next »

[Drof]

You enter a thread where Drof says 'What if' a bunch

I've had some really good experiences in MUDs and other largely text-based games. But, I was wondering what are the common pitfalls of this area? What really are the reasons this isn't super mainstream? I really want to try to tackle this, but I can't quite put my finger on why I can relax by reading a book, but can't stand a non-visual Choose Your Own Adventure.

If the walls of text were too much on people, what would it be like if the game made sure no description went on too long without player input? Even if it was just a super simple, <Press any key to continue> type deal we see in dialogue boxes for old RPGs?

If the text appeared to disgracefully, what if text appeared character-by-character, developing their own cadence?

If the play was too silent, what if there was music? Sound bytes for each character of the words coming in, atmosphere in the background so that the wind doesn't have to be described in the text? Effects for actions to give them more credence?

If it's annoying to type in a specific command, what if your options were controlled through number keys? Common actions like movement bound to WASD? Discard the need to press enter after entering a command - indeed, getting rid of typing out phrases entirely?

What if the text was formatted and displayed in such a way that they'd be screenshot worthy? So that if someone saw a trailer they wouldn't just say, "Whatever, just a text-adventure. They suck."

Any good examples of good text-based games?

[Ludipe]

Hey Drof!

I think you raise some interesting points. With text games you can build deeper systems because you don't need sprites or 3D models, we can come up with crazy things that we couldn't do (or at least it would take a lot of time) in a regular game.

I think there are two main reasons for it not being too popular (at least most of the time):

-Giant walls of text, instead of shorter messages.
-Ugly, or way too simple, presentation of the interface.

But there are already some quite famous modern text games, like the "Sorcery" saga or "80 days". I've also heard good things about "Sun Dogs" and "This book is a dungeon".

I once made a short text game over a jam ("Who the hell is Tom?"), it's super short and features little interaction, but it's just another way to display text (I'm working on an extended version of it as a sideproject).

Hope more people can contribute with examples and we can have a good debate about text games

[Drof]

Thanks for showing me those!

It seems like a pretty common theme here for the popular examples mentioned is that they aren't just text-based. They have simple, easy to digest visuals which go along with the text. And that makes sense - you can design an interest curve where the peaks are text and the troughs are abstract visuals.

Even in the days of MUD, there were some visuals like ASCII maps which you could use to move around, but it often felt like it wasn't really hitting that mark because every square you moved on had a different room description.

My favourite example of a classic Choose-Your-Own-Adventure is Howling Dogs. Without getting too much into the debate, I find it feels less like a game and more like a book that you can change the passages of. It's a literary experience that places you as one of the characters.

I'm wondering if classic text-adventures or MUDS could be modernized in a way that maintains it's homage while tackling the issues mentioned, or if there's not much to try down that path. I thought ASCII-based roguelikes would always appear unpalatable until I saw Cogmind.

[rav10li]

I've thought a lot about text-based games recently (having just re-played some of the old Infocom stuff). The games are great if you have the patience for them, but I'm not sure the modern market does. The freeform parser model is one I really really like – i.e, enter in your own text via keyboard and see if the game recognizes it (rather than choosing from a list of predetermined answers). But those tend to be rather obtuse and overly limiting, meaning if you aren't already invested and/or knowledgable about how freeform parsing usually works and what it responds to, you'll get frustrated quickly.

How to do freeform parsing in a way that allows for natural language and is flexible enough to interpret various response styles... that would be nifty. Maybe it's already been done??

As you said Drof, most modern "text-based" games utilize some form of imagery to keep interest and/or frame of reference, especially on mobile devices where the implication is that players are usually only playing in bursts.

[JWK5]

[valrus]

]I really want to try to tackle this, but I can't quite put my finger on why I can relax by reading a book, but can't stand a non-visual Choose Your Own Adventure.

