a collaborative game design model

[coder]

The brain wrote:

Reading this I got a kind of weird but interesting idea. There's a lot of writing involved, but right now creating erotic games (or let's keep it to interactive stories, for now) still seems like an individual effort. The stories remain small or limited. I feel collaboration could result in larger stories, or to put it into context: more branches. Of course the difficulty of collaboration is that it requires a lot of extra effort in organization.

So the idea that quickly entered my mind was this: what if you create a wiki-like structure for creating an interactive story. Basically give anyone the power to easily edit the story and its interactions. Obviously such freedom needs groundrules and active moderation, but for now it's more the concept that seems interesting.

Even for the simplest interactive story model (a tree structure, no loops, no variables) I'd expect you could get interesting results because of the size and branching factor that a single person couldn't achieve (or not as quickly). That is, of course, if it works... Vandalism could be a major problem, but in a true wiki nature that should be reversible by anyone (although a 'history' is not as straightforward in a tree structure, where complete branches may have been deleted). Another potential problem is inconsistent writing, different branches written by different people may have characters acting inconsistently with the 'global truth' (for example, personalities).

Adding static graphics (like Chaotic's games) would be a next step, but to also make that a collaborative process would certainly have difficulties (for example, inconsistencies are much more apparent in graphics).

That said, it might be a nice experiment (even text-only). I'd love to hear thoughts on this (I might be motivated to conjure up a prototype), although a large discussion should probably go into a separate thread (for now it's just an idea I found interesting :P).

[coder]

That's a very intriguing idea. I don't think a wiki is the best form though. Something like an open source model would be better:
I think you would need:
a small team of maintainers
a way to make suggestions (write new branches)
an evaluation model (I was thinking of a voting system and added discussion)

Creating a game-engine is not hard. I'd prefer html and javascrïpt, since it's cross platform and easy to expand. I already have made something which is quite easy for creating interactive stories with pictures.

We also could use the tads engine, but I don't know how hard it is to write a scrïpt and I never got it to show any graphics before. And personally I never really liked text adventures. To much typing and to much guessing which commands to use. From the coding point of view it also takes a lot more work.

For creating a game creation platform, you would need a website. Suggestions on a form for it would be very welcome.

[TheBrain]

My main thought in suggesting a wiki-like nature is having a low threshold to contribute, basically 'register/login -> change'. Having a downloadable editor of some kind raises that threshold pretty high. This could be both a good thing or a bad thing.
The good being that it probably weeds out a lot of spammers (I'm imaging 13 year old boys who would find it funny to create shortcuts to sex or inappropriate sequences, etc.), while the motivated people wouldn't mind it. Another good thing might be that it's easier to create a graphical interface with more options (doing that in a browser would require stuff like ajax).
The bad of course is that it also might repel potential 'casual' contributors, the kind of people who wouldn't mind fixing spelling errors or adding short sequences.

But then maybe the two models wouldn't have to be mutually exclusive, there's nothing that would prohibit having a simple browser based editor and an advanced downloaded editor. It is just more work to do both.

Other thoughts:
- Maintainers: In my mind these would be little more than active contributors given a few extra privileges (permanently banning abusive users, stuff like that).
- The text adventure stuff: I agree, I was more thinking in a direction like Masq. (not that I dislike text adventures particularly, but in my mind it is simply to elaborate).
- Graphics: Like I said in my original idea, collaboration on such a thing would be difficult. Taking Chaotic's work as an example, the images result from posing one or more characters into position (with poser/daz3d, that sort of software), photoshopped onto a background and maybe also touched up with postprocessing effects. Needless to say, this is a complex procedure. But if you want collaboration to work, you'll need some other person to be able to access the sources of that image (instead of just the end result). Now I'm not really experienced in that area, but it seems to me it can't be totally automated without creating some form of an editor for it (but I'd love comments from someone who is experienced in that stuff). And without automation I'd think you're quickly subject to the creativity of the contributors, which can lead to inconsistent graphics (unless in some way the graphical contributors cooperate and use the same techniques).

