Habitating

Randy Farmer has posted a thought-provoking critique of The Business of Social Avatar Virtual Worlds that spends a lot of time talking about Second Life.

Where does that leave us? Are social/avatar virtual worlds doomed to business extinction? Is there any way services like Second Life can make it?

Perhaps.

Read on for more . . .

In a similar way to Richard's recent Paradigm propogation, Randy spends some time reviewing the lessons of Habitat and points out that:

. . . Second Life embodies several of the original lessons almost to a fault, specifically:

- Communications bandwidth is a scarce resource.
- The implementation platform is relatively unimportant.
- Detailed central planning is impossible; don't even try.
- And especially our Future Directions section which said to let the users create the content - both the world and objects.

However, Randy points out that all is not rosy, and discusses to his own experience in using SL:

I loved it when I was unemployed. It was nothing but fun and intellectual challenge to produce an invisible teleporting 100-round-per-minute auto-cannon that ripped havoc throughout the WWII online community that settled there. Creating a Blade Runner blimp that traveled the world and handed out teleport cards to the city of Little Tokyo meanwhile playing a custom Japanese audio track was the highlight of my citizenship.

But, as soon as I got a job, I stopped creating, and then I stopped playing.

Rand goes on to argue that content "consumers" aren't interested in consuming user-created content and that "experimenting with a bunch of user-generated experiences of varying quality is just too heavyweight for people who are used to Television, Instant Messaging and Email."

He concludes with:

Focusing on the problems at hand: Consumers want to be fed content, they may even pay for it and a good platform can enable many talented people to create content, it seems that the main missing components are a way to identify and promote the content the consumers want and a create way to deliver it to them with the least possible burden on the consumer's part.

If Second Life can accomplish this, they will be the first.

An excellent article that raises very important points that are very in-line with our current thinking -- why, why did you have to go to Yahoo, Randy :-)? -- and that echoes thoughts that I've raised as well.

More generally, it returns to the questions of how to create content that has appeal beyond the creator, to enable creators to build experiences rather than just items and to provide more experiences that don't require 15 hours to enjoy. The next few releases will be fun :-)!

(Oh, and to save Greg his usual disclosure post, I'm the VP of Product Development for Linden Lab and Randy Farmer did some great consulting work for us on Second Life's UI)


Comments on Habitating:

Andres Ferraro (ex: DivineShadow) says:

"It seems that the main missing components are a way to identify and promote the content the consumers want and a create way to deliver it to them with the least possible burden on the consumer's part."

Beyond of the gameplay elements, The Sims Online method of searching for content (overhead view, keyword and category searches and well as map markings) is a pretty nifty way to get to relevant content.

Posted Jul 17, 2004 12:23:03 AM | link

Tobold says:

Sounds very much like the internet in general: Tons of content, but very much subject to Sturgeon's Law: 90 percent of everything is crap.

The internet, with its message boards, individual home pages, and blogs, has lowered the barrier of entry into content publishing so much, it is now nearly non-existant. At which point one realizes why in earlier times there was a barrier of entry in the first place, with editors checking what authors wrote before it being published.

Sure, if everybody contributes content for free, you will get some very good content at very low cost. But it will be swamped in the kind of content nobody is really interested in. And that is true for virtual worlds as well. Maybe the users creations should at least be subject to review from a game developer before being published.

Posted Jul 17, 2004 1:13:33 AM | link

ren says:

SL is currently my primary VW and I’ve been wondering about how the model will pan out in the long term.

Just so long as there is population increase or churn then the creators will always have a new market, but at some point the population must start to stabilise to the point where the majority of people have been around for some time.

Now the content market in SL seems to work pretty much like it does in FL (First Life) there are trends, people like to have the latest clothes for parties, art for the house etc. But, of course SL has created the stay clean fabric, the light bulb that never breaks. So as stuff does not wear out, what happens when all the major styles have been done?

Where does it go, will minor variations on existing themes work, will people get to building bigger and more complex structures, will it just be parties? Or is pure social sustainable? Is the predominance of the economy as a driving force just a trend and might SL really become a place that people want to be?

SL is certainly looking more and more like a fully functioning place, I guess only time will tell.

