next »

[melos]

8/24/12 update: feel free to give advice if you want, but I'm working on the 2nd part to the linked post, and have a few things in my head on how to improve.

I'm not sure where it's appropriate to post this, or if at all. I don't think it warrants its own topic in Design given the nature of the topic, and this forum seems inactive enough where it would be okay, BUT, Move or delete it if necessary. Anyways I've been trying to write  posts about things I do with design in my current game, and then sort of expanding it a bit to be of use to others who are reading.

I'm not a particularly good writer, I say this because I don't really know much about writing other than stuff done though school, and, well, everyday attempts at communication. I was wondering how I could make my design posts more clear, coherent, and whether the points I'm making are even at all useful or tangentially practical. The posts of course are too "technical" (w.r.t. design or whatever) that I think getting general advice from people who don't think much about design wouldn't be as useful as say, here.

Anyhow, personally I think this is my most decent attempt at a post so far - the other ones I think are sort of incoherent and poorly organized, with not very much effort...

http://seagaia.wordpress.com/2012/08/11/old-zelda-like-dungeon-design-in-anodyne-part-1-of/

Let me know what you think. Specifically:

-Is the connection from thought to thought clear? Or do I stumble off the path a bit too much?
-Does the post just seem self-masturbatory about my own game, or does it have lessons or tips that could be useful to other devs? If the former, I can try to avoid that more but of course advice on how to do that would be great.

[rofl_mao]

Definitely will check this out when I'm home. I skimmed it and for me it's interesting to see your thoughts when designing dungeons, so it was definitely helpful.

[Graham-]

I liked it. Your writing flows well, and it doesn't tangent or ramble. Also you aren't repetitive, you don't mix your concepts, and your content is solid.

Your post is not self-congratulatory. The only annoying post is one that derails from content to say nothing positively, about something. Having content, no matter what it is, and being clear about it is all that matters, whether it's about your game, or your penis or whatever.

I'll focus on form, instead of content, because that seems to be your question.

----

You have a lot of this (and by this I mean the abstract concept I'm about to introduce, and hopefully clarify) stuff, where you confuse (which is something that totally happens really easily if you're not careful, especially when writing) the reader. Instead (as-in replacement of) you can separate out your ideas into groups.

like so...

You have a lot of stuff where you confuse the reader by trying to explain things in the middle of sentences. If you have an idea it's better to isolate it. Brackets should only show up sometimes.

It's a tough idea to grasp conceptually, but it's very easy to understand with practice. Like-ideas should go together. You can tell how things are related using your instincts. If it comes out naturally then that is a good sign. If it doesn't, that's a bad sign. In those cases, pause, jot down the things you want to say, then write them out cleanly. Write a topic until you come to a natural stopping point, as determined by your intuition, then hit enter twice and do it again. Keep going until all your thoughts are on the page.

Think of it as picking directions then committing to them. It's easy to circle back. Readers stay engaged if you stay engaged. Structure beyond that isn't so important. Ideas can come out in any order.

It's a good idea to have paragraphs that are just dedicated to explaining what you are about to explain next. I hit your 3rd paragraph and I still don't know what you are about to explain to me, not without re-reading.

....

My rewrite (of the opening).

I'm going to talk about old dungeon designs in (old) Zeldas, because I find them interesting. We'll call all dungeons in Zelda games appearing before Oracles "old," and everything else irrelevent. I'm going to show you how brilliant these designs are, how they work, and how to steal their lustre for your own designs.

(1 paragraph to introduce what the article is about).

A Zelda dungeon is: awesome. For the sake of clarity, we'll also call it a series of interconnected rooms. The player starts in some room, A, and ends up in some room, B, after passing through many rooms in-between, in some order that is interesting. That is a Zelda dungeon.

