Mapping Philosophy

[WJohnston]

So mapping, I think we can all agree, is a giant convoluted pain in the ass to figure out.

Oh, not the actual editor. Aside from a few traps and weird bits of nonsense to keep in mind, it's very easy to map.

The problem is actually figuring out a design that people like. This comes in 3 factors: Functionality (efficiency or lack of), aesthetics (visuals, style, theme), and gameplay (pulling the player into chaos, giving weak spots for traitors to exploit). Understanding how departments should be connected (or if they should be), how to allow traitors to make good use of bombs or being seen (the 15x15 tile vision mechanic, where bombs would do the most damage), and how the map should feel thematically for immersion is extremely important. The issue here is that these points aren't defined anywhere and making a map basically amounts to taking a stab in the dark with regards to all of these and hoping that players like it (and we all know how successful THAT method has been).



So I'd like to give out my thoughts on these points, and then invite you guys (mappers and players alike) to share your own ideas.

1. Function
A map should be capable of fulfilling its jobs and doing it well. This means things like hydroponics bordering the kitchen, which borders the bar, in order to supply them with materials. These two departments are essentially inseparable. Chemistry and R&D and robotics should all be connected to share resources, mining should be close by to provide minerals and so on.

Is what I would like to say.

The fact of the matter is, this just isn't what people want. At least, not to the levels I've been doing it. Efficiency station 1-3, despite having some very questionable decisions with their overly large maintenance and the 1st one's brig, had a consistent theme: That is, all departmental connections surrounded mining and eachother, such that material transfer would be very easy between them all. At the time I viewed departmental connections to be a bad idea, so I mitigated it. This turns out to have been a bad idea, as when people actually DID perform their jobs there rather than exploring maintenance, they didn't enjoy doing it as much as on box. I thought to myself and asked others "why is that?" and came with a simple conclusion. Jobs actually don't have very much to offer when you've done the same thing a thousand times. Outside of something like security, xenobio, and hydroponics your job simply isn't an engaging thing. The chaos created by the round type is. Where does chaos happen? Outside of your workspace, usually in the main hallways. What's the easiest way to get you to leave your department? Make the station inefficient as shit and have you walk everywhere, talk to people, and get in dangerous situations.

This ties into the other two mechanics:

2. Gameplay
A map should allow traitors to be able to break into departments, it should take into account our game mechanics like line of sight, AI cameras, lighting, and our vision range all for security coverage reasons (or for others functional reasons like cryo and genetics). A lot of antagonists rely on what's called traffic flow: How frequently people visit an area based on its importance or the importance of its surroundings. You ever notice that nobody visits the doors just outside of virology? How the permabrig and gulag halls are always deserted? How the entire west hallway with arrivals and the dorms tend to be completely empty? This is where traitors come in to do their business, and this is where functional mechanics directly tie into this. Where the points of interest are in a department (for medbay in order of importance: lobby, chemistry, cryo, genetics) means where people will tend to move to, and this directly influences how a traitor can both play stealthy (requires near or complete lack of people) and go on a rampage (requires a steady stream, but NOT too many).

Currently, on metastation, box, and all efficiency stations, there's fairly high efficiency across all of them. Points of interest in departments such as chemistry, robotics, research, hydro and others all border the main hallway for convenience. This also unfortunately puts them extremely close to the front and means that there's very little to work with in the back that people actually want to visit. This is good for stealth traitors, but bunches up everybody in the same area and makes it nearly impossible for a rampaging traitor to last for more than a few seconds before the 3 security officers that are also in the pile pull tasers on him.

I was wondering then how this would change if the main points of interest were both very far in the back of departments and also very far from eachother. How would this change? Stealth traitor, as a first, would suffer. Since people would be moving around back and forth and effectively patrolling areas, this would not work as well. When you really think about it though stealth traitor is just contributing to the whole "boring, unchanging station" situation. If changelings are despised because their whole mechanic revolves around silently taking people out of the round, and nuke ops are loved for their extreme loudness that involves everyone, wouldn't it be more interesting to do away with stealth traitors? On the other hand, rampaging traitors would have an easier time in this type of design, as you would run into people one after another rather than into a crowd of 20 that ends your rampage. Furthermore, if you're not a traitor, you're the one who will be walking around all the time and getting dragged into these rampages (and therefore given something to actually DO that isn't your job). I think ultimately this may be a positive move, were this type of map be made.

