I have been writing and discussing a lot about the early microelectronics industry recently with other researchers. There are still a lot of mysteries about the formative period, before true foundations for single chip designs and mass production came into the picture. It’s been interesting to think about the true how and why people came to start using microprocessors as it wasn’t just the Apple story of rebellious democractizing of computerization. It first and foremost a tool by larger companies to create periphery devices, not to really by the heart of modern computing.
There are a lot of contenders for what might be called the first microprocessor game. Keith Smith has covered this on his blog in a series of posts and I will be covering two major contenders in my upcoming series. However there are stories that have yet to be synthesized and researched to a large degree, largely brought about (not that I’m bragging) from my uncovering of court documents and trade publication articles. We’ll go step by step to see where these experiments came from and what implication they have for the commonly accepted story of microprocessor mass market.
The Intel MCS-4 series first went on sale in mid-1971, providing the first commercial avenue for microprocessors (their status as ‘first’ in conception is disputed, but we’re going to fly past that right now). The 4004 CPU was of course made specifically to power a Japanese calculator and that may well have been the most intensive thing they ever expected it to do. Microprocessors were meant to be controllers for things like cash registers, never meant to do the grunt computing work of a standalone device.
However, it seemed that Intel did have some inkling of this possibility. Sometime in approximately 1972 they put together a demonstration of a visual display to show Magnavox, who of course were the licensees of the Odyssey technology. It’s unknown how Intel got contact with Magnavox, who created the demonstration, or any further details aside from Ted Hoff’s comments as above. “We talked to [Magnavox] about it and they were not interested.”
The replica display of this demonstration is shown several times in the video above. Running on a SIM-402 development board, a white spaceship flies against a black background in a diagonal pattern down the screen. Whenever it disappears off the edge of the right side, it will appear at the corresponding vertical line at the left side, though with no “wraparound” function as in Spacewar!. It’s not a game or even really a program more than a proof of concept. It’s running off the 4004 which is completely incapable of running any kind of sophisticated video game. It presented a lot more questions than answers, so Magnavox rejected it, though interestingly enough Intel would be Magnavox’s partner when they finally decided to go microprocessor with the Odyssey². That relationship doesn’t seem to have carried over though from these talks in 1972.
The Computer History Museum no longer houses this display, nor seems to ever have housed it. Ted Hoff is the only person who has ever mentioned this relationship or even acknowledged the demo’s existence. Ralph Baer says nothing about it in his book Videogames: In the Beginning, in fact noting that RCA’s microprocessor seemed to be the first candidate for the interchangeable system idea at Magnavox (Baer 118). The trail ends there with this small, off-hand quote buried in a presentation of more significant importance to most tech enthusiasts. Yet this was almost undoubtedly the very first concept of a commercial video game making it to software and I would certainly love to know more about it.
The first dated creation of a microprocessor driven game – perhaps more than a dozen of the first microprocessor-driven games – can be attributed to RCA. RCA is most famous to video game players for releasing the RCA Studio II, but that system was a legacy hardware going back to the creation of RCA’s first microprocessor in 1970. To spare the details, as much of it will be covered later in my series, Joseph Weisbecker led the creation of the RCA 1801 microprocessor and had also independently developed games about computer logic. Ergo, he wanted his FRED (Flexible Recreational and Eductional Device) computer system to have games (Weisbecker 1-2).
The very first game noted in Weisbecker’s notes is a tic-tac-toe game he created himself called Tic-Tac-Toe on November 22, 1971. This was before he even had a display, so it may not be best to call it a video game. Only a few weeks later though he had figured it out and on April 8, 1972 he created a NIM game called “BINIM” which used binary notation to play the classic thinking man’s board game. Erase, a variant of this, followed on May 19 and by this time he was able to start convincing his fellow engineers to help write programs with him (not that it was very hard) (Weisbecker pg 113-114; Hagley Museum and Library).
Paul Russo was the first, creating a checkers game on June 7. After that, a game called Space War (though no relation to the mainframe game, this was a fixed movement dueling game) was completed July 22 by a C. T Wu. By August 1972 Weisbecker and crew had created 18 programs and fully half of them were games (Weisbecker pg 114-115). Weisbecker didn’t restrict his thinking either. He made a prototype in May 1973 which featured a light gun, inspired by the Odyssey, attempting to expand the system’s possibility’s as it lulled in full production (Weisbecker pg 143-144; Hagley Museum and Library).
