World design in IF

Yeah – to be clear, I don’t think anyone did anything in a bad spirit! This was so obviously a well-intended attempt to help people out. And the original version isn’t very comfortably formatted.

It just happens to be an area where I’ve been warned I need to be careful about what I allow, if I’m going to publish some of this work elsewhere. The alternative would be to hold back content from my blog and release it only in paid contexts, which isn’t what I want.

I did! It’s in that PDF I linked. :slight_smile:

The .epub format is actually a renamed .zip file, containing the books pages as .css, .xhtml and image files. But in order to upload the .epub, I had to rename it to .zip, and when you extracted (instead of renaming) the thing you most likely ended up with a folder with the raw files from inside the epub.

Thanks! And I totally get that this was well-intended in the first place. I’d just been recently advised that I should be careful about situations like this, so I felt I should say something.

In general, avoid superfluous rooms. If you have lots of locations that don’t have anything to interact with, you can often remove them without harming the story. That said, exploratory games can be done well, but if the player needs to traverse the same five empty rooms over and over to accomplish tasks, that might warrant a re-think. You can effectively simulate a vast desert with one location.

Think of it as the same process where movies jump-cut ahead instead of showing 45 minutes of a character driving somewhere.

I think it could be a whole article in itself, when to (not!) include locations with no gameplay value!

A desert may be represented as a single location - still, it may be the right choice to add a supefluous “Side of the Road” between the road and the desert, to show that traveling out into the desert is a major decision.

Creating 24 seperate locations, to show that this is a really REALLY big desert, might be a bad idea.

Walking Sims consist of nothing but non-gameplay space!

in one of Eraserheads memorable scenes, the main character step into an elevator, press the button, and wait for a whole 13 seconds for the doors to close. That scene is repeated each time he returns to his apartment.

I imagine that obvious redundant space can be used (sparingly) to give a more uneven and interesting rhythm to the navigation. Also, if something mundane is given its own location, the player will get the notion that there is something going on:

I noticed that he concept ‘overworld’ was not mentioned in Emily Shorts article. She mentioned Hubs which seems like sorta like the same thing, and also talked about Street Maps, but she never said ‘overworld’. Nor did I see it anywhere else in the 430 page long IF Theory Reader.

Is overworld simply not an established concept in IF?

I would guess not, because it doesn’t fit the established style of most IF games. As I understand it, overworlds tend to consist of a large area that you navigate rather smoothly and painlessly, from which you enter areas where you find a certain challenge. So in VVVVVV you have a large mostly open area that you can move around without too much difficulty, and from there you can enter several much more linear spaces that are full of challenges and lead to the rescue of a crewmember.

IF generally doesn’t have the sharp distinction between open space and puzzle areas like that. Even if you do have puzzles in several different places that you can move between, it would be kind of rude to put a big separation between them; navigating from one place to another just isn’t as interesting when you’re typing n.e.s.s.w as when you’re walking and jumping across a bunch of obstacles. So what you get is more like a hub, where you can easily move from one puzzle location to another. Otherwise you just have the superfluous rooms Hanon was talking about.

(Emily Short’s Metamorphoses is one example of the kind of design I was talking about… and in another way, Porpentine’s howling dogs, which has an environment from which you move into all sorts of different episodes, though I think those are presented in a linear order. Pacian’s Rogue of the Multiverse does have a sort of overworld/exploration dynamic, though most of the really interesting stuff is going on in the overworld, in my opinion.)

I don’t recall it coming up in IF discussion.

I associate “overworld” with the idea of a different-scale map and probably a different UI. IF leans towards having a uniform texture and interface across the whole game. Not that you have to do that; Heliopause kind of has that distinction between large-scale travel-and-action and small-scale (on a planet) travel and action. But I don’t think anybody referred to the space-travel part of Heliopause as “an overworld”.

I think I get your meaning. Overworlds sort of exists out of time. If anything actually happened in the overworld, like the player got attacked by a creature, the game would have broken its contract with the player.

I think its hard to come up with something not already covered by Emily Short, but here’s a few ideas:

Brainstorm - There’s lots of brainstorm methods online, you just have to navigate past all those centered around business-meetings. My take: Close your laptop, find pen and paper, get wired up on sugar or whatever, and until the brainstormning ends, you are not you, you are David Lynch. Really, you are. After the brainstorm, you should have moral qualms about stealing someone elses ideas.

Research - even if everyone knows what a park looks like, you’re supposed to be one step ahead of the player. Google stuff, or go to the nearest park.

Use a modified snow-flake method. - That is, start out with the big picture of your world. What shape should the room connections form? Should the south area have certain features as opposite to the north? What impression should it, as a whole, give the player? How should the later unlocked areas differ from the first ones?

Forget the rooms - Try for a while to ignore the digital navigation mechanics. They can come later. Instead, try to envision a true-to-life dirty and vivid reality.

