Strategies to Eliminate Compass Directions from IF Narrative?

compass200t

The Problem

Traditional IF typically uses compass directions; North, South, East, West etc. to navigate the game. This has problems:

  1. Compass directions do not feel natural.

  2. The location description is forced to mention the exits in terms of these compass directions, which often makes the text feel lumpy and awkward.

  3. It’s hard to write good exit descriptions without becoming repetitive.

Possible Solutions

1. Using Location Names

eg

  • go into the dining room
  • go to the main road
  • go into the toy shop

This works, but has the drawback of being tedious to type.

2. Using Choices

Inject navigation choices, so that the above appear as “clickable” choices.

Some problems with this:

  1. Where story choices exist too, the list of choices can get too long.

  2. For parser games, players are lead toward the choices rather than typing commands.

3. Using a Companion Map

If a visual map is given in the UI, it is possible to click on adjacent (or even remote) locations to navigate.

Clearly such a map would need to reveal locations only after they are visited or mentioned.

Problems with this:

The map would need to be visible on the UI in addition to the narrative. For large screens, this is not a problem, but for mobile devices screen space is too limited.

4. A New Navigation UI?

Traditional UI compass roses are used to navigate the N,E,S,W paradigm.

Could such a device be used for generic, non-compass movement? Perhaps a star shaped device labelled with the names of adjacent locations?

It would nevertheless still occupy significant screen space. Additionally many locations would benefit from name abbreviations, lest the text be too small.

Another problem is that the player has to continually switch from keyboard to mouse, then back to keyboard. Not a big problem, but it gets annoying unless there are also keyboard shortcuts - of which new players would be unaware.

5. Intelligent Completion

A twist on the idea of typing the entire location name, would be for the UI to detect, go ... or enter ... and automatically offer relevant text completions.

This would obviously require a custom text input box in the UI, and moreover, mobile users would not be able to use their familiar system virtual keyboard, using an in-game custom input area instead.

Nevertheless, the latter could offer a better class of auto-correct and game specific word completion than the system keyboard could achieve.

Discussion

Are there other ideas, better ideas?

Looking for a general solution rather than something that worked in a specific type or style of game.

This has been discussed before, but there weren’t any real conclusions. Maybe because the answer needs help from the UI (as in most ideas above) rather than from authoring techniques. In which case, what would be the ideal UI facility?

2 Likes

This has been discussed many times and all of the options you’ve identified have been implemented to varying degrees of success.

I believe this has to align with the story you’re telling. If your story can be told/played without locations or compass directions, then you don’t need them. If your story has a map of locations, then a compass or go to location strategy is likely the best solution.

3 Likes

I grew up with maps and a compass. It feels natural to me. And, it is traditional.

However, I am all for the alternatives you mention as long as it flows smoothly.

Long live parser based text IF…

3 Likes

I’m working on a “forward, back, left, right” system. Although it requires a little more work to adaptively change what goes where. It’ll be great for mazes! …if people liked mazes, which generally they don’t.

2 Likes

i find it especially intuitive the way advent does this:

i am… absolutely horrible with directions no matter how they are represented, soo i tend to navigate exclusively with landmarks.

There’s nothing wrong with “all of the above”. Options are complementary.

TELEPORTATION POINTS
This can be magic words, transporter, or public transit.

JOYPAD
Useful if you have on-screen map, either 2D or 3D.

This can be avoided by using “hub” points. Use CITY, MOUNTAIN, SEA. Then go in the city, allow more choices, market, housing, park, etc.

I don’t see a problem as long as you can also type things.

SEE MAP
List of choices displayed, select destination.

GO (place)
You go to (place)

You can turn the map on and off. It doesn’t have to be constant on.

I would advise against alternate navigation as substitute for traditional ones. OK if you want it as alternative/addition.

@libele teresting! Landmarks or not this adds another idea;

You type some kind of reference, either the name or some synonym or tag of some kind. But it would be a kind-of fuzzy match that didn’t use the traditional parser. So;

go beach

Would use the normal parser, but additionally;

beach
sea
coast

Would all translate to, go beach.

Furthermore, immovable objects in that location could also be used.

sand

might also work, assuming “sand” was not something you could move about (in this example). as might waves etc.

@ramstrong Good ideas! Perhaps something new can be done with joypad or mouse. Kind-of navigation mode?

