Dynamic POIs
Way back in Pre-CU [Star Wars Galaxies] while ‘walking’ from Eisley to AnchorHead a Twi’lek (I think) stated my avatar by name (could be wrong) and gave me a disk then some stormies spawned and killed her then came after me.
Anyone ever finish this quest? What was it like?
This was a rather complex quest. Does anyone know how this was coded? Why would my avatar be chosen over others?
The Twi’lek slave girl quest was part of what we called “dynamic POI’s.”
A normal POI is a “point of interest” — something to break up generic wilderness. it was a term we used back in the UO days that we got from Richard Garriott, and was probably older still. POI’s are normally placed by hand, of course; you sculpt a location for them, add a little bit of something unique or flavorful, maybe some interaction, and there you go. They can be as small as a little faerie mushroom ring, or as large as a bandit camp or something. In other words, they are the static content of a world… usually not the main quest lines, but just “interesting stuff.”
Of course, adding these in by hand is excruciatingly slow and requires an army of developers. That’s the cost of content. In a game as large as SWG, we had a real issue here. At one point, there was a large roomful of junior developers who did nothing but put down little interesting locations on the maps… and it was nowhere near enough, particularly since they had no interactivity with them.
Part of the solution that we wanted to try, then was dynamic POIs.
These were modelled after random encounters in D&D. Not the plain old table of boring ones that you see in say, a JRPG, where you roll up a combat encounter every few steps, but something a bit more intricate. In Ultima Online we had done this with orc camps and mage towers — spawns that included a building.Ā Since the map in UO was static, you needed a lot of empty space to get one of these spawns, and housing eventually used all the space up, so you never saw these happen. And there was no plot or story to these — they were just combat encounters that happened to come with a building.
SWG was more flexible. So instead we wanted the sort of thing that you saw in the better-written modules: random encounters with a little bit of plot, a little bit more context. Don’t roll up just a bandit; roll up a little bandit campsite, with a tent, a campfire, three bandits, one of whom hates one of the others, a young bandit who isn’t actually a bad guy but has been sucked into the life because he has a young pregnant wife at home… In fact, maybe have an assortment of bandits — twenty possible ones maybe. Then pick three for your camp. That way you always get a flavorful but slightly different experience.
This path led us to create dynamic POI types based on “plot” — such as “escaped prisoner”, “poisoner” etc. These could then be placed in different “theaters” — little layouts of buildings and objects that could be spawned on the fly and placed on the map. Because the SWG map was procedural, we had the theoretical capability to spawn these anywhere, by affecting the underlying terrain and terrain textures, flattening the space, and placing the objects on top of it. When the POI was complete, it would wait until no players were nearby, and delete itself and all associated structures and spawners, thereby restoring the underlying terrain.
I spent several weeks building theaters. A sandcrawler on fire with sand people attacking Jawas. Lots of little moisture farms. A small village. Campsites of various sorts. I also made the plot skeletons for things like simple murder mysteries, family feuds, etc. One plot involved two villages having a Romeo & Juliet moment. Another was a simple “who’s lying” sort of puzzle.
The very first example, however, was a tease: a dynamic spawn whereby an NPC slave girl was spawned who ran up to you and said “Help! They’re after me! Quick, take this, don’t let them have it or all will be lost!” You would be handed a data disk, and then the girl would run off, right into the arms of some Stormtroopers who shot her dead in front of you. You’re left with a disk you cant read and a mystery.
It was pure tease — originally, the disk did absolutely nothing (I want to say that later, the disk was actually tied into something else… a collection quest maybe?). But it was an example of the sort of modular interactive narrative that could easily be created. Swap out slave girl for other sorts of escapees — and maybe have two or three types of escapees — good ones, bad ones, etc; and two or three kinds of hunters: pirates, Empire, etc. You can read about Haden Blackman’s experience running into one of these during the testing phase here.
So what happened? Well, honestly, content is hard to make. A dynamic POI system is not any easier to craft than a usual encounter. In fact, we never cracked a way to make it fully data-driven, which meant that it was actually harder. The dynamic spawning system was problematic. Scripting in SWG in general was too hard, so we couldn’t spread the burden of creating the dynamic POIs, and each one was a moderately tricky custom scripted encounter. In the end, the dynamic POIs came out of the game, and all the theaters I made went unused.
