Save-Point
Dungeon Theory by Brickroad - Printable Version

+- Save-Point (https://www.save-point.org)
+-- Forum: Games Development (https://www.save-point.org/forum-4.html)
+--- Forum: Tutorials (https://www.save-point.org/forum-19.html)
+--- Thread: Dungeon Theory by Brickroad (/thread-2896.html)



Dungeon Theory by Brickroad - DerVVulfman - 03-11-2009

Reprinting an article I had for a couple
years. I didn't write it...


Dungeon Theory
by Brickroad


What dungeons are and why they exist.

Dungeons are integral to a traditional RPG. Dungeons make up a large part of the gameplay, second only to actual combat. Because of this, any successful RPG must practice good dungeon design. Dungeons cannot be viewed as a perk, or a bonus, or the icing on the cake. Dungeons must be the focus, the star, and the cream-filled center.

A dungeon is defined as a place that is hostile to the player characters. Dungeons are populated with enemy characters, are the source of the bulk of an RPG's treasures and secrets, and are where most of the bosses are encountered. Dungeons offer the player a series of challenges, which must be overcome; they must enter prepared before entering and they must sustain themselves until the end.

Think about that for a moment. Why do heal spells exist in RPGs? Because otherwise, the PCs could not conveniently heal themselves in dungeons. Why do save points exist? Because otherwise you couldn't save in the dungeon. Why do towns exist? To give players a chance to recuperate from the last dungeon and prepare for the next one. Why do bosses exist? To provide the player with the final, most challenging step in a dungeon.

There are, of course, exceptions to every rule. But I challenge the reader to find an RPG that is an exception to all of the rules. RPGs are geared towards dungeon exploration. You don't have to think of many reasons to justify having dungeons in your game - the genre already justifies them for you.

For something that is so important to the genre, it's a shame so many RPGs don't give much love to dungeon design. The goal of this article is to examine why that is, and to offer some examples of how to remedy it.


Dungeon skeletons... spooky!

The skeleton of a dungeon is what it looks like before you add the graphics. Strip away the music, visuals, monsters, and story cut scenes from a dungeon, and it looks like any other; lines on paper, connecting areas. Maybe with some notes off to the side about puzzles and such, but a skeleton nonetheless. This is how every dungeon should start.

To the casual observer, it looks as though there are two kinds of dungeon skeletons. One skeleton is linear; the PCs follow a path from entrance to exit, collecting treasure and fighting progressively difficult enemies along the way. The other skeleton is more freeform, with many branches and multiple paths. For an example of the former, play most any Final Fantasy game; for an example of the latter, try Suikoden 1 or 2.

Either of these dungeon layouts is acceptable. In some ways, they are your only options; either your dungeon is linear or it is not. Either there is only one path or there are more. There's not a whole lot of flexibility to take advantage of here.

However, there are ways to make the skeleton more interesting. If you're using a linear skeleton, remember that lines don't have to be straight. You can loop the path around and through itself. You can put an early part of the path enticingly close to a later part. You can even require the player to cross a previous path, effectively creating a loop allowing the player to explore the dungeon in both directions.

You have these same options with a non-linear skeleton, except on a looser basis. With some clever twisting of paths, any part of a non-linear dungeon can be linked to any other part. A non-linear skeleton is really just two or more linear skeletons ravelled up inside each other, and the hero has the option to jump back and forth between bones.

The shape of a dungeon can be used to enhance the experience as well. Remember in the original Zelda how the rooms in the levels were put together in such a way that each level had a distinctive shape? When the mysterious old man said "SECRET IS IN TIP OF NOSE", awful Engrish aside, you knew where to hunt for the secret because you could see where the nose was on the map.

At the same time, the shape of a dungeon should be logical. A haunted mansion wouldn't consist of twisting passages and dead ends because mansions aren't designed with twisting passages and dead ends - they're designed with hallways that branch off into individual rooms. However, you could justify having secret passages or dilapidated rooms in a haunted mansion that would serve the same purpose as aforementioned dead ends.

Thus, the skeleton of a dungeon should give you an idea of how it looks from space. A successful dungeon will keep the shape and structure of the dungeon within the realistic confines of its setting (more on setting below), and will offer a path that is interesting to follow.


Setting and theme... they aren't the same thing!

The setting of a dungeon is its physical location. Caves, castles, canyons, mountains, mines, monasteries, battleships, airships, spaceships, towers, towns, tundra, wells, waterfalls, and whale's bellies are all potential settings for dungeons.