[ExLibris]

There are a couple of collaborative interactive story sites already established if you want to see how such a thing could work in practice

See here:
http://editthis.info/choose_your_own_adventure/Main_Page
and here:
http://www.chyoo.com

As far as sharing files for collaboration goes, would something like Dropbox (http://www.getdropbox.com/) be useful for that?

As for using a text adventure authoring system, I know there's a library for TADS that allows you to create CYOA adventure games and I've played CYOA ADRIFT games as well. But I'd agree that it's probably more complex than you need if you're not going to be dealing with things like inventory and so on. Writing IF doesn't lend itself well to open collaboration anyway, unless one person is doing the coding and everyone else is writing the text.

[Franco]

There is also a cyoa mode for Inform. You might want to look at the Interactive Fiction Archive at http://www.ifarchive.org/. There are probably hundreds of game development systems available there. Though most of them will probably not be feasible for what you want to do.

You also might want to look at the Quest Markup Language at http://www.questml.com/.

It might be hard to include point and click with the Tads or Inform cyoa modes. I am not sure how easily that could be done with Qml either. I expect it can be done with all three, but the question is how easily.

[ExLibris]

There are indeed a lot of development systems for interactive fiction, but if you do go that route I'd suggest picking one that has a lot of users you can ask for advice if you run into any problems.

TADS can handle hyperlinking, but not true point and click. RAGS is mouse-driven, but again not true point and click. I've never seen anything of that nature for Inform, but that doesn't mean there isn't an extension for it.

If you're going for something CYOA oriented, I'd suggest looking here: http://www.ifwiki.org/index.php/Category:CYOA

Quest is a possibility from the little I know of it. You might also want to look at DreamPath (http://drakevision.com/projects/dreampath/index.html) or SUDS (http://www.sudslore.org/).

For true point and click there's systems like Wintermute (http://dead-code.org/home/) or Visionaire (http://www.visionaire2d.net/cms), although I haven't used either.

The main thing though is probably to determine what your requirements are. If your aim is essentially hypertext fiction, then most of the above systems would probably be more complicated than you need. It also depends on who would be doing the coding. If they're more comfortable with one system over another then that's the one you're probably going to end up with.

[TheBrain]

ExLibris wrote : There are a couple of collaborative interactive story sites already established if you want to see how such a thing could work in practice

See here:
http://editthis.info/choose_your_own_adventure/Main_Page
and here:
http://www.chyoo.com

I probably could've guessed someone already came up with the idea (I admit not trying very hard to google the concept). The first link would be pretty much the most basic implementation of the concept (i.e. just taking a standard wiki and impose a few rules). I haven't had a really in-depth look at the second one, but it seems similar, just with a little added structure to focus it on interactive stories.

That said, it's interesting to see that the concept is already out there. Like I said I looked mostly at the first site, so naturally what I'm about to write is mostly taken from looking around there (although I expect the second site is not different enough to make a big difference).

I initially wrote a post which got kinda long, so instead I'll just summarize my observations:
- The sites offer a lot of stories, some big, but many small and unfinished.
- While the 2nd site offers ratings I didn't find it helpful in finding the good interactive stories (rather many I viewed were surprisingly uninteractive).
- There is a big difference in writing style between stories: some offer only a few lines of text between choices, others literally have pages of text (I personally prefer the former).
- The stories are in principle 'unguided': I didn't notice any efforts in trying to define a background story. As such the additions of a contributor have a lot of freedom (how characters would react to certain things: their personality, is a good example).
- From checking the wiki history at different branches in stories I got the feeling there is not a lot of collaboration going on. It feels like the stories are still largely individual efforts (but I could be wrong). It seems there is no large scale collaboration going on.
- A lot of branches in the stories end in dead ends (text with unwritten choices, but not an apparent 'ending'). I found these frustrating.


