News | Forum | People | FAQ | Links | Search | Register | Log in
How Do You Design Your Stuff?
I was enjoying the talk about mapping methodology in the RMQ thread and I personally think it could make an interesting thread in itself...

I plan a lot (far from all) of a basic layout on paper first, just for a rough flow overview, usually changing it a lot when adding monsters and such and then finding out it doesn't work as I expected. Although I often make something out of throwing together one area/fight I like and bolting bits onto the sides. As I said in the the RMQ thread I dislike doing architecture and it's usually the last thing on my mind, apart from thinking about grander elements such as large buildings or centerpieces... this btw is one of the main reasons I haven't properly finished anything yet :P

I don't have a scanner so sorry for cack pic

Anyone else care to share their ideas/techniques? :)
 
I do almost everything organically, i.e. I just build. It's working o.k. so far, but I can understand why someone would want to plan a map's layout. 
 
my plan is not having a plan! :) 
Nice Dev Drawings 
I have a bad habit of using only my minds eye. The plan or sketch in my head can change a lot whilst I am mapping. But I often think I would do better if I completed a proper planning phase first..... 
Rambling On 
very cool concepts zqf.

I usually do not do that; I have an acceptable memory and I use that. I also noticed my layouts tend to be a lot flatter when I use paper sketches.

I realized that it is important to have a central challenge in a level (activate the bridge, activate the slipgate). Opening the exit can also be turned into a challenge.

I design levels around such a central task (or one bigger, one smaller task).

Ideally, the main challenge should be an interactive setpiece, and the player should understand what the task is before being sent off on their button-pressing / key-finding / monster-killing quest.

The finding stuff part can be as long as you like, and you can torture the player along this path in any way you see fit.

There can be any number of sub-tasks (between 1 and 4 would be reasonable, which is why RMQ has 4 standard keys and 4 custom keyitems, for example).

Do a lead in, establish the theme, then show the player the challenge. Typically the way is blocked somehow, and they must remove the obstacle. Then make them go and collect keys or whatever.

It is nice if the map wraps around itself in such a way that you can look into different parts, but not touch until you get there.

Apart from that, it's a repertoire of standard techniques to build stuff that just developed over a couple years. This goes especially for architecture / brush detail; build flat walls first, then break them up... once you get something that looks good, _do_ absolutely shamelessly reuse it everywhere, just vary it a bit. Be cheap and effective. If you built a cool pillar, copy and paste it! It'll become a recognizable feature. Same for door frames, wall supports etc. Try to come up with unique stuff for setpieces two or three times, though.

It is also nice to have parts of the level where you just kill monsters, no switches or keys involved. The map should not ramble on endlessly though, it should be structured. It's nice if there are ambushes or hotspots alternating with less combat.

The level should be largely linear, because players like knowing what to do next, and it makes it easier for the designer to predict the player. However, don't make it utterly linear. Include choices for the player at some points in the map and have little dead ends with goodies in them. Hide goodies everywhere.

A map doesn't need to be huge or extremely nonlinear in order to be fun to play; just reuse some areas or make the player stay in one area longer.

3D combat is good. Flat maps are boring. Slow lifts are bad. Long / frequent cutscenes are bad. Having to wait on a lift or train is bad. Centerprints get annoying quickly. Running around endlessly without completing a challenge is bad.

Quads are fun. Blowing stuff up is fun. Ambushes are fun, if used sparsely.

If you use experimental things that require the player to think, slow down the game etc., consider providing an alternative to them that's also somewhat challenging, but more straightforward.

Platforming can be fun, but failing a jump shouldn't be punished too hard, although it depends on skill level.

Make it easy to get around the map; wind tunnels help to get from A to B quickly.

Returning to an area from a different angle or on a different height is always nice.

It's good to have two or three things in a map that just look interesting, provide an unusual type of challenge etc, just to give it character. Think in terms of hotspots.

Too many switches or keys are bad. Only use them if it makes sense, don't go overboard. Use a unified look for your switches or buttons.

Traps are totally underused. Combine traps (like nail shooters with crushers). Combine traps and platforming (jumping puzzle with crushers). Use monsters as part of traps - a flak ogre can be a great nail shooter replacement.

Ladders are a nice way to traverse the Z axis if you don't have room for a full staircase. Speaking of stairs, reuse them as well. Save yourself the effort of building tons of different stairs. Build one nice set and use it in different configurations.

