It's ironic that one of the easiest elements to design, develop and test in a game is also the least satisfying-- prescripted, repetitive interaction. Without having an army of content developers constantly updating your game (hence the restriction in the thread title), what are some ways that you can build dynamic experiences into a game? Don't worry about type of game, I'm making this wide open on purpose for the sake of brainstorming. Some thoughts I had: Context - Even if I'm doing something repetitive, as long as it's not mind numbingly so and the reason for doing so shifts, I'm can have a different experience due to a shift in my motivation. Hence why story / quests / missions can be effective. Varying the Simulation - If attack / damage stats in combat change, or the rate that blocks drop speeds up (or slows down due to a powerup), my experience of the game can change. Items, events and state shifts can all be the cause for the simulation to change. Impressive Variation - If I'm wading through goblins and get to one twenty feet tall, the very fact that a pattern has been established and then broken can shift my experience (especially if the simulation varies in proportion). Any others or other thoughts on this?[Edited by - Wavinator on August 2, 2008 1:07:18 PM]

My suggestion would be to create a simulation with as many interlocking game mechanics as possible. Then hopefully emergent gameplay will appear. For involuntarily hilarious results, see Dwarf Fortress.

Randomised scenarios or dungeons.

The layout and design of the dungeon is static. NPCS, encounters, traps, monsters, their positions and dialogues are randomised.

Warning: Wall of Text

When I designed a little 'MMO' server in NWN, I randomized all the dungeons and I randomized all the quests. I also got rid of all dialogue for the most part, and filled it in with 'types of responses'.

There was not a single pre-scripted quest on the whole server. Just a bunch of different NPC types wandering the towns that represented different types of potential jobs. The player could talk to any of these NPCs and be offered a random job. They could accept or decline it, and have up to 3 active quests stored on their character at any one time.

The quests could be anything, and new types of quests were come up with all the time. A random location was chosen from anywhere in the game's world, and then the quest would spawn there when the player entered the area.

You'd talk to a shop owner and see 'This shop owner is offering you __ to go and take care of a group of __ who were last seen heading __." [Accept][Decline][More Options]

A thief completing a mission to steal some art in one town might trigger a counter quest to go and steal it back later. The world had 4 major regions and players had a rep with all of them. Doing jobs for one faction would lower your reputation with the opposing faction.

I made the towns people react to players based on their reputation. They might greet them, or taunt them, or do a number of things. If you got to be on one factions good side, you'd be a hero in the eyes of the locals, and possibly get higher caliber quests to go and mess with the other faction.

I let dead bodies of NPCs and slain players be used for various things. One of those things was leaving the corpse on the ground, impaled or otherwise for other players to stumble across with messages left on them.

All the gear was either randomly generated, or player created. Players were free to create the look and stats of their own items.

While it was up, everyone who played loved it. My system created a nice space that let the RPG players actually role play. They got to define their character, make their own gear, and actually do whatever they wanted to do.

If you wanted to be a monk who wore a pink and blue outfit and only ever attended to matters of the forest, you could do it. The game would keep making you new random quests until the end of time.

Same with any other role you wanted. The game treated all the players as wandering adventurers, and created random adventures for them as if they were characters on a weekly fantasy adventure series like Xena or Hercules. There were no level zones in the world, all the spawns were based off the levels of the characters in that zone. So it was always possible to stumble upon some strong monsters. The danger was always there.

The respawn options were very forgiving too, allowing the player to choose a suitable respawn option, instead of having a strict set of rules that would inconvenience people in different situations.

You never knew what would happen in any one session. It was just another chapter in the continuing adventures of PlayerName, the wandering adventurer. [lol] I just put the world there, and let the players make what they wanted of it. I never put dialogue in their mouth and let them respond to NPCs in general terms, and imagine their own words. Even to the point of making quest item drops usable for other purposes in case the player chose NOT to turn them in for quest completion, which was sometimes the case.

I was always coming up with new ways to use these items. It might be more fun to use that dead corpse of the enemy to make a booby trapped scarecrow with a message on it warning the other players not to pollute the nearby forest with their unwanted drops. [lol]

If you rescued a child they would follow you. You could return the child to the mother, return it, but demand more money, take the money, keep the kid, bring the kid to a roaming caravan and sell them to slavery, etc,etc..

