Methods of planning maps

[NoodleCollie]

I was wondering this evening what kinds of methods you guys like to use when planning out maps; I know some people like do a couple of rough sketches whereas others will fill out a complete blueprint before even touching Hammer, and still others will jump into the editor with reckless abandon.

I have a diagram theme that I first invented while mapping for 007 Nightfire and Half-Life 2 that revolves around squared grid paper. I wanted to be able to get a concrete sense of measurements for maps which you don't get from freehand sketches - if you plan areas according to the paper grid, you know exactly how many units in Hammer to make a brush and are able to draw out new areas by comparing the scale to that of pre-existing areas.

The following is basically what my method is made out of:

One side of a square - this can be 64 or 32 units, depending on how much space you're anticipating to take up on the page. I used to use an 8mm grid but have since switched to graph paper to be able to fit more on a page without having to sellotape multiple sheets together (and then you start to get problems with drawing on the top of tape...).

Stairs - consist of their physical area filled with vertical or horizontal lines depending on the direction of the stairs. An arrow is also present along with with the number of units these stairs ascend/descend. The arrow always points up the stairs. This method of marking stairs implicitly connects two different levels on the page, so if you have lots of stairs (particularly where they're not near each other) it can be easy to lose track of the z-level of an area in relation to the one it's connecting. You also have to calculate how long a set of stairs should be in relation to the height it ascends - I usually use the standard 2:3 ratio of rise:run.

An example room - walls are simply lines, with the brush volumes implied behind them. the physical arrangement of the brushes needed to make the room are not considered here. Corridors can be made by leaving gaps in the room walls, or by using the style below. The height of walls is not considered here, and is only implied in relation to other geometry the room may be near or part of.

Lines perpendicular to a wall specify a doorway. This can be used interchangeably with simply leaving a gap in the wall but is useful if you want to insert a corridor into a wall that you've already drawn.

Drop down - a curved arrow over a pre-existing line implies that a ledge or platform of some kind exists. The difference on the Z axis is specified. The arrow always points from the higher geometry to the lower.

Window - a thin rectangle whose length is the length of the window. Window height or distance from floor is not considered here.

Gate/fence - Some form of vertical grated plane. The height is not considered here.

Ramp - similar to stairs, the arrow points up the ramp and the distance ascended is specified. Ramps are more reliable to be able to draw that stairs, since their rise/run doesn't need to be considered as much (though care should be taken to ensure the ramp isn't too steep to walk up).

Control point - getting the dimensions right on the control point makes it easier to estimate the scale of the geometry as you draw it. I almost always start planning by drawing a control point and working outwards.

Payload track - follows lines, can curve. Isn't always easy to build a track that follows the dimensions of the track pieces in TF2 exactly.

Floor grating - specifies some sort of grating that is not vertical.

Bridge/tunnel - implies that the areas the two ends of the arrow connect are accessible to each other underneath the other geometry.

Health/ammo - H/A specifies the type of pickup, s/m/l specifies its size.

An example VSH map I planned out a short while ago:

Note that this uses other arbitrary symbols like the "S" (standing for "Spongify" - this was going to utilise the textures and some of the mechanics from the old Harry Potter games). Also note that there's a big attic area that I wasn't able to draw here because it would be drawn completely over all the stuff I've already put.

Advantages of this schema:
- Simple and easy to sketch out accurate blueprints for maps.
- Measurements can be known if a scale is decided on.
- Working with a grid helps to learn the scale of a map when creating.
- Irrelevant details don't have to be decided on until creation in Hammer.

Disadvantages:
- There is a tendency to make areas too small (I think this is because you're focusing on an entire map on paper, so it's easy to "zoom out" in your head and think a small enough area is adequately sized).
- Isn't very easy to implement height differences, or draw large areas above/below a previously drawn area as you end up drawing over the top of the diagram and things get confused quickly.
- Easy to lose track of relative heights of areas.
- In certain cases symbols can be unclear (for example, if a doorway is put into a wall and the width of the corridor it connects onto is the same width as the doorway, two of the "prongs" of the doorway symbol are drawn on top of the wall lines of the corridor).
- Drawing large maps can get very unwieldy, as can looking at the pieces of paper when you go and transfer into Hammer - I once ended up with a load of taped-together 8mm grid sheets that covered an entire dining table.

