Psychochild's Blog

A developer's musings on game development and writing.

10 June, 2007

Weekend Design Challenge: Dynamic worlds
Filed under: — Psychochild @ 8:03 AM

This week, someone emailed me in an interesting challenge they wanted me to post up. Being the lazy sort, er, I mean… generous sort, I decided I would post it up for them.

This week’s challenge involves the question of dynamic worlds in MMOs. It was sent in by Bobby Thurman.

Here’s my challenge:
When making content for a MMO, the designer must be careful to allow players to progress independently through the world. This allows the world to become more personal for the player.

You also want players to be able to affect one another, group up, and change the world as a result of their actions. This is at odds with the quest giving NPCs remaining static for the younger players versus the world changing for the sake of the accomplished older players.

How do strike a balance between individual player experience and the shared world altering experience?

This, of course, is not a new topic. The dream has often been to try to have a dynamic world, but Bobby correctly points out that this can be disrupted, especially to new players. We’re running into this a bit in my current project, because we want the world to be very dynamic, but without some touchstones for players to reference, it can be overwhelming. Even logging off then logging back on to a radically different world during patches can cause people to get cranky. Now imagine that happening between the time you log off for a night and then when you log on the next day.

So, what do you think? Where is the balance that needs to be established between having a dynamic and exciting world and having something players can understand?

I’ll post some thoughts a bit later.

« Previous Post:


  1. A lot of that is being able to pin down precisely what defines the gameworld at its core, its major regions/characters/elements, and then ensuring that the dynamics apply only to the traits peripheral to those core identities. It’s not even necessary for the player to be consciously aware of the core characteristics, as long as identification is still natural and immediate when the core object changes.

    For example, I’ve never been to New York City, but my understanding is that it is divided into major regions. Manhattan is upscale, Harlem is poor, while Brooklyn is largely industrial (maybe I’m wrong about that, but let’s say it’s hypothetically true). Now imagine the city re-imagined to fit a Star Wars style of flying vehicles, strange futuristic architecture, and every building reaching up into the clouds. If the regions are still sharply segmented by their core characteristics (upscale, poor, industrial), old street names and corporate logos remain, and dialects vary from one region to the next, the city may still be recognizable as New York. In hindsight, a city’s a complicated example, but you get the general idea, I hope.

    Nothing is wholly inviolate. Even the most central characters and places of the game’s lore can change dynamically without disturbing the player, but you have to distinguish peripheral traits from core traits and (generally) allow only the peripheral to be altered. The more dramatic the change in character and closer that change is to the character’s core, the more necessary it is for that change to be witnessed by the player directly, rather than after the fact.

    Comment by Aaron — 10 June, 2007 @ 9:40 PM

  2. I think that a lot of the problems comes because you want the game play experience of the new players to be the same as the game play experience of the old player when they first logged in. Why?

    Why not make it so that after a certain amount of change those new players will get to max level (to pick a levelling tread mill example) much quicker. Why not make it so they ‘catch up’ with the people at the bleeding edge.

    Equally I think this is because your trying to have “world” changing events – and frankly people don’t work well at the world level. So shrink your vision and have community changing events (150 or so people). What I mean by this is that guild of uberness can kill – permmentally – the lord of nastyness and for there community he’s dead. Totally and utterly dead. But for another guild/community – the guild of newbness he’s still very much alive.

    Comment by mavis — 11 June, 2007 @ 4:17 AM

  3. Time for a new computer

    [...] weekend challenge for the week is one of those that really resonates with me this time, related to dynamic content [...]

    Pingback by Voyages in Eternity — 11 June, 2007 @ 4:48 AM

  4. I have also been musing on this precise aspect. I think every game developer’s dream is to develop their own world one day, ie a MMO.

    Anyway, as you stated it might become harder and harder for new players to enter the world if the world is indeed mostly dynamic…

    One solution could be to simply have a “single” player tutorial of sorts and then unleashing the player into the world, or a sectioned zone in the world where you can’t change the dynamics.

    I myself would love to see a proper dynamic world, but I hate it every time when I wasn’t there when something awesome happened, ie like World-Events in WoW (which isn’t even very dynamic) to the extremes in Eve-Online. I guess you have to take the good with the bad in a dynamic world… Everything changes and you can’t hope to be there for everything.

    Comment by Tr00jg — 11 June, 2007 @ 5:11 AM

  5. Have a GM film events so that people who couldnt attend can watch what happened. :)

    Comment by paul — 12 June, 2007 @ 3:29 AM

  6. You need a dynamic world but you also need recyclable content. I see a few ways to make this work:

    1. Dynamic respawns – Near the starting town there are a set of caves (let’s call them the Caves of Chaos :P). When the game stars, there’s a group of bandits inhabiting the caves. The bandits have a leader who’s bigger and tougher than the normal bandits. As long as the leader is alive, the bandits will respawn. But kill the leader and the respawns stop. The bandits eventually despawn (if they’re not killed) and leave, never to return – at least not these bandits. After a suitable time, the spawning system kicks in and “re-seeds” the caves with a new enemy set. For example, now that the bandits are gone, a colony of giant ants has moved into the cave. Or goblins. Or a rogue necromancer and his undead minions. And so on. In short, you provide the permanence but you still recycle the content.

