The difference in speed between the Space Shuttle and a snail is the same as the difference between ROM silicon chips and a CD-ROM, a difference of about 2 million times. […] The next time someone tells you the CD-ROM is the wave of the future, tell them that the future doesn’t belong to snails. (Nintendo Power #59, April 1994, 108)
On July 3, 2015, the Internet awoke to a weird story. Someone had recovered a long-lost prototype of the “Super NES CD-ROM.” Pictures were posted, news sites reported all kinds of things, and authenticity was discussed, questioned, proved, and disproved, all at the same time. The Internet was ablaze with a fire so strong it could have roasted the top half of the exhibited machine—only because the bottom half of the machine was yellowed out, a normal consequence of the flame-retardant additive incorporated in its plastic shell, just as with the SNES. But the top half was pristine gray presumably because the flame-retardant chemical had not been incorporated in it. The Internet being what it is, message boards and comment spaces disregarded the facts and soon filled with “Obvious Photoshop is obvious,” “Disgusting smokers, why didn’t they put that Holy Grail in a safe,” and so on.
Given that I was finishing the first draft of this book, and specifically this chapter, which focuses on the SNES-CD, I was surprised, puzzled, and worried that some heretofore unknown fact would turn up and completely invalidate everything I had written (these are the risks of working on the history of secretive corporations, I suppose). But what soon struck me the most was how the dead peripheral—a stillborn, no less—elicited a capital of sympathy among gamers. Soon, alternate histories turned up, various writers speculated on what would, might, or could have happened, how the (nonexistent) games would have stood out from their peers, and so on. The spoony bards hath awakened, and verily, their lyres pluckethed! The amount of interest the prototype generated is commensurate with the impact of its failure on Nintendo and on video game history in general. In a sense, this story shows that the SNES CD-ROM debacle encapsulates everything about the transformations of video game technology, culture, and marketing that would put an end to Nintendo’s Silver Age.
The Future Was Multimedia
In the late 1980s, when NEC announced a CD-ROM player for its upcoming PC-Engine system, Nintendo got scared into developing a follow-up to the Famicom. This attests to the perceived importance of CD-ROM technology at the time. In the early 1990s, CDs were seen as a vital component of the “Future of Computers,” a technological trajectory at the end of which lay vast fields ripe with possibilities.
Giovanni Dosi (1982) defined technological trajectories as trails of solutions that try to answer some engineering problem. That problem, in turn, is framed as part of a certain technical paradigm. In this framework, the home video game industry was operating at the time according to the technical paradigm of storage space maximization: Games were produced to fit on cartridges and thus had to maximize their usage of the limited storage space afforded by the cartridges’ ROM. Game developers worked in that technical paradigm and sought solutions. One such solution was found, for example, in the technique of bank switching for Atari 2600 game development (Bogost and Montfort 2009b). Altice (2015) explains in great technical depth how Nintendo managed to cram so much data in its game cartridges to have its seminal Super Mario Bros. fit in only 40 kilobytes.
“Thinking outside the box,” as the saying goes, eventually changed the terms of the problem by changing storage media altogether in favor of the CD-ROM. This completely reversed the technical paradigm. Game developers were not operating from the paradigm of storage space maximization but rather from the paradigm of storage space exploitation; developers moved from being constrained by cartridges that packed up to a whopping 32 megabits of space (4 megabytes), a technologically impressive feat for 16-bit standards, to CD-ROMs that offered at least 650 megabytes of memory. The new problem was that all this additional space had to be used somehow to demonstrate a game (or a platform)’s superior technology in the face of competitors. If Super Mario World could have 72 levels in 4 megabytes of storage, then having a Super Mario World CD with 11,638 levels just wouldn’t cut it. This set in motion (quite literally with Full-Motion Video games) a trajectory of innovation that had been lying around dormant for years but could be made flesh thanks to the CD-ROM, a path that Nintendo had treaded in the past: the game console as multimedia entertainment device.
The idea that game technology would do something else, something more than only playing games, was not a new one when it started gaining traction in the early to mid-1990s. In fact, Nintendo had been actively pursuing it throughout the Famicom’s life in Japan and attempted to do so as well for the NES in the United States. The Famicom’s designed expandability could accommodate auxiliary services and complementary goods and functions. Japan was already seeing accessories meant to expand the system beyond playing games. The Family BASIC package comprised a cartridge that would allow users to program in a BASIC environment using the supplied Family Basic Keyboard and to store these user programs on the Famicom Data Recorder (a standard cassette tape recorder drive). Another high-profile peripheral was the 1988 Famicom Modem, which adults could use to trade stocks, bet on horse races, and access weather forecasts. Although these systems were not a huge success, they attest to the pursuit of such audience expansion goals by Nintendo, which attempted to reach out to family members other than the children and for activities other than playing games.
The World of Nintendo, as Provenzo noted in 1991, was bound to reach out into all kinds of other activities, which in hindsight works as a ghostly comeback of Nintendo’s forays into other sectors in the 1963–1968 period, when the firm tried branching out into instant rice, taxis, a chain of love hotels, and so on. The idea of game consoles becoming multipurpose machines was seductive in the late 1980s, in part, because the market for home computers was still bustling—in Europe, personal computers such as the Commodore 64 and Sinclair ZX Spectrum were all the rage, just as the MSX computers in Japan had been a few years earlier. A trajectory of game consoles expanding to become home computers was in place, as Newsweek reported: “The biggest philosophical question among manufacturers at this year's CES: is the video game a mere toy, or a new communications medium? Nintendo clearly plots a course beyond entertainment for its machines” (Newsweek, June 17, 1990). The Economist reported on Nintendo’s grand plans for the Brave New World of Nintendo on August 18, 1990, with an article titled “Wham! Zap! You just made a million”:
This autumn Nintendo—the world’s largest supplier of video games, with around 80% of the market—is set to launch a professional version of its best-selling “Famicon” (Japlish for “family computer”). Its Super-Famicon will be nothing less than a powerful business computer masquerading as a game-player.
