Designing interesting Quests

Quest archetypes are not important. As long as the story is told with the Therefore/But principle, the quest will be good and interesting.

First, some story telling theory. What is a quest? A quest is a series of events. A series of events is also a story. Therefore, make a good story and your quest will be good. To make a good story, use the Therefore/But principle.

The Therefore/But principle is the rule that to tell a good story, each event of your story must be linked together by a Therefore or a But. In a good story, each event is the cause of the next event. If an event is the cause of the next, they will be linked together by a conjunction of causality like Therefore or But. If your events are linked by a conjunction of addition like And, this means there is no causality between the events.

For example, here is a simple quest process within a game: I talk to a homeless guy. He tells me to kill 2 goblins and collect some water. I find the goblins and I kill 2. And then I find the water source, and I collect some. And then I talk to the quest giver again. and I receive a sword. End of quest.

Now, if I take the same quest process, but I tell it with the Therefore/But principle: I talk to a homeless guy. He tells me he is very thirsty. Therefore, I think it should be a good idea to find him some water. Thefore, I go find a water source. But, 2 goblins guard the water source. Therfore, I kill them. Therefore, I can collect some water for the homeless guy. I give the water to the homeless guy. Therefore, he brings out a sword. He tells me he has no use for it, therefore, he gives it to me. End of quest.

Again, quest archetypes are not important. You can have any number of archetypes, if the story is told properly, the quest will be interesting.

I sadly must disagree that theory for some reasons of happening. I played a game yesterday where there is a town that s tourmented by pirates. The citizens decide to trap the pirates and collective form a plan to kill them once at all. You need to talk for the towns tinkerer that makes the decision to assemble some bombs digged into the beach to blow the pirates up but how to get them there? You meet an alchemist that gives ya a recipe for fool's gold but wait, how to get into the pirates camp without getting catched and killed? So you will also have to catch one of the pirates and ask him for there camp's location and watchword, also you need to mask and act like a pirate.

This is just a summary of the quest line, more or less interesting story in theory but in practise all what you are doing over two hours was collect X of Y, collect another X of Z, talk to X, talk to Y, talk to Z, collect X of Z, carry W to another W, collect X of XY and a lot of kill X of whatever. In summary, a really boring gameplay sending the party over a really (really) small map where it just took only 2 minutes to reach the other end.

What about this, how did this fit into your theory? How could have this be made better and worht playing if you say that it dosent depend on the archetype of the quest itself?

Edit Dont ge me wrong, it dosent mean that any of those constellations have to be boring, in games like Dungeons & Dragons Daggerdale you are also just smashing a dozens of goblins with your party and run from the dwarfs to the goblins camp back to the dwarfs but this has made fun sloughting through the hordes of goblins and there wasnt any great told complicated story involved here (cause the game has had a total time for playing of not even 6 hours) but this was a lot (and I mean A LOT) more fun than what we have played yesterday. Even the boss fights (were as trivial so I'm not even sure these were boss fights but just some stronger foes to defeat) did not feel as they really were boss fights

I also agree that Micheals answer that text to a boring quest is still a boring quest.

Autogenerated / trivial quests are typically hard to do anything about. It fills MMORPGs with some content, but people do them for the loot and xp as you said. No simple solution to this...

In guild wars 2 you dont even talk to someone and (skip) the quest text. You just enter an area and "tasks" show up in your HUD and you start killing and collecting. Streamlines but doesnt really solve anything in my opinion.

In that quest you told, I see two problems. First, where are the conjunctions of causality? There's barely any. Also, the player is treated as a second class citizen, when he should be the star of the game. None of what happens is decided by the player, it's all imposed to him.

