PDA

View Full Version : Text-Based Adventures



Neftren
2012-09-10, 02:39 PM
So, does anyone here have any experience with creating/writing Text-Based Adventures? I've been tasked with putting one together (with my friend), but I'm a bit stumped on where I want to start. For starters, what formats work well in an interactive, text-only medium?

Rakmakallan
2012-09-10, 03:41 PM
Clarification is needed. Do you mean GMing adventures in text-based environment, or actually designing an interactive adventure like the old Infocom games?

Totally Guy
2012-09-10, 04:51 PM
It sounds like you are talking about the Parseley games by Jared Sorensen (http://memento-mori.com/online-store/parsely-games/).

The Action Castle incident! (http://www.youtube.com/watch?v=Jo8R4s0AY4o)

Ravens_cry
2012-09-10, 05:21 PM
Clarification is needed. Do you mean GMing adventures in text-based environment, or actually designing an interactive adventure like the old Infocom games?
This question is mine as well.

Neftren
2012-09-10, 06:51 PM
Clarification is needed. Do you mean GMing adventures in text-based environment, or actually designing an interactive adventure like the old Infocom games?

Ah, sorry. I mean the latter: actually writing an entire text-based interactive adventure.

Ravens_cry
2012-09-10, 07:18 PM
Ah, sorry. I mean the latter: actually writing an entire text-based interactive adventure.
Unfortunately, you would probably find other forums more suited to this elsewhere.
Still, there is a certain similarity in writing a module and writing a text adventure I imagine.
Many of the early ones, like Zork and Colossal Cave Adventure, i.e. Adventure, took many ideas from D&D.
The exact particulars are a closed book to me, however.
This (http://www.brasslantern.org/writers/)and this (http://www.textadventures.co.uk/quest/)may be of help, however.

EccentricOwl
2012-09-10, 07:26 PM
I sure do.

My first recommendation: start with a purely text-only adventure so you can get the hang of writing. Deal with the programming later.

Take a look at a program called "TWINE" for making such simple adventures. The output is basic HTML . It's sweet.

NichG
2012-09-10, 08:45 PM
I'm a fan of TADS as a system for these. It is general enough that you'll pretty much never hit a situation where what you want to do is strictly impossible within the system, but you can also do really simple thinks without having to master the language first. There are also a lot of intelligent defaults you can just take advantage of, so when the player tries to do something unexpected there's a good chance the engine has at least some kind of response ready (stuff like 'look thru/into the aquarium' is handled naturally if the aquarium is specified to be both a Container and Transparent or whatever the TADS mix-in class for that was).

The one place I've found TADS to be slightly inconvenient is that it has something like three options for NPC handling and learning how those work can get complicated. Basically you can choose between a sort of 'Option 1, Option 2, Option 3' kind of thing, a topic based thing (ask X about Y), or some mix of them, but because its all written where the dialogue choices are objects just like trees or apples its a little counter-intuitive.

Neftren
2012-09-11, 11:12 PM
I'm a fan of TADS as a system for these. It is general enough that you'll pretty much never hit a situation where what you want to do is strictly impossible within the system, but you can also do really simple thinks without having to master the language first. There are also a lot of intelligent defaults you can just take advantage of, so when the player tries to do something unexpected there's a good chance the engine has at least some kind of response ready (stuff like 'look thru/into the aquarium' is handled naturally if the aquarium is specified to be both a Container and Transparent or whatever the TADS mix-in class for that was).

The one place I've found TADS to be slightly inconvenient is that it has something like three options for NPC handling and learning how those work can get complicated. Basically you can choose between a sort of 'Option 1, Option 2, Option 3' kind of thing, a topic based thing (ask X about Y), or some mix of them, but because its all written where the dialogue choices are objects just like trees or apples its a little counter-intuitive.

Yeah we've settled on using TADS. The tutorial was rather amusing.


I guess what I'm wondering is how much environmental interaction should be put together (e.g. examine axe vs. "an axe lies in the corner"), and just how much of the world really needs to be filled out. I could imagine the complexity of the whole thing to scale rapidly with more and more settings.

NichG
2012-09-12, 12:13 AM
Yeah we've settled on using TADS. The tutorial was rather amusing.


I guess what I'm wondering is how much environmental interaction should be put together (e.g. examine axe vs. "an axe lies in the corner"), and just how much of the world really needs to be filled out. I could imagine the complexity of the whole thing to scale rapidly with more and more settings.