Following the general availability of the COSMAC 1801 microprocessor, two game specific projects were underway at RCA. One was to be an arcade cabinet (discussed in my series) and the other was a home specific TV oriented game system. The January 1975 memo showcases diagrams for the device as well as eight games already created. Jackpot, Space War, Match Game, Acey-Duecy, Spot Speedway, Deduce, Life, and the combined Bowl/Tag Game. The first program in this list was created by Weisbecker’s daughter Joyce, who was not an employee of RCA which made her the first independent game programmer (Hagley Museum and Library).
The failures of RCA to capitalize on this pre-built library were numerous, but the games themselves were constrained by the system as well. The drawing routines on the RCA devices were not able to product smooth motion without flickering due to an over-reliance on the CPU. They also failed to get the system out soon enough to beat their competitor’s on novelty (Hagley Museum and Library). The brilliance of the team in creating what was almost certainly the first fully-fledged video games for a microprocessor though are to be applauded, as was Weisbecker’s commitment to the project and history by maintaining these documents until they were preserved at the Hagley Museum in Delaware.
Quite close to Intel in Silicon Valley, it’s no wonder that Atari would be exploring the options of a microprocessor early. More specifically, their advanced R&D group Cyan Engineering were looking at the opportunity as far back as 1973. The most well-known of these experiments was their Delta Queen pinball prototype. In mid 1974 they were actively testing a pinball machine with a microprocessor in the field, but had too many service issues and therefore didn’t continue with the project. What isn’t well known is that this was already the third step in their process of exploring the technology, all right in the midst of creating Atari’s earliest successful products.
The first evidence of Cyan’s experimentation comes from a publication Systems Engineering Today in February 1974. An Intel flyer showcases the uses for the current Intel “microcomputer” OEM parts, with the tagline “From electronic games to blood analyzers”. It features a picture of the rare Color Gotcha machine with a small blurb talking about how the installed microprocessor allows the player to win every so often. Earlier in the ad it mentions pinball machines as a possible function of the device as well.
Gotcha does not have a microprocessor, nor does it seem to have ever been intended to be the first game featuring a computer opponent. Cyan Engineering did the major engineering on the game but by February of 1974 they had only just received their first Intel development kit. Steve Mayer recalls Intel being very aggressive in sending reps their way (Mayer 2018), though it appears his interest was pre-emptive, as he sought out information in November of 1973 and attended and Intel seminar the following month (Thomas, pg 28-31). This blurb may have been a marketing angle based on an idea that perhaps Nolan Bushnell had rather than a reality. It would have been quite a leap to create even a primitive AI for a fast game like Gotcha, and certainly far too expensive on an already not hugely popular game.
However Mayer and Larry Emmons were interested in using the microprocessor as a an interface for game controllers, so they wanted to create some sort of hardware system to test the component’s potential. The first pen to paper experiment was a pinball machine, but it wasn’t Delta Queen. Instead they rigged an MCS-4 series Intellec 4 development kit up to Bally El Toro model. An engineer named Gregory Cox was tasked with creating the software for this system as his first microprocessor project (Cox 1981/09/11 pg 45-46). It was mainly a test bed since the development system was rigged right into the pinball machine, but it was probably the first of it’s kind (Cox 1981/09/18 pg 95-96). Mainly the problem was to do with electrical noise that they couldn’t properly isolate, something that all the early microprocessor pinball games would struggle with (Cox 1981/10/12 pg 10-14).
After El Toro’s completion in about May of 1974, Cox also started experimenting with creating a box with an LED display that could play small games like blackjack and craps. In a sense this was the first handheld game display, but the four games that he completed were – just like the El Toro electronic pinball – never seen outside the Cyan Engineering lab (Cox 1981/09/18 pg 94-96).
Right about this time there were two video game projects that Cyan began experimenting with. Cox created a first person flying game (possibly called “1st Person”) that used a joystick though it was never finished before he left due to Atari and Cyan’s financial difficulties (Cox 1981/09/18 pg 107-108). The other was a project called “Frenzy” which was specifically being worked on for Kee Games. Steve Bristow described this game as “the thing with the pounding buttons” (Bristow 1981-08-05 pg 58) and also that it may have been called “Reaction tester” which implies a sort of pressure sensitive control testing. Both titles are listed in a project memo dated April 11, 1974 by Steve Mayer (Deposition Notice pg6). Also listed on the memo is Touch Me, which according to Steve Bristow was also microprocessor controlled, probably just to register the inputs (Bristow 1981/08/06 pg 32).