Putting these observations into perspective of possibly creating such a site:
- I believe there should only be one story (at least initially). This focuses effort. A new user would play 'the one' story a few times and may decide to want to contribute, and if so he will contribute to 'the one' story (at the mentioned sites the potential of contributors would distribute among all the different stories). I feel this is essential to creating a large consistent interactive story.
- Users would have to be guided in their efforts. Freedom can be good, but too much freedom will probably result in a lot of unfinished branches. There should be global background story defined, including an actual 'roadmap' with limits on where the story can go (a high-level storyboard, locations, characters, etc.). Again to focus efforts and have the story remain consistent. Of course such a roadmap would be open for discussion, you wouldn't want contributors to feel like 'writing slaves' (I'm not sure of a good term, the IT world analog would be codemonkeys :P), it would have to remain a community based effort.


To shortly respond to the suggestions of text adventure languages: as ExLibris already mentioned they are not particularly suited to large scale collaboration. And with compilers and interpreters that need to be downloaded they are not the most accessible either. A browser based approach would be ideal to attracting potential contributors.
On top of that, the CYOA style is merely a suggestion because it is the most simple interactive story model. The actual implementation is trivial and not really the point of discussion.

[coder]

Interesting input. I'll add my opinions.

I agree with TheBrain that we should just have one story. The websites ExLibris gave are interesting, but they miss evaluation and guidance. Dropbox could be an interesting tool, especially since it's cross platform and it also allows to post on the web. I'm gonna check that one out to evaluate the usability for this idea. Google wave could also be a good tool, maybe even better. I haven't tried it for myself yet though and I don't know how stable it is. I would be preferable to do everything online, since it would be the lowest treshold to contribute and that's the main thing you want to stimulate.

My goal would be to create a game that's similar in style to arianeb and vdg. It would be easier to create a game like crystal. So that would be a good start. Though simple stuff like buying things could easily be implemented and it's also easy for the contributors to keep the overview (did you buy a camera, a dress, a dildo, did she change clothes etc.) It would make the game more interesting to play. Adding clickable areas to a picture would also help to make the game more interesting.

The hardest part would be adding graphics and have collaboration there. You would need to agree on a tool to create them and share the source. If a graphic artist could create a model in commonly used tool and some backgrounds could be chosen to start with. I'd think you could kick off and others could expand on that. I believe that's possible in the 3d design programs, but I might be wrong there.

This way you could add locations and characters as the game evolves. So I'd think the only starting point would be a character descrïption of your first character. And only in a general way, like sportive, musical, daring, shy etc. No details on what sport, which instrument she plays and so on.

Some limits on the sex I'd also deem necessary. I also don't like rape, incest or sex with tentacles, animals, children or dead bodies. That would stop me from contributing. I don't mind girl on girl sex, not at all I would say. If someone likes it, they could add gay sex. S.M. I wouldn't mind either. But I'd say those things are for later. If this idea would start off and it works.

The idea I have at this moment is a game with two extra buttons. "Something 's missing" and "incorrect grammar"in the top right corner. If you press the "missing" button you enter developing mode. You are presented with other options given by other volunteers and can either add a choice or evaluate or comment on the other choices. I'd like to see both a choice there ("you drop a pencil.") and the result ("She bend over and you have a good look at her well formed ass").
That ways others could say: No, she should say: "Do you think I'm that stupid?"

The "incorrect grammar" button would allow you to point out a spelling mistake, a grammatical error or a word with an incorrect meaning.

Later on you could also add a translate option, so the game would become multilingual. So if you prefer French the game would be French until you enter a scene that has not been translated and it would switch back to english.

[TheBrain]

coder wrote :
My goal would be to create a game that's similar in style to arianeb and vdg.

Yes, that would be pretty much the style I would like to see as well. You might add Masq to that list, although a little more complex with branches that merge after every episode. The main reason to mention it though is that it has a very consistent feel to it: playthroughs can be very different in what you experience, but the story still progresses mostly in the same way (there are some events that will almost always happen, but your perspective on it can be different).

