Detail first, Gameplay later

Many people agree that focusing on gameplay among other things is the key to good map design. For me however, sometimes that is easier said than done. Often I may create a scratch room with the types of monsters I want to see choreographed together, and once ive got the monsters and items arranged the way I like, I'm left with an assortment of amorphous blob sectors that I have no idea how I'm going to make an actually good looking map with.

So I'm curious, does anyone find that designing the layout, and detailing the map so that it looks like the kind of place you want your map to look like first before placing any monsters works better for you?

Most map reviews considered, a really good looking map with remotely mundane gameplay is often excusable just because its a pleasure to look at while you kill easy monsters with basic weapons. I don't necessarily mean to condone maps with unremarkable gameplay (eg maps with enough health and ammo but mostly easy monsters that always approach you from the front, lacking in traps, places to run, jump across, explore, etc.) but I am curious as to what point testing and finalized thing placement comes in the process of mapping for some people.

I think focusing on good gameplay scenarios is an essential part of creating good maps, but for me I find it helps tremendously to have an empty map to run around in first before I get a feel for the kinds of monsters i am going to want to fight, and how they will present themselves to the player.

I'm absolutely unable to make rough sketches. I can only create fully designed areas one-by-one with some of the most important monsters\objects already there, and when everything is done I place the rest of the things in 30 minutes or so. After that only minor tweaks to the gameplay, move\add a few things here and there. I guess I spend less than 5% of map making time on thing placement.

That's pretty interesting considering how well received your maps usually are, and also how critical you can be on other peoples maps :)

I think I've been stumping myself a lot lately because ill play a map with something like mancubi used in an interesting way, and then ill open the editor with no goal in mind other than "I want a map where i get to blow up a bunch of fatsoes with rockets" and ill create a giant assortment of startan rooms with piles of mancubi on tall pillars and ledges and all the rockets i could ever want, and by the time I'm finished my imagination basically switches off as to how to make a decent looking map out of what I just made.

Perhaps a major part is simply having an awareness for how much space the player needs to deal with certain types of monsters, and preparing closets and traps in advance for places where you think you might need them.

40oz said:

I think I've been stumping myself a lot lately because ill play a map with something like mancubi used in an interesting way, and then ill open the editor with no goal in mind other than "I want a map where i get to blow up a bunch of fatsoes with rockets" and ill create a giant assortment of startan rooms with piles of mancubi on tall pillars and ledges and all the rockets i could ever want, and by the time I'm finished my imagination basically switches off as to how to make a decent looking map out of what I just made.

I wish I was able to make maps that way. My mapping is all about adding rooms, details and things at the same time, and then tweaking and reshaping them until I'm either satisfied (and then I go adding new ones), or bored / out of ideas. All of my maps were made this way, I think.

It is difficult to say what should come first, level architecture & detail or gameplay... which is not only monsters placement but also locked area and doors with key.

When I build the level I already have in mind the type of battles and enemies - generally speaking. I add rooms and details, and then the enemies.

The problem is that if a battle is not "staged" well in a area, I amy have difficult to reshape the area.

scifista42, do you ever reshape an area in order to improve the battle in it?

Layout, basic detail and functionality, monsters and other thing placement, finish up visuals. That's my process.

For me, the most intuitive way is to draw and place monster as I advance. However, I think it's much more comfortable to postpone the placement of interactive objects after the level is laid out, since I no longer have to worry about balance while the design isn't finished. On the other hand, it can be less enjoyable not being able to casually play the map I'm building just for the fun of it. However, I think this one detracts from the productivity.

A problem with laying out the map before placing any monster is that it's possible to make the map very ill-prepared for monsters: they can get stuck, have unfair visibility to the player and generally not be fun to play against monsters. I'd probably have to readjust sectors then.

Like most of 40oz's threads, now I'm inspired to continue working on my current level, by applying this layout-before-monsters methodology.

Angry Saint said:

scifista42, do you ever reshape an area in order to improve the battle in it?

