Postcard from GDC 2005: Fable: Lessons Learned

Author: by Daniel Sánchez-Crespo Dalmau

Peter Molyneux appeared for the second time in this year's Game Developers Conference for a public postmortem of one of his latest creations, the game Fable. The session ranged from "making of" style comments to pure post-mortem material in the form of the rights and wrongs of a project that took over five years, several major redesigns, and some extreme resizing, to complete.

In the beginning, the idea for Fable came at a time when Molyneux was a big fan of the RPG Wizardry. The first vision of his creature was quite simple an RPG designed to be played on PCs with the subject of battles between mages, and both a single and multiplayer modes. The year was 1998, and the allocated team was four people. The game found no publisher at this stage, so the design was stored for some time.

Approximately a year later, the first major refinement of the design took place. The focus of the game became the morphing of an avatar who wants to become a hero and can choose between the good and evil paths. This morphing should affect both the character, but also the world he lives in and the perception of those who surround the player, so their reactions reflect the player's alignment in the story arc. This core design, that remained unchanged for the rest of the project, was christened Project Ego in the year 2000, when team size expanded to 15 people. The subject of moral choice began to surface, as did the concept of the hero's age and the world aging also as the story advances.

It was in these days, with the core elements in place, that the game began to grow, as the design team filled their seminal vision with more and more elements, enriching both the environments and the actions to be carried out within them. This expansive phase reached its peak at about 2002, when a team of 23 had a huge vision and the daunting task of mapping it to a series of prototypes, so the game could begin its life in trade shows and media appearances.

The turning point in the Fable story came in 2003, with a team of 28. At this stage, internal scheduling took place, only to discover the game would be completed by the year 2103 due to its excessive size and ambition. The team called for help, while deciding the final name of the game, Fable. The help came from Microsoft, who would end up publishing the game in the Xbox platform. Under pressure to have a first playable at E3 2004, the game began shrinking, mainly in content but not so much in design or features. Team expanded inversely, reaching a size of 50, but still missing the E3 2004 milestone. As a result, the team decided to redefine and completely resize the project. A crisis team was established with members from both the publisher and developer, to make sure a game similar in ambition but not in scope to Fable actually reached the shelves in a reasonable amount of time. Both Microsoft and Lionhead believed in Fable, and as such they began to work together around the clock, setting clear goals and sticking to them.

It was decided to keep all core elements, which were combat, villager AI, scripts and story, accessibility, ease of use, music and cinematics. All other items were reexamined, scaled down, so the first playable version of Fable was finally completed December 19, 2003.

Molyneux recognized Fable dropped many elements in this sometimes painful process. For example, he openly stated that the final version of Fable is about 15 times smaller than the game world that was originally planned, including the omission of one full continent and all its story and gameplay elements. Some other interesting features, such as a multiplayer mode, followed a similar fate. In the original Fable, up to four players could cooperate in a story mode. One gamer would be the hero, and the other three would be alter egos or henchmen. Whenever a henchmen won 100 points (through quests and missions), the player lost its status, and the henchmen would become the hero. Although the team had prototyped this and had a working version, the multiplayer mode was totally scrapped much to Molyneux's disappointment. Ironically, some features crept in, such as Xbox Live Support. Still, the game remained the same at its core, including the morphing hero, living world and accessibility. The way Molyneux likes to put it, "a blind child with no hands should be able to beat the first set of levels".

The rest, as they say, is history. A much smaller version of Fable was finally released and, despite sales success, received with mixed emotions among players. While some thought it was a great game, others considered the game was seriously decimated along the process due to its excessive initial ambition. Luckily for Fable fans, a PC version is in the works and will be released later this year, that includes some of the content that did not make it to the retail Xbox version. Still, it is interesting to see how even the biggest industry luminaries such as Molyneux can sometimes overshoot and be ambitious beyond what's imaginable or, what's more important, feasible within reasonable time and budget constraints. Facing these issues, Molyneux responds that if you aim for the sky and end up constructing just a piece of it, it will still be an awesome piece of entertainment. However, if you are too realistic and un-ambitious from day one, you will end up with a mediocre product that does poorly in the market. Although partly true, the sort of budgets in use in the industry today force us as developers to be rigorous. And we urgently need to develop approaches that do not stifle creativity while allowing well-rounded, perfectly finished products to ship on time. Especially in an era of 8-digit budgets and hardware that can make designers dream of the impossible or, even worse, make them believe it's actually possible.

______________________________________________________