The adoring crowd of 300 engineers executives, and reporters fell silent. They had gathered this Friday night in the bright new site of Sente Technologies Inc., in Milpitas, Calif., to pay tribute to their leader, a man many of them regarded as the philosopher-king of Silicon Valley. In the last decade, he had founded an industry and parlayed seemingly outlandish ideas into not one but two multimillion-dollar companies. Now he stood before them, puffing on his pipe and flashing his puckish grin.

"My name is Nolan Bushnell, but I'm not God," he told them. "I need to build factories."

If not God, then Prometheus -- about to be unbound. At midnight on September 30, 1983, a seven-year noncompete agreement, which Bushnell had signed when he sold his pioneering Atari Inc. video-game company to Warner Communications Inc. in 1976, was due to expire, and a grand party had been organized to celebrate his release. Streamers and banners adorned the walls and ceilings. Bushnell mingled gregariously, clutching a can of Budweiser and a little blue notebook with a picture of a clown on its cover. Waving the notebook about like a wand, he declared that it was bulging with ideas for video games far more creative than anything then on the market.

"The game business, much like the music business or the carnival business, gets in your blood," he pronounced. "For the past several years, I have been scribbling down ideas, waiting for today." Sente, the latest in his series of glamorous enterprises, would carry the video-game torch into the future. "My grandchildren and their grandchildren will still be playing video games."

The crowd believed. This was vintage Nolan Bushnell, the Valley's fun-loving, boyish genius, who summoned successful companies like rabbits from a top hat. After Atari had come Pizza Time Theatres Inc., a restaurant chain that expanded so rapidly it twice made INC.'s list of the 100 fastest-growing public companies in America. After Pizza Time came Androbot Inc., which had dazzled the press and the public with a mechanical-man prototype named Bob. There were others, too -- each one a piece of its creator's entrancing vision.

As for the creator himself, his charisma was unmatched anywhere in the entrepreneurial wonderland of Silicon Valley. "Nolan walks in the front door with that smile on his face and puffin' on that pipe, and it's like a whirling dervish walked in," says the former president of a Bushnell company. "People's hair stands on end. Their eyes get like saucers. And they flock around him like J.C. the Man just walked in."

Such, indeed, was his reception at the Milpitas party on that magic night in 1983. Only six months later, however, the magic was gone. By then, Pizza Time Theatres had filed for protection under Chapter II of the Federal Bankruptcy Code. Androbot, touted by Merrill Lynch in the summer of 1983 as one of the country's hottest new issues, had lost more than $4 million. Sente itself had been sold to Bally Manufacturing Corp. in a desperate attempt to raise cash for the parent company. And Nolan Bushnell had lost around $5 million from his own pocket, plus many millions more in the paper value of his stocks.

How did it happen? In answering the question, it is tempting to see Bushnell as he was in Milpitas, clutching his clown notebook and smiling under a shower of accolades: a sort of overindulged child prodigy. But there is more both to the man and to the collapse of his corporate domain. Bushnell's companies weren't ordinary start-ups built around one or another market opportunity. There was a unifying theme to them a mixture of technology and fantasy that Bushnell hoped would elevate him to the status of entertainment mogul the Walt Disney of the '80s, an entrepreneurial sorcerer whose visions permeated American life. That he ultimately turned out to be the sorcerer's apprentice, overwhelmed by his creations rather than in control of them, is his tragedy.

The story of the collapse is also a parable for Silicon Valley, a region increasingly obsessed with technology and troubled by the swift punishments now being meted out in the marketplace. Even more than Apple Computer Inc.'s Steven Jobs or Intel Corp.'s Robert Noyce, Bushnell was the spiritual leader of Silicon Valley. He believed, still believes, that one day all of America will in some ways resemble the Valley, a suburban Oz built from California redwood and Spanish tile, where giant billboards advertise computer spreadsheet programs and where the gas stations run promotional programs with names like "Protech." New technology has always been at the center of Bushnell's entrepreneurial vision; like others in the Valley he came to believe it would never fail him. That he was wrong seems no small lesson in a decade that will likely be remembered not for its wars or its cultural artifacts, but for the dazzling technology it has unleashed.

The picture in the newspaper captured Bushnell immersed in a hot tub next to a beautiful girl who could pass for half his age. Clutching his pipe, his face hidden behind a ragged beard, Bushnell looked in 1976 like a high-tech Hugh Hefner.

The image did not sit well with the directors of Warner Communications as they pondered the wisdom of acquiring Bushnell's Atari company. Warner's film and record divisions reportedly were doing poorly at the time, and the entertainment conglomerate desperately needed a strong performer. Atari was that, all right, but the video-game industry -- which Bushnell had virtually created when he invented Pong, an electronic tennis game -- was in its infancy, and while it was growing fast, some considered it a fad. Disney and MCA, two of Warner's richest competitors, had already passed up the chance to acquire Atari. And now Atari's founder and chairman, a lapsed Mormon, was wallowing in hot tubs like a blissed-out Hollywood star.