Adding clickable areas to a picture would also help to make the game more interesting.

Clickable areas is good, but how would someone writing a new branch cope with it, given that the graphics are not there yet? Would you assume the writer can imagine what sort of graphic will be added and therefore can anticipate clickable areas? Or do you turn it around and expect the artist to try to mold the written choices into the graphic (where possible)?

The hardest part would be adding graphics and have collaboration there. You would need to agree on a tool to create them and share the source. If a graphic artist could create a model in commonly used tool and some backgrounds could be chosen to start with. I'd think you could kick off and others could expand on that. I believe that's possible in the 3d design programs, but I might be wrong there.

I think the ideal would be where the artists could download one some sort of 'project file' which could recreate one image exactly by just 'running' it. The main thought behind this is that if one artist creates a 'creepy smile' (as we've seen in Chaotics games now and then), another artist can fix it by simply editing the pose and 'running' the project again. Ideally this would all remain in the same program (but I have no idea whether you could easily add backgrounds and maybe postprocessing effects in a program like Daz3D, let alone make it so it would recreate one specific image).

So I'd think the only starting point would be a character descrïption of your first character. And only in a general way, like sportive, musical, daring, shy etc. No details on what sport, which instrument she plays and so on.

I think I would disagree on 'no details'. While you could start off with a descrïption like shy/musical, once someone establishes somewhere in a branch that she plays the piano it becomes a ground truth: to remain consistent it means that in every possible branch the instrument she plays would have to be the piano.
It can be a descrïption that expands over time, but it cannot remain without details.

The idea I have at this moment is a game with two extra buttons. "Something 's missing" and "incorrect grammar"in the top right corner. If you press the "missing" button you enter developing mode. You are presented with other options given by other volunteers and can either add a choice or evaluate or comment on the other choices. I'd like to see both a choice there ("you drop a pencil.") and the result ("She bend over and you have a good look at her well formed ass").
That ways others could say: No, she should say: "Do you think I'm that stupid?"

While I'm not entirely sure about which buttons should exist, etc. I do know that there is some danger in adding choices and that is that I do not expect a contributor to worry about combinatorical explosion. The problem is that some choices can branch out quickly and never end. The dropping of a pencil would be relatively innocent because I would not expect it to have side-effects later and so it can merge right back into the existing story. But say that a dinner story exists and someone wants to add 'accidentally' pushing over a glass of wine to stain the girls dress, this has side-effects (her clothes are ruined and need changing, plus she might become mad) and because of this it cannot merge back into the original story.

This brings me to a point I wanted to make in my last post when I mentioned the 'dead ends'. I consider a dead end to be a 'work in progress' issue and as such it shouldn't be reachable for a casual player who is not intending to contribute (yet). The upside here is that it should be relatively easy to indentify choices leading to dead ends automatically, given that a 'proper ending' is marked as such. In a non-development playing mode all these choices can be made invisible, reducing the story size, but also making it 'complete'.
Once in development mode I assume you are no longer playing just for the experience itself but also want to contribute, or at the very least you accept that the extra choices you get lead to dead ends.

[coder]

About the clickable areas: I presume that graphics and story progress together. So there's a picture with some options and the contributor wants to add a choice. Or take Arianeb, let's say you want to invite the girl in the gas station. You enter development mode, make the girl a clickable area, and offer a new branch. A graphic artist would then need to create a picture and the story continues. In this case you could also continue with the same picture.

Concerning the consistency: In development mode you should see the history. So if you threw a glass wine over the girls dress, stole a camera or bought her a new dress, you should see that info. If a new branch depends on one of those things you should add a dependency to the branch. You can only take pictures if you have a camera. So to enter that branch you would need a camera.

Also: once she plays piano in the story and she tells she can't stand the sound of a saxophone, this should be added to her descrïption. I just meant to keep it general in the beginning, until someone decides to fill in those details. So we agree on that too.

