Since my last post was pretty theoretical, I thought I’d bring this back to earth a bit.

The MMOG market continues to be very hot, and possibly all but impervious even to our current economic chaos.  I continue to see MMOGs in development for ever broader demographics and more obscure (or focused) niches.  Despite the difficult times for some and the demise of others, investment and development in this area continues to be strong.

And yet technology continues to be a huge thorn in the side of any developer.  There are a number of middleware suitors trying to woo developers, but recently an unusual one has appeared on the field.  Can it be that Facebook will save MMOG development?

One of the enduring questions for any potential MMO developer is, what technology to use as a platform?  Until a few years ago, it was common for MMO developers to create almost all their own technology, with the possible exception of a 3D engine for their client.  This necessity has greatly increased the time, cost, and risk of almost every MMOG project.  I have likened this in the past to someone saying, “okay, you want to make a movie?  To get started, you first have to learn to grind your own camera lenses.”  The intricacies of 3D scene graph management, scalable data base construction, billing systems, login servers, customer relationship management software, world creation tools, artificial intelligence, story creation, and a slew of other major areas, while all necessary for an MMOG, are typically far outside of the core competencies of any MMOG developer (trust me, no MMO developer rubs their hands in glee and says, “great, a new MMO! I’ve been wanting to try out a new account creation mechanism!”)

In response to this, many MMOG middleware providers have sprung up — companies like Bigworld, Simutronics (with the HeroEngine), Icarus, and others focusing on providing all-inclusive solutions as platforms for creating MMOGs.  In the past few years a few others have popped up too with a slightly different model — companies like Multiverse and Metaplace, supplying middleware tools, but focused on hobbyist users and professional developers.

Each of these suites of tools has its strong points and weak points — either the cost is seemingly astronomical (approaching or exceeding a million dollars for a license), or the tools are incomplete or immature.  No professional developer wants to risk an already expensive, risky project on technology that may leave them high and dry mid-way through development, so everyone is looking for the first big project that is completed using some form of middleware.  In effect, all of us are waiting for someone else to take the first step and show that these tools can in fact create viable third-party MMOGs. So far, we’re all still waiting.

Amidst all of this is the typical industry blather that MMOG budgets are going up to unattainable levels — frothy comments about $50M and $100M budgets are not believable (and not necessary, if you’re building your game wisely) in my opinion, but even a more “reasonable” $15-20M budget for a professional MMOG can be difficult to fund.

So as often happens in nature, evolution finds a way.  If the budgets for “mainline” MMOs are absurdly high, then someone will figure out a way to make these games with absurdly low budgets, and still be successful.

And this is where the odd miracle of Facebook comes in.  I’m not going to go over the history of Facebook or other social networking sites enabling external apps; you can find that all over the web.  ‘Social gaming’ is also all the rage right now, though even most of those creating these games really seem to have little idea what it means, based on the thin threads of socialization present in these games (attaching a chat channel or leaderboard does not make the game suddenly “social”).  But there have been a few interesting bright spots, the first indications of an evolutionary change, in games like Mob Wars (and all its many copycats) and YoVille.

These games aren’t what any die-hard MMOG player would recognize as “their game” and in that regard may well represent the next stage in MMOG/VW evolution.  Mob Wars has “jobs” (essentially quests) but no geographical world as such.  YoVille has apartments and a geogrpahical world, and been described as a scaled down version of The Sims (or maybe Habbo Hotel — this gets back to the unnecessary conceptual divide between MMOGs and virtual worlds that I’ve already covered).  In fact YoVille seems remarkably similar in its intent and gameplay to The Sims Online, which shows perhaps how these new games are recapitulating old lessons without having entirely learned them first.

The main things these games have in common is that they play in the browser (as do many other MMOGs in the current crop), they are free to play (ditto), and significantly, they have let Facebook do the heavy lifting as a social platform and in part as a purely technological platform.  That is, tasks that are distracting, tedious, but necessary like account creation, user verification, etc., are handled by Facebook.  Flash (as a bare technology or using existing client-server wrappers) enables the client-side.  And since the games exist within the context of a social setting already, the ‘reach’ issues of any game, but especially any free-to-play one, are greatly reduced. Smart designs encourage players to bring in their friends — though notably without trying to trick players or spamming their entire address books as earlier putatively social efforts have done.

