Atari: The Golden Years -- A History, 1978-1981

Following his article on Atari's genesis, game historian Fulton returns with an amazingly detailed piece on Atari's 'golden years', from the rise of the Atari 2600 through Asteroids and Battlezone.

[Following his article on Atari's genesis, game historian Fulton returns with an amazingly detailed piece on Atari's 'golden years', from Asteroids through Battlezone.]

Atari was founded in 1972, but its crowning accomplishments in console gaming and computers were the Atari Video Computer System (or 2600) and the Atari 400/800 line of personal computers.

This four-year period -- from 1977 to 1981 -- contains some of the most exciting developments the company ever saw in its history: the rise of the 2600, the development of some of the company's most enduringly popular games (Centipede, Asteroids) and the development and release of its first home computing platforms.

This comprehensive look back, filled with quotes from the original creators and other primary sources, offers a detailed peek into the company that popularized video gaming as the '70s turned into the '80s, and created the first viable market for home consoles.

For more detail, be sure to read Gamasutra's first Atari history article, which covers the period of 1971 to 1977 -- the latter date being the year that the Atari VCS was first released.

Innovate, Kind Of Like You Did Last Year

"One of the guys at Warner said... I had made a proposal to make a really interesting set of games. I can remember him not even blinking and looking at me and saying 'Nolan, why don't you innovate kind of like you did last year, none of this new stuff?' He did not understand what he said, he was so out of tune with what the nature of innovation is, and I've been thinking I was going to get that put into needle-point sometime." i

- Nolan Bushnell

Innovative leisure. It was a concept that Atari, under guidance of Nolan Bushnell, had cultivated for its entire existence. The engineer entertainers of Atari either invented or were driven to invent by competition, some of the most mind-blowing gaming creations of the 20th century.

"These people were my friends and co-workers and we were sort of united in this quest for cool stuff." ii

- Nolan Bushnell

However, dreaming up ideas is only one part of a successful business -- you also need to find customers to buy them. The objective pursuit of engineering cool stuff is almost entirely at odds with the subjective nature of marketing it.

"The marketing department had never played a video game... marketing thought the programmers were lazy, the programmers thought marketing was stupid... we didn't like them, they didn't like us." iii

- Rob Fulop (Atari VCS game developer)

When Atari was focused on coin-operated games, marketing was not as much of a factor. The small audience that needed to know about its games (arcade operators and distributors) could be reached fairly easily through publications like Replay magazine and Coin Connection, mail-outs of advertising flyers, and trade shows.

However, with the Atari VCS effort, the company was firmly working outside the confines of its old business, and it required a more sophisticated marketing effort than Atari could manage on its own. Warner Communications brought this marketing focus to the table -- as well as the vast amount of money needed to make Atari's ideas come to life.

"Warner put a lot of money into the company, which certainly helped pay for marketing and manufacturing the games and computers." iv

- Alan Miller

However, the sudden oil and water mix of engineers and marketers at Atari almost single-handedly created a computer age cliché that is now common place in many failed technology companies: engineers and marketers can't coexist peacefully for very long.

"If there is anything engineers despise, it is dumb marketers defining the impossible." v

- Nolan Bushnell

Instead of celebrating the successful launch of the VCS in 1978, Atari was lamenting missed opportunities and mistakes that held sales back. While management scrambled to find ways to save the consumer business, programmers struggled to come to grips with the new technology, and R&D looked towards the future.

The eternal struggle of the marketing and engineering was set alight, and put on course to explode within the company, taking many of the pioneers along with it.


As 1978 started, Atari was having trouble on several fronts. While the VCS had sold well during Christmas 1977 (upwards of 350,000 - 400,000 units), sales were stunted because of production problems that had the VCS units delivered late to retailers, resulting in a $25 million dollar loss for the period. vi

Atari was also hobbled with warehouses filled with unsold dedicated Pong units, the stagnation of the coin-op business, and an increasing divide between Warner brass and existing Atari management.

