Condensed History of Video Games: The Rise and Fall of Atari (1975-1984)

This is part of a series. Please also see:
Condensed History of Video Games: From Tic-Tac-Toe to Pong (1952-1975)
Condensed History of Video Games: Nintendo Seal of Quality (1984-1989)

We ended our discussion of early video games when the market for both home video game machines and uprights appeared. However, while the period between 1972 and 1977 witnessed a flourishing of home Pong machines and arcade video games, the market for home machines was necessarily limited by the appeal of the single home game available: Pong. It hit the market in 1975 and was an immediate success, but home video game innovation was caught up in the craze of this game, and stagnated for a few years.

The missing link was a video game system that could easily swap game programs using some tangible form. The home Pong machines of the mid-70s were not capable of this, and always played the same game each time. As addictive as Pong was, it failed to stay popular indefinitely. By 1977, the public was thoroughly bored of Pong and all of the Pong clones.

Magnavox Odyssey with original packaging

Ralph Baer's Magnavox Odyssey was capable of switching onboard games by inserting different PCBs into the console. It is important to note that these did not constitute unique games; all games were pre-programmed into the logic of the machine, and the only effect of the PCB was to connect the appropriate jumpers that would link the signal generators and logic for each game. No programming was done on these game "cartridges". A further weakness of the Odyssey was that it only displayed in black-and-white. In order to get "color graphics", the Odyssey came bundled with translucent colored sheets which could be taped onto the television screen. It sounds as clumsy as it was, and with only two TV sizes supported, it likely didn't add much weight to the cause of the old Odyssey. To summarize (and recap), when the first Pong home machines hit the market in 1975, they forced the unloved Odyssey off the market.

Odyssey "active card" that was to be inserted into the machine

Baer had some very innovative ideas, such as "active cartridges", which contained additional logic and modules that could implement features such as sound and variable position and speed of onscreen objects. This thinking became archaic as soon as someone could release a complete video game system that did not require updates from games. In order to make a successful home video game business, a company had to have a complete system that could be sold at a loss, and the games themselves should be a simple program stored in memory that could be read by the game system, so that each game would be cheap and the real money would be made from the game sales. The crucial step needed was the simultaneous invention of the microprocessor-based video game system and the ROM game cartridge.

Baer's ideas for developing the potential of the machine through hardware updates would be revisited later. Anyone remember the expansion pack for the N64? In today's world of constant development, it's often not good enough to put a system to market and let it ride for 6 years unchanged. Updates are expected. In the current seventh generation of game consoles, the Playstation 3 and Xbox 360 have modular designs featuring variable-sized HDDs. These systems have both undergone major revisions halfway through their life and still sell well in 2012. Nintendo's Wii is pursuing a different course: it is considerably cheaper than its competitors, but it comes in only one form and it is not designed to be upgraded or modified. Still, Nintendo does offer a bewildering array of peripherals. In this generation, based on sales, it seems Nintendo was right and Sony and Microsoft were wrong, since the Wii has significantly outsold both of its rivals, and at times in the past few years has outsold both of them combined, particularly outside the United States. As we shall see, the Odyssey approach of constant development is adopted most heartily by Microsoft, while the Atari approach of making the best system possible and then keeping it in the market until a full replacement is available, is what Nintendo does. Happily, in the modern world there is room enough in the market for both approaches.

Fairchild F8 microprocessor

But back to the Seventies. The first company to nearly get the recipe for a modern console right was Fairchild Semiconductor (part of Fairchild Camera and Instrument), a legendary semiconductor firm formed from the "Traitorous Eight" who left the draconian Shockley Labs in 1957 (which should be known as the location of the first transistor production). Fairchild developed their own F8 microprocessor in 1975, a few years after Intel invented the market with their 4004. The F8 was (surprise surprise) an 8-bit processor. It had some advantage for cost purposes in that it did away with an address bus, which made it cheaper, and added 64 bytes of scratchpad memory, which meant that very small programs (such as simple games) did not need outside RAM. It was a great microprocessor for a video game system, and that's precisely for what it was soon used.