There was some discussion here about using the numpad to move around.

The numpad would work best as a substitute for cardinal directions. I’m trying to get away from those, so that players are not forced to remember the spacial layout.

Using a joystick or mouse sounds the same, but not necessarily. Perhaps they could navigate destinations rather than directions.

Of course, i guess also the numpad could work, where you’re actually “speed dialling” a destination as opposed to 4=West, 6=east, arrangement.

Additionally, what if they don’t have a numpad?

IMO, the problem with a general-purpose solution is convincing people to adopt it.

Blue Lacuna only has optional compass directions. (You pick up a compass well into the game, and only then are you even permitted to use compass directions.) I think it’s great! Indeed, Blue Lacuna has a bunch of great parser innovations, including default verbs (type a thing to examine the thing; type a place to go to the place), and the author, Aaron Reed, made his innovations available as Inform extensions for anyone to use.

The problem is: hardly anybody did use his extensions. Instead of adopting the “superior” solution, authors continued to ship games based on compass directions.

To truly eliminate compass directions, you’d have to not just create an alternative that’s equally good; you’d have to create an alternative that’s obviously, overwhelmingly better. Your alternative would have to be so much better that if anyone shipped a game with boring old compass directions, players/ reviewers would report it as a bug. “This game is pretty good, but it’s still using compass directions instead of the newer, better thing.”

I also predict that the new style would have to be built into Inform, so naive authors would automatically take advantage of it. I don’t see how that could even be possible until the open-source version of Inform ships. (At NarraScope 2020, Graham said that would be well before the next NarraScope conference, but there wasn’t a NarraScope 2021, and we can only hope it’ll be ready before 2022…)

2 Likes

A very good point, but perhaps the title of my thread here is a bit misleading.

My aim is to eliminate compass directions from the narrative, not actually from the whole game.

I should explain;

Narrative games with an interesting story hook you in with situations and events, then they “spoil it” by saying things like;

To the west is a path leading out of the garden.
This is the main high street leading east-west through the town.

They do their best to hide the words EAST and WEST etc, but to me, these stick out like a sore thumb, because “normal” stories do not have these compass words. What’s more, writers are continuously struggling to come up with natural sounding ways to disguise these unnatural constructs.

So what I’d like to do is eliminate these from the prose and instead have other bearings, such as landmarks, places or objects. But at the same time, the game would nevertheless still have geography which would mean point A is east of B etc. Which would be borne out on any visual map.

This means compass direction would definitely still have to work, but those directions are not made clear to the player directly. In which case there needs to be a navigational interface in place that works with what the player was first told in the narrative.

Perhaps this is what you’re saying. Whether you mean authors adopt it or players adopt it. I’m interested in what players would want.

2 Likes

Yeah, this is the thing I’m interested in too! When I think about this problem, I primarily relate to it from an author’s point of view – as you say, it feels annoying and artificial to work this in, rather than just describing the landscape and what you see without taking the player out of the story. So I totally see the upsides there. On the other hand, when I’m playing a game, even one with GO TO LOCATION implemented, I usually just wind up typing the directions, since it’s typically many fewer keypresses, with fewer opportunities for typos – like, GO TO BEACH is equivalent to six one-letter movement commands, which is pretty far in most games! And if movement requires moving my hand off the keyboard to the trackpad or mouse, that adds even more friction.

I also recently introduced my wife to some IF, and was surprised that the directional movement system seemed to come really naturally to her. She’s basically played no video games of any kind previously, and doesn’t naturally use cardinal directions when navigating in real life, so I was expecting her to struggle. But after I explained it to her once, she easily was able to pick out the possible exits and move around the map (inside, outside, and enter were harder to figure out). And after thinking about it for a bit, I can see how despite the conventional wisdom that things like GO TO X are more newbie-friendly, it could be that for some people it’s the opposite way round: usually the problem with learning how to play a parser game is that you’re not sure what to type so that the game will understand, and while there are ways of instructing the player in the ABOUT text, and coding the GO TO command to be robust to synonyms, it’s still maybe less intimidating to know you just need to type a direction.