Posted Jul 17, 2004 1:15:31 AM | link

Hellinar says:

"the main missing components are a way to identify and promote the content the consumers want and a create way to deliver it to them with the least possible burden on the consumer's part."

"It was nothing but fun and intellectual challenge to produce an invisible teleporting 100-round-per-minute auto-cannon that ripped havoc throughout the WWII online community that settled there."

I think those two quotes encapsulate a big problem with unrestricted user generated content. Some very clever people will spend their time making content designed expressly to increase the burden on other subcribers. I'd guess that Randy's annoying invisible gun drove away some of the WWII online community. So Randy's subscription may have been a negative business proposition for SL. Which raises the question of how your content tools both give power, and enforce comensurate responsibility. Without that 80 precent of everything is crap, and the last 10 percent is actively toxic.

Posted Jul 17, 2004 2:55:53 AM | link

F. Randall Farmer says:

Hellinar quoted me:
>"...to identify and promote the content the consumers want and .. deliver it ... with the least possible burden on the consumer's part."

[And mentioned my little slaughtering machine.]

>I think those two quotes encapsulate a big problem with unrestricted user generated content.

Exactly! I'm thinking some future piece may be entitled:

"The Tyranny of Geography in Avatar Worlds."

The balancing act between a global sense-of-place and local control of permissions for user-constructed worlds is still a mess.

My blimp caused a controversy because it irritated some people when the voice of a Japanese travel commercial started blaring over the sounds of their medieval alchemy shop’s wind chimes. They responded by setting the bit that prevented all foreign scripts from running on their land and that problem went away, but now my blimp (and peoples air taxis, and dogfight-at-3000-feet-games , etc.) literally halted execution and ‘froze’ at the edge of their land, creating a patchwork of unusable airspace. Linden responded by limiting the altitude that the ‘no scripts’ bit would work, allowing my blimp (and other air traffic) to pass through, while effectively disabling the sound. All in all a seeming decent compromise.

But the point of my post (and building Hellinar’s excellent comment) is that even the implemented compromise is wrong for consumers. You can still see the giant airship as it comes thru Medieval Land at the minimum 50 feet. That isn’t a good consumer experience. There is more than one reason that Disney wants to control the airspace over their park: The most important one is not terrorists, but that air traffic tampers with their customer’s suspension of disbelief. New Orleans Square is ruined if you can see modern city high-rise buildings from the street or if a jumbo jet flies 200 feet overhead on its approach to the airport.

You can see this pressure in the growth of the Second Life world map. There are more teleport-only islands than ever. This limits some of the “geography” effects: Neighbor visual and audio pollution as well as “fly overs” and the a-little-too-pervasive trebuchet-launched-farm-animal.

Posted Jul 17, 2004 12:31:41 PM | link

Hellinar says:

Randy >But the point of my post (and building Hellinar’s excellent comment) is that even the implemented compromise is wrong for consumers.You can still see the giant airship as it comes thru Medieval Land at the minimum 50 feet. That isn’t a good consumer experience.<

I think A Tale in the Desert has made a pretty good stab at resolving the conflicts induced by powerful creation tools. The Law making system allows passage of laws that are then hard coded into the Server. Some limits to annoying creativity have been introduced through this. Not many though. I’d ascribe that to two factors (though the reasons are hotly debated in the world forums). One is that the Server based laws are always enforced, so can have unintended consequences. We usually rely on the law turning a blind eye in “silly” cases. Consequently, most laws beyond feature requests are voted down. People can see scenarios in which the law would have painful consequences. The other is that the subscriber base is small, and many activities in the world require cooperation. Getting a bad reputation is usually deterrence enough, and a law is overkill. I would say though that ATITD attracts more producer types than consumer types. A consumer would likely expect the game company to provide the laws, and police them.

Perhaps alternate geographies might reduce conflict? We know that the familiar Flatland is custom made to produce geographical disputes. For my own world, I’m considering a tree structure. Extra “land” can be sprouted in any popular spot. I’m hoping it will make for a more peaceable place.


Posted Jul 17, 2004 11:15:21 PM | link

Cory Ondrejka says:

Randy> The balancing act between a global sense-of-place and local control of permissions for user-constructed worlds is still a mess.

