Development Timeframes and Structure

A typical game project has a very different timeline than some public policy or government projects. While high-end projects can take two or three years to complete, most games are developed over a 12-18-month timeframe. Very little of this time is spent on traditional R&D priorities. Most game developers do the bulk of their R&D alongside basic development work; even if their R&D work is done separately it is usually measured in months, not years. The result is that game developers have little appreciation for traditional research and development work.

The Council's report also highlighted the fact that many games have a short lifespan and that little effort is given to upgrading or improving them as they age. Both of these points remain valid, although perhaps less so than in 1996. In most cases these problems arise due to the nature of the entertainment market itself, not because of specific practices adhered to for other reasons. The game market has always brought out sequel products such as John Madden Football 1999, 2000, 2001. However, only recently have developers begun to respond more favorably to updating previously released products with a series of upgrades and extensions over their shelf life of the original product.

Valve Software's Half-Life is a great example: the core product has seen numerous free, and paid add-ons developed for it by Valve and developers outside of Valve. This new practice is limited to PC products as the architecture of consoles makes the cost of upgrades prohibitive. Many top PC-oriented developers such as id Software, Valve, and Maxis all have released several upgrades to existing hit products. The existence of the Internet and high-speed lines has made this increasingly easy to do. Developers who have pushed this practice have benefited from the upgrades and, along with other developers, there is an emerging appreciation within the industry for the idea of continuous development on pre-existing titles. As more developers realize the benefits of frequent updates, extensions, and progression of their core titles, this will become standard practice. As a result, developers will provide integrated tools and technologies to further automate this process.

Many titles or projects that would be developed in cooperation with government and non-governmental organizations would be designed to have much longer life spans with more continuous development and maintenance than many commercial products have. In the game market, graphical capabilities have grown so rapidly during the last five years that titles driven by their visuals become quickly antiquated, thus driving frequent upgrades. But for many modeling and simulation projects, visual effects will not be as significant. The lesson is that this is less a cultural issue than an issue of contracts, budgets, planning, and markets served with little or no competing products. Unconstrained by market demands for newer products, the latest graphic advances, and sales competition, game developers are more than capable of delivering a product that is developed with longer-term horizons in mind.

Concerning R&D issues, developers are again constrained by the market and their budgets, resulting in little formal R&D by developers and publishers alike. R&D is usually built into an overall products' development and is rarely conducted as a stand-alone process. This means developers are truly focused on finishing the desired deliverable and will not get bogged down in various R&D that may not contribute directly to the project at hand. However, the industry as a whole doesn't necessarily investigate much beyond its immediate needs, resulting more often in innovations that are built on applications of pre-existing research in AI, graphics, sound, etc. than on groundbreaking new discoveries.