 

[Fruity Snacks]

I usually just do semi-emergent design for my levels... I did just discover SketchUp though, and may be using that more in the future. It allows me to import models and also basically build brushes piece by piece quickly so I can get an idea of the map layout. Also, let's me see the overhead layout relatively quickly and without the need to compile.

 

[Crash]

I usually do a handful of top down sloppy-sketches on scraps of paper and then reference those when I start a new design up. Usually they don't last long in the process and I end up diverging from them quick, but they help me get a basic idea of what I want to do with a map and I usually I drawn in some core concepts and "hooks" that end up going into the final product.

I don't use grid paper, because it makes me design in squares rather than more interesting shapes.

 

[Fantaboi]

I just do some really simple sketches. After I get a number of sketches, I decide the best parts of each map and create a final sketch. I then create a page of notes that I want to include in the map.
After I get an early build of the map, I take a top-down photo and use paint to manipulate parts of it (adding lines and stuff) so I can see how it would change the map.

For those interested in my $2 budget design process here is koth_workstation's early sketch
http://i.imgur.com/6ca1aR4.png
Since the map was symmetrical I only drew half the map.

You guys have really cool design processes, I think I might use them in my next map.

 

[UKCS-Alias]

i generaly first need to have an idea and then i just draw a specifc area in paint.net using layers. Layers so i can easily add information at any part. Ofcourse, some of the drawings will look crappy at some point due to copy pasting areas or just moving them. And sometimes i either fade a layer when i notice i have to edit a part, but i can still see the original to make drawing slightly easier.
Some of my images from the skullcove design:

This used to be the overal map layout at first. However, the start felt boring and too much like decoy hence it needed a change:

Thats the only drawings of skullcove that realy mattered.

Yes, the text is dutch since im dutch and did the planning with a belgium player which made the initial alpha shape the first months (although that went very slow but wasnt a problem since it was still just planning).

 

[Ælement]

I start with a good idea, and then i simply build a bit onto it, check if it seems to work together, then i expand a little further, check if it's working once again.

I repeat that till i have a complete layout.

Lets look at sludge:

I started out with the idea for the spawn area; a glass facade, a ledge overlooking the hatch, which is placed in front of the glass facade.

Then, i defined the plateau, that the hatch is located on. I also made another exit from spawn.

Now i started splitting stuff up a bit. One route going below the plateau, while another one runs along on level with it. I decided that it was time for a structure now, so that the visibility could be obstructed a bit.

I then came to the conclussion that it would be a good idea to make a structure by the platform overlooking the hatch as well, to sorta connect things a bit more. With these two structures being quite harshly seperated at the moment, i decided to bridge them.

Originally, i wanted the sunken path to only run for a short lenght, before regaining height, but with the two new structures, i decided to make it run much further.

Along with the decision to lenghten the sunken path, i decided to make a third structure to bridge the land on the other side of the path with the rest of the structures. I also filled the pathway with water to give it more reason for existing in the chosen environment.

Now, i wanted to find a way to connect the first structure with the ground without obstructing the tunnel to the watery area. I found this solution suitable (tho, it later proved itself a sightline-fest)

At last, i began building the bots drop in area. I wanted a forwarded upgrade shp that was clearly visible, without being exposed. After a lot of creating and deleting, i ended up with a simple building hidden behind some cover. I placed a building for the players to defend from, but wanted it to be relatively difficult to hold. Therefore, i made a staircase that the bots would easily be able to use.
A rock makes sure that spies and other ambushers can easily break the line of sight with the bots, and a tunnel makes for a lazy tank entry point

At this point, the map looked like this.