Even so, Bushnell was positive that, with the VCS, Atari had a winner on its hands. It just needed to find enough talent to make games for the system.

"I see us as having built a record player and now it's up to our creative people to decide how many records there will be." vii

- Nolan Bushnell

In his mind, if he could make it work, the sky would be the limit for Atari's game system. The profit potential for a system like the VCS was one of Bushnell's crowning achievements for Atari.

"The thought of taking something that cost $3 and selling it for $20, or selling it for $40, I take great pride in that as a concept." viii

- Nolan Bushnell

At the same time, Bushnell found himself butting heads with Warner's executive VP, Manny Gerard. Bushnell and Joe Keenan disappeared from Atari for some time after the VCS was released, but popped back in at times to give their opinions on the business. This frustrated Gerard.

"You can't disappear and walk in six months later and say 'let's do this.'" ix

- Manny Gerard

Also, Gerard was positive that Atari was spending far too much effort on engineering and R&D and not enough time trying to sell and market its products.

"They had no sales, no advertising, no marketing, nothing but R&D." x

- Manny Gerard

"We had a very powerful engineering team working on a lot of projects -- a lot more than Manny thought we should have." xi

- Nolan Bushnell

In February 1978, Manny Gerard encouraged Bushnell to find some help marketing Atari's products. When Bushnell was slow to respond, Gerard suggested Harvard educated Ray Kassar, an ex-marketing VP from Burlington textiles. Kassar was exactly the button-down, straight-laced businessman that Bushnell was not. Kassar began as a consultant, with his directive from Warner to find out if Atari should be dumped altogether. What he found was not encouraging.

"It was a disaster." xii

- Ray Kassar

However, Kassar was impressed with one thing at Atari: the VCS xiii. Instead of recommending liquidation to Warner, he set out to develop an integrated marketing plan that would save Atari.

At the same time, Bushnell and president Joe Keenan found themselves struggling to hold on to the company that they had created. Instead of leading Atari in new directions that would build the business even further, Bushnell was constantly clashing with Manny Gerard and Ray Kassar over the future of Atari's products and especially R&D.

"Where we became unglued was when Manny started killing the research projects. I saw that as building a very fragile company." xiv

- Nolan Bushnell

1978: The Languishing VCS

The VCS group started 1978 with a crew of about 12 programmers, many of them new hires -- including David Crane, Jim Huether and Warren Robinett. They all found the consumer division to be a rewarding place to work.

"We had a lot of fun, Warner had owned it for a while, but Nolan was still running it. He's an engineer, and he ran the company as an engineer would run it and that's why Warner bought it. But he would still isolate the engineering department. He'd say, 'You guys go over there and have a lot of fun. We'll come back and talk to you every once in a while.'" xv

- David Crane

"It was all sort of started by Nolan Bushnell who was the founder and he kind of instilled this family friendly feeling, parties every Friday... but you had to get your job done or otherwise you didn't hang around too long." xvi

- Jim Huether

"It was a cool job to have. For me, it was like dying and going to heaven." xvii

- Warren Robinett

The pressure in the consumer division to come up with new games and get them to market to support the platform created a department with a different face than the R&D heavy Coin-Op division. While it was still a casual environment, there was a cut-throat edge.

"People in the engineering group worked very hard. It's true that it was a casual environment and the kind of clothes you wore was not important in the engineering group. Results were what counted." xviii

- Alan Miller

"(It was) an environment with little discipline, but yet with clearly stated goals." xix

- Nolan Bushnell

VCS games took roughly 6 months to create from start to finish, with each programmer working mostly alone on their projects.

"When I started they just said, 'We want you to do a game in about six months... you have no set hours, we don't even want to see you until the game is almost done'... It was great." xx

- Jim Huether

Since the VCS was so difficult to program, only people truly dedicated to their projects were successful.