I've seen games that were successful with very little, and games that were really lush and anything you wanted to look at would have a description, which you could then 'think about' or whatever.

Honestly the most important thing is to keep the level of detail more or less constant. You want the player to develop expectations about how much time they should spend trying to examine every little thing, and then you want to try and meet those expectations throughout (so if you're going to do sparse description, don't later have a puzzle that requires looking at the blemish on the doorknob on the door or something).

I'd say perhaps that a good way to proceed is to try to get as much of the 'game' parts of the game in before worrying too much about describing the windowsills and potted plants. That way you can decide how much it needs by playing through (and more importantly having others play through for the first time) and seeing what they try to look at/etc.

It makes me think a cool thing to do would be an interactive text adventure with an online component, almost like a wiki. You play it, and if you do something that has no corresponding response (e.g. looking at the potted plant) then the game prompts you to enter the description of the thing. This could obviously be vandalized, so it'd have some moderation aspect, but I think the most important part would be that it would directly and in real time tell the author what people are actually trying to do.

Starbuck_II
2012-09-12, 03:37 PM
So, does anyone here have any experience with creating/writing Text-Based Adventures? I've been tasked with putting one together (with my friend), but I'm a bit stumped on where I want to start. For starters, what formats work well in an interactive, text-only medium?

Well, here is an example text adventure: http://forum.rpg.net/showthread.php?646942-Let-s-Read-XS2-Thunderdelve-Mountain

We are currently playing it at the site.

Neftren
2012-09-12, 07:48 PM
Well, here is an example text adventure: http://forum.rpg.net/showthread.php?646942-Let-s-Read-XS2-Thunderdelve-Mountain

We are currently playing it at the site.

Awesome, thanks for the link. I'm reading through it now.


As for actually writing the whole thing, I'm wondering about the setting right now. The initial idea I've been kicking around is some form of an insane asylum.

Are these things typically written from the perspective of one or a set of characters that are player controlled, or am I more designing a world in which people can bring their own characters in (a la MUD style)?

NichG
2012-09-12, 09:19 PM
Awesome, thanks for the link. I'm reading through it now.


As for actually writing the whole thing, I'm wondering about the setting right now. The initial idea I've been kicking around is some form of an insane asylum.

Are these things typically written from the perspective of one or a set of characters that are player controlled, or am I more designing a world in which people can bring their own characters in (a la MUD style)?

Well, its much more writing to let people bring their own characters in, unless you make a strong separation between the characters they bring in and the actual game content (as most computer RPGs do to solve this issue).

Honestly though, the interactive fiction genre has a very strong experimental side, so you'll find stuff out there written in all sorts of nearly inconceivable styles and ways. I read a paper describing interactive fiction where the story was told in reverse order, or where stories were told asynchronously and not monotonically (i.e. older facts were repeated dynamically), and other such stuff. I've seen ones where you jump characters at certain points. I would not be surprised to find one where you are the narrator, but I've yet to see that in particular (I guess you'd be writing commands like 'Eliza look at flower pot' rather than 'look at flower pot').

Neftren
2012-09-14, 09:59 AM
Well, its much more writing to let people bring their own characters in, unless you make a strong separation between the characters they bring in and the actual game content (as most computer RPGs do to solve this issue).

Honestly though, the interactive fiction genre has a very strong experimental side, so you'll find stuff out there written in all sorts of nearly inconceivable styles and ways. I read a paper describing interactive fiction where the story was told in reverse order, or where stories were told asynchronously and not monotonically (i.e. older facts were repeated dynamically), and other such stuff. I've seen ones where you jump characters at certain points. I would not be surprised to find one where you are the narrator, but I've yet to see that in particular (I guess you'd be writing commands like 'Eliza look at flower pot' rather than 'look at flower pot').

Hmm. So I've settled on having the player control just one character (from a first-person perspective).

I'm wondering how the passage of time works? One of my friends suggested putting in dream sequences and the like, but personally, I'd find that rather boring to read. It seems kind of gimmicky and I don't want to railroad a player along by forcing them to read a big dream scene, or play through one.

NichG
2012-09-14, 12:26 PM
Hmm. So I've settled on having the player control just one character (from a first-person perspective).

I'm wondering how the passage of time works?

I'm not sure I understand this question in the context of the rest of the post about dream sequences. How are these connected?

Generally what I've seen are that these games are split into sub-modules. Within a given sub-module, there may either be a timer or not, where a timer means that you have to complete things within a certain number of moves or you fail. Generally speaking, timers make for a much tenser but also much less forgiving game, especially if things like 'look' take a tick off the timer! Personally I don't care for timed puzzles in text games, but it is something that you can do.