 

[Egan]

Before this becomes irrelevant, I found this quote online that might whet the appetite of some:

Think 3D - An effective 3D level cannot be laid out on graph paper. Paper maps may be a good starting point (though even that's under limited circumstances). A 3D game map must take into account things over the player's head and under the player's feet. If there's no need to look up and down - constantly - make a 2D game!

http://www.gamesindustry.biz/articles/2013-09-04-warren-spectors-commandments-of-game-design?utm_source=newsletter&utm_medium=email&utm_campaign=us-daily

 

[Grizzly Berry]

Before this becomes irrelevant, I found this quote online that might whet the appetite of some:

Think 3D - An effective 3D level cannot be laid out on graph paper. Paper maps may be a good starting point (though even that's under limited circumstances). A 3D game map must take into account things over the player's head and under the player's feet. If there's no need to look up and down - constantly - make a 2D game!

http://www.gamesindustry.biz/articl...n?utm_source=newsletter&utm_medium=email&utm_ campaign=us-daily

I like that. I've been planning on graph paper lately and I've noticed it makes me less likely to use angles and things like that. It's also not until I've actually mapped out in hammer I see where I should have used elements vertically as well.

 

[Prestige]

close your eyes and dream your map

 

[Toxindude]

I usually start with a single idea even just a room I dream up and go from there. I use props and spawn ent's for scale. I do sketch sometimes just to get the idea out of my head but you can only do so much with a 2D top down. I could go all day long building a single section of map in my head itching to get it out when I get to map again

 

[RaVaGe]

I have no idea how i'm doing it, but it works.

 

[Valkyrie]

It may sound odd but I have started planning maps in minecraft xD I use each block as 4 HU and with that I can easily see my areas take shape but in a very quick and rough manner. People have said "well surely thats slower than just planning it on hammer". Well I work slowly on hammer and can work really fast on minecraft, so it just works for me

 

[NoodleCollie]

That's a good idea actually. Voxels would make it very easy to change the shape of areas quickly.

 

[Valkyrie]

It really does help more than people give it credit for, yeah, if you are slow it can take time, but it does not have the draw backs of planning on paper, you can easily see your map in 3d

 

[UKCS-Alias]

The problem i many times have with building in 3d directly is that i get a boring shaped map. There normaly isnt alot of interesting parts in that (although that wouldnt be for everyone - for some it works better). Its ideal for connecting parts but thats about it hence i dont realy use them.

I first need an idea of a shape, one that has no ingame relations. all it needs to be is interesting in shape. This can be some area arround a tower, some indoor room, whatever side view of pathing. Anything can work, but i already have to think about what i want to create before actualy making it.
If i randomly would just edit a part it often comes down to mapping for the sake of mapping. And the creations that come from that are often more horrible than listening to porcelain of the red hot chili peppers (only cabron is worse - but for that i would have to carve a sphere into a sphere).

Especialy in ramparts it was visible which parts were actualy thought out and which were made by mapping for the sake of mapping.

Still, if you have an idea but are unsure of depths minecraft indeed can help alot.

 

[Tarry H Sruman]

Got me a white board and a pad of like 1x2' graph paper.

I sketch things on those and then go into Hammer and make things that dont resemble the sketches.

 

[LeSwordfish]

Okay, let's use sundown as an example.

First, I get an idea for something. With sundown, I wanted palm trees inside, and sun umbrellas and deckchairs outside. I also wanted a lot of cool curves, and pools. Thinking about pools gave me other ideas- namely using a pool as a reset area. That would probably mean sloping up to the top, and outting it opposite a door, so bots couldn't just climb out and there was a good place to airblast them in.

So. I had a checklist of things for the spawn area.

• Door, high up, opposite a pool.
• Cool curves
• The standard MVM spawn things- sight lines, a spine, an alternate route.

With those done, I began searching for images of period resorts, as inspiration. These helped solidify my ideas, including using planters as cover and having a central "bar" as the forward supply area. With those ideas done, I began to sketch.

