Good Dungeon Design Principles

[Jakegilla]

Hello everyone!

I'm new here, been working with the editor and lurking the forums since the tools went live earlier this month, and would first off like to thank you all for the (indirect) help in learning the editor.

I was hoping some of the more talented designers might be willing to share some design principles for building good dungeons in Grimrock. I'd be interested in hearing things on how to introduce mechanics, puzzle and enemy density in levels, and maybe even things on actual layouts, anything really.

Anyone have some wisdom they're willing to share?

Thanks again!

[Neikun]

People will like to see new things. Whether it's new models or new features, doesn't matter. If your dungeon features something they haven't experienced before, you will garner attention.

The number one rule of design in anything is "Does it make sense?"
For example, when applied to a dungeon,
On the starting location, is the a clue as to how you got in there, why you can't go back.

Do decorations lend to the concept of the room?
For instance lets say you have a corpse on the floor, you may want to have a note either on the corpse on in the room that has some words jotted down like the poor guy's thoughts just before he entered the room he met his death in.

People are looking for engaging things, and the best way to engage you audience is to give them several little bits that all fit together.

[Tyroie]

I wouldn't call myself experienced, but something I've been putting a lot of thought into is how difficult any given challenge should be, whether it's a fight or a puzzle. If everything is very easy, some players will just get bored - but others want that. But if everything is too hard, people will feel like it's too much and just give up - but some people won't be satisfied unless it's hard. So what do you do?

Tough stuff can be optional - let the good players feel like they're going the extra mile without holding back the others.

You can put clues in slightly later in the dungeon so people can then go back with a better idea about what to do.

Also I think people are more willing to do tough stuff later on in the dungeon, and that's worth taking advantage of I think.

There's probably a lot more to do with this...

[Decayer]

One of my principles is to never put anything required to progress to the next level in a secret area (the kind that makes a sound and makes a message pop up); they should only contain optional items, monsters, puzzles and/or similar content (and the occasional shortcut).

[msyblade]

I like to mix simple puzzles with no hints and difficult puzzles with hints, so the less determined player (that may just quit the first time they get stuck) feels like they have accomplished something complex. The downside is that the hardcore crowd (other modders, old school gamers) can feel like the answer was handed to them on a platter, and don't feel the intended sense of progression, cause they know the tricks. Finding the balance between these two extremes can be a solid foundation of design.

[Drachir]

Whatever you do, the most important thing IMHO is to get someone else to test it to destruction.

No matter what you do, how many times you run through, someone else will find a bug,find a particular puzzle unsolvable, a trap unavoidable etc.

The more you test it yourself, the more assumptions you make

[msyblade]

That is all so true. Its impossible to playtest your own mod because you know exactly what you intended and can pull it off seamlessly every time. That doesn't mean it works gameplay wise,as a player trying to solve it will try all sorts of different approaches in figuring it out (backtracking, triggering other switches, leaving to a different area for awhile,etc)...A beta version is usually necessary before release because You NEED people to play through it, and tell you how the strange approaches they attempted in different parts made your dungeon seem "broken" in some context. Of course, after getting the input you always feel like "Duh, surprised I didn't think of that"....until 15 or 20 of them come through and you understand that there are far too many possibilities to foresee while your head is "in the game", and the intention you have in mind is not the same as the players you are ultimately wanting to play a polished product.

[HaunterV]

I've learned with my dungeon that Trial and error is not a valid mechanic to expect from the player to solve door/torch puzzles.

If you play grimrock again and pay attention, all puzzles can be solved or have the solution presented to you via observation alone.

[spectre900]

Whatever you do, the most important thing IMHO is to get someone else to test it to destruction.

No matter what you do, how many times you run through, someone else will find a bug,find a particular puzzle unsolvable, a trap unavoidable etc.

The more you test it yourself, the more assumptions you make

question is what is the best way to get any sort of feedback? just uploading the map doesn't seems to give anything, even if gets a couple of subscribers.

make a thread maybe? but if everyone made a thread about their maps then maybe only a few would ever be looked at.

it's kinda hard to continue or improve a map without any sort of feedback.

[Komag]

I think the best way is to recruit a couple folks to specifically test your map and get back to you on it