The first of the Kyoto-based company’s applications for adults will be a database package for its own distributors, who will be able to use their game-machines to dial in and find out what software products are available for Nintendo machines. This is only the beginning. Nintendo has teamed up with Fidelity Investment, an American fund manager, to develop software that will allow users to play the stockmarket or to manage property. The Japanese company has also joined forces with AT&T to provide a communications network so that American households with a Nintendo in the living room can send video messages to one another. (The Economist, August 18, 1990, 60)
This trajectory, however, soon came under criticism. Couldn’t the “Future of Games” just be full of games? The Economist’s report from January 19, 1991, tellingly titled “Back to Earth,” nicely illustrates this quick turnaround. After pointing out that Nintendo’s share price has dropped by half in 6 months and that the Super Famicom is a risky gamble for Nintendo, it weighs in against the multimedia strategy:
Perhaps the company’s best hope is that Europe, where it only recently began selling its older model, will prove to be its next big market. This seems more promising than its other strategy of adapting machines for home banking or education. Video games are about addiction and fun, not learning.
Home computers constitute the first technical paradigm in the technological trajectory of video game hardware serving for multiple different purposes1, and the modern networked video game consoles with social media integration, video playback, and media server capacity are the latest (Sony’s PlayStation line benefited greatly from this trajectory, with its consoles doubling as music CD, DVD, and Blu-Ray players). In between the early home computers and the later integrated games-and-media-and-network machines, however, was a substantial technical paradigm: the CD-ROM-based gaming console.
From 1991 to 1993, we find many game machines that were designed and sold as general multimedia devices: the Philips CD-i (1991), Commodore CDTV (1991), Tandy Visual Information System (1992), 3DO Interactive Multiplayer (1993), and Pioneer LaserActive (1993). They were all able to play games but also to host edutainment software for the children, museum tours, encyclopedia, art galleries, and so on—incidentally, all kinds of things that would soon be available to everyone for free thanks to the Internet. There are probably many reasons that those platforms did not meet with success, but the rising availability of the Internet in the mid- to late 1990s (especially in the kind of technophilic market these machines were aiming at) is surely one of them. Incidentally, I contend that these machines, taken together, should constitute the “generation 4.5” of video game historiography that I alluded to in the book’s introduction. Their technological make-up, marketing initiatives, and cultural positioning exhibit distinct and common features that make them almost irrelevant as part of the “fourth generation of video game consoles.”
Nintendo would be given the chance to partake in the multimedia future but would ultimately reject the SNES CD-ROM. The Super NES would concentrate on its status as a game-playing machine, in the classic sense of what video games had been up until now, upholding tried-and-true lessons. Still, the firm tip-toed in diversification efforts for the console through a range of accessories. The vast majority of them were intended for traditional gameplay purposes: specialized gamepads with auto-fire, wireless communications, slow-motion mode, programmable buttons, flight stick shape, and so on. A single-handed controller was released in Japan, specifically for playing RPGs or strategy games. Multitap adapters allowed up to five players to play together. Light guns included Nintendo’s own Super Scope bazooka or Konami’s Justifier. However, one of them pushed in the direction of multimedia: the Super NES Mouse, included with the important Mario Paint. The mouse was the quintessential controller device for PCs that allowed one to browse documents, navigate complex graphical user interfaces, and have fine-tuned manual control over whatever it was they were doing. Mario Paint was a playful art studio where users could color, draw, make icons and sprites, animate pictures, and compose music. It was a suite of basic software like those found on a personal computer. It stands out from the SNES library and continues to enjoy success, two decades into the 21st century, thanks to musicians creating and uploading music made with the Mario Paint composer module.
Unique as it may stand, however, Mario Paint couldn’t substitute for the wide range of software that CD-ROMs brought to the computers and multimedia devices of the 1990s. Unfortunately for Nintendo, rejecting the CD-ROM would not only close off the trajectory of multimedia computing but would also bring dire consequences to the traditional form of video games.
From Smokescreen to Vaporware: The SNES CD-ROM
When Nintendo formally announced its plans for a follow-up console to the Famicom in 1988, it was mainly a smokescreen in an attempt to combat the hype around NEC’s PC-Engine. Ken Kutaragi, an engineer at Sony, initiated contact with Nintendo to share a sound chip he had been working on, which would be a dramatic improvement on the Famicom’s. Asakura shows, through Kutaragi’s notes from 1989 (Asakura 2012, 46–48), that he was pushing for a partnership with Nintendo as part of a long-term plan to get Sony in the games industry to concretize his vision of a “playstation,” a mirror to the computer “workstations.” Because NEC had announced (and eventually released in December 1988) a CD-ROM player add-on, the PC-Engine CD, Nintendo was afraid to lose its edge. Discussions with Sony turned to the development of a CD-ROM player for their next-in-kin Super Famicom, a logical solution given that the CD-ROM standard had been codeveloped by Sony and Philips.
Nintendo would get from Sony a Super Disc add-on for its Super Famicom, a heir to the Famicom’s Disk System. The possibilities of CD-ROM storage would both expand the interactive possibilities of games and increase the technological appeal and life cycle of the console to the public. In return, Sony would get the possibility to develop and market its own Play Station machine (mind the space), a multimedia device that could read music, entertainment, educative, and multimedia CD-ROMs, as well as featuring a cartridge and controller port for playing Super Famicom games from Nintendo.
True to its business model of controlling the production chain, Nintendo asked Sony to develop a proprietary format for their software instead of using generic CD-ROMs. Sony complied, putting together the Super Disc format, a hybrid CD-ROM-in-a-plastic-disk-casing format, on the condition that it would control manufacturing and licensing for it. Surprisingly, Nintendo agreed, relinquishing an uncharacteristically high amount of control for its habits. At Sony, the terms of the agreement with Nintendo were largely seen as advantageous, leading to widespread disbelief (Akagawa 2013, 35–37). Why had Nintendo agreed to such conditions?