For text games with a puzzle structure, I think part of it might be pacing, and expectations about pacing when reading text.  Games that have a puzzle structure involve some amount of time where you're stuck, but in text I find it more frustrating.  (Beyond the natural frustration of parser problems.)  Maybe it's that we expect text to keep flowing and not repeating itself.  It's a bit frustrating to keep walking down the same street in Monkey Island or something because you're stuck; it's more frustrating to keep reading the same description of a room.  Looking at the same landscape fifty times is ok, reading the same paragraph fifty times is awful.

In more choice-based adventures, on the other hand, I think there's some aspect of Fear of Missing Out.  Like if you don't choose just the right thing (and it could be a seemingly irrelevant thing) you're going to miss out on the really cool part.  Take 80 Days -- a good game, mind, but there are definitely sequences where one route is really interesting and another pretty uneventful, or sequences where there's a significant plot buildup but nothing happens because only one of three choices (and you can't be sure which) leads to the conclusion of that plot.  This is another thing that goes contrary to the expectations of readers.  I can relax reading a book because I know I'm not going to inadvertently screw up somewhere and somehow get a non-ending.  (Eg., "Sorry, you should have chosen to say 'You don't say!'  Not having done so, a conversation doesn't happen and you don't learn the clue that reveals the murderer.  Everyone goes home and nothing happens.")

I think the most important thing in a text-based game (well, beyond good writing) is the idea of Something Interesting Happens Regardless.  Sorcery! is good at this; you don't end up with many paths that get you stuck doing the same thing again and again trying to get out of a loop (problem 1) and there aren't many story arcs where one possibility leads somewhere really interesting and the other is just "In the end, nothing happened" (problem 2).  (Well, since Sorcery! has easy rewind, maybe there are, but they don't matter as much as in 80 Days.)

Put another way: A problem of text-base games is when the punishment for non-optimal play is that the player has to read boring, repetitive, or unsatisfying text.  That's compounded if the game isn't very fair (obtuse text puzzles, irrelevant choices with bizarre consequences, etc.).  In that case, I'd much rather read a book, because when a book gets boring, repetitive, or unsatisfying it doesn't pretend like it's my fault

[valrus]

How to do freeform parsing in a way that allows for natural language and is flexible enough to interpret various response styles... that would be nifty. Maybe it's already been done??

Y'know, come to think of it, there are some techniques that are possible now that I don't think have really been used much.

Like assume we've got analytics running in the background; you could determine what people are typing in at what point in the game, and in each update ease the frustration curve by adding in good solutions that you hadn't thought of, common alternate expressions of the intended solution, tailored responses to nonsolutions, etc. 

Possibly some people do this to some extent, but it'd be interesting as a design *methodology*.  Don't spend your time budget by dreaming up responses to things you think people might enter; you're probably spending a lot of time on things people will rarely type and missing a lot of common ones.  Instead, do it empirically.  Start out with barebones interaction, get what people are actually typing, then concentrate on responding well to the most common responses.

Another possibility, and I don't think anyone's done this, but it could work.  Get a bunch of actual user inputs like the above and the responses that you think they should get.  Don't parse the input in the old-school way; treat it as an NLP classification task.    Like say that you have a set of 150 user inputs that cover 97% of what people are actually typing at decision point 158b, and you decide there are 12 consequences you want them to have.  Convert all the inputs to a vector representation (like in a high-dimensional semantic space).  Then, when the user types an input that you don't already have covered, convert that to a vector representation, and classify it according to which of the 12 consequences it might have.  (Like take the angle between it and the known vectors and use that to do a k-nearest-neighbors classification.)

It might be awesome, or it might be a hilarious disaster, but worth a try maybe.

[rav10li]

Put another way: A problem of text-base games is when the punishment for non-optimal play is that the player has to read boring, repetitive, or unsatisfying text.  That's compounded if the game isn't very fair (obtuse text puzzles, irrelevant choices with bizarre consequences, etc.).  In that case, I'd much rather read a book, because when a book gets boring, repetitive, or unsatisfying it doesn't pretend like it's my fault

I think this is a very good point, maybe the ultimate pitfall of text-based games that rely on specific player interaction to move forward. If the player chooses "wrong" for whatever reason, the consequences must still be narratively interesting. But then, what's the incentive to try and choose "correctly" as a player when everything works out, regardless of what you do?