Having tried (and failed) many times to construct interesting, believable and gameplay-enhancing worlds for what I’d think of as „a large variety of media“, here are a few lessons on worldbuilding in general and on doing it for IF in particular that I learned from experience. This is what I personally wind up doing every time, regardless of the way I started out:

  • When I begin working on something new, I have mostly the aesthetics of the world in mind, so gameplay considerations are very much secondary initially.
  • That said, I try to avoid the well-known pitfalls.
  • I do as much research as I can during my initial brainstorming, but I always have to get back to it later on regardless. Still, for me, it’s a waste of time to try and make up something even remotely believable from fantasy.
  • When writing IF, I start the actual building of my gameworld with a series of rooms with names, but without descriptions. The reason for this is that writing actual descriptions demands effort to produce an acceptable literary standard, while the visual conception I have off all these locations does not. This is never the final layout.
  • This rudimentary world can already be tested for some of the spacial arrangement effects it creates (i.e. does it confuse the player? Does it involve tediously long walks? Are these tedious enough?)
  • As I begin implementing interactions with the world, the descriptions of places and objects follow naturally. This has the advantage of limiting me to the kind of things I can actually express without having to throw a lot of previous work out of the window while still providing me with a kind of skeleton to work in.
  • When I grow tired of the small details, I go back to expanding (or pruning) the world skeleton.
  • Invariably, I have to trim down the number of rooms I created radically several times during the development. In my recent IF comp entry, I ended up with about half of them.

The downside to this „approach“ is that it tends to produce very visual descriptions. Regardless of whether this is a good or a bad thing, it has its limits, because I definitely want my room descriptions to be brief, especially if I have a lot of them. Thus, I err on one side or the other in every single room I write.

I have experimented with giving natural names to rooms as well as managing a systematic internal naming convention (e.g. L-1_station_platform). From my experience, natural names are better for several reasons: 1. The artificial creation of different „dungeon levels“ is not a good way to envision a naturaly grown environment (not all stairs are the same length). 2. Whenever I want to create fair mapping puzzles, I have to take more care than to just follow a naming convention. 3. If I have namespace collisions, My map is usually too big. 4. If I end up moving a room around the map, I don’t want to have to rename every reference to it.

Getting procedural world generation right is even harder than doing it by hand. It gets a lot easier if I can restrict myself in some way (for example, all rooms in a dwelling are normally of similar size, making it much easier to re-use parts of descriptions. So maybe if I think hard about it, I don’t have to go outdoors after all?)

I’d really like to get away from this simulationist spacial world model for once, but again, I always end up using it in some way eventually.

Good idea, I never considered this! Since a room can be created with a single line (in Inform 7), it makes sense to just create the world and try it out.

Huh? I didn’t get that. Can you try again?

What I meant was:
it is, of course, possible to construct and express a story, or even just an environment, without resorting to a series of “rooms” - that is locations indoors or outdoors, with props, scenery, backdrops etc. that form a geography to be navigated in some way by the player, as is customary in parser IF. In contrast to this convention, I cannot, off the top of my head, think of any good novel that even tries this. Instead, many authors of non-interactive fiction rely on dialogue and/or interior monologue alone to convey anything from setting and character to plot and message. The same is done by many writers of choice-based IF.
[rant]There are those who have done otherwise to some degree - say, for example, Tolkien, in his entire oeuvre. As you can readily see from this failure, which is nothing but an endless, embarrassing indulgence in Kitsch (to put it mildly), among which I count the worldbuilding, this may not be a good idea in general.[/rant]
I realise that the ability of the player to enter anything and everything into the command line often leads authors of parser games to implement such a “simulationist” world model. The existence of this apparently natural (really?) dynamic makes avoiding it all the more interesting to me, even if only for a short work, as an experiment.

Okay, I think I get it now. To be fair, Tolkien was NOT the only one! It was the norm in those old novels to use lots of pages to describe exactly what the landscape looked like in meticulous details (Not all of them, mind. ‘The Three Musketeers’ was critizised for its lack of ‘atmosphere’ - because there is no mentioning of the weather and such - in C. S. Lewis essay “On Stories”, 1947)

I think old parser games attraction was the same. The enjoyment came from exploring the world at your own pace, and drawing maps. The static descriptions were a neccesarity due to gameplay reasons, but they were also a feature.

I have no patience for books which opens each chapter with several pages of environmental descriptions. But I’m not sure wether the form is bad as such, or if I’m just not able to read it right.

Going slightly off-topic, one of the greatest works of Portuguese literature is Eça de Queiroz’s “Os Maias” (“The Maias”. It’s a family name). Among other things, the book is infamous for spending ages - ages! - describing furniture.

But when you actually read the book you see that Eça is not just describing furniture. Ostensively, he is; but in fact he’s using that furniture as an excuse to tell us an awful lot about the characters that sat on those chairs, on those couches, that ate at those tables.

Anyway, this is probably not what you mean, you’re probably talking about straight-up environmental descriptions which don’t actually develop characters. But I thought I’d mention it, because hey, it’s also a valid technique for IF… maybe not strictly 100% on-topic, but worth a sidetrip, I thought. :wink:

Yes it is considered bad form in modern writing.

There is a rule in the Writing World, “Show and not tell!”

It means the your write about the character interacting with the world and not studying and describing the world.

Not just the writing world! :slight_smile:

Is this the lost chapter of The Princess Bride?

My Dad love James Michener, and we are always joking about how “the first three chapters are about a bubble rising from the ocean floor…”

Personally I want to keep the amount of text to read between being able to interact with the world (choices or parser) as short as possible. In a parser-based game in particular if I want more information on something I can x-amine it myself, so no need to stuff the room-description full of details. My ideal game would read much more like Howard than like Tolkien.

On a personal note, agree. I used to think I liked prose-rich games, then I realised that no, I like well-written games - and games which overdid the room descriptions and the prose did nothing to my enjoyment.

If you have not done so, I recommend checking out Ryan Veeder’s games as a lesson in descriptions that point the player toward only what is interactive without being too minimal.