17
Solid State of Mind

On September 26, 1974, Bally executive vice president of technology John Britz led a delegation to Milwaukee and was ushered into a room at the offices of Dave Nutting Associates containing two Bally Flicker pinball tables that Dave Nutting had received from the company the month before. The two tables looked identical save for one tiny detail: the typical scoring reels on one machine had been replaced by a readout composed of light-emitting diodes (LEDs). Nutting demonstrated both machines, noting how they played and behaved exactly the same. Then came the dramatic reveal as the games were opened, and the one with the LED readout contained just a small circuit board connected to a few wires as opposed to the rat’s nest of wires, relays, steppers, and other parts one would expect to find in a pinball cabinet. The shocked Bally executives immediately began searching the room for the giant computer that must have been running this specially modified machine only to discover that all the components were arrayed on that single board inside the nearly empty cabinet. The future of pinball had arrived.

The Flicker demonstration came at the end of a period of transition for Dave Nutting. His original coin-operated amusement manufacturing company, Nutting Industries, entered bankruptcy in 1971 following the failure of its educational subsidiary,¹ after which he established a new firm with his head of manufacturing, Dan Winter, called Milwaukee Coin Industries or MCI.² The new company released an updated version of Nutting Industries’ Red Baron filmstrip target shooting game called Super Red Baron that incorporated electronic sound alongside a few minor updates and subsequently developed several variations on the theme over the next two years with names like Flying Ace, Blue Max, and Desert Fox.³

¹ Nutting, 2017.

² In Cash Box, Dave Nutting announced the move to MCI as merely a name change for Nutting Industries, but he has stated in more recent interviews that Nutting Industries went bankrupt and MCI was a separate company.

³ Former MCI employees Keith Egging and Dan Winter remember MCI doing a limited run on a Pong clone as well, but this has not been corroborated by any other sources.

The novelty game business proved volatile, so Nutting began searching for a steadier revenue source. Pinball was still considered the staple of the industry even with the advent of the first video boom and seemed a likely choice, but Nutting did not believe he could break the stranglehold of the big Chicago companies unless he came up with a novel approach. Therefore, he decided to explore a solid-state pinball design. Through his ill-fated Modec subsidiary at Nutting Industries, Nutting had experience deploying solid-state machines, so in late 1973 he asked one of his engineers, Duane Knudtson, to adapt that technology to create a solid-state version of MCI’s latest game, Airball, a novelty piece in which the player guides a ball through various targets by manipulating a column of air. The prototype never worked properly, so the game was released using traditional electromechanical components instead.⁴ Undeterred, Dave Nutting continued to attack the problem with a new hire named Jeff Frederiksen.

Born on July 5, 1944, Jeffrey Ellis Frederiksen attended St. Thomas College in Minnesota for three years as a math and physics major before transferring to the University of Milwaukee to pursue electrical engineering and computer science. He left school after a year to join the U.S. Air Force in 1966. Beginning his service as a communications technician, Frederiksen later became a keypunch operator for a mainframe on an Air Force base in southern Turkey. Frustrated by the lack of error-checking capability on the computer, he requested a program be created to handle the task. The base could not commit resources to the project, so a technician gave Frederiksen a programming handbook and told him to do it himself.⁵

Frederiksen left the Air Force in 1970 and took a job with a firm called Radio Communications as a technician repairing two-way radios while finishing his computer science degree part time at the University of Milwaukee. In 1973, while working for another communications firm called Ken-Com, he began consulting for MCI as it looked to diversify out of its filmstrip games into solid-state products. In October 1973, he joined MCI full time and designed a game called The Safe, in which the player turns the dial on a mockup of a safe door built into the cabinet in an attempt to crack the combination within a time limit.⁶ Within three weeks of joining MCI, Dave Nutting asked him to start preliminary work on a solid-state pinball game and a solid-state update to the IQ Computer. Both would be powered by a microprocessor.⁷

⁴ Nutting, 1983, 570–572.

⁵ Paul Thacker, “Jeff Frederiksen Interview (Part 1),” Bally Alley, September 29, 2011, http://www.ballyalley.com/ballyalley/interviews/Jeff_Frederiksen_Interview.txt.

⁶ Jeff Frederiksen, testimony, Bally Manufacturing v. Williams Electronics, No. 78-2246 (N.D. Ill), January 3, 1983, 63–64.

⁷ Nutting, 1983, 586.