I mostly agree, although I think that this is more related to the tensions between global and local needs than to the sense of place. Ephemeral chat rooms might not have permanence or place but still have tensions between global rules ("no profanity") and local desires ("we want to talk about the movie Pulp Fiction").

Also, it is a mess, in part, because the current method of a priori creation of code rules to decide all of these issues. ATITD is an interesting approach to putting more power into the hands of the residents themselves. Many types of user conflicts that are easily discovered and/or resolved by humans are extremely difficult to detect or to resolve via code.

Posted Jul 17, 2004 11:31:14 PM | link

MM says:


Part of SL's problem with consumers is not really the consumers, but the producers. It is further aggrivated by the complexity of the scripting language.

The producers of the really top quality builds see their creations as art perhaps, and tend to keep those items to themselves. The fishbowl in Avalon is an example. I think that bowl would sell a copy to everyone who owns a house. But it isn't for sale.

This is aggrivated by the c-based scripting language. Since it's just a scripting language, you get all of the negatives but none of the positives of C. Something simple like a jack-in-the-box is really rather difficult to create. And that is artistic concerns aside. If the language choice was made to increase the value of scripts, I suppose it does do that. But it is at the greater expense of excluding too many potential scripters.

The scripting bit is a design issue though, and I don't think is indicative of a faulty model. I still have a great deal of faith in user created content. This is still in it's infancy.

Posted Jul 18, 2004 1:28:07 PM | link

Richard Bartle says:

F. Randall Farmer>even the implemented compromise is wrong for consumers. You can still see the giant airship as it comes thru Medieval Land at the minimum 50 feet. That isn’t a good consumer experience.

It seems to me that the people who hang out in Medieval Land don't want to play SL, they want to play Medieval Land. They're using SL as a hosting system for their own virtual world, which they are implementing in SL.

>"The Tyranny of Geography in Avatar Worlds."

In a shared space, everyone has to play by the same rules. The Medieval Land lot feel their magic circle is being broken by regular SL players, who in turn feel their SL experience is being diminished by people who want to exclude them. The solution would seem to be to rip out Medieval Land and put it on its own server. This means no shared space, no shared geography, and both groups of players can play according to the declared rules of their world.

Whether Linden Lab would like to find themselves running such "game worlds" is another matter, of course. It doesn't seem to fit their general ethos of inclusivity, their business model might be affected by it, and if the "game world" servers attracted more players then they could find themselves at the beck and call of the people who designed those game worlds.

It's interesting to note that people breaking the magic circles of game-like worlds have reflections that impose magic circles on social worlds.

Richard

Posted Jul 18, 2004 3:58:58 PM | link

A concerned blogizen says:

Time out! Everyone. Get over your dorkiness. (Especially you Richard--you are smart. But stop.) Even though Second Life made a mistake in calling themselves that ("Second"), digital worlds are a tool. "Role players" role play, in-world and out. But most people want a platform that lets them go about their day. You go Linden Lab. You go. Kill a dragon? Start a business! You tell me? No. I tell you! That's why I like you.

Posted Jul 19, 2004 1:21:00 AM | link

Richard Bartle says:

A concerned blogizen>Time out! Everyone. Get over your dorkiness.

You're basically asking us to close down the blog, here..!

Richard

Posted Jul 19, 2004 2:21:05 AM | link

A concerned blogizen says:

All apologizes, Richard! I don't mean that. :) What I would like, though (knowing that what I like here is not so liked) is a little more David Gelernter, Mirror Worlds. Is all.

Cheers!

Posted Jul 19, 2004 3:46:04 AM | link

Al says:

Sometimes I find it hard to get a handle on these things....

I frame Second Life content creation by analogy to Linux development - it's basically done by hobbyists for the fun of it or as a way of practicing/creating something they can point at as an example of their skill. Is that a fair assessment? I think there may be some sort of explicit financial incentive for some of the Second Life people? Is it fair to say there are similar problems in that 'boring but necessary' content tends to lose out to 'cool but not hugely useful' or is it too early to say?

Aside from licensing agreements, what's to stop the Second Life approach from maturing into a number of 'flavours' of virtual world, each including content in much the same way as Linux Distributions include the best applications? I'm aware of but not familiar with SL so a newby 411 would be much appreciated if I'm way off base!

Posted Jul 19, 2004 7:13:33 AM | link

Al says:

A concerned blogizen> digital worlds are a tool

Can't help but bite on this hook....