Moreover, while the revenue model on Facebook continues to evolve, the purchase of virtual goods is gaining acceptance quickly, and truly novel monetization methods are gaining a real foothold.  For example, in games like Mob Wars, players can earn points usable in the game by filling out partner surveys or visiting other sites — so-called Cost-Per-Action advertising (assisted by companies like OfferPal and SuperRewards) where the advertiser gains useful information and actions, the game developer gains dollars in payment, and the player gets in-game tokens of one sort or another.

How successful is this?  Mob Wars reportedly is the top Facebook game, making more than $20,000 per day (external estimates of $7-15M annual revenue) on about 500,000 daily active users.  I don’t know how much Mob Wars cost to develop, but given it’s lack of animation, geography, heavy quest trees, or many other infrastructural elements, I’d be surprised if it was more than a few hundred thousand dollars at most — probably a whole lot less.  Even though Mob Wars is the top game, it’s not as much an outlier as WoW is for the rest of the traditional MMOG pack — a compelling ROI case for any online developer.

In short, I think we can expect to see a lot more MMOG/VW-like apps on Facebook and other sites (but Facebook continues to lead the way at this point, not yet ceding the field as other SN sites before it have done). If your friends are already there, why would you go someplace else to play games with them?  And if you want to play games with them, why not persistent games that have real social content as well?  Facebook (plus Flash, readily available database software, evolving revenue models, etc.) make MMOGs there more than possible — they make them all but inevitable as the field continues to evolve.

You may be familiar with Maslow’s hierarchy of needs (more on this below the cut).  Maslow’s theory has heavily influenced the architecture of our AI technology, which is why I’m attuned to discussions of it or instances that support or undercut it.  Recently I ran across a theory in education known as “CBUPO,” an ungainly acronym for “Comptence, Belonging, Usefulness, Potency, Optimism” designed by Richard Sagor at Washington State University (an accessible introduction can be found here (pdf)). Sagor’s theory suggests some interesting modifications to Maslow that have consequences for how we understand ourselves — as well as the motivations for gamers and AIs.

(Warning: psychological theory leading to AI and game-relevant thoughts below.)

Maslow’s theory that says we each have motivations that change as we develop: first we’re concerned with physiological needs, then safety and security, then social acceptance, then skill-esteem, and finally our ability to contribute, what Maslow called self-actualization.  This has been an important theory in much of psychology, though it’s sort of fallen by the wayside during the cognitivist revolution.

Sagor references Maslow (and others such as Glasser, who has a similar Choice Theory of motivation) in the background to his “CBUPO,” idea.  In doing so however he creates a fundamental difference: Sagor essentially says that functional belonging requires basic competency — mixing up Maslow’s “belonging” and “skill” layers.

So whether you’re a kid in school or an adult in the office, you don’t just accept someone “for who they are,” you accept them for what they bring to the group.  You may accept them provisionally (“hey new kid, come sit with us”) but their full acceptance will depend on whether they show basic social competencies (how they dress, what they eat, their familiarity with cultural references, etc.).  People who don’t appear competent to the local standards are rejected from the group.

Which makes me wonder about Maslow’s clean division between a desire for social acceptance and the need for skill esteem.  When I first started thinking about this, these two levels seemed conflated, and possibly really confused or even inverted.  The more I’ve thought about this though, the more I’ve begun to think this is actually a new illumination of an important part of Maslow’s hierarchy (that at least in my reading, he doesn’t speak to clearly).

(Yes, this gets back to games and AI, hang on.)

