What Went Wrong
1. Planning the story. We underestimated how long it would take to construct and write the story element of Black & White. The free-form nature of the game required an unfolding tale to give it some structure and lead it to a conclusion, and in October 1999 we began to work on the story. We thought it would take no more than two months, but after a while we realized that we didn't have the skill set needed to take care of this vital aspect of the game. I contacted James Leach, who'd been the in-house games scriptwriter at Bullfrog and had worked on Syndicate Wars, Dungeon Keeper, Theme Hospital, and many others. He was working as a freelance ad copywriter but gladly came on board, again in a freelance capacity, and turned our ideas into a fully plotted story line, wrote hundreds of challenges and quests, and wrote all the dialogue in the game. It ended up being more than 60,000 words, the size of a novel.
Hiring James meant that we got a sense of continuity, consistency, and style throughout the game. It also meant that we could describe what we wanted, or even write placeholder text, and he would rapidly turn it into finished work. Sections of the game that were still at an early stage seemed more easy to understand, get a feel for, and work on when we used dialogue and text which seemed, to us, finished. Of course, another pass was usually needed to make it accurate and sometimes to polish it, but having a dedicated scriptwriter made this a simple task.
Storytelling in games, as elsewhere, is an art. If a story line flows easily and naturally, that's because someone has worked incredibly hard at it. I'm a great believer in the emotion and immersion that can be added to a game through good story and dialogue. It can't make a bad game good, but it can make any game better. And when the script was looked at by Hollywood scriptwriters and film directors from the BBC, we knew we were on to a winner.
Another by-product of using a professional scriptwriter was that we morphed the in-game advisors, the good and evil guys, from being just sources of information and guidance into stylish, popular characters who are now bankable properties in their own right.
2. Fixing the bugs. After canceling our Christmas party on December 26, 2000, we managed to hit Alpha, which as any developer knows is a very loose definition, but at least we could say that all the game features were now locked. After a well-deserved Christmas break, we came back to find that we had more than 3,000 bugs. We had six weeks to reduce this to zero, but the thing about bug-fixing is that you can solve one problem but in doing so create three more. So although we worked as hard as we could, the overall figure crept down slowly rather than dropped at the rate at which we were actually sorting out the bugs.
We tried to make the micromanagement of the villagers as user-friendly as possible.
By this stage the team was very tired. The only things that kept them going were the sense that the end was in sight and the fact that they could now play the game and actually experience what we had created. Bugs, of course, could have killed the game, so there was no way around it but to fix each and every one. We had bug lists circulated to every member of the staff, and we put up a chart on the wall which was updated daily. Some days we had more bugs than the day before, and that was like looking at a mountain which was growing quicker than we could climb it. But there came a moment three weeks into this process when we felt we'd broken the back of the major bugs, and the numbers fell steadily. Of course, the irony was that the last 10 bugs were the hardest to fix, and with every one there were four more created. It was as if the game just didn't want to be finished and perfected.
3. The project was too big. Black & White got to be so large that we almost felt lost within the code. In fact there are well over a million lines of code within the game. Loading up even the most simple of the smallest tools would take many minutes, and compiling the entire game took over an hour. This meant that toward the end of the development phase even a tiny change could take a whole day to implement.
Checking in changes and rectifying errors was a nightmare. We eventually decided to limit the checking-in to one machine, and we implemented a buddy system whereby nothing was done without an onlooker checking it at every stage. This put a stop to tired people checking in changes at four in the morning and finding that, instead of fixing something, they'd actually caused further problems.
Another worry about the project's size was that we didn't think the game would fit on one CD, although we were desperate for it to do so. The audio files are immense. Music, dialogue, and effects are all compressed, but of sufficiently high quality that we refused to reduce them any further.
And with 15 language versions to get translated and recorded, we had to do the biggest localization job I've ever seen. This landed on Lionhead Studios at the very busiest time, and although our publisher did an excellent job of handling it, we were needed to check and answer questions and to provide explanations for some of the more arcane elements of the game.
4. Leaving things out. The idea of the game didn't really change much over the course of its creation. But I do have some regrets that features we thought would be great proved unworkable. I expected this, as it happens with every project, but I thought the problems would be caused by software or even hardware limitations. In fact, it came down more to emotional issues.
