In the public eye, the legacy of the Atari 2600 home console has always been the videogame crash of 1982. The image of thousands of unsold copies of E.T. the Extra-Terrestrial and other shovelware being buried in landfill will always overshadow the consoles’ successes and innovations. And yet, as many seem to forget, some of the art’s most critical and defining moments took place during the 2600’s reign over the market. It is no secret that older consoles’ most severe limitations often inspired their most creative solutions – Howard Scott Warshaw of E.T. fame, for example, began his developer career with Yar’s Revenge, a title which functionally and graphically broke more new ground than almost any of its contemporaries. Yar was the first game which could be paused, the first game with “filmic” (in Warshaw’s own words) audio-visual design, the first game with a full-screen explosion and, most famously, the first game to use its own source code to create an on-screen graphical element. It was also, incidentally, the first game for which the developer was openly credited. There is a common theme here: Yar’s development, like that of many other 2600 titles, saw its developer attempting to replicate features from other art forms, and other kinds of games. Where these developers arguably made most progress was in creating increasingly elaborate worldspaces, and situating the player within.

What fundamentally separates video or board games from other arts is this interaction between players and their worldspaces – early tabletop RPGs evolving into computer-generated text adventures (most notably with Will Crowther and Don Woods’ seminal Colossal Cave Adventure) faced the issue of creating a descriptive interaction between player and environment which resembled actual fiction, and even reality. The challenge thus presented to Warren Robinett, who three years later sought to bring Adventure into a new visual medium on the Atari 2600, was significantly greater. Many of the techniques Robinett exercised for the first time now form such an integral part of game design that one would be forgiven for believing them to be common sense: instead of North-South-East-West, for example, the player’s movement was mapped to Up-Down-Left-Right, and when a particular ‘room’ or ‘screen’ was left on one side, the next would be entered from the other side. Rooms were coloured differently to indicate a progression of environments, and a ‘dark room’ effect was achieved by blackening the screen with the exception of a small square around the player icon, representative of the character’s limited vision. Many of his techniques, too, remain even now more unconventional tools of game design: while the room system operated on a grid, many areas overlap impossibly or link up on opposite sides, as notably practiced later in the original Legend of Zelda, though the implementation here being much more advanced. This system made navigation purposefully disorientating at times, and paper mapping (common practice in adventure games even today) completely futile.

Though developed for Unix systems, Michael Toy and Glenn Wichman’s original Rogue played a similarly significant part in the move from text-based to graphics-based adventures. Adopting many of Robinett’s principles of navigation though and interaction with the world, Rogue can be seen as a much more complex elaboration on Adventure’s ‘Mode 3’, a mode which scattered the game’s key items randomly throughout the world. Rogue’s world, though Spartan in its appearance, was a step-up from its predecessor in that, for the first time, its layout and contents were generated completely procedurally. Functionally, this provided a completely unique experience on each playthrough and was bound by the same rule of feasibility as ‘Mode 3’: keys always spawned outside their respective doors, potions and scrolls were placed on the same floor as their respective puzzles, and enemies were progressive in their dangerousness to the player.

Two years later, back on the 2600, David Crane’s extremely popular Pitfall! platformer adopted the principles of Adventure and Rogue, but applied them to an entirely new context – that of the action game. Though similar to Adventure in its use of a grid, or a series of ‘areas’ with defined edges, to display the worldspace which could not be confined within a single screen, Pitfall!’s move to a side-on view substituted up-down-left-right movement for North-South-East-West (the difference being that while the on-screen axes of movement remain the same, the implied real-world axes are shifted adjacently). The genre now moved from single-screen to large-scale platforming territory – in effect bridging Donkey Kong and the later Super Mario Bros. – and from representative icons and boundaries to true visualisations. Blocky, ASCII dungeons here became colourful, detailed jungles and dark, atmospheric caverns; meanwhile the player character, represented by a simple square in Adventure and an ‘@’ symbol in Rogue, is in Pitfall! a fully-fleshed-out figure, whose running animation was the product of hours of Crane running round the Activision offices and sketching himself. Yet further to this, the procedural generation which defined Rogue and Adventure’s ‘Mode 3’, for the very first time, extended to the game’s cosmetics – as well as the physical layout of the landscape and placement of items and enemies, the game’s more aesthetic features such as tree formations and underground structures were similarly randomised (a feature which was massively expanded upon in the game’s expansive 1984 sequel). For the first time in a videogame, both the player and the world they inhabit were real, organic entities.

One particular natural progression made from these early attempts at ‘organic immersion’ seemed to be replacing the player icon or figure with the player themselves, shifting from a bird’s-eye or side-on perspective to first-person. Though first-person ‘walkers’ predated the console by several years, kicking off with Steve Colley and Greg Thompson’s 1974 MazeWars project, the genre found its first comfortable home on the 2600. Obviously these early titles could not accomplish 3D-POV in the way that later PC titles like Doom would, nevertheless a whole decade earlier remarkable progress was being made. The principles of navigation and immersion are carried forward here: 1983’s Tunnel Runner, for example, spiritually akin to MazeWars, utilised coloured floor panels for progressional points of reference in much the same way as Adventure used coloured walls. The implementation is nonetheless unintentionally disorientating and extremely difficult to grasp (especially for those used to the luxuries of modern first-person movement mechanics), and the reliance on a map screen does little but confuse the player. Ed Rotberg’s Battlezone implements the same ideas, albeit much more successfully. Using coloured ground to meter movement direction and distance, and providing an indication of the player’s orientation in the form of tank tracks at the bottom of the screen, in combination with analogue steering and movement, Rotberg is highly successful in physically placing the player within a faux-3D worldspace where intuitive navigation and visual feedback allows for greater immersion than ever before.

3D worldspaces, detailed graphical feedback and procedurally generated gameplay on the 2600 came together at perhaps their most successful moment in Doug Neubauer’s Solaris in 1986, though by this point the videogame crash had long since passed and the Atari’s influence over the market had continue to dwindle throughout their subsequently unsuccessful launches of the 5200 and 7800. The Famicom’s popularity in Japan had allowed for the launch of the Nintendo Entertainment System in the West, and the hardware limitations of Atari’s machines sadly prevented any further significant innovation. The 2600’s loss of relevance to the NES marked a shift in the industry: a shift away from the kinds of shovelware, knock-offs and dreadful licensed titles which plagued Atari systems, but also away from the kind of wild experimentation which made titles like Yar’s Revenge, Adventure and Pitfall! so unique. Nintendo’s subsequent domination of the market has led to a relatively unfavourable remembrance of Atari’s place in the art’s development – only in retrospect can we look back at these titles and observe how much, and on how many fronts, they truly pushed videogame design forward.

Further reading: Ian Bogost & Nick Montfort – Racing the Beam (MIT Press, 2009); GDC Postmortem Talks with Howard Scott Warshaw, Warren Robinett and David Crane