Postmortem: ngmoco/Demiurge Studios' WordFu
August 5, 2009 Page 4 of 4
4. Give Me More, More, More
When ngmoco first approached us with the concept of WordFu, we thought a dice-based word game sounded simple enough...
What started out as a six-week jaunt, incrementally progressed into a four and a half-month expedition. Thanks to the ease of App updates, we're in our eighth month, and we're still cranking out new features and improvements.
We were consistently on-time and didn't miss any major deadlines, but the project expanded to be much larger than we expected. While this is a success in many respects, feature creep (the cause of these extensions) becomes a nightmare for project planning.
Every time we showed the game, either internally or to ngmoco, the reply would always be "That's great! Wouldn't it be cool to do this, too?" We got lots of great ideas to improve upon the game; no single one was very big or time-intensive, but they certainly began to add up.
We quickly fell into a feature creep loop, since each time we added a new features, it received positive feedback. It's easy to scale back when the game is visibly becoming overly complex, but breaking the loop is much more difficult when each time you add a new feature, both the developer and the publisher like the addition.
To compound matters, we didn't have a clear vision of when the project would be feature complete. Thus, it became very difficult to plan when or what type of developers would be needed - some features were more art-intensive, others required engineering. Staff always had to be on their toes to come on to or off the project on a weekly basis.
On top of staffing issues, the game itself started to become messily constructed. Features were engineered without long-term visibility, so they were implemented to be acceptably functional and not necessarily optimal. Then, more features were piled on top before an optimization pass could be taken. The system would have been more stable and ultimately faster to make if we had known the scope from the start.
In the end, the game came out better than we could have imagined, with lots of features and ongoing updates. However, it was an unexpected resource commitment, affecting the progress of other projects at the studio.
5. Too Many Cooks in the Kitchen
Feature creep and the lack of vision on project scope lead us into another problem. We had many artists rolling onto and off of the project as we needed them, and the consistency of the artistic vision was difficult to maintain.
Naturally, each artist has his or her own style -- a problem that seemed to be more noticeable on a title with fewer assets than the large-scale games we're used to developing. With four different artists having worked on the project, maintaining a consistent style was a challenge.
For example, when we added special power-up dice, a new artist was tasked with creating the effects and text. Text like "Double Score!" needed to be both exciting and interesting, while matching the existing art. Finding the right balance of uniformity and contrast was an iterative process made more difficult by the fact that we had new artists thrown into the mix.
The outcome of the art style ended up being something that "went right", but the process of getting there bogged down development on a title to which we were already over-committed.
We seem to thrive on journeying into unfamiliar endeavors, and this project was no exception. We learned a lot about what we'd do differently, as well as the same, in upcoming iPhone games. It was interesting to see how many design and process principles were still applicable to this platform.
We're entering a new generation of technology. Smart phone technology is becoming more and more ubiquitous. Soon, everyone will have a powerful computer in his or her pocket. It's going to change how people think about how they interact with information and with each other in a big way.
We only touched the surface of this with how WordFu interacts with social networking systems. As this platform and its market grows, the coming years will be very exciting, to say the least.
Developer: Demiurge Studios
Release Dates: February 26, 2009 (v 1.0), March 31 (v 1.1), April 21 (v 1.2), and June 29 (v 1.3)
Platform: iPhone and iPod touch
Development Time: 8.5 months
Number of Developers: 3 at any given time, 8 total
Longest possible French word in WordFu: constitutionnalisassions
Page 4 of 4