Jess Lebow's MMO Story Hour: Design By Bug

One of the cornerstones of game development is iteration. The old adage: if at first you don’t succeed, try, try again—this is more than our motto. It is our manifesto.

It has to be.

MMOs regularly take more than five years to develop. (Some even longer.) Creating a fully realized virtual world just takes a long time. Even if you have a big team, there is still so much to do, that it’s not uncommon to have companies with well over a hundred employees working on just that one game.

With that many people, it’s often difficult to get everyone moving in the same direction. Don’t get me wrong. Every company has a core group of creatives whose job it is to create and maintain the vision for the game. But if these people are doing their jobs right, and they are trying to make a game that isn’t just a copy of something else in the marketplace (the serial numbers neatly covered up with a new art style), then they are pushing to break new ground. They are looking for new and interesting ways to innovate, come up with fun interaction types that give players a new experience, or find new ways to enhance old mechanics.

To do that, you have to come up with a lot of theories—things you think will be fun but just won’t be sure of until you see them in action. But you’d be surprised how often a good idea on paper isn’t as much fun in reality as it sounds. And in MMOs you always have to be conscious of exploits. Even if something turns out to be fun, it might not make it into the game because it breaks the economy or some other major system.

When we encounter these sorts of roadblocks, we start over.

And over.

And over.

It’s inevitable. Some things are going to work. Some are not. And you just have to keep plugging away.

One of the ways we keep up our momentum is by generating and collecting ideas as a team. I’m talking the whole company. Everyone inside the building with a signed NDA chimes in. Because at the end of the day, everyone on the team has a vested interest in seeing the game succeed, and while each person on the team has a different role, we all have something in common—we all love to play games. We’ve all played a lot of games, and collectively we have a lot of ideas about what would make a game fun.

So we make a list, a huge list, (believe me when I say huge, I mean: titanic, gargantuan, enormous—really, really big), and we distill it down. We punt the things that we know won’t work, the good ideas that will cost too much or take too much time, and the stupid ones we just don’t like because they are, well, stupid. That leaves us with the ones we hope will work, but we still aren’t quite sure if they will be any fun. We decide which ones we have the time and money to experiment with, and we simply start putting them in the game.

One of the things that always ends up on the list is some variation of the escort mission. On paper it’s a great idea. It allows players to continue core game play (I.E. explore the world, kill monsters, take treasure, learn more of the lore, all the while gaining XP). But in reality, we all hate escort missions. Either the guy you have to escort runs off and aggros a bunch of mobs, or he’s too damn slow and can’t keep up with you, or he gets stuck on a prop somewhere and you have to go back and find him, or he’s always in your way and you just wish you could put your own warhammer through the back of his skull.

But it comes up, so inevitably, we put it in.

One time we put it in fairly early. It was like the second or third full-blown mission in the game. This time we’d chosen to let the NPC run ahead of the players, effectively making him a timer. If you can kill your way through the mobs before the escort gets himself killed, then you can clear the mission.

In he went, and we all followed, running the mission to see if we had created an encounter with that exquisitely delicate balance—if we have managed to walk the fine line that separates fun from frustration—or discover if we had in fact laid a giant digital turd.

About half way in the NPC we’re following comes to a stop, turns to face us, and shouts the word “window!” Okay, so he didn’t really shout. But a word bubble popped up over his head. And he might as well have shouted because none of us were expecting this to happen.

We knew we had word bubbles, but at that point we hadn’t figured out how to fire them off at pre-specified locations in the world without the player interacting with the NPC. If we had known how to do that, I certainly would have written something more interesting for him to say than “window.”

This was, of course, a bug. Well, it was actually two bugs. But what we learned from the experience was that escort NPCs who run up ahead of the players have the opportunity, if you design it right, to deliver interesting story information in a rather organic fashion. Since it doesn’t slow the action of the game and you’re not forced to read a huge block of text, it feels very natural for an NPC to tell you a little bit about himself or the world around him while you complete the quest.

Did we solve the problems inherent in escort quests? No. But we did find a way to make them at least a little more interesting—a result we hadn’t planned for or intended.

I guess it’s fair to say that there is another reason for lots of iteration, other than simply proving that a good idea on paper is also good in the game. Sometimes you find things you never thought of in the first place.