It just seems like players need to be invested in the story and/or game arc before they start playing... otherwise a handful of "wrong" choices with negative (in this sense meaning narratively boring) consequences will very quickly eliminate any player enthusiasm. Text-based games don't usually have the luxury of quick-stimulus mechanical gameplay to keep the player entertained. Suddenly the reason for so many modern text games incorporating fancy motion graphics makes sense 

[s0]

tbh i think it's because reading, and especially reading prose, takes more effort than looking at pictures. if you look at the text based games that gotten some kind of hype outside of the usual niches like the aisle, gun mute, a dark room etc., one thing they share in common is concise writing.

[Drof]

For text games with a puzzle structure, I think part of it might be pacing, and expectations about pacing when reading text.  Games that have a puzzle structure involve some amount of time where you're stuck, but in text I find it more frustrating.  (Beyond the natural frustration of parser problems.)  Maybe it's that we expect text to keep flowing and not repeating itself.  It's a bit frustrating to keep walking down the same street in Monkey Island or something because you're stuck; it's more frustrating to keep reading the same description of a room.  Looking at the same landscape fifty times is ok, reading the same paragraph fifty times is awful.

This is a big problem I have with the standard adventure game formula in general. The appeal of these games usually is to solve problems using unconventional means. If you need a key to a door, you can bet that key is underneath a sleeping tiger you have to lure with meat. It requires you to pay attention to your surroundings and rewards you for following logic. Which is great!

That is, until you ineveitably come up with a different solution to the puzzle. I can't count the number of times where I would come up with a solution that I thought would work, get frustrated when it wouldn't, and then 'solve' the puzzle later as the game condescendingly gives it to me. I think that's just the nature of practical puzzle solving.

The thing is, adventure games don't have to emphasize this. I don't even really think 'puzzle' is the right word we're looking for here. You could do Final Fantasy style combat in text based, just as you could do reflex-based or logical puzzles. It's the legacy which is holding most text-adventures back. Adventure game like Zelda, rather than Monkey Island.

How to do freeform parsing in a way that allows for natural language and is flexible enough to interpret various response styles... that would be nifty. Maybe it's already been done??

This may seem cynical, but related to the above point, I'm not sure if this will ever work. I think it's only 50% related to the actual commands - it's more to do with their interactions. How do I know
>pick up nail file
>use nail file on guard

means stab them in the throat instead of do their toenails? Some of these fires you can put out without too much effort, but even with analytics as Valrus said, this will get waaay out of control for something which, in my opinion, isn't better than having a decent multiple choice system.

If you want to keep the typing part, there are games which do exactly that. In Emily is Away, you select dialogue options with 1-3, but then you mash the keyboard and see your response, character-by-character. Actually a pretty good game to recommend a look at, even if I think it has a lot of room for improvement.

In more choice-based adventures, on the other hand, I think there's some aspect of Fear of Missing Out.  Like if you don't choose just the right thing (and it could be a seemingly irrelevant thing) you're going to miss out on the really cool part.  Take 80 Days -- a good game, mind, but there are definitely sequences where one route is really interesting and another pretty uneventful, or sequences where there's a significant plot buildup but nothing happens because only one of three choices (and you can't be sure which) leads to the conclusion of that plot.  This is another thing that goes contrary to the expectations of readers.  I can relax reading a book because I know I'm not going to inadvertently screw up somewhere and somehow get a non-ending.  (Eg., "Sorry, you should have chosen to say 'You don't say!'  Not having done so, a conversation doesn't happen and you don't learn the clue that reveals the murderer.  Everyone goes home and nothing happens.")