Otherwise, its usually assumed to be more or less static time within a module. Then when you complete a bridge event (something that logically takes you to the next module), the time and setting can change. Usually you don't go back to the previous module, at least not in its old state, so you also have to make sure that a precondition of leaving one module is that you have everything from there you will later need to solve subsequent modules.

Many games are just one static-time module, so the entire area is always explorable and you might have to backtrack to find an object. This leads to a sort of pack-rat behavior of 'I'd better pick up everything because it might be needed to solve a puzzle later'. If you do have bridge points, you can basically reset the player inventory at those points (they went and put the tuna can, the rolled up newspaper, and the candle they got from the ruins of Al'zamir into their trunk o' memorabilia, and now they're carrying a flashlight and a shovel they didn't have before because, hey, next step is to dig up some graves at night).


One of my friends suggested putting in dream sequences and the like, but personally, I'd find that rather boring to read. It seems kind of gimmicky and I don't want to railroad a player along by forcing them to read a big dream scene, or play through one.

If you do dream sequences, I'd suggest making them actually interactive though, not just big blocks of text. Tomorrow for instance does this with flashbacks, where you actually play through the flashback. The outcome is predetermined, but for most segments of the game due to the combinatoric effect you need that to be the case anyhow.

Its a computer game so railroading is really inevitable to a much larger degree than any tabletop game. Keep in mind that each 'real' choice the player has roughly doubles your workload, and this compounds exponentially. The key thing is to place those few 'real' choices you can afford at the right spots to make each option interesting - you generally don't have the development time to spare on enabling boring event sequences just because theoretically the player could choose to do something that takes things in that direction.

Neftren
2012-09-15, 02:29 PM
I'm not sure I understand this question in the context of the rest of the post about dream sequences. How are these connected?

Well, for instance, let's say you have a "day" of in-Adventure time, and at the end of each "day" the player goes through a dream sequence connecting it to the next "day" (roughly speaking). I didn't come up with the Dream Sequence idea, and my partner wasn't super clear about it, so I'm not 100% sure what he means by it either.

I suppose my question is how to reflect the passage of time?


Generally what I've seen are that these games are split into sub-modules. Within a given sub-module, there may either be a timer or not, where a timer means that you have to complete things within a certain number of moves or you fail. Generally speaking, timers make for a much tenser but also much less forgiving game, especially if things like 'look' take a tick off the timer! Personally I don't care for timed puzzles in text games, but it is something that you can do.

Otherwise, its usually assumed to be more or less static time within a module. Then when you complete a bridge event (something that logically takes you to the next module), the time and setting can change. Usually you don't go back to the previous module, at least not in its old state, so you also have to make sure that a precondition of leaving one module is that you have everything from there you will later need to solve subsequent modules.

Ah, okay, that makes sense. So time is relatively fixed within each submodule, but by completing submodules, the character progresses throughout some timeframe (be it a month, or a year, etc.)?


Many games are just one static-time module, so the entire area is always explorable and you might have to backtrack to find an object. This leads to a sort of pack-rat behavior of 'I'd better pick up everything because it might be needed to solve a puzzle later'. If you do have bridge points, you can basically reset the player inventory at those points (they went and put the tuna can, the rolled up newspaper, and the candle they got from the ruins of Al'zamir into their trunk o' memorabilia, and now they're carrying a flashlight and a shovel they didn't have before because, hey, next step is to dig up some graves at night).

Yeah I'd like to avoid packrat behavior. So far my initial premise involves a set of role reversals, with the main character having one psychological condition initially, and a sort of therapist that is perfectly sane, and over the course of the narration, it flips, and the therapist goes insane, and so on...

With that in mind, I was thinking of dropping clues all over the setting. For instance, examining the shelves of a bookcase might reveal a number of titles involving _____ subject, which is partially indicative of each character's state of mind.


If you do dream sequences, I'd suggest making them actually interactive though, not just big blocks of text. Tomorrow for instance does this with flashbacks, where you actually play through the flashback. The outcome is predetermined, but for most segments of the game due to the combinatoric effect you need that to be the case anyhow.

Its a computer game so railroading is really inevitable to a much larger degree than any tabletop game. Keep in mind that each 'real' choice the player has roughly doubles your workload, and this compounds exponentially. The key thing is to place those few 'real' choices you can afford at the right spots to make each option interesting - you generally don't have the development time to spare on enabling boring event sequences just because theoretically the player could choose to do something that takes things in that direction.