For example, the original idea of the Creatures was that a player could choose to make any living thing a Creature. We wanted the player to be able to select an ant and grow that, or a human being from a tribe, and raise him or her. Christian Bravery, one of the artists, spent a long time drawing concept work and sketches depicting what the Creatures could look like at various stages of their development. This of course included humans.
We soon realized that people would have certain expectations from a human. Players wouldn't expect a turtle to learn as quickly as a man, but if we dumbed down the people, they'd seem like a proto-hominid race from eons ago, and we didn't want that. Also, discipline in the game involves slapping your Creature. We certainly couldn't have the player slapping a child or a woman or, really, even a grown man. The emotional feel of raising a human, teaching him or her to eat what you want, and leading him or her around in a speechless environment was all wrong.
The game's underlying detail is immense but never overwhelming.
Christian's work in visualizing humans as player Creatures was all for nothing in the end, and we dropped the idea. We also dropped the notion of turning any living thing into a trainable Creature, as ants, butterflies, fish, and other nonmammals would have caused big problems. A flying Creature would change Black & White into a totally different game.
I also regret that we couldn't use color as a dynamic concept a little more. The landscapes in the game are gorgeous, and our sound and music man, Russell Shaw, suggested that various spells could drain the color out of areas, or spread different colors around. We liked this idea for its surrealism, and we thought about having color wars with other wizards (at this stage you weren't a god, you were a wizard battling others on a land). The idea lost momentum when we thought about how the land would actually look, and how it would seem like something drawn by a preschooler. I still like the idea of color wars, but I think children's TV has also cottoned on to the idea, which means we won't be going there.
5. Talking about release dates. I have to admit, ruefully, that I have a reputation for being, shall we say, optimistic about when the projects I'm working on will be completed. I opened my big mouth and announced that Lionhead Studios would finish Black & White and get it released at the end of last year.
I just can't resist talking about whatever I'm currently working on. This has been a problem I've experienced with every game I've ever developed. But the thing is, when I think something is going to be finished in December, I really do believe it. People at Lionhead were telling me that we had to build in time for bug-fixing, and I knew this was true, but the truth is that there seems to be no formula for working out how long things will take. The best thing to do, I guess, is to take the finishing date I first think of and move it twice as far away—and then not announce it until we're halfway there.
It's a function of working on products which could literally be endless. Unlike a film, where once the footage is shot, you edit it with an idea of where you'll end up, you can add completely new features to a game and then balance it and change it radically right up until the last minute. I'm sure that there were many people who didn't believe me when I said we'd finished making Black & White and were only convinced when they saw a box with a CD in it.
Black & White is unlike any other game ever written. That was our goal, and we achieved it. We wanted something more beautiful, more complex, more emotive, more innovative, more clever, and more, well, just more.
As you've read, it was beset by problems. We nearly drove ourselves crazy solving them. Nothing worthwhile is ever simple, though, and for every minute spent thinking up wonderful ideas to include in the game, there were probably 20 hours of sheer hard effort trying to get them to work.
Lionhead we were perfectionists, but if we were, the game would never
have been finished. It's not a perfect game. Our next game won't be, either.
But because there's no such thing as a perfect game, we'll just try to
do something different, and do it as well as we possibly can. Someone
asked me recently what drove us to work so hard on this and to spend so
much time thinking outside the box. The simple truth of my answer only
struck me afterwards. With Black & White, we made the game
that we wanted to play.
Publisher: Electronic Arts
Full-Time Developers: 25
Budget: Approx. £4 million (approx. $5.7 million)
Length of Development: 3 years, 1 month, 10 days
Release Date: March 30, 2001
Platforms: Windows 95/98/2000/ME
Hardware Used: 800MHz Pentium IIIs with 256MB RAM, 30GB hard drives, and Nvidia GeForce graphics cards
Software Used: Microsoft Dev Studio, 3D Studio Max
Notable Technologies: Bink for video playback, Immersion touch sense for force-feedback mouse
Project Size: Approx. 2 million lines of code