"I believe that Atari in the early days succeeded because the games were labors of love by the programmers who worked on them. At least that was the case with my games for me. In those old far-off days, each game for the 2600 was done entirely by one person, the programmer, who conceived the game concept, wrote the program, did the graphics." xxi

- Warren Robinett

The dedication of the programming team started to show as the sheer number of releases for 1978 rolled out (eleven titles in all), even if the games themselves didn't set the world on fire.

Since most of Atari's original coin-ops were simulated in the VCS launch titles, the second wave of games suffered by comparison. While there were still a couple of notable coin-op translations, many games from second wave were based on traditional games, and not their coin-op brethren.

Space War, programmed by Ian Shepherd xxii and released in May, was a version of the original Computer Space coin-op. However, the button controls of the coin-op version did not prove to translate very well to the VCS.

"The game is uncontrollable. Joystick movement is sluggish and clumsy, and likely to drive an experienced player to irritation."

- The Book Of Atari Software, 1983

Hangman was also released in May. Programmed by Alan Miller, it was notable because it required more ROM to store data for the game than any previous cartridge.

"My game, Hangman, was the first 4K byte cart for the VCS, but the extra space in that cart was simply used to store additional words. Being the first one to use that part, I had to electronically qualify it with the ROM vendor, Synertec, to make sure it met the timing and current requirements of the VCS." xxiii

- Alan Miller

Home Run, released in June 1978, was one of the first home versions of baseball ever attempted. Programmed by Bob Whitehead and David Rolfe, it was a simplified version of the sport that, for nearly five years after its release, remained the only 1-player version of baseball available on any home video game system.

"Jamming a baseball program into 2K of memory forced the designers to prune away many of the elements of real sport. There are lots of factors that worry Earl Weaver that won't lose managers in Home Run a minute's sleep." xxiv

- Arnie Katz and Bill Kunkel

Codebreaker, also released in June, was a version of classic game Mastermind plus the logic game Nim, Hunt & Score, released in June and programmed by Alan Miller, was another attempt at a traditional game brought to the VCS.

"The title of an Atari VCS game I programmed called Hunt & Score was also later marketed by Atari under the title A Game of Concentration. In the early days of the VCS, Atari frequently changed the cart titles for games marketed under the Sears Tele-Games brand. Hunt & Score was called Memory Match for Sears." xxv

- Alan Miller

Slot Racers, programmed by Warren Robinett, was released in July. On paper the game sounded exciting: "A pair of skillful drivers with the killer instinct drive up and down the streets of a city firing hood mounted cannons at each other." xxvi However, the execution left a lot to be desired. The cars resembled shoes, and the "city" was no more elaborate than the maze in the Tanks variation of Combat!, a game that all VCS owners already played for free.

"It would have never gotten published in any normal situation, but Atari needed product and published everything the programmers produced in 1978." xxvii

- Warren Robinett

Brain Games was released in August and programmed by Larry Kaplan. It contained 6 different thinking games, including the VCS conversion of the Touch Me coin-op. It was successful in execution, but its subject matter did not pose to set the world on fire. Polo was developed by Carol Shaw as a cross-promotion with Ralph Lauren (also owned by Warner Communications). It played like a two-player soccer game with horses. The game was finished, but never released.

However among the misses and near hits and unreleased gems of 1978 were several classic VCS games that showed the system had real promise. Outlaw, programmed by David Crane, was more like Midway's Gunfight! than Atari's own Outlaw coin-op game.

Flag Capture was programmed by Jim Huether and released in August. Different from nearly all other VCS titles, it was an engrossing strategy-style game. Your job was to find your flag before your opponent on a grid of squares. Each time you clicked on a square it would give you a clue as to where the flag might be.