That’s just one anecdote of course, but I think there are many things like this that players experience differently from authors. Like, a similar thing that annoyed me when writing my IFComp entry for last year was the way significant items usually get broken out into their own line rather than being naturally included in the description paragraph. So I made most objects undescribed or wove them into the description, which felt much more aesthetically satisfying. But when I looked over transcripts, I think many players wound up struggling with that and the game was harder than intended, because the idea that the most obviously significant objects are going to get called out separately is a familiar convention, and also one that’s got some real ease-of-use even if you’re a new player completely unfamiliar with the convention.

I remember hearing that Blue Lacuna’s ease-of-use upgrades were based on watching a bunch of people try to play parser games, and implementing solutions that would remove some of the pain points. Having that kind of breadth of data on what works for players, especially newer players, would be super helpful, I think, because I feel like it’s easy to make assumptions, and then second-guess those assumptions. I’m curious whether IF events that might have a little more mainstream attention, like the various itch.io game jams, save transcripts and would be a place to look – the IF Comp ones I think are really helpful, but primarily reflect how experienced players tend to approach a game.

2 Likes

Those are exactly my motivations, although you’ve explained it better than I have; thank you.

Perhaps i am wrong, but my “feeling” about how someone, who has never played any IF before, would react to directional navigation. As it would seem completely artificial. I definitely feel that, even if they adapted to it, the story is made worse for it and it detracts from the entertainment.

Furthermore, I imagine the idea that they might have to make a map to play a game
a total dealbreaker for them.

Nevertheless you make some good points that, if faced with the “guess the word game”, they’d much prefer a N,E,S,W system, albeit bogus. But if that’s the case, the real problem is the input system instead.

I agree and also dislike the idea of objects being separate from the main text. I much prefer to have them mentioned “inline”, and only separate later - but only then if totally necessary (for example, do you really need to drop anything?)

I am trying to think from a blank page, rather than be tainted by preconceptions. I’m sure people on this forum feel quite at ease with the current way,

but what does the wider, more mainstream and less IF familiar, player want?

My gut feeling is that the traditional compass directions are preserved largely because they can be accessed with a single letter - or in the case of the intercardinals, two. And since up and down can be abbreviated to U and D…

If you’re familiar with the system, you can move about very efficiently. A more intuitive system, such as landmarks, requires much more typing.

2 Likes

True enough, but this sounds fundementally an interface problem not a conceptual one.

To me the big win for compass directions is that you can build a picture of where things are in relation to each other, on a larger scale than “this thing is next to that other thing”. And people seem to pick up on that pretty easily and it just works. I’ve seen a few games (maybe Zarf’s Bigger Than You Think?) which use like up/down/inward/outward instead of N/S/E/W, so I don’t think the names of the directions really matter (though NSEW aren’t very evocative), but having directions means you can think, OK, this place is a couple levels down and a couple rooms outward from where I am and navigate on a larger scale.

And I know there are people (notably Chandler Groover) who don’t seem to think directionally and spatially and somehow can build a picture of a place from just the connections, but I think that’s somewhat uncommon. And I think that even in a theoretical algorithmic sense it’s a more computationally intensive thing to do.

And I’ve also seen games where you don’t care at all about the spatial relationship of places and navigating by landmarks or place names is fine, and it’s often more convenient because you don’t have to keep the map in your head, even roughly, but as @Melendwyr said it’s often more typing. Skies Above is a good example of this IIRC, and has pretty good abbreviations too.

3 Likes

Uh, yeah. I have trouble with that. Big trouble with Capital T. So much so that if you do it that way, I will not play your game. I think many people think like you do, “why not pretty up the text?”, but what will happen is that players will get confused.

Then there are players who can do it just fine, but I think that we can all agree that @Warrigal is an expert player!

This is just one person’s opinion, of course, but if the form of fiction is very important to you, I suggest Choice games is a better fit. Room based Parser games heavily relies on compass direction convention. You may add to it, but replace it at your peril.

@JoshGrams Now this is interesting, because it might reveal different ways people’s heads are wired!

For example, in real life, i tend to navigate from point to point. So routing is basically a chain of locations.

When i talk to cabbies, they navigate a lot more spatially (ones that don’t use sat nav anyhow). So that helps them re-route a lot easier as they must essentially have a map-style or compass-style mental model.

What this is saying, is that there might be case for both styles.

2 Likes

I don’t see a problem with compass directions.

3 Likes