Your idea to hide the unfinished branches for the player until they lead somewhere is also a good idea. Otherwise a player would always feel he's playing an unfinished game. Well maybe you could add an alpha mode. Maybe it could be integrated in the development mode. The later would probably be the better choice. Maybe one button to enter development mode should suffice. And there you could choose to correct the english, add a translation, add a new branch etc. And also evaluate the work in progress.

I also think it's a good idea to limit the possible choices. I don't think there should be more than 4 choices under the picture. In the picture there could be more choices. But pixel hunting is not nice, so the area should be big enough.

I've been experimenting with a simple raster and it's pretty workable for creating a story. 3x3 works quite okay, maybe we should use 4x4 or 5x5 but I don't think it's necessary to make it any bigger.

The big remaining question is how to handle the cooperation on the part of the graphics. I guess the idea with a downloadable project file should work. This would also mean contributor should be able to upload. In first instance you could use rapidshare for that. Later you could add an upload option to the site.

I just don't know how big the chances are, you would get graphic contributions. It's a lot more work, than just adding an option and a few lines of text.

[TheBrain]

coder wrote : About the clickable areas: I presume that graphics and story progress together. So there's a picture with some options and the contributor wants to add a choice. Or take Arianeb, let's say you want to invite the girl in the gas station. You enter development mode, make the girl a clickable area, and offer a new branch. A graphic artist would then need to create a picture and the story continues. In this case you could also continue with the same picture.

Well, I suppose that up to a certain point it is true that they will follow eachother. I could imagine that if we did follow through with this idea we'd probably start with writing one linear story and then convincing one or more artists to join and create the graphics for it (creating a baseline in the techniques and models used for the graphics).

But on the other hand I can also imagine that if it becomes 'a hit' we'd end up with way more writers than artists. And then the story would continue to stay ahead of the graphics (provided the artists work in a 'first in, first out' fashion).

Concerning the consistency: In development mode you should see the history. So if you threw a glass wine over the girls dress, stole a camera or bought her a new dress, you should see that info. If a new branch depends on one of those things you should add a dependency to the branch. You can only take pictures if you have a camera. So to enter that branch you would need a camera.

It's funny that you mention a 'history'. I've been playing in my head with rather the same idea, being that a 'page' in the story should not only contain text and actions (and side-effects, if you include variables), but also some form of notes which can contain details about events that are important to know when expanding the branch. In development mode you could simply append these notes into a log, creating the history you mention.

However, there could be some overlap with variables, assuming they are available (I'm beginning to think there should be variables, but very basic (only booleans?), to keep real 'design issues' away from contributors). If we take the spilling of the wine as an example, this could be entered into the notes or could set a variable like 'dress ruined' to true. The first would assume that the issue is going to be resolved before the branch merges back into the main storyline (without lasting side-effects, also not graphical (!)), or that the branch never merges back (i.e. there will probably not be a good ending after that event).
Using a variable instead will pretty much open up any option you want, as long as the main storyline has dependencies on 'dress ruined' the branch can merge back after setting the variable. The great danger is that adding too many variables can cause a 2nd type of combinatorical explosion. While the number of pages can remain lower, each page will get more and more difficult to create when you have to account for a lot of variables. Considering that we are thinking about the premise of having a lot of writers it will probably become inevitable that some page ends up inconsistent with respect to some variable (but then again, anyone can fix it if he comes across it). It would make me think that variables should not be allowed to be created freely, but rather have it in the hands of the maintainers (I suppose).

Your idea to hide the unfinished branches for the player until they lead somewhere is also a good idea. Otherwise a player would always feel he's playing an unfinished game. Well maybe you could add an alpha mode. Maybe it could be integrated in the development mode.

I would just keep it limited to development mode. Although I just thought of another form of pruning and that would be graphically: only show branches that are both leading to 'proper ends' AND also contain images all along the way (although initially you'd probably end up with an empty story this way :P).

