Icarus Studios News

Software development timelines are inherently hard to predict: some aspects of design have well-understood ratios of person-hours to functionality, while others can take anywhere between weeks of work and a moment of inspiration. For the former, assigning more personnel or using overtime crunch can allow you to meet a milestone on time no matter how far behind you are at the moment. For the latter, the typical tools of acceleration can just result in delaying the creative individuals that might have quickly found a solution if left alone to think.

A surprising amount of virtual world development fits into the latter category. An area just doesn’t feel right until a designer comes up with a brilliant new layout. An NPC patrol goes widely off course every few dozen times until a programmer finally realizes how five unrelated systems are interacting to produce the bug. Something about a particular creature model is sending viewers deep into the Uncanny Valley, and the art and animation teams have to work out how to fix it to make it the friendly helper it was intended to be. In many of these circumstances, the team is forced to make a hard decision: do the minimum to deal with the issue and accept the reduction in quality, or take the necessary steps to improve the game even if it requires many hours of extra work.

The old engineering aphorism of “good, cheap, or fast: pick two” is almost applicable, except that throwing more money at the problem probably won’t help: your choice is often just between good and fast. This creates a dangerous alchemy: the sooner the game is released, the sooner the investment in its development starts recouping costs—but the more likely it is to suffer poor reviews and word of mouth, losing money in the long term. If you’re lucky, your customers will complain about “paying to be in the beta,” but stick with you as the designers improve the experience. If you’re unlucky, an unfinished launch means poor retention and customers that will never return. And it’s almost impossible to predict which way the coin will fall: focus groups are notoriously unreliable, and most artists will never be completely satisfied with their creation enough to call it “done.”

In general, if the money will support it, it’s worthwhile to wait to launch until your designers will grudgingly admit that they can no longer make dramatic improvements with just a little more time. There are many cautionary tales of virtual worlds that launched too soon and failed to find and hold an audience. Meanwhile, fans bemoan delays in hotly anticipated properties but usually buffer their desire with an understanding that they will be more rewarded by a later launch. The most hotly anticipated games these days have had a series of delays, and the dominant MMO producer in the world is famous for not giving anything but the vaguest of predictions for when they’ll release new games and updates.

In the end, virtual worlds design is more like cooking than baking: unless you have hungry people demanding to eat right away, it may be worthwhile to leave it on the stove for a little while longer.

-Stephen Cheney, Virtual Worlds Team