"what is a narrative driven game? just a miserable pile of stories. but enough, have at you!" -dracula
why people play games. people play games because they are fun. we tend to play the games that are most likely to appeal to us. I like puzzle games, brent likes party games, we all have some sort of preference.
I think that the same is true about stories, and narrative driven games. there is a certain way that people prefer to play these games based on what they like. some people may want to play with a more action focused style, or others may play by communicating with non player characters and trying to leverage that. if someone is forced to play in a way that they do not like, they may grow bored of the game, or worse might just get stuck.
what can be done about this? well in traditional table top role playing games this really isn't an issue. there is a human there that acts as the Game Master that will work as the rest of the universe. They are able to change things on the fly so that they are are engaging the players. If the players get stuck on a puzzle the GM might make it easier. if they keep getting stuck on puzzles the GM might just use less puzzles in the future. Since the players are not aware of the fact that the world is tailor made to them they still perceive it to be realistic and enjoyable. If they knew that the world is subtly changing itself to get easier or to skip challenges they might feel that they are getting a sub-par experience. (this is also a reason as to why GM's should not tell their players how things could have happened if they did stuff differently) This isn't really the case in table top games because the story is very fluid and on the fly. but for video games their stories tend to be a lot more rigid.
In most narrative games there is more than one path to take. you can defeat the dragon through waltzing up to it and fighting, or you can try to talk to it, and use the time to steal a cup and find a weak spot. but these things are usually rigid choices. when the player is told that "you can go fight the dragon head on" or "you can sneak around". they may choose to do the thing that is the safest even if it is not the thing that they would find the most enjoyable. you can obscure these effects behind many layers of other choices (see, Witcher 3) but this is still just masking the problem, and a player playing the game again would be able to do the exact same things.
So how do we solve these problems of getting a player to do the things that they will find interesting, but also feeling like they are doing the actions? Drama management! Drama management takes the form of two parts: building a model of the player, and using that model to figure out what to give the player. Modeling the player can be done indirectly by what actions the player takes. If the player talks to a lot of people in the game their corresponding model may reflect that. Using that model we can then figure out what to give the player in terms of content. A player that tends to pick aggressive choices in conversation may be given the chance to be more aggressive in the future.
The problem is if the player acts in a certain manner that isn't a reflection of themselves. this could be an issue in the way that the model is created, either putting too little or too much emphasis on certain metrics, or it could be the player just acting in a way that may not apply to future situations, like getting angry at a bully does not mean the player is an angry person. This leads back to the problem that the player may feel like they are forced into playing a certain way, that they might not be comfortable with. Or the player simply wants to change their play style. Players are fickle creatures by nature and how they tackle problems may change all of a sudden. With a faulty or outdated player model the DM may give the player choices that only strengthen its wrong idea of what the player wants, not allowing it to change it player model enough to give the player a chance to do the things it wants. (this can be related to youtube suggested videos, if you watch one video on politics by mistake you may be flooded with them from then on)
The thing that we want to preserve is the feeling of agency. We want the player to do things that the player want to do regardless of the DM's interference. In poor drama management systems this can be a problem if the DM gives the player a hallway of locked doors. Each door may lead to interesting things for the player to explore, but the drama manger only unlocks one. So every place that the player may want to explore is locked off. This can be terrible for a players sense of agency, and may take them right out of the game.
This is where our system comes in. The idea is that the player is given a way out if they are not doing things that they want to do. With these options we lessen the issues with updating the player model, as they represent choices that distinctly only the player can make.
This is done with a couple ways. First the stories (or quests) need to be designed to take advantage of this system. They need plenty of branching paths and different ways to complete them as to try to give players as many options as possible. These paths are ways to get ot the main objective, but we also need side objectives. Things that can distract the player and be used to tell if the player is not engaged. when the player is not making progress towards the main objective we can use that opportunity to try to get them to switch on to paths that they may take more interest in.
p1 q1
p2 -> q2
p3<- q3
p4 \-q4
p5 q5
Players can be on path p initially, but if they fiddle fart around too much and not make progress, the DM might try to give them the opportunity to switch to path q via an event. For example if the player is looking for a cat and the path they are on requires them to talk to a lot of people, but they just aren't all about that life, an event will activate that will show them a trap that can be used to catch that cat, and allow them to take a piece of the trap. Now the path they are on is to find bait for the trap and set it up in the area the cat was last seen to try to catch it.
The data that the player is not willing to talk to a lot of characters may be used model the player in future quests to give them quests more suited towards that play style.
We can also think of the player as a set of properties that we have a certain amount of confidence in. When our confidence in a certain area decreases we can then try to offer options for the player to take in those areas. The hope of this is to explore areas that we are not sure about and to exploit what we are confident in, until we are no longer confident in it. This should allow the system to reassure itself that it is on the right path, and to change when it is not.
What we have so far:
currently we have the text adventure for players to actually explore in, and a single partially completed story for them to play through. There is a system to determine when they are bored, but its simplistic, and a method to generate the player model, but again simplistic.
what to do going forward: more stories, replace simple systems with complex. that should be easy to do because just about everything in the system is designed to be easily replaceable with the rest of the features still working.