3. Aesthetics
This one's a rather arbitrary point, but I'd like to go over what I feel would make a map look best. A map to me that looks good is one that emphasizes the fact that you're in space, and one that emphasizes how shitty your surroundings and your employers are. A lot of windows to space not only helps with it feeling more removed from just an office on earth, but it ties in nicely with the second point: Lighting. Lights, as they are right now, are entirely too common and too bright. A single light tube illuminates up to 8 tiles (and a few of these is all that's needed for an entire hallway to be fully lit). I think that making significantly more use of light bulbs and light tubes would greatly help making things feel run down rather than clean and in order. Finally, I'd also like to point out that I feel that too much detail actually detracts from the station's aesthetic quality. I don't want there to be something on literally every single wall and every single tile, whether that is a colored tile or the third air alarm in the room or whatever. I think box has this down very well, and that metastation is a bit overkill in this.



That's my thoughts on map design. If I try making a new map I think I'll try out this gimmick and see how people like it, if it even gets a chance at being played.

I'd like for this to be an open discussion, both mappers and experienced players should express their own ideas on how maps should be built and why.

[lumipharon]

A few points to go over.

There is a difference between a stealth antag, and an antag that doesn't do anything.

A stealth antag does shit without attracting too much attention or get caught. This is interesting and intelligent.
Lings and some traitors don't do anything that could get them caught in the first place. This is boring and shitty.

And in terms of distributing traffic, you have to look at functionality pretty seriously.

Perma doesn't get much traffic because it's a dead end corridor for dumping shitter and antags instead of/or killing them. There is no reason to need or want to mke this a high traffic area.

Stuff like chemistry and whatever having desks accessible from the public is because they're designed to service the public in the first place, not jsut exclusively their own department.
If people had to get into medbay to access chemistry desk, 50% of people wouldn't get healed, or would have to get cryo'd. That is not good game design, that's just frustrating for the players.

Keeping with medbay, sleepers, cryo and cloning are all close to each other because they need to be. (except sleepers, they're literally useless for anything except morphine OD'ing people until they get upgraded)
If you were having to drag people all across medbay just to heal them, that would be a huge time waster and again, basically just be unfun for the sake of 'making people walk around more'.


The big thing with box station, is that service, medbay, science and escape are all along one corridor, which is the majority of the crew.
Now I'm not suggesting it's an issue, or it should be changed, but if you wanted to change traffic flow, on a new map you could easily seperate service from med/sci (which are connected because genetics).

[WJohnston]

Yeah, I'm definitely not going to be happy if people are looking for some arbitrary middle ground. I think that trying out a map that goes to the other extreme (incredibly inefficient) to force movement where they can get dragged into traitor shenanigans would be interesting in its own way. A middle ground is even less defined than maximum efficiency or no efficiency at all, and that's worth a whole discussion by itself.

We can say with certainty that extreme efficiency is definitely not good due to players sitting still and doing nothing but performing pretty shallow jobs.

We CAN'T say with certainty that the opposite is better or worse than that. I think it's worth trying out, if only to see what would happen and collect some data so a potential middle ground can finally be established.

[Reimoo]

Aesthetic wise I think it is a good idea to actually imagine what a room would look like in first person and use examples from real life when you're designing it. It helps identifying rooms and their purpose more easily.

Also I couldn't agree more on lighting. There really is an abundance of light on the station; I feel that appropriate lighting even in a 2D spaceman simulator goes a long way in setting mood for a room. An example would be how the desk lamp is the only light source in the interrogation.

[Celdur]

I don't think stealth is entirely related to traffic. A lot of traitors myself included do their sneaky stuff in very public areas, just takes timing really.
I mean, some objectives demand it. Getting the captains laser for example means you have to break in right in view of one of the most visited hallways, but that doesn't stop people from being able to do it stealthily.
And come to think of it, how many traitors really go to arrivals to do anything? Sure you could buy stuff from your PDA or hide stuff there, but there isn't much there for you to get so most of the time no one is there, not even traitors.

I think one of the biggest problems from box is that the entire population is on the east wing and the west wing is pretty much barren aside from a few cargo techs.
Spacing out departments across the station is the best way to get an evened out population across the station, and in that way you lower huge clusters of people in the same spot, making crazy wizard/esword rampages a bit more viable without being instantly overpowered.
And I don't think it would hurt stealth too much, because I'd say stealth is more about timing than anything else.
I mean, you don't wait for your target to go to arrivals to kill him, you wait till he's alone in some room in his department and then you take your best shot.
That said, departments that make sense together should still be somewhat nearby each other, such as med/sci and engineering/atmos.

And maint is also very important for all antag types. Having one form of access should be reserved for only a few rooms like perma and the AI core and whatnot.
There should always be wiggle room for antags to get around somewhat concealed, and have an alternative route for escape. But I think most people realize that pretty well.