"This was the first game I did for the 2600, so it was a big learning experience. It was difficult to do a game in 2K bytes of ROM, 128 bytes of RAM including the stack, and writing to the screen on the fly. I wanted to do something like Stratego, but realized I couldn't do it on a single screen. So I turned it into a capture the flag type of game. It took about six months to complete. The graphics were pretty bad, but the gameplay was very good." xxviii

- Jim Huether

Basketball, programmed by Alan Miller (himself a basketball player) and released in December was a bit of breakthrough with its pseudo-3D playfield and furious 2-player action, even if the ball was a square and animated players looked a bit silly.

However, there was one VCS game that stood out among the rest: Breakout. Released in November, just in time for Christmas, Breakout was a mostly true-to-form translation of Atari's hot coin-op from 1976. Brad Stewart won the right to program the game by beating fellow programmer Ian Shepherd at the coin-op version of the game in the Atari break room. xxix

"Another programmer, Ian Shepherd, became available at the same time. Since Breakout was one of the titles we were going to do, and since there was a coin-op Breakout game in the coffee room, Ian and I decided to play for the coding rights. I can't remember which of us went first, but I managed to knock down both walls of bricks with one ball, then leave the game in "lock up" mode where the ball continues to bounce off the same place on a motionless paddle and retrace the same path over and over. Ian missed when it was his turn to play, so the coding rights went to me." xxx

- Brad Stewart

Ray Kassar's aggressive marketing plan for the VCS showed its fruit in Q4 1978, when Atari's first slogan-filled commercials ("Don't Watch It, Play It") arrived on TV screens across America.

Not only could people see Atari VCS games in their homes, but could see actual celebrities such as Don Knotts, Pete Rose, and Kareem Abdul Jabaar shilling for VCS games (Breakout!, Home Run, and Basketball respectively). In total, Kassar spent $5,000,000 on advertising in 1978. xxxi

Kassar also insisted on creating a QA program to help calm nervous retailers' concerns about defective returns, and used his extensive experience in manufacturing to ramp up VCS inventory for Christmas 1978.

However, it was not just Kassar that helped make the push to get VCS units manufactured for Christmas. Nolan Bushnell pulled out all the stops with his employees to get the products manufactured and shipped. Bushnell rounded up managers, supervisors, and other corporate, salaried employees to help fulfill all orders. Most worked four to eight hour shifts beyond their own work day. xxxii

The other good news for Atari was that Nolan Bushnell's strategy to tie-up the n-channel semiconductor manufacturers with Atari work to stave off competition was holding. No major for the competitor for the VCS arose in 1978, and Atari sailed into the Christmas season with 800,000 units ready for sale.

1978: VCS Follow-Up: Colleen & Candy

When work started on the follow-up to the VCS in 1977, the engineering team Joe Decuir, Jay Miner, and Steve Mayer were tasked to create a machine that could both follow up the VCS, and double as an entry into the burgeoning personal computer market.

"(The follow-up had to) support 1978-vintage arcade games. We knew we would need to leapfrog the 2600 before somebody else did. (It also had to) support home computer character and bitmap graphics. We saw the Apple II, Commodore, and Radio Shack appliance machines coming." xxxiii

- Joe Decuir

The project was soon split into two separate projects, dubbed "Colleen" and "Candy" after two particularly attractive secretaries These products were later re-named the Atari 800 and 400 respectively. "Colleen" was to be the full-fledged computer, while "Candy" was more suited as the game machine follow-up to the 2600.

Both were based on the same basic hardware design with four separate silicon chips that handled different parts of the computer's operation: the 6502 CPU running at 1.8 MHz, the ANTIC display microprocessor, CTIA graphics chip, and POKEY sound chip. The power of these multiple processors pushed the 8-bit computers power far beyond that of the VCS.

Jay Miner, as system architect, led a group that included Joe Decuir, George McCloud and Francois Michel that designed the ANTIC microprocessor for processing display information and the CTIA graphics chip to put it on a screen xxxiv.