There is also an impression out there that these sorts of content were all that SWG was supposed to have, which is incorrect. Dynamic POIs cannot replace good old-fashioned static content. A theme park has a mix of wandering performers and rides, after all, and all dynamic POIs do is supply that element of surprise that human statues or other “street performer” things do. (At one point I tallied up all the sorts of content SWG was planned to have, and it was something like seven different types, with dynamic POIs representing just one type. Others included lairs, true random encounters, static quests, etc).
In conception though, the dynamic POI system is pretty simple. Given proper resources applied to solving the problem, and most importantly, the large quantities ofĀ content required for it to work, it still seems quite viable. Today we see a modern form of it in Trion’s new MMO Rift: Planes of Telara, which is making dynamic content a big selling point, a full 14 years after the first crude ones went into Ultima Online, and long long after text muds had procedural zones with shifting content.
42 Responses to “Dynamic POIs”
Sorry, the comment form is closed at this time.
I have two distinct memories of your dynamic POIs. First, a corral which spawned on top of me, as I was running across Naboo. I suddenly found myself surrounded by fences, and since we couldn’t jump, got stuck. I had to page a GM to teleport me out of it. š
Second, while standing outside Theed, I saw a message that a DNA sample had just been extracted from me. I turned around, and saw a probe droid. It immediately launched itself into the sky, and disappeared. I found the experience rather disturbing.
Haha. Well, the first was a bug, as dynamic POIs were never supposed to spawn on top of people, but instead spawn on the edge of view between where players were.
The second sounds more like a random spawn than a dynamic POI, properly speaking.
Great Stuff.
I’ve always been a fan of this concept since I encountered it (as “Brief Encounters”) in Imagine, the brilliant 80’s TSR UK magazine.
Back in the days, there were a few RPG supplements that were built around this idea, like some of Catalyst’s City Books, The Companions’ “Places of Mystery” or GURPS Supporting Cast (where NPCs came with short adventure seeds).
I’d like to see more of these, even in single player games, where the games could tailor dynamic POIs to the PC’s state. Then again, maybe they already are heavily used and are so good that I don’t see the difference with static content š
I designed and implemented the slave girl POI and a couple others.
Dynamic POIs were a failure and reflected larger failures in how we built content on SWG.
Dynamic POI spawning was problematic for a long time. One algorithm used would spawn content in areas near players…every player was given an allotment of POIs. The spawning system would try to place the POI in the path of the player, so that as they explored they would run into some interesting even unfolding in the world. This was incredibly naive. Most players congregated in cities, so you’d get a crust of POI content around the edges of the city — just beyond the radius where they weren’t allowed to spawn. The player’s content allotment would then be met. The result was a thick ring of creatures and events and then the rest of the world being mostly barren.
The idea generation / content creation / idea validation process on SWG was broken. There _was no methodological idea validation_. We would have weeks of content creation and no one was actually playing that content. Designers would implement dynamic pois without really knowing if they were fun, if they would function correctly, etc. When playtests did happen they were overwhelmed by fundamental problems in the game design (broken combat, broken professions, etc). Getting a clear evaluation of secondary systems like dynamic content spawning was extremely difficult. As a result the system shipped in a broken state.
Another failure of dynamic POIs were the cost. They were expensive to implement, but only interesting the first time they were encountered. Once the player had played them they became more annoying than useful. They are, essentially, quests pushed to the player. The system doesn’t care (or even know) if the player had seen that content before. The dynamic POIs often involved complex branching dialog trees, multiple characters interacting on unknown terrain, etc. This involved a lot of implementation time for a payoff that didn’t scale to the investment.
Another failure was the fragility of any complex scenario when spawned in a random world location. A dynamic POI might break if it spawned an Imperial soldier near a rebel base — the two would start to fight interrupting the scenario. The designer had no idea where the scenario would be spawning or even if it was a appropriate location for the story they were trying to tell or the npc actors involved.
I better stop now before I have flashbacks.
My intent wasn’t to postmortem SWG! I agree the dynamic POIs never really worked, for many of the reasons you cited. That said, I think that many of the reasons they didn’t work were external to the core concept (example, the spawning routine you describe, or the fact that they are “pushed” quests — both solvable problems — in some cases, eminently solvable).