(1 paragraph for an important definition/concept)

There are a couple of others things that need to be in the dungeon too, for us to classify it as a "Zelda" dungeon (of the kind we care about). I will show you these things now. Pay attention.
  1. You (the player) must have to defeat a number of enemies in your journey (from A-to-B). These enemies have to impede your progress in some way, necessitating your actions to destroy them i.e. a bat that tries to kill you.
  2. You must solve puzzles. We'll say that these puzzles are constructed from a series of "entities" that need to be manipulated in some way to produce a necessary result i.e. pushing a block to unlock a door.
  3. You must find things. These things must be necessary to overcomming other obstacles in the same dungeon i.e. a key that opens a locked door.

(1 paragraph to introduce what I am about to explain, followed by that explanation.)

That's enough for now. Your structure is similar through-out. If you want to improve at readability, do this:
  1. Write as you have done.
  2. Write it again, as I have done.
To do this you may want to jot down the "core" concepts in each section of your post as you go. Sometimes I write, then analyze later, writing down the key concepts, then return even later to put it into a form that flows. Sometimes the breaks are necessary to gather perspective.

If you do that enough it starts to become natural. I have to work hard at it too.

The best part about clarity is that it focuses you on your own ideas. You walk away with a much clearer understanding of what parts were really important. Then leveraging your insight becomes more natural. Communicating is always a good exercise for honing an idea.

That's why I'm replying. Whoa.

[melos]

ah, forgot to reply. Thanks, toast_trip, the advice is helpful. reminds me of what few math textbooks i would read - some definitions/axioms, then a theorem/lemma/corollary statement, then the proof, and so forth.

good idea with the brackets, too - the writing could always use more planning in terms of splitting apart ideas.

go communication!

[Graham-]

Good writing is all about being straight-forward, at least with what you want to express.

It's like you're always saying exactly what you want to say, and nothing else. It can help if you do some prep-work before-hand, that isolates the key concepts that you want to address.

Outlines can be like this:
  . most important point
  . other most important point

Then:
  . intro
  . point
     . best example
  . point
     . best example
  . recap

Repeat a bunch of times.

It's like you grow from a little tiny baby, that is the ultimate essence of what you want to communicate.

When you finally sit down to write then you can just write naturally, in the order the thoughts come to you, and you will:
  . be clear
  . be natural
  . communicate the most important things

It's like building a good game... with words.

[Blademasterbobo]

now if only you'd listen to your own advice

[ஒழுக்கின்மை (Paul Eres)]

i feel that good writing consists of two things

the first is that the clearer of a thinker you are, the better of a writer you are. most bad writers are also bad thinkers, whose brains are like a jumble of unrelated and unorganized thoughts, so when they write, they write down a jumble of unrelated and unorganized thoughts

the second is that good writing has to go through several revisions to become good; nobody writes good writing on their first draft, it takes three or more drafts before something becomes good. so in a very real sense good writing isn't good writing at all, it's good re-writing

as for the article itself, i too am making a zelda-like game, but because i don't have "dungeons" per se (except in the sense that the entire game is one huge interconnected dungeon, like SoTN), and because it seems as if your article's design philosophy deeply clashes with my own, it wasn't that interesting to me. i didn't really understand what you were trying to say, no. that may not be entirely your fault, though, because i also wasn't that interested in reading it (i'm sure that if i focused on reading it through a few times rather than just once i'd be able to figure it out)

[Graham-]

now if only you'd listen to your own advice

Yeah, I walked into that one.

My TIG posts are more like a stream-of-consciousness.

p.s. proper drafting is the most important skill.

[melos]

i feel that good writing consists of two things

the first is that the clearer of a thinker you are, the better of a writer you are. most bad writers are also bad thinkers, whose brains are like a jumble of unrelated and unorganized thoughts, so when they write, they write down a jumble of unrelated and unorganized thoughts

the second is that good writing has to go through several revisions to become good; nobody writes good writing on their first draft, it takes three or more drafts before something becomes good. so in a very real sense good writing isn't good writing at all, it's good re-writing

yeah, it's all about the iteration on drafts, like a lot of things, it's a matter of discipline in reiterating. i've noticed this on very small scale writings i've done outside of game stuff, guess I should force myself to do it more than once on these posts.