By January 1974, Frederiksen’s work on a solid-state game system had progressed sufficiently for him to convince Dave Nutting to buy an Intellec 4 development system from Intel so he could build a complete solid-state pinball hardware around a 4004 microprocessor. The Intellec was still relatively new and hard to come by, but Nutting secured one by convincing his Intel rep that his firm was an advanced R&D unit for Bally.⁸ At this point, Frederiksen told Dave that continuing to work on both a pinball machine and a quiz game would not be possible, so they decided to concentrate on the Super IQ Computer because it could serve as a superior testbed for their theories on how to incorporate a 4004 into a coin-op machine. The quiz game was completed and demonstrated for Bally in June 1974, but the company was not interested in manufacturing the product.⁹

Soon after the Bally demonstration, Nutting found himself changing companies once again. Back in 1972, MCI’s marketing director, John Ancona, suggested a new tactic to compensate for flagging sales of the company’s novelty game products: emulating Aladdin’s Castle by opening a chain of shopping mall arcades under the name Red Baron. Before long, MCI controlled roughly 20 arcades stretching from Ohio in the east to Arizona in the west. By mid-1974 these arcades had proven so profitable that the company’s investors decided to terminate game development and transform MCI into an arcade chain.¹⁰ An operating company did not require a game designer, so Nutting was asked to leave.¹¹ He formed a new partnership with Frederiksen called Dave Nutting Associates (DNA) to continue designing games, which he funded by entering into a development contract with Bally.¹² For the next three months, the solid-state pinball system became the primary focus at DNA, with the ultimate goal of presenting a complete working system to its new corporate benefactor.

⁸ Donovan, 2010, chap. 4.

⁹ Nutting, 1983, 587–590.

¹⁰ Donovan, 2010, chap. 4.

¹¹ Nutting, 2017.

¹² Nutting, 1983, 582.

While the Bally executives who attended the Flicker demonstration in September were impressed by the Nutting system, the company ultimately decided it should develop its own solid-state hardware in house rather than pay DNA a royalty. Therefore, nutting needed a new partner to bring a solid-state game to market and turned to John Bilotta, a distributor who had been mentoring him in the coin-op business going back to the introduction of the IQ Computer in 1967. Bilotta introduced him to an Arizona company called Mirco Games.¹³

Mirco Games was a subsidiary of an electronic testing equipment manufacturer called Mirco founded by former General Electric employees John Walsh, Bruce Kinkler, and Bob Kessler in 1971.¹⁴ Before founding Mirco, Walsh worked for GE in Germany, where he was exposed to one of the most popular coin-operated amusements in France, Germany, and Italy: table soccer, also known as foosball. In 1967, he started a side business with another GE employee named Dick Raymond to ship foosball tables to the United States. Soon after, Raymond returned to the United States and established Arizona Automation to serve as a distributor for the foosball tables that Walsh continued to import from Germany.¹⁵

In 1971, importing tables became less economical after the devaluation of the German Mark, so Raymond began building his own tables domestically. Meanwhile, Mirco started looking for additional sources of revenue to fund its primary business and purchased Arizona Automation in 1973 to enter the coin-op business.¹⁶ The company’s first video games, Champion Ping Pong and Challenger, were uninspiring two- and four-player ball-and-paddle games that did not perform well at first, but after the company converted Challenger to a table-top format, Mirco rode the cocktail boom to success.¹⁷

After the introduction by Bilotta, Mirco signed on to produce a pinball game based on the Nutting system, though it chose to make its own modifications, including the use of a Motorola rather than an Intel processor. The resulting game, Spirit of ’76, attracted intense interest and large orders at the 1975 MOA show, but the game was still in the prototype phase, and Mirco suffered through various technical difficulties and production delays. While a few units were made available in November, they often failed to work correctly, so the company had to shut down production. By the time Mirco resumed shipping in March 1976, many distributors had cancelled their orders, so the company only sold 140 units.¹⁸ The solid-state pinball revolution was placed on hold. In the meantime, the microprocessor had a transformative impact on coin-operated video games.

¹³ Nutting, 2017.

¹⁴ Smith, 2016, 286.

¹⁵ “The Table Soccer Phenomenon,” RePlay, February 28, 1976, 17.

¹⁶ Ibid.

¹⁷ Smith, 2016, 287–288.

¹⁸ Ibid, 289–290.