In general, players will notice some things more than others. You can get away with reusing smaller stuff. It will actually make the level seem more cohesive.

For lighting, I usually stick to one main type of light per area. I like using sky windows. I use a bright sun but a dim sunlight3 with AguirRe's tool. I use minlight during development, but eventually switch it off and rather place more light sources in parts that require them. I stick to a maximum light value of 250 (it gets clamped anyway) and rather use .wait or .delay 5 if I need certain lights to reach further. I only do this on the main lights though, but there are exceptions. Cheat if it looks better.

If something is weak or superfluous, have the courage to cut it. If you can get away with less, consider doing so.

Present simple stuff (door opening etc) as a spectacle. Make them work for it. 
 
I realized that it is important to have a central challenge in a level

This is the key I believe - an environment is always better if there is a logical theme - Rubicon 2 is a perfect example - it is a giant hydro-electric power station of some sort. This gives a sense of purpose to the world. So to plan the level with what the player is going to do whilst there in mind is even better. 
 
What Gb says above is hard to default - that's a pretty German-efficient method for building a level, and its hard to find fault with.

I'll play devils advocate and also suggest that building a series of itneresting interconnected areas will get you a decent map, but not the best one possible.

Typically when building a level or sketching it out a hundred ideas will occur of stuff that could be cool or memorable.

The tricky task (and one that seems to bog lots of mappers down) is choosing the strongest few and basing the map around them.

So whilst you can build a functional world, you also need to capture the player's imagination, get them focused and immersed in what you're presenting.

I find this is best done in two ways -

1. Never leave them unsure of what they're doing.
2. Make what they're doing part of the theme, and fun to do.

The first point is pretty straight-forward. It doesn't mean don't allow them to never get lost (although typically the biggest turn off in any map is getting lost) but rather don't allow them to wonder 'now what?' because immediately they're wondering how to finish your level and get it over with.

Better that they wonder 'ok, I killed the dragon and it smashed into the cliff up there, and I need its heart to unlock the gate... how do I get up that cliff?'.

This means you've clearly presented things in an idiot proof manner - players tend to do exactly the opposite of what you'd expect, so you need to plan the flow of the level carefully.

In the dragons heart example this would mean at the very least showing them the dragons heart lock / door before they fight the dragon. For best results they'd see the dragon fly overhead throughout the level a few times, before fighting it.

Which ties into the second point. You want the player to get involved in playing your level and forget that they're hunched over a screen moving a mouse around.

There are many methods for this - storytelling, atmosphere and creating a rich world to explore. All of these (and I'm sure the list is much bigger than those three) boil down to the level design holy grail - immersion.

You want the player so immersed that when they finish, they're bursting for a pee that they've been holding onto for the past twenty minutes.

Once they've gone to the bog they should then be thinking 'I wonder what would have happened if I got all the secrets?' or similar.

As you might have guessed, I always like to have a story with my level design. Typically this story has no direct effect on the level, but it helps me a lot to build it and lets me create areas with purpose - not necessarily 'and this is the castle kitchen' but more 'and here is where the forgetful ogre cook dropped the key in the privy'.

These micro-stories are fun for you to build around and even if the player never catches on, they're still aware (maybe subconsciously) that something is leading them onwards. 
Wow, Syntax And Grammar, And Spelling 
Also, I tend not to sketch levels, which is a bad habit.

This comes from my maps for the past few years either being speedmaps or based off id1 layouts.

My other maps are usually ones that I work on after another designer has got the ball rolling.

One thing I do do, and seems to work very well for me, is keep a pad nearby when I play test my levels, and jot down every little thing that bothers me about it, or every little idea that occurs.

It avoids breaking the flow by alt-tabbing to a text file and means I get closer to whatever vision for the map lies just around the corner. 
 
The basic idea for the whole map I keep in my head. I've always wanted what I had planned initially in this form and appeared at the end of creation. But it never worked.
Say, the appetite comes with eating. Same for the mapping. Planned to do one thing, but when you start to build, I always get a lot of thought and begin to deviate from the original plan.
On paper, the overall plan will never happen. All in my head. But if I'm away from the computer, I can jot down on paper some individual rooms.
Another feature I never in a hurry. If I need to think about a spot on the map, but I do not know yet how it will look like, then I stop doing. Or make a pause or switch to a different part of the map.
Coherence map can be and is, in architectural terms, but I can not think of a scenario some idea. 
Hahah, Very True 
Any creative process is a process - it has no clear end.