This is my first sketch. From this, I began building it in hammer. Hammer building is the only way I can judge scale, and by the time I was nailing down a reasonable scale it was becoming apparent that some changes would need to be made, most notably the positions of the "bar" and the ramp. This generated a layout that kinda worked, I think.

((I shall put a picture of the map here when I can, on iPad right now.))

So with half the map built, I began to look at the other half. What had I done? What hadn't I done? Well, there were some nice big open areas, but limited smaller or indoor areas, so I added one of them, based around my palms-in-planters idea. Height advantages or disadvantages- I designed the planter room around having two levels, and added a staircase between them. Once again, scale became an issue, so I moved the staircase to the other side. For the final area, I once again included height variations, and I included an area designed for bots to snipe from. I needed another ramp down, for the tank, and so I twisted it around the edge of the room. Most MVM maps provide the players with a small height advantage over the hatch, so I added an extra gantry and stuck the spawns on that.

So yeah. TL, DR: I design cool spaces and then make the other areas fit around them.

 

[Bombcartography]

I used to jump in with reckless abandoned. But I always ended up painting myself into a corner. If I ever got close to alpha they couldn't be optimized or they were a nightmare to edit. So a while back I started studying the Valve maps intensely and noticed a few things.

The basic geometry was rather blocky. It seemed like the maps were built upon very basic blocky shapes. And the maps seemed generally designed around areas that could be sectioned off by area portals.

Also, the maps could be broken down into four basic areas: Rooms, Fields, Spawnrooms, and Set Pieces. The spawnrooms and the set pieces are basically just functional. Spawnrooms are just where the players start. And the setpieces divide areas and fill in spaces. The fields and rooms are the meat of the map where all of the gameplay happens.

With this analysis, I developed a strategy of first building a very basic blocky outline of the map I wanted to make and then build upon that.

The blue and white striped blocks are blue spawns. The red and white striped ones are red spawns. The red and black blocks are rooms. The maroon blocks are setpieces. The dark tan areas are fields.

It looks deceptively simple but it takes about 20-30 hours to build a block map like this. In the photo above, I see much more than just blocks. I've painstakingly arranged these blocks, analyzing how they will relate to each other in terms of game play. And I'm kind of seeing what they're going to look like in the end.

For me, it makes editing much easier. Psychologically the map just always feels extremely simple no matter how complex it may get.

 

[UKCS-Alias]

Another thing to note. Most areas are linear. the only moments they make a turn is to provide better optimizing and to keep the map compact.

Allmost all 5cp maps feel as a linear map. and they many times barely have any turns. If they have turns its the middle part between 2 open spaces. For 3 stage maps it makes a center stage that leads to a side, followed by making a circle arround the map. Even the popular 4cp payload maps already show that pattern (specificly upward and badwater show this clearly).

Even then, it still doesnt mean its good. The thing that makes a map good or breaks a map is down to bottlenecks only:
A map shouldnt have any strong bottleneck. Not even at the final pit. This is for example what dustbowl, goldrush and barnblitz show that bottlenecks are problematic on any pub (there is a reason why the final point rarely gets capped). Badwater although it provides bottlenecks doesnt realy suffer from it since the bottleneck is mostly controlled by blue. Which is exactly why it plays better. Red should never have a strong bottleneck. Its too easily outcamped.
Note that for competetive this isnt counting since those actualy use respawn times the way it should, most pubs these days have instant or reduced respawn times already. Combined with extra players 2 demoman are able to allmost fully lock down a path. And if they are near each other lock down both together.

When planning a map to me its most important to break down bottlenecks. Adding a side path often doesnt do it, the whole part needs a remake. And thats exactly the thing the Set Pieces Bombcartography mentions should do. And what many times goes wrong is that they end up as a bottleneck:
Keep them open as much as you can, look for example at foundry or badlands where those set pieces are somewhat equal in size of walkable space to the control points themselve. Only the doors are bottlenecks and they are usable equaly on both sides there.

 

[Fish 2.0]

Actually, I wonder if you could build a map in Minecraft, test it in Minecraft and have those results give an idea of the actual map gameplay.