Fairchild VES/Channel F

The Fairchild VES (Video Entertainment System) was released in August 1976. It will be remembered as the first video games system using a microprocessor and the first to use ROM cartridges as well as the first to have AI sophisticated enough for players to compete against a computer rather than against a human opponent. Thanks in part to the 250,000 sales of this system, the Fairchild F8 was the world's best-selling microprocessor from 1975-1977. When Atari launched their VCS (Video Computer System) in 1977, Fairchild immediately changed the name of their system from VES to Channel F. Fairchild knew that Atari was a more well-known video game and they did not want to be perceived as a rip-off unfairly.

The game system almost bears more in common with the earlier Pong consoles than it does with the latter Atari 2600, because of its unambitious games. In fact, in that vein, the Fairchild Pong game was extremely good and remains one of the best recommended games for the system. Contemporary reviews found most of the games mediocre, but the educational games were said to be very good and effective for children.

Nevertheless, it is surprising that it only lasted for a single generation, since it was technically the first. It did not enjoy much lasting success; few modern game players will be familiar with the Fairchild VES, or the latterly-renamed Channel F. Why is this?

1. Timing. Fairchild entered the market after the Pong craze, and it had difficulty gaining popularity as the public was largely unconvinced that video games had more potential for unique fun. Some refer to this period as the North American Video Game Crash of 1977; it was neither the first nor the last time that informed journalists and reviewers claimed that video games were a fad that had passed and should be abandoned.
2. Technical limitations. The Channel F had half the memory of the Atari 2600 and it could only take eight onscreen colors: white, black, and two shades each of red, green, and blue. Complex colors like orange or purple or others were not supported.
3. Shoddy controllers. The controller is one of the the most-recurring gripes from later reviewers who looked at this historical oddity. The controllers were upright joysticks with no base, so they had to be held constantly. The "cap" atop the stick was the only moving part; it could be twisted, pushed in various direction of motions, or pressed inwards. The controller had no other buttons. Most heinously, the controllers were integrated directly into the machine, so if one or both controllers broke, the system had to be replaced. What a design oversight!
4. Name recognition. Fairchild was a parts supplier that would have been well-known to manufacturers, but not to the general public. Atari, although a much younger company, had the advantage of  marketing its name directly to the public. It was largely the same advantage that made the IBM Personal Computer so successful compared to competitors. American buyers preferred familiar names to market new products.
5. Scale of commitment. Fairchild was happy to launch when there were no competitors, but they did not want to take the ruinous losses that Atari was driving them through during the price war of 1979. They concluded that video games were a fad that would die out, so they left the business rather than take losses from the game division.

Given that Atari was already planning to launch the Video Computer System with a microprocessor-based logic, and using ROM cartridges, the fact that Fairchild got a slight headstart on them did not mean that Atari copied anything from the semiconductor company. The main historical impact of the Channel F was not to force other game makers in the same direction (Fairchild did not have the clout to be the standard-bearer of the game industry) but rather to force Atari to speed up and release the 2600 as soon as possible.

Atari hit the market with its Video Computer System (known by 1982 as the 2600 or 2600 Video Computer System) in October 1977.  They had some long-term advantages over all their competitors, especially the Channel F.