The lower needs in Maslow’s hierarchy are what he calls “deficit” needs – the less you have, the more you want to fill it.  This is true of hunger, thirst, basic human contact, feelings of physical security, and to some degree, socialization.  The “higher” motivations are termed “growth” needs – these don’t have an urgency when not filled, but represent human growth.  We’ve modeled both deficit and growth needs in our AI, though we’ve taken a slightly different approach when modeling the latter.  We’ve said, in effect, that the higher motivations, skill esteem and contribution, are quiescent until they’re used, at which point they kind of “perk up” and reward the individual (an AI in our case) with feelings of satisfaction.

So as a young kid you may not be particularly motivated by enhancing your skills, but the first time you tie your shoes, say, you want to do it again — you have a skill-based motivation that is suddenly more part of your overall motivational set.  If you don’t have many positive skill-related experiences you won’t develop this motivation much; on the other hand if you have many positive experiences of this type it can become the overriding motivation in your life (see: professional work-aholics).

One hazy part of Maslow’s theory, in terms of trying to build it in as part of an operational AI, is in the transition from deficit needs to growth needs.  How and when does this occur?  Sagor’s ideas may shed some light on this.  If we all have a desire to belong to a group (once our basic needs for today and security for tomorrow are more or less taken care of), then we have to confront what such acceptance requires of us.  We all want to believe in “inherent acceptance” — a mother’s love, for example — but even when we have this in our lives (and far too many do not), it does not replace the need for more general acceptance and belonging.  As children, as we begin to branch out from our families in to society, we want to be accepted there.  And belonging with others socially, as Sagor points out, requires various forms of competency: in effect, “okay kid, what can you do for us?”  This is as true in church groups and science clubs as in violent gangs.

And not incidentally, this leads to the activation of the growth-based need for skill-esteem.  In effect, our lower-level desire for belonging is able to be satisfied more completely as we show that we have some locally required level of ability.  It’s noteworthy that people will do whatever they need to find this sense of belonging: they will excel academically if they can; extend themselves socially (wearing new clothes, using new slang) if they must; learn obscure literary, movie, or musical references; or, if nothing else works, show their ability to dominate in terms of more basic needs via violence and aggression.  Ultimately it doesn’t matter if your group accepts you for your knowledge of medieval literature, your ability to quote Star Trek, your skill in hunting deer, or your willingness to jump in on a fight — its your basic competency, evaluated in socially local terms, by which you gain acceptance and keep satisfied your need to belong.

This may all seem pretty theoretical, but it’s interesting to me in terms of designing effective motivational models for AI, and in terms of designing effective social experiences for people playing games.  It’s also interesting in terms of what it might mean for using games in learning situations.

In social gaming, we have thus far taken the shallower, weaker track of equating activities like “chatting” with socialization.  In MMOGs this is less the case, where someone’s value to a group or guild is often more clearly tied to their in-game abilities (they’re an effective tank or healer — or they can lead the party itself well).  But in many “social games” today we have a pretty shallow view of what belonging and acceptance means (consider how we’ve turned “friending” into a verb, and the concept of a friend into a neutered version of its former meaning).

I suspect that there’s an opportunity there for creating increased engagement and satisfaction in social gaming if we provide mechanics and dynamics that support both provisional acceptance and skill-based belonging.  Enabling people to accept (“friend”) others for something meaningful to both people and to the group they evolve into — where ‘meaningful’ depends entirely on the context of the game: as in life, it could be your ability to take targets down with a head-shot from far away, your ability to rally troops, grow crops, teach others new knowledge and skills, or lead the group overall.

In games even more than in life, we have the ability to determine the kinds of skill expertise will gain people acceptance and belonging.  If we construct social games where knowledge of ancient Egypt or the German imperfect tense are valuable to the group, this will work just as well as a motivation for play (and not incidentally, skill increase — a.k.a. learning) as if the task is destroying buildings or head-shotting enemy soldiers.

From James Portnow’s blog, a terrific Zero-Punctuation-style video on the Uncanny Valley.  You probably know what that is, but it’s worth watching the video and passing this on to others who don’t.  And if you don’t know what that is and how it applies to games and AI, you really should watch it.