Cyan felt they could do better than El Toro, so right after the pinball was completed they set to work on acquiring the Bally Delta Queen around June of 1974. In a quick turn around they repurposed their initial work into something they could show at the MOA that November (Bristow 1981-08-05 pg 51-52). Bristow recalls they had to take apart the El Toro machine to get the Delta Queen to work, which had been relegated to the sunny porch of the Grass Valley facility (Bristow 1981-08-05 pg 15, 75). Five units were created and shown privately at the exhibition to gauge potential interest (Reissue Examination pg 6-7).
It does not appear that the Delta Queen got a lot of traction in sales so they decided to test the play value at a local hangout called Frank’s Pizza Parlor at year’s end. Their fears with the El Toro machine resurfaced as the machine was constantly inoperable due to electrostatic noise rendering the whole thing completely inert. One of the main issues was that the power to the reset circuit inside the machine caused outages which completely messed with the internal logic (Schoeffler pg 52). Mayer and Emmons had to go back to the drawing board to create some sort of protection for the device, elongating production time (Trial Brief pg 157-158). They showed the game once again in April 1975 at a DNA conference, of all things, called Asilomar (Reissue Examination pg 7).
Problems persisted though and Cyan would have one last go at making a solid-state pinball game, though it appears to have initially been an afterthought. Mayer procured a William Super Flite pinball game and wanted to modify it to work for four players (though interestingly it already had a four-player model). Into this he plugged their aged 4004 system and around the end of 1975 sent it down to Atari production engineers to have a look. It was still impossible to get working and incredibly finnicky. This limited design was abandoned once and for all, leaving way for The Atarians pinball machine to be produced in 1977 (Trial Brief pg 158; Schoeffler pg 54-56).
All this experimentation with the microprocessor had made Atari realize that they needed to start finding people who could work in software when they eventually made the transition. Their first was a fellow named Tom Hogg and the second was David Shepperd (Smith). Shepperd had actually been intrigued by video games for a long time, and as a techy in Silicon Valley he was keenly aware of the new technologies coming out. He was a very early adopter of the Altair 8080 computer which started shipping in early 1975 and took to the idea of modifying his expensive toy into a video game (Shepperd).
The two player game he created was vaguely Pong-like, but was not a Pong clone. He couldn’t clearly remember what it’s function was. In truth it wasn’t even really a microcomputer-driven game either. The major video processing was handled by a series of about a dozen TTL boards and the Altair simply set some registers at game start-up, but either way it was a game that did use a microprocessor. When he interviewed at Atari in late 1975, he was told to not speak about this project for legal considerations, but he got the job probably on the strength of being one of the few people in the world to have experience creating a microprocessor software game. Shepperd continued to hold this hardware up until recently when he donated it to the Strong Museum of Play in Rochester, New York (Shepperd).
Atari’s early but not quite speedy enough introduction of microprocessor technology into their business had as much to do with wanting to do things right as it was simple fumbling. They had definitely the most sophisticated development system for their early programmers compared to all the other arcade companies, having an actual mainframe to work with. Their early failures had provided their mechanical engineers with a test bed and their programmers with a scope of what was possible. Of course, right at this same time they were also making their steps into the home market with the 6502, segueing nicely into the next candidate.
The Hialeah, Florida based Allied Leisure had always been ambitious on exploring new frontiers, but they lacked production stability to really make their products stand out. Even before the microprocessor they were putting together a solid-state pinball machine which would eventually be known as Dyn O’ Mite, named after the memorable catchphrase from Good Times (Smith). It was a startling leap forwards that caused a lot of people to take an interest, but they did actually have a microprocessor pinball going at about the same time (Play Meter 1975/06).
It was none other than Chuck Peddle, mastermind behind the MOS 6502, who was building this machine for them. Peddle had come to Allied Leisure to try and sell them on the 6502 and offered to build a pinball machine for them to prove the component’s usefulness over their TTL design. Peddle’s main intention behind this demonstration was to show off the 6502’s capabilities, especially the I/O ports, but it’s likely also that Allied were interested to see which of the two solid-state projects could be the most viable. The 6502 may have been cheap compared to other microprocessors on the market, but in 1975 it’s worth had not yet been proven. Every other company had come by to offer them a chip, but no one offered them any assistance until Peddle (Bagnall pg 40-41).