I think this is a seperate issue really. If anything, it's much easier to tackle these problems in text-based. But choice-based is very appealing from a purely text-based medium, since it's really the cheapest opportunity you're going to get to implement it.

The breaking point for me in terms of choice-based was that every choice has to feel like it matters, but they don't have to feel like they'll change the ending of the game. The Witcher 3 did this really well. The actual 'ending scenarios' are relatively straightforward, but who you befriend and what situations spiral out of control on the way makes it feel so personal anyways.

I actually was thinking something along these lines when I was dreaming up a text-based which looked pretty. But I'm thinking something along the lines of The Matrix, and maybe an emphasis on ASCII art.

Let's face it - not having a screenshot that immediately grabs the audience is going to vastly impact a game's success.

tbh i think it's because reading, and especially reading prose, takes more effort than looking at pictures. if you look at the text based games that gotten some kind of hype outside of the usual niches like the aisle, gun mute, a dark room etc., one thing they share in common is concise writing.

I think you're right about prose. It might just not gel very well with interaction. Short, utilitarian stabs of description suit play. I think the trick is to get the player to invoke the scene in their imagination.

Maybe having succinct lists are the way to go?

An Abandoned Dining room
Meals, swarming with maggots
Mahogany table, dusty
Windows, boarded
Young couple, rotting
Stairs, creaky

[Jordgubben]

I'm currently working on a game that is mostly text. I've never really been into text-adventures, CYOAs, MUDs or even VNs. My main motivation was to find a way to make games on a shoe-string budget.

The biggest challenge (in my experience so far) for text-games is in repeating mechanics (ie. most game mechanics except adventure-games). Having the same lines of text appear over and over is not just dull, but actually makes it harder to remember what just happen. Making the text more varied on the other hand makes the mechanics harder to understand. The correlation between input and output gets lost when the output is filled with aesthetic noise. In any case clarity is lost. The player is both bored and confused at the same time.

Complementing or replacing text with illustrations and icons, paying attention to layout, typography and use of colour helps a lot to improve clarity. Good visual flair is not just eye candy, it is information density.

[Drof]

Complementing or replacing text with illustrations and icons, paying attention to layout, typography and use of colour helps a lot to improve clarity. Good visual flair is not just eye candy, it is information density.

I think that's particularly true for mechanics which appear often, like combat. If the player is always thinking one step ahead, I'd guess they're less likely to read text. If a lot of combats are going to be more or less the same, then it's better for that information to be easier to parse.

Also, it can raise the tension of a scene as there are more actions and reactions per minute. Choosing 'attack' and then having to read a paragraph can kill that momentum. But choosing 'attack' and seeing a health bar go down with maybe an accompanying short stab of text keeps the player reacting.

[valrus]

One thing that I find annoying when changes are expressed textually is the need (real or perceived) to reread descriptions to see what (if anything) has changed, especially when you're stuck.

It might be nice to have color-coding on the text to indicate where something has changed.  Like when you get

> look windowsill
There is a small box on the windowsill.  The box is closed.
> look box
The box on the windowsill is hexagonal and made of red and black lacquer.  It is closed.
> shake box
You pick up the box and shake it.  It appears to be empty.
> look box
The box on the windowsill is hexagonal and made of red and black lacquer.  It is closed.
> open box
You open the box.
> look windowsill
There is a small box on the windowsill.  The box is open.

That's kind of a trivial example but you get the idea -- taking care to notify the reader that something is different this time.  It also makes it easier to go back and get a quick synopsis -- reading only the red text rather than rereading each paragraph.

[s0]

most of the better IFs use shortened descriptions of rooms you've already been to anyway.

[pelle]

I love text-based games. I even read/play some paper gamebooks now and then. I do not understand why they must add so much artwork and fluff to them, like some childrens books? Just plain text in a nice font would/should be enough. And some apps get really annoying with rolling animated 3D dice all over the screen while I try to read and enjoy the story.

