View Full Version : Character Generation via Choose Your Own Adventure (idea)

2015-11-02, 01:51 AM
Has anyone tried to generate character classes based on a short choose your own adventure story which would essentially also act as the characters backstory?

The idea was to make a quick and interesting way to generate "not so special" characters for high mortality rate games that give characters quick backstories as well as defining their "class" of sorts.

So if at some point in the Choose Your Own Adventure (CYOA) story the character grabs a grenade rather than say a sniper rifle to eliminate an enemy machine gun encampment (just as an example) their characters stats once they begin play would be better suited as a Grenadier than a Marksman. Of course not all options must be exclusively relevant to Character Attributes and may instead determine how their comrades or enemies see them. A scenario where they have the choice to execute, imprison, or set free a captured enemy squad may cause their allies / enemies to see them as bloodthirsty, weak, or compassionate giving the character a little flavor and making the process a little more fun rather than just selecting the best number-generating options for min-maxing.

I know Traveller has famously had something along these lines, and even Morrowind had a questionnaire that assigned you a class based on your answers, but the first one was less "Choose" and more "roll dice" and the second one is more hypothetical without reflecting an actual backstory for the character, so I don't think I've ever seen this done before. Like all things where "I don't think I've seen it done before" my first question is "Why not?" because I doubt I was the first person to think of it, so maybe there's something I'm missing that causes people not to take this route? Is it not entertaining enough or too complicated to implement in an entertaining fashion?


2015-11-02, 03:17 AM
IIRC, the Morrowind questions are literally lifted straight from the earlier games in the series, in which they were presented as more of a backstory for your character.

Some possible reasons why they haven't been implemented in table top games is probably because of the work involved to create such a list.

here's a quick challenge. take all of the classes in core DnD, (or any class based system). Now come up with say ... 20 questions. Each relating to the character's backstory. Now, each of these questions has 3-5 answers. Now, since these all relate to the character's backstory, they have to all be able to relate to one another in a way that makes sense. Take 10 out of these 20 questions and present them to the player. They then give you their answers and now you have to determine which kind of class they are. I think the same would go for skill based systems as well, or point buy systems.

The above might be fine for a one player game, but for multiple people, the backstories will probably be similar in someway. "Oh my gosh! Your village was attacked by goblins too?" "Why yes, I used my pitchfork to fend them off." "I see, well I used my kitchen knife!" "Huh, no wonder you like daggers..."

And in the end the players catch on to the pattern and just end up picking the "right" questions to generate what they want. Thus it ceases to be interesting after the first or third time, and just becomes a chore.

Take your "eliminate machine gun" point. If the players were given a choice between, grenade, sniper or shotgun. Players are just going to pick what they want to have as a skill in the first place. Why make them jump thru hoops to get what they want when they can just pick and choose the skills they want their character to have. Or maybe they mess up once, and they get something similar, but not quite what they wanted. And then they're just annoyed.

Even in the Elder Scrolls titles, where the choices weren't so obvious. There was still a pattern. Choosing honorable/direct/bold answers scored you warrior points. Neutral scored you mage and careless/dis-honorable answers scored you thief points.

Sure, you might get a player who likes to just "see what they get", but then, there are systems that use dice to randomly generate characters, which provides a lot more combinations a questionnaire.

If you're talking about a highly lethal game .... then why bother generating backstories? Backstories are not necessary to have fun in a table top game. Kicking down the door of a dungeon, moving from room to room, is just as valid a play-style as a drama-filled epic. If you wanna have them? great! If not? then don't. If they're not suited to the game/style then don't bother. Besides, often enough, players can come up with a rudimentary backstory when necessary in the first place, without needing to go through 10 or so story options.

2015-11-02, 04:17 AM
I like the concept, at least in principle. It's got some issues, but I think if it's implemented correctly it could be fun.

As far as what that implementation should look like:

This would probably work best if all of the PCs are assumed to have some commonality in background. If they're all survivors of the same goblin raid, it wouldn't be nearly as odd for them to have gone through a lot of the same events. Focusing it like this also lets you get more specific within that framework, which means you can work in more of the fun, interesting little sections that would make this idea pop.
I would probably not make it decide something as important as class. That would inevitably lead to people being stuck playing classes they don't want to play, which can ruin their fun. Rather, it would probably work best to have it generate a list of the classes which are most suitable to the playstyle they've described throughout the process (e.g., "You're 67% compatible with rogue, 54% compatible with ranger, etc.") The actual mechanical effect of the result they get should be small (so that it's not a huge problem if the CYOA gives a result far from what they want to play), interesting (so that they care), and unique (so that it can be interesting).
Use the CYOA to present information that the PCs would have before the game, rather than doing handouts or whatever. To use the example of a goblin raid from earlier, maybe one route through the background generator lets the character see the shaman cast fireball, so now the player is better equipped to fight the shaman. Obviously this would be most effective on the initial batch of characters, but if you put in enough of these tidbits they could keep finding new ones even after they've gone through several characters each. Be careful with this, though; if you put in too much branching, it can get to be a lot more work to design the CYOA.