So what about the Uncanny Valley?  What are the best ways to approach one of the two peaks (on the near or, gulp, the far side) of the Valley?  How much does emotional and social resonance in computer characters help?  Are we pushing a huge rock up an ever-steeper hill by pursuing incredible graphics for characters that are effectively lobotomized and becoming creepier and creepier?  Would more broad-stroked graphics along with really effective (and affective — emotional, social) AI be more compelling?

And in case anyone is wondering why compelling AI matters, it’s like this: what we find emotionally and socially resonant we find compelling (see any Disney movie).  What we find compelling sets the stage for meaningfulness.  And what we find meaningful we trust — with our time and our money.  So yeah, it’s about monetization: create compelling characters in meaningful situations, and people will line up to interact with them, paying for the privilege over and over again.  Photorealism can stand in for meaning for a little while, but not for long, and not nearly as well as emotional and social resonance.

People like games.  People also like learning — mostly.  And of course adults like it especially when their kids learn.  Many valiant attempts have been made to use games to teach kids or adults, but with few real, intentional successes.  This is largely an unknown art, and one where when learning does occur, it seems almost accidental.

For example I learned about the geography of the Caribbean, I’m abashed to say, by way too many hours spent on the old Pirates! game; and my son learned a surprising amount of history by playing Age of Empires.  Many people have fond memories of Oregon Trail, and this often comes up in discussions of “games used for education,” but still this area has languished rather than flourished.

Why is it so difficult to make games for learning?  Is it the topics we’re choosing, or a too-pedantic approach, or something else?  I don’t have any solid answers on this one, and would love to hear others’ opinions.  What do you think?

What’s the difference between a virtal world (VW) and a massively multiplayer online game (MMOG)?

No, the answer isn’t “lipstick.”

VWs and MMOGs are like estranged siblings.  They share a common background and future, but right now at least, they don’t talk too much.  I think a lot of this is artificial or the result of odd circumstances having little to do with the online worlds themselves.  And as with many estrangements, I don’t think this one is particularly healthy.

So why are these separate, and what will bring them back together again?

Loosely speaking, MMOGs grew out of text MUDs, and VWs grew out of chat rooms — and less game-like MOOs — but both have their roots in the online social experience.  MUDs and MMOGs have, at least through the apex of that form in World of Warcraft, catered to achievement-oriented users, people who wanted to go out and “do stuff’ (more precisely, “kill stuff”) in an open-ended game-like setting.  VWs have from the start been more social, more about chatting and hanging out and… stuff.  Often, from the beginning, “stuff” included a range of social activity from crushingly banal chat to often repellent (to others) ‘cyber-sex’ chat.

From the beginning there’s been crossover.  Was Habitat a proto-MMOG or a proto-VW?  Both, I think, having significant social and game aspects.  As technology marched on we had worlds that were purely social such as Club Caribe, Worlds Chat, and Active Worlds, and we had games that, while social, depended on their gameplay to keep them interesting, such as Meridian 59, The Realm, Ultima Online, Everquest, and of course World of Warcraft.

As these forms grew, they divided a bit, though mostly without broad recognition.  First, the gamer types were all about the gameplay, and pretty much continue to be.  For them the social aspects are secondary (though any MMOG designer will tell you that it’s the community that keeps people playing long after the content has lost its shine).  Social and community aspects remain peripheral to the MMOG experience in almost all games (that is, the guild you’re in — and note that it’s usually still singular — has no overt effect on your gameplay, whereas the armor you wear clearly does).

Meanwhile the “social worlds” went through incarnations like Onlive Traveler, Active Worlds, There, and eventually to the peak of this form so far, Second Life.  In all but the last of these there was little to do but “chat” (in AW you could build buildings but couldn’t do anything with them).  One of the key differentiators in SL is that while it didn’t provide “gameplay” per se, it removed a lot of the sterility of the other “social worlds” by allowing users to create scripts to be run in the world itself.  And, not incidentally, Linden Labs focused heavily on creating enormous buzz in the mainstream media for Second Life, so that now even moms and pops and executives who want to jump on this “innernet thing” can at least wave their hands about virtual worlds… like Second Life.

