Design Theory Discussion: Dungeons

Trying to decide how to design the actually dungeons in my project.

Do I go all Link to the past style with a wide variety of "Rooms"

Could go like Pokemon Mystery Team and a block area for levels or floors.

I liked the way Neo Online was but I don't really know how to recreate that style in RPGMVXA

Spiral Knights has an awesome look and style but idunno about RMVXA way to do that.

Wondering how other RM uses do it?

 

It might help the discussion if you describe the way those programs make their dungeons for those people who never played them.

For example, I don't know any of the four games you mentioned, so I can't comment on wether they made a good job or not.

For me, a dungeon needs to be logically constructed - you don't place a lava room in a thieve's den below a forest and so on...

 

That sounds like a challenge to me, I'm off to create a masterpiece! =P

Really though on topic, the dungeons design itself should solely dependent on the kind of dungeon it is. Square rooms and corridors may work for old ruins but doing that in a Cave (Although it will get the job done basically) is kinda out of place as caves tend to follow the organic patterns in life more then geometric.

PMT uses the squares and corridors for everything (from the one I played on GBA) and that's more because it's the quickest easiest way to do it then anything else, I think recettear did the same. But those are procedurally generated which is where the whole easiest way takes hold really. For hand crafted the shapes should really depend on the kind of dungeon it is.

Are you talking about doing a procedural thing? because if so I can think of a off sort of way of getting the best of both worlds but I wouldn't know how RMVXA could actually do it.

 

Like Andar said, I think dungeons should have a logical design. Also, I like puzzles that get harder and more complex as you progress through a dungeon. If all a dungeon consists of is walking around fighting and collecting loot, it can get old quickly. It's cool if the puzzles fit the theme of the dungeon as well. For example, pushing blocks around in a fire themed dungeon would be awkward while using pails of water to put out small fires in a particular order would fit better.

 

My general philosophy on dungeon design:

• Even if the dungeon is linear, it shouldn't feel linear.  Have the path double back, loop around, and have a few dead ends.  Have the player take a different path through a room they were in earlier (such as a bridge over a lower path).  Include things that are visually interesting and interesting to explore.
• If your dungeon isn't linear, have different paths offer different experiences.  For example, one path may feature logic puzzles while another features combat and the third features a platforming section.  This encourages the player to try all the paths anyway or choose the path that best fits his favorite gameplay experience.  Either way, player choice and possibly gameplay extension (without padding).
• At least half of dead ends should have a reward of some sort.  In general, the further the player has to go out of his way or the harder he has to fight to get it, the better the reward should be.
• If you have puzzles, do not include random encounters in puzzle areas.  Visible encounters are maybe okay, depending on the puzzle and number of encounters (and how they respawn).
• Dungeons should be memorable.  If all your dungeons are interchangeable save for wallpaper, your player's experience in them is not going to stand out that much.
• If backtracking to previous locations is desirable, make a "shortcut" through the dungeon that is unlocked after completing it.  If the dungeon ends with a giant dead end, provide a shortcut such as a teleport back to the entrance.  No one likes being forced to trek back through the same dungeon they already explored, unless:
• Include optional areas that may be too difficult or blocked for the player right now (an optional miniboss (that you warn the player is too tough), a locked door to a room with treasure and optional plot content, a treasure chest that you can't get until you get the rock smashing hammer, etc).  That allows you to reuse your dungeons and give them a bit more utility, extending playtime and fun for the player.

 

In my view, the most "puzzling" areas should be in the midgame. In the early game, you want to establish themes and emotionally engage the player. Plus, there's fighting to be done and loot to get. But you gradually want to ramp up the difficulty. Toward the end, though, there are diminishing returns on the excitement generated by difficult puzzles. But by the end, having established the story, and challenged the player with difficult puzzles in the midgame, for the last part, you can focus on resolutions, repeating themes established earlier, and most of all, providing payoffs for their hard work. 

 

[SIZE=12pt]A design aspect I would like to point out, which hasn’t yet, is the idea of only one puzzle in a dungeon. I find that the best dungeons with puzzles actually only have one type of puzzle but its equations progressively get harder as you solve them. This is good since it teaches the player the rules of the puzzle by placing out a super easy problem as the first one; this one acts as a tutorial of the puzzle so that the harder problems of the puzzle aren’t unfair to the player since they’ll know how to figure it out. The Tales and Pokemon series’ tend to pull this trait off very well. [/SIZE]

As for what kind you should use for your game, it really matters on the type of game you're making. [SIZE=12pt]They have to make sense to the plot and feel of the part said dungeon's at. They generally…[/SIZE]

[SIZE=12pt]Shouldn’t feel like filler for the game.[/SIZE][SIZE=12pt] I’ve found many RPGs that failed at this miserably, whether the dungeons were just unnecessarily long, or seem to come out of nowhere and don’t contribute to the plot of the game whatsoever. Exceptions are for optional dungeons such as the caves that could be explored in the Pokemon games.[/SIZE]

[SIZE=12pt]Should have story throughout.[/SIZE][SIZE=12pt] A lot of dungeons tend to perform a pattern of going through the whole dungeon; then, revealing the next plot point which can have a negative effect on the story’s pacing. This idea should generally apply to longer dungeons so that the theme of the dungeon doesn’t get distracted by the plot.[/SIZE]

[SIZE=12pt]Should reflect the mood of the plot at the point of doing them. [/SIZE][SIZE=12pt]What I mean by this is if all the party is doing in a dungeon is going on some happy-happy-joy-joy treasure hunt then the feel of the dungeon probably shouldn’t be too depressing, unless the 2nd point is applied to suddenly change to tone of the plot. Likewise, if the tone of the game is depressing, then walking into a dungeon with pastel colors and (unfrightening) clowns everywhere probably wouldn’t fit very well unless it’s using the 2nd point. This really isn’t required, although it does make a dungeon turn from good to great real easy. [/SIZE]

[SIZE=12pt]Overall, what I’m trying to get across is that the aesthetic\technical design of a dungeon should make sense in the part of the game it’s in[/SIZE].

 

I think it's important to nail down the dungeon's character initially and then a lot of the pieces start falling into place and a lot of the general dungeon rules become much less restrictive.

Now, when I say character, I don't mean fire cave, tomb, underwater ragefest, I mean to say a great dungeon really is a huge NPC and you're crawling around in its belly. Is it a cold, stoic relic silently awaiting challengers to take on its trials or is it a wild beast that actively hates any who dare plumb its fire-spewing depths? These are two dungeons that can offer incredibly different experiences and applying too many general rules are going to water down those experiences into the same old grind. The first dungeon would prolly be a lot more linear, giving you an obvious path from one challenge to the next and encounters would prolly be sparse, but a lot tougher than what you're used to thus far. I'm thinking the second dungeon would be a big windy mess with events trapping you onto certain paths and leaving the party narrowly escaping from red hot doom at a few different points, pissing off the player as it baits them with treasure and snatches it away at the last moment.

Of course there are those general rules that others have illustrated really well here, but don't be afraid to bend them or change their priority if it suits the concept. It really is possible to make a strait line enjoyable if you have a solid idea backing it up.