Some of those were exhausting to implement, but the players always had a ton of supported choices in all steps of the random quests.

I'm very proud of all that design. I wish I still had the module file around. Later NWN patches opened up even more possibilities.

This is as opposed to Morrowind, where you are supposed to make your character, and then do what you want in a sandbox, but the sandbox is very limited, and the content is static and limited. You can't be a career thief in Morrowind, because there's only a dozen or so thieves guild quests before that entire avenue of gameplay is shut off.

I just don't understand why people make MUDs and MMORPGs the way they do. With level range zones that you play through one until you get sick of them, and then move on, doing the same pre-scripted storyline missions over and over again, like they were in some amusement park. How are people supposed to role play in a game where NPCs are like the characters you speak with at a drive-thu?

This type of thing works fine a game like Final Fantasy or Dragon Quest, but not in a persistent online universe. When I was in the WOW beta in my friends on the dev team of the NWN 'MMO' server, we'd laugh at how primitive and limiting the design was. Not to say that it wasn't well done, or fun, as it was both.

In our module, you could be who you wanted, look how you wanted, make your own gear, and do whatever the heck you wanted. In WOW, you pick a character that looks like everyone else, you progress like everyone else, and you have half a dozen realistic equipment choices, making you further appear like everyone else. If you want to go adventuring, you go to the predefined adventuring area for people or your range and you fight the same dozen bandits that all of the 250 players in that region are also fighitng, for the same '1 of a kind' drop. [rolleyes]

So much for immersion.

And if I want to try a new class, I roll up a new character, and do the same old stuff all over again. Catch up with those same dozen bandits who seem to be stuck in a time warp, constantly stealing the same 1 of a kind rare drop. [lol]

---

Next Genre: One of my favorite games of all time, FPWR, gives you 500 character edit slots and lets you make over the game in your own image. There is a very deep AI editor, and creating characters and playing with them, or watching them go at it never gets old.

Daaark, I completely agree.

That is exactly what the MMO needs. And you could even apply it to recent developments in the industry, like WAR's public quests! Imagine dynamic Public Quests!

And on top of that you would have various factions launching attacks and defending across various parts of the world, which provides obvious and easy-to-implement quests, such as "Defend the town!", "Siege the walls!", and "Help cnstruct a barricade!".

Wavinator, you mentioned that you would need an army of content developers constantly updating your game in an MMO. Using dynamic, procedural content generation, you could apply the exact same system to either a single player, or multi-player game. Generally, I think of 'quest' generation as "Context" Dynamics. And the more quest structures and variables you have, the more complex and fun the generated quests become

I imagine creating a world where you could just let the economy/ecology/faction warfare run itself, creating conflicts, and thus objectives for the player. Perhaps a townsperson decides to become a bandit, gathers followers, then begins to prey on passing caravans? A fisherman hasn't caught anything in a week, so he packs up and heads to another shore. A clan of giant spiders have finally spawned enough children into the brood, and launch a full-scale assault on a neighboring NPC faction for fresh meat in the winter.

All the developers would do is occasionally improve the generator, and add the really epic, high-level quests and world-objectives themselves. Developers create a new race of insect-beasts? Suddenly the ecology and the other factions begin to react to it. Launching attacks, capturing and studying the creatures, bolstering defenses on certain borders, establishing new outposts.

No system is so complex that it can't be broken down into a computer algorithm eventually.

Quote:

Original post by Humble Hobo Daaark, I completely agree. That is exactly what the MMO needs. And you could even apply it to recent developments in the industry, like WAR's public quests! Imagine dynamic Public Quests! And on top of that you would have various factions launching attacks and defending across various parts of the world, which provides obvious and easy-to-implement quests, such as "Defend the town!", "Siege the walls!", and "Help cnstruct a barricade!". Wavinator, you mentioned that you would need an army of content developers constantly updating your game in an MMO. Using dynamic, procedural content generation, you could apply the exact same system to either a single player, or multi-player game. Generally, I think of 'quest' generation as "Context" Dynamics. And the more quest structures and variables you have, the more complex and fun the generated quests become I imagine creating a world where you could just let the economy/ecology/faction warfare run itself, creating conflicts, and thus objectives for the player. Perhaps a townsperson decides to become a bandit, gathers followers, then begins to prey on passing caravans? A fisherman hasn't caught anything in a week, so he packs up and heads to another shore. A clan of giant spiders have finally spawned enough children into the brood, and launch a full-scale assault on a neighboring NPC faction for fresh meat in the winter. All the developers would do is occasionally improve the generator, and add the really epic, high-level quests and world-objectives themselves. Developers create a new race of insect-beasts? Suddenly the ecology and the other factions begin to react to it. Launching attacks, capturing and studying the creatures, bolstering defenses on certain borders, establishing new outposts. No system is so complex that it can't be broken down into a computer algorithm eventually.