• There is a town that's tourmented by pirates. ( are you forced to go there? do you want to go there as the player? )
• (therefore), The citizens decide to trap the pirates and collective form a plan to kill them once at all. ( the player should decide, not the citizens )
• and You need to talk for the towns tinkerer that makes the decision to assemble some bombs digged into the beach (and) to blow the pirates up but how to get them there? ( the player doesn't decide again )
• and You meet an alchemist and he gives ya a recipe for fool's gold but wait, how to get into the pirates camp without getting catched and killed?
• So you will also (also is synonym of and) have to catch one of the pirates and ask him for there camp's location and watchword, also you need to mask and act like a pirate.

That quest is just a list of tasks given to the player, it's not a proper story. A proper story doesn't need any words or speech. A proper story is when each event of the quest are linked together by causality.

Witcher 3 is nothing but "collect X of Y, collect another X of Z, talk to X, talk to Y, ...", yet it succeeds at having the best side quests in an RPG because each time, it tells a proper story. The same for the Mass Effect series. There is no magical quest archetype that makes it good, it's all about the story telling (again, no need for words, need for causation).

The problem is there is usually so MANY quests that they get bland and feel like shores. Maybe try to make fewer, larger quests that can seem more unique or worthwile can work?

I felt the witcher 3 side-quests got boring pretty fast as well. But im not normally to focused on story in games anyway.

I agree the player is forced to may tend to a list of ergotherapy tasks to keep the gameplay up. But again, in similar games like Dungeons & Dragons Daggerdale where you are just sloughting goblins and other stuff for the full 6 to 8 hours of gameplay, it is similar but less boring (to me at least). You are thrown into some sort of "campaign" without any choices, hunted by a giant golem into a portal and forced to help some dwarfs. Here again, the game decides what to do not the player but this works and the other game dosent seem to work.

Same as in Skyrim, you are forced to play for the or the other side even if you didnt want to go for both. The game decides to get both fractions onto a table to talk, not the player.

The only real choice is that a player may or may not accept a quest if possible while in most games you didn't even have the choice (unless you stop playing)

I can't believe that "a reasonable story" on its own makes the game, it must be something additionally !?

14 hours ago, suliman said:

Autogenerated / trivial quests are typically hard to do anything about. It fills MMORPGs with some content, but people do them for the loot and xp as you said. No simple solution to this...

This is right, for that reason I asked here to investigate how complex the generator has to be and what kind of rules should be applied to it. I could imagine it depends on the kind of context a quest is running into if it will fit or not.

A quest that is like I described; "go to the place I marked on your map and collect 12 X then get it back to me" where the area is just half a minute away so the NPC himself could have been going there by its own instead of talking to the player is useless while giving the player a 'real' task might make it here. It will still be a collect X of Y task but as a side quest the player can decide when and how it will be done or the real task should be not collecting Y but discover where/how to get Y.

I also think the fun is not seeking for the guy with the sign above its head rather than The Edler Scrolls is doing it: talk to people, listen for there stories and you might get some rumors to investigate, reading books and reveal new places on the map. Here it is all about talking, no one gives ya a quest just cause.

Correct me if I'm going wrong in this thoughts

BTW: The 'real task' with the auto generator will be to get NPCs dialogs different and a kind of responsive that a player might think it acts like a real character instead of the two or thrre standard clauses , but this is a more technical issue than a game design related one

On 4/10/2018 at 3:05 AM, Shaarigan said:

BTW: The 'real task' with the auto generator will be to get NPCs dialogs different and a kind of responsive that a player might think it acts like a real character instead of the two or thrre standard clauses , but this is a more technical issue than a game design related one

Generating the dialogue leads to interesting design challenges. For example, one of the questions I had to answer when developing Quest Machine was where to balance the human author's control versus putting faith in the generator. On one end, you could let the author write very detailed, motivation-imbued templates that the generator fills in Mad Lib-style. They're almost indistinguishable from fully hand-written quests, but they're not flexible, and the format quickly becomes obvious to the player unless you write hundreds of them. On the other end, the generator could very flexibly produce a shopping list of steps (goto X, get Y, give to Z) without any flavor text or sense of motivation.

