What Went Wrong

1. Selection of Partners is Critical -- A Mixed Success

Dream Chronicles was unlike the eight other games that we had made at Kat Games. Where we previously had been able to handle all of the development tasks required to complete a project, we knew from our early brainstorming on Dream Chronicles that one of the major challenges we would face would be appropriately resourcing the project. It was simply more than our four person team would be able to manage.

As we hadn't previously needed to rely on external partners to create our games, we initiated conversations with outside teams once we had signed our agreement with PlayFirst. The most direct area of impact to the overall cost and schedule of our game was the art production, where we had a difficult time keeping the project on track from a quality, budget, and schedule perspective.

In the end, we "touched up" many of the scenes that were received from our external partner and that required more time than we had originally planned. This wasn't all bad, as we learned a lot about the process of working with external teams and have been able to use the lessons we learned in Dream Chronicles in subsequent projects.

Other areas of the project required third party partners as well and we again learned a great deal about the process of collaborating with multiple teams to create a unique and innovative type of product. All of the audio was produced in collaboration with a third party audio production company, a script writer was hired to author the story and text in the game, and a freelance Flash resource contributed to the cinematics.

All in all, we had a challenging but rewarding experience here, but it definitely is one of the areas that we will take into consideration as we work on future projects.

2. Iterative Design Process (part two)

This point is so prominent in any discussion of Dream Chronicles that it deserves billing in both the good and the bad of the postmortem discussion. While I appreciate the iterative design process that we experienced while making Dream Chronicles, it also presented some challenges that are worth noting.

Originally, we did a lot of experimentation because we could spend the time thinking through the game's story, user inputs, and puzzles. Eventually, however, production became strained because of the number of questions that were left unanswered during the up front design phase.

The dream jewels and game diary were not completely integrated into the gameplay until near alpha and work on the story, puzzles, and cinematics (especially for the later levels) was delayed more than we would have liked.

Fortunately, our team had the foresight to build a level editor for the game, which allowed us to make changes during the late stages of the project. Once that level editor was available, any number of issues could be tested and tuned as we finalized the design of specific elements in the game.

Another area in the iterative design where the project had some strain was the script writing and integration. Because English is not our team's native language, the decision to hire a script writer was an easy one that we made with PlayFirst early in the project.

However, we found ourselves in a somewhat clumsy cycle of version control around the script, where the test team was opening bugs during testing of the game before the revised text had been delivered by the writer and integrated into the game! This caused some amount of back and forth with the publisher that we could have avoided with a more streamlined process.

One last area of the design that was impacted with the ongoing iteration was the game's length. Because of the unique nature of an adventure game, it was difficult to quantify the amount of time that the end user was going to spend on any given scene and in turn on the entire game.

We experimented with different techniques (including the beta testing metrics implemented with PlayFirst's proprietary game engine) to try and make the best guess at the average amount of time consumers would spend on the game, but ultimately we wished that we could have increased the length of the game by some reasonable number of additional screens.

3. Technical Challenges

New to us for Dream Chronicles was the fact that we would be developing the game with PlayFirst's development platform Playground SDK. We saw several advantages in using the platform, most importantly that it allowed us to release a Mac version of the product, which we had not done with our previous games.

However, we hadn't planned up front to use a new technology to build Dream Chronicles and we had to take some time up front to learn how to us the framework. It took some getting used to, but in the end we really liked it. The documentation and support that we received from the Playground SDK team was very strong and all the requests we made were either fixed in an interim release of the technology or received attention through direct response from the engineers.

Next, because our vision was to create a new game genre showcasing a unique Art Nouveau style, we also all wanted to include some fresh cinematic sequences in Dream Chronicles. From the game design and art standpoint, these sequences allowed us to round out the story world that we were creating.

From the technical standpoint we utilized Flash as the method of delivery and ran into some issues around integration of the cinematic sequences with the game engine. In the end, with a little extra coordination around the settings being used by the creator of the cinematics, our implementation team, and knowledge about what the engine required, we were able to work through this issue. But, it was something to include on the list of things that we would improve on subsequent projects.