I have grown tired of walls of text. I like shorter texts. Same goes for linear stories though. Never read much fiction really, but I have found almost all the fiction I read now is things like old pulp. Even when they are longer stories, there is usually not a lot of crap in the text. Just tell me what happens and let me imagine what it is like instead of forcing hundreds of words about how someone smells down my throat.

One interesting variation I would like to see more of are choice-based games that still have a world-model (like typical parser-based games). Tiny Text Adventure (on Android) is like that. You can tap noun-words and choose actions to do to them. You can navigate around a world and can pick up and use objects and so on. There are even a few old paper gamebooks that are sort of like that, with free movement instead of just a linear run through a story (Fabled Lands; although that is the only example I know of outside of some obscure Windhammer Prize entry/entries).

[Deidril]

Hi,

i ran a mud for like 15 years (from 1995 to 2010). I've been also a mmorpg player since the beginning (if Meridian 59 and Ultima Online could be seen as the beginning), so here is some feedback from my experience about the matter.

The reason people were playing my game was :

It was years 1995-2000 : there was no mmorpg, and later there was no free mmorpg. You could play muds from the school/university computer room also. Most of the players were software engineers doing their classes.

In years 2000+, most of the players were blind, or vision deficient players using text to speech software like JAWS to play online games.

There was always exceptions but basically, when the generation of software engineers completed their classes and got a salary, they left  muds for mmorpg with monthly fees, and when free mmorpgs raised, broken players left also.

From an admin point of view, games features from mmorpg were welcome (like having strategies to set up to kill a boss), but everything that slow down the game speed was rejected by players, like ascii map displaying a large world instead of having a smaller world of connected rooms. Even if you put random things so travelling from A to B through a world map is also an adventure, players prefered a smaller world becaues it's quicker to consume content and to achieve goals.

This last point, rejecting everything that slow down the game experience, made me making this personal theory :

Let's have Bob, a player that enjoy reading as much as viewing a movie. We could think that Bob would enjoy playing a well written mud as well as a skyrim or any rpg with nice gfx, but it's not the case.

Let's say there is a text based mud and a mmorpg with exactly the same content to consume (world & quests), and same rpg rule engine, except rendering is in text for one, and 3D rendered for the other, and let's say Bob try both games.

Let's say also than both games have the same community, and from a social point of view, you would enjoy people the same on both games.

Reading text and understanding things like ascii map is slower than analyzing a graphical rendered world. We play rpg games to enjoy experiences, worlds and lifes we couldn't have irl. Therefore, if Bob plays 10h the text based game and 10h the gfx based game, Bob will explore more content in the gfx based game. We could say that

 Bob->HappynessFromContentExploration(GfxGame) > Bob->HappynessFromContentExploration(TextGame).

Reading previous answers, it's a shared thought.

I have grown tired of walls of text. I like shorter texts.

I agree : walls of text, even well written, are not enjoyed. Players prefer a dynamic adventure with surprise, fun, difficulty, and content than reading a piece of art. Feedback from players were more 'Wow, that puzzle was hard to figure', or 'Wow, that fight was hard but fun', but never 'That location has nice text'.

An Abandoned Dining room
Meals, swarming with maggots
Mahogany table, dusty
Windows, boarded
Young couple, rotting
Stairs, creaky

I ended with the following display :
-=[ Abandoned Dining room ]=-
3 lines of description max.
Attributes : YOUR_HOUSE, NO_FIGHT, <location flags obvious to the player>
Exits : N S E W
Entities : Gondegor Gran<Zombie>, Ewilan Gran<Zombie>
Players: Rozanna <No guild>, Ewilan Lockert<Guardians of X>, PCName<Guild name>
Objects : a dusty table, a large window, (2x) a rotten meal

Screenshot (in french) : http://ensimud.net/images/shootnewversion/19mai_01.gif

About interaction with npc, i used everquest way, or surrounding with brackets keywords that matter, so the player doesn't test every keyword used in the npc words, trying to guess what to do with this NPC.