It was a very potent combination, giving the Colleen and Candy the most sophisticated graphics for any microcomputer at the time. The ANTIC took graphics information in the form of Display Lists. Display List Interrupts allowed the screen to be cut horizontally into multiple parts, allowing for almost limitless display options.

These instructions were processed and displayed by one of multiple graphics modes by the CTIA (later GTIA) processor. The design also included hardware based sprites (Player-Missile Graphics) for creating games, a character-set that could be completely redefined in code, and per-line fine scrolling. xxxv In terms of sheer horsepower, the graphics capabilities of these new machines made the output of the TIA chip in the VCS look primitive by comparison.

After the 6502, ANTIC and CTIA, the 4th chip of the design was the POKEY. POKEY was a dedicated sound processor started by the core team and finished off by Doug Neubauer.

"The Atari 800's architecture evolved as an upgrade of the 2600. Conceived primarily by Steve Mayer, Joe Decuir and Jay Miner before I arrived at Atari, the original plan for the POKEY chip called for keyboard interface, audio and paddle controllers." xxxvi

- Doug Neubauer

The chip had four distinct sound channels, with the ability to set volume, frequency and waveform per channel. This gave Colleen and Candy sound production abilities far beyond the speaker beeps of most other personal computers at the time.

As the hardware was being finalized, Atari started working on the software required to get the computers up and running. After announcing late in 1978 that the new computer systems would be on display at the January CES in 1979, the scramble was on write software that would run the machines.

"Atari had been designing a personal computer for a couple years and had a group of programmers working on the OS for a long time. Atari then pre-announced that the computer would debut at the January 1979 CES. " xxxvii

- Alan Miller

To meet this date, Atari tapped some of the best programmers from the VCS team, Alan Miller, Larry Kaplan, Bob Whitehead, and David Crane to work on the operating system, while outsourcing the job of creating a version of the BASIC language for the computers.

"There is a period at Atari when there were no games coming from Larry Kaplan, Alan Miller, Bob Whitehead, and myself. As the most senior designers at Atari, we were tasked with creating the 800 operating system. This group, plus two others, wrote the entire operating system in about 8 months. A funny story from this time that Al Miller likes to tell has to do with the Atari BASIC cartridge that was to ship with the system. Atari had contracted with a young programmer named Bill Gates to modify a BASIC compiler that he had for another system to be used on the 800. After that project stalled for over a year Al was called upon to replace him with another developer. So, while Al is the only person I know ever to have fired Bill Gates, I suspect that rather than work on Atari BASIC, Gates was spending all his time on DOS for IBM. Probably not a bad career choice for him, do you think?" xxxviii

- David Crane

The BASIC language was finished by SMI corporation in time for CES, as was the internally developed OS.

"I'm very proud of the OS we created for the Atari 400/800. It was similar in complexity to QDOS -- the OS that Microsoft licensed a couple of years later from Seattle Computer Products and renamed MS-DOS for the IBM Personal Computers. However, the Atari OS was much better designed in terms of its user friendliness and it had a much, much richer graphics subsystem and many fewer bugs." xxxix

- Alan Miller

1978: The Coin-Op Business

At the beginning of 1978 Atari's coin-op business was still coming to grips with the microprocessor technology now used in most of its games, while pushing forward with game and cabinet innovations.

By that time, the coin-op group had expanded to four core teams (adding fresh talent such as Mike Albaugh and Ed Logg to the ranks), each with consisting of a team leader (usually a senior electrical engineer), plus a couple of programmers, a couple electrical engineers, and a couple technicians.

Unlike the solo work on VCS games, coin-ops were a team effort. The focus was still on technology, because the graphical capabilities of the hardware were still fairly primitive.

"Graphic and cabinet design were a 'shared resource', and 'graphics' meant physical stuff, not pixels. Teams did their own pixels, which we called 'dots' at the time, none of us having taken a computer-graphics class." xl

- Mike Albaugh

Atari started 1978 with an aerial action game named Sky Raider released in March. Still working from the "military" theme of earlier games, Sky Raider had the player fly over a scrolling landscape, dropping bombs on enemy targets. It was one of the first games to keep a high score (though without initials).