Most video game histories chronicle the events as follows: Late into development, when the CD-ROM add-on was ready to be revealed to the press, Hiroshi Yamauchi, for some reason, suddenly realized that the terms of the contract were too advantageous to Sony and declared this deal to have gone sour. Nintendo then secretly maneuvers to develop a partnership with Philips instead. Sony presents the SNES-CD and Play Station in a press announcement at the summer Consumer Electronics Show of 1991 on June 1. On June 2, Nintendo announces that it is rather pursuing a CD-ROM project with Philips. Scandal! Treason! Sony is humiliated. Later, at Sony in Japan, everyone has suffered from the “Philips Shock”; Ken Kutaragi pleads for Sony to pursue the game console project, meeting dissent from the conservative Sony board. Then he says, “Surely you won’t let Nintendo get away with what they did to us?”, and president Ohga slams his fist on the table, full of sound and fury, exclaiming “DO IT!”—that is, go on and do the PlayStation (minus the space, hence a wholly different product), and make Nintendo pay.
A number of crucial pieces are missing from this puzzle, however.2
Between a Rock and a Hard Place: The Sony-Philips Deals
The reason that Nintendo had agreed to the Sony deal was that it wanted to integrate a lockout chip in the Super Discs to continue with the noble tradition of its CIC and 10NES. Because Nintendo would own the chip and code that would serve as key for the software to run, it would effectively control licensing and production. Sony may have controlled the disks’ hardware, but Nintendo would control the hardware-and-software lock needed to run the software on the disks. Sony, however, wanted such a lockout chip to be placed in the game console hardware rather than in the software disks and planned to protect the games and other software with data encryption instead (Dikmen, Rhizlane, and Le Roy 2011, 17). This solution was sound, and it was letting Sony control the flow of software—a big no-no in the Nintendo Economic System.
Sony may have had trouble understanding Nintendo’s worries. At the time, Sony was not interested in producing video games (after all, Kutaragi had to work against the majority of the Sony board to develop the audio chip technology for Nintendo) but rather was pushing its multimedia and entertainment technologies and assets from its Sony Music and Sony Pictures subsidiaries. As Akagawa (2013, 34–35) relates, Sony’s Play Station was envisioned as a way to get into the highly lucrative domestic karaoke machine market, a goal that makes perfect sense for the CD, music, and film powerhouse that the firm has always been. Moreover, as a manufacturer of televisions and VHS players (and co-developer of the CD-ROM format), Sony could stake an increased claim over the domestic, general entertainment electronics market and push videos on disc (especially music videos) to consumers. In this sense, the Super Famicom compatibility might have been little more than a Trojan Horse for Sony, a way to seduce consumers that their children’s video game machine might be replaced with a general device that would still allow them to play their games, as well as allowing adults to play or otherwise enjoy some entertainment.
If the future was to be all multimedia players, then Sony could reasonably end up being the best positioned firm in the world to profit from it. Since starting the collaboration with Nintendo in 1988, it had started moving from a technology company to being a technology and content company, acquiring CBS Records in 1987 and Columbia Pictures in 1989. Harris (2014, 136) gives this as the reason Nintendo backpedaled from the deal with Sony: It was afraid Sony would start developing content for the system, encroaching on its software-sided business model. It didn’t help that Sony opened a software publishing branch in 1989, Sony Imagesoft, which would market games for Nintendo’s platforms. That can’t have been good news for Nintendo, which, as a self-party firm, treats third-party publishers as competitors more than allies, as we’ve seen in chapter 2.
These combined factors led Nintendo to seek additional leverage against Sony. Well, that and the legal bills that awaited Nintendo from Philips. One of the most interesting analyses of Nintendo’s CD-ROM antics can be found in the book La Bible Super Nintendo (Audureau et al. 2013, 38–43), which is worth summarizing. Why would Nintendo reject the Sony project so late in the process and so unceremoniously? Incurring Sony’s wrath could have led them to stop supplying the sound chip for the SNES, a key component to Nintendo’s system. “In announcing its turnaround in June 1991, Nintendo not only publicly shames Sony, but also gambles on a considerable industrial risk” (Audureau et al. 2013, 40). It turns out that there’s a longer history linking Nintendo with Philips, one that begins with an important event: Philips acquiring Magnavox in 1974, gaining them the all-important founding patent for video games that Ralph Baer had filed (U.S. Patent No. 3,659,284, known as “Television Gaming Apparatus”). Philips asserted the patent’s reach by suing every major video game developer to obtain royalties on every video game system that was connected to a television.
When Nintendo of America entered the home video game market, Philips’ lawyers proceeded to move against them for patent infringement. Nintendo of America, however, had just won their major case against the movie giant Universal Studios, which had unsuccessfully attempted to sue Nintendo for plagiarizing King Kong with its Donkey Kong game. Nintendo chose to fight Philips and the unbreakable Magnavox founding patent that had won against Atari, Sega, Bally, and others before. Nintendo strategized and attempted to sue Philips for inequitable conduct, demonstrating that cases of prior art existed for computer games before the filing of Baer’s patent, and that they should have been disclosed when filing for the patent.3 It didn’t fly because Baer’s patent was granted on the basis of making interactive applications on raster scan displays; TV display was the central point, and there was no prior art of games using TV displays. Checkmate.