If Bushnell was not yet a national celebrity, he was made one by Warner's decision in late 1976 to ignore his antics and buy out Atari for $28 million. Only 33, he was now a rich man, with a personal fortune totaling some $15 million (his share of Atari's sale price) plus a stake in Warner Communications. And Atari was a highly visible company on the cutting edge of a popular new consumer industry. At the same time, of course, Silicon Valley itself was becoming an American Eden of technological genius and entrepreneurial daring.

Bushnell had taken to this garden like a man escaping the desert. Born to a religious middle-class family on the dreary flats of the Great Salt Lake, Bushnell arrived in California a conventional electronics engineer, with a wife, kids, and a suburban tract house. But in the early '70s, he began to tinker with new lifestyles as well as new ideas. On New Year's Day, 1974, just as Atari was beginning its explosive growth, Bushnell left his family. After the sale of Atari, he took to flying around the country in his own Lear jet and sailing on a yacht called Pong. He bought a 16-acre northern California estate surrounded by old stone walls and complete with riding stables, a swimming pool, and tennis courts. He also acquired homes in Paris and in Aspen, Colo.

Atari, too, became an emblem of the Valley's promise. Employees showed up for work at the company's Los Gatos headquarters at all hours of the morning, often dressed in blue jeans and T-shirts. At Pajaro Dunes, a coastal vacation resort, company executives gathered regularly for open bull sessions, where new, often wild ideas were bandied about in the ocean breezes.

Bushnell was the immensely popular leader of these New Age shock troops, and his growing legend attracted the Valley's most precious human resource: engineers. Trained in electrical engineering at the University of Utah, Bushnell had designed Atari'; first video games himself, laboring with logic boards and a black-and-white television in a converted bedroom of his house. He had faith also, the wild entrepreneurial faith to keep plugging away when his earliest inventions were scorned by investors. As Atari grew, he became less involved with hands-on engineering, but his enthusiasm for the company's technology burned brighter than ever. "The day you go to lunch without playing a game to decide who pays, you know the game has lost your interest," Bushnell explained to a Washington Post reporter who asked him about Atari's success. It was the kind of answer that endeared him to some of Silicon Valley's most talented engineers, who would later follow Bushnell from company to company as his empire grew.

He was not a good manager. He admitted it, and few resented it. It seemed a beguiling corollary to his creativity. Besides, his other half, his partner, Joe Keenan, did have managerial talent.

By the late '70s, however, the company was growing so fast that Keenan by himself could scarcely contain it. "Absolute chaos," said venture capitalist Don Valentine, one of Atari's underwriters, describing Bushnell's video-game company at this time. Some of the problems stemmed from the marketing of Atari's Video Computer System (VCS), the home console unit Bushnell had been developing when he sold the company to Warner in 1976. After the acquisition, Warner put up $120 million to develop and market the system, and Atari brought it out in the fall of 1977. Sales were sluggish at first, and remained flat through 1978, leaving Atari up to its ears in unsold consoles.

In late 1978, with the crisis worsening, Emanuel Gerard, the Warner executive who had engineered the Atari buyout, proposed a reorganization. Bushnell would remain a director of Atari but would be replaced by Keenan as chairman. Raymond Kassar then in charge of the company's home-games division would take over as chief executive officer.

Bushnell was dead set against the plan. The slow sales of the VCS were due to overpricing, he argued. Gerard countered that Bushnell had little experience in, and knew even less about, the national marketing of consumer electronics products -- a charge that was at least half true. The nasty climax came at a November budget meeting in New York City, in which Bushnell and Gerard showered each other with obscenities. A few days later, Bushnell was ousted from Atari. The following month, he resigned from the company's board.

And so the break with Atari was complete. Bushnell immediately set about building Pizza Time Theatres -- his second great success and his first great failure. I'm sort of a honeybee. I like to fly from flower to flower and pollinate them." The words actually belonged to Walt Disney, and the fact that Bushnell quoted them so frequently reflected nothing so much as his profound identification with the great wizard of family entertainment. After college, Bushnell had tried to land a job with Disney, but was turned down; later, in search of a buyer for Atari, he had turned to Disney again in the hope that he might become a part of his favorite company. But Disney's people weren't interested. Unable to join them, Bushnell decided to beat them. Maybe it was only a bit of whimsy that, to take on the proprietors of Mickey Mouse and Donald Duck, he would let loose a rat. But a rat it was, by the name of Chuck E. Cheese.