[ஒழுக்கின்மை (Paul Eres)]

okay i read through the article a second time. the main thing for me is i couldn't understand the point of your post, even though i understood individual parts of it. like, what is it supposed to accomplish? it doesn't seem to be fundamentally saying anything about dungeon design, it touches a lot of little things in little ways but has no main "thrust" to it

but let me try to walk through it

- you define old zelda-like dungeons. but you don't explain *why* the division exists: you don't mention what makes the ones you name different from modern zelda dungeons. is it merely that the old ones are in a "grid" and the new ones are not? or what? if so, why does that difference matter?

- some of the ways you phrase things make no sense to me. e.g. "concurrently implementing" -- why couldn't you just say 'while implementing' there? and what does "chunking up the map" mean? you didn't explain that. those parts could be written more clearly

- in your discussion of scale, you don't actually say anything interesting or even informative. you say that scale can vary from x to y rooms, and they vary from x to y rooms in zelda, and you say that if your room size is too small you may come off as a lazy designer to the player. but you don't actually say *why* you'd want a large or small dungeon, or how to decide on the scale of one. it read that part and felt like i went away knowing *less* about the scale of dungeons than i did before i read it

- in the dungeon flow structure part, i got the sense that you view dungeon design as a series of hoops to jump through in a specified order, or at least (being kinder) as something that is supposed to deliberately and gradually teach the player how to use a new mechanic, which is something i reject in my own designs and don't like when i see it in a game. it feels too overdesigned. i prefer more organic structures for dungeons. you also mention (briefly) non-linearity and complexity of a dungeon, but don't go into it, it's only a passing remark inside another paragraph, even though i feel that's probably one of the most essential things to understand about dungeons

[melos]

@Paul: I appreciate your honesty and the amount of things you've given to think about. I have a quick question. can you talk briefly about what you mean by "more organic" dungeon design? Do you mean more than one set of goals you can achieve to reach the "end" - or are there multiple endings?

I'm thinking about the dungeons I already have, and they all did start off as a graph (in the mathematical sense) - while there are multiple ways to reach the end, they all involve achieving the same subgoals (you *always* have to get these 4 keys and open these doors, but they don't happen in the same order). In that sense, I can feel the over-design in my dungeons.

what's an example of you achieving the organic-ness with something in your current game?

I will probably have more ?s later.

[Graham-]

Organic is when the player can develop on his own.

This kind of thing comes up in Zelda discussions every so often. What is a good Zelda? Why are modern Zeldas so constrained? Why is Mario cooler etc. Then Gimmy says Mario is Zelda and I don't know what's going on.

So "organic development" is something I strive really hard for in my designs as well. I'd rather eat my face than rope the player in a particular direction. (Not doing) this is achieved by being awesome, and is no small feat. Often Zeldas do it with back-tracking. They tie hints to the solutions to puzzles all over the place, so that when you slog through each area your impression of all the things you need to understand slowly develops. Each trip around the ring gives you something, and hopefully by the end you know enough to not have to repeat things until they feel redundant.

I don't know what the secret to "organic-ness" is. I've been doing this so long I don't even know who I am. I can't even tell if you know, but Paul brought it up so here we are.

Structure is like this:
  1. Player acquires item X.
  2. Player learns basic skill with item X.
  3. Player learns advanced skill with item X.
  4. Player overcomes big challenge relying on his skill with item X.
  5. Player acquires item Y.
  6. Player...

Sounds like FF13 doesn't it?

