Norman, didn't you specifically design a storyless sim survival game?? Why are you concerned about interactive storytelling now? I'm of course very interested in interactive storytelling, but maybe you should save it for your next game?

Now having watched the Extra Credits, I have to say that quests, and particularly "good quest design" have a really different meaning to me than that. I strongly don't like the kind of quest designs described in both Everquest and The Secret World. On the other hand I definitely do admire the player-directing functions of questing in WoW. If I was going to make my own Extra Credits style episode, I'd start by saying that quests organize the player's work within the game. Work can be mainly split into three categories:

1. Gathering materials from the world whether by drop-hunting, growing crops, or gathering

---1a. the degenerate case of this is gathering XP as combat 'training', like a student ninja or knight or something

2. Developing your social reputation in positive personal ways by doing favors for NPCs, in intimidating ways by killing boss monsters and revealing who was behind crimes, and in professional ways by earning reputation with factions and pleasing class-related mentors in order to gain new class-related skills and ranks.

3. Achievements of dexterity and/or intellect such as solving puzzles, crafting complex objects, amassing collections, high scores on minigames, stunt jumps, PvP rank, making money in the auction house, breeding a rare pet... what's possible here varies widely by game.

Quests should, at the very minimum, serve as tutorials to introduce the player to the fact that it's possible to do all these different kinds of work within the game. Quests are much like the memory of the Avatar (or other playable character), which should be distinctly separate from the meaning of the player. The quest log represents the character's memory that they have agreed to do X, Y, and Z, where they need to do it, why they need to do it, etc. Most games don't allow players to create quests for themselves, but if they did it would be very much like writing a To Do list for yourself; like a grocery list of the mats needed for a craft and what monster to kill where to get each one. A few games allow players to set shopping quests for themselves, which is basically declaring item X to be your next major purchase, and showing the market price of item X compared to the amount of cash you currently have.

Beyond tutorials, quest should serve as an external source of suggested goals to motivate the player to play various particular parts of the game, in an order that provides a natural and satisfying increase from simple to complex, and gently, regularly reminds the player of the variety of things available to do within the game while also cheering on the player in areas the player prefers to focus a lot of effort and time on. Quests are also very important as positive feedback and recognition for the player's accomplishments; the quest structure is what allows the game, as a relatively dumb computer program, to recognize such abstract things as when it's appropriate to praise the player.

Now for interactive story. Quests have two main opportunities to allow interactivity. The first is before the player accepts the quest - the player can be given a choice between variant or alternate quests, and I'd love to see a game that allowed the player to express offense at a quest being offered and actively reject the quest, perhaps resulting in combat with the offering NPC. Choices between quests are generally considered an unnecessary development expense (like interactive story in general) but this type of choice can allow the player to express a lot of things in a way the game can recognize and respond to. Things expressed can range from personal philosophy/morality, to personal gameplay style preference, to which NPC or faction the player most wants to build reputation with, to which specific goal within a game-supplied list of options the player would like to pursue. The second point at which quests can be interactive is the details of how the quest is accomplished. This is more problematic for the game to be able to recognize, unless the quest explicitly specifies 2 or 3 methods the player must choose one of, which isn't great for immersion and verisimilitude. With playtesting, this type of interactivity can be made to work most of the time, but it's probably always going to be somewhat buggy.

>> Norman, didn't you specifically design a storyless sim survival game??

BG! it started as one day, out of the blue it just came to me that: "hey! wouldn't it be cool if there was a game where you were a caveman and could craft a stone knife and take on a saber tooth tiger?"

don't even ask me where it came from i have absolutely no idea! it just sort of popped into my head fully formed. you 'd have a cave, which would mean you could take over caves. you could go outside your cave and wash your face in the pool of recent rainwater, and see your reflection in the water, and play with the water and reflections and ripple effects. all with killer lighting and high quality plant models. IE a highly immersive environment (give me another decade and i may get there! maybe... <g>). then you would craft your knife all highly immersive, with particle systems for the flakes flying, reflecting light and casting correct real time shadows as they go and such (lighting is everything). about the time you finished, your buddy shows up (multiplayer co-op), and you head out to where the tiger is. that was (and still is, i guess) the original vision for the game.

but up to that point, i'd only written arcade games, flight sims, wargames, and text mode and 2D RPGs. i had no idea how to write a.... what would you call this thing? an open world survival sandbox stone age fpsrpg / person sim? its grown to the point were "paleo-world simulation" is probably the most accurate description. in fact i'm considering calling it: "Caveman v3.0: a paleo-world simulation".

so now i've just about finished building the world, and all the mechanics, except for final graphics and an audio remix. all the actions and objects are in except for the few that require lots of new graphics. with the world largely done and playable as a finished or at least a late beta product with place holder graphics, it time to give the player some optional quests.

