In many ways the Saturn was on the back-foot from day one. While significant volumes of Virtua Fighter were sold during its first few months of sale in Japan, it was a game not without its criticism, with many citing the game's "glitchy polygons", possibly caused by a rushed development timeline but more likely a lack of understanding of how to effectively render 3D graphics on the hardware.
It is a common misconception that the Sega Saturn was developed to be a 2D games machine. While it is certainly true that it handles 2D graphics better than much of its competition, every Sega game demonstrated on the hardware during its development - even in Japan where the 3D craze was not as significant - utilised real-time 3D graphics, and were pushed as selling points of the system. It was, however, very difficult for third-parties to meet expectations during its first six months of sale, leading to Sega AM2's Tadahiro Kawamura creating the Sega Graphics Library (SGL), under orders to ease arcade-to-Saturn conversions by Yu Suzuki.
By mid-1995 the old Sega Saturn Programming Boxes were replaced with cheaper CartDev units, and shipped to developers alongside an SGI Indy workstation, Softimage, SNASM2 and AM2's new SGL software. In a sense this meant that while Saturn development was technically feasible, prior to this date developers were forced to do a great deal of hardware research and library building of their own, slowing down game development considerably.
But despite all of these faults, worldwide Saturn sales were very strong up until the middle of 1996. By April of that year, mostly on the shoulders of Japan, the Saturn has been cited as owning 51% of the worldwide 32-bit console market. But the momentum was with Sony, and Sega quickly lost its commanding position over the summer. Conflicting reports in fact suggest that by the end of 1995, 3.4 million PlayStation consoles had been sold worldwide versus 3 million Saturns, Sony likely having overtaken Sega much earlier. Sega were said to have been aware of the Saturn's poor performance and were hoping to regain ground by February 1996.
From as early as Winter 1995, a persistent rumour suggested that Sega were planning a "64-bit" add-on to the Saturn, codenamed "Eclipse" - something which was publicly denied by Sega. Eclipse was thought to be either a stand-alone system or a unit designed to make use of the Saturn's expansion port. The Eclipse project would have come into fruition in 1996, but is thought to have been held back due to poor response from third-party developers wary of a second Sega 32X debacle. Other factors may have included a delayed Nintendo 64, which would have been the main competitor to the Eclispe.
Sega had dismissed the idea of expanding the Saturn by mid-1997, after finding it could not deliver meaningful improvements for a reasonable price.
After the holiday shopping season in 1996, the Saturn had fallen behind the PlayStation and Nintendo 64 in North America and Europe, prompting calls from Sega of America's senior management for a new platform. Analysts were predicting that only 4 million Saturn consoles would be sold globally (versus 10.9 N64s and 10.8 PlayStations), and production was slowed to 3 million units for 1997 versus the originally planned 4 million.
For most of the 1990s Sega of Japan and Sega of America were at loggerheads. Japan had intervened constantly since the Saturn's launch, to the point that by 1997, there was no official Saturn development happening stateside. Sega's US operations comprised of SegaSoft and various forays into the online PC market, and so unlike the Genesis (and even the Mega-CD and 32X), technical support for third-party developers was difficult to come by. This meant that Saturn ports became less and less of a priority for developers, and many planned releases were cancelled across the year.
Sega's idea was to get the jump on the next generation of systems ahead of their competitors, and so by E3 1997 Sega had already begun talk of the new system, codenamed Katana, which would eventually turn into the Sega Dreamcast.
As Sega started aggressively moving the Katana project forward it caused something of a rift between Sega and many third party developers. The Saturn was more than holding its own in the Japanese marketplace where the vast majority of Sega game development was based. As a result many Japanese developers saw little to no reason for Sega to rush another platform to the market, which would in the process, effectively kill the Saturn despite its large user base and many active development projects.
Meanwhile in the West, Saturn projects were abandoned en masse during the summer of 1997, including all development by Shiny Entertainment (with Sony paying for Wild 9's exclusivity), Eidos Interactive and its subsidiaries (including sequel to one of the best selling Saturn games, Tomb Raider II and Saturn-led Ninja), Virgin Interactive and THQ (which were mainly concerned with ports of Psygnosis games, namely Destruction Derby 2, Sentient and Tenka (with Wipeout 2097, Krazy Ivan, Discworld and Assault Rigs missing their US launches)). Konami and Eidos Interactive also abandoned the Saturn around this period, and Electronic Arts also cut down development.
Towards the end of 1998, newly elected president of Sega, Shiochiro Irimajiri stated in an interview with Yomiuri Shimbun that production of Sega Saturn hardware would cease by the end of 1998, with software production ending by mid-1999. Instead the focus would be on the Katana console, now known as the Sega Dreamcast.