The first Chuck E. Cheese Pizza Time Theatre opened in May 1977, in San Jose, Calif., as a division of Atari. Early growth was slow: Eighteen months later, only six new restaurants had opened. Shortly before his final battle with the powers at Warner, however, Bushnell had bought back full rights to Pizza Time for $500,000. When he left Atari in early 1979, he set off to devote his full energies to the animation of Chuck E. Cheese.

Although the Disney influence was obvious from the outset, Bushnell's Pizza Time Theatre restaurants more nearly resemble Las Vegas casinos for kids than Walt's sublime fantasyland. In the theater dining room, families sit at picnic tables and devour pizza while a cast of large animal robot characters, led by head rat Chuck E. Cheese, performs loud cabaret routines based on original songs and dialogue. The music is punctuated by buzzes, bells, and whistles emanating from the nearby Fantasy Forest Game Preserve, an average 2,500 square feet of arcade games, video games, and kiddie rides. Most restaurants also sport a merchandise booth, where T-shirts and toys are peddled to customers.

What Bushnell seemed to admire most about Walt Disney was the reach of his company's entertainment empire. Indeed, he regarded the Disney company as a Pizza Time competitor, and took a close interest in its new ventures and overall performance. Yet as much as he was fascinated with Disney's success, he didn't follow Disney's formula. As Pizza Time Theatres' indentity evolved, it always retained the atmosphere of a neon-lit, cacophonous summer carnival -- much like the Utah amusement park where Bushnell himself had worked as a youth, hustling the crowd into milk-bottle-throw and ring-toss booths. Where Disney's fantasies were always distinguished by their innocence, Bushnell's never quite lost the undertone of hucksterism.

Hucksterism, of course, has never stopped kids from dragging their parents to arcades, and they promptly began dragging those same parents to Pizza Time Theatres. Chuck E. Cheese aside, the restaurants' early success owed much to the flowering of the video-game business, which Bushnell had seeded almost a decade before. Now, as Pizza Time rolled out, video games were becoming an $8-billion-a-year industry. In the first two years after Bushnell left Atari, the number of Pizza Time restaurants grew more than twelvefold, from 7 to 88. About half were company-owned, the rest franchised In April 1981, Pizza Time went public, issuing 1.1 million shares at $15. Within a year, as the chain mushroomed to 204 outlets in 35 states and three foreign countries, the price of the stock more than doubled.

As chairman of Pizza Time's board of directors with a substantial ownership interest in the company, Bushnell was now one of the richest men in Silicon Valley. "Business is the ultimate game in which you keep score with money," he was fond of saying, and he was a big winner once again. The only question was what rabbit -- or rat -- the sorcerer would pull out of his hat next.

Soon after Pizza Time Theatres had reached the height of its success, Bushnell went to lunch with Herb Caen, the well-known San Francisco Chronicle columnist. Bushnell arrived at the appointe place, a ritzy San Francisco restaurant, in a limousine. He sat down at a quiet table and, as Caen later reported, exclaimed, "I'm mad for robotics!"

Caen soon gathered that the robots Bushnell had in mind were personal robots, household helpers designed for consumers, not the usual industrial variety. Why, Caen inquired, would anybody want to pay some $2,000 for a robot?

"Companionship!" Bushnell exclaimed again. "Robots are good company. Nice to have somebody around the house. Suppose you're watching a football game on TV and you don't want to walk into the next room to get your pipe or a beer. You don't want to press a button for one of the servants, either. So you get the robot to do it.

"And then there's the ego thing. Wouldn't you like someone to come into your room in the morning and say, 'O great and omniscient one, are you ready for your coffee?"

Madness was perhaps an accurate description of Bushnell's attitude toward robots. For years, he had longed to build and market robots, but he had never managed to translate this fantasy into a sound business plan. Then, in the fall of 1981, a group of ex-Atari engineers approached him about funding a robot company. The group was led by Walter Hammeken, who had worked as an executive at several Silicon Valley companies, and included rack Larson, who had been a key Bushnell employee at Atari.

The group had lots of enthusiasm, but no money. Bushnell had both. He agreed to form a company, line up investors, and pay the engineers' salaries while he went to work on a marketing plan. "I guess that's the way the idea became his," says Larson.

Androbot Inc., as the company was called, was more than just a fun new plaything for Bushnell: It was part of the entrepreneur's evolving strategy for the '80s, his attempt to build an entertainment and consumer-product empire to rival Disney or even Warner Communications in size and scope. Pizza Time, for the moment, was the centerpiece of this strategy, Bushnell's point of entry into the hearts and minds of consumers. But Pizza Time was already a triumphant success, and Joe Keenan, who had left Atari, was running it on a day-to-day basis just as he had Atari.