i have no problem with story. i think i would be the coolest thing if i could come up with some plot point reveal system / engine that would allow multiple optional main quest sized stories (and smaller stories too) to be discovered, and followed, and played though by the player. with new DLC stories from time to time. i'm thinking it might be doable. that holy grail that lets you play stories. i mean we all know what we want, we want a holodeck program. we just haven't figured out exactly how to code one easily yet... but i'm not planning on a story engine at the moment (at least not yet).

right now the generator for a quest consists of an init routine, a run routine, and routines to return map locations. last night i figured out how to add a dialog handler and an AI handler for quest specific dialog options and AI behavior. so you can have quest specific dialog options such as "i heard someone here is offing a reward?", or quest specific NPC behaviors such as an NPC walking up to the player and initiating dialog, or leading the player to the "someone here offering a reward". so i now have all the parts required for branching storyline quests with multiple endings, custom dialogs, and custom AI behavior.

quests can be used for teaching how to do things in a game, and can also be used to teach what there is to do in a game, as can achievements, or exploration (go everywhere and click everything).

but they can be more than that. tutorials can teach you how to play. game guides or a decent game encyclopedia can tell you what there is to do in the game. quests can be epic adventures - or not - depending on the purpose and writing.

>> Quests are also very important as positive feedback and recognition for the player's accomplishments; the quest structure is what allows the game, as a relatively dumb computer program, to recognize such abstract things as when it's appropriate to praise the player.

my action handlers all do that. "Shana: Congratulations! You made a spear!". if i wanted to, i could go into more detail, commenting on the quality of the work, how much experience they got and in what skills, and any mood boosts based on success or player defined special interests (such as crafting). but i prefer to keep the front end minimalist and keep the back end detailed. we should see effects, not be obsessed with numbers.

tracking player progress and providing feedback simply requires tracking player progress and providing feedback. a quest might be one way to do this, but its not required to do this.

it almost seems as though folks have discovered a number of new useful applications for "quests" and have so fallen in love with these new uses for quests that they have forgotten the original use for quests: epic adventure.

well, Caveman has a dedicated tutorial game already, and a caveman-o-pedia is planned, so i'm guess i'm going to use quests for their original purpose: epic adventure. beats the heck out of yet another "X gp for Y hides" game.

wish me luck!

>> I think after talking to the right guy and showing him the the lighter I got my weapons back and that's why I remember killing Benny with a cattle prod and a combat knife. Wasn't terribly satisfying though. No idea what level I was.

you should try the "at how low a level can you kill Benny and friends" challenge. its makes for a pretty good and satisfying play thorough. of course, there may not be much left of the world in your single minded pursuit of benny. almost everything you encounter brings you one cap closer, one gun closer, one more round of ammo closer to the big moment. and when the moment comes its like getting to play that scene from the matrix - you know the one. ultra cool - i die at least ten times before scraping through with one HP left - AFTER burning through two dozen stimpaks!

happy gaming!

>> Norman, didn't you specifically design a storyless sim survival game??

BG! it started as one day, out of the blue it just came to me that: "hey! wouldn't it be cool if there was a game where you were a caveman and could craft a stone knife and take on a saber tooth tiger?"

don't even ask me where it came from i have absolutely no idea! it just sort of popped into my head fully formed. you 'd have a cave, which would mean you could take over caves. you could go outside your cave and wash your face in the pool of recent rainwater, and see your reflection in the water, and play with the water and reflections and ripple effects. all with killer lighting and high quality plant models. IE a highly immersive environment (give me another decade and i may get there! maybe... <g>). then you would craft your knife all highly immersive, with particle systems for the flakes flying, reflecting light and casting correct real time shadows as they go and such (lighting is everything). about the time you finished, your buddy shows up (multiplayer co-op), and you head out to where the tiger is. that was (and still is, i guess) the original vision for the game.

but up to that point, i'd only written arcade games, flight sims, wargames, and text mode and 2D RPGs. i had no idea how to write a.... what would you call this thing? an open world survival sandbox stone age fpsrpg / person sim? its grown to the point were "paleo-world simulation" is probably the most accurate description. in fact i'm considering calling it: "Caveman v3.0: a paleo-world simulation".

so now i've just about finished building the world, and all the mechanics, except for final graphics and an audio remix. all the actions and objects are in except for the few that require lots of new graphics. with the world largely done and playable as a finished or at least a late beta product with place holder graphics, it time to give the player some optional quests.

i have no problem with story. i think i would be the coolest thing if i could come up with some plot point reveal system / engine that would allow multiple optional main quest sized stories (and smaller stories too) to be discovered, and followed, and played though by the player. with new DLC stories from time to time. i'm thinking it might be doable. that holy grail that lets you play stories. i mean we all know what we want, we want a holodeck program. we just haven't figured out exactly how to code one easily yet... but i'm not planning on a story engine at the moment (at least not yet).