The theme of a dungeon is its mood, its flavor, and its soul. So you've got a forest... is it a dark, spooky forest full of rotting stumps and spider's webs? Or is it a beautiful spring forest full of fairies and unicorns? Or a woodsy nature hike full of birdcalls and blowing leaves?

All dungeons have a setting; that's a given. But not all of them have a theme. The theme is something extra. In FF4 you go through the Misty Cave, the Watery Cave, the Antlion's Cave, the Baron Cave, the Magnetic Cave, the Eblan Cave, the Sealed Cave, the Sylph Cave, the Monster Cave, Bahamut's Cave, and a half-dozen Lunar Caves. They certainly don't all look and feel the same. Each cave has at least one (and often more) additives that spice up the deal. In the Magnetic Cave, you have equipment restriction due to strong magnetic fields. In the Sealed Cave, you must walk across tightropes and fight dozens of trap doors. In the Sylph Cave you have to navigate damaging tiles, and you can even see the level underneath you through the organic web-like floor.


A good dungeon is always described in at least two words. It isn't enough to have a "palace". Is it a Sealed Palace or a Crystal Palace or an Ocean Palace? A strong bridge or a rope bridge? An occupied well-guarded prison or a decaying abandoned one? It's your basic adjective-noun scenario; the noun is where your dungeon is, the adjective is what your dungeon is.

The combination of setting and theme is stretched over the skeleton of the dungeon to form its skin.


Gettin' gimmicky with it.

Even the most interesting skeletons with the most impressive skins can be boring. Dungeons are oftentimes just excuses to fight lots of random encounters, collect a few treasures, and fight a boss. In this light, no amount of variation in setting and theme will make one dungeon different from the others... they'll all run together in the player's mind.

To help prevent this, some dungeons offer the player something that keeps their mind focused on the dungeon itself rather than the battles therein. These are called gimmicks, and they provide good opportunities for the player to interact with the environment. In other words, gimmicks are puzzles, diversions, tricks, obstacles, and illusions that provide ways to trip the player up other than just having the path split in multiple directions.

Sometimes dungeons have one all-encompassing gimmick. The aforementioned Magnetic Cave is an example; the player simply cannot use certain types of equipment in that dungeon. In Kefka's Tower, you have to simultaneously control three parties in order to reach the goal. Certain gimmicks in Breath of Fire include spinning the dungeon wildly so the player becomes disoriented, areas with invisible walls, and arrow tiles that phase the PCs in various directions.

Other times dungeons have a series of gimmicks. Each dungeon in a Zelda game is a sequential collection of puzzles that Link must solve by using his screen full of toys and gadgets. Lufia 2, a title which I hope requires no introduction, is the king of gimmick dungeons; it's chock full of block puzzles, clever switch puzzles, twisting passageways, and various environmental elements which sets it above and beyond all other RPGs in dungeon design. An aspiring RPG designer could do much, much worse than to take some notes from Lufia 2.

Gimmicks can also be used to strengthen a dungeon's theme. An ice cave with slippery floors and brittle walls seems icier than an ice cave without. Swinging rope vine mazes would enhance any jungle dungeon. Quicksand for deserts, rolling boulders for mineshafts, keycard-protected airlocks for space satellites. Dungeon gimmicks designed with setting and theme in mind add flavour without looking out-of-place.

Taken a step further, many times a gimmick will reward the player for mastery. They must solve the gimmick to (x) specification in order to continue in the dungeon, but if they solve it to (y) specification they get a treat. Say you've got a crumbling bridge the PCs have to cross. The door to the next room is halfway across, but if they manage to make it all the way they get some treasure as well as the ability to continue. Not only does this restrict access to the more powerful treasures in more meaningful ways than just putting them at the end of a long tunnel or mini-boss fight, but it also gives the player some incentive to hang out in a dungeon longer than usual, thus creating a more memorable experience.

Gimmicks symbolize the organs of a good dungeon; they're what give dungeons life.


Roar snarl snap!

Combat is as important a factor in dungeon design as anything mentioned in the above paragraphs. The monsters in your dungeon have to fit right; you wouldn't see evil bird-men living deep underground, or fiery ghosts in an arctic mine. Fortunately, there is little more than creating monster variations than swapping their sprite, so it's not much of an issue.

Of more important interest is the makeup of dungeon's population. How many different types of monsters a dungeon should have, and what they should be able to do, is a huge consideration. Here are a few useful monster archetypes:

- a monster with little HP that isn't much of a danger, but has enormous defense and is coded to escape at the earliest available opportunity.

- a monster that has significantly more HP and attack power than any other monster around, but always appears alone.