***

In 1975, pool topped the inaugural earnings chart of the new coin-operated amusement trade publication RePlay, narrowly beating out pinball and finishing well ahead of video games.¹⁹ While arcade video games recovered somewhat from the market collapse in 1974, sales did not return to the heights of 1973, with an estimated 53,000 units sold.²⁰ Atari dominated the market with sales of 25,000 units, while Midway ran a distant second at 10,000. Next came Allied Leisure – newly returned to video games after taking a year off – at 7,500, followed by Ramtek at 4,000 units. Between them, these four companies controlled over 85% of the coin-operated video game market.²¹

The big hits of the year were Atari’s Tank and Midway’s Wheels, while the single biggest earner on location was likely a massive racing game Atari deployed through its Kee Games subsidiary called Indy 800. Housed in a 4.66 × 4.66 × 7.25-foot cabinet that accommodated two players on each side for a total of eight, the game incorporated a color monitor so that each player’s car could be distinguished by a different color. Supporting eight players required eight circuit boards, which pushed the cost for operators to buy a cabinet to nearly $9,000, but the game proved capable of taking in over $300 a week on average and reached earnings as high as $5,000 in a single month in some locations. Due to the immense size and cost of the game, Atari planned to manufacture only 200 units, but when those sold out quickly they built more.²² Even so, the production run remained low compared to other games.

In 1976, arcade video game sales surged to levels not seen since 1973 as the widespread adoption of the microprocessor allowed manufacturers to overcome the realistic limits of TTL hardware and displace the electromechanical novelty game market. Early driving games and target shooting games like Gran Trak 10 and Tank had been hampered by restrictions on the number of moving objects that could appear on the screen at once, as more gameplay elements meant more circuits, bigger boards, higher costs, and greater service headaches. In contrast, microprocessor-based games could transfer some, though not all, of this burden to software. This allowed manufacturers to reduce the cost and size of game boards and introduce more sophisticated gameplay and smoother animations than games created solely in TTL hardware.

¹⁹ “Replay Route Analysis,” RePlay, October 1975, 38. Replay was launched in October 1975 by veteran coin-op industry journalist Ed Adlum, who had covered the amusement beat for Cash Box since 1964. In Cash Box’s own survey for 1975, pinball narrowly beat out pool for the top spot. Both surveys placed pinball and pool well ahead of video games. “1975 Amusement Machine Route Survey,” Cash Box, October 18, 1975, C20.

²⁰ Frost & Sullivan, 1976, 87. This figure does not include the sale of cocktail-style games, which were estimated at 26,000.

²¹ Ibid.

²² Smith, 2016, 150.

Jeff Frederiksen at DNA pioneered the use of microprocessors in coin-operated video games by improving on his pinball system. Initially, Frederiksen attempted to use the same 4-bit 4004 processor found in the Flicker machine only to discover it was not powerful enough to drive a display. He turned to the 8080 microprocessor instead. Frederiksen created a hardware system around the chip in which a frame buffer housed in dynamic RAM contained the instructions used by the monitor’s CRT to draw images on the screen, with the microprocessor providing instructions on which individual pixels would be active in each frame. This system, the first bit-mapped display created outside of high-end computer research labs, allowed for more elaborate graphics and smoother movement and animation than any video game yet released.²³

After Frederiksen completed his system, he turned to one of his former instructors at the University of Wisconsin named Richard Northouse to recommend programmers. Northouse funneled the company two of his students, Tom McHugh and Jaime Fenton.²⁴ While Fenton worked on a home conversion of the Bally pinball game Fireball, McHugh joined Dave Nutting on a project to redesign Western Gun, which Bally subsidiary Midway had recently licensed from Taito. While the executives at Bally and Midway were pleased with the overall concept of the Japanese game, they were less impressed with the squashed, cartoony graphics they felt would not go over well with an American audience. DNA’s new microprocessor system appeared to provide a perfect solution to this problem.

Dave Nutting designed the new version of Western Gun himself and handed off his concepts to McHugh to implement on Frederiksen’s hardware system.²⁵ The duo redid the graphics so the cowboys were more realistically proportioned and eliminated the fixed playfield of rocks and cacti. Instead, one cactus appears on the screen before the first kill, then two after the second, then three after the third, and so on up to a maximum of six, which are slowly replaced by trees in later rounds. After five kills, a stagecoach also begins moving across the center of the screen to serve as an additional obstacle. Each time a player is killed, the words “Got Me!” appear over his head. The controls are reversed from the original, with a small eight-way joystick providing movement and a large lever used to change the angle of the gun. Though a slower paced game than its Japanese counterpart, the superior graphics and animations of the DNA version, dubbed Gun Fight, proved a huge draw, and the game was a sizable hit for Midway that sold 8,600 units.²⁶

²³ Thacker, 2011.