1. More onboard memory. Atari had a separate RAM/IO module (MOS Technology 6532) which had 128 bytes, double the 64 bytes of the Channel F's on-chip memory. It was still miniscule, and sounds ludicrously small by today's standards, but it was just enough to work.
2. A better processor. Atari used the MOS 6507, which was a cheaper, simpler version of the 6502 which went into many of the first microcomputer builds of the 1970s. Although this was slower than the F8 (1.19 MHz versus 1.79 MHz), the 6507 was capable through hardware tweaks and continuous development to run increasingly ambitious video games. Very soon after launch it had surpassed the complexity of the most advanced Fairchild games.
3. Bigger games. Channel F carts were only 2 kB while Atari carts could be 4 kB, meaning more could be done in the game. Actually, the 6507 processor could address 8 kB but was cut down to 2 kB for money-saving reasons, but 4 kB carts were released later in the life of the Atari 2600, and were ran using special workarounds.
4. Better peripherals. Atari joysticks were not bolted to the machine. They could also be replaced with twist-top "paddle" controllers or special driving controllers for a few games. All of them worked very precisely for the given game; controllers were a strong point of the 2600. Atari even released a wireless controller in the early 1980s, almost 20 years before it became common in consoles.
5. More experienced programmers. Atari's best advantage came in the form of its people, who came from diverse backgrounds. In the pre-2600 era, many creative developers flocked to Atari for the opportunity to make it big in game development. Although the Atari 2600 was one of the most complicated devices for which to program, because of its quirks, many effective solutions were introduced whenever a more ambitious goal was desired. Unfortunately, the Atari management style did not allow individual programmers to receive recognition for their work or compensation appropriate to the magnitude of the massive sales they created. Consequently, many of them became frustrated, left, and never returned, especially those who founded Activision in 1980 and started the very idea of third-party game development.

Why the name 2600? The internal part number for the Atari VCS was CX2600. After the release of the 5200 in 1982, it was necessary to differentiate between the two, so they called the older system 2600. Why did it receive this part number? I really hope that such an iconic game system didn't just get it randomly. There is no reasonable hardware explanation for it: 2600 does not describe the processor speed in kHz, the memory space on game carts, the number of games released by then, the number of pixels displayed, or anything like that. I've heard that it's an homage to the phone phreakers who used a 2600 Hz signal to hijack long-distance Bell telephone lines. It's amusing to note that both Steve Jobs and Steve Wozniak (who was a phreaker with alias Berkeley Blue) worked for Atari at a stint during the 70s before they made the Apple I. But this explanation is unlikely, since Atari was never part of the hacker scene, and its management was rather conservative by the standards of tech company startups. The true reason may never be conclusively known.

Amusingly, the successors of the 2600 were also given the same numeric nomenclature, with their product names (Atari 5200 and Atari 7800) being mere multiples of 2600. That's almost a whole decade of being the number one game company, without spending a single dime or extraneous thought on the right names for its flagship game consoles. (But have times changed that much? PS1, PS2, PS3, PS4...)

En route to their dominance of the early 80s, losses by the company in 1978 caused it to seek outside support from Warner Communications. Nolan Bushnell left Atari that same year and founded Chuck E Cheese. His successor as CEO, Ray Kassar, was appointed by Warner and universally described as authoritarian by the old guard of Atari. He was utterly oblivious to any of the demands of his programmers, and described them as "high-strung prima donnas." In 1979 four of the most experienced engineers at Atari requested commissions on their million-selling video games and Kassar replied memorably: "You are no more important to the game than the guy on the assembly line who puts it together."

Important as the manufacturing line is, and with no intent to denigrate line workers, let's note there were only a few dozen people in the world who had the background and technical skill to write good Atari games. For the CEO to show this much ingratitude was a slap in the face of stunning magnitude, and was basically an invitation to resign.

The new hires by Warner had a background in business, and they lacked the engineering and artistic background that the original staff of Atari had possessed. As more outsiders were forced in, the original spirit was slowly crushed. Despite this, they were more eager than ever to squeeze every dollar possible out of the company's video games. Some have credited Kassar with single-handedly ruining Atari by prompting massive turnover among both programmers and management.

The 2600 launched with a price point of $199, but under Atari's business model, such a high price point was probably not necessary even at the start. By 1979 Atari was periodically slashing prices as low as $99, cutting profits for the sake of increased market presence and more game sales. The initial launch was not auspicious, but it got better each year until the roof caved in.