Example :

Gongegor Gran says to you 'Hello fellow adventurer. Welcome to the prancing sword ! I'm the best - and the only - weaponsmith around. Therefore, you'll have only very little options about my weapons : [buy] or leave my shop.

> say buy

Gongegor Gran says to you 'Fine, here is the list of weapons you can buy ...'

 Screenshot (in french again) : http://ensimud.net/images/shootnewversion/19mai_02.gif

This way, the player knows immediately the only interaction he can have with Gondegor is buying weapons. He doesn't spend time trying 'say prancing sword' to check if there is not some background stuff to learn.

In a quest, when a speech option should become available at a npc, the engine add the information as if you already asked the npc about the topic.

Gongegor Gran says to you 'Hello fellow adventurer. Welcome to the prancing sword ! I'm the best - and the only - weaponsmith around. Therefore, you'll have only very little options about my weapons : [buy] or leave my shop.'
Gongegor may know something about [the lost sword of Nardell]

That's because it's not fun to loop over all quest each time you meet an npc and check keywords.

[Drof]

(I typed out this whole damn post and then accidentally deleted it, take no. 2)

Reading text and understanding things like ascii map is slower than analyzing a graphical rendered world. We play rpg games to enjoy experiences, worlds and lifes we couldn't have irl. Therefore, if Bob plays 10h the text based game and 10h the gfx based game, Bob will explore more content in the gfx based game. We could say that

 Bob->HappynessFromContentExploration(GfxGame) > Bob->HappynessFromContentExploration(TextGame).

I think that's a really important thing to note, although I may disagree with the conclusion. Most descriptions of situations will be slower in text-based, therefore the player will consume less of the game, which conflicts with a lot of design goals. If you want to fill the player's life with adventure, you want them to go through the rooms as quickly as possible. After all, a picture tells a thousand words. But pictures can't tell every kind of story.

A game like A Dark Room really capitalizes on what makes the medium great. If we took the graphical equivalent of that game, we'd probably have to see objects sort of pop in, and we'd have to give information about the environment that would pretty much ruin the mystery of that game. That's because we're used to learning new information in each line of text, but when it comes to pictures, we usually learn everything there is to know all at once. A Dark Room leads the player on just enough so that they are learning more about the world, even though as a character these conclusions would be made pretty quickly. It uses minimalist text, but I think that's just an element to it's success, not the recipe for it.

This debate may as well be 'books vs. movies' because they have the same content disparity as text-based vs. graphical. However, if actually interacting with the fiction drains more energy than just reading a book, that might make things different.

At the heart of the matter, room-descriptions are not akin to books. Books will only fully describe a room in a paragraph if that room is very important. I feel like in games, this introduces a pacing issue and just a general engagement problem. I really like Howling Dogs, and there are barely any room descriptions in that game, and I really don't think it needs any. Instead, it takes the dramatic beats of the story and makes THOSE the rooms. This character gets angry, or sad, or hung. Choices are made because players care about the characters.

Maybe the drive to push immersion and roleplaying by giving these room descriptions is getting in the way of the actual play and fiction, and that's why these games get so tiring?

[pelle]

Thanks Deidril, that was interesting info.

I think the comparison book vs move vs graphical vs text game must be considered while thinking about how very few, if any, books have been word-for-word translated to movies. The movie and book of the same story are often vastly different, and for good reasons. I would not expect Fallout or Quake to work as text-only without massive changes. But the fact that many people still enjoy books at least to me implies that it should be possible to find one type of text-only games that can reach out to a larger audience than what current text-only games do, even if that game has yet to be created (or it was created but not marketed enough to the right people). Many probably overlook just how amazing things your brain can think of when you see a few words rather than a 3D-rendering of a scene?