After Nintendo’s claim was rejected, Philips proceeded to sue them. The case was set to court in 1990 against impossible odds; it was just a matter of time before Nintendo lost, like every other console and arcade manufacturer before them. Nintendo wanted to settle and so had to offer something. It had just signed, on January 1, 1990, the contract for the Super Famicom CD add-on with Sony (Asakura 2012, 48) and was negotiating with the co-inventor of the CD-ROM. What had looked like a checkmate might turn around and offer a way out. Ralph Baer reveals that Nintendo paid $10 million to North American Philips in 1991 to cover all past patent infringements, a sum that he qualifies as “cheap” to resolve the issue. In all likelihood, that symbolic amount was part of a bigger deal between the Japanese and European firms which called for Nintendo to abandon its partnership with Sony and let Philips develop the Super NES CD-ROM add-on instead, as well as granting Philips the rights to publish video games starring two of Nintendo’s most valuable intellectual properties, Mario and Zelda (which resulted in the infamous CD-i games Hotel Mario, Link: The Faces of Evil, Zelda: The Wand of Gamelon, and Zelda’s Adventure). This episode led Audureau et al. to conclude: “More than a temporary nuisance, Philips is in truth a powerhouse in the industry, from a technological and legal standpoint” (Audureau et al. 2013, 41). The SNES-CD story certainly seems to support it.
After Nintendo’s backstabbing episode, Sony nevertheless continued working with them. Dikmen, Rhizlane, and Le Roy claim these second-round negotiations happened because Sony threatened to sue Nintendo for breach of contract (2011, 18), whereas other accounts speak of Nintendo suing Sony instead. According to Akagawa, it would later be proved that the whole Philips deal was only intended by Nintendo to stall development on Sony’s PlayStation (Akagawa 2013, 39). In any case, development work on the SNES-CD advanced substantially. Electronic Gaming Monthly mentioned definite plans in June 1992:
One of the biggest surprises at the Winter Consumer Electronics Show was Nintendo’s announcement of some of the specifications for their upcoming Super Nintendo CD-ROM drive. Their press release stated that their unit would be in the stores as early as January 1993, and that it would sell for only about $200! Add in the fact that almost all of the specifications they published equalled or exceeded the ones for Sega’s Mega CD-ROM, while the price was only about half of what Sega’s unit was selling for at that time in Japan ($370). (EGM #35, June 1992, 48)
What ultimately happened is that Nintendo dealt with both firms to honor its agreements simultaneously. In October 1992, Sony and Nintendo announced the Play Station, set to “combine Nintendo's Super NES home video game system with a CD-ROM drive” (The New York Times 1992), for which Nintendo would control licensing over all game software, while Sony would deal with non-game software (multimedia encyclopedias, music, educational titles, etc.). In parallel, Philips would be manufacturing a CD-ROM add-on to the Super Famicom and Super NES—the SNES Nintendo Disk Drive—whose software would be compatible with Philips’ CD-i player.
These plans were all set to go. Electronic Gaming Monthly had a feature on the Super NES CD in their March 1993 issue, discussing technical specs and claiming that development had been recently finalized at Nintendo of Japan. The general “gaming gossip” section even stated: “As you’ll read in this ish, the Super NES CD-ROM is far from vaporware!” (EGM #44, March 1993, 46) Nothing materialized, however, and the project of a CD-ROM add-on just went up in smoke. The whole project had started as a strategic necessity to stay relevant in an era dominated by CD-positive games futurology, which Nintendo’s competitors NEC and Sega fully embraced. But as months passed, Nintendo observed the occasional misfortunes and general low performance of the Sega CD add-on, which was adopted by fewer than 10% of Mega Drive owners (Screen Digest, March 1995, 60). Nintendo passed up on CD-ROM technology, pulling through the years with expansion chips in its SNES cartridges and pushing for its next console, a 64-bit system initially known as “Project Reality” that would become the Ultra 64 and finally the Nintendo 64. It would use cartridges, good old proprietary cartridges, with their high barriers to entry, guaranteed profit margin, and manufacturing locks, on which the Nintendo Economic System hinged.
The CD-ROM debacle led to a new discourse surfacing; because Nintendo wasn’t going to have CD-ROM technology, it could reject it wholesale. Articles appeared where Nintendo decried CDs’ loading times and argued for a “purity” and “back to basics” approach—a natural ambition for a firm whose business revolves around reiteration rather than innovation. Nintendo Power famously compared the difference in speed between cartridges and CD-ROMs to the difference between the travel speed of a space shuttle and a snail (as seen in the chapter’s epigraph). Even better, the appeal to speed was also a recuperation of Sega’s main war horse in marketing the Genesis as a system that was all about speed (see chapter 3). Nintendo’s newfound discourse was, in a sense, a jujutsu move on Sega’s brand identity, using its own strength against itself.
Although Nintendo did have a point, it was given somewhat of an undue importance by the Japanese giant, especially when compared with the digital soundtracks, sprawling environments, lush prerendered 3-D cinematics, or Full-Motion Video clips that CD-ROM games allowed. Nintendo might not have been interested in making games based on these new, CD-dependent features, but many of its third-party licensees were. Additionally, cartridge production costs were much higher to begin with and inevitably translated to a higher price point for the consumer, which together with Nintendo’s stringent quality assurance process stymied experimentation and the development of risky, innovative game concepts. Silverware and spoony bards were fine as a dining experience and brought in good money, but a variety of other offers would simply never exist in this context. Many third-party game developers were enthused by the possibilities of CD-ROM storage and ultimately left Nintendo and worked mainly or entirely on rival consoles—notably Capcom, Konami, and Squaresoft.
The Rise and Fall of the Licensees
Initially, Nintendo had set on the path of closed platforms with the Famicom. It had opened to others by necessity, not by choice (Gorges 2011, 49), and put up stringent control mechanisms to tolerate licensees. Now, after Sega had shown up, it had started working on its attitude with the Super NES, but it amounted to little more than accepting them in its garden. It sure didn’t make a platform for them. One place where this was perfectly clear was in the documentation for developers, which Nintendo provided.