Peddle spent about a month writing the program for the pinball machine in Florida, assembling it with the help of the company’s engineers. Bagnall reports he showed it at the National Computer Conference, presumably the one in May 1975, just two months after Dyn O’ Mite had been shown to coin-op distributors (Bagnall pg 45). It’s not clear why Allied didn’t actually go forwards with this design. Peddle doesn’t speak to any of the difficulties that many of the other early builders encountered but he is also not very forthcoming on many of his inventions’ failings in other sources. It’s likely that because the 6502 was a better CPU than the 4004 that it didn’t have as many difficulties in operation, but I find it doubtful that he made anything resembling a production-ready version.
Ramtek was a likely source for early use of microprocessors. They were in the monitor field which was one of the earliest market segments to take the circuit as a full-on support and their base was only a mile or so from the main Intel building (McEwan pg11). Charles McEwan made the decision that Ramtek was going to enter the pinball business by the end of 1973 and so hired an experienced pinball designer out of Williams Electronics, Bob Jonesi (McEwan pg 12-13). A team consisting of Howell Ivy, Tom Adams, Roy Nesi, Ken Acgard, and a consultant named Ray Holt (who claims to have invented the microprocessor prior to Intel) were set to create a 4004-based pinball machine from scratch that they deemed Lucky Dice in September of 1974 (McEwan pg19; Schoeffler pg 64-65).
Unlike many of the other experiments going on at the same time, they were able to make Lucky Dice self-sustained after a short while. They disconnected it from the main development computer and got it running on it’s own, reducing the need for so many development kits (McEwan pg18). They built three of these devices, all in different colors, and tested it internally at Ramtek around May of 1975 (McEwan pg20-21).
However, Lucky Dice was not a very lucky machine. According to all who worked on it, it functioned only inasmuch to show there was electricity running through it. The scores were wrong, the lights went off with reckless abandon, and it wouldn’t register the tilt. Despite this they still attempted to showcase it at the October 1975 MOA show in a private hotel room. It didn’t get enough of a positive reception and Ramtek wasn’t making the money it needed at the time to fund a new full scale manufacturing operation, so they left it behind and would spend the next year or so working on a video game microprocessor hardware system, possibly the first of it’s kind (McEwan pg 25, 29-30, 48-51; Kayton 57-58; Electronic Engineering Times pg 153-154).
The mystery of the San Mateo-based company Major Manufacturing is pretty much all but unknowable at this point. What is clear is that they were founded around 1974 and produced two cocktail tables, a pong clone called Fascination and Lunar Module. However, buried in obscure mentions in the coin-op trade publications, at the MOA show in 1975 they were mentioned as having for sale three arcade games which used a microprocessor (Vending Times 1975/10 pg 80).
One of the named products, though not expressly named as microprocessor-based at the time, was Demolition Derby (this is especially confusing because there was another Demolition Derby available at this same show from Chicago Coin). This product was the providence of a Jerry Lawson, engineer at Fairchild. A long-time player of games and computer programmer, Lawson used a Fairchild F-8 to create a game in his spare time and behind the back of management (Smith).
Lawson never described how this game played, but a description of the game when speaking about Major’s upcoming products describes it as “[D]emotion derbies, wherein fenders ‘fall’ from electronic vehicles” (Vending Times 1975/10 pg 95). He did note that the cabinet he built had a feature to defeat people attempting to trip the coin switch on the machine. He sold this game to Major Manufacturers who tested one unit in a pizza parlor, but according to Lawson never put it out. He wanted to sell the game to another coin-op company, but by this time the people at Fairchild knew about his experiment and were roping him into the project which would become the Fairchild Channel F (Smith).
But, it does appear that Major Manufacturers at least attempted to interest distributors in the product. According to president Bill Kinsel, the other games were sourced from an internal development staff, but also said that 25% came from outside submissions. It’s unclear how they solicited these games, but at least one of the other microprocessor games appears to have been a Football game, though whether this was a Pong variant or not is unclear (Vending Times 1975/10 pg 95). Chuck Arnold, who had picked up the sales arm for Major Manufacturers about a month earlier than the MOA show, remembers precisely nothing about the products that they were offering. The experience at Major Manufacturers for him was not a pleasant one and he only recalls ever selling the Fascination cocktail table (Arnold).