• 1977: Atari launched 2600. Sales by year's end were 250,000, which already beat out the Channel F, but it was below expectations.  There were nine launch titles, of which the most popular was Combat.  2600s made early in the 1977 production run were made in Sunnyvale, CA and can be identified as "Heavy Sixers" with their six-switch form factor and heavy RF shielding. The Heavy Sixers are still the most prized by collectors because of the solid construction and bulky materials. These are well-made systems and are the most likely to last well. Even in the late 1970s, it was becoming rare to see assembled modern electronic devices in the US, and this is one of the very few examples of a mass-produced, American-built video game system.

• 1978: Atari sold 550,000 units on production run of 800,000. The nation was proving still slow to come out of the crash of '77. Nolan Bushnell left the company amid disagreements with the new management regime. Some more unique content was released, like the popular Breakout arcade port. Production was outsourced to Hong Kong and this production run was known as "Light Sixers" for their lighter components, but still keeping the six switches and woodgrain of the original. The outsourced production models are likely to still be reliable, but they are far more common and much less valuable. (By the way, this is the model I own.) Fourteen games released this year. 

Not my favorite, but here it is: Adventure

• 1979: Atari outlasted Fairchild, which ceased to develop new titles for the Channel F and left the market altogether. In Christmas of 1979, Atari's strategy of attrition finally worked, because they captured virtually all of the increasingly growing market, and sold over 1 million units for the year. It was rumored to be the most popular Christmas gift of the year. Twelve new titles were released this year, including the famous Adventure, which I've never much cared for. Adventure was the first known example of a video game "Easter egg": a room in the game contains the name of the developer, Warren Robinett. Such high quality exclusive content was a major factor in the increased sales.

• 1980: Atari released Space Invaders in January as an exclusive home title; it was the major factor in selling more than 2 million units for the year. The systems were made cheaper with a redesign that included only four switches; the four-switch models are even more common and less valuable than the 1978-vintage redesign. A total of eleven new first-party titles were released this year, including Night Driver, the first first-person racing game. The first third-party games hit the market, spearheaded by Activision (made of former Atari employees who had chafed under Kassar), which released popular titles like Dragster, Boxing, Skiing, and Fishing Derby. Atari promptly sued Activision over the rights to develop games for a system that they did not develop. Although third-party game makers did help to sully the reputation of the 2600 later on, it is important to note that Activision, made of former Atari employees, is never included in this characterization by historians. Activision games from 1980-1984 are considered some of the finest games available for the Atari 2600 (including the legendary Pitfall) and have been sold in compilation discs on PC and modern video game consoles up to the present day.

Warlords for 2600: one of my favorites

• 1981: More than 3 million units were sold for the year. Only six new Atari games were released, but among them were some former arcade blockbusters: Asteroids, Missile Command, Warlords, and Video Pinball. All of them were well-received. More third-party games were released by Activision and Apollo, including Tennis and Freeway. On the other hand, Atari bumbled badly in its release of Pac-Man for the 2600, since its poor graphics and bad controls are a far cry from the joy of the arcade original. Pac-Man for 2600 was critically panned. Sales were still high, but failed to meet expectations. On all other fronts, spirits were high in 1981. They were predicting major success in 1982.