Nice article on a good idea that just wouldn’t work. It is nice and refreshing to see all the ideas and dreams you had for SWG that still haven’t been realized by any other MMO, and probably never will.
Tragic really. You were thinking so far out of the box, and WoW comes along and shoves everything back in the box and nails the lid shut.
[…] Tagged: mmorpg, raph koster, Star Wars Galaxies A recent blog post from Raph Koster discusses dynamic POIs in SWG and why they never really worked right. One of the comments in particular, from an old SWG […]
Ah the spawning code. I can’t count how many times I asked for it to be cleaned up. I still think it was the basis for a lot of fun, it just got sidelined by bugs and other priorities.
My hope is Telara (and perhaps Guild Wars 2) will provide a good working model for dynamic POIs so we can see more of this in the future. The effort involved may not be worthwhile to make it a primary focus for many games, however there is an important amount of awe generated from finding unusual content.
Thanks for the explanation. I always like the idea behind doing this but as noted, the execution was lacking. This seems on of those systems which needs to be well thought out from the very start and incorporated correctly.
Still this type of system really does have potential for more varied content.
Wasn’t the sand people attacking the Jawa crawler turned into a static POI near the krayt dragons?
Yes, a lot of the content was repurposed.
It is nice and refreshing to see all the ideas and dreams you had for SWG that still havenāt been realized by any other MMO, and probably never will.
Easy on the cynicism there. Posting it online makes it a lot more likely someone will adapt it to their own MMO. …if, you know, people still make the things.
Postmortem SWG? That’s funny, I’m still paying a subscription fee… I guess I didn’t realize it had died.
That’s an interesting choice of words, Mr. Koster.
pwh, don’t read too much into it. The term “postmortem” is widely used to describe the process of evaluating the game after it ships. So in some sense, a postmortem of SWG’s development is years and years overdue. š
I beta’d SWG, and… damn, it was so beautiful. So much potential. If we had only had six more months, maybe, to pick it apart and put it back together again, we could have got the major systems polished and spent some quality time on things that got back-burnered, like the dynamic POI system.
The counter-argument is that if it were up to the beta testers, nothing would ever ship. There may be some merit to that point. But hard ship dates with no flex are spawning premature births left and right these days.
In regard to dynamic content:
Star Trek Online has the Genesis system for spawning dynamic missions, but at least at this stage, it’s bland, rote, flavorless.
It feels as if what we need at this point is somebody who’s an expert on Aristotle’s dramatic unities and Campbell’s work on heroic myth structure, and who also has the chops to stitch together a database of plots, characters and locations into an infinite variety of rich, compelling adventures personalized for each player.
Tall order. WoW clones are easy. You’d really have to be some sort of crazy cockeyed visionary to tackle something like this…
@Yukon, said visionaries bring about great things. Henry Ford, the Wright Bros., Lord British….
That really is what we need.
From the sounds of how the Rift: heroes of telaran is going maybe a viable way to do things like POI that would help negate some of the hurdles SWG faced.
From the videos and things I have read it looks like they are using some kind of phasing tech on lines of what wow is starting to play with from wotlk and cataclysm.
Something like this type of phasing tech helps because now once a POI/rift/whatever spawns that area is now in its own phase. So if it phases in things in the existing phase such as existing rebel bases would not exist in that phase so not effect the POI or hinder it.
You could also potentially tie a phase to a person or group so instead of having the gauntlet of POI around areas where a lot of people roam you could have the system only display it to the group/person its intended for.
I think the technology is coming along enough that sooner or later somebody will be able to do more dynamic POI effectivly in a fun way.
I enjoyed the first few I saw in SWG back in beta like when I was running around the middle of nowhere and then I see a lamda class shuttle and imperial troops around it. It felt very starwarsy but as has been mentioned the times it worked right were outweighed by the times it just could not function correctly.
Still when aiming for something like this you would have to take great care to do a lot of testing to make sure the content is actually as fun or more fun than a more static area such as an instanced dungeon. If done wrong it could make for a lot of effort that gets wasted or just not appreciated by the consumers of the content.
It was really great reading about your ideas behind the dynamic POIs in SWG, Raph, and also Brandon’s thoughts on what went wrong. I thought that dynamic POI’s/missions in SWG were amazing, and one of the many things that made it revolutionary to an extent that no other MMORPG has even tried to attempt and probably wont for a long time.