Not radically, but at least I try to. I add monster closets or ledges, and fiddle with shapes that affect player/monster movement, line of sight and paths of projectiles (for example changing size and height of windows or similar obstacles). But the problem is, I've never created particularly elaborate battles that would work well, and I feel that results of my mapping are half accidental.

you've been making quite a few "meta"-mapmaking threads lately, 40. :p

40oz said:

Perhaps a major part is simply having an awareness for how much space the player needs to deal with certain types of monsters, and preparing closets and traps in advance for places where you think you might need them.

this. I'd like to believe I have a decent sense for it because my design process usually goes: conceive a few set pieces and/or a basic layout construction and/or a visual motif --> allot appropriate space for each area within layout --> detail/polish/add-functionality/whatever to each of these pieces (can be done in parallel) --> play through once or twice to finalize balance. in extreme cases balancing may involve scrapping or reworking an area, but in general it's just playing around with monster/ammo/health positions. I pretty much never have to rescale anything anymore, amount of space is usually always decent.

If you already know which sectors in a particular area are going to be dynamic (e.g. will lower to reveal monsters, will raise to create steps/lifts/etc, will host teleport sectors) you can feel free to splurge on detailing the static areas to your heart's content. It only ever gets annoying if you decide you just have to have an elaborate piece of detailing overlap functional sectors (in my case it's usually pretty ceiling detail that intersects some floor stuffs), at which point it's time to bust out 10000 overlapping linedefs and a bajillion control sectors. I like to make sure I have everything else in that area polished to my liking before pulling the trigger on stuff like that though.

to explicitly respond to the thread title: detail and gameplay in parallel. I find myself much more interested in making a map if it has a clear identity. If you can establish visual aesthetic and defining gameplay motifs of a map early (within the first few room prototypes that get vomited into db2) the rest of the map practically makes itself. With a strong enough atmosphere and strong enough style of enemy placement you can pretty much just draw random shapes and instantly recognize "well this would be a great place for a balcony, this would be a great place for some overhanging ceiling detail, monsters could tele in from here after player crosses here, I could put some of those pretty pillars I used 2 rooms ago over there, monsters could lower from over here, if I wanted to be a total dickweed this would be an awesome place for an AV, etc etc etc...."

Gameplay first, details later. I barely put any details on my map, good lightning plus architecture will automatically do it for you, especially in vanilla when you work with a budget. Important gameplay pieces should always comes in mind when designing a level. Usually I find it harder to put roaming monsters to keep the action flowing rather than set-piece encounters.

Ribbiks said:

you've been making quite a few "meta"-mapmaking threads lately, 40. :p

To me it seems he has been making them since ever on this forum. :D

I just seem to make details first, to get an impression of what my map would look like. Then, I would not work on it, pick it up and eventually finish it. Especially when I'm making something for a community project and there is a deadline looming around.

I usually design the world with a lot of the basic scenario I want to create in the back of my mind. Always conaidering the monster AI and how it will work with diffferent layouts. Once the world design is lain out, as well as the odd specil encounter, I'll start working on the actual gameplay. This is usually the most involving process taking months of continuous testing. Some maps I've finished the design of in a weekend and then been testing and balancing fpr some six months after.

kristus said:

I usually design the world with a lot of the basic scenario I want to create in the back of my mind. Always conaidering the monster AI and how it will work with diffferent layouts. Once the world design is lain out, as well as the odd specil encounter, I'll start working on the actual gameplay. This is usually the most involving process taking months of continuous testing. Some maps I've finished the design of in a weekend and then been testing and balancing fpr some six months after.

What problems do you run into that makes it take that long, if you don't mind me asking? It seems like an awful long time to be moving things around, changing soundblock lines, etc.

For me it depends on how much of the design of the map I have foreplanned. I will put in enough elements so that the key mechanics of the map work correctly, whether that be layout, set pieces or working parts. It's unlikely detail will be a key mechanic so I will put this in last. However, if I have not planned what is happening ahead of time and I'm not feeling particularly creative, I will detail areas that I've already mapped out.