• 1982: More than 4 million units were sold for the year, bringing the total of units sold to over 10 million. Few who had been observing the video game landscape five years ago would ever have predicted that the 2600 would make it so big. In a landmark decision, Atari lost its case against Activision, and the floodgates were opened to any third-party manufacturer releasing games for any system they wish. The contribution of quality games from some outside makers like Activision were actually the bright point in the year. Atari failed massively in 1982 by releasing the poorly-conceived and buggy E.T. video game. The other major disaster was the new-found prevalence of poor third-party game makers, emboldened by the failed suit against Activision. On the positive side, Atari had minimized costs to the lowest level possible. Systems sold for an average of $125 nationwide but cost just $40 in raw materials. Games cost $4.50 to $6 in material cost and $1 to $2 in advertising cost, but were priced at an average of $18.95 and even at that price, the popular ones often sold out. E.T. was not one of these. Atari anticipated strong demand and made 4 million cartridges before sales begin; only half a million were ultimately sold. A store at J.C. Penney noted that in order to move the product, they discounted it five times from its hyped-up $49.95 launch price to just $1. Atari lost $100 million on unsold cartridges and the marketing blitz, earning just $25 million in sales. Atari also launched the ill-fated 5200, which was to be a competitor for the more expensive ColecoVision and Intellivision consoles. To that end, it incorporated many more buttons (a full 0-9 digit keypad) and an analog stick mounted on a controller similar in form factor to a TV remote control. Atari's biggest problem with the 5200 was that it never got enough unique treatment. It was fully compatible with 2600 carts, but since that system was selling better, it was often left to the more-expensive 5200 to get updated versions of old games. The 5200 was novel for incorporating a pause button, but it was otherwise critically panned for not having enough unique content, as well as its clunky controller.

• 1983: Although Atari published more than twenty titles this year including some previously popular licensed titles, the company started losing money throughout the year. They were forced to bury over a million unsold E.T. cartridges in New Mexico. Major disastrous third-party games of the year were the infamous Custer's Revenge, where the shockingly offensive object of the game is to lead a naked, erect Custer across the screen to rape a tied-up Indian woman. Crude mature games like Custer's Revenge caused untold damage to Atari's reputation as the company was seen to be allowing illicit content on the game machine that they had been advertising for years as clean family fun.  Idiotic product-placement titles like Chase the Chuck Wagon didn't help either. The size of the video game market peaked at $3.2 billion in Q1 of 1983. But by Q4 1983, Atari announced losses exceeding half a billion dollars and entered bankruptcy soon thereafter. The crisis is known as the North American Video Game Crash of 1983, although its full effects were not felt until 1984. Also in 1983, Atari also made one of the stupidest business decisions of all time: After nearly completing a deal with Nintendo for exclusive release of their Famicom (later Nintendo Entertainment System), Atari myopically severed the deal because of irrelevant improprieties committed by Nintendo's other partner Coleco. Nintendo did not attempt to renegotiate, as they saw Atari's fortunes fading fast. This was the last chance for Atari to remain relevant and they blew it. Nintendo would go it alone in 1985. Like with transistor radios, motorcycles, and finally cars, the Japanese makers of video games would independently seek to conquer the American market.

• 1984: Atari reeled from a completely destroyed home video game market. They pulled the 5200 from the market. Total industry revenue from home video games shriveled to about $100 million, having shrunk 97% from its earlier peak. Warner Communications broke up Atari and sold the computer and home video game divisions (Atari Corporation) to Jack Tramiel, founder of Commodore. Warner kept the arcade division to itself (Atari Games) but later divested it. Neither of these separate entities would ever have control over the video game industry again. By 1984, about 13 million Atari 2600s had been sold worldwide. The successor, the 7800, was announced in spring of 1984, but due to the continued lack of video game sales, Atari postponed the project indefinitely. It would not be launched until 1986. By that time, there was a new and overwhelming competitor with which to contend.

April 2014 update: This article was written in 2012 with the best knowledge I had at the time- it made reference to E.T. games being buried in New Mexico because that is what my research indicated. I included a small screenshot of the Alamogordo Daily News in the next post of this series (I did not think I had license to publish the full paper). The recent rediscovery of the trashed E.T. games in New Mexico is being covered by news agencies as a story as though it were simply a legend. The author is stunned at this revelation, since the dumping was reported at the time it occurred, and it should have been regarded as historical fact rather than some kind of myth. Locals of Alamogordo with long memories know the games were dumped there, and Howard Scott Warshaw's denials should never have been taken as fact. Will future generations regard the existence of The Wizard, a movie with a shameless tie-in to Super Mario Brothers 3, as a myth?