Theoretical Discussion: Deciding What To Build

I'm always curious about how other mappers work. What I'm wondering is how you decide exactly what to build -- when you don't have a specific type of encounter in mind. The "don't" is the very important part.

I open basically every map I play in the editor, and large maps like those in BTSX awe me with their combination of purposeless anonymity yet macrostructural coherence. (Hopefully that sentence is at least 5% understandable -- I woke up recently and can't English well.) As an example, here is map10 of BTSX E2, things temporarily removed.



Part of the appeal of designing a map around encounters, to me, is knowing exactly why every major structural element makes it in. I can say to myself, "Okay, I'll set up a ledge right here because I want revenant snipers. I'll carve out some nukage pools right there because I want to exert some kind of space restriction. I'll erect some pillars right here that'll serve as turrets for chaingunners and then as cover for the later archvile fight". The feeling is one of control. Of course, I still have to improvise in certain ways -- especially wrt exactly how these various elements will be shaped and positioned with respect to one another -- but the heuristic of "make the shapes aesthetically appealing from a top-down perspective" generally satisfies me.

But a lot of the combat in exploration-centric maps is frontal, incidental, free-flowing without being staged as specific traps and set pieces. When I design stuff like this, I'm essentially just "winging it". The thought process might be something like:

1. I want a semi-open layout with some sort of spatial contrast and height variation.

2. Okay, I'll make a room shaped like this because it looks nice.

3. I'll make a room shaped like this because it looks nice.

4. I'll make a room shaped like this because it look nice.

5. I want this journey to culminate in a jump into a pool inaccessible from ground level, where the player grabs a soulsphere or something, so I'll build a room with a pool and design stairs.

6. I'll make a room shaped like this because it looks nice.

7. I'll make a room shaped like this because it looks nice.

(I'm using room in the loose sense that means "one of the units that makes up an area", not necessarily a literal room. Cave chambers and stuff can count as rooms too.)

I don't really see how I could build anything resembling the snippet of BTSX e2map10 I showed you without essentially just winging it the whole way, subject to the whims of my aesthetic tastes, with logic in the distant backseat.

I guess you could try to build and then adhere to some sort of narrative. Like, "Okay, this sort of place would have a set of triangular shaped rooms with a shrine in the center, and hallways leading to them, and a ring-shaped hallway around that". Or "adjacent to this crate room, there should be a small control room with computers and a dead body stowed away".

Sure. But even if you do that, how exactly do you decide how these individual rooms and hallways are shaped? In a way that satisfies you.

Here's a snippet of a map of mine.



I like the way this area turned out. But I can't say that I had much of a grand plan beyond "Make a semi-open layout with rooms connected at multiple junctures with shapes that look appealing to me" and then I winged it (and later attached alcoves for various reasons). And as a result, I emerged from the process with a somewhat empty feeling, not sure how consistently I could replicate this.

So as I look into designing larger maps, I want to have some sense that I can exercise some manner of control over the exploratory fragments that I build, and not just design semi-random things that get in because they seemed like a good idea at the time and were good enough to meet my quality-control requirements afterwards (I delete or overhaul a decent amount of what I design, thankfully).

So I'm looking to you. How exactly do you build your stuff! Windbaggery over. :D

I'm personally not arcade-minded, preferring game-play based on exploration and puzzle-solving, using monsters mostly as ambient threats, so my rooms usually aren't designed around specific encounters (and in fact, if there IS a specific encounter in my mind it usually revolves around robbing good players of their usual advantages by, IE, creating tricky places to walk or spamming frustrating & random enemies, mainly lost souls) so most of the spaces I create follow an overall level narrative - I'm always trying to make some statement with a map, no matter how obtuse it might be. I notice I usually get the criticism of, "I don't think this map achieved what it set out to do" which is usually hilarious, but I guess I could practise mapping more intelligibly.

With room shape I usually try to create shapes that imply there's more to see around the corner. I love dead ends and doors for this reason, as well as pillars and low ceilings. I've also found that creating some segment of a room, rotating it to a non-orthogonal angle and then building off that creates shapes the player really has to explore properly - grid-based squareness usually allows the player to feel a little too comfortable, or at least that's how I feel.

dunno if this answered your question at all, hee :) I can also admit that not all the shapes I've mapped make sense because they might have been done in a moment of procrastination... detailing and gradiating a hallway while I feel unenthused about drawing what's meant to be on the other end and so forth. That comes from the way I compose - get one segment of the progression down, then listen to it a lot of times and work out ways to introduce variation.

edit: I can admit to some amount of combat thought, since after I test a map a few times I do put slow-damaging liquid sectors in some of the most obvious sniping positions. I try not to make the levels look too kaizo, though.