Organicis...ici..cism is like this:
  1. Oh my god this game rules.
  2. What?
  3. Player dicks around.
  4. Player gets maybe item X, Y, or Z.
  5. Wait, maybe he got X AND Y.
  6. Holy shit.
  7. Maybe gets Z too.
  8. No wait, he learns skills with X.
  9. He learns even more skills with X.
  10. No he's grinding for some reason with his X skills to gain some useful currency or xp or something.
  11. Woah he just beat the boss that needs the Z, but with X!
  12. Now he left.
  13. No, now he's back.
  14. He strokes the X.
  15. Now he eats it. Is that allowed?
  16. Now he practices with the Y, but he sucks so bad that the game laughed at him.
  17. So he gets another X and beats the dungeon with that.

ps. Dark Souls: super organic. Why? Because the structured challenges always hinted to the player at what areas of their skill-set they should be focusing on. Very, very good game. Organicniss... cis.. fuck, is about not determining the player's progress through an area. It is about giving many goals simultaneously, then letting them progress towards each one somewhat according to their will. Then you pour in structure to keep it interesting. You can say, "thou shalt not spend too much time doing Z," "thou shalt always do a little P with B... and ever 3rd time with J."

Does that make sense? I don't know.

It's like if you graphed a player's progress towards solving a puzzle, or getting an item, or gaining a particular skill, or learning something, all individually, the path for each would be unique to the player, and better yet, all would be traversed simultaneously. Then you can graph their tension, and the kinds of challenges encountered... and whatever. What? Yeah.

Organicism....

[ஒழுக்கின்மை (Paul Eres)]

by organic i mean the design process. i feel as if your design process was over-analytical -- i.e. i feel you planned out consciously each dungeon, imagined the player's path through it, and imagined how they would gradually learn about that dungeon's mechanic or theme, and the things that they had to do. this tends to make levels feel too planned out or mechanical, as if the player were just a rat in the game designer's lab, doing exactly what the designer planned, with little freedom. for an "extreme" of this you can play portal 1

how i design levels is that i try making something in the level, just playing around with a few game elements. i play it. if it's fun, i keep it, if it's not i remove it. i then make another part of the level, sometimes related to the previous one, sometimes not. i connect the different areas in a way that feels natural, rather than according to a map. after i finish making a bunch of areas in a group (the equivalent of your "dungeons" -- e.g. a forest, or whatever) then i test the set of areas together as a whole, and change things based on what i like and don't like. this is sometimes called "iterative design"

this method emphasizes the subconscious rather than the conscious mind when designing areas, and emphasizes organic growth from previous areas and "evolution" rather than "intelligent design" (that's why i call it organic). i hope that this method makes the world feel like a real place, as if it has its own life to it, rather than feel like something intentionally created according to a plan. my goal is to make a world that feels alive rather than an obstacle course or a test

i also try to have as many secrets to discover in my areas as possible, but i try to avoid those secrets being too formulaic (e.g. i don't want a huge number of secrets that are too similar to each other, like the burning trees / bombing walls things in zelda1, but instead i want each secret to be discovered in a different way, ideally). i try to make the areas diverse and lively, each area has to have its own "character" and feel different than all the other areas in a game. if two areas feel too similar, i try to improve/change one until it's distinct from the other one

if you want to watch this process in action, you can watch how i designed an area for my current game (this video is a bit old, but you can get an idea of how i make an area from it):

-- you'll notice that i didn't actually have much of a plan when creating it, instead i had ideas as i made it, and put new stuff, changed the positions of stuff, etc., largely at the subconscious level. i focused on what felt or looked right, sometimes even moving an object slightly a few pixels to get the feel right

[Graham-]

Ah, you're talking about the process for design. ... the results, though, are the same. I design that way too.

Often I'll design something small then just start adding stuff all over the place. Then I remove all of it except the most important bit. You do this enough times, stripping and building, then you end up with a "core." For example, you could try 10 different variations of sword swinging mechanics, then try all sorts of enemies and room layouts and physics interpretations and so on until you had something you liked. In the process you'd iterate the physics, the sword mechanic, a couple enemies, a couple room layouts etc.

Each time you try the sword with a different enemy you get a new idea of what the most "important part" of the mechanic is: the thing that makes it the best in every situation you try it out in. That becomes the core of the sword. You'd get the same for a few enemies and so on.

