A problem of compass directions

Authoring General Design Discussions
1

I don’t know how silly this question sounds, but I’d rather have a seasoned game designer dismiss it than let it dangle…

As an exercise, I attempted to turn an existing location (a tavern by the sea) into a map of rooms for a hypothetical IF. Let’s say the plan is something like this (including some room planning, but keeping the proportions):

IF Orientation Problem (1)
IF Orientation Problem (1)1000×586 59.7 KB

The player will then perceive the rooms’ map like this:

IF Orientation Problem (2)
IF Orientation Problem (2)1200×601 60.3 KB

All good so far. What if the compass were like this, though, rotated by 45º? It could be a location that is part of a bigger picture and one can’t have only N-S alignments, right?

IF Orientation Problem (3)
IF Orientation Problem (3)1000×586 62.6 KB

Then, the player would have to struggle with this:

IF Orientation Problem (4)
IF Orientation Problem (4)1000×630 55.7 KB

Am I the only one thinking this is getting ugly? Is this the way to go? Or do designers tend to stick to N-S alignments?

3 Likes
2

For my own part – speaking entirely personally here – I think that N-S/E-W passages “feel better” as a general rule: they take somewhat less cognitive effort to map out mentally and remember how to traverse, and especially if the player is going to need to explore and learn the space, and/or move through it repeatedly, it’s easier if most of the connections are in cardinal directions. That said, a map that’s on a four-direction grid exclusively “feels artificial”, and it’s good to have some map connections in orthogonal directions. This helps to promote … not actual realism, but the semblance of realism, as an effect, which is much more important, generally.

I think that, typically, there’s little reason to worry excessively about adhering exactly to the actual location you’re modeling, unless you anticipate that your target audience is already very familiar with that location. So if the players of your game are primarily, say, the denizens of the tavern you’re modeling, strict adherence might matter; but Joe Schmoe, who has never been to the tavern, is unlikely to care if your modeling is perfect; he cares more about whether it’s easy to learn to get around in the virtual world. (Similarly, it would be weird to set an otherwise-realistic game in Manhattan, only rotated so that everything is laid out along a NE/SW-NW/SE grid, and to put Harlem in the southern part of the island; that violates a geography with which many people are already very familiar, and would really only be a good idea, in my opinion, as a well-thought-out effect; otherwise it looks like a simple error. Or, as someone who lives in Denver, I found it incongruous in last year’s IFComp that a game set in Denver mentions the Rocky Mountains being visible to the east, which is never true in real life. It did not ruin the game by any stretch of the imagination, but it was rather jarring.)

So I guess what I’m saying is: it’s generally wise to do what needs to be done to make a good game, rather than what accurately models real life, because it helps to focus on the fictional needs to the story you’re creating, rather than a perfectly accurate model of an existing place. To that end, I’d rather see a model of the tavern created that’s easy to navigate, and have the realism tuned more toward description and detail and objects than to whether the map is oriented exactly in the same way as the real-life space on which the map is based.

Does that make sense?

8 Likes
3

I agree with what has been said. Design for playability and good fiction. Realism can be part of that, to a degree, but shouldn’t be the end goal.

In static fiction, people speak clearly using lines of dialogue that are complete sentences. That’s not realistic at all, but it serves the story. Every character is distinct (there’s only one Mr Anderson, one Ms Smith, one vicar, one butler, …) just like in IF there’s only one object of every kind (and not, say, two identical bottles on a shelf). Again, this serves the story at the cost of realism.

The map of the real tavern is fine in any orientation as a starting point, but I would expect to redesign it gradually during the writing process to fit the story and, where applicable, the puzzles.

8 Likes
4

I agree with @lft that you’ll want to adapt your map to your story, so do what feels best. I would encourage you to enable landmark navigation in parallel with compass-based navigation, which could lessen the issue for many users.

You may ignore the rest of this rant. It’s an argument I’ve made more than once over the years.

Compass-based navigation in parser IF is an odd artifact of the original Adventure that, unfortunately, was copied in work after work.

Adventure did not require compass directions until you got a few rooms into the cave. Above ground you could use landmarks to navigate. From the game’s help text:

Most of my vocabulary describes places and is used to move you there. To move, try words like forest, building, downstream, enter, east, west, north, south, up, or down.

(Only four of the ten examples given are compass directions.)

