By Dr Hakan Ozalp
About the author
This article is based on Hakan’s recent co-authored publication ‘Platform Architecture and Quality Tradeoffs of Multihoming Complements’ in Information Systems Research (CABS 4*).
In 2005, a new generation of video game console competition began. This was the year Microsoft released the Xbox 360, looking to compete in a market that Sony’s PlayStation consoles had previously dominated.
The Xbox 360 came to the market one year before Sony’s next generation of console, the PlayStation 3, was released. The delay in the launch of the PlayStation 3 was due to Sony’s extensive research and development (R&D) process.
Led by IBM and Sharp, R&D for the PlayStation 3 began in 2002, with a focus on the new central processing unit (CPU) technology, Cell. The belief was that this CPU technology had the capability to be so advanced that competition from the likes of Microsoft would be impossible. In 2006, the PlayStation 3 was released and, as promised, its CPU technology was incredibly advanced. However, being the most technologically advanced appeared to come with its own set of disadvantages, particularly if you are a games developer.
Coding games – for the many or the few?
Multihoming games are games developed for multiple competing platforms with a version of the game developed for each platform (eg Xbox and PlayStation) and most games in the current market are as such. In a market sense, this removes competition in terms of gaming exclusives, but it relies on the platform being optimised for this shared resource.
For the Xbox 360, the CPU technology used meant that games developers could use generic development approaches. However, for PlayStation 3, games developers needed to use a specific instruction set for programming specialised CPUs and allocate tasks amongst them. This concept is best understood if you think about it from a game developer’s point of view:
A games development company wants to create a new game. This company has a huge team of coding experts, who all have experience using generic development approaches to make multihoming games. Due to the architecture of the Xbox 360, the games developers know exactly what is needed to make a new product for the market.
The new PlayStation 3 arrives on the market with its advanced CPU technology. This technology means that if the games development company wants to make a game for the PlayStation 3, it will have to train all its coding experts for the specifics of that new architecture. Alternatively, this company can continue to develop games using generic approaches, which is compatible with major consoles such as the Xbox 360, and find a way to modify their new game until it’s good enough to run on the PlayStation 3 – this removes the need to retrain the team.
The latter example, where a games developer kept to the generic development approaches and tried to adapt the game for the more specialised architecture, was often the case for the PlayStation 3. This potentially meant that the superior CPU technology developed by Sony for the PlayStation 3 was in fact a limiting factor behind the capabilities and experience of certain games on this platform.
To investigate this concept further, Dr Hakan Ozalp and his colleagues took a closer look at the complexity of a platform’s architecture.
Do the advanced capabilities of PlayStation lead to a better gaming experience?
In order to test this question, multihoming game releases in the United States for the 7th Generation consoles between 2005 and 2010 were analysed using critical scores. Critical scores are the review scores of games, much like movie review scores. These scores are then combined and standardised on Metacritic.com.
Looking at cross-platform differences in critical scores for identical games in each platform, ie games that are exactly the same for Xbox and PlayStation, such as Fifa 2007, authors found that platform architecture does matter. Multihoming games often received a lower quality score on a more complex platform (such as PlayStation) than on a less complex one (such as Xbox).
An opposite case? The PlayStation 2
When the PlayStation 2 was released in 2000, it sold more units than any other platform. Today the PlayStation 2 is still the most popular console of all time, currently holding the sales record of over 155 million units.
The PlayStation 2 also had a complex CPU design, but nevertheless dominated the market. And in contrast to what’s just been discussed, its complexity was in fact part of its success. So why was there such a different story for the PlayStation 2 than the PlayStation 3?
Analysis of the data for the PlayStation 2 shows that multihoming games again receive a lower quality score, yet the PlayStation 2 shone through in its gaming exclusives, such as Metal Gear Solid 2: Sons of Liberty. As mentioned earlier, most games in the current market are multihoming, developed for more than one platform, however, during the reign of the PlayStation 2, games were often exclusives, which were developed for only a single platform. This meant that games developers actively looked to co-specialise with platform technologies and allowed the PlayStation 2 to thrive on its ability to offer games and functionality that could be delivered through tight integration with the complex architecture of the platform.
In summary, research suggests that in a gaming era when exclusives drove sales and games were often restricted to a single platform, the highly specialised technology of Sony’s PlayStation was a winning formula. However, in today’s market where multihoming games are commonplace, it pays to keep your platform technology simple and easy to use for games developers.
So what does that mean for the following generation consoles, which we still play with today? All of them ended up with architectures quite similar to PCs (except for Nintendo) – sharing most of the development code and resources across all platforms. And what will happen for the next generation of consoles? For now, we’ll have to stick to reading the rumours.
To hear more about Hakan’s research into platform technology and multihoming complements, you can watch his recent presentation as part of the Festival of Ideas (starts at 18 minutes and 32 seconds).