Ultimately, annotating the game world (what's there and what can be done with it) and helping NPCs establish good world models was the key to making interesting quests. Quest Machine runs a STRIPS-style planner (a little slower than the HTN planner I started with, but more flexible) on smart objects to generate a plan, and then uses NPC-specific text lookup tables to compose the dialogue in a manner that fits the NPC's personality.

This shifts the design focus to how interestingly you can annotate the game world with things to do. So you can see that just one aspect of something like procedural quest generation can have broad design implications. Note that this is for a system that generates quests based on the current state of the game world. There are some academic quest generators that take the opposite approach; they generate a quest and then change the game world to fit that quest (e.g., spawn a new cave of monsters).

On 8-4-2018 at 9:31 AM, Shaarigan said:

What kind of quests or quest components could be generated too that are not listed below to make the game more interesting?

• Collect X of Y
• Carry X to Y (you maybe need to find X first)
• Kill X of Y
• Talk to X
• Kill X

You have listed "collect" "kill" and i suppose "talk", which seems to be the kind of mechanics you 're aiming for.

To make the quests more interesting you wil need to change the smaller details of the quests themselves; the kill-quests can be repeated many many times for different(not necessarily stronger) enemies, an enemy could have have ranged attacks or only melee attacks, he may have stun/poison effects, maybe he even runs away and heals himself, or maybe the player has to assasinate the mayor of a town which is quite easy, but then all the townspeople try to kill you and you still gotta get to the safe-point.

Carry-missions are more an excuse to send the player one way or another, unless there are specific skills involved that the player needs to manage(for example carrying-capacity, maybe he needs either forging or trading to acquire an object or else he needs to steal it etc.)

Talk-missions are mostly in games to explain mechanics(tutorial) and present story(aka present the other quests) if you WANT to make them more interesting multiple choices are key, though i would only do that for side-quests(because they 're optional and you made the player choose). (They can still be made more interesting with skills as well)

On 4/14/2018 at 4:36 AM, Tony Li said:

(...) There are some academic quest generators that take the opposite approach; they generate a quest and then change the game world to fit that quest (e.g., spawn a new cave of monsters).

That's interesting. Why you call it "academic"? I'm seen different games where I'm convinced they do it this way. I'm also venturing down the same road where the story comes first and the locations required for it to play out are put in place as needed.

48 minutes ago, RPTD said:

That's interesting. Why you call it "academic"? I'm seen different games where I'm convinced they do it this way. I'm also venturing down the same road where the story comes first and the locations required for it to play out are put in place as needed.

True, some real games, such as Bethesda's, create a story first and then modify the world to support it. There's absolutely nothing wrong with that. Bethesda's goal is to drive players to unexplored areas. Placing a quest there that has context to their character is a great solution. But the academic projects I examined, like Jonathon Doran and Ian Parberry's, and Young-Soel Lee's, model it in a little more depth. Then again they didn't need to fit it into a AAA game running at 60+ fps.

I like the idea of consequences in an open world. So in my implementation quests rise organically out of the current world state. (Starbound is another good example of this approach.) I think this shifts the focus more toward player-generated stories versus imposing a story on the player. If the player introduces rabbits to a new land, he might come back later and receive a quest to exterminate some rabbits that have overpopulated the land and are crowding out native fauna. Compare this to a generator that first creates a quest to exterminate rabbits, and then spawns a bunch of rabbits. In this case, there isn't as much player-driven cause-and-effect. But I will admit that the drawback is that it takes more work to annotate world objects with interesting things that can be done to them, and to annotate NPCs with a motivation system that lets them recognize what objects they want the player to manipulate and why. Without sufficient annotation, quests can feel drier than your story-first approach.