Innovative Coin Corporation
A former Atari engineer named Doug Hughes left the company around the beginning of 1975 to create an independent arcade game manufacturer called Wizard Kings in California with partners Don Cera, Jerry Donaldson, and Mike Glass. Hughes had contact with Fairchild at the time and managed to procure an F-8 chipset from them, presumably a pre-production version as he recalls finding bugs in the microcode which they helped fix. As the team started working on a plane game, which would be called Spitfire, Hughes found a local partnership called Innovative Coin Corporation that would produce the game when it was finished (Hughes).
Hughes’ game on a gameplay level wasn’t too dissimilar from Atari’s Jet Fighter, but the under the hood technology was very different. Hughes noted that Atari’s motion circuit patent used counters to determine horizontal and vertical positioning on the screen, but Spitfire used a different type of vectoring which didn’t violate the patent when Atari lawyers came sniffing around. One other difference to Jet Fighter was that Spitfire had a tower in the center of the screen, creating a more dynamic play area as in Tank. They built this game out of Hughes’ and Glass’ garages into a wire-wrap prototype (which Hughes still has to this day) and gave the product to Innovative Coin Corporation in late 1975 (Hughes).
Though Innovative Coin Corporation would release this game in January 1976, after the famous 1975 MOA show, it was a concurrent development with the other races to the starting line on this list. The unscrupulous company would go down in flames and attempt to drag Wizard Kings with it, but Hughes was saved by Hank Ross of Midway Manufacturing (Smith). Subsequently he went to join a development arm under Bally and work with the Dave Nutting Associates group for a while exploring the microprocessor, but that was well past the debut (Hughes). The game would be remade by Mike Glass for the Fairchild Channel F console in 1977 (thanks to Alex Smith for noticing this!).
The very first known amusement device to use a microprocessor was Bally Alley, a wall game released in April 1974 (Cash Box 1974/05/18). While in name it was modeled after an enclosed bowling game from the 1940s, in practice it was far less tactile. The game used bulbs to indicate the area and trajectory of the ball, lighting up which pins were knocked down. Up to four players could grab joysticks which they would hold and release a button of to compete for strike supremacy (Foster pg 111-114).
The invention was the work of an Australian contractor named Colin Foster, an experienced gambling game designer (Foster pg 6). He had formerly been employed by Bally before establishing his own firm called Emu Electronics (yes, really) in Illinois to serve as a contractor for the local game companies (Foster pg 23). Foster designed a very densely packed and complicated game running on an Intel 4004 that didn’t have so many issues with basic operation (Foster pg 49, 58, 74). According to Lane Hauck, designer of Gremlin Industry’s microprocessor game Fooswall which came out the following year, microprocessors were a huge boon for wall games because they bypassed a lot of the basic timing issues of circuitry-based games (Smith).
Bally also had several other potential contractors which were coming up with new technologies, and the most pivotal of those for them (and the entire industry) would be the contact they made with David Nutting. At this time David Nutting was at his second company Milwaukee Coin Industry (abbreviated MCI) and was growing increasingly unhappy with the direction of the business. They were moving towards arcade operation and away from self-sustained manufacturing, having brought in several consultants themselves to supply game concepts (Nutting 1984/01/06 pg 86-87).
One of these initial independents was a local engineer named Jeff Fredriksen. One of Fredriksen’s first tasks with MCI in 1973 was to judge the viability of using a Fairchild F8 microprocessor in a video game, which MCI had not created up to that point. He found it feeble and so they looked at other options like National Semiconductor who wanted to sell them a big self-sustained computer board. Finally they contacted Intel in December 1973 and evaluated that they were the proper chip to pursue (Fredriksen 1984/01/03 pg 130-131, 134-135).
Nutting’s mind went immediately to a pinball machine, but he was a little weary about if a single chip could drive a full game. Fredriksen built him a demonstration and they decided to work on a simpler and more familiar concept. Fredriksen would first modify an IQ Computer game, a game Dave had helped create back in 1966, and see if he could get the 4004 to control the switches and lamps inside. Nutting okay’ed the purchase of an Intellec 4 system for about $3,000 which would arrive in April of 1974 (Fredriksen 1984/01/03 pg 151-157; Nutting 1984/01/06 pg 105-110; Fredriksen 1984/03/19 pg 123).