- a monster with little in the way of physical attributes but with one or more attacks that hit the entire party.

- a monster that gets two attacks each round.

Having all four of these monster archetypes in the same dungeon might make for some interesting battles. The player has determine which threat is the largest and eliminate it. This goes a long way to make battles more interesting than just "attack over and over until everything is dead".

I would like to take a moment here and discuss a phenomenon I call experience plateau. In simplest terms, an experience plateau is the most amount of any given monster in a single fight that still makes that monster worth fighting, given the risk. The reason experience plateau exists is because experience gain is additive while monster difficulty is multiplicative.

Take the monster above that gets two attack a round, and call him DeathFrog. Say DeathFrog is worth 100 experience. Two DeathFrogs, then, are worth 200, three are worth 300, and n are worth 100n. For each DeathFrog we add, the player gains another 100 experience if he wins the fight.

Say one DeathFrog is capable of hitting for 50 damage per hit; that's 100 damage per round since he gets two attacks. The player might think 100 damage is a worthwhile trade-off for 100 experience. Adding a second DeathFrog, however, and the player isn't going to take 200 damage like you might think, but rather 300 damage. He'll take 200 the first round where he kills one DeathFrog, then 100 the next round. 300 damage for 200 experience. With four DeathFrogs, the player stands to gain 400 experience but take 1000 damage total (and more than likely burn off healing items in the process). Notice that four DeathFrogs means four times the experience but ten times the damage.

Obviously this phenomenon is less dangerous the more people there are in the player's party, the more multi-target attack spells they have, and whether or not they get to act before the monsters do on a regular basis. Indeed, it's a useful tool if used correctly (you can re-use older monsters in larger groups to match the current dungeon's challenge level, therefore diversifying its population, for example). The point is, it's very easy to unwittingly make a dungeon far more difficult that it's supposed to be by ignoring the experience plateau factor.


Oooo... shiny!

Without adding rewards to a dungeon, it's nothing but a path to the boss. The most common kind of dungeon reward is treasure. This article recognizes four different types of treasure that can be found in dungeons.

Exceptional treasure includes powerful weapons, great armor, new spells or special abilities, incredible items (such as one that restores the HP/MP of the whole party), and anything else that gives the party a considerable gain in strength. Exceptional treasure should never just be thrown around; great stuff like this, the player has to work for. Puzzles, mini-bosses, or even just long treks off the beaten path are good ways to "guard" an exceptional treasure.

Good treasures are somewhat hidden or out-of-the-way, but are worth the trouble to get. For example, the player might not mind backtracking through half the dungeon for a treasure box filled with 5,000 gold. A Hi-Potion is a good treasure if all the player can buy in shops at the moment are regular Potions. A weapon or piece of armor that the player otherwise can't buy until the next town (or maybe just couldn't afford in the last one) is another good example of a good treasure.

Common treasures give the player slight incentive to explore your dungeon, but are little more than minor momentary victories. Common treasures are often hidden in plain sight, and contain things that, while helpful, won't give the party any kind of advantage. A regular Potion, a few pieces of gold, or inferior armor (good for nothing other than its resale value) classifies as common treasure.

Poor treasures are literally anything that is not worth the trouble to get. The player won't like fighting a powerful mini-boss for an Antidote. Nor will they be happy using up 500 gold worth of potions just to get to a treasure box containing 100 gold. If the treasure causes the player an overall loss in power (or at least doesn't do him any better than breaking even) it qualifies as poor.

Different games have different treasure makeups. Each dungeon in a Zelda game has at least one exceptional treasure (a new tool or weapon), several good ones (skulltulas, treasure maps, compasses, fairies) and dozens of common ones (rupees, hearts, bombs, arrows). To find them, though, Link must remain ever vigilant and seek out every nook and cranny of a dungeon.


The light at the end of the tunnel.

I have discussed the anatomy of a dungeon, various types of dungeon layouts, and how to make a dungeon interesting by giving it a theme that enhances its setting. I've also discussed how to utilize gimmicks and monsters to offer the player an interesting challenge. Finally, I've touched on using treasure as incentive to explore the dungeon in total.

The potential for a dungeon is limitless. It can be made to be hellaciously fun or gruesomely irritating. It can be made laughably easy or bone-crushingly difficult. They come in all shapes and sizes, all moods and colors, with all different manners of monsters and rewards.

It bears repeating: dungeons are crucial to interesting gameplay in an RPG. Without them, there's nothing to do but run around the world map fighting monsters.



Dungeon Theory by Brickroad - Sporky - 03-20-2009

That's a pretty decent article :) Thanks for reposting it!