I agree with Shayde when he says that ‘You were thinking so far out of the box, and WoW comes along and shoves everything back in the box and nails the lid shut.’ The timing really was a tragedy. I think if WoW didn’t come out so soon after SWG or if it did not become so popular that many MMORPGs would continue in the direction that SWG had started. Instead the entire industry did a 180, and SWG was left on the far edge of that turn. I really hope I get to see a day when Raph directs another big MMORPG and can achieve something to the same level as SWG did for many of us.
If you want a timeline that lead to WoW’s global domination, to my memory it seems something like this:
UO ships. We love it. We also hate it, because of the primitive graphics and unchecked bloodsport outside every guard zone.
Then EverQuest ships. It’s got no PKs and 3D graphics. UO is no longer the only game in town, and it hemmorages players to EQ. The fact that EQ is an amusement-park game with a class-based system is secondary; the winsauce is in the eye candy plus no player-killing. But most of the clones that follow EQ don’t draw that distinction; they follow all its primary design concepts slavishly in hopes of recapturing its success.
Some time later, SWG rolls along, and it’s poised to shift the paradigm again with the best of both worlds – the loose, mostly ‘sandbox’ style of UO, a non-class-based advancement system, yummy eye candy and ganking confined to a faction system. But it ships unfinished and never recovers its momentum.
By the time WoW rolls around, they don’t even attempt to shift any paradigms. They essentially lift every concept from EverQuest and its offspring, polish it to a high gleam, and they finish the product before they release it. It ships clean, brightly-colored, easy to learn, and it holds your hand from cradle to grave.
The irony is that players outgrowing WoW are fueling a resurgance of interest in many of the core design concepts of UO and SWG. I see recognizable bits and pieces in many of the new titles hitting the market. There’s a definite demand for the deep game, and I find that a heartening counter-trend to Farmvilification.
It turns out that it is easy to create a complex experience. It is very hard to create an adaptive one that is also resilient given the amount of memory that has to be semantically evaluated dynamically given the number of feeds shaping it.
There is a significant interest in such topics in the public safety (say NRF) industry where scenarios fed by live or recorded data are emerging from the mapping systems almost bottom up.
http://www.vanbreda.org/adrian/resilience.htm
A trick is to extract that from the feeds using appropriate framing metadata being supplied by the system responsible for coordinating and ordering the feeds. The problem generically is such metadata frames tend to be domain-specific. So the inspiration has to be in the simplifying assumptions one can make about the metadata classes that enables one to group the dynamic resources sensibly for real-time dispatch/allocation given an event/incident type.
Where is the facebook option to ‘dislike’ or ‘hate’ your posts? I hate not having options…
Just some loose comments.
SWG seemed to me to have a pretty good level grind to it, much like EQ. (I could be wrong since I didn’t stay long after seeing the first few “levels” advance so strongly.) It seems to me then, that players for the most part are going to level grind away, and anything else takes a back seat to that need to reach for more and more power. I don’t know if this actually happened in the community at large, but it’s always been a suspicion of mine based on the other games like EQ and WoW.
A thing that bothers me, being a depth and sandbox lover, is that too often games are made thinking “game” instead of “world”. This case is a classic example, I think. The POIs were made thinking that it’s important to supply game to players, and thus centering the activity around where the players currently are. You see the problem, all the activity from this system was around the central hubs of players. But more importantly, the players didn’t have to move out of that central hub. Land locked, or better yet game locked, into not moving out to explore and spread out. I think it would have been far better to think in terms of the game world, instead of the players. Let the players go out and discover the world, and I think it would be a lot more interesting than simply having content gamed for them all the time.
And another comment that I forgot to add.
Making content that games to the players seems very difficult, using metrics and all. Does it have to be that complicated? Again, if the thinking is “world”, as opposed to “game”, it seems to me that things would be much simpler. Thus avoiding this sort of complication.
Yeah, but the roleplayer/roll-player conflict will always play out in favour of the roll-player group when business-decisions are being made. Business does not equal art. Art makes a more fulfilling experience, but also a more demanding one. Dynamic POI is kinda trying to satisfy the roll-players within an artistic world vision. Making that work as a whole is not trivial (but probably possible).