Nowadays, a slew of measures to help third-party developers support a platform are considered to be a baseline requirement. However, it was not the case for earlier platform owners. In an interview with Masami Ishikawa, who designed the Sega Mega Drive, two answers are given that illustrate the mindset of the time:
How did developers create games for the Mega Drive? Did Sega supply development kits or frameworks?
MI [Masami Ishikawa]: As far as I know, they were using ZAX Z80 emulators to develop game programs. It was nothing like modern-day software development environments, which are equipped with libraries and SDKs. I recall that programmers were studying the source codes of the test program I made for debugging in order to develop each function.
[…]
Did games developers and designers have any input into how the hardware was designed?
MI: The process was not like it is today—we did not ask software developers for opinions. We simply had a one-way meeting when we finished drafting the specs. (Stuart 2014)
Nintendo and Sega’s documentation for their 8- and 16-bit consoles, for instance, consist of little more than lists of addresses and dry data, devoid of any explanations and contextual information. Moreover, the documents have been translated from Japanese into an overly cryptic and byzantine form of English that anyone who has played 8-bit video games will readily recognize.4 As a matter of fact, Australian developer Beam Software got its authorization from Nintendo and became one of the first foreign developers for the Famicom by reverse-engineering the Famicom and producing quality documentation in better English, which Nintendo eventually started distributing to its third-party developers.5 Parsing out Nintendo’s Super NES documentation may have felt more like deciphering an inventor’s private notes than reading an instruction manual explicitly written with an effort to be helpful, resulting in a difficulty of access that was further compounded by the “silverware” architecture of specialized components requiring specialized knowledge to handle correctly (to say nothing of the platform’s legal and financial barriers to entry described in chapters 1 and 2). As a result, all it took was a developer-friendly platform and business model to steal Nintendo’s thunder, leaving it alone to dine on its precious silverware. This is what happened between the mid-1980s and the mid-1990s.
The following figures compare data on the sales of games by publisher for different platforms. For each platform, I took the listed total worldwide, cumulative-to-date sales for software titles, as tallied on the VGchartz website. I then made a list of all million sellers (titles that have sold more than 1 million copies), a common benchmark for noteworthy and profitable games, the goal of any game publisher in the hit-driven video games market.6 I then tallied up the total count of software sales for each publisher by adding up the sales from these million sellers. Figures 7.1–7.4 represent not how many games were made but how many “hit” game copies were sold by each publisher.
Figure 7.1 Publisher shares of total sales of million-sellers for the Famicom/NES. (Total: 233.68 million copies)
Figure 7.4 Publisher shares of total sales of million-sellers for the PlayStation. (Threshold: 5 million copies; total: 426.65 million copies)
Figure 7.2 Publisher shares of total sales of million-sellers for the Super Famicom/Super NES. (Total: 143.6 million copies)
Figure 7.3 Publisher shares of total sales of million-sellers for the Nintendo 64. (Total: 142.75 million copies)
As the data show, the spread of sales is limited on Nintendo’s consoles, particularly when compared with Sony’s PlayStation. The Super Famicom is also the platform where third-party licensees sold the most copies of games compared with Nintendo—in other words, where they encroached on Nintendo’s software-side garden the most. The Nintendo 64 numbers show the isolation in which Nintendo found itself. Why was there a lack of third-party game developers to support the Nintendo 64? Because by the mid-1990s, most of them had left the Nintendo playground for the greener pastures of Sony, who wowed them with two key technologies that were all the rage in the 1990s: high-powered real-time polygonal 3-D processing and the CD-ROM as a high-capacity storage media. Beyond the “wow” factor, however, Sony seduced them because it actually courted them.
The Sony Creative Expansion (SCE) Model
When Sony entered the video game industry in 1994, its policies shifted the positioning of third-party developers and publishers from outside the platform ecosystem to inside its boundaries. In this respect, Sony leaned further away from vertical integration and more toward the horizontal form characteristic of the network organization. This was entirely different from Nintendo’s system and pushed Sega’s openness further. The PlayStation was not there primarily for Sony to publish its own games (symptomatically, its launch in the United States did not include a pack-in game, contrary to about every other major video game console so far) but was a game machine destined to play a wide host of different games, just like the record, VHS, and CD players were not manufactured for a particular film studio or record label to push its entertainment products.
Like Nintendo, Sony would control the manufacturing and distribution of licensed games. Like Sega and unlike Nintendo, it would sell the system at a loss to build market share and make money on subsequent software titles. Unlike both Sega and Nintendo, it would not make its own games the core of the PlayStation’s business proposition but rather put extensive effort into making things easy for third-party developers and publishers. Where Nintendo was a reluctant gatekeeper that carefully screened anyone who showed up to enter its walled garden, Sony not only opened wide the licensing gates (like Sega) for any game to be made as long as the developer paid the fee, but went an extra mile by offering rides and guided tours to make sure everyone would be comfortable.
From Platform Owner to Platform Provider
Although this shift in emphasis may seem subtle, it leads to a major difference in the role of the platform owner. Whereas Atari, Nintendo, and Sega were platform owners—game developers and publishers that also developed and sold game machines and accepted some licensed third parties for complementary goods—Sony stepped in as a platform provider. Instead of being a game developer, the platform provider is a technological supplier and production facilitator for game developers and publishers, playing the role of the middle man between consumers buying games and the parties that make them. It had to develop its own games, of course, to kick-start the adoption of its platform, but internal game production was a means toward an end rather than the other way around.7 Whereas Nintendo distinguishes itself from other platform owners through its software orientation, Sony can be said to have adopted a service orientation. As a technology company with no expertise in game development, it approached the video game industry from the perspective of a hardware manufacturer and standard bearer, providing services to other firms that would produce the content. Sony’s template became the new blueprint for subsequent entrant Microsoft, which adopted the model with its Xbox console.