Oh, I see... I guess that was my preconception, then, sorry. Normally the people I have encountered who strongly want to make and play sandboxes are the same people who are anti-premade-story in games. Typically games labeled 'sandbox' have no story and minimal quests. And normally the people I have encountered who want both story and sim gameplay (like me) prefer a sandpark structure instead of a sandbox one.

Aside from that, I'm not really a fan of 'epic' stories myself - I like stories with a more personal scale, which are about individuals' thoughts and feelings. I can pretty much agree with the idea that a holodeck program is the goal though. I usually think in terms of directed dreaming as a recreational activity, as seen in Tanith Lee's Biting The Sun, but it's a lot the same as a holodeck, except easier for you to take the role of something with a completely different body, like an animal, and I've never seen a scenario where the holodeck is simulating magical abilities. But ok, we can ask ourselves, "I'm in a paleo holodeck program, what do I want to do?" Me, I'd probably want to find a place no one else was living and build myself a treehouse like Swiss Family Robinson. Then, maybe I could get an impressive fur cloak and make myself some colorful facepaint and put it on. Since you can't learn to do magic in this game, maybe I'd want to become a master of javelins or throwing knives (like the sling in Clan of the Cave Bear). For bonus points I could tame a pet raven or fox or something. Now I'm ready to meet other people, since I can look impressive and defend myself while doing so. I don't have aspirations of being a leader or collecting followers, so the point of meeting people would be to decide if I want to seduce any of them. I guess they might also have interesting information I could learn from them, but I'm not sure what that might be, since there's no magic, no mystery about the world's past or my personal past, no big bad guy trying to destroy the world, etc.

If the concern is to have interesting quest activities, then I would agree with the previous statement that the crucial thing is "non-trivial" quests. Gameplay is gameplay and by its nature will either be fun to the player or not depending on its challenge/etc. To make the quest interesting means to make the story context surrounding the gameplay interesting, and interesting stories start with interesting characters. As such, I think your focus should be on how to make memorable and unique characters and how to invite the player to develop a relationship with them through your gameplay interactions.

The player will of course be limited by what gameplay you have programmed in, but even if your task WAS to collect 10 wolf hides or what have you, the quest itself can be made very interesting by getting into the details of the characters involved in the quest, their relationships, and how those relationships are changing as a result of your involvement and decision-making. The more personable, recognizable, and fun character you can put into the "people" of your games, the more the player will actually be immersed in their conversations with them, meaning that they will be able to "believe" that whatever task you've assigned to them has meaning in the narrative and is worth their effort to invest in.

That said, it certainly HELPS when you have a variety of quest ideas to use for gameplay tasks. +1 for sunandshadow's suggestions on quest types:
>>

1. Gathering materials from the world whether by drop-hunting, growing crops, or gathering

---1a. the degenerate case of this is gathering XP as combat 'training', like a student ninja or knight or something

2. Developing your social reputation in positive personal ways by doing favors for NPCs, in intimidating ways by killing boss monsters and revealing who was behind crimes, and in professional ways by earning reputation with factions and pleasing class-related mentors in order to gain new class-related skills and ranks.

3. Achievements of dexterity and/or intellect such as solving puzzles, crafting complex objects, amassing collections, high scores on minigames, stunt jumps, PvP rank, making money in the auction house, breeding a rare pet... what's possible here varies widely by game.

Getting that variety in will take a lot more development time though, so it may be better to focus on a small subset of activities, but make those activities deep systems that result from the intelligent combination of several small and simple systems (ex. Portal's puzzles, Super Mario enemy/level design escalation and sequencing, etc.). That'll get you the perception of gameplay diversity while not too wildly blowing up your production budget as you try to create gameplay content for any story/character development you work in.

>> And normally the people I have encountered who want both story and sim gameplay (like me) prefer a sandpark structure instead of a sandbox one.

sandpark vs sandbox - please elaborate

>> I think your focus should be on how to make memorable and unique characters and how to invite the player to develop a relationship with them through your gameplay interactions.

hard to do in a quest generator - think "procedurally generated content" how to you program memorable and unique characters into a quetsgen that will make dozens or hundreds of quests for the player? how do you make them memorable and unique? backstory, dialog, character graphics, and AI behavior are the only way to convey character info to the player. do you generate a backstory, generate graphics, generate dialog, and generate AI behavior? sure - hard coding all this stuff is trivial - but generating it? not that it couldn't be done. a storygen would make some backstory. some sort of dialog gen would make dialog - not sure how that wold work, maybe have a dozen different custom dialog sets and pick one at random. but with those types of approaches, unless you have a very large number of "parts" to choose from, it eventually becomes repetitive. same approach as used for dialogs would probably be used for character graphics and AI - IE mix and match random components or sets of components. unless the entire system had tons of parts to choose from, the limited number of permutations would be noticeable .

>> but make those activities deep systems that result from the intelligent combination of several small and simple systems (ex. Portal's puzzles, Super Mario enemy/level design escalation and sequencing, etc.).

an admirable design goal, but how would you apply those concepts here?