A tool for what purpose?
What is the tool - the world or the inhabitants? Either? Both? Something else?
A tool in its final form or still developing? Towards what?
A tool that might be usable in new ways beyond what were initially intended? How?
A tool that is also a place? Where is it? What is it? How does it mesh with the real world?
Who owns the place - providers, coders or inhabitants? Depends? What about objects in-world?
What rules govern the place? Who sets them? Are they enforceable in the real world? How?

Terranova tends to have a law/philosophy/economics/game design twist, looking at virtual worlds from an abstract and theoretical ("dorky") angle. It tries to scratch a little deeper.

Posted Jul 19, 2004 7:38:46 AM | link

MM says:

A tool for what purpose?

Lot's of purposes. VWs that model reality can be used as simulators. Testing bombs gets the most press coverage, but there are applications in medicine as well. Oh yeah, and games too.




What is the tool - the world or the inhabitants? Either? Both? Something else?

I suppose it depends on the application, but for purposes of your quiz, I'll say the world is the tool.




A tool in its final form or still developing? Towards what?

Still evolving.... towards the holodeck if we are lucky.




A tool that might be usable in new ways beyond what were initially intended? How?

I can make a world to do whatever I want, so I'm not sure I need to improvise. But I can unlock a door with a butterknife, so I'm sure we can expect all kinds of easter eggs from the future VWs.




A tool that is also a place? Where is it? What is it? How does it mesh with the real world?

This is all about definitions... what is a place? If it's only in my head, is it still a place? It meshes with the real world as much as the inhabitants say it does. (eg, message boards can effect the stock market)




Who owns the place - providers, coders or inhabitants? Depends? What about objects in-world?
What rules govern the place? Who sets them? Are they enforceable in the real world? How?

The debate rages on... read the thread about EULAs.

Posted Jul 19, 2004 9:07:08 AM | link

weasel says:

I've always considered direct user content-creation (in the SL vein) to be the worst case intersection between the problems of Sturgeon's Law and conflict in roleplaying 'styles'. (E.g. the dwarf with the 14 page Norrathian backstory meets an elf insisting he's a cyborg from Earth)

Unless an authoritarian figure is giving the 'yea/nay' to content based on theme, as well as quality, it's going to create serious friction between players. This friction is either going to lead to splintering of the community, or outright erosion of it. Eventually you'll wind up with one or more small communities, joined by theme and ruled by social fiat.

The amount of freedom provided almost guarantees factionalization to the point that I think 'massive' communities are unsustainable. The reported rise of teleport-only islands seems to follow my expectation.

Don't get me wrong: I seriously applaud Linden's achievements, and I respect their product. But I think that the market problems of such worlds is in large part due to the very fullness of player freedom that is their foundation.

Posted Jul 19, 2004 9:11:50 AM | link

Avi Bar-Zeev says:

Randy's (or was it Cory's?) comment about the design of Disney parks is a good one, but there's more be gleaned from it, I think.

Disney parks cater to many different themes in a relatively small space, and do it very well IMO. The walkways and skylines (trees and all) are designed such that you can't see one land from another. Turn a corner, and you're in a new land. This is even true walking around Epcot's big multi-national lake, where you can see across the lake, but your nearest "country" is still fairly immersive. Yet, physically, the distance is small (not that small, having walked it several hundred times..)

The key difference between a physical Disney park and a VW is the limitation on geometry in the real world, requiring very clever design and designers to keep the lands distinct. VWs don't have many imagineers.

Teleport-only islands are perhaps a way to bend the cartesian/euclidian rules in a virtual world that's perhaps too rigid in imposing uniform geometry on space.

There's no reason geometry must be euclidian or continuous, even for people who walk or fly, as long as the transitions aren't disconcerting.

I think VWs would do well to allow groups of users to effectively warp or fold space. One simple way is to add filter bits to content and lands/servers, such that sci-fi land and fantasy-land can sit side-by-side, yet not impose on each other. Call them magic curtains, walls of illusion, portals of passage, or whatever. Making CG constructs that filter out one set of content vs. another depending on where you stand seems a no brainer (and it's good for improving rendering performance ;)

Deciding how and when to apply such a feature is more complex, but perhaps not much more than allowing sims to set their rules such that Randy's blimp would only appear if it had the fantasy bit set. If it did, it might even show up as a different creature entirely, a flying carpet, for example. Avatars could similarly have multiple context-sensitive costumes as well. There are a ton more ideas in this vein, but the key is less reality, more creativity.