Sony’s plan had all the right tools to accomplish this revolution. On the development side, they assembled an internal team solely dedicated to helping out third-party developers and listening to their feedback. Moreover, they supplied a software development kit to game developers, with proper documentation, sample code, and detailed instructions to help them work with the intricacies of the hardware. More important, they also innovated by providing software libraries to developers, modular parts of code supplied to deal with common features that developers often needed to implement—a kind of basic building block. This new proposition in the world of home consoles initially required adaptation from developers, who had to trust the platform provider and work from the tools it provided them instead of writing every single part of their code from scratch. Developers quickly got over their sentiment of losing control however when they realized how streamlined it made development: They could now stage a structure with Lego bricks instead of spending their time melding their own plastic pieces.
This was a different kind of brick-handling than what licensees had been subjected to in the Kyoto firm’s imperial garden.
Bricks in the Network Wall: The Nintendo Economic System
The Nintendo Economic System, with its various criss-crossing policies, constructed a hybrid organizational structure that straddled the categories of hierarchical and network structures in a demonstration of what Melissa Schilling qualified as modular organizational forms relying on contract manufacturing to engage in “loose couplings” (Schilling 2003b). Jennifer Johns (2005) has similarly detailed the global and international production networks of video games for both the hardware and software sides of the market.
Hierarchical and bureaucratic structures allow a firm to be more focused and efficient in attaining its goals. A common form of business-to-business relationship in this context is for a firm to attempt vertical integration, which consists of buying or taking control over most or all stages of a given product’s life, from production to consumption—for instance, a console manufacturer deciding to acquire a semiconductor manufacturer rather than purchasing its semiconductors. The downsides of vertical integration are that it decreases a firm’s flexibility and concentrates risk because the corporation owns every segment of the product chain. For example, if a new technology suddenly renders semiconductors obsolete, then the vertically integrated firm is stuck with useless semiconductor factories and needs to spend resources dealing with this. On the upside, vertical integration increases coordination and control over the market, supply, and profit margins; a vertically integrated firm depends on fewer intermediaries, which is a rare and valuable freedom in business.
Among the many other types of organizational structures, the network is the most pertinent to understand Nintendo. The network structure increases cost-efficiency by having component suppliers, distributors, and all other actors involved along the product delivery chain undercut each other through competition; it also distributes risk among all participating firms. In the tech industry, this advantage is particularly important because technology firms need to venture into research and development (R&D) and new technological developments all the time, always risking that its manufacturing facilities get stuck on a bad technology gamble. This is why the tech industry is particularly volatile for investors. It’s also worth noting, following Casey O’Donnell, that although network structures are typically regarded as being less hierarchical, it is not an automatic principle:
Too often networks are talked about as inherently open, better, or different than hierarchical systems, yet networks can be just as hierarchical. There is nothing fundamental about networks that make them naturally flat or more open. They must be constructed in ways that enable flatness or openness. (O’Donnell 2011, 95)
Nintendo has made extensive use of the network structure in developing a “fabless” (“fabrication-less”) production model: “At Nintendo, we do not own the production factories that manufacture our products. All production processes are outsourced to external suppliers and production factories (production partners)” (Nintendo Co. 2016b), Normally, the trade-off is that the network structure limits the firm’s control over its partners’ processes, thus preventing tight integration and resultant economies and introducing delays in manufacturing as the different partner organizations each negotiate their own schedules. A semiconductor manufacturer may reduce the production volume of a chip Nintendo needs to start producing a new standard, leaving Nintendo empty-handed or facing chip shortages—which is exactly what happened in 1988 (Lazzareschi 1988).
But Nintendo’s business architecture has managed to minimize almost every disadvantage and achieve a best-of-both-worlds situation, a vertically integrated network—a wall. The first, counterintuitive element we need to acknowledge to truly understand Nintendo’s positioning is that Nintendo is not a technology company. It lets technological firms such as Cisco, Intel, Silicon Graphics, IBM, and so on innovate and take the risks, along with the commensurate rewards, to develop the right technologies. Then, as these firms sell their new technologies at premium prices to recoup their R&D expenses, it looks at the “last-gen tech,” those seasoned or withered technologies in an industry where everyone else is using the top-of-the-line novelties. In this, Nintendo occupies a unique place as an entertainment company with a “software orientation,” contrary to its modern competitors Sony and Microsoft, which are technology companies that leave most of the software to third-party licensees.
Nintendo’s reliance on hardware suppliers organized as a network would normally rob them of the advantages of vertical integration—controlling costs and supply. But in fact, Nintendo still manages to control costs and supply by resorting to “old” tech because the rate at which new technology costs go down is appreciable. The other way of bringing down costs is to place large orders, which allows streamlining of the process, stability, and predictability—all desirable features for a production firm. The resulting economies of scale can confer enormous advantages that make ownership of the production process through vertical integration almost irrelevant. As an indication of the scale we are talking about, Steve Jobs explained that the same 68000 processor used in the Apple Lisa had its price shaved off by 80% for the large quantities ordered for the Apple Macintosh (Sen 2012, 32:00). These are not questions of details. Nintendo’s phenomenally large initial order for the Famicom from semiconductor supplier Ricoh resulted in such rock-bottom costs that the firm managed to market the Famicom at a price point no competitor could match.
Now, if Nintendo’s strategy is so good, then why don’t other firms opt for it too? Part of the answer is that Sony and Microsoft are large tech companies, with games being only part of their activities; developing, manufacturing, and selling technology products is their main business. The other part of the answer lies with Nintendo’s unique fiscal policy of accumulating savings and holding onto them. Whereas corporations typically take on strategic debt and invest capital in the hope of creating growth and generating ever-increasing returns on investment, Nintendo stockpiles cash and assets. Inoue wrote in 2010: “In December 2008, its cash and cash-equivalent assets totaled more than one trillion yen” (Inoue 2010, 100)—the equivalent of $11 billion. That’s with absolutely no debt to repay, an anomaly in the modern corporate world. These cash reserves are essential for Nintendo to place large orders from committed manufacturing partners: “For a fabless company that doesn’t own manufacturing facilities, the cash is important as a guarantee of credit” (Inoue 2010, 101).
