. That is a revealing technical reason for the game’s long gestation.
The early design document also shows ambitions that did not all survive in the form originally imagined. Platforms and modes changed. The lesson is not that the document predicted the finished GTA perfectly; it is almost the opposite. The project’s identity emerged through iteration. Reading the document alongside developer testimony shows a design being negotiated rather than executed from a single immutable blueprint.
The 1990s mattered because games were renegotiating scale
Grand Theft Auto arrived when 32-bit consoles, faster PCs and CD-ROM distribution were changing what designers could put around the player. Many celebrated games of the period pursued 3D spectacle. DMA’s first GTA took another route: it retained an overhead viewpoint but used a large navigable city, numerous vehicles, pedestrians, audio and missions to create breadth. Its novelty was therefore not simply graphical realism. It was the density of possible actions inside a readable urban space.
That choice also helped the game support three differently named American cities without requiring the detailed 3D geometry that later entries would demand. The player could understand blocks, junctions and districts quickly from above. The abstraction was a limitation and a design advantage at the same time.
Crime supplied a flexible reason to use every system
Once the project moved towards Grand Theft Auto, the criminal framework gave its mechanics a common purpose. Taking cars explained why the player could constantly change vehicles. Jobs justified rapid movement across the map. Police pursuit created pressure. Rival figures and mission contacts supplied direction without requiring a deeply scripted central protagonist. The theme joined systems that might otherwise have felt like separate experiments.
This does not mean the original GTA invented open-world design or that every later sandbox descends directly from it. Such claims require more evidence than resemblance. Its historical importance is narrower and more defensible: DMA combined urban navigation, vehicles, pursuit, mission structure and optional disorder in a form that its own sequels would radically expand.
The most important invention was a relationship with the city
Looking back from modern GTA can make the first game seem visually remote, but its central relationship is recognisable. The city is simultaneously map, obstacle, toy box and stage. Players learn where roads lead, which vehicle suits a situation, how police react and when abandoning the planned route produces something more interesting.
That is why the origin story of Grand Theft Auto is more revealing than a simple tale of controversy or branding. A racing-and-chasing concept encountered an overhead city simulation, and the team discovered that the space between objectives could be as important as the objectives themselves. The 1997 game was rougher and more abstract than its descendants, but it established the question that the series would keep asking: how much of a convincing city can be turned into something the player can actually use?
Development snapshot
| Stage | What it shows |
|---|---|
| Race’n’Chase design, 1995 | Racing, crashing, pursuit and multiplayer ambitions around an overhead city engine. |
| DMA iteration | Developers discovered that free driving and interacting city systems were compelling in themselves. |
| Grand Theft Auto, 1997 | Crime missions, three cities, vehicle theft and police response gave those systems a coherent framework. |
Sources used for factual grounding: the surviving March 1995 Race’n’Chase design document; Dave Jones’s Game Developer retrospective; Mike Dailly and Paul Farley’s Guardian interview; contemporary release information cross-checked against historical catalogues. Interpretive claims about design are identified as analysis rather than private developer intent.