    2. Territorial control/influence – Let’s suppose that our game world has two nations that are competing with each other for resources. These nations are seperated by vast wilderness at first, and, of course, the wilderness is inhabited by many monsters. Players can use a “city-building” system to help establish outposts in the wilderness that are affiliated with their chosen nation. The players build the outposts, and the nation provides troops and services. As this goes on, the wilderness becomes more civilized and the monsters are pushed back, but the other nation also starts to get nervous about the outposts encroaching so near its borders, and may send agents to attack or sabatoge the outpost. This can also be played into classic RvR PvP.

    3. Meta-discoveries. Suppose you have a game set in space. Your players spend their time fighting space pirates, making interstellar trade runs, mining asteroids, and so on. When the game starts, players can craft only basic technology – fusion reactors, ion engines, small lasers, etc. But the game includes a research system that allows scientifically-minded players to perform research using various looted/mined elements from the galaxy. Each successful research experiment performed has a chance of unlocking a discovery, in the form of a schematic for crafting a more powerful version of an item. This is the immediate reward for experimenting. But in the background, there’s a tally going of the number of discoveries made, and once a certain discovery or type of discovery has been made often enough, it become standard and all players and NPCs thus inclined get the schematic. As time goes on, players advance the galactic level of technology through their actions.

    4. Lore-based scripted events. Suppose that throughout your game world there are monoliths of ancient power. No one knows what they do or what they are for. But there are clues hidden away in the form of ancient tablets and artifacts in various ruins or other places. Players who collect the clues might figure out that a certain ritual can be performed to tap the power of the monoliths and use them as teleportation devices. When they undertake the ritual, the monoliths do indeed open up rifts – and out spills an army of demons. Now the players have to find a way to close the monoliths across the world and fight back the demon armies. After the monoliths are all closed again, they remain dormant for a period of a few months, and then maybe something else will happen…

    There you go, just a few ideas on making a dynamic world :)

    Comment by David (Tal) — 12 June, 2007 @ 4:29 AM

  7. Multi-cyclical quest chains.

    The scenario would run like this… when the game launches, you have a town and a camp of bandits out in the woods that are at odds with each other. Both sides have quests, each rewards with exp, faction, money, items, just like typical quests. The town gives you quests for random things as well as “spy on the bandits”, “clear the woods of thieves”, etc, things that oppose the bandits. The bandits give out quest like “gather supplies”, “steal from town”, “kill a city guard”, etc. Basically, if you are creative enough you can think of tons of quests that oppose each other. Each quest completed adds to a “power counter” on the server side, and new quests pop up as the other side gets stronger. If both sides reach a certain level of power, a war event triggers, players participate, war ends when the bandits or town runs out of soldiers… if the town wins, things go back to normal, bandits are reduced to “camp size”, town is where it is, minus some power equal to percentage of soldiers lost. If the bandits win, they take over the town, the factions switch, and the old “good guys” now because the resistance hiding in a cave. Now the town is “evil” and the camp is “good”. Two new sets of quest chains that work like the old ones, but with a different flavor, the good guys are trying to retake the town and save the townspeople, the bad guys are trying to maintain control of the town.

    When a player logs in, if the balance of power has shifted since they were gone, they’ll get a page of text or a cut scene animation that explains “The bandits, under the leadership of their warlord have stormed the gates and taken the town! What forces remain of the city guard have moved to a hidden cave to plan their resistance…” and the game will log you in to the appropriate faction location for your character. But that cut scene is key, you have to make sure that major events that happened are related to the player BEFORE they actually get into game, at least on a high level… once inside you can have an NPC that’ll say “The bandits make have taken the town, but we fought well and they suffered great losses.” Maybe, as paul suggested above me, on the GM side have them be monitoring the status of the cycle and be there when the event triggers so they can record and post videos of it on the official site.

    Comment by Jason — 12 June, 2007 @ 4:36 AM

  8. I really like Jason’s first suggestion. Having MoB leaders makes griding more interesting and more akin to solving a puzzle.

    I’m still having problems that are much smaller than the large solutions presented here. I’m having to use the following guidelines here in a multiplayer world:

    1) Quest givers can’t participate in combat because another player may walk up between NPC spawn times and get confused. This might be a source of CSR support issues too.

    2) Quest givers can’t talk out loud most of the time to a player because it may spoil/reveal a later quest.

    3) You can’t ever really save the princess because as soon as you begin to escourt her back to the castle, she’ll respawn at the tower again.

    These are some pretty tiny problems with dynamic worlds, but these activities are the bread-n-butter for a quest driven MMO. Some of these issues can be resolve with instancing, but this isn’t always feasible, and instancing is super expensive. I know that WoW (all bow before great WoW) violates a bunch of these rules, but even so, it leaves a bad taste in my mouth. I’m starting to wonder if I should just hold my nose and swallow, or is there some better approach.

    Comment by Bobby Thurman — 12 June, 2007 @ 8:29 AM

  9. 3) You can’t ever really save the princess because as soon as you begin to escourt her back to the castle, she’ll respawn at the tower again.