Apart from maybe "it's good enough, release it".

It's what I was trying to say before with the ideas jotted down while playtesting. Though my designs tend to fall short of what proper level designers can accomplish. 
 
in quake mapping i never make layouts on paper. i start from planning the main idea of the map and the general layout in my mind, since i don't think that a complex 3d layout can be made on 2d paper list. but i can do a paper sketching for some certain areas.
but if it goes to mapping for realistic world games (like half-life2 or anything) the paper plan for building interiors is a must for me. 
 
One thing I do do, and seems to work very well for me, is keep a pad nearby when I play test my levels, and jot down every little thing that bothers me about it, or every little idea that occurs.

Same here, scribbled notes help me keep track of what I need to change. 
What I'd Like To Think I'd Do 
if I ever make some single player experience is plan it as a succession of thoughts and feelings I want the player to go through and then build the world around that.

Then you could also measure how well you do by interviewing your testers. 
Thats 
Kind of my approach, but I think there is a set structure for it. Especially when in a game like Quake there are very specific emotions you're wanting to provoke.

A lot of it has to do with the flow of a level. A high is not a high when its swamped in highs.

On the other hand, Quake is a game about running around lightning fast and blowing the shit out of monsters.

I suppose its a balance.

For such an emotional design concept I think you'd have to have a big team of testers, but I don't think it'd need custom progs if done right. 
 
I think as feelings are involved variety is a key point. Combat is the central aspect but not all combat is the same, so it's about pacing the player and taking them through an ebb and flow of building up to a bigger fight, and then allowing them to relax again afterward. This is certainly the main focus of my base map :E

Usually points out a bad map for me is one where the combat remains of the same form and same intensity all the way through. 
True 
But I'd say thats a mediocre map. A bad map is one where you get lost and don't know what you're doing.

Or worse, you do it, then wander around because the exit wasn't compelling. 
Awesome Post, Gb 
and ijed, I think these ideas will help me, and others :-)

When mapping I get the most immediate gratification from building little "nibbles" of gameplay; maybe 1-2 rooms with 4 or 5 monsters. Usually starting with browsing through the id1 textures and finding something inspiring, and then building on to it until I have something that looks cool. This is fun, but I struggle with making interesting gameplay and have trouble building full maps...

Right now I'm working on a small map pack, and trying to do a better job of top-down planning so that the maps make sense and have better gameplay and progression 
Cat Out Of The Bag Time 
Something I've noticed, both through playing episode 4, and remaking one of the maps, is that episode 4 has a lot of flat or fairly flat maps. There aren't many places that you can go over/under. I do wonder if this has something to do with Sandy Petersen being a Doom mapper before being a Quake mapper...

For both the levels that I've worked on since joining the RMQ team (the aforementioned ep4 map and a chainmap with gb) is that sketching out the levels can often mean you end up with a flat, Doom-like map. Yes, you can have height variation. But there's very little over/under unless you actually put some in. Which can be difficult to visualize on a piece of paper.

Having multiple height levels running throughout your map not only means you can do a lot more revisiting, but it also allows you to get to different parts of the map a lot easier. There's less travel time, especially if you provide a way of getting up faster (down is usually less of an issue).

that's a pretty German-efficient method for building a level

Vorsprung durch Technik as they say on Trac... 
 
Something I've noticed, both through playing episode 4, and remaking one of the maps, is that episode 4 has a lot of flat or fairly flat maps. There aren't many places that you can go over/under. I do wonder if this has something to do with Sandy Petersen being a Doom mapper before being a Quake mapper...

Nail on the head I think. On the other hand e3m7 is one of the most vertically designed maps in ID1 and most of the time you don't even notice it. 
 
Ha, I saw e3m7, decided to launch Quake to see what map it is then already planned to reply to say how much I love The Haunted Halls with its vertical gameplay (and creme de la creme metal looks). 
 
See I'm a fan of verticality when it's about creating interesting combat scenarios, but since I like fairly tight control over player position and progression, creating vast open vertical towers of stuff isn't that inspiring for me.

I should perhaps give it a go though, I'll try making an Oil Rig style level :E 
 
vast open vertical towers or generally big open areas are boring and broken in Quake so thanks god you don't. I don't think anyone meant that either. Quake was designed as a tight and close-up game so obviously that's what works best (and to me by that is most fun). Still you can have a lot of vertical elements in small space. 
 
Ceremonial circles is pretty tight, but has lots vertical space.