I guess we all begin to imagine how an algorithm like this might be constructed when we talk about ideas like this. Maybe the really hard thing though is to rein down the potential chaos of the system into a consistently enjoyable experience.

Quote:

Original post by Humble Hobo Using dynamic, procedural content generation, you could apply the exact same system to either a single player, or multi-player game. Generally, I think of 'quest' generation as "Context" Dynamics. And the more quest structures and variables you have, the more complex and fun the generated quests become

I think there's an unseen tradeoff here, though. Aside from testing for bugs and playability, which would have to be a huge amount of work, I think there's a problem with meaning. How do you do something like this without eliminating the narrative and explanatory content that would give context to the events in such a shifting world? Spiders are invading because of such and such, but how do you communicate that to the player in a convincing fashion? Zoom into the player's perspective and ask what keeps the fisherman's movement from being a cryptic "huh"-- one minute he's there, the next minute he's gone.

I think it's a cool sounding idea, but I'm pretty pessimistic about the implementation details. One possibility could be to expose the simulation to the player up front by using conventions, I guess. See lots of spiders? That means they've got enough brood to expand, not that an cryo-hydra has invaded their lair or whatnot. Then the player knows roughly what is happening and why, and hence what it means to them and what to do.

I handled that by having the guards of a city or forest encampment tell the players that they are getting out of control and they will be rewarded for kill 'x' amount of them. I used the player's actions to keep giving them the kind of content in reaction to their own actions. If you sell rescued kids to slavery, or harvest body parts off them, or anything else, then the towns people start to tell stories about you, and you are known as someone who can't be trusted around children. [lol]

I think people who have never implemented something like this give the theory of less meaning too much consideration. It's a given that this is a side effect of continuous randomly generated content, but every system of implementation has it's bad sides.

But, it's fun! It may not work in every genre, but it's fun to play. Not every scenario needs a giant amount of work for voice acting and story. The spiders are invading because they are. That's all there is to it. Kill them, or find something else you'd rather do.

I read something about this on a RPG site about real DND DMing while I was programming my NWN MMO system and putting a lot of thought into the emptiness of the little plots. The article went along the lines of pointing out that to most players, and to beginning players, the simple plots of saving the town from some orcs was good enough. They were invading or threatening the town, and that reason was good enough. Killing them and saving the day made for a fun game, and that was all that was needed.

But some long time players would want to look too deep into it, and DND was getting to the point of everything needing to have an overly complex plot, and every enemy you encounter needing to have a long thought out motivation and being a part of something bigger. The eventual point was that all that stuff wasn't really needed, and the simple plots that focuses on the fun gameplay were all you needed for a good session.

In my experience, my depth of player options more than made up for the depth of plot. (There wasn't even a plot, I let the player decide). The goal was simply to have fun, and enjoy the company of the other players while they did it.

Either way you can't please everyone. All my players liked it, but I'm sure some people wouldn't have. Just like some people don't like PvP servers, or certain game genres.

You just have to make the randomized system more deep than making quests that can be completed one way and only one way. Let the player choose their own way to resolve the problem, and let the reward item be useful in many ways. A quick turn in object, a new quest starter object, and maybe a few alternate turn ins locations that get different results, and maybe let them keep it as a power-up object, or an ingredient in something else.

Grand Theft Auto would be been a bit better if the content didn't simply end, leaving you with a big world to roam and nothing to do. They could have added random missions like the ones I described, especially since the functionality was already shown to be in the game!

-edit-

I keep forgetting to add in a point I wanted in the original post. The games that people play for a long time over and over again are usually games that only have little bits of content but allow lots of freedom or interaction, with addictive gameplay systems.