2 hours ago, Tony Li said:

True, some real games, such as Bethesda's, create a story first and then modify the world to support it. There's absolutely nothing wrong with that. Bethesda's goal is to drive players to unexplored areas. Placing a quest there that has context to their character is a great solution. But the academic projects I examined, like Jonathon Doran and Ian Parberry's, and Young-Soel Lee's, model it in a little more depth. Then again they didn't need to fit it into a AAA game running at 60+ fps.

I like the idea of consequences in an open world. So in my implementation quests rise organically out of the current world state. (Starbound is another good example of this approach.) I think this shifts the focus more toward player-generated stories versus imposing a story on the player. If the player introduces rabbits to a new land, he might come back later and receive a quest to exterminate some rabbits that have overpopulated the land and are crowding out native fauna. Compare this to a generator that first creates a quest to exterminate rabbits, and then spawns a bunch of rabbits. In this case, there isn't as much player-driven cause-and-effect. But I will admit that the drawback is that it takes more work to annotate world objects with interesting things that can be done to them, and to annotate NPCs with a motivation system that lets them recognize what objects they want the player to manipulate and why. Without sufficient annotation, quests can feel drier than your story-first approach.

I always felt Bethesda quests to be very boring and repetitive though especially since they pulled a random dungeon from a list (no matter if visited or not) and plunged the "go kill this to get me that item because I'm too lazy to do it myself" in it. In the end this all felt out of place in the sense of "why is this thing 'there' in the first place? makes no sense".

I think what you mentioned with the actors reacting to changes in the environment instead of generate unrelated quests is the better approach. In the end this is the approach I'm actually venturing to. And you are right, it needs a more elaborate AI system to do. But where would be the challenge in boring simple systems, right?

Agree, I ever felt that it was not a good idea to explore every dungeon in the game of Skyrim/Oblivion just to get an NPC that a few hours later want me to go in and fetch something for it. All enemies dead but the one "Ring of Ultimate Whatever" has spawned there so again walk through feeled 100 floors just to find the ring in a chest I looted just an hour or too ago? Yeah, thanks but no!

What I'm twiddling arround at the moment is an approach where there are auto generated quests as same as human written story snippets that are just that generic so that the content generator could assemble it too. Anything, any NPC in the world should have a background story why he/she is there at the moment a player discovers it. Partially auto generated but also as some kind of principle of cause and effect. Based on that background-story (without going too deep into technical details) quest will be spreaded arround the world to make them "more plausible" to play.

I personally like the Bathesda games for there "talk randomly to someone that tells you something you might find interesting enougth to get a quest for" principle. Regardless of if it is reading a book and gather some knowledge about some palce anywhere in the world or find people that need help with something because one heared from another one that there is someone that knows someone to need help

I personally also dislike the typical "you have no choice because you are just the hero and go get the situation fixed that a mass of people has broken for the past 10 years" scenario. Especially in (bad designed) multiplayer/online games where each player is something special and the ultimate hero the citizens have waited for a century or two. What is the motivation for playing such kind of game? Sometimes it the story but mostly I feel it is just to explore the world.

I can't immagine that WoW players play the game for the background story rather than exploring and raiding dungeons to beat down there bosses.

So what I'm finally destinated to with this topic (and technology) is not making the story in a complex main quest but anything arround it so one might or might not tell his/her story and the content generator fills the world with anything else or quit on a main story but still provide a living world with all these interesting things to explore.

I also like the idea that quests should react on the world and not vice versa, what would (I think) make them more plausible than the other way round

I personally never liked this concept of "you are the hero we waited for ages". It results in NPCs loitering around (or just standing around) waiting for the player to blow them in the face to tell him he's the one and only they waited for. It's the main reason such game worlds feel artificial and not credible at all (Bethesda suffers a lot from this).

Unfortunately it's not so easy to get away from this concept without getting a DM involved.