I agree; it’s going to be difficult, labor-intensive, and rife with fail potential to make dynamic POI pan out. The only counterweight to that is if it pegs the Cool Meter so hard that the bean counters see dollar signs.
I haven’t checked out Rift yet. I had associated it with the old Palladium game Rifts, which left me with a bad taste in my mouth for reasons I don’t entirely remember. But there doesn’t seem to be any association between the two.
At any rate, if they knock the ball out of the park with their implementation, it may open things up for other developers who want to explore the concept.
Roleplayers vs. roll-players is based on two false premises, in my opinion.
1) Yukon said it
2) The vast majority of MMO gamers are new to the massive, having played mostly single player games. Single player games are best made in the roll-player theme. Even so, there’s always been a lot of complaints that might lead one to see players expectations for “worldly”. Look at the complaints for Balder’s Gate because of the forced paths. Look at the excitement of the physics of Morrowind. Players do want “worldly”. But the whole thing was directed to roll-player by circumstance.
Oops, actually I meant that the false part was that most gamers want roll-play, as opposed to roleplay in a “worldly” environment. The “false” part of those premises is in the relationship between the two kinds of gamers. Flip the majority from the current view that roll-players are the predominant sort.
Well, in this context roll-play was meant to refer to game mechanics dominating the world experience. The world taking the backseat as a backdrop rather than as The World. Creating a World is a much more risky business proposition than creating a Game.
Ola, I think we’re saying the same thing as far as “roll-play”. I specified the cause, maybe? That being the level grind system pushing players into that style by the natural tendency to want to grow.
Does that mean growth in powers and skills only? That’s all that’s offered in the level grinds. Other than points and collections, which are meaningless to large segments of gamers. At least, as is.
The risk seems to me to be shifted. Who can compete with WoW? It’s been done, and done exceptionally well. WoW also now has years of continued polish behind them. And it’s also become boring to experienced gamers who’ve been there.
A worldly sandbox is where it’s all going anyways. Look at this topic. This feature is just one of many that tend towards that end. And all the talk of “what’s next” is also. But the tendency of game producers (in general) to want to make it all like previous kinds of games is holding it all back. That’s why we need the visionaries. people who can think in “worldly” terms to free up the gears that are currently grinding slowly.
On āworldinessā — Iām one of the folks who naturally prefers that kind of thing over being led by the nose. OTOH, I donāt mind a good story now and then as long as itās brief and a reward for effective exploration. So Iām all for both static stories and dynamic POIs.
But I think a lot of people talk about āworldinessā as an inherently Good Thing compared to something like WoW, and itās not that simple. What are people really thinking of when they say they want a more worldy game than the clones currently available?
In other words, how would the typical player experience in a very worldy game differ from the typical experience of playing one of todayās MMORPGs with the usual set of buff/aggro/cooldown/class/level conventions? Could any one developerās vision of a very worldy game be generally appealing for the majority of us who ache for such a game?
Personally, I wish some publisher/developer would just plow ahead and make such a game. It would at least be the proof-of-concept that there really is a market for such games. Supporters of this approach might do well to look for a developer who is known for making relatively worldy single-player games and encouraging them to bring that expertise to a MMORPG.
I hear that Zenimax Online, sibling company to Bethesda Softworks, has licensed the HeroEngine MMORPG builder…. š
Finally, the saga of SWG deserves not just a post-mortem but an entire book in the mold of _Hackers_ or _The Soul of a New Machine_. Preferably it would be written by someone whoās not only a good storyteller but who also understands something of the art and business of computer game development, and who can persuade the major and minor players to speak freely. That book is crying out to be written.
Personally, I wish some publisher/developer would just plow ahead and make such a game. It would at least be the proof-of-concept that there really is a market for such games. Supporters of this approach might do well to look for a developer who is known for making relatively worldy single-player games and encouraging them to bring that expertise to a MMORPG.
I just realized that I never did get around to building that collaborative design site I had wanted to do for worldy MMOs. And I don’t quite remember the details of my plans for that.
Would anyone be interested in participating in such a beast?
