• DigitalFlux
• June 15, 2010    7:56 PM

So, I was sitting in a somewhat small room in Vegas attending the IMGDC conference when a developer for a AAA MMO stood up and talked a bit to his fellow audience members about the challenges facing Independent developers looking to create MMOs. The problem, he said, is that content is expensive, and an MMO needs lots of content. Also, he added, “don’t fight us [AAA companies] on content, because you’ll lose”.

I disagree.

To clarify, content is really the hardest part of all this development. We’ve had features in for months, and the conversation system works, and has been getting revisions and improvements for a while now. At the risk of invoking Murphy’s Law here, I’ll state for the record that features are not the hard part in making an MMO! They just aren’t. In game development circles, the project is considered daunting due to the scale of the game once completed, but honestly, the features aren’t too much different from what you could find in a single-player RPG (ahem, the network code and architecture can be a bit rough though).

But take a walk through a city in your favorite MMO and take a good, hard, look around: Buildings, plants, NPCs, signs, tables, chairs, props of all kinds and the textures that love them- they’re all around in abundance. And then if you walk out into the woods- rocks, trees, plants, animals, streams, lakes, etc. That’s a lot of stuff, and it all had to get created by hand. Now, if you have any experience with Photoshop or even MS Paint and have created bitmaps from scratch before and know how much time it can take to create something of quality, take that experience and apply it to the gameworld. That’s a lot of work.

In terms of lightening the content load for a team, there’s just not a whole lot that can be done with that. The bigger your art budget, the more details you can cram in there. The smaller it is, the more you want to lean towards a stylistic approach (as we have with Epic Frontiers). We’ve skipped some polygons and gone with a style that we feel conveys the attitude of the game while also allowing us a realistic chance of creating content for it. But still, there’s a lot more that goes into the game than artwork.

Consider dialog, for example: Quests in any game require reams of paper worth of text. Our game wouldn’t be any different- actually, it is worse, with our NPC conversation system, we’re putting in dialog to cover fairly broad-ranging conversations by NPCs (in the beginning, we considered using canned dialog for the players too, but that was just way to unruly to be feasible, much less fun). In order to keep this under control, we’ve implemented some “generic dialog” which is similar to the generic dialog in real life. Sometimes, things just don’t need to be typed into the system twenty times by us. It gets tedious.

Some of the stuff I do on the other side of the DigitalFlux Entertainment house is tools creation. Specifically, I create a lot of tools that deal with generating procedural content. It’s a fancy name, but basically it means semi-random content. You set parameters and rules for the tool, and then it goes about either producing something that is random within the range of options given, or churning out every possible combination of those settings.

So, say you need swords: You know- “rusty copper sword”, “shiny copper sword”, “gleaming copper sword”, etc. But in addition to the varying names, you also need to vary the stats on the swords, which means that you can easily be entering a thousand variations of the sword into the database. Who the hell wants that job?? Not me- and I usually get stuck with it, being lead developer and all, so instead of spending days on that, I spend hours on tools to address the issue.

I can go into a lot of it technically, but that’s boring, so here’s the meat of the blog:

• Missions: I have to say that I admire Anarchy Online for pulling off mission generation like they did. What we’ve done is taken that two levels higher and not only generate different missions that don’t all have to do with combat, but we also generate the dialog for the NPCs involved in that mission. We also generate those NPCs, as a point of fact- and everything it needs. It’s personality-based AI settings, it’s conversational, combat, and skill abilities, its inventory, its name (we can generate 24,000,000 unique Charkritian names for this purpose), and even its backstory. Then we drop it into the gameworld for you to do what you will with it (probably kill it, but do us a favor and chat him up first- makes us feel better about the work done on backstory generation).
• Conversation: With the need to generate missions dynamically, we sort of had a problem- our NPCs talked. So we solved the issue by generating dialog for missions as well, and you’ll be able to tell when you talk to NPCs about the mission that they want to give you- it’s generated on the fly so that mission assignment is done by talking to NPCs and answering Yes when they ask you if you want to do something for them. Other “procedural” dialog includes dialog where the NPC informs the player about certain AI states (“I hate you, <player>” is probably a good example of what that looks like), and “templates” of dialog that relate to missions. It’s like giving an actor their lines while they’re on stage, right through their brain…
• AI: One of the good things about the AI system we’ve developed is that it lends itself very well to procedural generation of NPCs. What we’re doing is developing modifiers to the attitudes of the people of a locale, region, or even a faction, but nudging various traits just slightly. The same thing happens when we account for gender, age, profession, and even backstory (you didn’t think Johnny Orc grew up the same way your paladin did, did you?). The result should be a varied and deep representation of the game world through NPCs who all see the world just slightly differently.
• Instances: You’re going to have fun with these. While we’re on the content generation bandwagon, why not use it to supply the player with endless raid content? Generated and instanced “dungeons” (they’re not dungeons, but that’s neither here nor there) keep those who enjoy the grind of combat thrilled with areas created solely for them, and keep them challenged. And not all instances have to take place underground or indoors- but that’s something for a later blog…
• Points of Interest: Another method for providing content are Dynamic Points of Interest (POIs). POIs are areas in the game world that call out to the gamer (or maybe scream to the gamer in a blood curdling way) to resolve a particular issue. The POI is not always there, like running into a strange traveler on a lonely road who sends you careening on an epic quest that you had no expectation of going on (the best kind!). Of course, that stranger isn’t always going to be on that road, or be that stranger, or send you on that quest. And sometimes, the event you witness is something strange, rare, and worthy of some story telling of your own…

As development continues, we’ll be exploring all of these kinds of dynamic content in more depth, so keep a sharp eye out!

• Tags: