« Rent-A-Vatar | Main | Casual Dress »

Jun 11, 2005

Comments

1.

Um...isn't that what we would normally call a "single player game" ?

Sounds to me like that's what you're describing...

2.

First off, speedy recovery to you!

Diablo launched in Jan '97 and it was an MMO: It had persistent chat features and a partially persistent economy (the junk sold to the vendors with bottomless pockets wasn't, but the trade in high-end goods was). All of the content was instanced.

Compare this with the current MMOs like WoW: The chat channels are persistent, the economy for high-end items is persistent (the rest of the junk is still sold to vendors with bottomless pockets), and much of the content is instanced. The only substantial difference is that you are in a graphical chatroom and have to walk to the instances in WoW whereas the chatroom was textual and you only needed to click "join" to get to the instance in Diablo.

Yet most people claim that Diablo isn't an MMO and those same people are quick to call WoW and MMO (Sir Bruce i'm looking your way!). Between the launch of those two products there were many MMOs with lots of persistent content, but the trend seems to be moving back toward heavy use of instanced content -- WoW, GW, COH anyone?

The general attitudes from players seems all for this trend. I've seen more than 1 thread on the WoW forums asking for instanced auction houses. Instancing does help players feel more like the hereo, and does resolve some issues about griefing, kill stealing, loot stealing, etc. IMHO the cure is worse than the disease.

Dan do you have anymore of those pills?

~don

3.

Well, ow, Professor. Hope you're all right.

To the point, I think there are a fair number of people who would enjoy it if everybody else were NPC's, but I think that would eventually get boring...people's likes and dislikes rarely seem to me to be that static. Even a fully adaptable system would seem stale and flat after a while.

4.

Ouchers. Hope you get well soon-ish! My wife can attest to the wonders of Percocet (A-list drug for C-Sections), and the unfortunateness of being allergic to it...

Playing alone is an interesting topic. The current games are trying to solve the social problems from the previous generation, but I agree there can be a troubling downside. In a way, it corrolates to "forced grouping". If a game provides no reason to group, then a goodly amount of players will not. If a game provides no reason to be social, then the result is similar.

I think the dichotomy is in the game play though. There was no reason for persistence in an EQ-like game. Players will fight over limited resources, and they lacked any real means to do so (the basic justification for Shadowbane). Instancing solves this particular problem: removing other players from getting in the way of individual/small-ish raid-group fun.

Conversely, SWG (and Eve, ATITD, et al) is built upon the entire concept of "forced" interaction. Players can play a Diku-like experience, but everything they buy and use is from another player at some point in the chain.

What players want is that gray area in between still being explored :)

5.

Dan--I was sorry to hear you won't make it to Supernova--Kevin Werbach let me know on Friday... Hope you recover soon!

On instancing and definitional things.

Diablo is not a MUD by my definition because its shared core does not use a spatial metaphor. To me, a spatial metaphor is a core element of a virtual world.

WoW does use a spatial metaphor for its core world. The instances are then embedded games. A virtual world is fundamentally a spatial environment that you can wander around in. The presence of games is irrelevant, definitionally speaking.

This may seem like an arbitrary or academic distinction, but for me it is the dividing line that keeps out things like IRC or web-based games as well.

As far as heavy use of instancing--at the extreme, it can remove the given software from the realm of "virtual world" altogether, as in Diablo. When all the action, so to speak, is in the instances, then the feature-richness of the core can be reduced until you don't have much of a core--perhaps only a textual chat space.

It's a useful tool, but many of the answers it provides are essentially cop-outs. Let's look at the typical list of reasons to use instances.

1. in order to provide a single-player experience.

2. in order to provide a limited multiplayer experience.

3. to prevent interaction from by other players.

4. to provide an embedded game experience that cannot be provided using the standard game metaphor.

5. to permit the illusion of massive changes caused by players, that do not actually disrupt other players

#4 is really the only GOOD reason. The others are varying degrees of surrender to the problems presented by massively multiplayer environments. Looking at them briefly:

- you might want to provide a single-player or limited multiplayer experience because you have a particular story to tell. However, methods exist and have been used for quite some time to permit that in a multiplayer environment.

- the argument that you need to put world-affecting changes in an instance for social reasons doesn't wash either given the genre's history. Instead, it's usually for technical reasons. You want to blow up a bridge at the end of the instance for a dramatic reason, but leave the bridge intact for everyone else. Well, there ARE other ways to achieve that, they're just more work. One of the greatest potentials for the medium is shared worlds where players have an impact; reducing all the impacts to illusions is not the direction towards realizing that promise.

- copping out on solving the issues of interaction and interference is just going to lead to a retreat from the potential of massively multiplayer environments in general. Just to be clear: the reason to be massively multiplayer is SO THAT PLAYERS CAN INTERACT. Pursuing design choices to prevent it as much as possible is essentially a retreat from the genre.

This doesn't mean that there aren't many many legit uses of instancing. One of the most common is to reduce load of content; sharding is instancing, after all. Embedded games have huge potential that has barely been tapped, but their greatest potential comes in embedded games that are DIFFERENT from the main event. WoW's battlegrounds are good example here.

-Raph

6.

Hmm. I’m moved to post because I’m deja-vuing a lot of Terra Nova these days and because I too have wondered what is the deal with this Cameron Diaz thing anyway?

Here’s the deja-vu part, from TN of a year ago:

"In the future, everyone will have their own virtual world, just like everyone currently has their own imagination. Some people will borrow from other people's worlds/imaginations, just like they do now. The current implementation of personal and instanced missions/dungeons will expand as memory/bandwidth/and such expand.

"In the future, you will be able to choose exactly to what degree real-life others interfere/interact/inter-whatever with your virtual world. If you want to interact with griefers, you can -- for thirty minutes or so -- then you can turn that part off. If you want to do the hero's quest, you can, for days or weeks or months, or you can just fast-forward through the whole thing. If you want the x-rated content, almost certainly you'll be able to get that -- for a price; almost certainly for a price, but there you go.

"Connections will be cheap, memory will be cheap, relationships will be cheap, everything will be cheap but the profits and the code, which (unless the revolution comes or the meteor hits) will remain as proprietary as all hell, and an extraordinary daily churn will not cost anyone anything -- if indeed churn even still exists after the great vw consolidation into something that looks like ma bell."

***

And, meanwhile, here is what I think is a little designer hubris:

"One of the greatest potentials for the medium is shared worlds where players have an impact; reducing all the impacts to illusions is not the direction towards realizing that promise… Just to be clear: the reason to be massively multiplayer is SO THAT PLAYERS CAN INTERACT." – Raph (above)

I rather think that those “shared worlds where players have an impact” are fairly illusionary to begin with. And I would think that players in general much prefer their own illusions over the illusions of others, up to and including the illusions of well-intentioned game designers.

But all this was discussed earlier in the AI thread: The illusion of interaction (or AI) is what you need, not the real thing. And, in fact, in almost all contexts of play, the illusion of interaction is preferred over the real thing. Interactions with living and breathing others, after all, more commonly disrupt than reinforce play.

So, if anyone wants to create a shared world where players have an impact, I suggest they think of it as constructing the illusion of a shared world where players have an (equally illusionary) impact. MMO designers are, after all, most fundamentally creating (I would say REcreating, but that’s another issue) human experiences, not shared worlds. And if human experiences are -- at some basic level of perception and awareness -- individual experiences (i. e., cognitive), then the construction of those experiences would seem validly and, in fact, preferably instanced.

7.

And if human experiences are -- at some basic level of perception and awareness -- individual experiences (i. e., cognitive), then the construction of those experiences would seem validly and, in fact, preferably instanced.

When we can create AI that is as engaging, unpredictable, deep, emotional, frustrating, lovable, variable, and valuable in social situations as are actual people, then this might have some validity.

But -- and I say this as one hip-deep in advanced AI and game work -- that day isn't any time soon.

8.

Thanks to everyone for the well-wishes. Everything looks fine for the long term, may the gods and Perkies be willing.

I don't have much to contribute in my current haze, but I was particularly drawn to Raph's and Dave Myer's comments as better-explaining some of my contradictory thoughts on this. There are strong game design reasons for instancing (and for better AI interaction, etc) but it seems to me perhaps to be a step towards a solipsistic end that we may not quite anticipate. Everyone dancing in their own little instanced world, while representations of others watch and applaud every move we make. Who wouldn't want to be the celebrity in their own little world?

I don't want to sound like Chicken Little--or even Cass Sunstein's strangled imitation of Chicken Little--but the move towards instancing seems indicative of some sort of trend towards isolation. Maybe we are actually Greta Garbos after all: it's just that we want to be alone in a cognitively-compelling fiction of togetherness.

9.

Raph:
>Diablo is not a MUD by my definition because its shared core does not use a spatial metaphor. To me, a spatial metaphor is a core element of a virtual world.

The chatchannels in Diablo's lobby were distinct places. Players could only be in a single place at a time. They had to "move" from one place to another. Sure their movement wasn't from forward/backward or WASD, but through the channel join feature. Perhaps i'm confused but at what point does a virtual environment become a "space"? Does it require a description akin to that in a MUD? The desriptions are brief in Diablo, but they were present ("Adventurers Wanted" for example). Or do they need virtual coordinates?

What about a world that is 100% instanced? Sure enough it might meet your spacial requirement, but is it really a PSW (persistent state world) if nothing remains when the players log off? Where does this put Guild Wars? The cities are sharded and all of the content is instanced. Players Meet in graphical areas. Players can chat with others and trade. These areas are very close in function of the chat channels in Diablo where players typically chatted, brokered trades, and found willing adventurers. The cities in GuildWars provide essentially the same functionality -- the main difference being a graphical one. Are you stating that GW makes the cutoff because the cities are spaces but the chat areas in diablo aren't?

The all of the exploring, gaining xp, and fighting mobs occurs outside the cities. Everything outside of the cities is instanced. This is analogous to Diablo. In GuildWars you walk your avatar in a graphical context from the city to the instance portal. In diablo you clicked the create or join buttons to enter an instance.

Perhaps we are splitting hairs here, but Diablo and GW are nearly the same from an instancing standpoint, yet most everyone rejects Diablo's MMO status. I have heard some rumblings over GW's inclusion in the MMO world, but a great number more accept GW in the group. I have a hard time accepting one while denying the other. WoW isn't substantially further behind GW with its heavy use of instancing.

With the addition of BGs (battlegrounds) in WoW, the idea of a persistent world is even further blurred. The BGs are all instanced. In addition when a player gets in line to join an instance, he/she will be instantly teleported from their current location directly to the BG. This yet erodes the importance of all the persistent space inbetween. True enough not all content in WoW is instanced, but when a game rises to a given percentage (such as GW with 100% instancing) at what point does it stop being an virtual world?

Diablo: 100% instanced, textual chat channels without coordinates, not an MMO?
GW: 100% instanced, graphical chat channels with coordinates, maybe/maybe not an MMO?
WoW: <100% instanced, graphical chat channels with coordinates, MMO for sure?

Again i have a hard time accepting that Diablo fails the test but GW meets it because the chat channels in Diablo lack map coordinates.

Raph:
>copping out on solving the issues of interaction and interference is just going to lead to a retreat from the potential of massively multiplayer environments in general. Just to be clear: the reason to be massively multiplayer is SO THAT PLAYERS CAN INTERACT. Pursuing design choices to prevent it as much as possible is essentially a retreat from the genre.

/agreed

Raph:
>This doesn't mean that there aren't many many legit uses of instancing. One of the most common is to reduce load of content; sharding is instancing, after all. Embedded games have huge potential that has barely been tapped, but their greatest potential comes in embedded games that are DIFFERENT from the main event. WoW's battlegrounds are good example here.

/agreed, but not all MMOs use shards. When I left EVE Online last year it made no use of shards or instancing. True enough max concurrent load is 12k users, and it has "only" about 60k subscribers. The use of shards is similar to the use of instancing: It's the easy way out. Indeed "mini games" have huge potential in VWs, but that doesn't mean they need to be instanced. If your MMO turns into a collection of instanced mini-games, then how is that different than connecting to a website with a list of games to play? Microsofts Game Zone for example...

I'm trying to chip away at your definition based on the spacial-metaphor requirement and replace it with one of a persistent-world argument. A space without persistence is nothing more than playing UT2003 online. Again most of Diablo is isntanced, but the chat channels and the high-end items in the economy are persistent. Even if everyone logs off the battlenet the economy for high-end items remains intact. I content that it is the persistence -- the economy and the chat channels -- that makes it a VW. That isn't to say it makes it much of one, but it does help make a clear line in the sand between VWs and other online games.

~don

10.

I rather think that those “shared worlds where players have an impact” are fairly illusionary to begin with. And I would think that players in general much prefer their own illusions over the illusions of others, up to and including the illusions of well-intentioned game designers.

Right now, they largely are. The number of true persistent state worlds is very small at the moment, and few want to tackle them because of the huge design and database efforts involved. Instead, the "extended character state" model seems to be the most prevalent these days, essentially an evolution out of the "character state" model but encompassing some elements that give the illusion of persisting world state.

That said, the advantages of true world state seem clear in the long run. For example, witness the gradual rejection of character state models as regards player shopkeepers. Players want this function of gameplay to work in a world state model, and they are not shy about saying so.

The illusion of interaction (or AI) is what you need, not the real thing. And, in fact, in almost all contexts of play, the illusion of interaction is preferred over the real thing. Interactions with living and breathing others, after all, more commonly disrupt than reinforce play.

That would, I think, be a bit of a surprise to everyone who plays real life sports or games.

I don't mean to minimize the hassles involved in interacting with others. I have a lengthy set of lofty reasons why interaction with others is critical to our health as a species and as a culture, but that's all beside the point. Suffice it to say that if you're after the illusion, that is what single-player games offer already. If you are working in massively multiplayer, you are doing so because the medium is about the interaction. It is definitional. Remove the other people, and it's not massively multiplayer anymore. I find it hard to believe that anyone would contest that.

I DO frequently see that players want a world where they can play in limited multiplayer fashion but OBSERVE others going around from time to time. It's still a multiplayer environment, however.

So, if anyone wants to create a shared world where players have an impact, I suggest they think of it as constructing the illusion of a shared world where players have an (equally illusionary) impact.

Such as, say, Animal Crossing? Actually, there was a REAL impact there, and merely a time-shifted world.

This could be accomplished by simply having random crap change in the (single-player) world, and subsequently claiming that the activity was due to the actions of other players.

MMO designers are, after all, most fundamentally creating (I would say REcreating, but that’s another issue) human experiences, not shared worlds.

PLAYERS create human experiences. Game designers (and artists in general) merely create the conditions under which they hope those experiences will arise. Every creator has had the experience of their carefully crafted creation providing a completely different experience to the viewer/user/reader than they intended.

In those media where you are most trying to impose an experience upon the user--what I call impositional forms of narrative--I agree that instances is the best tool. Games, however, are by nature less apt for impositional narratives, compared to media which offer the reader/viewer/user less agency. In other words, games are expressive media, and lend themselves to expressive narratives. Multiplayer games are more expressive than most.

And if human experiences are -- at some basic level of perception and awareness -- individual experiences (i. e., cognitive), then the construction of those experiences would seem validly and, in fact, preferably instanced.

There is quite a lot of scientific evidence to suggest that the cognitive experience changes significantly based on the presence or absence of other individuals.

There are strong game design reasons for instancing (and for better AI interaction, etc) but it seems to me perhaps to be a step towards a solipsistic end that we may not quite anticipate. Everyone dancing in their own little instanced world, while representations of others watch and applaud every move we make. Who wouldn't want to be the celebrity in their own little world?

I don't want to sound like Chicken Little--or even Cass Sunstein's strangled imitation of Chicken Little--but the move towards instancing seems indicative of some sort of trend towards isolation.

The trend is in the opposite direction, IMHO. The existence of truly isolate single-player games is a historical aberration resultant from the presence of computers without networks. Ever since networks came along, videogames have gradually introduced more and more elements of the network back into them.

It did result in a massive rise in one specific type of game; what you might call the interactive entertainment experience, which is largely a story with bits of gameplay in between the plot points. Those who are wedded to games as storytelling media are the ones who are most likely to espouse instancing as a solution, because they are trying to practice that particular craft.

The chatchannels in Diablo's lobby were distinct places. Players could only be in a single place at a time. They had to "move" from one place to another. Sure their movement wasn't from forward/backward or WASD, but through the channel join feature. Perhaps i'm confused but at what point does a virtual environment become a "space"? Does it require a description akin to that in a MUD? The desriptions are brief in Diablo, but they were present ("Adventurers Wanted" for example). Or do they need virtual coordinates?

To me, all that is needed is to employ a spatial metaphor. In other words, yes, if Diablo's chat channels were named as places and not as chat channels, and were described as places, and were assumed to exist in some sort of space, then yeah, people would tend to call it an MMO.

I realize that sounds like an arbitrary and even irrelevant distinction, but that has been a distinguishing factor every time I've really pushed hard on myself to ask why a given game "feels" like an online world or not.

What about a world that is 100% instanced? Sure enough it might meet your spacial requirement, but is it really a PSW (persistent state world) if nothing remains when the players log off?

My full definition includes persistence.

1. A spatial representation of the virtual world
2. Avatar representation within the space
3. A sandbox to play in that offers persistence for some amount of the data represented within the virtual world

But my full definition is many pages long. I refer you to the book I never finished because Richard beat me to it:

Insubstantial Pageants

Where does this put Guild Wars? The cities are sharded and all of the content is instanced. Players Meet in graphical areas. Players can chat with others and trade. These areas are very close in function of the chat channels in Diablo where players typically chatted, brokered trades, and found willing adventurers. The cities in GuildWars provide essentially the same functionality -- the main difference being a graphical one. Are you stating that GW makes the cutoff because the cities are spaces but the chat areas in diablo aren't?

Yep.

WoW isn't substantially further behind GW with its heavy use of instancing.

I beg to differ. But I suspect that your view may be skewed by the type of gameplay that you are doing in WoW. Log in as a newbie, and the distinction should be clear. You don't even SEE an instance for many hours worth of gameplay.

In addition when a player gets in line to join an instance, he/she will be instantly teleported from their current location directly to the BG. This yet erodes the importance of all the persistent space inbetween.

Teleportation has been a feature of virtual spaces since the beginning. That's not an issue. The issue is whether the spatial metaphor remains intact.

not all MMOs use shards. When I left EVE Online last year it made no use of shards or instancing. True enough max concurrent load is 12k users, and it has "only" about 60k subscribers. The use of shards is similar to the use of instancing: It's the easy way out.

I think I have made the same point; the reason why most games use sharding is purely from a lack of content, not from a lack of technical capability.

Indeed "mini games" have huge potential in VWs, but that doesn't mean they need to be instanced. If your MMO turns into a collection of instanced mini-games, then how is that different than connecting to a website with a list of games to play? Microsofts Game Zone for example...

I think anyone who visits Second Life (which is well on the way towards being a collection of embedded mini-games) will agree that not only is instancing not necessary for embedded games, but that it's pretty different from web-based game collections. Again, the spatial metaphor is pretty compelling.

I'm trying to chip away at your definition based on the spacial-metaphor requirement and replace it with one of a persistent-world argument. A space without persistence is nothing more than playing UT2003 online. Again most of Diablo is isntanced, but the chat channels and the high-end items in the economy are persistent. Even if everyone logs off the battlenet the economy for high-end items remains intact. I content that it is the persistence -- the economy and the chat channels -- that makes it a VW. That isn't to say it makes it much of one, but it does help make a clear line in the sand between VWs and other online games.

By that definition, Amazon.com or eBay are MMOs. After all, you have persistent characters, and you have persistent data for economics, for goods, and so on.

Persistent databases are used all over the place, and nobody thinks that they represent virtual worlds. But if I say "add avatars and actual hallways to Amazon, but leave everything else the same" everyone will instantly agree that it's a virtual world.

There's a reason why people always attach the word "world" to these discussions. And world implies spatiality. It may rub you wrong, and I haven't really examined WHY it is so, but scratch just about anybody deeply enough and you'll find that this is inherent in their internal unexamined definition.

It's worth pointing out that the thought experiment of a game with no chat channels of any sort, no avatar persistence, and no economy, but with a spatial metaphor, would likely still meet everyone's internal definition.

11.

Regarding the spatiality of chat channels as opposed to Guild Wars-style restricted shard city-areas - I'd say personally that they were both spaces, but different kinds.

I think of it using a MUD-style metaphor - in a Diablo-style chat channel, it can be conceptualized as a room, but it's a small one, where everything happens in the same place. Effectively, it's a dimensionless point, or at least has only a time dimension. (If it allows subdivisions, such as private messages and clan channels, then it's got a little bit more dimensionality, but not a great deal.)

But in Ascalon (the start city in Guild Wars, effectively) it's more like an area - you can see space, you can move around into different areas (not delineated ones, but they have boundaries just the same as 'this part of the room' is distinct from 'over there' in a real-world room) and when you move around people and things you couldn't see before are revealed to you.

You also get a lot more information about the character of the space and the people in it, visual and aural clues, and (speaking partly as a MUD builder) one cheap and easy way to make a room look 'real' is to make sure it has walls and a floor and ceiling (or not, as appropriate).

Of course, the other way is to look around at the other people and see how they treat it - Raph's point above, about psychological experience being affected by other people, covers this one. If I look around in Ascalon and see people standing in groups, walking from NPC to NPC, dancing or sitting near a fire, then that looks very place-like to me. But a bare chat interface doesn't support that sort of complexity and information density, there's nothing to build world-dwelling behaviour on, without a lot of cooperative work. But with the right kind of community, evolving without significant opposition, it can happen and sustain itself.

One of the most 'real' online environments I've been in was a very plain HTML-based chatroom, with public messages and private messages and nothing else, but the regulars decided what was there, put a lot of effort into maintaining the shared world and its contents, and the communal memory allowed a form of persistence. ("Joe mentioned we had rafters last week, I'll go sit on them. Oh, what's this I've found up here, it must belong to Susan.") Partly, that was because at the top of the page it said, very specifically, 'Enter the Underground Corridor', so that sets up the initial conditions - but someone long ago decided there were windows, so everyone remembered and referred to them. So the initial conditions weren't all that constraining.

A brief thought, also, on single-player/instanced/massively populated environments - it looks like it changes from "it's all about me" through "it's about us" to "it's about itself". I do see a parallel between Richard Bartle's perennial theme of the Hero's Journey, moving from your comfortable environment into a larger space with more degrees of freedom and larger tasks to do, where you can (potentially) have a serious and persistent effect on the virtual-physical, economic, or social world itself - but it's interesting that it seems to work in the opposite direction in so many cases, trending from the start in the big world to a largely instanced end-game.

12.

Dan>I don't want to sound like Chicken Little--or even Cass Sunstein's strangled imitation of Chicken Little--but the move towards instancing seems indicative of some sort of trend towards isolation. Maybe we are actually Greta Garbos after all: it's just that we want to be alone in a cognitively-compelling fiction of togetherness.

Hi Dan - I hope you're feeling better soon and enjoying your "augmented reality" in the meantime. I had the benefit of those pills a few years ago when I had kidney stones. Amazing! I'm sure you are heeding the usual caveats in the package about heavy machinery operation and making Important Business Decisions! Maybe MMOG's should have the same warning? :)

The instanced world seems to be heading past Sunstein's notion of the "Daily Me" with the "perfect filtering" problems you previously noted, and into the world of mimetics where intelligent agents will soon tailor their on-line advertisements, pitches, editorials, reviews, and solicitations not only to the interests of the viewer, but also to the viewer's involuntary gestures (facial muscles, eye/pupil movements, body language etc.) so that not only is the collection of materials received individualized, but also the content of each piece within that collection. Therefore, even if something gets through the "perfect filter", the content will still be instanced.

Peter


13.

To me the whole "is it a MMOG" can be brought down to a simple experiment...

remove ALL the instanced things... completely - remove from code and cant ever access them...

is there any game/world left? is it persistent?

If yes - you got an MMO.

Eve-online has no instancing - its an MMO for sure.

WoW gets a bit less exciting, but sure as heck its still a functioning game - its an MMO.

Guild Wars - all you are left with is the towns. There would be no way to advance levels, get new gear etc... not really much of a game left, only chat channels... To me this falls (just) on the "not MMO" side of the fence.

Diablo - you have IRC left, effectively, once all the instancing is gone - definently not an MMO.

14.

We all come to the table with our own definition and filter of what a world is and should be.

Regardless, we can still have the isolation of individualized experience with or without instancing. Peter already explained how this can occur without instancing.

To elaborate further, with the advancement of customized client technology we can filter, translate and annotate our own perception of the world client-side. We can create our own Percoset-induced experiences. Server sends a “hello” and I receive a “greetings”. Server sends an NPC opponent with X ability and at Y difficulty and I receive it as a Spider with X1 ability and Y1 difficulty. We can create our own individualized illusions all at client-side.

Thus, instancing as a technology is not needed and is not the force behind isolation or a more single-player experience. We can happily have our own individualized experience within a networked world and we can create the illusion of persistence and other definition of virtual worlds/MMOs whether the server or the network is designed as such.

We can play alone.

But as Raph stated, we play online to play with other people. But we also want to do it at our own terms and within our own social comfort zone. Instancing is just a workaround for lack of content and the lack of ability to play with others but within our own terms.

Well, at least this is how I non-Percoset-ly see it.

Get well Dan. May you rest, recover and gain great insights you are at it.

Frank

15.

I'd argue that the point of sharing is not necessarily interacting, it's emotional validation. All achievement-related content could be instanced, but so long as the evidence of achievement has some effect on how we are perceived in a shared social/physical environment, the achievement receives a powerful validation from the social reputational effect. That's why mog content can be more dull and yet more motivating than a single player game. It's dull, but people will see that you've done it. So it feels more important.

I actually see a future like the one Dan describes, where each player has a huge mass of AI devoted just to him, yet where the results of his actions also register with other real people.

16.
Magicback wrote: But as Raph stated, we play online to play with other people.
Many play online for the opportunity to play with other people. The distinction there is that players (both as individuals and as groups) look to control when that opportunity occurs. When they can't, they use the vernacular that's developed over time, things like "kill-stealing", "ninja-looting", "farming {my} content", etc. Those are all situational events that result from a lack of control, the breaking down of the "game" that can no longer be mastered by the individual that got used to doing so. This, to me, is the event instancing tries to "fix", protecting the essence of the game experience players expect, the traditional CRPG.

Non-instanced games aren't driving to deliver CRPGs. They're looking to do more than that. Some refer to this as a sandbox. I consider it more of the larger playground. People are on swings, seasaws, monkey bars, all doing their own thing, sometimes with other people. Spontaneous goals arise, as does the necessary pecking order (Leaders), outside governance (GMs), and sometimes overall goals (organized mini-games), sometimes not. MMOs try to add longterm elements that compel players to keep coming back, but right now, these games are still as fleeting. Players change, along with their preferences, until they eventually move on. Even the most embedded leaders eventually move on because the parameters that allowed for their leadership have changed (sometimes because all the people changed and brought with them new leaders).

Edward Castranova wrote: ...so long as the evidence of achievement has some effect on how we are perceived in a shared social/physical environment, the achievement receives a powerful validation from the social reputational effect.

There's a wide gulf between exclusively-text-based "badges" that require specific interfaces just to see and that glowing sword/cape/helmet that everyone sees if they're in the same virtual space. Being able to brag about a reward is only part of the compulsion. Involuntary "bragging" (showing off just by donning something) has a similarly powerful cache. That duality may have been to what you're referring though :)

17.

“That said, the advantages of true world state seem clear in the long run.”

If there were a “true world” state, perhaps. What I am saying is that the “true world” state is an illusion, or best, the perception of a true world state. The desire for that state may be real, but that’s a different (and an individual) matter.

There is really no firm evidence of such a state – even in the true “true world.” Our common and shared beliefs about the true world are just that: beliefs. And it takes a passel of lawyers and judges and mass media folderol to keep those beliefs in a psuedo-stable state.

Ergo, we live within the perception of stable and persistent world. And, therefore, in order to construct a persistent world analogous to the one in which we live, we are forced to construct its perception. Perceptions (and awarenesses of perceptions) are individual and cognitive… and so on from there until you get to the instancing part.

*

“That -- [Interactions with living and breathing others, after all, more commonly disrupt than reinforce play.] -- would, I think, be a bit of a surprise to everyone who plays real life sports or games.”

Ah, but real life sports or games are different from real life PLAY.

It’s my observation that mmo designers have this notion that playing a game and playing are equivalent. My notion is that these two are quite distinct. Playing a game (or, more pointedly, abiding by the rules of a game) binds and restricts play.

Sure, interactions with others reinforce the rules of a game -- which, in an mmo context includes the game CODE. That’s precisely why there are so many guilds and guild-like “interactions” in mmos – to maintain order beyond the (current) capabilities of the code to do so.

But then, from the point of view of the guild leader/game designer, there’s all this darn PLAY stuff that disrupts my rules/code. From the point of view of the player, the same thing looks like: there’s all this darn GAME stuff that disrupts my play.

But do consider real life sports/games: These activities are very heavily instanced. There are boundaries of space and time; there are rules; there are referees to enforce those rules; there are, in general, rigid social and cultural conventions that determine what is appropriate and inappropriate play.

You need social interactions in sports and games, sure, but you need to be able to CONTROL those interactions (very similar, in some important ways, to how you need to control interactions in a narrative). Which means you don’t need (or want) a massively multiplayer environment. Not nearly enough control. What you want is a massively multiGAMEplayer environment, which, to me, implies a level of control that either doesn’t (and, perhaps, thankfully can’t) exist or exists only in some perversion of free and individual play like, say, war, or hysterical dancing [http://www.pbm.com/~lindahl/lod/vol3/dancing_mania.html], or maybe bladderball [http://www.explore-sports.com/sports/B/Bladderball.html] – all of which, despite their imaginative graphics, realistically rendered 3-D environments, and lush and detailed player-created narratives, truly suxxors.

*

”PLAYERS create human experiences. Game designers (and artists in general) merely create the conditions under which they hope those experiences will arise.”

Well, I agree totally with this one. Well, almost totally. Might actually be play, not players, that creates the human experience. But close enough. I like to say games evoke experiences – though sometimes I don’t know exactly what that means.

And Animal Crossing is a reasonable example of a sort of illusionary mmo. But, of course, having random crap happen is not sufficient to reproduce (or REcreate) the human experience. Maybe if you could pull the ears off some cute little squirrel or grief a couple of penguins – and they could do the same to you. Then you might have something.

*

The only real impact that matters, in the end, is the real impact on the individual, not the world. Thus the claim: shared worlds are superfluous; instances will suffice.

18.

Edward Castronova wrote:

I'd argue that the point of sharing is not necessarily interacting, it's emotional validation. All achievement-related content could be instanced, but so long as the evidence of achievement has some effect on how we are perceived in a shared social/physical environment, the achievement receives a powerful validation from the social reputational effect.

I agree completely. I suspect this is largely at the heart of the matter for many people who oppose RMT because it "cheapens their achievements". They want to be able to show off the evidence of the difficult things they have achieved... the task of accomplishing that goal could be entirely instanced, so long as the social-interaction portion is not.

magicback wrote:

To elaborate further, with the advancement of customized client technology we can filter, translate and annotate our own perception of the world client-side. We can create our own Percoset-induced experiences. Server sends a “hello” and I receive a “greetings”. Server sends an NPC opponent with X ability and at Y difficulty and I receive it as a Spider with X1 ability and Y1 difficulty. We can create our own individualized illusions all at client-side.

I've been wrestling for some time with a pseudo-instancing concept along these lines... it's a difficult problem.

One of the reasons that I see people calling for instanced content is the aspect of non-consensual disruption of your play experience by others. I'm sure may people would be happy to have others play with/near them if they could do so in a manner that didn't ruin your own experience. One of the major drawbacks to the highly anonymized nature of typical MMO player interaction is that it seems to draw out the worst in some people. Griefing, harrassment, ridicule, circle-breaking dialog, just plain being a jerk because you can... all of these and more contribute IMO to driving people towards wanting instanced content so that they can get away from this.

Current implementations of the ignore feature only go so far in helping to remedy the situation, since you're still sharing a spatial metaphor with someone who, it seems, is hell-bent on making other people uncomfortable. It's apparently how they get their kicks. We need look no further than the M-rated review of WoW at Pointless Waste of Time to see this player type in full regalia.

When players like this can abuse the spatial metaphor to inject their own crassness onto others, the /ignore feature is not enough. I've been trying to come up with a workable extension to the /ignore concept, that extends fully into the spatial metaphor. I'm hoping that if I can get it working, we can have the ideal blend of player-specified partitioning.

As it stands, we have a "banish" command in the works. Utilized in a similar manner to /ignore, players gain the ability to have not only chat communications cut off from someone they don't like, but the other player's avatar and associated world-actions disappear too. We make this a reciprocal command, so if you banish someone, you disappear from their world-view as well. They no longer get the ability to interact with your avatar in any fashion. In effect, you gain the ability to put them on a different shard from you.

No more jackasses dancing in the fire at the auction house, no one camping a stack of mounts over the top of the mailbox, no more spammed AoE spells lagging you down to a few frames per second. Each player gets to make their own "partitioning" of the spatial metaphor to exclude (and be excluded from) those that are bothersome.

There are, of course, natural problems of scaling when managing this, as the number of potential relationships rises as the square of the number of participants in a space. I'm hoping that I'll be able to find a way to minimize the performace impact, as this seems like an effective solution to at least SOME of the desires to "play alone"... I don't think it's so often wanting to be alone, as be free from jerks ruining your gameplay.

I think we'd see people naturally partitioning themselves into like-minded spaces... you gain a world where the people you interact with are the ones from whom you gain a positive experience.

19.

Dmyers> It’s my observation that mmo designers have this notion that playing a game and playing are equivalent. My notion is that these two are quite distinct. Playing a game (or, more pointedly, abiding by the rules of a game) binds and restricts play.<

I think this is a good point. I would take it further: The emphasis on the VW as a game space hobbles its possibilities as a social space. For me, a big part of the impulse to solo in WoW comes from the lack of opportunities to find the people I would want to group with. What tools there are, meeting stones etc. select only the game properties of an avatar, not their players social character. After a few groups in which most of the characters have a wildly different playstyle to mine, its easier to give up and just solo.

A solution that perhaps would work for me is “whole world instancing”, in other words, servers with different rulesets that attract different groups. Examples include a server that is only up one particular night of the week, for the really casual player. A server with severe limits on rate of item and experience acquisition. A server in which all dropped weapons and armor bind on pickup. Etc. Each would attract a different sort of social group, giving players a much better chance of being with people of similar playstyle. And such servers would only require a few lines of code to make the change.

Of course, on any of these servers, the “game” would suffer some. Its tuned to a particular set of conditions. But on servers where the “game” is background to social play that would matter very much. I just don’t think the Diku style game is that good a framework for building a community. Servers that select for other shared interests or playstyles would I think reduce the incidence of soloing.

I’m not at all convinced that most people would want to live alone in their own little instance. It sounds kind of deadly in the long run, like taking too many happy pills. The surprises other people bring are part of what keeps life interesting. But that doesn’t mean I want to be chained to the same server as people already living in a totally different world. Something less than a random selection of all humanity, and larger than a six person instance seems about right. “Niche” servers could be the right size to me.

20.

Edward Castronova > "I actually see a future like the one Dan describes, where each player has a huge mass of AI devoted just to him, yet where the results of his actions also register with other real people."

Of interest here: Will Wright's new game, Spore, won't be massively multi-player, but the content populating every local world will be evolved by the exterior player community. You'll never interact with another player directly, but you'll collectively build each other's environments.

If anyone wants to check it out, video of Will's demo-heavy Spore presentation is up at GDCTV (free registration). I've made this time index of demo events for something I'm working on, which you can use to jump around:

-12:30 for the first stage of Spore (simple, underwater organisms)
-17:45 (macro body plan, crawling out onto land)
-30:00 (tribal, memetic, technological)
-33:00 ("the city game")
-38:00 ("civilization game," pulling out to see the world)
-43:30 (spaceship)
-46:00 (zoomed all the way out to the solar system and colonizing other planets)
-51:30 (out in the galaxy, SETI tools for discovering other intelligences)

21.

Raph>I refer you to the book I never finished because Richard beat me to it:

I did this in part by ripping off the sections you did write!

Richard

22.

Sam Kelly> One of the most 'real' online environments I've been in was a very plain HTML-based chatroom, with public messages and private messages and nothing else, but the regulars decided what was there, put a lot of effort into maintaining the shared world and its contents, and the communal memory allowed a form of persistence.

Well, they weren't based on HTML, but we were using such spaces 20 years ago -- they were called "BBSs." ;-)

But let's get to the real question: instancing. Barry has once again nailed it: instancing is what designers do to prevent non-consensual disruption of one's play experience by other players. Or more directly: "I don't think it's so often wanting to be alone, as be free from jerks ruining your gameplay."

In short, instancing is a mechanism for temporarily limiting interactions between a player or a small group of players and the rest of the currently active player base.

So why do this? How can cutting players off from other players be desirable? As Raph says, the point of these games is that they're massively multiplayer -- if you're not interacting with a bunch of other people, you're not playing the game. In a way, the interaction IS the game.

The problem is that some kinds of interaction spoil the game. To expand on Barry's point, I'd suggest that there are actually three specific forms of unwanted player interaction that MMOG designers are trying to minimize:

* interruption of storytelling
* social griefing (harassment)
* camping of static spawns that drop high-end loot

To solve these related problems, developers have turned to instancing because, in addition to being relatively simple/cheap to implement:

* it permits coherent narrative/exposition
* it reduces social griefing (harassment)
* it guarantees access to high-end loot

What makes instancing a hard issue to address is that these goals aren't all bad, or all good. If you reject instancing because it seems like a cop-out (as opposed to finding a design solution to the loot camping problem), you wind up also losing an effective way to allow NPC characters to interact with your players in a controlled way for engaging storytelling.

Let's say that Darth Vader is telling you what a wonderful servant of the Empire you've been, and is about to clue you in on the Emperor's plans for your sector. If that exposition can be interrupted without consequence by another player sending you a message saying, "hey d00d, i need sum money," then maybe doing this kind of narrative through brief instancing isn't such a bad idea after all.

But is this really the main reason why developers use instancing? Somehow I doubt it. As far as I've seen, the #1 reason for instancing is loot. When the game is designed to allow some high-end loot item to drop from a statically-placed NPC or container, instancing is "what you have to do" to insure that all your high-level players can get their very own copy of that loot item. Instancing is just the most obvious workaround mechanism for a game design that allows ninja looters and spawn campers.

Storytelling and harassment prevention might be given as justifications for instancing. But the real reason seems usually to be considerably less noble: it's about insuring player access to phat lewt so they'll stop whining (about that, at least).

Is it really impossible to design a popular game world that doesn't rely on static spawns that drop uber loot?

Really?

...

Finally, there's still the question of whether this trend toward instancing is merely an artifact of needing a quick and dirty fix to guaranteeing high-end loot, or is another confirming instance (so to speak) of the "collapse of communities" theory that Dan referenced in the first message of this thread.

I suspect it's the former. The 20th century saw various persons expressing concern that the migration from urban life to the suburbs and exurbs would lead to the destruction of society by breaking generational and community relationships. Suburban life in particular has been vilified as a kind of social instancing popular among "I got mine" isolationists.

There's some reason to think that our society was in fact damaged by this trend toward a more individualized life, though the kind and degree of these effects are still being debated. Even so, they seem not to have happened to the extent feared by those who could only imagine highly-interactional city life as worthwhile, and who considered any deviation from that model to be a threat. Our society may have been fractured, but it survived. The republic endures.

Now we're in the process of building new channels for interaction, only this time they're based on networked ones and zeros instead of physical presence. If 20th century transportation technology allowed social instancing, 21st century communications technology seems to be allowing citizens to create their own shards by actively choosing the members of their social groups.

Perhaps a similar technology will allow MMOGs to likewise reverse their movement toward isolated play experiences and return to being about shared social experiences. Maybe the future is "persistent instances for you and your friends, linked to other instances."

When players have the tools to create such personalized but connected communities, will they do so? What will such game worlds look like, and what new rules will be necessary to support fun gameplay in these worlds?

--Flatfingers

23.

There's a ton of stuff to reply to in the above, and I think some are misconstruing some of my comments... but to reply briefly to Flatfingers, loot is a TERRIBLE excuse. Making sure that everyone gets a copy of a given item upon slaying a mob is trivial and does not require instancing in any fashion whatsoever.

24.

Raph: "Making sure that everyone gets a copy of a given item upon slaying a mob is trivial and does not require instancing in any fashion whatsoever."

Really? How would you do it then?

Give an equal share to everyone who contributed damage to the kill? Not fair on the healer who stayed back and kept the fighters alive, and open to abuse by J. Random Stranger who jumps into the battle just long enough to do a few points of damage and then runs off.

Divide up in proportion to the damage each character dealt? See healer above. Also unfair on the caster who keeps the mob ensorcelled while the melee specialists kill it, or the tank who soaks up damage to pin it down while the casters and archers do the real damage (both common tactics in Guild Wars).

Equal share to everyone within X metres of the kill? All the same problems only worse.

Assign them at random (the Guild Wars approach)? Without instancing you just run into all the same problems in deciding who's eligible for the lottery.

25.

Raph, I think we're in agreement that instancing isn't a perfect mechanism for insuring that players have a fair chance to get high-end loot.

But if this isn't the primary reason for the trend toward more instancing, then what is? If storytelling and griefing prevention are also not why developers are increasing their use of instancing, then what is the real reason?

I wouldn't be surprised if my list of pro-instancing reasons was incomplete; I just can't think of any more. If there's some more compelling reason than guaranteeing loot why instancing is gaining in popularity among developers, I hope someone will mention it here.

--Flatfingers

26.

As far as I know instancing has been introduced to reduce the immersion-shattering practice of camping, lining up for spawn points, and seeing popular dungeons or hunting grounds having been essentially clear-cut by roving locust-like bands of players. This devolves to loot production and sharing, but it's more about the player experience than anything else.

I think instancing can also be used powerefully for storytelling -- e.g., you go to find the small green muppet-like wiseman in the swamp, and you don't have to do it with twenty others or wait in line for your climactic conversation with him (which is depressingly like all the others'). I don't know if any games thus far have put it to this use.

27.

I listed storytelling and preventing interference on my list too. (To my mind, storytelling is a form o embedded game, btw).

I've just never heard anyone say it was because of giving out loot before.

The obvious solution to making sure everyone gets the same loot is to give the same loot to every member of the group--hand out as many Vorpal Snickersnack Swords as there are party members. There's games that do that right now in non-instanced situations. Instancing doesn't buy you anything in this regard; the group exists either way. I can buy that it reduces things like ninjalooting and so on. But it has little to do with whether or not everyone gets loot.

28.

To clarify: when I say "giving out loot," I'm not talking about dividing loot between members of a successful party; I'm talking about regulating how loot is provided to entire groups.

Instancing (as noted) has nothing to do with divvying up loot within a party. But it does seem to be gaining in prominence as the preferred tool for regulating how entire groups are allowed to try to obtain particular loot items. In particular, instancing is being used to address the problem of spawn camping, which ultimately is a "loot" problem.

Are static loot-dropping spawns really the only or the best loot distribution mechanism? Is no other design possible? Have designers truly exhausted all other possibilities for how to distribute loot, and instancing is the only effective solution?

Or is instancing used because it's an obvious, quick-fix solution that players think of and demand? If so, is satisfying this particular demand really a better long-term solution than designing a better loot distribution model?

--Flatfingers

29.

On second thought, I don't like the term "loot distribution model" -- that sounds too much like dividing up loot within a party, which is not the concern here.

"Loot regulation model," maybe? Is there already some industry-standard shorthand term for describing the mechanism that controls how players access loot rewards?

--Flatfingers

30.

"Drops" is the usual term.

If the concern is that not enough people get a given drop, first you make it more frequent. If it already drops at max frequency, then you add it to more mobs. This is a trivial exercise.

The reason why designers don't want to do that is to preserve the alleged storytelling function of having a given loot item dropped ONLY by a given mob. This is purely a matter of illusion (for one, many--most!--items are functionally identical, except in name). The "story" value from the designers' point of view is arguably close to non-existent; it's greater from the player's after-the-fact mythologizing point of view, actually. But I suspect players would mythologize quite ably were the items found on different mobs--there could even be plausible fictional reasons.

This is assuming, of course, that you buy into the whole "unique loot on a specific special mob" thing, which I don't. It's one design pattern, but it's not the only valid one (after all, it leads to loot camping!). Holding that one design pattern so dear that you go down the chain of causality to end up at a single-to-limited-player RPG within an instance feels like having your priorities wrong. There's lots of other ways to structure drops.

31.

I guess, ultimately, it comes down to the difference between focusing on ways to make masturbation more satisfying, vs. actually having sex.

The former is a poor substitute for the latter, and, even if technology someday provides a convincing illusion of the most interactive of all human experiences, it will remain a sterile exercise.

Note that I would fight to the death to protect anyone's right to focus their time, energy and resources to design better teledildonics, if that is what gets them off.

It just seems to me a rather foolish and self-centered exercise, particularly in a world full of real people--not to mention a waste of talent and energy that could be devoted to more productive purposes.

We here have a rare opportunity to create truly new experiences for large numbers of people, and we have a limited lifespan in which to do it. At the end, does one want to look back and say that one devoted one's energies to further isolating and dehumanizing society, creating large sets of parallel but nonoverlapping virtual experiences, or does one want to look back and say that one helped community happen?

32.

> Holding that one ["unique loot on a specific special mob"] design pattern so dear that you go down the chain of causality to end up at a single-to-limited-player RPG within an instance feels like having your priorities wrong. There's lots of other ways to structure drops.

That's precisely the feeling I had. Thanks for confirming it.

Unless there's some other good reason for instancing, it seems to be getting overused lately. I'd rather see more innovative ways to regulate loot drops given a chance.

For the 99 that fail miserably, there'll be the one that makes everyone say, "Oh -- well, but that was obvious." ;-)

--Flatfingers

33.

Instancing is probably more about isolation of social gameplay than about loot distribution.

I am sure you have heard of the phrase "get a room!" or "let's find some quiet place to talk."

So I think it's more about getting optimal population density then anything else at the fundamental core. EQ2 and other new games seem to have designed instances to address this aspect.

Frank

34.

There is a short story by Stephen Baxter (I think thats him) called Glass Earth that deals with this topic.

35.

magicback> Instancing is probably more about isolation of social gameplay than about loot distribution. I am sure you have heard of the phrase "get a room!" or "let's find some quiet place to talk."

Hmm. I suppose that's possible. But my off-the-cuff reaction is that, with most games offering various kinds of private chat channels, using instancing to provide a sort of in-game chat room seems like horrible overkill.

Your point about sizing is something to consider, though. If the maximum number of players allowed in an instance is keyed to the maximum group size, then -- assuming a small group size, say, 5-20 players -- I'd bet it was more about guaranteeing uninterrupted access to high-end loot drops.

But if we're talking instance sizes on the order of 100 players or more (especially if they're not grouped in any way), then you could be right; that might indeed be more about managing population densities.

Which brings us back to: What do social structures tend to look like when the player base is partitioned by design?

--Flatfingers

36.

They look cliquier, is my guess.

I don't buy the "get a room" argument either. Not even the most crowded MMO lacks places where people can go to be alone. It's not that sort of social driver.

From the player side, it's about exclusive access to content.

37.

I'd consider the desire to be more along the lines of "non-disrupted access" rather than "exclusive" access.

I suppose one definition of "exclusive" would be "non-excluded", but I think most people take the concept of "exclusive access" to mean that you're the only one who can access it.

Typically, instancing isn't used today to block others from accessing the same content... after all, everyone can have their own group's instance and access that content too. Instead, they are generally set up to allow shared access, but you're only sharing the experience of YOUR participation with those who have your consent to do so.

The distinction (to me at least) is as sharp as that between consensual group activities with people you KNOW, and the same activities with random strangers able to force their way into your activity whether you like it or not (and especially when the strangers are very ill-behaved).

(This non-consensual aspect gets especially touchy when viewed through the prism of galiel's "masturbation vs. sex" analogy.)

So I'd call it consensual sharing of experiences versus non-consensual sharing. I don't think the social interaction value is necessarily diminished just because you choose who to share your experience with...

38.

I think my quotes were taken with a different context.

It's more about consensual, but about optimal social/poplulation density.

The social dynamics I attempted to explain in terms of optimal population density (variable according to game and player) is the one Clay Shirky described in his A Group Is Its Own Worst Enemy essay.

My take on it is that in a house party, the house have different rooms and a few entry points. People come in through the entry points and may gather around first to grab some food & drink (content), but will move to different areas of the house to find their optimal population density (OPD). Loners will have an OPD of one and may go lock themselves in the bathroom. Socializers will have an OPD of MAX and find the biggest room.

Loners (using the loner-socializer axis) may required pre-defined content: food, drink, TV, PS2, etc. However, socializers will require OTHER PEOPLE. Thus, my conclusion that OPD is the underlying desire. You can argue that the actions of other people can be classified as “content” for socializers, which is valid. But, I think you lose some explanatory power at this lower denominator.

Take another look at the in-game auction house. They may be globally-connected or localized. I would argue that the localized auction houses are “instances” as shards are “instances”. Why not have everything globally-connected across server? Technology limitations, provision of content, or OPD? What happend to Utra MMOs?

So, we may be talking in circles about the same thing. However, I’m attempt to nail down the core issue and terminology.

Dan in his opening comment write about an drug-induced half-phase OPD of one (similar to two people in the same room playing the same single-player game on two separate unconnected computers). I think this experience is what Will Wright is attempting to leverage on with Spore.

I think this is the core issue.

39.

I have never posted here before but I am reading this as it was linked from another page and wanted to reply. To the original poster, I hope your face is feeling much better and the percosets are treating you well.

To hop back up to midway through this page I would like to explain my thoughts on the Diablo vs GW and whether or not they are MMOs question.

The difference between the sharded towns of GW or the chatrooms of BNET is fairly simple in my opinion: Intent. I never got the impression that the chat rooms of BNET were *intended* to feel like part of the game world. Whereas with GW it is pretty obvious that the sharded towns were *intended* to feel as though they were part of the world. I wanted to make this distinction because I consider MUDs to be MMOGs and in many cases they are not much more than chat rooms. There is no graphical representation of space however the illusion of space is created through the text descriptions etc.

That being said I feel that GW leans more towards being an MMO than Diablo does because it is linked through part of what is intended to be the game world. The towns allow interraction on a "massive" scale within the game world itself. Diablo does not offer this.

So does this make GW an MMOG? I personally do not feel that it does. While the town certainly displays some of the characteristics of a MMOG, in my opinion the majority of a game must display these aspects to consider the game as a whole as an MMOG. The majority of GW is a single player or small multiplayer game with only a very small portion of the world linking it all together. As someone mentioned above, if you take all of the instancing away from any MMOG you are still left with a playable game in most cases. GW is not such a case. It is the exact oposite of something like WoW where most of the game is MMOG and a small portion is instanced.

That leads back towards the end of the thread and discussing instances. Instances are perhaps the worst thing that has ever happened to the MMOG market imho. As has been mention over and over above they limit social interaction which I feel is the only thing that seperates MMOGs from other game genres. The why of this is pretty simple I feel as well. It makes design so much easier.

Instances give control to game designers that they cannot easily achieve in the great outdoors of an MMOG:

-Right off the bat you can limit the number of players which makes it much easier to decide upon difficulty levels and insure the content provides a challenge but is not too difficult.

-Again based on how many folks will be in one instance you can more easily achieve metrics on how intense the graphical representation within can be without stressing the engine and/or creating massive lag for the player.

-You can get by with getting more out of less content. Creating an instance that only allows a single group of 5 to complete basically creates an effective 10X as much content as a typical dungeon of the same size (assuming of course you have say 50 people who want to go there).

-Someone above had mentioned loot and I think that instancing allows the designer to specify a degree of difficulty that needs to be surmounted before delivering loot. In most "outside" situations loot gathering can be trivialized by the "zerg".

-I could go on but the point is it simplifies design so just assume and etc etc etc here.

On top of simplifying design its very nature provides some easy solutions to common MMOG complaints such as kill stealing, interruption of immersion, outside interferance etc etc and many more things which have already been described above more eloquently than I could manage.

All of this seems pretty win win to me from a design perspective and it does offer some pretty substantial benefits for minimal effort. As someone posted above there are some things you can easily accomplish in story telling or player effect upon the world that would be nearly impossible in a non-instanced setting. At least not without some serious coding and logic development. So as this all sounds great and wonderful why do I still feel that instancing is a complete cop out?

It's because I think that instancing steals the very "soul" of the MMOG. The one thing that seperates an MMOG from the rest of the gaming world. The completely random, unscripted, unique, and LIFELIKE experience of social interraction. I once tried to explain to someone why I absolutely loathed GW but could still have fun playing WoW. The short short version is that in WoW when I am out "adventuring" there is a chance I might encounter another living breathing person who might change my gaming experience from what it was the last time I was in that same locale.

To me this social "spark" is what brings a persistant online world to life. While improved AI may eventually provide a more lifelike environment I doubt it will ever trully be able to replace the experience of interraction with another human. Be the experience good or bad it is the variety and randomness that bring spice to life. In locking the player inside an instance you lock the life of the game out.

just my 2 cents *shrug*

The comments to this entry are closed.