I have trouble a lot while mapping as I keep switching back and forth between making sectors for cosmetic reasons and making sectors for gameplay purposes. I used to pretty much always go with the former and let the gameplay create itself. But I've been so attracted to difficult maps that just piling monsters up in a room isn't gonna cut it.

I need my ledges to put monsters on, monster closets, pillars to hide from archviles or chaingun guys, and my details smoothed out so as not to impede movement, etc. I think one of the reasons I get so few maps done lately is that as I get closer to completing a map, i can't seem to shake the feeling that I can do better than this. These ledges dont make any sense and these cool lighting effects make it tricky to see the ammo i dotted along the walls, etc.

I consider them to be my two modes of thinking and i end up either making a fun map that looks stupid or a good looking map with nothing exciting or innovative in it.

Ah, this is a topic I find very interesting to discuss, because there's alot of varied approaches.

Some mappers are calculated, planning the map before they start. Some mappers are do-ers, taking less concern over the general layout and flow of their maps and just compose something as they go.

I view myself as a mid-point between the two areas.

The starting point - theme
So, my mapping process may seem odd to you but to decide what to build I need a theme. For this topic, let's use Egyptian / Aztec as our theme.

So I will make a couple of practice assets and dummy rooms with no connection between them. These are to help me learn a bit about the textures I plan on using, the features of the theme's architecture; ways I can make rooms / areas feel different, yet consistent.

So now I've established a theme, got a basic texture scheme going (Brown bricks, sand, vines) and know of a couple of landmark objects I can use (such as obelisks & tombs), I can start to make a map.


The layout
Now, the layout part - I never, ever plan a layout fully before I map. Instead of layout a map, I theorize a few objectives and leave them open to adjustment as I progress through mapping.

In our fictitious Aztec map, my objectives are as follows (I came up with these on the spot, much like I would when planning a map)

- Locate a hidden tomb
- find a key in the tomb to enter another place seen earlier in map
- Beat a 'set-piece fight' that makes use of pain elementals
- find a desert oasis and beat a boss



Start mapping
Now I grab one of those rooms / areas mapped in the practice session; this will become the map's temporary (potentially final) starting point. Smash in 8 player starts and build from there.

This is where my no-plan layout method works. I often try to apply a 'death-match' rule - at least two exits to any one location should keep the map flowing. So I might start with a corridor, and before I've detailed it I address what directions the map should go in. I leave holes in the wall for continued corridors, or a door to a room, etc. Then I detail this room - detailing isn't a quick process normally, and it allows me time to visualize the next area(s) of the map - I try to be mentally prepared for two or more rooms ahead of myself.

After a few rooms are done, assuming it's a predominantly indoors map, I feel the need to break up any sense of monotony - time for a huge room and/or an outdoors area - this can serve as a great non-linearity tool, too, making it a hub that connects to several areas of the map.



I know this is a brief overview of how I map but maybe it'll inspire you guys, or alternately, open me up to criticism so I can improve my methods.

I only need two things in order to map: an established theme, and broad descriptor of the map and gameplay type.

In fact I'm so OCD about coherent themes within a map that almost every map I've made to date contains 50 or less unique textures, some less than 40, and even a good portion less than 30.

I tend to also gravitate heavily to creating believable structures that feel purposeful or convey a strong sense of place. Immersive, if you will. I think it also limits some of my actual architecture though, as I tend to stay away from curves unless it's outdoors or a natural object. In theory it makes sense, why would someone build a place with a bunch of crooked hallways? It's inefficient, and I think that's another facet of the OCD I have.

Now, when it comes to building the actual map, deciding what to build is quite a procedure for me. Sometimes I have a dream and then recreate that dream as the map, I've done this for about half the maps I've produced. The rest of the time, I spend hours perusing all the different combinations of monsters and situations involved spatially in order to create a unique encounter or section of a map. Once something pops into my mind it becomes absolute and I proceed to making it in the editor.

A couple of times I have just 'winged it', and it worked out. My entry in the first map roulette was pretty decent. It's also looking like that's what I have to do this next time, cause I have nothing and haven't come up with anything since I joined it!

rdwpa said:

Sure. But even if you do that, how exactly do you decide how these individual rooms and hallways are shaped? In a way that satisfies you.

on that point I usually try to avoid long sightlines, partly to avoid a visplane overflow but mostly because I think it's more interesting if you can't see too much of the level all at once. the idea is to keep the player wondering what's around that corner & so on, so I try to block the player's view with something. it can be done lots of different ways like with narrow maze areas or rocks like in E3M2, etc. it gets back to one of romero's rules about contrast between cramped & open areas. there's also the possibility that a hallway, for example, can frame something at the end in a dramatic way. I think that's called a terminating vista.

I'm impacted so much by the level format/game-type that it's hard to know where to start really. In Quake 3, I literally had a system of standards that I came up with that defined the structures. All stairs fit on the grid a certain way, certain heights had to be met, only certain kinds of arches were used, rooms had to be certain widths and lengths...I made it so stringent so I could stick to my own insane theme and I was intentionally trying to push the limits of the compiler.

In Doom, my approach has changed completely. Everything is a-symmetrical, everything has curves (just about) I wanted things to look as much as possible like they came straight out of a sketch in a notebook (and they did) but also, the fact that adding another linedef is practically an afterthought, you're free to do all kinds of wonky shit.

In what I'm working on now, I've actually put together a really loose story line that takes place in several locations. These places provide the themes for the maps, and from there..it's details.

If I need to build a room that's a command center, I reference pictures of real command centers, contrast that with what doom's actually capable of, and construct many....many iterations...play through a ton of times to see how it fits, and if it makes the level a more enjoyable experience, I keep it.

Similarly, with the aesthetic of something ..if I'm REALLY applying myself in that department, I'll pester myself with the way something looks until it meshes well with the environment. It's very much a theme-driven spaghetti test.

I think most of the "don'ts" I've uncovered thus far are obvious ones. Spaces that are too big make the player seem slow and are practically impossible to decorate well. Spaces that are too small hinder freedom of motion. Spaces that are too tall don't add anything to gameplay past a certain point.

I don't wanna trash the visibly hard work someone did on a WAD I was playing the other day so I won't call it out by name, but throughout the whole thing, the level design looked great, and it was interesting because it kept crossing back over itself in unique ways...you had to think about where you were going. The gameplay, however, sucked. Each room was a serious-sam kind of thing with a hoard of either zombiemen, imps, or shotgun guys..over and over again for like 7 levels before you even saw another type of enemy. At first, it seemed like popping bubble-wrap, there were little bursts of gratification from clearing rooms out...but once you realized it wasn't going to change, it got so flat I was just cheating my way through to see the rest of the levels...because playing it wasn't fun enough anymore to justify how long it took to get through.

@yakfak -

You mentioned:
"preferring game-play based on exploration and puzzle-solving, using monsters mostly as ambient threats, so my rooms usually aren't designed around specific encounters"

I'm fascinated in what you do while I personally, completely disagree with it. To me, the best moments in doom revolved around encounters and the combat that stemmed from the space you're in. I like a puzzle here and there, I'm actually hoping to build some into my current project, but my love for doom wouldn't be what it is without those encounters.

@Rayzik -

You said:
"I tend to also gravitate heavily to creating believable structures that feel purposeful or convey a strong sense of place."

I lean this way as well, I build hunks into my maps that could be part of the story of a map...like starting a level in a small outpost building or finding an item in a computer room.


As for my process start to finish....

I sit at my desk at work, often times I do my best to ignore all really dumb office discussions going on around me, and I draw. Eventually those drawings evolve into full blown maps, and when I get home, I transcribe those drawings into doom builder. I add a piece, play test it to death, and do it all over again until it feels right....or I get sick of a certain item or room and focus on a totally different part of the map. My starting point for creation is building something that I want to play. I've looked at other wads...and lots of other games since doom came out, and I've derived an idea of what I don't like in FPS games. Modern shooters taught me enough of that. I try to keep action at a pretty constant boil in my designs. I don't recall where I heard it, but at some game development thing ages ago, someone said "if you don't think it's fun, no one else will either" so I bring my own reference for what 'fun' is, and go from there.