Japanese Game Development: The Path Forward
November 23, 2010 Page 3 of 4
Rethinking the Structure of Large Creative Teams
Most of the big Japanese video game development projects were described to me as taking place in rigidly structured, top-down organizations where employees are divided into sections by discipline and do not feel empowered to suggest their own ideas as they relate to areas not under their immediate jurisdiction.
I feel that this method of complete control from a single section leader can work in smaller-scale productions, but as we approach the team sizes required to make a big budget game in today's market, the number of decisions that need to be made on a daily basis can easily become overwhelming and difficult to coordinate. In these cases, inter-department communication suffers as each team specializes in its discipline to the exclusion of others.
A postmortem of Final Fantasy XIII published in the October issue of Game Developer magazine describes this problem succinctly:
As the project's scope increased, the traditional development style of dividing the team into specified roles, such as character modelers or texture artists, started to present issues as well. This problem of over-specialization presented itself in each discipline. The biggest problem was that the project became bloated with the increase in staff within each department. And because roles were so specific, the communication flow became faulty and information was not being shared properly.
To anyone with experience in large-scale game development, these difficulties will seem familiar. We have all heard the typical complaints about what went wrong during a big game development project such as "the programmers didn't do enough" or "the artists didn't understand the system".
How can stumbling blocks like these be addressed? One of the promising alternative methods of team structure being explored is organization by feature instead of by discipline (this structure often comes to game developers via partial implementation of Scrum methodology).
For example, a "hero character team" may consist of a small interdisciplinary group of a concept artist, a 3D modeler, a texture artist, a rigger, an animator, and an AI programmer. These team members would be charged with ensuring the hero characters of the game looked, acted, and performed their best.
The small size of the group would enable the team to be physically located next to each other and have regular cross-disciplinary communication. For example, the animator might have ideas about how the AI programmer could handle blends between certain behaviors. Or the rigger could have feedback on the proportions of the character as designed by the concept artist.
These various concerns are more easily brought up, discussed and addressed in the smaller setting of a specialized "character team" as opposed to the typical array of large teams separated by discipline.
One of the criticisms of such a team structure is that disciplinary standards can become more difficult to maintain. If an animator is part of the "character" team, and there are other animators spread across the studio on the "enemy" team and the "cinematics" team, animation standards and direction may become more difficult to disseminate.
In this case, animation team meetings could also be held with the animation director supervising and approving all animation across each feature team. In other words, an individual animator might be understood to posses "dual citizenship" to both the larger animation team and to the interdisciplinary feature team to which he or she is assigned.
Personally, I believe this approach is preferable because it also increases the involvement and engagement of the typical rank-and-file team member. High levels of technical skill and artistic talent are certainly present on an individual basis within Japan's developers, but these qualities are not always combined in such a way as to make a well-rounded final product.
In such a case as this, structural changes are best method to go about releasing a team's true potential. Additionally, individual employees will have better morale if they feel their personal ideas and proposals are given attention and taken into account.
Improving Morale Through Quality of Life
Morale is another issue that many of my interviewees brought up. As I noted above, part of this has to do with a general feeling of pessimism in the country. But even though it may be pervasive and difficult to address, it is not a problem that can be ignored, either. Poor morale is often a self-defeating prophecy: if team members believe the situation is hopeless, then they won't try to work for a good result, and the end product will display this attitude.
Unfortunately, it is very easy to confuse "high morale" for "blind fanaticism". Many game development teams, not just in Japan but across the world, make a point of encouraging and accepting only the most extreme forms of dedication. They wear their unhealthy work conditions as a badge of honor instead of trying to do anything to fix them.
To prospective new employees, they display the attitude that "you have to want to work at our company badly enough to put up with the environment that we have here". In other words, instead of trying to create a company that welcomes all kinds of creative professionals, the grueling schedules are seen as a kind of test of loyalty to the cause.
In Japan, this combines with a corporate culture that is already well-known for placing a strong emphasis on overtime hours and having a high-stress environment. Combined with the fanaticism I mentioned above, a situation can arise where only the most obsessively devoted video game enthusiasts are the ones who are creating them.
While this may sound desirable initially, in the long term it leads to an unsustainable industry. Game creators tend to make games that appeal to themselves, so if this is the only type of employee hired by the company, the company's products will have a hard time breaking out from the world of the people who made it. Instead, they will extract ever-smaller returns from a dwindling customer base.
(A comparable situation might be found by looking at the current state of the Japanese animation industry. Conditions such as extremely low pay combined with overwork are driving talent away from the business. The well-documented difficulty of working as an animator in Japan means that skilled artists often leave to look for jobs in other fields. This in turn reduces the capacity of the industry to create truly unique new works. That diminishes the potential audience of such works, and that, finally, squeezes the industry even further.)
In some quarters, the downturn in the availability of fanatically dedicated workers has been explained by saying that Japan's younger generation is less motivated and doesn't care much about achieving success. To look at it another way, though, one could also say that it is the current prevailing style of working conditions that must change, because they do not serve the personal fulfillment and happiness of tomorrow's creative contributors.
In this case, decent wages and working hours that accommodate a healthy lifestyle would be an important part of making game development a desirable career choice for the younger generation. And in the longer term, taking care of workers to ensure they do not become disillusioned with the industry and quit should be an important consideration, especially as Japan's population ages.
Working conditions that accommodate more relaxation can also bring unexpected benefits. For example, some experts claim that the most creative ideas and solutions come to people during "downtime" when they are not physically or mentally at work but doing something completely unrelated, such as taking a walk or cooking. Therefore it is important to enable game developers to have hobbies, enjoy time away from the computer screen, and otherwise lead "normal" lives.
Page 3 of 4