When it's Done!
By Rowan "Sumaleth" Crawford
nother option is to actually set release dates, but once that date is reached with the game still far from complete, you simply set another date. There's no point singling out any companies/games, but I'm sure we can all think of half a dozen without having to loose focus from that player coming at you with a railgun. This is a risky way to make games, and while no one intentionally sets a date they don't intend to keep (usually), this practice probably makes investors more than a little nervous. With so many games failing to recoup the cost of development these days, the added costs of a game missing it's deadline, along with the greater risk of it being out of date when it's released, means that missed deadlines represent a serious risk to developers.
The other option, which is going to seem pretty radical in this day and age of the Internet and all it's wondrous possibilities, is to keep games quiet until it's really close to release. While it does avoid having to say "when it's done" or setting a date which is later missed, it doesn't get around the problem of games these days being too complex to judge development times.
With the Internet gaming scene expanding exponentially, and gaming magazines often spending more space on previews rather than reviews, the days of silent production are probably mostly gone. Additionally, pre-release hyping of a game is virtually essential today if you want to sell enough copies. So the challenge ahead of the gaming industry now is to find a way to merge all these competing elements - artistic integrity, financing, publicity, musical chairs etc - into a system that works for everyone.
One such solution may well be based on the movie industry solution. The gaming industry has been becoming more and more like the movie industry over the years, but there is still one big difference; actors are hired on a movie by movie basis, whereas game developers usually aren't.
Perhaps this is the way the gaming industry will go, with artists, animators and coders all hired individually to work on a game. It'll be different, not better or worse, just different. It may work well, and a quick look at the lists of people changing projects already suggests that hired developers may be more comfortable working in this fashion. Game creation would never change fully towards this method - there will always be core teams - but I suspect that we'll see this practice becoming more and more frequent in the coming years.
Another approach may be a system similar to the one used by the advertising industry. With advertising, the advertising agency plan out an advert, which may include things like live action, actors, special effects etc, and once they have that planned to a reasonable level of detail (and the client are happy with it), they approach a number of production studios to make a pitch for the job.
The point here is that the finish dates are pre-set, with on-air time booked well in advance. The production studios must take this into account, working out how many people it will take to produce the commercial at the quality required and in the time given. From those details they work out a cost and that is their pitch to the agency. Often their pitch will re-evaluate the details of the design to make things cheaper, or quicker to produce, for example, using computer generated effects instead of traditionally (hand draw) effects. The agency collects up these pitches and decides which pitch has the best balance of speed, cost and results that they wish to go with.
If you imagine this approach applied to the gaming scene, you might have a game company approaching a number of 3D animators with a breakdown of the models needed and the time allotted, and the animators pitch for the project. The cost of missed deadlines are then placed squarely on the shoulders of the animator which breaks the game creation process down into many smaller, and far less complex, individual systems which, when the output from each system is combined, results in a game.
I don't seriously believe that this approach will be fully adopted by the gaming industry, but that would be a plausible solution to the problem. Advertisement deadlines are rarely missed, but at the expense of fun and creativity. The gaming industry thrives on these two factors, so a change like this would almost certainly be for the worse. However, the "milestone" approach to development, mentioned below, actually boroughs a number of ideas from this system.
There is, of course, another obvious way for developers to set a deadline and then meet that deadline, and that is to simply stop production at that time and release it. Christmas is a deadline that a lot of developers and publishers feel compelled to meet, and with good reason too; Christmas is a time when presents need to be bought, and what better than a well priced computer game? With games being such complex beasts, as previously mentioned, stopping production and tying up all the loose ends into a neat bundle quite often proves to be just as difficult as meeting the deadline in the first place, and invariably leads to a product which is both messy and not reaching it's full (and originally intended) potential.
A messy release is no good for anyone, but with the Internet slowly easing it's way into our everyday life, a slightly "truncated" game isn't quite the problem it once was. So long as the original release covers just enough bases to keep reviewers and gamers generally satisfied, the practice of adding new features and maps later may well prove to be a feasible, if not perfect, approach.
Recent examples spring to mind, particularly the splitting of Quake 2 into the main "single player" game which was released by the aforementioned Christmas deadline, with the fine-tuned multiplayer elements released a few months later. This practice was met with a lot of skepticism, yet in the end it wasn't such a bad arrangement; how many players would have preferred to wait 2-3 months for the deathmatch side to be finished just so they could play the single player game? Not me, I really enjoyed Christmas that year.
In recent times, a new approach to development has appeared which aims to put more responsibility on the developers in a similar way to the advertising industries production system mentioned above. The "milestone" system, in use by many companies right now, divides the project up into a number of smaller milestones, with the developers under contractual commitment to have the required results at the end of each period. The price of missing a deadline is a loss of royalties.
Most developers have always worked to similar internal milestones, the difference is that with these now being set in conjunction with the publishers, they are far more concrete than they have been in the past. Developers really are putting in the heavy work required to reach these milestones.
A number of interesting effects appear from this approach. Firstly, developers are far more likely to reach these deadlines than they are from a single "final" deadline, due both to the financial commitment and the subsequent confidence boost gained in reaching these deadlines. As with most industries, it's the people who do the most work in the gaming industry that earn the least amount of money, so it's a bit worrying to see even more risk placed onto the shoulders of the people who already put in the most hours. Game development isn't a 9to5 job; 10+ hour work days are common place, and we've all heard the stories of developers working 24 hours a day, sleeping at work, and the like, just to reach these deadlines.
Time will tell whether developers continue to be comfortable with this approach, but with the acquisition of publishing deals so difficult at the moment, the leverage is definitely in favor of the publishers.
Worth noting is that some companies working by the milestone approach, meaning that they do have intended deadlines, are still reluctant to give out release dates, preferring to instead use vague phrases such as "in the fall" or "3rd quarter", and even "when it's done". So even with this breaking down of the development process into bitable chunks, developers are still finding it difficult to be completely sure of deadlines. Also interesting is even with this system being used, deadlines are still being missed.
The majority of the difficulty is based around the phenomenon of that other catch phrase, "two weeks!". It's a hard process to write a game, but the hardest part, by far, is finishing it up. The end always, sincerely, appears to be no more than two weeks of work away. It's impossible to predict the unexpected.
Game developers are in it for the creativity, publishers for the money, and players for the entertainment, so there are always going to be clashes between what everyone wants. It's only natural that everyone priorities are dependant on which category they fit into, and it's probably optimistic to think satisfying all parties, all of the time, is a simple task.
If history is anything to go by, you can be sure that developers and publishers will eventually find some even ground. A system will be found where making a game isn't as risky for the publishers/developers as it is now, nor as constraining for the developers, although there are always going to be developers keen to work by the "when it's done" principle. It's their choice.
The trick is simply to accept that it is the way it is.
- Rowan "Sumaleth" Crawford is a member of Team Impact. His most recent project has been the upcoming Quake 2 Internet Pack: Extremities.
|Credits: Illustration © 1998 Mike Sanzone. When It's Done is © 1998 Rowan Crawford. All other content is © 1998 loonyboi productions. Unauthorized reproduction is strictly prohibited, dangit.|