Bushnell needed a new challenge, and Androbot could provide it. But, he wondered, why work with just one company at a time? His long-term vision called for many companies. Moreover, Hammeken's team wasn't the only group of Valley entrepreneurs to seek his money and expertise those days. And so, in late 1981, Bushnell formed a proprietorship that he called Catalyst Technologies. It was one of the most innovative business ideas ever seen in Silicon Valley.

Catalyst was conceived as a protective holding company that would eventually mass-produce not products but small high-technology enterprises. In early 1982, it began to provide wide-ranging support to a number of start-ups. Initially the support was to include seed capital, business plans, office space, accounting, and other operating services. Later, when the sponsored company was ready to fly on its own, it would get a second round of venture financing.

Androbot was the first Catalyst company, and Bushnell's favorite. Fueled by a $1-million investment that he arranged, it moved into the Catalyst building even before heat and electricity had been hooked up. Bushnell spent increasing amounts of time there, consulting with the engineers and developing a marketing plan. He was excited about the prospect of building truly personal robots, friendly mechanical pets that would sell for $1,500 to $2,000 and could fetch a beer when the owner "didn't want to press a button for one of the servants."

As it happened, Walter Hammeken had had quite a different idea: He wanted to build industrial robots, perhaps to serve as mobile guards or fire-watchers in plants and warehouses. Bushnell, however, was not to be stopped, and this time around it was he who owned the company. After a series of what Hammeken calls "severe disagreements," Hammeken left the company abruptly.

With Hammeken gone, Bushnell plowed full speed ahead. He recruited a new president for the company: Tom Frisina, a Los Angeles sales executive with a strong background in consumer electronics marketing but no experience in Androbot's kind of engineering. Frisina recalls that his own attitude at the time was: "I know what makes people tick in homes. I know how to motivate them to buy things. Let's build some personal robots that can satisfy people in their homes. And let's not worry about how difficult they are to build, because we'll find some way of licking the problems."

That, of course, was the Nolan Bushnell way. Hundreds of people were crowded onto the Las Vegas hotel convention floor, milling from exhibit to exhibit at the January 1983 Consumer Electronics Show (CES). Many of the visitors were congregated near Androbot's booth. There, the darling of the entire show, a three-foot-tall electronic creature named Bob, was walking purposefully about, talking to the guests and following them around like a three-year-old child. Bob was Androbot's first prototype. When Bushnell had unveiled him to 200 reporters from around the world at the start of the show, the reaction was electric: It was as if everybody's science-fiction fantasy had materialized on a makeshift stage in a Vegas hotel. Bushnell, who was manning the Androbot booth with Frisina and Catalyst president Larry Calof, rubbed his beard and said, "You know, this is going to be the next Atari."

When the executives returned to California, they were in high spirits. Bushnell called a meeting to discuss how the company should proceed, and told Frisina that he wanted to develop 10 different robots over the next 12 months. Some would be as intelligent and mobile as Bob, others would be less sophisticated but could plug into a personal computer; still others would be targeted to educational markets. Bushnell also wanted to draft a business plan that could be used to raise $13 million. Although the company had not yet sold a single robot, Bushnell wanted to take Androbot public.

Frisina had also been seduced by the adulatory crowds at CES. But as the executive responsible for implementing Bushnell's ambitious plans, he worried that Nolan wanted too much too soon. Ten products, he told Bushnell, was too many. "I should have taken a strong stand at that point," Frisina recalls in hindsight. "I should have said, 'Nolan, it's hard enough to do one product right in a start-up." Instead, Frisina compromised with Bushnell, committing the company to develop four separate products with nonoverlapping technologies. Bushnell, for his part, was satisfied with the compromise and not at all worried that four new robots was too many. He says now that he believed in early 1983 "that the lower-end three products were a real slam dunk." Slam dunk. As easy as that.

But even as he revved up Androbot for its public offering that summer, he was simultaneously preparing to make a Napoleonic return to the video-game business, following the seven-year exile imposed by the noncompete agreement with Warner. The vehicle was to be Sente Technologies, a company he had named with care. The name "Atari" comes from the Japanese game Go, and describes a move similar to check in chess. As Bushnell's publicists relentlessly pointed out, "Sente" is the only move in Go that can overtake "Atari." The noncompete agreement would expire in October 1983. Now, in January, just after he returned from the celebratory CES in Vegas, Bushnell closed a deal to buy the assets of Videa Inc., a game software company run by ex-Atari engineers, for $2.2 million. Bushnell made Videa and its engineers a part of Sente, and set the whole operation up as a wholly-owned subsidiary of Pizza Time Theatres.

It must have seemed to him, back in the winter of 1983, that all of his maneuverings finally were coming together into the grand master strategy, a tour de force of technological and entrepreneurial wizardry. Video games, in Bushnell's estimation, were the key to Pizza Time's success. But because of the noncompete agreement, Pizza Time had been little more than a national game distributor, in effect a middleman for the companies like Warner that manufactured the games. In October, though, Bushnell would be free to create his own games and distribute them exclusively through hundreds of Pizza Time restaurants. What's more, Bushnell did not intend to sell his Sente video games, as most other game manufacturers did; instead, he would lease the games' hardware and software to distributors and retailers.