If you actually want to make something like this, I highly recommend the ChoiceScript (https://www.choiceofgames.com/make-your-own-games/choicescript-intro/) programming language. It's free, fairly easy and intuitive to write, and the games it generates are much easier to use than an actual printed book.

2015-11-02, 10:37 AM
Has anyone tried to generate character classes based on a short choose your own adventure story which would essentially also act as the characters backstory?

2015-11-02, 03:58 PM

The amount of work to put into it to make it varied enough that people don't end up going down too similar a path is probably the main thing that worries me. I think it can be done but I have my doubts in my ability to make each path interesting enough.

I fully anticipate people finding a "path" that leads to answers they like, or perhaps taking a very similar path but changing one answer just to "See what happens" so I've considered writing a simple one at the start of each campaign (or each chapter maybe) so they're slightly different and ideally players would only run through the options two or three times before a new one appears. But like I mentioned, that could be a lot of work even just making a number of smaller ones rather than one large one, and it could all be for nothing if they decide not to explore the options anyway.

I also should have pointed out that my idea of "high lethal" isn't quite as bad as say The All Guardsmen Party (http://www.giantitp.com/forums/showthread.php?386908-The-All-Guardsmen-Party) and is more like a character or two dead per session. So a relatively simple backstory wouldn't be out of place. But that's my error for not being clear on that.


Thank you, I didn't know about Choicescript but that could be extremely helpful. I was doing the entire thing in a Word doc and was already getting frustrated with the difficulty of tracking choices.

You are probably right on all points. It'd be more convenient as the writer to let them assume the CYOA takes place in a battle that each of their characters were present, making it not far-fetched to have similar experiences. It's probably a better method of class suggestion than it is class assignment so people can just treat it that way and play what they want to. It's best purpose would probably be setting the story up for the players and giving their characters a role in that setting by the time they take official control of them.

So far it seems like the best way to go about this is to use ChoiceScript to not overwork myself. Write out short CYOA's for each "Chapter" so that they do not become stale and predictable and so the setting is understood in a unique way by each player. And use the answers to suggest classes rather than assign them. I'm also considering granting "titles" in the CYOA to people based on their decisions, like if the option comes up to release enemy captives or slaughter them, releasing them grants the title "The Merciful" and slaying them grants "The Black / Monster" or something along those lines. This would be essentially the "end" of the CYOA and just a way to reflect upon those choices. I could also see this as some incentive to play them "just to see what you get" rather than following a clear path, but I could be wrong.

I also may have been misleading with my "Machine Gun Nest" example since I would like to write this as a Fantasy setting. That was just the first example that came to mind. Not an important detail, but if I finish this and want to show it off I don't want anyone to be mislead. :smallsmile:


Oh, how about that? I didn't play 4e so I never even knew this was a thing.

2015-11-02, 05:38 PM

Yeah, I'm a huge fan of ChoiceScript. I'm actually about 90k words into writing a game with that language, so I can probably help if you have questions. As far as the actual implementation goes, these would be my thoughts based on your latest post:

If you're running the game in distinct chapters, you could do worse than to write a new CYOA for each chapter. Maybe have it take the place of spoken narration detailing how they get from one to the next, so that they have a bit more engagement with those sections without it taking as long as fully playing it out. I would definitely not try to apply one CYOA to multiple campaigns, because unless you're running extremely generic campaigns, the background won't be specific enough to bring them in.
Class suggestions are better than class requirements. As a GM, I think one of the worst things you can do is tell a player "you must play this." Even if it's what they would have played anyway, telling them they have to can really ruin the experience, especially for people who don't like to be told what to do. If you just list suggestions, you might point out an option that hadn't occurred to them, and if they disagree with the CYOA, they can do something else.
The titles idea is a good one, but I would associate it with some mechanical benefit. Something small but interesting. I don't know what system you're using, but the traits system from Pathfinder is about the relative degree of power that I would suggest. Make it something interesting, something that provides a perk that they can't get otherwise, but not powerful enough that getting a bad one is a serious problem. Maybe give everyone two traits, one they pick and one selected from a special list by the CYOA? That way if there's one they really need they can pick it up themselves, and then the other one is just a bonus.