²⁴ Fenton transitioned from male to female in the late 1990s. At the time of the events in question, she presented as male and went by her birth name, Jay.

²⁵ Tom McHugh, interview with Ethan Johnson, April 20, 2017.

²⁶ Jarrell, 1976, 50. Gun Fight is usually identified as the first microprocessor-based video game, but there is another candidate for that distinction called PT-109 by Mirco Games, in which four players control boats and attempt to shoot each other while navigating a playfield full of islands, mines, and other obstacles. Mirco debuted the game at the same MOA show that Midway bowed Gun Fight, but which one shipped to distributors first is not known. Unlike Gun Fight, PT-109 was not successful. Smith, 2016, 288–289.

In 1976, DNA developed four more games for Midway using Frederiksen’s system. These included Tornado Baseball, in which the player graphics are superimposed onto a painted model of a baseball diamond through use of a mirror; Amazing Maze, in which the player navigates a complex maze drawn from over 1 million patterns within a time limit; and a target shooting game called Sea Wolf, programmed once again by Tom McHugh based on a design by Dave Nutting. Derived from earlier electromechanical target shooting games like Periscope and Midway’s own Sea Raiders, Sea Wolf challenges the player to peer through a periscope and fire at ships and other objects moving across the screen at various speeds in order to score as many points as possible within a time limit. Capable of earning over $200 a week consistently for weeks on end, the game proved so popular that Midway brought it back for a second production run in early 1977 after it had been discontinued.²⁷ The company ultimately produced 10,000 units of Sea Wolf, making it one of the best-selling video arcade games of the 1970s.²⁸

The fourth and final game DNA developed in 1976 was a driving game called 280 Zzzap. Unlike early driving games like Gran Trak, Speed Race, and Demolition Derby that portrayed the game world from a bird’s eye perspective, 280 Zzzap took advantage of the microprocessor to present the action from a first-person perspective instead, with smoothly animated white rectangles outlining an undulating road. The basic idea for this game did not come from DNA, however, but originated with a German named Reiner Foerst.

²⁷ “Sea Wolf Back on Midway Line,” RePlay, March 1977, 60.

²⁸ Bloom, Invaders, 1982, 16.

An engineer with both a master’s and doctor of engineering degree from the Institute of Technology in Darmstadt, Foerst was working at a German wire manufacturer called Trakus in 1971 when he took notice of early driving simulators, which consisted of a computer hooked up to a projection screen and the controls of a car for the purpose of training individuals in motor vehicle operation. Foerst became fascinated with the engineering behind these simulators and built one of his own. With both computers and displays being prohibitively expensive, he used analog circuitry and lightbulbs to achieve a primitive driving effect.²⁹

With the advent of the Pong craze in 1973, Foerst adapted his prototype to incorporate a monitor, though the machine remained largely analog. After completing two prototype units in 1975, Foerst attended the International Exhibition for Coin-Op Games in Berlin in March 1976 to display them under the name Nürburgring 1. The game generated some interest at the show, but due to a lack of money to fund a full assembly line, Foerst was only able to produce machines at a rate of one per week. By the time he had split from Trakus to form his own company to sell the game in September 1976, only 30 units had been sold.³⁰ This meant copying by a more well-equipped factory was inevitable.

The first company to notice Nürburgring 1 was a small American manufacturer called Micronetics, which grew out of a game repair shop called Amusement Device Engineering established by Bill Prast and Steve Holder in early 1974. In June 1974, the duo founded a new company called Digital Games to enter the booming cocktail market and experienced immediate success by selling over 7,000 units of various ball-and-paddle games.³¹ When that market collapsed, the company entered a period of financial difficulty, so in June 1976 the owners wound it up and established Micronetics to replace it.³²

Digital Games/Micronetics made several attempts to break out of the ball-and-paddle market but were largely undone by a lack of engineering or manufacturing experience. This situation changed in August 1976 when the company dispatched a recently hired engineer named Ted Michon to Germany to fix a defective shipment of its latest game, Air Combat. While in Germany, Michon saw one of the few Nürburgring 1 units at a bowling alley in Düsseldorf. By chance, Foerst himself was showing off this unit to his sons, and the two engineers ended up discussing the inner workings of the machine. When he returned to the United States, Michon created a digital version of Foerst’s game.³³

²⁹ Smith, 2016, 189–191.