The model for the entire system was the pre-1950 film industry. Back then, the major movie studios not only controlled the pictures they produced, they owned the theaters as well. "It was such a good business that it ultimately was broken up by the Justice Department," Bushnell says. "I felt that the analogy was such a powerful one, of having unique, exclusive products in a retail establishment, that it was just an overwhelming profit potential." And indeed, if one accepted the idea that video games were equivalent to motion pictures, Bushnell's new empire would have it all: theaters, games, even a pizza concession.

"We're talking about bedazzling new games that will also be educational. We're talking about putting a player inside the video game. What we come up with will be well worth waiting for," Bushnell proclaimed. Even Androbot had a crucial place in the entertainment juggernaut he planned to unleash. At the same time he introduced his revolutionary new video games, Bushnell intended to debut robot table service in Pizza Time Theatres. Such an innovation, he said, would be "astounding."

The truth about magic is that it works: A man's fantasies can change the world. But the truth about the world is that it won't stay under anyone's spell for long. As early as the summer and fall of 1982, sales at Pizza Time's restaurants were "relatively flat," and in mid-March of 1983 the company warned that first-quarter earnings would be lower because of a 20% drop from the previous year in average store sales. Pizza Time executives blamed the hit movie E.T. and exceptionally rainy weather for the poor showing. In fact, the flat sales reflected a national waning of interest in video games. Despite Pizza Time's explanations, the price of the company's stock dropped dramatically; The Wall Street Journal's "Heard on the Street" column reported analysts' concerns that Pizza Time's concept was wearing thin with consumers. The analysts, said Bushnell, "have their heads up their asses . . . . You're talking to fools, or people who are totally misinformed."

He could not so easily dismiss the lawsuit filed by Atari around the same time. The suit charged that he was violating his noncompete agreement by priming Sente Technologies for entry into the video-game business. Bushnell settled the suit in May, but the litigation hampered his efforts to get Sente off the ground.

At Androbot, too, there were signs that all might not go according to Bushnell's plan. The Las Vegas CES had proven to Bushnell that there was tremendous interest in consumer robots, but the show hadn't demonstrated the existence of a market. Moreover, Bob was nowhere near ready for manufacturing. As a "brains on board" robot -- that is, one whose memory and functions are entirely self-contained -- Bob presented monumental engineering problems, particularly if it were to sell for $2,000 or less. And once the glow of the CES reception had faded, no one at the company was entirely certain what tasks Bob should be programmed to perform. "That's when the questions came," a former engineering supervisor recalls. "'Great. It looks terrific. It's fun. But what will it do?"

The more Frisina's sizable marketing staff tried to find out what consumers wanted from a robot, the more problems it created for the engineers. Consumers wanted a robot to wash windows or mow the lawn. If they wanted a pet, Frisina's research seemed to indicate, they would go to the pound and save $2,000. Bushnell himself was aware that the robot somehow needed to be useful, but occasional remarks -- like the one about pushing buttons for a servant -- suggested that his understanding of ordinary consumers' needs might be limited.

In any event, there was no way Androbot's engineers could come up with a manufacturable robotic handyman under the gun of a six-month deadline. Some of them doubted they could create such a product in six years. Bushnell's decision to create four different robots simultaneously, moreover, had sown dissension among the engineers. By the new, post-CES schedule, Androbot's first product would be a "brains off board" robot named Topo -- a creature that needed a personal computer to operate it. It was hoped that Topo would be sent to market in late spring, carrying a price tag of $800.

But few of the Androbot engineers were interested in Topo; they had signed on to build a real robot, not a computer accessory. In addition, precious engineering resources were now being diverted to develop the company's two other products: Fred, a brains-off-board robot targeted for the educational market; and Androman, a more ambitious consumer robot, described as a kind of three-dimensional video game. Two camps of engineers began to form around Topo and Bob, severing the spirit of comradery and adventure that had characterized Androbot before CES. This splintering led to quality-control problems: When the prototype of Topo was finally marketed in May, a large number of the robots were "dead on arrival" and many others quickly developed engineering snafus.

Bushnell was largely unaware of the dissension; he and Androbot's other top executives were preoccupied with financing the company. Frisina, Androbot's ostensible day-to-day leader, spent three solid months working almost exclusively on taking Androbot public. He, too, lost touch with product development.