Well, some MUDs have more of that worldlieness (with scars and warts…) to them that any commercial 3D world IĀ“ve seen or read about. The MMOs get the worldlieness wrong because they start with an image of RPG, the social 3D worlds (active world/SL etc) get it wrong because they start with a 3D editor. A True World has to grow as an eco-system, not as a game or as a tool… I could go on ranting about this… forever. š
I think this is critical. All other things need to be based off this and interact with it. There does need to be all the other stuff too. Combat, magic or science (or both), monsters to fit, exploration (and discovery needs a big boost), mystery and lore, all that stuff.
Of course, there can be good sandboxes and there can be bad ones. The more in the game, and the better the design concept and coding, the better it will be. I’m assuming that the game we’re talking about is a top notch effort.
I think the very first thing you need to do is design a game world that’s highly interactive for the player. Being able to reach out into the game world and make things do things is very important. Even for theme park MMOs, and it’s very much missing. WoW has some in, but limited to specific circumstances (quests and skills). I’d love to have to lower a drawbridge every time I want to cross that raging river. Opening and closing things, picking them up and dropping them, moving things, pulling things, pushing things, there needs to be a worldly sense to the entire game world.
Within this interaction you can place loads of game content. Mysteries, historical meaning, events, clues, monster reactions, player field defenses and camps, let your imaginations run wild, let your code monkeys out of their cages.
For example, UO had the mysterious numbers 3-4-1-5 in a dungeon. Evidently broken, no one that I know of knew what this meant. Suppose in your game you had a statue in this dungeon, and if you stood on one side you could *push* it and it would slide over revealing a floor safe (pushing from any other angle does nothing). The safe has 5 buttons, so the player can push them in that sequence, it opens and the player finds a note starting a small quest.
Now you have *push* in your game, and players can use this to move tables, push wagons stuck on a muddy road, even push a player out of the way when he’s blocking a doorway.
Of course, such a game offers players plenty of opportunity to be a jerk to other players. *Pushing* a player into a dangerous position probably came to mind. That’s why I’ve always believed in a justice system that’s fluid enough to allow for some retaliation before going into criminal status. And criminal status needs to really offer the possibility of real punishment. Games seem to want to protect players who commit “crimes” within their worlds, and I never understood that. Those players are not only committing virtual crimes in a game, they are affecting the game play of other players. Even driving them away from the game. They need a good spanking if that’s what they want to do. š
Yet, in my opinion, you don’t want to remove the possibility of crimes. That adds to the worldliness.
Michael, Iād be interested in that conversation.
And Ola, Iāve made some of the same points in the “Living World” design concept Iāve been working on. (http://flatfingers-theory.blogspot.com/search/label/Living%20World) The basic principles are: 1) such a thing needs to be understood, designed, and implemented from Day One as a coherent system of deeply-integrated systems, and 2) a sufficiently interesting world-environment naturally generates opportunities for different kinds of enjoyable gameplay.
Thatās definitely a different approach to game creation than whiteboarding a list of rules-based gameplay mechanics, then building just enough world to enable those mechanics. Itās got a different set of risks and potential rewards than a mechanics-first, world-second, focused-game experience. And I suspect there are gotchas that someone whoās actually tried to build such a worldy game could point out that just havenāt occurred to theorists like me.
Even so, the vision of a deeply-realized computer-mediated place where fun gameplay can happen remains incredibly alluring. The strongly positive response to efforts along those lines by pioneers like Raph confirms to me that there really is a market for truly worldy games.
And then Winch Gate went and made all the code and art assets of the MMORPG Ryzom open source: http://dev.ryzom.com/news/13/
This doesn’t include world/server data, which stays private to Ryzom. But the client and server code (which apparently includes some ecological processes) will be available for inspection and modification, assuming you’re OK with the terms of GNU’s General Public License.
Shazam.
Just pointing out, you can have the exact same fun elements of the current games inside a sandbox, if you design it that way.
Hack and slash.
Level dings.
Travel area to another (forget zoning and direction, just have it with freedom.)
Umm….that it?
I’m not sure how much you’ve been paying attention, but what Guild Wars 2 seems to be doing looks like something of a compromise measure that solves many of the issues that SWG had.