Hmm, alright. I think I'm going to try and drop the dream sequences if possible. It doesn't seem like it would be an integral part of the story if I'm just using them to connect two submodules.


I'm wondering how long I should make this now...


Edit: I'm wondering a bit about incorporating a no-win scenario. Has this already been done? For instance, I'd like to portray the main character slowly descending into insanity (a la Call of Cthulhu style perhaps)...

Ravens_cry
2012-09-15, 03:33 PM
Personally, I dislike timed (or limited turns) adventure games.
It's very, very easy to create an unwinnable state, and it's no fun to have to restore, if you even have a workable restore, solve puzzles and situations you've already done, all because you didn't look under the overturned coffee mug for the allen key you need four in-game days later.

Nepenthe
2012-09-15, 07:44 PM
No-win scenarios are a great way to improve player agency and provide a sense of urgency, as long as the player knows that it's a possibility. I suggest including a "bad" ending very early in the game; like a game over for immediately walking away from the asylum rather than entering it. However, it is extremely important that these no-wins eventually reach an ending. You don't want to leave the player wandering around indefinitely with no way to progress, but it's perfectly ok to reveal that something the player did in the first five minutes ruined his chances for success hours later. Classic text adventures were not intended to be completed on the first try.

EDIT: To clarify, it should be possible to discern the exact act that caused the no-win. It doesn't have to be obvious, but it shouldn't be completely random either.

Adoendithas
2012-09-16, 09:32 AM
If you haven't tried it already, I recommend Inform 7 for your first few games. Graham Nelson implemented a basic D&D-like game ("Woodpulp and Wyverns") as an example and released the source code on the site.

But I agree with what was said before. Most (of my favorite) text adventures do something new--the spell system in Enchanter, the robots in Suspended, almost everything in Adventure and Zork. It's hard to make an interesting game unless you have a creative idea for how it's going to be different than the other ones out there.

Neftren
2012-09-17, 10:50 PM
No-win scenarios are a great way to improve player agency and provide a sense of urgency, as long as the player knows that it's a possibility. I suggest including a "bad" ending very early in the game; like a game over for immediately walking away from the asylum rather than entering it. However, it is extremely important that these no-wins eventually reach an ending. You don't want to leave the player wandering around indefinitely with no way to progress, but it's perfectly ok to reveal that something the player did in the first five minutes ruined his chances for success hours later. Classic text adventures were not intended to be completed on the first try.

EDIT: To clarify, it should be possible to discern the exact act that caused the no-win. It doesn't have to be obvious, but it shouldn't be completely random either.

Hmm, but why would walking away from the asylum be a game over? Wouldn't that imply that the player has no interest whatsoever in actually playing through the text adventure?

I was thinking that the Professor might offer you pills (e.g. Ibuprofen) to help you "feel better," where the pills would be some type of medicine for crazies. Perhaps continually taking the pills would further worsen your condition... I think I want to structure this as an exercise in how long it takes before the player figures the whole thing out and stops following the Professor's advice (or takes action against the Professor...).


If you haven't tried it already, I recommend Inform 7 for your first few games. Graham Nelson implemented a basic D&D-like game ("Woodpulp and Wyverns") as an example and released the source code on the site.

But I agree with what was said before. Most (of my favorite) text adventures do something new--the spell system in Enchanter, the robots in Suspended, almost everything in Adventure and Zork. It's hard to make an interesting game unless you have a creative idea for how it's going to be different than the other ones out there.

Hmm, well, I don't have all the time in the world to play through all these text adventures. I will definitely look at some of them, but in the meantime, could you maybe describe a bit more about the spell system in Enchanter, robots in Suspended, and so on?

I haven't really given much thought to game mechanics actually, so I'm now somewhat interested in how they might be implemented and applied.

Thrawn4
2012-09-18, 09:48 AM
For inspiration:

Choice of Game
http://www.choiceofgames.com/category/our-games/
They are basically doing what you want to do, albeit with a different script code (which is free, however).

Project Aon
http://www.projectaon.org/
Because Joe Dever created game books vastly superior to whatever Ian Livingston ever could put together. Coherent story, rich atmosphere.


General advise:
I've already tried to do this once or twice. Plan ahead, take care that there are actually different ways of solving problems, maybe create two routes so that people may like to play it again, don't let players die randomly but because they had it coming
and most important: Have someone to see your progress, e. g. test gamers. You need feedback in order to stay motivated. Putting ideas together is nice, but the actual work is tiresome.