Diablo 2 - random content, but not a lot of it.

Fighting Games - Not much content, but people play them for years.

Wrestling Games - When they are fun, and the very deep character creation systems that the fighting genre is just now picking up on.

Sports Games - Addictive gameplay only needs small yearly adjustments.

Games that grade performance and have unlockables and stuff for repeating through content and allow tracking of performance stats.

By far the simplest way to add meaning and context is pure violence.

If an NPC faction is attacking another NPC faction, you have an instant, easy motive for the NPCs there to ask you to help defend. Dynamic warfare and points of military interest (a magical power source, a fort, a key bridge or tunnel) make for obvious context.

The reason why I like the idea of a public quest, is that you don't need an NPC to invent quest text that no one will read, just to assign you the quest. It's sort of an area of effect distribution of quests.

And yes, the more complex the generator, the more fun and meaningful the objectives become, but it will take so much longer to create, debug, and balance the generator in development. It is a trade off, but a worth-while one because it is an investment.

----- edit -----

Of course, all of this is not to suggest that you can't have an overall plot objective.

My theoretical single player RPG uses dynamic content like we have discussed, but also has objectives that must be reached to get to the next "Chapter" in the story. It might take some players longer to reach checkpoints, but NPCs will also be constantly helping the effort.

An example:

Chapter 1: Forest folk discover demonic corruption in their forest. The chapter is broken down into world objectives.
Stage A: Trace the source of corruption to demonic presence.
Stage B: Reclaim X out of Y key points of military interest. Cleanse points of magical interest.
Stage C: Eradicate the Race of *something* from the forest that has been corrupted.
Stage D: Activate the forest defenses, defeating the old guardians of said defenses to do so.
Final Stage: Shut down the lesser demonic portals, and defeat the forward commanding demon at the main portal.

And so on... you can accomplish some stages in any order, and some must be done in a certain order. Some objectives might even be accomplished without you, such as Stage A. If some NPC discovers the source first, and alerts everyone else, the objective will be fulfilled. There is still plenty of plot and depth, but the means by which you achieve the goals are entierly up to you.

Quote:

Original post by Wavinator It's ironic that one of the easiest elements to design, develop and test in a game is also the least satisfying-- prescripted, repetitive interaction. Without having an army of content developers constantly updating your game (hence the restriction in the thread title), what are some ways that you can build dynamic experiences into a game? Don't worry about type of game, I'm making this wide open on purpose for the sake of brainstorming. Some thoughts I had: Context - Even if I'm doing something repetitive, as long as it's not mind numbingly so and the reason for doing so shifts, I'm can have a different experience due to a shift in my motivation. Hence why story / quests / missions can be effective. Varying the Simulation - If attack / damage stats in combat change, or the rate that blocks drop speeds up (or slows down due to a powerup), my experience of the game can change. Items, events and state shifts can all be the cause for the simulation to change. Impressive Variation - If I'm wading through goblins and get to one twenty feet tall, the very fact that a pattern has been established and then broken can shift my experience (especially if the simulation varies in proportion). Any others or other thoughts on this?

Variations of the tools at hand....

Have more useable objects (tools) and interactive terrain (furniture) that effect the game mechanics and get reactions from the NPC/Monsters which are part of the quests. Have a variety defined and have some random set place about the quest area (combinatorics would make it reasonably fifferent each time).

Of course to prevent the 'kitchen sink syndrome' the objects shouldnt be removeable from the vicinity so that they could be reused as part of a 'quest solving toolkit'.

Items that create alternate solutions to the sequence of step that lead to quest completion can create puzzles (and even fun just experimenting) that will make the quest more than another 'kill X, take Y, and return' grind.

The more complex interactions required can vary from fairly simple 'goblins are afrid of light so carry a torch' to physics where you 'collapse the faulty ceiling onto the Balrog with a firegall blast to trigger it, but only if he is right under it'. Something I always wanted was just to be able to ram in those stupid doors that always seem to be locked.


These extra objects and interactions would be something 'logical' not arbitrary so that the players can figure it out for themselves. They should be generalized so that all objects are governed by the same interactive physical rules. Likewise the monsters/NPCs react to different actions differently (so that you cant use that one 'good' sword all the time on everything, but have to switch to an appropriate tool, etc...