One of the main problems I see is (as you mentioned in the beginning) the types of quests. Kill/Fetch quests all around the place are hardly helping in this situation. Coming up with quests which require more creative way of using your abilities would help. For example think of what could you do with your weapon else than just smacking goblins. Can you craft a situation where your ability helps solve a problem that does not first and foremost require killing something? This produces unexpected moments breaking the monotony.

I like Bathesda games but only those that offer a huge world where you could just explore and meet people so you have a total ingame play time of 10 hours before going to do the first main quest

But in the other points you are completely right and thats why I wrote somewhere earlier that it is more fun if an NPC just wants X of Y but the game dosent tell the player where/how Y is to obtain instead of pointing to a 2x2 meter area on the map spawning every 5 seconds an evil NPC that drops Y. It will still be a collect X of Y quest but with a little gathering background for information that could themselfs be obtained from other NPCs, books or experience of other players.

I also would always work with consequences. Killing an NPC prevents you and/or other players from finishing there quests. NPCs never respawn and as long as this isn't a main story required one, it wont have any impact on the game itself but will punish players if there is later a quest that has an option to be solved more easy by just talking to that NPC.

I think a game can sometimes be but shouldn't only be a place to just slaught anything that moves because 'we can do it'. I don't say that I never took my sword and simply slashed a guard or shot an NPC down but this shouldn't be a general approach to go (except for a relaxing turn Unreal Tournament )

So by the way, if anyone is interested in this discussion, I could still need some help writing and designing the core game mechanics and content creation rules! Just PM, Skype or Discord me

A few ideas could be:

• Travel to location X
• Interact with X
• Do not get noticed by X, but defeat Y and travel to Z

Quests are in an interesting part of Game Design. Whilst there is a variety of different styles of game play, they typically fall into several categories. Arguably, the best quest design incorporate the mechanics of the game and have several layers of depth. For an example, let's look at Pokemon. The main quest is to catch defeat the Pokemon league, which is the primary goal. The secondary goal which is completed on the way is to collect every Pokemon and train them. To do this the player must take advantage of their knowledge of the game's mechanics and exploit them in such a way to progress their journey. This quest subsequently journeys the player through a story.

There is no real answer to the best quest design, and to generate them whilst keeping them interesting would be difficult. I can however advise that to keep them refreshing is to incorporate the game's mechanics and keep side quests randomly generated, but the main story predetermined.

On 04/09/2018 at 2:41 AM, Michael Aganier said:

Quest archetypes are not important. As long as the story is told with the Therefore/But principle, the quest will be good and interesting.

First, some story telling theory. What is a quest? A quest is a series of events. A series of events is also a story. Therefore, make a good story and your quest will be good. To make a good story, use the Therefore/But principle.

The Therefore/But principle is the rule that to tell a good story, each event of your story must be linked together by a Therefore or a But. In a good story, each event is the cause of the next event. If an event is the cause of the next, they will be linked together by a conjunction of causality like Therefore or But. If your events are linked by a conjunction of addition like And, this means there is no causality between the events.

For example, here is a simple quest process within a game: I talk to a homeless guy. He tells me to kill 2 goblins and collect some water. I find the goblins and I kill 2. And then I find the water source, and I collect some. And then I talk to the quest giver again. and I receive a sword. End of quest.

Now, if I take the same quest process, but I tell it with the Therefore/But principle: I talk to a homeless guy. He tells me he is very thirsty. Therefore, I think it should be a good idea to find him some water. Thefore, I go find a water source. But, 2 goblins guard the water source. Therfore, I kill them. Therefore, I can collect some water for the homeless guy. I give the water to the homeless guy. Therefore, he brings out a sword. He tells me he has no use for it, therefore, he gives it to me. End of quest.

Again, quest archetypes are not important. You can have any number of archetypes, if the story is told properly, the quest will be interesting.

Absolutely, WoW is a great example of this, massive story lines all integrated out of a bigger picture throught the expansions till now, and 90% of quest has its own story behind a story to tell, always keeping that intruiging effect on players.