By the time the unit had arrived, Nutting, Fredriksen, and fellow MCI technician Dwayne Knudston had moved out of the MCI facilities proper into a warehouse that Nutting owned in anticipation of splitting with the company. Nutting and Knudston wired in the external Intellec to the IQ Computer’s lights and switches while Fredriksen wrote the program to control it all. The product was called Super I.Q, and it worked, but it was not something they were going to test in the field or anything. They did the first step, now it was time to move on to something they knew they could sell (Fredriksen 1984/01/03 pg 148-157, 167-171, 178-179; Nutting 1984/01/06 pg 111-112).
Fredriksen had been tearing apart a Gottlieb Magic Carpet pinball which Nutting had at home to test it’s signals so that he could be prepared for the challenges ahead (Fredriksen 1984/01/03 pg 157-158; Fredriksen 1984/01/05 pg 135; Nutting 1984/01/06 pg 101-102, 116-117). Nutting arranged a meeting with Bally in June 1974 to see if they would finance this independent game development operation. Fredriksen and Nutting showed Bally the Super I.Q with the promise that they could make a pinball machine from the technology, which Bally accepted (Fredriksen 1984/01/03 pg 171-173; Nutting 1984/01/06 pg 115). With a budget accrued from Bally management, they officially left MCI (along with a few other team members) to form Dave Nutting Associates at the end of June (Fredriksen 1984/01/03 pg 180-181; Nutting 1984/01/06 pg 117, 119).
Nutting asked Empire Distributing head Joe Robbins to send him two of Bally’s latest pinball machines which arrived at the beginning of August. These were two of the not-yet generally available Flicker pinball machines which Bally was just starting to put into production (Nutting 1984/01/06 pg 122-123). Excitedly, the small team of worksman got to work converting one of the machines to use the microprocessor. Just as all the rest, they had to rig up their 4004 development system into the machine with an extended cable and try to translate every function into software (Fredriksen 1984/01/03 pg 193). Nutting had floated ideas for more interesting pinball designs (Fredriksen 1984/01/03 pg 182-183), but he specifically wanted to just make a working solid state prototype, not a showcase of new features so as to not overawe Bally (Nutting 1984/01/06 pg 129-130).
Fredriksen was the programmer, as the only guy with any computer experience, and he modified a teletype to create the code to be read into their Intel EPROMs (Fredriksen 1984/01/03 pg 151). Nutting rigged up a Fairchild digitial display to the cabinet and hooked up every relay to the appropriate function. Knudston and Dave Stewart did the layouts and Paul Smith did the physical wiring (Fredriksen 1984/01/03 pg 192; Fredriksen 1984/01/04 pg 110-111; Nutting 1984/01/06 pg 124-126). On their first test, a week or two in, they tried the Flicker game out for the first time (Nutting 1984/01/06 pg 128). They received no blowouts when they first turned it on, practically a miracle for any electronics engineer. As soon as they started playing it though, they found that the score counter was completely out of whack. It was inconsistent to, a veritable nightmare for troubleshooting (Tzannes pg 10; Smith).
The solution turned out to be pretty simple though. They didn’t have issues with electro-static noise because they had already created a solid-state arcade machine at MCI called the Safe (Fredriksen 1984/01/03 pg 127-128, 198-200; Nutting 1984/01/06 pg 102-103). The problem instead was that the microprocessor was too quick for it’s own good. It was triggering the functions of the pinball table too quickly for it to handle itself. Fredriksen had initially worried about not having enough speed with a microprocessor, but now it was too much! (Fredriksen 1984/01/03 pg 132-134) Initially Fredriksen wanted Nutting to create a switch to remedy this function, but probably out of not wanting to over-complicate things he just asked Fredriksen to input a system delay. That worked flawlessly and they soon had a totally functioning Flicker machine (Tzannes pg 10; Smith).