³⁰ Ibid.

³¹ Ibid, 185–187.

³² Ibid, 191.

³³ Ibid, 189.

With Micronetics continuing to face an uncertain future, company management decided to license the game to Midway, which debuted its version at the newly renamed Amusement and Music Operators Association (AMOA) show in October. Designed by Jamie Fenton at DNA, the Midway game incorporated a microprocessor that allowed the white rectangles outlining the road to scroll quickly and smoothly to provide a better sense of speed than previous driving efforts. Midway originally planned to call its version Midnite Racer, but it secured a licensing deal at the last minute to name it 280 Zzzap after the Datsun 280Z automobile. Micronetics also showed its version of the game at the AMOA under the name Night Racer and released it in December, but the company folded soon after.³⁴

Atari also embraced the microprocessor in 1976. Like DNA, the company began experimenting with microprocessor technology in 1974. That year, Steve Mayer and Larry Emmons at Atari’s Cyan Engineering subsidiary launched a project to replace the electromechanical parts in a pinball machine with solid-state components. They began converting an old pinball game called El Toro to a solid-state design before switching the project to a Bally table called Delta Queen in the middle of the year.³⁵ The game was tested at a local pizza parlor, but Cyan proved unable to iron out all the technical glitches, so no game based around the system was released.

Meanwhile, Atari began hiring programmers so it could transition to microprocessor-based video games. The first programmer, Tom Hogg, was hired in 1975 and began working on an eight-player version of Tank as a follow-up to the highly lucrative, yet prohibitively expensive, Indy 800. Dubbed Tank 8, the game fit on a single circuit board rather than the eight required for its predecessor because every element no longer had to be rendered in hardware. Hogg also produced a microprocessor-driven trivia game called Quiz Show that debuted alongside Tank 8 in April as the company’s first microprocessor-based products.³⁶ Atari’s second programming hire, Dave Sheppard, tackled the company’s answer to Midway’s Tornado Baseball, dubbed Flyball, and then turned his attention to Atari’s version of Nürburgring 1,³⁷ which became a modest hit under the name Night Driver and sold roughly 2,100 units.³⁸

³⁴ Ibid, 191–192.

³⁵ Stephen Bristow, deposition, Bally Manufacturing v. Williams Electronics, No. 78-2246 (N.D. Ill), August 5, 1982, 12–14.

³⁶ Smith, 2016, 221. Both games were released under the Kee Games label, which persisted for several years after the merger with Atari as the company continued its fight against distributor exclusivity. The Kee engineering group was also kept separate until Atari split into consumer and coin-op divisions in 1976, and the Atari and Kee teams merged. Frank Ballouz, interview with the author, July 28, 2009, and Keenan, 2018.

³⁷ Smith, 2016, 228.

³⁸ “Atari Arcade Production Numbers,” internal document with release dates, prices, and units sold, c. 1999.

Atari’s most successful game among its early microprocessor releases was a variation on Gran Trak 10 dubbed Sprint 2, which was designed by Tank creator Lyle Rains and programmed by Dennis Koble. While conceptually similar to previous games in the Gran Trak series, Sprint 2 broke from convention by removing both the brake pedal and the ability to drive in reverse and added a fourth gear to transform the game from one of maneuver into one of pure speed that emphasized drifting around the curves of the track. The incorporation of a microprocessor allowed not only for smoother animation, but also for the addition of three computer-controlled race cars that moved around the track in a pseudo-random fashion. These cars served as additional obstacles rather than opponents, and the goal of the game remained completing as many laps as possible within a time limit.³⁹ Released in November 1976, Sprint 2 moved an astonishing 8,200 units and remained one of the top earning games on location for three years.⁴⁰ Despite this success, Sprint 2 was not Atari’s biggest hit of the year. That honor went to the last significant TTL hardware game released: a ball-and-paddle variant called Breakout.

³⁹ “The Making of Sprint 2,” Retro Gamer, no. 89, 2011, 38–39.

⁴⁰ Atari Production Numbers, c. 1999.