Sometimes I have felt like the old text interactive fiction (and MUDs?) were just substitutes for not-yet-created graphical games. Even the more recent wave of hobby i-f designers were very quick to start including graphics in their stories when popular i-f engines started to support it (I base this on looking casually at entries in the annual i-f compo; I have not done any proper statistics). It's like both many players and many designers were doing text-only simply because of a lack of possibility of doing graphics, rather than that they wanted to make text-games. It made me a bit sad when I started seeing designers jump from text-only to graphical adventures for the if-comp or when I see gamebooks adopted to tablet apps that are very heavy on graphics instead of relying on just the text like books for adults tend to do.

Imagine if books were almost all written by people that wanted to make movies but could not afford a camera, so they tried to give the movie-experience in text instead? Maybe there is a degree of that sort of issue with most text-only games today?

There are text-games written by people that obviously care a lot about text and that can write well. Those are usually not my favourites though because that tends to end in quite linear games with walls of text.

[Deidril]

Sometimes I have felt like the old text interactive fiction (and MUDs?) were just substitutes for not-yet-created graphical games.

I agree 90% with you, although it took time (years) to accept that.

So what about the 10% left ?

It's not totally fair to say MUDs were just substitutes. First there is a population of players who can't play 3D games (vision deficient players). Second there is people who prefer play muds rather than graphical mmorpg for various reasons.

As far as i know, there is ONE commercial company running paying muds, Iron Realms Entertainment, including Midkemia, licensed from Raymond Feist's Rift War. Adapting a book license into a text online game is maybe something worthy although i have no idea of what could be keys to a product that could be appreciated from readers. I'm not sure there is enough players around for another. There is muds with donation and cash shop also but they are ran by one programmer (or a very small team) and have only one product.

Drof, i also agree with you. Games you describes are indeed using text in a smarter way. Considering today video games industry, they are games thought to be text based and are new experiences.

Muds were started because, basically, any wannabe software programmer could download the basic game engine, implement few changes to match its ideas, and run it over any linux computer connected to internet. Also, there was times where universities were ok to let students run such games. Also, most of us when we were running muds have no clues about how running a player community, balancing games, review new zones ,etc ... Everything was student work, or should i say it, very poor quality.

[Drof]

I think the comparison book vs move vs graphical vs text game must be considered while thinking about how very few, if any, books have been word-for-word translated to movies. The movie and book of the same story are often vastly different, and for good reasons. I would not expect Fallout or Quake to work as text-only without massive changes. But the fact that many people still enjoy books at least to me implies that it should be possible to find one type of text-only games that can reach out to a larger audience than what current text-only games do, even if that game has yet to be created (or it was created but not marketed enough to the right people). Many probably overlook just how amazing things your brain can think of when you see a few words rather than a 3D-rendering of a scene?

I think this is a part of why text-based is botched. People write the game like it's a book but give enough information for the player to make decisions about complex contexts such as environmental clues, or just to incite immersion. I don't think in practice that actually follows through - book storytelling doesn't translate to that scenario very well.

It's not totally fair to say MUDs were just substitutes. First there is a population of players who can't play 3D games (vision deficient players). Second there is people who prefer play muds rather than graphical mmorpg for various reasons.

This is really at the heart of why I started this thread. What benefits from game design can we get from text-based?

To speak in terms of MUDs, specifically, I found that the two things that appeal to me are:

• Descriptions can bring your attention to a lot of the finer details of an environment. The sounds, smells, tastes, and small vignette scenes which would be ignored in a graphical setting. Not to mention, that freedom for the creators can really broaden the creativity of these locales.
• Players themselves can roleplay with extremely high fidelity. Not all MUDs do this, but being able to say, "Drof readjusts his glasses nervously, and says, 'Well, I'm not sure about that'" is MUCH MORE POWERFUL than a /dance emote.

I think the former doesn't really appeal to many others, and I don't think really makes up for the pacing problem. The latter we're pretty much only going to see in contexts where humans can interpret these emotes, which means they only will ever be in MUDs, MOOs, etc.