"Sky Raider will be a super attraction was first previewed at the ATE show in London where the overall reaction was superb" xli

-Frank Ballouz, Atari Marketing executive

One of Atari's earliest coin-ops from 1978 was Avalanche, programmed by Dennis Koble and released in April. Avalanche was a sort of a "reverse Breakout", in which rocks fell from the top of screen and had to be caught by the player.

Soon after, in June, Atari released the Fire Truck coin-op. Fire Truck was one of the first cooperative multiplayer coin-ops. One or two players would guide a hook and ladder fire truck around a city.

The game cabinet consisted of a seat for the driver with a steering wheel for the cab, and a platform directly behind the seat with another steering wheel used for the second player to control the ladder portion of the fire-truck.

"Fire Truck was a follow-up to Superbug. The idea was to create a two player cooperative driving game. In Fire Truck, one player drove the front of the truck, and the other player drove the rear. It was trickier than it sounds. " xlii

A single player version of the game named Smokey Joe was released in August 1978, intended for locations that didn't have room for the massive Fire Truck cabinets.

July 1978 saw the release of Skydiver, programmed by Owen R. Rubin. The game consisted of a player jumping from an airplane and guiding their avatar to hit a spot on the ground.

Rubin, a pinball fan, added a "pinball feature": the name of the game on the marquee would light up when you landed successfully. If you were able to light the full name, you would win a bonus life, points, or a free game. Rubin had to fight his superiors to get this added to the game. xliii

"I was a pinball fan, so this is the first video game to have a pinball feature. When I asked for it, I had to fight to get it added because they thought I was crazy."xliv

- Owen R. Rubin

When Ed Logg joined Atari in 1978, he immediately started working with Atari engineer Dennis Koble on the ill-fated and unreleased coin-op Dirt Bike. When that project was cut short, he moved on to Super Breakout, a follow-up to the established 1976 Atari hit Breakout. Released in August, Super Breakout was a sizable hit with nearly 5000 units sold. xlv

The game improved on Breakout in many ways: it was written for a microprocessor (6502) instead of discrete logic, and it included 3 game variations -- Double Breakout (two walls separated by a space), Progressive (continuous walls that scroll down the screen), and Cavity (two balls embedded in a standard wall).

"The original idea was to have six variations on Breakout... However, in actual play there was one overall favorite, Progressive Breakout". In the end we put three variations in one game: Progressive, Double, and Cavity Breakout." xlvi

- Ed Logg

Other coin-ops released in 1978 were Sprint 1, a 1-player version of Atari sister company Kee's Sprint 2, Ultra Tank, a version of Tank inspired by Atari VCS Combat, with eight game variations, and Orbit, a raster version of Space War developed for the European market by Owen R. Rubin in six to eight weeks. xlvii

As well, there were several notable unreleased games in 1978, many with military themes. These included Wolf Pack, PT Commander, and Captain Seahawk, plus the decidedly non-military Mini Golf.

Latest Jobs

Sucker Punch Productions

Bellevue, Washington
Combat Designer

Xbox Graphics

Redmond, Washington
Senior Software Engineer: GPU Compilers

Insomniac Games

Burbank, California
Systems Designer

Deep Silver Volition

Champaign, Illinois
Senior Environment Artist
More Jobs   


Register for a
Subscribe to
Follow us

Game Developer Account

Game Developer Newsletter


Register for a

Game Developer Account

Gain full access to resources (events, white paper, webinars, reports, etc)
Single sign-on to all Informa products

Subscribe to

Game Developer Newsletter

Get daily Game Developer top stories every morning straight into your inbox

Follow us


Follow us @gamedevdotcom to stay up-to-date with the latest news & insider information about events & more