Frayed Knights - How I Build A Dungeon
It's been a little while since I have posted an honest-to-goodness update on Frayed Knights - the comedy-based indie RPG from Rampant Games. So I guess we're due.

Today, I'm going to talk about how I build a dungeon. I do not pretend that this is the correct - or even a competent - way to do it. It's just how I do it. Members of the development team who are also working on dungeons for the game (Kevin and Brian) may do stuff differently - and much better than me, I should add.

But it's still me who ends up populating and scripting them.

First of all - my design document (when I update it) doesn't have much by way of specific information about each dungeon (or "adventuring area", if it's an outdoor location). Really, I don't have much more than some basic themes and objectives for the location. Maybe some specific events that need to happen here (like, "a prisoner has a hint as to the identity of Moonshadow.") And then some key "signature" locations that serve those encounters.

For the guys working on levels for me, I have to provide a bit more up-front work. Or a lot more. But for myself, I start out without much more than just that. From there, I sketch up a very rough map. And start planning encounters / events. More on planning encounters later.

From there, I build a walkable floor plan. I've recently gotten to the point where I just build the floor plan first and export the entire area to the game engine - floating in the air in a test area. And I use it for a walk-through as the level evolves. This helps me get a sense of scale, what's working, and what doesn't. I like to use vertical elements in my levels, and helps to find out early that I don't have enough headroom in one part of the dungeon and I need to kick things out a bit. These days, I'm doing all my interior-building with Torque Constructor. It's ugly and old-school, but it works well for this engine.

After I'm satisfied with the floor plan, I build the rest of the location around it. This can be pretty time consuming, especially with making sure textures line up correctly, adding some detailing (I'm saving the really careful detailing and "prettyfication" for the end). Changes do take place at this stage (really, at any changes), and large swaths of my construction may be changed, deleted, or expanded. The floors may get shifted and broken into smaller pieces as I try to vary the geometry to make things look more interesting, etc. In fact, the whole process is very iterative - at any stage I may get new ideas that I want to try, I may go back and change something to make an encounter work better, or I may get a new idea for an encounter / event by working on a particular spot. It's not exactly free-form, but it's not a strict assembly line (or "waterfall model," for programmers) either.

Now, the big trick here is planning encounters / events. This is also very iterative. Some get planned out (in rough detail) before anything else starts. The dungeon design may center around them. Others don't get planned out until the very end. I find myself with a relatively boring area without anything else happening, and I think, "What would be cool to have happen here?"

I thought this would be a lot easier to create these encounters than it is, given the amount of time I've spent planning pen & paper adventures or making Neverwinter Nights modules, scripting MUDs, and so forth. But it gets very, very easy to run into a game designer's version of "writer's block." Maybe it's all the pressure of making something intended for a much wider audience than my usual circle of friends. I don't know.

What I do know is that I don't just want endless encounters of monsters that just hang out in their rooms waiting to fight adventurers who happen to kick the door open in a straightforward brute-force slugfest. Sure, we've got that too - but I don't want that to be the meat of the game. I want puzzles. Tactical challenges. Memorable encounters. Unfortunately, that requires me to "be creative" - not to mention spending a lot of time scripting / programming.

What I recently discovered was that one of the things causing the block was that I kept thinking in terms of the existing code and capabilities, and what I could do with it. My real breakthroughs came when I just thought through the dungeons in a lot more free-form fashion, and then later started figuring out how I might implement it with existing (or new) code. Sometimes, by attempting to frame the idea in terms of existing limitations, I find myself with an even better idea.

The problem is that these "out-of-the-box" ideas often require a lot more scripting, code, and / or content support. But I think the final results will be worthwhile.

Now, to answer a couple of questions I've been been asked. Answers subject to change without notice.

#1 - When's the next "alpha test" of Frayed Knights?
The pilot was an early experiment to help gauge where we were going with the game early on and make course corrections where things just weren't working out. Right now, we're going full-out to make the game fully playable, but not necessarily pretty or ready for human consumption. I don't want to take the time out (yet) to make a new playable demo. So probably not until we're literally alpha and switching gears from pure building.

The next one may not be quite as "open" the first one was. But I will desperately need testers.

#2 - When do you expect the game to ship?
I'm hoping 2010. But I was originally hoping 2009, so apparently my ability to estimate this project is for crap. I keep expecting to hit this magical zone where development accelerates speed because it all becomes "old hat" to us, but that's not happened.

#3 - How Big Is This Game?
Too Big.

In an interview at RPGWatch, I said I expected the full game to be "six to ten times larger than the pilot." As it turns out, that's just the first act. Of three. Feature Creep Strikes Again.

But dang it, I do love this game. Warts and all.

#4 - Have you considered breaking the full game into three parts and selling them individually?
Um... Considered, yeah. Made a decision? No, not yet. It wasn't in the original plan, so I haven't worked out how I'd try to break it up that way yet. But if we do go that route, I'd want to be a lot closer to completion, so there wouldn't be a long wait between installments.

If you have particular opinions on that subject, let me know.

Labels: Frayed Knights, Game Design

- posted by The Rampant Coyote @ 7:00 AM (Permalink)