There are a *lot* of world events, but I’m not sure you can legitimately call them dynamic POIs, at least not as they would’ve been considered in SWG, since they’re bound to specific locations. They’re more or less static POIs bound to dynamic triggers, some of which are deterministic (killing all the deer in the forest can trigger something, as can succeeding or failing at a previous event) some of which are time based (only occur at night or day in the game world) and some of which are purely random. The big thing is that triggers are bound to the world, not to a specific character. This gives the illusion of a living world, since you’re never quite sure what you’ll see, even though you can only ever see things in specific areas. This provides some decent benefits as a solution: it avoids the issue of seeing the same content over and over to a greater degree, there’s more of an ability to control content progression since it’s bound to zones not characters, and it provides a method of continuity which makes it feel like you’re actually having an impact on the world. It does mean that some of your content may never be seen, and you still need a ridiculous amount of it if you’re going to populate the entire world, but since they seem to have removed all of the traditional busywork quests to focus on stuff like this, leaving the directed experience to the story missions, it’s no more of a challenge than building a WoW like number of quests; expensive but doable.
There are a few additional benefits to that approach, beyond making the world feel more real, especially in the context of the very loose kill rights and dynamic content scaling that they’re also implementing. World events are more social than quests since anyone can participate, it encourages exploration in a way that quests do not – if you’re on a quest you rapidly develop tunnel vision because that’s the most time efficient way to progress – and because the mechanism for events is trigger based, it’s relatively easy to add or adjust existing content without creating a sea of question marks or being concerned about breaking content flow. You can even adjust things based on what players seem to be more interested in, and it’ll organically flow into everything that’s come before. If the players are consistently beating an event, you can swap it out for something else, add a new event that ties the old content to the new, and move on with things. (As a side note, one really interesting adaptation of a system like this involves creating a portion of the live team whose sole reason to be is tracking event triggers and adding new things based on outcomes. You’d run it through a bug tracker like system. With some refinement and enough resources it’d be about as close to a pen and paper GM running things as you can get when you’re dealing with huge populations)
There are some downsides of course; not everyone is going to see everything you’ve done, and you still have a very high content requirement. But the former issue isn’t so much a problem if you start looking at content differently; decouple the events from progression and all of a sudden content isn’t something you do to keep individual players entertained to the end game, but rather something to keep your entire player base from running out of things to do. Once you look at it like that, the goal is to make sure that at any given moment in every given area, there’s something going on for all of the players logged in. You need a high enough content threshold to pull this off, but once you reach it, it’s relatively sustaining. Content will rotate out due to not all of your triggers going off, making this actually more efficient than static quests that are consumed once. Any given player will be unlikely to experience the same events play session to play session, even if the aggregate content is lower, and because the triggers are out of the player’s hand, there will be surprises on a regular basis.
This actually brings us to the biggest advantage of a system like this: as long as the content generation pipeline is efficient, the random nature of availability for consumption allows you to roughly keep up with the needs of the playerbase. You can add events at a rate that allow for there to be new things that a given player hasn’t seen yet on a semi-regular basis, no matter how quickly they try to consume content. You still need other things of course, you still need something that ties progression together, you need reasons to move through the zones so that players are bumping into things, you need other activities that provide some social glue, etc etc. But as a replacement to kill 10 boar and fedex quest filler, it’s certainly an interesting way to go.
@Bart, having an annoyingly hard time finding anything resembling an email address of yours online.
1) I have no idea how I never ran across your Living World RPG thing before. It’s very much what I’d be interested in.
2) It doesn’t look like I’m going to have the time I’d hoped for to build the design site I had thought about, so maybe we should just start up a forum or a wikia or something? (My idea for the site would have been a blend; Google Wave could approximate it.)
3) My email is not [email protected] (well, it is, but that was a long time ago; now it’s my spam funnel), but I’ll keep an eye out for your email and reply with my Gmail. š
Although it’s nice to know I’m running under the radar where spam is concerned, it does make collaboration a little more difficult.
My apologies for the trouble you had, Michael; I’ll contact you via your spam funnel in the next few minutes.
[…] system, a varied PvE and PvP environment, a sandbox approach where you just meandered around, dynamic POI content and more, but the thing that really got me was the hybrid skill system. If you were unlucky enough […]
[…] Star Wars Galaxies, wrote an interesting post on his blog at the end of April. In the post, “Dynamic POIs“, He discusses how, in Star Wars Galaxies, they constructed a method for the computer to […]
[…] its inner workings a surprise – if you want to read some of what influenced me, check out Raph Koster’s brilliant SWG design – but apply it to a randomly generated […]
[…] […]