Bushnell, however, had good reason to raise cash quickly: Androbot was burning his own money at a rate of several hundred thousand dollars a month. Frisina's business plan called for Androbot to go through about $10 million before the company broke even. Without outside financing, most of that would be Bushnell's money. A successful public offering would repay him about $2 million of what he had loaned Androbot. So Bushnell approached his investment bankers at Merrill Lynch, who in turn introduced Frisina to the firm's Capital Markets Group.

"It was pretty heady," Frisina recalls. "[Going public and getting rich] is the Valley philosophy. That's what makes this place tick. . . . The Merrill Lynch guys looked at the Androbot issue as a great way of getting into the high-tech [market] and making a real stand for themselves. These guys had been accustomed to dealing with smokestack America all their lives. And they wanted desperately to get into the Valley and make their mark."

As it happened, this was not the time for anyone to make his mark in the Valley. The first sign of the coming high-tech stock collapse appeared in early June of 1983, when Activision Inc., which at the time of registration was $17, came out at $12 and dropped to $9 in only six weeks. In July, just as Androbot's IPO window (the one-month period in which details of a new stock offering are circulated to investors) began, the electronic game and home computer industry began to crumble. Atari, Coleco, Mattel, and Texas Instruments all reported losses or disappointing earnings. Then word of serious troubles at Osborne Computer Corp. sent a sudden chill through the high-tech market. Retail investors had initially shown great interest in Androbot's $10 issue, but Frisina and Bushnell soon found that institutional investors wouldn't touch it. The offering was scheduled for the first week of August 1983. That week, Merrill Lynch withdrew the offering.

"And there we were, sitting with 105 employees, four products, and we were running out of money," Frisina remembers.

For a long time, Nolan Bushnell's head had been high above this gathering storm. Now, for the first time in his professional life, things began to look desperate. In May 1983, Pizza Time Theatres had sold $50 million worth of 8.25% convertible debentures, partly to finance its accelerating growth. A month later the company was forced to announce that it would do no better than break even for the second quarter of 1983, and that its research "indicated that our customers want better service and an improved food product." In July, stockholders were told that Pizza Time hadn't broken even in its second quarter after all. Rather, it reported a loss of more than $3 million.

"I was not planning to spend a significant amount of time at Pizza Time," Bushnell recalls. "I thought it was pretty much on automatic pilot. Then right after the middle of the year, it was very obvious that things were wildly wrong."

Wildly wrong, indeed. While fashioning a plan to make Pizza Time Theatres the center of a family-entertainment empire, Bushnell had forgotten about one thing -- pizza. By the spring of 1983, it was apparent that the company's "cardboard" pizza, as a former top Pizza Time executive described it, was turning customers away. So was the accelerating national decline in enthusiasm for video games; in just a few years, the industry had seen sales drop by $4 billion annually. Without a quality food product to fall back on, Pizza Time was rocked by the collapse. Declining customer counts at its restaurants, moreover, gave credence to speculation that consumers, especially parents, were tiring of Chuck E. Cheese and his friends.

While Bushnell was spending a great deal of his time at Catalyst, Joe Keenan and Pizza Time's board of directors were trying to respond to the assault on the company's revenues. They closed unprofitable stores, reevaluated their growth plans, and introduced a new pizza recipe. The 200-restaurant chain, insisted Bushnell, was "never a food concept." He wanted to attack the problems on his own terms, and he moved quickly to unveil his last weapon, Sente Technologies. Games, new and better video games -- they were the one remaining hope for his grand strategy.

Nolan Bushnell thrust his pipe into the air. "Free at last, free at last! Thank God I'm free at last!" The crowd at Sente's launching party in Milpitas on September 30 whooped and hollered. But if the party represented one of Bushnell's last opportunities to bask in the glory of his accomplishments, it was also a well-orchestrated attempt to distract public attention from the problems that were leading Pizza Time Theatres toward disaster. In early September 1983, Joe Keenan had announced his resignation as CEO "for personal reasons," and had moved to an office at Catalyst to pursue other projects. Bushnell had retaken day-to-day control of Pizza Time's operations, and had announced a major reorganization of the company into four divisions: restaurants, Sente, Zapp's, and Kadabrascope. Zapp's was a singles-bar restaurant concept in early development; Kadabrascope was an animation project that Bushnell hoped would develop state-of-the-art cartoons for Pizza Time restaurants. During 1983, he bought a $1.5-million computer for the division and hired a team headed by a renowned ex-Disney animator named Jack Nichols to run it.

What was taking place backstage, however, was a story considerably more dramatic. With Androbot decimated and Pizza Time bleeding cash, all that remained of Bushnell's grand strategy was Sente. But few, even within his own company, shared his nearly obsessive belief that Sente's games could save the day. Keenan, in particular, disagreed sharply with Bushnell's intention to lease his video games rather than sell them, a decision that would do little for the company's cash-flow problems. Keenan argued that Pizza Time was in no position to bank on the long-term success of the games, that it needed to get its money out of Sente as fast as possible. But Bushnell was not about to mortgage the future of his entertainment empire. He suggested a compromise: sell the games' hardware to distributors, but rent the games themselves as replacement software cartridges, like movies on videotape. Keenan remained adamant: The company could no longer afford to engage in marketing experiments. There was, in the words of one company director, "a lot of conflict." The conflict led Keenan to resign.