The Saturn would also be discontinued in late 1998 for Europe and April 3, 1999 in North America (though in both regions the system had been, for all intents and purposes, "dead" since mid-1998). Sega's history would damage the Dreamcast's reputation, with notable publishers such as Electronic Arts refusing to back the system, having made losses on the Saturn. It is widely considered that the Saturn was simply not up to the task of competing effectively in the fifth generation of video game consoles, joining the likes of the Atari Jaguar, 3DO, CD-i and Sega's 32X as hardware casualties.
Incidentally Sega held a stake in Atari Games during the 1990s and also worked closely with SNK. In both cases, the plan was for Sega to produce titles for both the Jaguar and Neo-Geo consoles respectively, in return for support on the Saturn - neither deal really materialised and Sega's games stayed on Sega hardware, though SNK kept its side of the bargain by releasing various Neo-Geo fighting games on the Saturn.
Estimates for the Saturn's worldwide console sales range from 9.5 million to 17 million - significantly less than both the PlayStation and the Nintendo 64 (despite the latter arriving more than a year late). 8.8 million console sales and 80 million games can be accounted for, with Sega having reached this number by the end of its 1998 fiscal year (1998-03-31).
In Japan, the Saturn continues to stand as Sega's most successful console outing, outselling the Nintendo 64 by a considerable margin, and having an initial lead over the PlayStation. However, the PlayStation became the mass market machine of all regions of the world, trumping both the Saturn and Nintendo 64 and becoming the defining system of its generation.
While strong in representing 2D video games, the Saturn is considered to be ill-equipped for the world of 3D gaming born in the mid-1990s. When it was designed, 3D games were still a novelty and texture-mapped polygons even rarer, however by 1997 3D was no longer the just the future, but very much the present. 2D gaming was seen as a thing of the past - something associated with "16-bit" consoles of the Mega Drive and Super Nintendo, not the $300+ 32-bit powerhouses. There was an expectation that platform holders would offer something truly new for the price, but the Saturn is not thought to have fully delivered on that promise.
While the Saturn frequently benefited from the experimental period of early 3D gaming (quadrilateral-based polygons (which the Saturn employed) were for a while, just as widely used as the triangle-based model of the PlayStation and later consoles (so in some cases, Saturn ports fared better than their PlayStation counterparts)), it was on the wrong side of the argument in many respects and ultimately failed to keep pace with technology. With the Dreamcast, Sega sided with graphics specialists and adopted more mainstream ideas and formats.
In mid 1995, Sega AM2 were demonstrating prototypes of Virtua Fighter 2 running at 704x480 at 60FPS, suggesting the capability was there to produce very competitive Saturn titles. However, AM2 were among the most skilled developers within the organisation and were very familiar with the hardware - third-parties rarely matched these statistics, and were unable to utilise the hardware fully.
The design of the Saturn was complex compared to the PlayStation, with various different processors (Sega of America claimed there were three, though you could argue there were as many as nine) all being used in tandem. AM2 wrote and released various development libraries to assist in Saturn development, but for months these were considered inadequate, and with many multi-processor setups, you could never achieve maximum efficiency when resources were being shared.
The Saturn was the last mainstream video game console to favour assembly language over higher level programming languages such as C (which was an option for developers, but the drops in performance made it unsuitable for demanding software). While there were many holdover specialists from the Mega Drive and Super Nintendo days, it was significantly harder and more time consuming to build a Saturn executable.
As a result, most multi-platform games 3D fare worse than their PlayStation counterparts. Simplified development tools were a priority for the Dreamcast, and the addition of Windows CE theoretically allowed Windows programmers to easily adapt to the console.
In 1995, Tom Kalinske publicly acknowledged that for most months of the year, Nintendo would "beat" Sega when it came to both hardware and software sales in the US. However, when it came to the "important" months of November and December where significantly more sales occurred, Sega would catch up and usually win. With the Saturn, this strategy stopped working, and played havoc with the US launch in May 1995.
This meant the Saturn (and the Dreamcast - it wasn't a lesson fully learned) often suffered in the summer months, as the system was not typically marketed heavily and the games schedule was spread thin. Conversely, while all video game companies skew themselves towards the Christmas period, other platforms were able to put out games all year round, or perhaps more importantly, give the impression that they were.
Though the Sega Saturn is not amongst the most fondly remembered video game consoles, it is still considered noteworthy. The 3D Control Pad set the standard for analogue control pads going forward, introducing the concept of analogue shoulder triggers and a thumbstick on the left hand side of the controller (something continued with the Dreamcast and later the Xbox line). Its six face buttons also made its controllers ideal for fighting games.