Through the next few weeks they would be able to remove the Intellec system from the interior and simply have the machine running on a single, tiny PCB that Nutting dubbed the “Bally Brain” about a week before the demonstration (Fredriksen 1984/01/03 pg 194-195, 198; Nutting 1984/01/06 pg 133-134). On September 26, 1974 four Bally executives were given a demonstration of this device and they were floored. Head of engineering John Britz was in such disbelief that he followed the power extension cable towards a closet looking for a hidden mainframe. Nutting and Fredriksen couldn’t help but feel inwardly gleeful about this as they showed Britz the circuit board inside. That’s all it was (Nutting 1984/01/06 pg 134-139).
Bally expressed immediate interest in the technology, but for some reason which was never outright stated, they did not want to go with the Dave Nutting Associates design despite Nutting’s insistence (Nutting 1984/01/09 pg 4-5; Fredriksen 1984/03/19 pg 114). Nutting presumes it was because they didn’t want to pay him a royalty (Smith), but one should be cautious at taking him at his word since his end with Bally wasn’t so amicable. Reading between the lines, it’s possible that they felt that going with the Nutting design would leave them unable to refine it in production. Given that they had managed to make Bally Alley, they didn’t want to be hamstrung by a potentially faulty design that they couldn’t make viable (Englehardt pg 189-192).
Instead one of the people who had attended that meeting, Frank Bracha, took away Nutting’s prepared materials and showed them to someone who had been working on a microprocessor slot machine project at Bally named William Englehardt (Englehardt pg 186-188). Englehardt was asked to head up the design of an electromechanical pinball machine using the Intel 4004 (Englehardt pg 188), and they decided to use the four player version of Flicker called Boomerang. Their model suffered through noise difficulties as well, but Englehardt had been a military engineer so he was able to solve those problems and actually managed to get some machines on test in May 1975 (Englehardt pg 198-201). Then they decided to do a larger run of units with a machine called Bow and Arrow, building about 20 and putting them side by side with electro-mechanical versions to see the coin take (Englehardt pg 194-196).
It would still be another year until Bally officially launched their own solid-state pinball games, but they had Englehardt and Bracha’s crucial patent and would eventually acquire the Dave Nutting Associates patent (Fredriksen 1984/01/04 pg80; Nutting 1984/01/06 pg 150). Nutting was encouraged to try and sell the technology to other pinball manufacturers. He tried Gottlieb first, but they were uninterested. He was then put to a company down in Arizona (discussed below) (Nutting 1984/01/06 pg 142-144). In the interim, they purchased several new microprocessor development systems and did some work with the Intel 4040 chip (Fredriksen 1984/03/16 pg 133-134). Fredriksen created a gun game with this system called “Weird Animal Kingdom” but it’s uncertain how far this got in design (Fredriksen 1984/03/19 pg 32, 40).
It was clear enough though that a 4-bit microprocessor simply could not drive a display on it’s own. So they bought an Intellec 8 development system for the 8080 chip and worked on a new proof of concept to show Bally shortly after the pinball proposal had fallen apart. Fredriksen, not having the best equipment or being the best programmer, was able to get some sort of noise on the screen driving the video display. It wasn’t much, but when Nutting showed it to Bally, they would greenlight the development of a video system which would of course become Gun Fight (stay tuned for that story in the series) (Fredriksen 2011; Nutting 2017).
Bally’s proactiveness in approaching the microprocessor yet reluctance to embrace it was reflected in all the other major pinball manufacturers. Gottlieb had been investigating solid-state design as far back as 1972 yet they were infamously the last to actually get on board. Williams went through several rounds of exploring the microprocessor market before they finally got on board, reaping the rewards of it. These tentative steps not only left a fascinating paper trail which forms the majority of this article, but an insight into the minds of the conservatism of the industry and the new blood that would bring it into the boom era of the 1980s.
Based in Phoenix, Arizona, Mirco was the perfect company to be getting involved with microprocessor games because they were already creating test equipment for mainframes and single components. They had entered the Pong boom in 1973 but their main game business was really reliant on foosball rather than video games. A new opportunity opened up through when David Nutting came asking if they would build this pinball table he had created (Smith).