"Maybe Nolan really expected a masterstroke with Sente," reflects Steve Lieberman, a Minneapolis video-game distributor and a Pizza Time director who resigned in December 1983. "In fact, it's really difficult to mix a manufacturing with a restaurant business, especially when the restaurant business has indigestion."

Lieberman was not alone in this view. By the time Bushnell took over as CEO, he was at odds with both his board of directors and his top executives. "It was pretty much Nolan versus the world," says Randall Pike, at the time the executive vice-president who coordinated Pizza Time's four divisions. "What we needed was to get a good food product in there. But his time all went toward Sente and the games."

The games. Where were they? By all accounts, Bushnell is one of the most dedicated, hardest-working executives in Silicon Valley. But there was not enough time in the day for him to exercise the kind of managerial control he seemed bent on at Pizza Time, Androbot, and Sente. His days were split inconsistently among the three companies and numerous other demands. Yet the more his companies spiraled downward, the less he was willing to delegate key decisions.

The previous winter, when he had first plotted the ambitious strategy that was now unraveling around him, Bushnell had planned to unveil Sente's first video games in October. Engineering delays and the Atari lawsuit sabotaged that idea. A new debut was set for December 9. Over at Sente, a team led by Ed Rotberg was toiling furiously to make that deadline with a game called "Snakepit." As time grew short, Pizza Time's cash problems were making Rotberg's job increasingly difficult. Tools and parts he ordered simply wouldn't show up. When he called suppliers to check on the delay, he often found that Pizza Time hadn't paid its bill in six months.

The company was now losing money at a rate of nearly $20 million a month. Customer counts at company-owned restaurants were plummeting. Randall Pike, Bushnell's second-in-command, was recommending that they shut down the vastly unproductive Kadabrascope division and sell off the Zapp's singles restaurants. But Bushnell refused to let Kadabrascope go. A Christmas cartoon would excite some holiday business at the restaurants, he argued, and Kadabrascope's cash needs were "minuscule" compared to the restaurant division.

Instead, Bushnell convened his board with increasing frequency -- once a week, at the peak of the crisis to wrangle over the company's problems. They met at Pizza Time's Sunnyvale headquarters where, for three or four hours at a time, Bushnell sat puffing on his pipe as he debated with the directors about how to keep Pizza Time from sinking. Some directors felt that Bushnell listened to their arguments and appeared to agree with their suggestions -- then when the board would reconvene the following week, it often seemed that nothing had changed. But what could they do? With 19% of Pizza Time's outstanding stock, Bushnell clearly represented ownership.

"Some of us heard things and interpreted them one way, and others of us heard things and interpreted them another way," says former board member Don Valentine, who owned 185,000 shares of Pizza Time worth more than $5 million at its peak. "Was it frustrating? Yes . . . . From my point of view, Nolan was just another player." Valentine refused to say why he resigned from the board in December, as did InterWest Partners venture capitalist Wallace Hawley. Steve Lieberman says he resigned because "board members shouldn't be in the position of running the company," and because he felt an increasing conflict of interest as a Pizza Time director and independent video-game distributor.

"How are you sleeping?" a Pizza Time executive asked Bushnell one morning at the height of the crisis.

"I'm sleeping fine," Bushnell replied nonchalantly. Other men might have cracked under the stress. By then, Bushnell was working 16-hour days. Androbot was in the midst of massive lay-offs and a bloody purge of part of the Frisina management team. Other Catalyst companies were proving to be a disappointment. These included Cinemavision Ltd., a big-screen TV company; and Axlon Inc., which made an unsuccessful data communications device. Over at Pizza Time, it was looking more and more as if Bushnell's beloved Sente would have to be sold, after all, in order to stave off a bankruptcy filing.

Nevertheless Bushnell pressed on with his plans. "Nolan is a very optimistic guy," Randall Pike explains. "He always seemed to think something brilliant was going to happen -- tomorrow. He tended to look primarily at the upsides and not quantify the risks . . . . [Pizza Time's restaurant problems were] sort of dull to him. There wasn't a quick high-tech answer to the problems."

Bushnell offers his own explanation. "I am not a person that sits around wringing his hands. I take a lot of personal responsibility for whatever happens, both good and bad. The Pizza Time situation, being my first what I'd call 'failure,' was a real disappointment. But I'm pretty resilient."