Unfortunately, many of the ports dropped much of the landmark-based navigation, which actually makes the beginning of the game rather unplayable. The descriptions have strong hints about which way you should go, but they’re often not stated in terms of compass directions. If you want to follow the stream, you have to unrealistically waste a lot of time guessing compass directions or have a map from a previous attempt. The original game, supported and encouraged natural travel commands like “go downstream” or “follow stream.”

Compass navigation is enforced in the cave because: (1) you can’t see many landmarks, (2) spelunkers exploring new caves are generally concerned with mapping them, and (3) getting lost in the cave was a puzzle to be solved–saying “go building” while you’re in the Giant Room would be a cheat.

Players spend most of their time in the cave, so they remember the compass-based navigation and forget the magical experience of navigating in a way that’s closer to how most people think about getting around.

In my own parser-based (probably-will-never-finish-it) WIP, I’m putting a lot of effort into enabling natural navigation to the point that the entire game is playable without entering a compass direction. Of course, I allow compass navigation because experienced players are used to it. And I drop compass directions into the text to help reinforce the map for those inclined to think about it this way.

4 Likes
5

Also from a practical play standpoint, if there is lots of map traversal and most of the directions are on the diagonal, the player is typing two-letter directions NW, SE instead of one letter. Not a huge distinction, but on hundreds of repetitions will feel less streamlined than cardinal directions N,S…

Compass directions are basically a convention of parser and most players accept them, so “realism” needn’t be the final word in map design over player ease of experience and comprehension of the map.

3 Likes
6

There are cultures which do predominantly or even exclusively use compass directions for spacial directions, so if the player character is from such a culture then this parser convention would become a feature of realism :wink:. But you’d have to check that you don’t say something like “your right hand” or else you might be departing from realism again! (There are cultures which use both of course.)

2 Likes
7

@aidtopia

Landmark based navigation is great and I love seeing that in a game, but I don’t think compass navigation should ever be dropped unless the goal is to intentionally disorient the player.

I understand that accurate compass directions aren’t realistic. When you’re at a friend’s house and you ask where the restroom is, they say something like “At the end of the hallway to the right”, not “go east then north” (oh, okay, let me just whip out my compass and align myself with magnetic north before locating the restroom).

That being said, while in reality we rarely know what compass direction we’re facing when walking about in a house, we at least understand the relation of one room to the next. If you were to create a mental map of the house, you’d be forced to assign a north direction to it (most likely the entrance door would be south), and so moving forward would be north, returning to that room would be south, etc.

Without the compass directions, relations of one room to the next are confusing. If you leave the entranceway via the living room and then enter the kitchen and then enter the door at the back of the kitchen, what direction are we heading? Did we walk in the same direction, or are we looping around to the entrance now?

Compass directions are also especially useful for overcoming the issue of having no idea what lies beyond a door. If this is your first time in a mansion, you shouldn’t know the door to the right leads to the music room.

Anyway, those are my thoughts on it.

3 Likes
8

This was an interesting point in the Roger Brucker cave history books (about exploration of the Colossal / Mammoth / Flint Ridge caves). The reason “Y2” was written on a rock is because someone took a compass bearing and distance measurement from Y1, and then another compass bearing and distance measurement to Y3, and so on – eventually compiling an accurate survey of that area.

9

To make an earlier point explicit: typing “forest”, “stream”, “clearing” hundreds of times would be very annoying.

3 Likes
10

That’s one advantage choice narratives have - you can make weird directions and interactions clickable.

You can also re-define new directions (in I7 at least) but that can end up being more trouble if you want the player to go CLOCKWISE or SUNWARD, for example.

11

It may be an artifact of Adventure, but I do think the reason it sticks around is because it works.

If I were in an unfamiliar building in real life, I’d probably get directions like “turn right, fourth door on the left, up the stairs, then turn right again”. I could follow that easily, and reverse it without any active thought.

But that reversal, for me at least, depends on a lot of deeply-ingrained instincts that don’t work in a text-based game—or even a graphical game that doesn’t give me full, FPS-style navigational control. It would take me at least a few seconds of actively thinking about it to turn that into “turn left, down the stairs, fourth door on the right”. If you’re like me, it probably took you a few seconds to realize that’s wrong: it should be turn left, down the stairs, turn right, fourth door on the left.

By comparison, “west, north, west, up, east” reverses trivially to “west, down, east, south, east”. I can visualize the Trizbort-style grid in my head without any issue.