    Solution – dynamic template-driven quest generation. See Anarchy Online or SWG’s mission terminals for ideas. It takes a robust system on the back-end to make it work though.

    The basic concept is that each player who approaches the quest-giving NPC gets a different quest. So my quest might be to save the princess, while yours might be to recover the enemy plans, and another person might get a quest to map the secret passage that leads out of the throne room for a future assault. And so on. You can feed this in to some of the stuff that Jason talked about too, where the quest templates change based on the progress of the ongoing storyline.

    The hard part here is that you’re going to have to be very, very careful about what types of quest storylines you build into dynamic templates. It’s possible with a multi-stage template to do some pretty involved quests, but things like “rescue the princess” that you really want to be one-shot world-changing events shouldn’t be offered via template. Instead, what you want is for the template quests to increment counters on the back-end that lead up to the special one-shot events and unlock them. So if you do enough missions against the enemy, then the special event becomes available. Ideally also you want the special event to be something bigger than a single quest so that as many players as possible can participate.

    Comment by Talaen — 12 June, 2007 @ 9:16 AM

  10. It really comes down to the mental model of the system. If the player is firmly set to believe that the world is static then the mental model is disrupted even by patches. If you can teach the player that the world is dynamic it wont really cause a problem even for newbies, but it will force you to design a different type of learning experience to your new players. And that starts to borderline with innovation which is a scary thing for big budget productions.

    Comment by Wolfe — 13 June, 2007 @ 5:53 AM

  11. 3) You can’t ever really save the princess because as soon as you begin to escourt her back to the castle, she’ll respawn at the tower again.

    Sure you can… in the same way I outlined the other idea… there is a quest to save the Princess, if she is saved, the “other side” now gets a quest to kidnap the Princess. Even better… when the Princess is rescued from the tower, a cryer appears in “bad guy town” and shouts “They taken the princess!” Now you not only have to save her, but the other side will be looking for you to try and stop you. Same thing the other way, “Oh no! they’ve kidnapped the Princess!”

    Comment by Jason — 13 June, 2007 @ 2:41 PM

  12. That’s a very creative and workable solution Jason!

    Comment by Bobby Thurman — 14 June, 2007 @ 6:57 AM

  13. The problem with creating dynamic content is ensuring that it doesn’t end. Every quest has to lead to “something” even if that something is to go back to the beginning. The reason I just multi-cyclical quests instead of single cycles, is that single cycles are easy for players to spot, bothers some of them, encourages others to farm it, etc… multi-cycle quest lines, especially when offered to two opposing sides in a PvP game leads to more competition.

    The other thing you need to further watch out for is being intentionally not completing a cycle to deny content to the other side. If that happens at any point, you need to adjust your quests to “time out” if left undone so that given a reasonable amount of time (a day, a week, a month, etc) the game will rotate the cycle. Using the princess example, if the bad guys never kidnap the princess and a week goes by, have the princess sneak out to go horseback ridding and vanish. NPCs have just kidnapped her.

    Comment by Jason — 14 June, 2007 @ 12:44 PM

  14. The outdoor pvp in WoW such as the towers in Hellfire or the pvp town in Nagrand has this type of “cyclic quest objective” with a pvp context. The actual effect we can see in the game is that the pvp context (such as kill enemy players while meeting the objective) does not mix well. What happens is that the opposing sides give the cycle away for free to each other.

    The reward system generally says that to get the reward you need to win the battle against the enemy, then the enemy can be rewarded by winning against you. The most efficient way to get many rewards is then to not defend when the enemy attacks. Which removes the pvp context, this confuse players who believe it is a pvp system.

    Nowhere in WoW have I seen a pve cycle yet, altho I think it would make more sense as game. I also believe it would make more sense for these types of things to be structured around player groups or guilds rather than hard coded sides (such as Alliance vs. Horde).

    Comment by Wolfe — 15 June, 2007 @ 1:14 AM

  15. Dynamic content

    [...] finally found a little time (due to insomnia, of course) to put together a little post on the dynamic content weekend challenge over at Psychochild’s blog. The comments on that post so far have already been top-notch, [...]

    Pingback by Voyages in Eternity — 15 June, 2007 @ 4:31 AM

Leave a comment

I value your comment and think the discussions are the best part of this blog. However, there's this scourge called comment spam, so I choose to moderate comments rather than giving filthy spammers any advantage.

If this is your first comment, it will be held for moderation and therefore will not show up immediately. I will approve your comment when I can, usually within a day. Comments should eventually be approved if not spam. If your comment doesn't show up and it wasn't spam, send me an email as the spam catchers might have caught it by accident.

Line and paragraph breaks automatic, HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Email Subscription

Get posts by email:

Recent Comments


Search the Blog


June 2020
« Aug    



Standard Disclaimer

I speak only for myself, not for any company.

My Book


Around the Internet

Game and Online Developers

Game News Sites

Game Ranters and Discussion

Help for Businesses

Other Fun Stuff

Quiet (aka Dead) Sites

Posts Copyright Brian Green, aka Psychochild. Comments belong to their authors.

Support me and my work on