SFoZ911 said:

Gameplay first, details later. I barely put any details on my map, good lightning plus architecture will automatically do it for you, especially in vanilla when you work with a budget. Important gameplay pieces should always comes in mind when designing a level. Usually I find it harder to put roaming monsters to keep the action flowing rather than set-piece encounters.

How do you create gameplay first? Do you decide which type of battles there will be and then stage them in the level?

For me it would be strange to fight enemies in a "void" level, I mean a map done only of temporary textures without details.

For a long time, I used to detail room-by-room, because I had a difficult time figuring out how to move on from an area if I couldn't see it as finished. Working that way caused me a lot of stress and frustration, because once an area has been detailed, it's much harder to commit to making changes to it, as it usually involves rearranging more complex pieces and throwing out a bunch of work. I would end up frequently mapping myself into a corner, or compromising on layout ideas that would have required me to rebuild an already-completed area.

In the past few years (especially since working on D2TWID) I've been moving further and further away from that, and nowadays I almost always try to get a functioning layout that I can run around in first. This way makes it much easier to work with nonorthogonal shapes and to implement large-scale moving parts. Plus, if I've mapped myself into an awkward corner somewhere, it's much easier to knock down some walls, reshape pathways, or set up some moving sectors to solve the problem if nothing's been detailed yet.

I typically will get the layout completed before doing much thing placement in it, but traps/setpieces, perches/windows/ledges, etc. are all major contributors to the way a map plays, that come largely from the way the map is laid out.

When I was building and detailing room-by-room, mapping would get pretty draining because of all the compromises I'd end up making, and the end result wouldn't accomplish everything I had wanted it to, because I didn't plan enough before building. I still don't do much actual planning, but roughing things out first allows for the more stream-of-consciousness "ooh, what if I did _____ here?" methods to be less inhibited by technicalities.

I think the biggest hurdle for me was being able to see a bare, undetailed map with two or three textures as "complete", and not having an immediate urge to start aligning textures and adding detail. Working with intentionally simple aesthetic guidelines for D2TWID helped a lot with breaking away from that mindset. I still want to be able to envision where the map is going to go thematically, but sector shapes and a few placeholdery textures can be enough to suggest a level's theme as "towering space city" or "ancient flooded village" or whatever.

I don't test my maps before texturing, why would I? I decide which textures will be used as I build the layout, It really depends on the project I'm contributing for(slot, theme, difficulty etc..). I create gameplay first by neglecting details, I use directional light and architecture to *detail* the map and proper sky usage and outdoor areas to create a sense of place. I also make a generous use of decorative stuff like gore and dead bodies, pillars, lamps etc(memfis calls this lazy detailing but I like doing it not because I'm lazy) to add colors and overall atmosphere for certain areas. You probably won't find random computer stuff and things like that on my maps. Almost everything serves a purpose. Oh and special sector light to make things look alive.

I'm not good at posting long comments so forgive my grammar.

esselfortium said:

In the past few years (especially since working on D2TWID) I've been moving further and further away from that, and nowadays I almost always try to get a functioning layout that I can run around in first. This way makes it much easier to work with nonorthogonal shapes and to implement large-scale moving parts. Plus, if I've mapped myself into an awkward corner somewhere, it's much easier to knock down some walls, reshape pathways, or set up some moving sectors to solve the problem if nothing's been detailed yet.

I typically will get the layout completed before doing much thing placement in it, but traps/setpieces, perches/windows/ledges, etc. are all major contributors to the way a map plays, that come largely from the way the map is laid out.

This is how I started approaching mapping when I returned to it.

I tend to layout a basic design for rooms, walk around in it, then add to it. Usually I like to make sure the map can be played start to finish with no things. After that I place enemies and items based on an idea I had for an encounter or for what seems appropriate.

And in my opinion, gameplay must take precedence over visuals. I have always thought that about all games I have played and Doom is no different. Like someone said, a map with mediocre gameplay is excusable if it has pretty architecture. That just isn't good enough. If I wanted to just admire the architecture I would turn monsters off and walk through it.

It's about striking a balance. And hey, if it has good gameplay AND visuals that's great. But one shouldn't be grossly compromised for the other especially when it comes to gameplay.