By this time, Nutting and Fredriksen were already attempting to transfer their logic to other microprocessors. They started with the Intel 4040, the next step up from their Flicker machine, and used that in a Gottlieb Flying Carpet machine for demonstration to Mirco. They were shipped a machine in August of 1975 with non-complete software, but Mirco felt that they could make something internally because they had in-house engineers who had come from Motorola (Nutting 1984/01/09 pg 10-12; Smith). The agreement with Nutting was maintained, but they didn’t use the design that had been provided to them exactly. Nutting and crew had been proposed by Bally to make a home pinball table, first off of the famous Wizard table, then off of the game Hocus Pocus (this project would eventually become the Fireball home table). They offered this F-8 powered game to Mirco, but the Arizona company was already full speed ahead on it’s new design (Fredriksen 1984/01/04 pg 54-56, 74-76; Nutting 1984/01/06 pg 144-147).
The Motorola 6800-based design was led by engineer Cash Olsen, one of those former Motorola engineers. They ripped off the playfield design of the Gottlieb Magic Carpet game they had been sent and created a few basic prototypes. Dubbed Spirit of ’76, echoing the coming United States bicentennial, it was the very first microprocessor-based pinball to become available for sale at the 1975 MOA show. For about a week. The unit shown at the MOA had been just a prototype, but the demand had been so strong that Mirco demanded it’s engineers start producing it immediately. The technical issues caused the units to just outright die and the fact that Mirco had never made a pinball machine before completely screwed over supply lines. They shipped some units in November 1975, then halted until March of the following year, but in all only sold 500 units, possibly all at a loss (Smith).
However, Mirco also had another game they were showing at the MOA show competing with Gun Fight: PT-109. The same engineer as the pinball project, Cash Olsen, had been working on this project before the pinball game. He knew his way around microprocessors from his prior experience and he evaluated the potential suppliers. They decided to go with the Fairchild F-8 because Fairchild gave far more support to potential developers, at least in their area of the world. This would be the start of a relationship with Fairchild which ultimately ended in disaster, but for now was quite useful in the rushed job that Olsen had to throw this project together at the same time as the Spirit of ’76 (Smith).
PT-109 was a submarine game that ultimately consisted of 300 total chips inside some packed logic boards. He developed the game using a timesharing GE mainframe with an incredibly slow turn around time, leaving little time to iron out particular bugs or make interesting gameplay changes. It was not a particularly exciting game in the grand scheme of things and it barely sold despite garnering positive reception at the MOA. It’s very likely that the Spirit of ’76 debacle hurt it’s potential reach and it’s format, only available in a cocktail table, was not appealing to arcade owners by the end of 1975 (Smith).
The takeaway from all this is not so much a definitive outline of the beginnings of microprocessor development or an attempt to determine a ‘first’, but rather the reactions and examinations of innovators competing with each other. The expertise of the high-end mainframe programmers trickling down into the amusement industry, soon encompassing it and providing it with a new direction. It was one heck of a moment in time and provides one of the most interesting looks into a development period for games that will ever be available to us.
Arnold, Chuck. Personal interview, 2017.
Baer, Ralph. Videogames: In the Beginning, published 2005 by Rolenta Press.
Bagnall, Brian. Commodore on the Edge, 2012.
Deposition Notice. Dated October 5, 1981.
Electronic Engineering Times. Gametronics Proceedings, Janaury 1977.
Englehardt, William. Trial testimony taken March 19, 1984.
Foster, Colin. Deposition taken October 5, 1979.
Fredriksen, Jeff. First trial testimony taken January 3, 1984. Second trial testimony taken January 4, 1984. Third trial testimony taken January 5, 1984. Fourth trial testimony taken January 6, 1984. Fifth trial testimony taken March 16, 1984. Sixth trial testimony taken March 9, 1984. Interview conducted in 2011 from Balleyalley.com.
Hagley Museum Collection.
Hughes, Doug. Personal Interview, September 14, 2017.
Kayton, Irving. Trial transcript dated March 28, 1984.
Mayer, Steve. Interview by Alex Smith on July 7, 2018.
McEwan, Charles. Deposition taken January 31, 1980.
Reissue Examination. Dated November 6, 1982.
Schoeffler, James. Trial transcript dated January 26, 1984.
Shepperd, David. Personal interview, 2017.
Smith, Keith. All in Color for a Quater, 2016. Unpublished.
Thomas, Ken. Deposition taken August 19, 1981.
Trial Brief. Dated Janaury 4, 1984.
Tzannes, Alexis. Gameroom Magazine “Found! The First Microprocessor Pinball Machine“, June 1999.
Weisbecker, Joseph. FRED System manual dated July 1972.