[WJohnston]

This is the kind of thing mappers have to wrestle with when coming up with a design for a map.

Add in physical limitations on where you can fit things in and an absolute ton of details like how piping, wiring, mulebot movement and a ton of other stuff and you have yourself a monstrously complicated project.

I'm curious on the whole bar/kitchen/botany deal. It seems like such a good idea, but does that promote those jobs standing still for too long? Are the only things that could possibly draw you out already connected? Would it be better to split those up and throw them in some arbitrary place, if only for the (inefficiency?) map's gimmick?

[newfren]

There's basically nothing to draw a botanist out of botany after the round start anyway - you can get like 12 bottles of mutagen from friendly chemists pretty easily, and that can last you a whole round if you're not trying to perfect every plant - and odds are if you are trying to perfect every plant you're daisy holmes and instead you go to cargo and get them to print 10 large beakers for you to fill with mutagen at roundstart so you never have to leave anyway.

The bar doesn't strictly need anything at all to be perfectly functional - the vast majority of drinks you can mix from your two dispensers easily. If you make the effort at roundstart to go to chemistry then they can provide you with chemicals for an extra 6 drinks that noone ever orders anyway, and mining can get you uranium for drinks that are only ever ordered by miners who are bringing a bar of uranium to you anyway. Botany then gives you all the various juices as well as reducing your reliance on mining/chemistry for other things - but again, noone ever orders any of the drinks botany allows you to make. If you did want to get the bartender to move around more then he'd definitely need to be separate from botany/a little further away from medbay, but it still wouldn't massively increase his movement in a round.

As for the chef, your round fucking sucks without botany, and with botany you never have to move anywhere for anything. If the chef and the botanist did start seperated, then the chef would have to spawn with plant bags somewhere in his kitchen for easy transportation of all the shit he needs to actually do his job. I mean if you want to increase the amount he has to move then tediously lugging like 7 plants each trip will make him have to move a lot and will also make the job absolutely terrible to play.

[Not-Dorsidarf]

When discussing mapping, it's important to keep in mind that camera coverage doesn't have to be universal.

[onleavedontatme]

Talking about traffic and stealth and whatever is completely pointless as long as we keep trying to design a game for 10-90 players. It's just too much variance.

When discussing mapping, it's important to keep in mind that camera coverage doesn't have to be universal.

Impossible. Are you sure?

Anyway remove main halls, replace with bridges and maint 2015. Just crazy enough to work.

[invisty]

WJ you and some interested mappers need to see about making a modular station that is populated based on the number of players in the previous round, or somehow re-arranges itself at start-30s based on the number of readied players. The modular station concept would be a great starting point. Alternatively, one or two smaller, simplified maps like some of the ships Bay has. I think it would be worth investigating by creating a test mini-station to see what the impacts are on tgstation's game modes.

Also, audio cues/audio range needs to be considered.

Kor makes a very interesting idea out of removing hallways. Worth exploring!

[WJohnston]

Celdur is literally making a modular station. I think it'd be interesting to see how it works, but I'm fairly certain it's against the whole "inefficient" design of a map I've been thinking of.

Removing hallways is a neat idea but we have to be realistic here: You have to throw in the limitations of mapping in that you can only cram so much stuff in so much space. No I mean really you can't have one department bleeding into another into another into another, let alone have that be visually appealing, gameplay-feasible (everything gets hit by a single bomb), or even functional in the basest sense.

Well, at least not with how our current access setup is.

You'd need drastic access changes, expecting almost everyone to be able to walk through command if it stays in the center, and security being only accessible through the bridge or some horrifying thing.

[invisty]

I agree there are certainly issues around limitation of hard-coded ID accesses. Surely a new map would qualify for map-based configuration of IDs, though?

I'm going to nerd out here for a bit and get all engineering-like on this map design stuff, so this is less philosophical than your thread might have wanted:

In thinking about map layout, the high-level station map can be represented by a force-directed graph, where the nodes represent resources/areas, and edges represent relationships between areas - most typically, the interactions of departments to one-another. Edges that pull two nodes together can be used to represent desired levels of efficiency or close transit distances, and repelling edges could represent desired isolation of an area from another area, or inefficiency. Arrivals might want to push away from escape, and hydroponics would want to pull towards the kitchen. Additionally, nodes can repel other nodes based on how much "space" they use, or a generally desired isolation (the singularity engine). From this, you can rearrange the positioning of nodes, crossing and uncrossing lines. If there were crossed lines on your graph, you could replace these with a hallway linking the crossed nodes. Access/transit redundancy could also be created by introducing loops in your all-access areas (like how you can go the other way around the central primary hallway if you find it's been bombed). You could also create another variable that identifies nodes as having a certain "access type", and then create a rule that says "no area (node) should be more than two similar access types away from an all-access hallway" (to represent concepts such as all access > department access > head access). Following this, you could simplify a map for low-pop by eliminating the less necessary areas (viro/xeno/prison) from the graph and improving the ID permissions of crew to broaden the scope of their job. You could also introduce more abstract edge relations to represent concepts like "An AI shouldn't be able to quickly flick their view over these three areas in succession" or "These areas should not directly adjacent to an all-access area" or a repelling node that defines "this area should be distant from security". In theory, you could go so far as to generate an entire station layout using this approach, defining how many tiles each area uses up, along with a minimum and maximum area, size and height, maintenance connections, so on and so forth.

It's all ultimately an engineering application of the concepts you're already familiar with, but takes out all the leg-work of having to think about everything during design. By abstracting the reasoning process of map layout, this could also provide answers to the more vague mapping problems, even some which mappers might not have considered. If you think it would be valuable, I could implement a force directed graph library to do this.

[Celdur]

How would a ever changing modular station even work? What would happen to the people that are in a place that gets swapped out for an upgrade?
Or maybe you just give small populations less stuff to do? Sorry, less than 30 people here, we won't get xenobio until we do, so just do R&D I guess.
I dunno, it's a neat idea but it's not very practical.

Each map just kinda comes with their own strengths and weaknesses. Even though there are obviously badly designed maps, I'd like to see people more willing to try to appreciate different maps(that aren't obviously bad) for what they are, instead of complaining about how they aren't like box.
I mean, servers could have map rotations right? I would enjoy that variety, but I'm sure everyone would complain about how they just want box all the time.

[allura]

wj wanted me to post here so heres what i have to say about my experience mapping, because i find it difficult to say i have a mapping "philosophy"

mapping is one of the most annoyingly frustrating and retarded things to do for any codebase ever. unless you are so severely autistic that every aspect of the process of mapping appeal to you, then you will hate it for a huge reason, and that's player to mapper interaction.
somehow every single player thinks that their opinion goes over that of a mapper, simply because they play. that is absolutely fucking stupid. if players were objectively right about mapping and mappers were obligated to listen to their babbling, we wouldnt get shitty comments like "camera coverage doesnt have to be universal." uh, yes it does. are you fucking stupid? additionally, does an architect have to live in the houses he designs for a year before he can sell them, otherwise its a shitty house?
reading comments in mapping threads is page after page of half-baked, retarded suggestions that are either "MAKE IT MORE LIKE BOX" or "THIS IS AN IDEA I JUST THOUGHT UP AND COMPLETELY IGNORED THE CONSEQUENCES OF, DO IT THOUGH." fuck you. jesus christ, the mapper's job isnt to make a station people want, its to make a station people like.
even worse can be mapper to mapper interaction. this isnt me being completely right, because its a matter of opinion, but when you get invested in mapping enough its hard to see a good map change. all you can think about is how retarded of a change that was and how shitty the mapper is.
i might post more later, but i need to go.

[mosquitoman]

we wouldnt get shitty comments like "camera coverage doesnt have to be universal." uh, yes it does. are you fucking stupid?

thanks for this exhaustive argument

[allura]

sorry, i assumed that the readers had the brain power to figure it out themselves, but sure. i'll explain for your sadbrain.

the ai has absolutely no form of movement, can have its only form of communication destroyed easily (one rwall into comms) and already cannot see a massive part of the station - maintenance. when making a map, if you deprive the ai of its own fucking eyes when thats literally the biggest feature of ai - seeing the whole station so you can help it - you might as well be saying "remember, scientists dont HAVE to have access to all of science." its absolutely batshit and it makes one of the least enjoyable roles even worse.

[Not-Dorsidarf]

The AI has a built-in intercomm and can always speak and always be heard via station intercomms

[Ricotez]

The AI has a built-in intercomm and can always speak and always be heard via station intercomms

Actually, it's a built-in headset with all channels that runs on Telecomms. The intercoms are on the walls around its core, and are station-bounced.

[Scones]

i still want to see a map that has no main hallways and instead is just departments surrounded by maint with some lit passages between them

[onleavedontatme]

i still want to see a map that has no main hallways and instead is just departments surrounded by maint with some lit passages between them

Space hulk station when

[Not-Dorsidarf]

Dee Station roars "GWWAAAAAAAAAAAARRRRRGH!!!"
[common] Lola Blabbleton exclaims "Halp station is space hulk!"