Posted Jul 19, 2004 8:30:40 PM | link

Ian Schwartz says:

Hi, long time reader, first time poster. Interesting topic. I play TSO and SL. TSO is called boring and not fun all the time, because there is only a certain point that you can reach, and there is no way to show off your creative output, other than house building, and roof art. But in SL, that is where I output my creativity. But for some reason, I am stuck in a boring point in the game. There is only so much you can do there, with your limited funds and a lot of items (i.e. clothing, guns, cars) that you can purchase. Every game seems to get a limit

Posted Jul 19, 2004 9:35:25 PM | link

magicback says:

Avi Bar-Zeev>I think VWs would do well to allow groups of users to effectively warp or fold space. ...

Avi Bar-Zeev>Deciding how and when to apply such a feature is more complex, but perhaps not much more than allowing sims to set their rules such that Randy's blimp would only appear if it had the fantasy bit set. If it did, it might even show up as a different creature entirely, a flying carpet, for example. Avatars could similarly have multiple context-sensitive costumes as well. There are a ton more ideas in this vein, but the key is less reality, more creativity.

As a follow-up to the above idea and my brief fly-by http://terranova.blogs.com/terra_nova/2004/07/paradigm_propag.html#comments”>comment on the Paradigm Propagation thread about creating a VW that reproduce the conceptual elements found in http://www.wizards.com/default.asp?x=products/dndacc/882420000”>D&D's Manual of the planes, let me expand further my idea:

My initial frame of reference is http://www.shadowrunrpg.com/”>Shadowrun’s concept of the virtual Internet, which the RPG called the Matrix. The basic concepts are that (1) we need iconography to help us make sense of the virtual world and (2) computers don’t care about icons. Thus , I would advocate the ability to create personal bit sets to translate all “foreign” objects. For Avi, the blimp can be programmed to appear as a flying carpet, while for me it can be programmed to appear as a mechanical bird. Both of us can toggle a switch to reveal the original bit set.

This addresses interpretive iconography, but the base virtual world (as set by the developer) also needs some kind of zoning policy. In TV-speak, it would be “a channel policy to avoid channel conflict and allow for channel guides to direct viewers to their desired content”. For this aspect, I like the structural reference used in D&D’s Manual of the Planes. Here is a brief overview of the relevant categorization of traits.

1. Gravity: no, low, normal, high, objective directional, and subjective directional.
Objective Directional Gravity “may be down toward any solid object, at an angle to the plane itself, or even upward, creating a chadelierlike world…” In a Subjective Direction Gravity, “the strength of gravity is the same, but each individual chooses the direction of gravity’s pull”. Zoning by physics may help alleviate the effects of a universal 50ft rule.

2. Time: slower, normal, faster, erratic, timeless
For current VWs, this is just a classification how time appears to flow. In an instanced dungeon, two days may have passed, but only two hours has passed in the base virtual world.

3. Shape and Size: Infinite, finite, or self-contained
In Self-contained worlds, “the borders wrap in on themselves, depositing the traveler on the other side of the map.” Teleport-only world would fall under Finite. Worlds with great expansive “body of water” may appear to extend infinitely, but is also a finite world.

4. Morphic Traits: Alterable, Static, Highly Morphic, Magically Morphic, Divinely Moprhic, or Sentient
The first two categories are easy to understand, but highly morphic worlds are further categorized by what can affect the "terrain" of the world: by anyone, by magic, by divine power, or by the world’s sentience. I consider SL as highly morphic, but am wondering whether a low level form of sentience can be coded. Randy, Cory, Avi, anyone?

With categorization of the different islands in an easy to navigate UI, I can create my own planar filter and map to fold virtual space and allow my blimp and dog fights to instantaneously bypass Medieval Land to the next available space without affecting my visual prespective. Conversely, the lords of Medieval Land can announce in their air-space usage policy statement that foreign objects can pass, but will be invisible; and my blimp will pass through Medieval Land visible only to users not bound by the policies of Medieval Land. This way, both parties are happy.

