Agency in Game Development


agency (ˈeɪdʒənsɪ)
— n , pl -cies
1. Action, power, or operation: the agency of fate
2. Intercession or mediation

It is probably safe to say that the vast majority of people go into the business of making video games not because it offers the best avenue to making money, but rather because they believe it offers a great avenue to exercising creative expression in front of a very large audience.

Also, it’s cool.

The reality, of course, is much more complicated. As with any large enterprise, decisions are made at differing levels of the development hierarchy depending upon the magnitude of the decision, the seniority of the staff, and the perceived impact of said decisions on the business side of the operation. Decisions which are thought to impact the business side will tend to be reserved for fairly high up; development teams with a higher proportion of senior developers will tend to be more comfortable with decisions being made at a relatively lower level.

Intruding on all of this are issues of employee retention, morale, and the enthusiasm that occur when a development team is staffed with developers who are self-driven to make the game they are working on the best that it can possibly be.

All of which brings us to one of the central challenges of game development: agency, or the granting of maximum practical authority and responsibility to as low a level as is beneficial.

When a development studio has imparted a high degree of agency at a relatively low level of the development hierarchy, developers feel empowered and part of the product; they willingly work both hard and smart, because they identify with both the product and the company. If the product does well, they feel to be fundamentally part of that success. The benefits of this, thus, should be obvious.

The flip side of this, however, is to understand why agency isn’t generally granted at the lower levels of the development hierarchy. The answer to this is simple: trust.

With vast amounts of money it takes to bring a video game to completion at stake, those who are responsible for this money want as much certainty in their investment as can be had. Even more viscerally, it may not be a question of ROI (return on investment), but simply a question of survival – many development studios are literally a failed game away from the entire studio being dissolved. These two reasons, in fact, are exactly why sequels and retreads are as popular as they are. Such sequels and retreads may be very unlikely to be blockbusters, but they are more likely to achieve a basic level of return and economic security for the studio.

What engenders trust?

  1. Reputation. Sid Meier of Civilization fame can pretty much write his own ticket, because he has a proven track record of putting out blockbusters. On a less grand level, a developer who has worked on a recognized, successful game is going to tend to be granted more agency than otherwise. Worked on EverQuest? Here, take over the development of this system.
  2. Related to reputation is experience. Game development is a very specific discipline, with specific offshoots each with their own lessons to be learned in mobile, MMORPG, RTS, RPG, console, platformer and FPS arenas, not to mention many others. Anyone who has been in the industry for more than a couple of years can relate stories of the new guy who came in with grand ideas that won’t actually work because of fundamental technical, marketing, or process limitations that are very hard to understand for someone not in the trenches of game development where the sausage is actually being made.
  3. Confidence. In much the same way as people who radiate confidence do better in avenues as diverse as dating and politics, confidence engenders trust in a fundamental manner that is hard to measure but nevertheless undeniable. While this is most evident at the higher levels of the development hierarchy – we all know the stories of the CEO or CFO who confidently assured the investors of aspects of the product for which such confidence wasn’t really warranted, this happens up and down the entire development hierarchy. Even the intern or QA guy who confidently asserts that they can do such-and-such job that is normally outside their area of responsibility is much more likely to be given a chance than the guy who is hesitant about this. Sadly, of course, the line between warranted confidence and unwarranted confidence is a fuzzy one, and knowing which is which is often impossible unless you know the track record of the people involved.
  4. A final one is security, specifically the feeling of security and confidence enjoyed by the executives and managers of a development studio. A studio that feels secure economically is much more willing to take bold gambles and interesting risks, and is more comfortable delegating agency to less senior members of the staff.

There is, as well, another legitimate flip side to the application of agency: vision.

If there is one failing that has sunk more development efforts than any other, I would argue strongly for it being the vision for the game. A fragmented, inconsistent or unrealistic vision for a project means wasted cycles, wasted money, frustrated employees and overdue milestones. Vision can be amplified all along the development hierarchy, but ultimately it must be owned by a very small group of people, often a single person. The degree to which the vision of a project can be communicated and consistently enforced is critical to the success of the game. Sure, it can and will inevitably evolve, but it should do such cautiously if one wishes to avoid disruption of the process.

This, then I would argue is the fundamental question and the keystone that a studio must achieve balance on: Agency versus Vision.

  • Establish a consistent, clearly communicated vision that is generous in the “what should this feel like” department and vague in the specific implementation department. Make sure that up and down the development hierarchy second and third tier managers and supervisors understand, internalize and accept this vision such that they can be counted on to accurately amplify and reinforce the central vision.
  • Promote agency on the implementation of the vision to as low a level as is feasible. No, you don’t give responsibility for the architecture of a fundamental game system to the guy who’s been in the industry six months, but you can almost certainly give such – with appropriate review, obviously – to the designer who has been in the industry a couple of years. Give people a chance in a controlled environment with small enough sections of authority that they are set up to succeed rather than to fail, and give them more agency as they are able to establish a reputation for success.

In all of the above, note that I haven’t given specific examples from specific projects – this is a deliberate decision on my part. The truth is, no team I have ever been on has gotten this balance exactly right in my estimation, and even different projects at the same company can have vastly different strengths and weaknesses. Some of the thinking on this comes from my own direct experiences over the years, but just as much comes from the experiences of friends and colleagues in the industry.

At the end of the day, those of us in the video game industry are interested in making games we can feel as passionate about making as the consumer is passionate about playing, and in the process enjoying the process.

To re-use an expression used earlier, “knowing how the sausage is made” can be illuminating; understanding why agency is often jealousy guarded and why visions become fragmented is incredibly useful to figuring out ways to improve both. Malice or malevolence are actually quite rare as true motivations in the industry; more often the culprit for development frustrations is simply miscommunication, scarce resources, and hesitation in the face of the price of failure for a project.

Comic by Cameron Davis at Funny Web Comic.
Definition by Collins English Dictionary: Complete & Unabridged 10th Edition.