Back in early 1983, when he seemed at the top of the world, Bushnell was asked by Success magazine about the temporal nature of achievement and ambition. He puffed contemplatively on his pipe and told the story of Steve McQueen's character in the movie The Cincinnati kid: "He always beats the little black guy pitching coins, but then he loses the big game, and goes back, and the little guy beats him. Suddenly he can't beat anybody. I think of that often, y' know . . . ."

The end to Nolan Bushnell's big game, when it came, was mercifully swift. Bushnell had argued that December would be the month Pizza Time turned around. In past years, the company's business had been relatively slow during October and November, after kids returned to school, and had picked up again dramatically during the holiday season. In December, moreover, Sente's "Snakepit" video game would debut, and Kadabrascope's Chuck E. Cheese Christmas cartoon would be distributed to Pizza Time restaurants. The combination, Bushnell believed, would attract hordes of customers.

The hordes never came. Nobody noticed the Chuck E. Cheese cartoon. The "Snakepit" game, unveiled to the press in Milpitas on December 9, was not a flop, but it was hardly a success. Cash problems continued to worsen: Between September and December, Pizza Time lost more than $75 million, over $12 per share. Shortly after the first of the year, Bushnell announced his resignation as Pizza Time's CEO. "When there wasn't the recovery," Bushnell explains now, "I said, 'OK, maybe I don't have what it takes in this particular situation.' And also . . . other fish had to be fried."

By now, though, Androbot, too, was pretty well cooked. After the public offering was rescinded in August, Bushnell took greater control over the company's day-to-day operations. By January, most of Androbot's original founders were gone, as was much of Frisina's eager marketing team. Fred, the educational robot, hadn't gotten out of the engineering lab. Androman was sold for $1 million to, of all places, Atari, but the deal subsequently fell apart.

Bushnell retained a tiny engineering crew at Androbot, which continues even now to work on robot products. He fully intends to reintroduce one of them within the next several years -- "after what I would call a 'false start." The debacle, he adds, was; matter of too little money. "It's the only time I've been four and a half million dollars short on the cost of development. Hindsight, I probably wouldn't have done one of the products."

He predicts that Sente, too, will rise from the ashes. When Bushnell quit Pizza Time, Randall Pike took over as CEO. His first act was to shut down Kadabrascope. A few weeks later, Sente was sold to Chicago-based Bally Manufacturing. Bushnell remains Sente's chairman.

Even the $3.5-million sale of Sente, though, was not enough to save Pizza Time. "We had hoped Sente would be a bigger asset," Pike says. "In fact, it didn't turn out that way." At the end of March, Pike resigned. On March 28, 1984, Pizza Time Theatres filed for protection from its creditors under Chapter 11 of the federal bankruptcy laws. It owed 5,000 creditors more than $100 million. In anticipation of the filing, several institutional investors in Pizza Time filed a class action suit against the company's officers and directors, charging that they knowingly deceived the public about Pizza Time's condition when they sold the $50 million in debentures in May 1983.

What has Bushnell learned from all this? ln the face off a failure that would rock most entrepreneurs to their souls, he observes, "It was a good, serious lesson in spreading yourself too thin. I think the Pizza Time experience was very valuable to me. I think it's important to really understand time budgeting a little bit better, and to find some ways to monitor and to intercede at earlier times.

"I really think that, looking back, I basically delegated too much responsibility in Pizza Time at too soon a period."

Bushnell says that -- despite losing at least $5 million in the Pizza Time and Androbot fiascos alone -- his personal solvency has not yet been threatened. "Well, I can't say it doesn't hurt. But I have transcended that, and in fact, and this will sound very utopian, but I've always felt that I'd rather lose my money than [that of the] people who have invested in me. Fundamentally, my pride is more important to me than my money."

Since resigning from Pizza Time Theatres, Bushnell has busied himself at Androbot and Catalyst, where several companies are ready to launch products in 1984 and '85. He has also been working to expand the Catalyst concept around the globe, and expects to announce an offshore Catalyst proprietorship soon. "We think the Catalyst concept is working so well . . . we're expanding it and training it for various places," he says. "In fact, if we wanted to, we could probably set up three or four in the next little while."

One of Androbot's successors as the Great Catalyst Hope is a company called Etak Inc., which is working on a computer navigation system for automobiles. It works like a magic carpet: A driver has a television monitor on his dashboard that displays electronic maps of the area he is in. A blip on the screen indicates where he is going, and another blip shows where he is now. "All you really have to do is drive toward it [the blip]," Bushnell says. Once asked why anyone would spend $2,000 -- the anticipated initial retail price -- for such a device, Bushnell explained: "Let's say you're in your car and you want to go to dinner. You've got this box on the dash. You punch in 'Japanese,' then 'cheap,' then good sushi.' The box takes over and guides you to a place."

Where Etak will guide Nolan Bushnell seems these days a little beside the point.