And if you do want to use ChoiceScript, the way I would do this is as follows:

Create a stat for each class, and another set of stats for personality. Don't bother with boolean variables to track individual choices they made; you don't need to have a record of the exact path they took through the game.
Keep the branching limited, and write about a paragraph for each choice. You should have a bit of descriptive text about the world, some information that's useful or interesting on a metagame level, and then the next choice. Keep it to one scene; longer than that is more work than you probably want to invest in this.
At the end of the game, find the classes that have the highest stats and report them as a statchart. I don't know how many classes are in the system you're using, but I would probably not report more than the top 10% or so; more than that just turns into clutter. The easiest way I've found to do this is actually with the set by reference commands from the Advanced ChoiceScript section, but you can also use a more elaborate setup that doesn't require them.
Also list the single highest personality stat title and the associated trait at the end. I would not list other titles/traits that they had relatively high scores in; that isn't necessary, I don't think.

The Mentalist
2015-11-03, 09:16 AM
I freaking LOVE this idea. I think that it works a bit better if you're going with a common starting place. The only survivors of an orc raid or something and I think it works better still if as a DM you're willing to draft one up at the start of each, or build a batch and randomize for each short campaign. I don't think this works for a one shot and I don't think this works for any game over 15-20 sessions but a 4-7 session game is where this idea could really shine.

Additions I recommend:
1. Adding a few pieces of special loot to the story where you can only chain to pick up one or two. Smallish things that might be outside the price of a level one. A bag of 2d4 thunderstones or a masterwork weapon, that sort of thing. That way even if they get a little patterned you're still getting something out of it.
2. Partial point buy + bonuses from the choices, so you can still customize, even if you do make a wrong turn.
3. Pretty much everything steinulfr said in their post.

In fact I think the best way to do this is to make it a benefit to go through it. Make even your munchkin powergamer smile a little when they see the CYOA booklet come out of the bag.

2015-11-03, 09:18 AM

So I stayed up all night doing... something (http://www.filedropper.com/mygame).

<that link is to the "mygame" folder of the "finished" choicescript attempt. It's hosted on filedropper.com so if you don't trust them or if it doesn't work or you want the file a different way, let me know.>

I followed your advice and most of my time was spent learning the commands and interactions. (selectable_if can be a real jerk.) and it kind of shows that it was a learning experience I think, since it gets into some slightly more advanced coding later on that I could have done much earlier to make the whole thing a little more smooth. But for going from "I've never heard of this" to what I've managed to put together there, it's not bad I think. It needs polished, the coding could be cleaner and the game itself hints at concepts I actually haven't put much thought into yet, but I think it's a good baby step.

I didn't have time to put stats into effect for the Titles, but some titles are generated, you just can't see them until the end.

This is definitely an idea I can run away with. So thank you for showing me this.



I had sat all night with this thread just gathering dust in my google chrome tabs and it didn't update until I went to reply to it some hours later, so I missed your post until I had already responded, hence this edit.

BUT. You're right about adding loot within the options, that could be an interesting idea. I also immediately started thinking of ways to implement "point buy" into choicescript and it wouldn't be hard to do at all. This is fun.

Mark Hall
2015-11-03, 11:14 AM
I know Lifepath generation was common back in the late 90s. The one I really recall it with was Fading Suns. You chose your background (alien, church, one of the noble houses, one of the guilds), then your training, then a couple of generic choices representing your post-training choices.

Not quite a choose your own adventure, but definitely constructing your character step-by-step.

The Mentalist
2015-11-03, 11:57 AM
Also, glancing through 5th, they have some in the book that is a little like what Mark was saying, no mechanics but some good roll for history and personality based on archetype.

2015-11-03, 03:49 PM
My currant active project is basically this. I approached it more from a perspective of binding your backstory to your stat selection. Sure you can game the system to get the stats you want, in which case some structure has been given to your back story. Sure I would rather you go backstory->stats but I don't intend to enforce it.

Also, I'm printing mine out so it is really more of a questionnaire you answer from your character's prospective than a choose your own adventure game.

2015-11-03, 06:29 PM
What I take from this is that it has been done before successfully, but not often with a "branching paths" angle that feels more like a CYOA rather than a questionnaire or multiple choice building characters.

And in my very limited experience so far it is not that hard to put in a game, but it is time consuming and easy to do in a way that is either exploitable or not enjoyable. Which some people may consider not worth the risk, or may just not be interested in doing.

Which is probably why it's been done in similar fashions, just so it can be interactive but easy to keep track and predict outcomes, but without being so time consuming or creating an exploit in character creation.

So lesson learned I think.