Then you have your pieces. Then you can start putting stuff together. Often you want to use "structural design" as an analysis tool for constructing good levels. You want to generate the levels themselves using what you feel is obviously true. Often you can iterate the elements of the design to the point that the correct and incorrect fit for any given element is immediately obvious once you try it out. A good rule of thumb is if an element doesn't slot perfectly into its place into the design, and you don't know where to put it, then something hasn't been iterated enough.

Once you have enough components you'll know how they can fit together, so then you'll feel far more confident building a dungeon in which the player can do 10 different things at once.

I guess that's not a very simple explanation... I don't know.

It's like, Miyamoto nailed the friction of Mario running, then coin boxes became obvious. He had Mario jumping, because that was his thing. Then he had blocks because he needed something for Mario to jump on. Then it was like, "this is _fun_, it would be more fun if I could jump into the blocks instead of just on them." So breakable were made. And then it was like, "it would be more fun if there were surprises inside." So then there were coin blocks and hidden coin blocks. Then a bug one day made a coin block infinite and it became obvious that a single block with multiple coins in it was fun in a different way. Players liked to jam it. So they kept it in.

That's organic design. Idea, execution, test, strip fat, adjust/add idea, repeat. Maybe it's best to think of dungeon design as playing with a bunch of legos rather than figuring it all out ahead of time. Except unlike with Legos, in game design each time you build a new structure you can redesign the Lego blocks themselves based on some new insight.

I love Game Design, most of the time.

A lot of AAA experiences had a wooden feel for this reason exactly. They are these super-tankers, they plan everything, they don't test and evolve, and you get a hollow gameplay experience. Assassin's Creed is my staple here: great animation/art, wicked-broken design... boring.

Then, at some point. hopefully, you can slot pieces together that individually provide a lot of variety. Mario is elegant because each piece was made independently, based on what Miyamoto (et al.) decided would be best at the time it was introduced. They refined each element until it was no more than what was necessary, then filled levels with them. Mario appears simple on the surface but it's design is actually insanely brilliant. Minecraft, Mario, and all these crazy games are actually so deep it hurts, just because they refined their core elements first, then found ways to stick them together. When you do that you don't even need design theory. Your player's experience will become organic all on its own.

I'm out.

[ஒழுக்கின்மை (Paul Eres)]

i agree that the results are the same and that dark souls is a good example of what i mean. i find that if you design organically, the levels also have multiple solutions (many of which the designer didn't even imagine), and that the player can complete the game in more customizable, creative, personalized ways, rather than the one intended way that the designer planned out. but yes i was referring more to the process, i feel games should "grow" like plants or forests rather than be "built" like houses or skyscrapers

which is not to say i never plan anything, i make outlines and to-do lists and a design document all that: you don't want the structure to be so loose that you never finish the game, but even with those i keep them flexible. most of the things i've implemented have differed markedly from their original descriptions in the design document

i believe that our subconscious mind is smarter than our conscious mind is, and that just as in composing music or writing a novel, overplanning/overstructuring can make a game or a level feel lifeless. i feel the best way to design a level is just to put stuff into it, test it, keep the fun parts and improve the bad parts, test it again, and iterate like that, gradually "growing" a level, as if from a "seed". i feel that the conscious mind isn't powerful enough to accurately predict what will be fun ahead of time, and fun can't easily be "planned", so that it's better to design levels through iteration. almost as if you are trying to *discover* a level (figure out a great configuration of the elements that compose it) rather than trying to build it

[Graham-]

"growing" best represents my method of development too. Sometimes I use "evolution." I may use that one more.

Often it's good to think of the player as a thing that evolves too, with his own struggles and developments.

Plans are nothing; planning is everything - Churchill (I think)

The real purpose of planning is like a fuel for the subconscious. There's this analogy by Steven King where he describes himself as a day laborer who does all the work, and his subconscious is this man with a magic bag who just sits around. But if he does the work the magic-bag man might offer up something beyond attainable in any other way.

You could say non-iterative development is an ego trip, disrespecting the subconscious.