Landmark-based navigation is all well and good, but in my experience, it’s only really intuitive for very linear maps. Let’s suppose I wanted to modify Adventure’s early game: you go downstream, then into the forest, then toward a clearing. How do you get back to the stream now? Maybe I went south along the stream, west into the forest, north into the clearing, so going east should take me back to the end of the road, but I would have no way of knowing those relative directions until I was close enough to see the road and the building again.

Tl;dr I believe compass-based navigation has stuck around, not because nobody’s ever tried an alternative, but because it makes it less taxing for players to navigate and easier for them to build a mental map. In real life we have various biological systems that make it easy to keep our bearings (e.g. it’s trivial to face the direction we entered a room through even if we’ve done other things since then); the convention of using compass directions is a way of replicating that, even if we don’t always use those absolute directions in real life.

6 Likes
12

After building a lot of choice-based exploration in Seltani, I decided that the clickable exits (in room descriptions) were also annoying. The next idea is a clickable map for each area, but I never got that far.

2 Likes
13

It is difficult to find a good alternative to compass directions entirely. In a cave it is almost a necessity. The same for large buildings and structures. Most structures are fixed in place and it works. What about structures and imaginary places that change their orientation? For example, shipboard directions, ie. POSH, port over starboard home. Or a place like Hogwarts that have movable staircases and necessary rooms that play hide and seek? Also, sunward mentioned in an earlier post. Sand Dancer? You have to provide clues that account for the changing direction of the Sun.

I just do not like clickable options in IF. For me, they remove the immersion and abstract thinking that parser games provide.

3 Likes
14

This notion of natural directions sounds amazing. I’m sure it can get cumbersome after a while, but just having the extra choice to GO TO RIVER (alongside the compass directions, of course) is just too cool to ignore.

I am very curious about how such a thing can get implemented… My guess (for Inform 7) is one has to define a new action (eg HEADING TOWARDS) applying to a room.

Perhaps one can also define a similar action to “any thing” and then find the room that thing is enclosed.

Thus, both > GO TO OASIS and > GO TOWARDS COCONUT TREE will find the quickest way towards the Oasis, the room enclosing the coconut tree, and then advance the player one step towards it.

Am I guessing correctly?

2 Likes
16

Example 306 in the Inform Recipe Book has some code that you can use to get started with this:

http://inform7.com/book/WI_17_10.html#e89

1 Like
17

In Inform 7, Emily Short’s Approaches extension gives you the support for this for rooms. Though not one step at a time – it tries to complete the journey to the destination, and reports how you got there, if you get there, or how you got to where you ended up if you don’t. After reading such a report, you could use compass directions to get to the inbetween places if you wanted.

You can extrapolate from Approaches to create the same action for things that aren’t rooms (e.g. coconut tree) though speaking from experience, this is a lot more work. There are usually way more non-rooms in the game than rooms. It hugely multiplies the number of ‘go to’ phrases the player can use and then you need to program carefully to make sure the game tries to take smart, logical journeys instead of dumb, illogical ones.

-Wade

2 Likes
18

That’s the idea behind Distant Movement, which tries to make the process of navigating to a place as seamless as possible (while you’re moving toward something, you can press enter with no command to keep going, or type a command to do something else instead).

1 Like
19

There are games that does not uses compass with somewhat success. I would recommend The lost legends of Redwall: Escape the Gloomer.

If you have a manageable environment, where you don’t need to orientate yourself, you can build that just using “doors” or capturing the action going into any object in Inform 7.

Imagine a fork where there’s an upper trail and a lower trail. The location has both objects lower trail and upper trail. Both can be examined, and in both you can enter, where you being teleported to the proper next location. Maybe it could be a little tedious to program, but it is a proper way to do compasless IF, using only the verb GO TO.

Also… I remember this hot debate, some time ago:

But I still in my opinion, the compass is unbeatable as an interface: Recently I’m playing the last CROCODRACULA, and the Ryan’s port shamelessly states the exists as this

NORTH SOUTH EAST INSIDE.

At the end of every description.

It is a commodity. It is easy to learn, and easy to get the bearings. It is easy to use. It is only rock and roll, and I like it.

3 Likes
20

I thought GO TO was considered harmful?

…

I’ll show myself out.

7 Likes
21

No. No?