Neftren
2012-09-22, 10:18 PM
What are everyone's thoughts on a text-adventure with no win condition? I'm thinking of structuring this somewhat like a Call of Cthulhu game, in which the only objective is to survive as long as possible...

NichG
2012-09-22, 10:23 PM
What are everyone's thoughts on a text-adventure with no win condition? I'm thinking of structuring this somewhat like a Call of Cthulhu game, in which the only objective is to survive as long as possible...

I think its a bad idea if there's no specific ending line, i.e. if you just go until your resources drop to zero. A bad ending is more palatable if there's actually a purpose behind it.

Neftren
2012-09-22, 10:37 PM
I think its a bad idea if there's no specific ending line, i.e. if you just go until your resources drop to zero. A bad ending is more palatable if there's actually a purpose behind it.

Well, the basic premise is that there's this crazy professsor, who is trying to convince the player that (s)he (the player) is going insane. There would be some behind the scenes counting, but this would function in a manner similar to Call of Cthulhu, where the player's sanity is kept as some value decreasing towards zero. The final scene in the game would involve a choice between following an instruction or walking away from it.

erikun
2012-09-23, 04:54 AM
Hmm, but why would walking away from the asylum be a game over? Wouldn't that imply that the player has no interest whatsoever in actually playing through the text adventure?
That certainly sounds like a good reason for the game to be over. :smallsmile:

Adventure games have traditionally been very big on giving players all the options feasable in any given situation. Unlike a video game, were getting into your car isn't an option because of "But Thou Must!" or because the car doesn't interact, an adventure game that does allow the player to get into the car and drive home makes it feel more realistic and more like an actual setting or world.

Also, as Nepenthe mentioned, it does highlight to the player that the game can end at any time due to player choice. The player is going to be less upset about a game over if they felt like it was due to a choice they made.


Well, the basic premise is that there's this crazy professsor, who is trying to convince the player that (s)he (the player) is going insane. There would be some behind the scenes counting, but this would function in a manner similar to Call of Cthulhu, where the player's sanity is kept as some value decreasing towards zero. The final scene in the game would involve a choice between following an instruction or walking away from it.
If there is a "final scene" then your game does have an ending and a "win" condition - getting to the final scene.

If instead you're just talking about an infinitely long game (or at least one that is longer than possible moves the player can take) and the "final scene" is the player hitting 0 sanity... I'm not so sure. There are rougelikes (video game) and other video games that have no defined end, but they are capable of continuing indefinitely due to acquiring rewards that can extend playtime. I don't think that a game - adventure game or no - would be that popular if there was no end scenario. In order to do so, you would either need to give the game no goal (leaving the player to wander until they drop dead) or give a goal without ever the possibility of completing it - a frustration to some, certainly.

NichG
2012-09-23, 11:42 AM
If instead you're just talking about an infinitely long game (or at least one that is longer than possible moves the player can take) and the "final scene" is the player hitting 0 sanity... I'm not so sure. There are rougelikes (video game) and other video games that have no defined end, but they are capable of continuing indefinitely due to acquiring rewards that can extend playtime. I don't think that a game - adventure game or no - would be that popular if there was no end scenario. In order to do so, you would either need to give the game no goal (leaving the player to wander until they drop dead) or give a goal without ever the possibility of completing it - a frustration to some, certainly.

Roguelikes can do this because they have procedurally generated content. A text adventure could in principle do this, but its a lot harder to make it feel different I think since so much about a text adventure is in the richness of description and the actual prose written by the author, something much harder to have a computer do for you.

Think of it like getting stuck in a puzzle game. You know you're stuck because nothing you do gives you any new messages, and the puzzle world becomes eerily static until you find the next thing you have to do to advance the game. You can sort of tell you're frozen in a single moment, and its frustrating - you can feel like you've exhausted all the options other than whatever wall is in front of you at the time, and that playing the game more won't actually help you get anywhere. If you run out of content in your text adventure and the character is still sane, it will be like that situation, where you basically have an unsolvable puzzle that you're stuck before. It might be a good way of making the player feel their character's insanity, but its certainly not fun.

I think as long as there is a discrete final scene though, where the game ends 'one way or another', its fine. That is as erikun pointed out, still a 'win condition' for the player, even if the ending isn't happy. That doesn't necessarily have to do anything with the PC's background sanity score though.