[cut]I also think it's a good idea to limit the possible choices. I don't think there should be more than 4 choices under the picture. In the picture there could be more choices.[/cut]
Considering that having a lot of choices will create a story exponential in size, I'd say this limitation is pretty much implicit. In addition you'd probably want to formulate a rule something to the effect that a choice must have some logical effect on the story. Consider a choice where the date asks the protagonist (i.e. the player) what his job is. I can imagine that someone might view the current choices and could think "well, I don't wanna have job X, so I'll add job Y!", but where the difference between X and Y has no effect. (Of course in this situation we could also consider having the protagonist included in the background details, so the job is static and your choices could instead be 'tell the truth' or 'lie')

But pixel hunting is not nice, so the area should be big enough.

I've been experimenting with a simple raster and it's pretty workable for creating a story. 3x3 works quite okay, maybe we should use 4x4 or 5x5 but I don't think it's necessary to make it any bigger.

Pro: A raster would be easy to edit.
Con: Having some freedom in the size might allow some more 'obscure' options, for example Date Ariane's stargazing game.

The big remaining question is how to handle the cooperation on the part of the graphics. I guess the idea with a downloadable project file should work. This would also mean contributor should be able to upload. In first instance you could use rapidshare for that. Later you could add an upload option to the site.

I just don't know how big the chances are, you would get graphic contributions. It's a lot more work, than just adding an option and a few lines of text.

Support for uploading is not that hard, but should have some limits, of course. Beside that, I'm afraid this really is a 'try it and see' issue. Maybe there are some motivated artists out there, maybe there aren't... (but if you are and you are reading this, drop a line! )

[coder]

history

Boolean variables is a good idea and should suffice. That was basically the model I had in mind. A list of boolean variables. And it would be easiest if you only saw the variables that are true. (HasCamera, WineSpilledOverDress, GirlIsDrunk, GirlLikesYou etc)

You could make this model a bit more complex when you have stages for some stuff. GirlThinksYouAreNice -> GirlLikesYou -> GirlThinksYouAreHot -> GirlWantsToHaveSexWithYou

GirlDrankSomeAlcohol -> GirlIsDrunk -> GirlIsPukingInTheToilet (game over)

WineSpilledOverDress -> DressRuined / DressCleaned

And then you could add dependencies to a branch.
If WineSpilledOverDress -> suggest girl to take off the dress.
This way the option would only be visible if you've actually spilled the glass.

Some more complex checking would be nice and would make the game more interesting (like the "girl likes you"-idea I suggested above. So you could suggest it, but she would only do so if she drank enough and think's your hot or better. But I guess you should start with a simple model and if the project works out well, you can add stuff.

Graphics

I did not explain properly what I had in mind for the graphics and the design progress. I don't think it's much fun for a graphic artist if he just needs to create pictures with an already written story. It would become more interesting if there's some interaction. So if someone suggests a choice and a descrïption of where it would lead to. There should first be a picture for the new scene (which could also be the same picture), before you progress writing the story. This way you could add text choices and clicks in the picture at the same time.

Writing a few lines of text is easy and can be done in a few minutes. Creating a picture will take more time. I'd guess at least a few hours. This way you also prevent that there's a lot more text than graphics. Progress would be slower, but I think it's more satisfying than having a lot of unfinished branches where you need to wait for someone to add graphics to, which might never happen.

Limitation of choices

The main reason I said there should be only a few choices below the picture, is that to much text is not nice to read in this kind of game. I have the same thing with comic books. I want to see a little text as possible when reading a comic. If I want to read a lot of text, I'll read a book and not a comic.