Now, taking the perspective of Yahoo as a virtual world, where there are very defined channels/zones of content, but no explicit knowledge of other users’ movements in the virtual world, I speculate that Randy’s is working to take advantage of spatial referencing to make users more aware of other users.

Thus, I think the future Business of Social Avatar Virtual Worlds and Internet Portals will be somewhere in between the two. Both have much to adapt from each other.

Frank

Posted Jul 20, 2004 3:50:38 AM | link

Avi Bar-Zeev says:

Interesting. One could argue that in the real world, everyone has their own set of filter bits, seeing every object and interaction according to their own bias. Just listen to Talk Radio in the US...

I'm sure that any open and/or P2P virtual world would need that functionality at every level--basically, anytime incoming data might not be trusted. It's possible SL and others could benefit from the idea of treating user content according to trust levels too, doing the filter thing above (remove, or replace with alternate representation) based on the observer.

That's fine for visuals, but scripting is of course much harder. If a script fires bowling balls that knock something else over, filtering the consequences is essentially like forking the world at every decision point (which could also be aruged on a metaphysical level, but not here). For VWs that gets pretty complex pretty fast. So I'm just suggesting doing the filtering by zone, rather than by person. It's much easier to implement.

The other note I'd make is on the "time flow" item above -- it's a pet project of mine to design a time-travel sim, but the MP version of that is probably impossible due to one simple fact: in "real" time travel (including differential time flow), if one person goes back and moves an object, those in the present wouldn't notice the object being moved, but would just accept it's "new" position. Real users have real memories, which can't be rewritten by any code I know how to write. We also can't change the user's perception of time, except perhaps by showing exciting or boring content, which isn't quite the right lever to be pulling. Food for thought.

Posted Jul 20, 2004 8:42:05 AM | link

ren says:

Ren > But, of course SL has created the stay clean fabric, the light bulb that never breaks. So as stuff does not wear out, what happens when all the major styles have been done?

Some one challenged me on this point last night, saying that an economy based on this kind of innovation is sustainable because there will always be new designers, new fashions and the building of brand and exclusivity will keep things going.

This is fine in theory. But I think in practice there is a question about just how many people in a VW are going to care sufficiently about this level of variation. OK one could argue that if the community is large enough then we really are replicating pre-existing social structures and the kind of pyramid structure of a capitalist economy with a few rich trendsetters a the top will make a sustainable model.

But commoditised VWs are economies on speed. Plus they have, in some sense, a perfect memory. Fashions can sweep through a system very rapidly, generic versions can’t be that difficult to make. Hence back to my original point can fashion really drive a digital economy of this type in the long run?

Posted Jul 21, 2004 7:24:03 AM | link

Mark says:

Richard Bartle> The solution would seem to be to rip out Medieval Land and put it on its own server.

I have been toying with this concept for awhile. The Web3D consortium is sure taking their sweet time, and I doubt that it will end up being what it needs to be.

The webserver should be augmented with the worldserver, and people should be able to move from one world to the next, just as they do now with hyperlinks. Perhaps throwing in some p2p would make it even more possible for people to serve up 3d content just as they serve up more mundane stuff now?

Posted Aug 4, 2004 12:54:37 AM | link

Vaz says:

Hey everyone. (For those of you who don't know me, I was the producer for WorldsAway's Dreamscape and Director of Community at Communities.com and There).

Frank said, "...the blimp can be programmed to appear as a flying carpet, while for me it can be programmed to appear as a mechanical bird. Both of us can toggle a switch to reveal the original bit set."

The issue here is, who carries the burden of creating the alternate imagery? As a blimp builder, I'm not interested in creating a flying carpet, a steampunk ship, a spaceship and a wyvern to go along with it. As a medieval town creator, I'm not interested in deciding how to mask every foreign object that crosses my borders. To ask either to do so is a huge disincentive for creation.

Frankly, I'm a much bigger fan of the controlled experience (surprised Randy? :-) so my bias lies there. Personally, I think Avi's suggestion ("I think VWs would do well to allow groups of users to effectively warp or fold space.") is a great solution as long as it does not disorient or confuse your visitors.

Posted Aug 18, 2004 5:48:09 PM | link