The combined effects of these elements help us appreciate how Nintendo built its Super Power. Through its immense cash assets, it secured very advantageous prices on older and cheap components, alleviating the downsides of not practicing vertical integration and freeing it from long-term commitment to any technology. It also allowed Nintendo to market consoles without subsidizing and needing to recoup their hardware losses, and the nonstandard architectures gave Nintendo a lead time on third-party developers to sell its own games. This more than enviable position was the result of former president Hiroshi Yamauchi betting the bank (literally) on the Famicom/NES, investing all profits from the hugely lucrative Game and Watch portable systems to develop the Famicom (Gorges 2011). As the gamble paid off with immense profits, Yamauchi resisted the temptation of diversifying the company’s activities, creating more products, borrowing money, and making acquisitions (a decision probably informed by its failed diversification attempts in the 1960s and 1970s). Instead, he put all Nintendo’s money in a war chest to weather the inevitable storms to come in the video game industry’s notorious boom-or-bust business. He was going to need it, time and again, but especially so against the biggest shock Nintendo would face: Sony’s PlayStation.
Behind the “Sony Shock”: Technological and Commercial Innovation
What happened in the closing years of the Super NES’s life was a marketing and distribution revolution to which Nintendo failed to respond. This marked the second time Nintendo failed to react appropriately as the incumbent market leader (Subramanian, Chai, and Mu 2011, 230), following Sega’s successful Genesis push in North America. Both of these situations confirm the important impact of two factors studied by Melissa Schilling (2002): a learning orientation, which Nintendo lacked because of the glittering NES gold and shiny SNES silverware, and an appropriate timing of entry, which the big N was too slow in seizing. Innovation from competitors sent them reeling.
Veryzer’s model of technological innovation (1998) considers a firm’s technological capability as separate from its commercial capability; innovations may likewise inscribe themselves in technological continuity/discontinuity and commercial continuity/discontinuity. Subramanian, Chai, and Mu (2011) summarize technological discontinuity as taking place “when a new functionality is introduced in an innovation that is beyond the scope of the industry’s existing technological knowledge base” and commercial discontinuity “when existing customers change the way they perceive and use the innovation” or “if the innovation is capable of attracting a new set of customers as its potential users.” The Super NES, as should be clear by now, is continuous on both counts. Moreover, as I’ve demonstrated, Nintendo remained on course and resisted every new technologically discontinuous innovation that appeared during the SNES years, while also enforcing commercial continuity through its Nintendo Economic System as much as it could get away with. This may explain just how badly Nintendo was struck by Sony’s entry in the video games market, an entry that was founded on a double discontinuous innovation with the PlayStation, both technological and commercial.
Sony contributed to the structure of the home video game market by giving more creative support to game developers but also because its CD-ROM technology and distribution infrastructures changed the fundamental dynamics of the market’s supply and demand by dramatically lowering the manufacturing cost of game software and considerably augmenting the flexibility of its distribution. In this sense, Sony resolved some of the most important tensions inherent in marketing disruptive technologies: establishing customer-oriented distribution and distribution-oriented pricing (Moore 2014, 197–212). The production of cartridges involved long delays, high initial production volumes, and fixed shipment sizes, resulting in a general lack of agility in responding to the demands of the market and a need for nothing less than clairvoyance in predicting the market’s demand, as Malik (1997) details:
While the Sony PlayStation uses CD-ROMs, Nintendo 64 uses cartridge-based games. CDs are preferred by developers for a variety of reasons. For starters, a CD can hold up to 650 megabytes of data, while a cartridge’s storage capacity is only 16 megabytes.
In addition, Nintendo makes these cartridges in Japan and it takes about three months for developers to lay their hands on the blanks, which means the game developers have to second-guess the demand and run the risk of making a costly mistake. Blank cartridges sell for around $35, while blank CDs sell for about $6.
Wrong forecasts may leave developers with either huge excess inventories or not enough copies of a hot title. In comparison, CDs have a turnaround time of less than two weeks. “CDs give higher margins to third-party developers, one of the main reasons they are attracted to the Sony PlayStation platform,” adds IDC’s Zinsmeister.
A higher installed base, which means higher volumes, a bigger library of titles and a cheaper medium, gives Sony a price advantage. The company can sell PlayStation games for about $35 in retail outlets, while Nintendo games are in the $75 price range thus limiting sales. The problem is exacerbated by Nintendo’s demographics—most N64 players are in the 8-to-14-year-old age group. Kids generally don’t have $75 burning a hole in their pockets. (Malik 1997)
In contrast, “PlayStation CD-ROMs could be produced in required quantities in as little as one week, and shipped in less than 24 hours thanks to Sony Music’s efficient distribution service. And we could even produce very small orders of tens of copies” (Akagawa 2013, 106). CD-ROMs allowed a model of “repeat business,” where game quantities could be ordered repeatedly as they were sold instead of huge (and risky) quantities determined from trying to guess demand months ahead of time. This occasioned a paradigm shift from production- to customer-oriented distribution: Consumer demand would pull in additional production as needed instead of top-heavy production trying to push its products down on consumers.