Playing it in a mod that has Z-Aware Ogres is great. 
DM4 
is also a good example. 
 
But there's very little over/under unless you actually put some in. Which can be difficult to visualize on a piece of paper.

True, this is why I usually make no sketches.

It can also be good to just lay down a floor plan, then give your map to someone else who creates the height variation (decides where you should have stairs etc) and maybe adds a second level / walkway that winds around the map, connected by lifts for example. This will give you a pretty good amount of verticality often (and you can place sniping monsters up there).

It's also sometimes good to put different parts of the map at a completely different height level.

DOOM btw has a lot of height variation in terms of ramps, stairs and lifts, just no room over room.

Actually room over room is rarely required in Quake either, except where you have a lift or something similar. It's a big step to have a second level inside the same room - platforms with ogres on them, catwalk crossing the room etc.

Many commercial shooters are very flat these days. At most you'll see two levels in the same room, which is "enough" - but sometimes more can be nice.

It often helps to just cut out doorways higher up the wall and connect them. This allows you to reuse the area, as well. 
Ceremonial Circles 
is one of my favourite maps. 
Vertical Designs On Paper.... 
how about drawing a side view? 
My Idea Is... 
... each time to do something that has never been done... else, it is yet another Quake map...
Well, I do not have the talent of necros (as example), and I am a poor qc coder, but I am trying to innovate in each map in each part (topic, sounds, texture, etc...)

So basically, I try to put the rough architecture layout on the paper, as well as main player route (i.e SK/GK Doors, etc...) and the final fight must be epic as far as possible...
Once done, it clarifies what needs to be done, and then I start to map...
... and if I am lucky 6 month later I release my stuff.. but I guess nowadays, it will be longer, unless if I am not aiming a giant map ;) 
Thinking Sideways 
how about drawing a side view?

I find that harder to visualize than a top view. Also, there's more going on sideways; there's usually at least one "sideways" you can't show with a side view... 
 
there's usually at least one "sideways" you can't show with a side view.

So draw two side views... In most technical drawings, there's generally 2 or 3 views per layer that are
- top view
- front view
- side view

with all this you can have a good idea of what will be the 3D rendering of your stuff..

http://en.wikipedia.org/wiki/Technical_drawing#Types_of_technical_drawings for further details ;) 
Seems Like A Lot Of Work To Draw 3 Views 
I mean, you might as well just make a rudimentary version in editor. 
 
ya at that point if I still wanted to draw it I'd just do a perspective sketch :E 
 
same here, I'd rather do a perspective drawing than a side view.

It's faster to just make the stuff in the editor though.

Top down view is the one that gives the best overall view of the layout. 
 
top down is best for a horizontal layout. I think what metslime was driving at was that a vertical map or section of works better drawn that way.

As for doing more comlex stuff, i think a rough perspective would be most useful, and easy to do. 
 
I've got to try perspective sketches sometime. 
Rawr 
I just tend to go straight into the editor after looking at some reference shots of the kind of thing I want to make. I usually have a pretty good sense of the visual aesthetic already but things like layouts I just tend to build them as they come to mind.

Sometimes this means I have to rip up and rebuild areas as it becomes clear that they won't work, but I've never really been good at planning maps ahead of time so I just accept that it's going to happen :)

I also tried something new with my l4d2 map, I basicly finished the entire map before I even posted any screenshots online of it, as a trap I've fallen into many times before is making a really cool section of a map, getting excited about it and uploading a bunch of shots only to never work on it again. I found with the l4d2 map I had more energy to finish it the further in I got without actually showing anyone what I had made so far. Not sure why this was the case but it worked great for me! 
 
It's faster to just make the stuff in the editor though.

This, for me, 100%. If you just keep it rough blocks and ignore any aesthetics or detailing, you can easily block something out from scratch, and also experiment with different connections and verticality, and be able to see how it feels to move through, much more clearly than via any sort of drawing or 2d plan. 
Post A Reply:
Name:
Title:
Body:
message
question
exclamation
idea
flame
noflame
error
skull
beer
moon
pent
rocket
sheep
pacman
pig
cheese
worldcraft
gauntlet
crate
pitfall
pimp
smile
cool
sad
frown
oi
yay
tongue
evil
wink
neutral
q1
q2
q3
ut
hl
cs
doom
dkt
serious
cube
Website copyright © 2002-2017 John Fitzgibbons. All posts are copyright their respective authors.