Too many choices will also make the game unplayable. What would you choose, if you can take her to a movie, a theater, a strip show, a restaurant, a bar, a disco, go shopping or take a walk in the park?

raster or no raster
It would be possible to create options with an area where the player can click instead of using a predefined raster. The added freedom would add more possibilities, like creating a mini game and it would also be a bit harder for the player. (he can't just check if one of the 16 fields of the raster is clickable)

[TheBrain]

coder wrote : Graphics

I did not explain properly what I had in mind for the graphics and the design progress. I don't think it's much fun for a graphic artist if he just needs to create pictures with an already written story. It would become more interesting if there's some interaction. So if someone suggests a choice and a descrïption of where it would lead to. There should first be a picture for the new scene (which could also be the same picture), before you progress writing the story. This way you could add text choices and clicks in the picture at the same time.

Writing a few lines of text is easy and can be done in a few minutes. Creating a picture will take more time. I'd guess at least a few hours. This way you also prevent that there's a lot more text than graphics. Progress would be slower, but I think it's more satisfying than having a lot of unfinished branches where you need to wait for someone to add graphics to, which might never happen.

Interesting thought. I went at it from the angle of "what if I could do everything myself?", which would be: 1. write high level scenario, 2. write dialogue/story, 3. add graphics. Of course the "it's no fun" argument crossed my mind now and then, basically when I said that people shouldn't feel like slaves that was the thought.
It's a tradeoff though, creating lots of branches with unfinished graphics has as an advantage that the progress is faster, and having a lot of branches probably will mean that the high-quality branches will get graphics first (because they are the most popular? something like that). Disadvantages being 'not fun' for artists and probably an unfinished feel (unless you prune every branch without graphics in the non-development game).
On the other hand, creating graphics first from a global scenario, I can't help but feel that it's approaching the design process backwards. An image can mean a lot of things, while text is rather specific. It might end up like a 3-word story kind of thing? (a popular forum game where a story is made by having ever response containing 3 words that add to the story), as in, you write something but don't exactly know where it is going (and such stories end up to be very... random ) . Here that would be, the graphic designer creates an image while he doesn't know exactly what is going to happen. Of course not exactly the same because he knows a bit about the intentions that come with suggesting the action/response. Still I'm not sure...
The other thing of course is the slow progress, doing it like that really puts the bottleneck at the artists. And especially in an experiment where we don't know how many of each 'type' (writer/artist/general contributer) we might attract I'd say that is a very bad thing.

One other alternative I can see is if you create a sense of teamwork, by organizing contributors into small groups that each work on separate branches. Grouping a few writers and artists together might create a sense of shared achievement, that might add some fun for the artists. That said, they would probably still be creating pictures for situations that have already been written.
Teamwork like this however very much removes the wiki-like freedom I would intend such a project to have.

Another alternative would be to simplify the graphic creation enough to be able to reuse large parts. The only example of this I can think of now is having a 'library' of prerendered poses that you can readily position onto a background image. Something like this would really be useful in situations where the situation isn't interacting with the background, say any situation where you would be standing in front of the girl talking to her (with a variety of library poses, like neutral, smiling, mad, etc.). It is without doubt that an approach like this will reduce graphics quality, but it would greatly increase the number of graphics that can be added because pretty much any contributor can also create simple graphics, leaving the artists free for the 'complicated' scenes.

If I want to read a lot of text, I'll read a book and not a comic.

Yea, pretty much what I'd think. Though sometimes a little narration might add a bit of immersion.

Too many choices will also make the game unplayable. What would you choose, if you can take her to a movie, a theater, a strip show, a restaurant, a bar, a disco, go shopping or take a walk in the park?

The first? And then come back and try the others?
A much better approach would be to split these activities into two or three groups, and limit the player to chosing one or two from each group (a bit like the episodes in Masq). So you first go to either see a movie or go to a restaurant and then go to the disco or the park, like that (still 4 options, but a lot more possibilities).
But that's more a gameplay design issue. Nothing to really worry about now (in this context, at least).

[coder]

To start with it would be better to take your approach and see how it works out. At least we'd be able to start the project. We could always change the approach later.

Ik heb je trouwens een pm gestuurd.