Beyond Distribution and into Genre Innovation
This also translated into a platform more conducive to innovative games. Cartridges were produced using mask ROM, a type of memory that required the programming code to be locked down and converted into a mask, which would imprint the instructions into the chips themselves at the time of manufacturing. Producing the mask was time-consuming and costly. Combined with Nintendo’s restrictive supply planning, the Nintendo Economic System was elitist and exacting, built to support high quantities of relatively few titles by developers with enough financial means to support the upfront costs to meet minimum order quantities. Sony gave more creative freedom to developers by lowering the entry barriers to the market, and its model was based on providing access to a large quantity of games. It achieved this goal by maintaining the game developers’ profit margins and cutting its own, all the while cutting the game’s retail price to consumers by almost half. These terms and pragmatic conditions stood in stark contrast to Nintendo’s, as table 7.1 shows.
Table 7.1 Cost and price breakdown for producing a game on Super Famicom and Sony PlayStation, derived from Asakura 2012, chapter 4.
Super Famicom
PlayStation
Retail price (for consumers)
$98
$58
Manufacturing cost
$15
$1.50
Royalty (to platform owner)
$15
$7.50
Equipment amortization cost
$10
$10
Advertising
$6
$6
Risk insurance on stocks
$5
None
Developer profit margin
$10
$10
Wholesaler profit margin
$12
$6
Retailer profit margin
$25
$17
The model would work if consumers kept on spending as much money on games as before. The same amount would get them almost twice as many games, which would amount to total royalties to Sony about the same as Nintendo could get but would double game developers’ profit margins (and possibly distribute them across twice as many game developers). Hence, Sony’s strategy also worked as a mitigating factor to the video game industry’s hit-driven market, where it is typically assumed that 20% of the games make 80% of the profits (Kline, Dyer-Witheford, and de Peuter 2003, 113). The commercial affordances of Sony’s PlayStation promoted innovation and risk-taking, as much as its technological affordances of increased storage space in CD-ROMs and real-time polygonal 3-D graphics.
When looking at table 7.1, it would seem that wholesalers and retailers would be refractory to Sony’s pricing strategy, given the important reductions in their profit margins. Despite the raw numbers, however, the logistics of production and distribution proved highly desirable for them. Unlike product development and production, the sales and retail world revolves around product circulation and transition; inventory has to be managed, storage space is wasted space, and the ideal good is one that arrives in timely fashion, in the quantities ordered, and that flies off the shelf quickly to get more goods in and out. Nintendo cartridges were always selling but notoriously difficult to plan for, with turnaround calculated in months and quantities in fractions of registered orders. PlayStation games would net them less margin per unit, but they were priced at a sweeter spot for the customer, meaning more units would sell. They took less physical space, thus lessening the impact of unit margins in terms of volume. These combined factors led to Sony’s platform being more attuned to distributors’ pricing needs than previously with cartridges (especially proprietary cartridges exclusively manufactured by a platform owner who could blacklist them for going six cents below the suggested retail price, as seen in chapter 1).
This was, as Asakura (2012, 110–115) detailed, Sony’s PlayStation Revolution: a technological revolution that appealed to game developers and a commercial revolution that appealed to them as well as retailers and distributors. Gamers, as the end users, were all happy to join the movement when they saw the quality and quantity of games on offer. The bricks had been taken out of the Nintendo Economic System walls and reconfigured as a paved road on the ground. The sun shone through, its rays dancing through the vivacious gardens of Sony, basking in the glow of a new Golden Age.
Into Twilight
One by one, Nintendo’s former subjects, useful in bringing secondary revenue and expanding their installed base, became Sony contributors. All the weight they had gained by publishing on Nintendo’s platforms suddenly shifted around, magnifying Sony’s presence. Konami graced the PlayStation with emblematic releases such as Suikoden, Castlevania: Symphony of the Night, Metal Gear Solid, Dance Dance Revolution, and Silent Hill, while the N64 received the “alright” Castlevania 64, Quest 64, and Hybrid Heaven. Capcom developed its Resident Evil and Breath of Fire series, as well as its Mega Man and Street Fighter characters on Sony’s platform, whereas Nintendo had to wait more than a year for ports of Mega Man Legends and Resident Evil 2. However, nothing hurt as bad as Square, with whom Nintendo had collaborated for Super Mario RPG and had done tests for Nintendo 64 hardware. Square brought its Final Fantasy, Mana, and Chrono series to the PlayStation. In the end, these developers wanted CD-ROM storage and creative freedom, and Nintendo wouldn’t give it to them. Licensee support plummeted as everyone left the Nintendo garden and its obscuring walls for the evergreens of the Sony kingdom. There had been the Resurrection with the NES first, then the Second Coming with the SNES, and the Crusades against Sega; now, finally, the Exodus was playing out.
Nintendo had committed to its self-party strategy. It wasn’t interested in adopting technical standards such as data storage media from other corporations, preferring to retain absolute control over manufacturing instead. It opted for a strategy of technological leapfrogging (Schilling 2003a) with a 64-bit processor, looking to embed technological superiority into the name of the console. In retrospect that was quite a leap to attempt, given that 64-bit cores would reach widespread adoption in PCs only a full decade later in 2006, when Intel released the Core 2 Duo processor for mainstream computers. Here, Nintendo may have been swept in by the marketing “bit wars” (Therrien and Picard 2015), which focused on consoles’ bit power (as seen in chapter 3), or it may have stumbled into Wesley and Barczak’s “performance trap” (Wesley and Barczak 2010). Did customers need 64 bits of power? No, they needed great games. Nintendo was all about making great games, so it did not need a large supply of third-party games. It may well be that Nintendo underestimated the draft of partners it would have to face. After all, developers and publishers had always put up with Nintendo’s ways, eventually. But pride comes before a fall and arrogance was never too far off, as Jeff Ryan perhaps best captured when describing the transition from the Super NES to the Nintendo 64:
With enough great games, Nintendo would be able to ride out the lack of third-party developers. Who cared if the shelf was mostly Nintendo for the first few years? Most of the other games merely gave the illusion of choice. … N64 gamers, like SNES and NES gamers before them, wanted Nintendo games. They wanted Mario, and Link, and little else. (Ryan 2012, 188)