And then a couple of odd things happened.  First, those who had spent more times in MOOs and early social worlds saw how many non-gamers were flocking to online social destinations, and thought it was high time social worlds got their due respect — after all, Second Life was now the media darling, and many exuberant reports predicted it would crush World of Warcraft in time.

Second, the highly successful Austin Game Conference was sold to CMP.  The AGC was widely recognized as the premiere conference for MMOG developers, if maybe it didn’t focus so much on purely social worlds.  CMP bought the show and turned it into an extension of the annual Game Developer’s Conference, wanting to maintain and increase the focus in the Austin conference on MMOGs.  AGC’s former owner, Chris Sherman, then went on to create new conferences, including, not surprisingly, the Virtual Worlds Expo (now Engage!).  Well-founded rumor around the game industry is that part of the terms of the sale of AGC to CMP was that Sherman’s company would not run new conferences in the MMOG space.

This sounds perfectly reasonable from a business point of view.  So the VWE focused on Second Life, primarily, with a smattering of other non-game and quasi-game companies respresented.  The show has catered to those in education, business, government, Hollywood, and elsewhere who want to explore this whole online world thing but maybe without getting into “men in tights” (the semi-derisive and often self-applied name for traditional fantasy-based MMOGs).

Unfortunately, the combination of this split in emphasis (social and game) and conventions to talk about them has driven a true but completely unnecessary wedge between these views of online worlds.  Virtual worlds now are a bit like Canada to the MMOG’s United States: beautiful, expansive in their own way, great places to go — but maybe with a bit of an inferiority complex about their much more numerous, loud, boisterous neighbor.   Some VW commenators maintain the difference between these and MMOGs almost implicitly despite the greater commonalities than differences between them.

And this to me is the ironic part.  As MMOGs advance beyond the traditional “kill monster, get gold” model, and as more and more people finally recognize that they want more to do than dress up their avatar and chat about the same old subjects, these forms are drawing together, not apart.  MMOGs do not flourish without a robust social and community aspect, and developers are finally recognizing that, like usability or combat system balance, this isn’t something you can just slap on at the end.  In a complementary way, while millions of people have visited Second Life at least once, the vast majority have not stayed — they’ve found the experience too difficult to get into, or too boring to sustain over time.

Many of the newer online worlds recognize this: destinations like Maple Story, Club Penguin, Dofus, Habbo Hotel, and others (many still in development) are complete hybrids supplying both fun gameplay (even if not in the “here’s another quest for ten rats” form) and support for compelling, demographically appropriate social interaction.  Sure there are new “chat destinations” and gamer-games like Warhammer or Age of Conan, but these I believe are stuck in the past.

So what’s the future?  I believe that this goes well beyond “better chat” or forums or the like.  I believe this goes to allowing and encouraging –often via gameplay– new levels of strong community formation in games, and more things to do, more ways to affect the world, in social situations.  Ultimately I believe that we’ll look back on this divide between virtual worlds and massively multipalyer online games as a historical oddity, an overemphasis on one leg at the expense of the other, when both are necessary to create compelling, long-term, and yes, monetizable experiences online.

Finally, I’ll toss out a theme that will likely become recurrent: I believe that if MMOGs and VWs are two sides of an arch leaning in toward each other, the keystone is what might seem like an unlikely factor, but one that is actually crucial: artificial intelligence.  That is, artificial agents who are part of each of our social contexts, but who literally live in the online worlds we only visit.  “AI shopkeepers” is not just the tip but the smallest sliver of the iceberg.  Socially aware agents who know each other and who know us, who interact with each other and with us, who live in the worlds we create, will finally give dynamism and life to the online worlds we have now, freeing MMOGs from their static purgatory, and VWs from their sterile desolation.

As always, I’ll talk more about this in the future, and I welcome other thoughts and comments.

I’ve been talking a lot about “social AI” recently as a way to differentiate what we have been developing from “typical” or traditional AI.

The easy way to say this is “we don’t do pathfinding.”  Which isn’t entirely true (we have a simple but effective pathfinding mechanism), but it shows where our focus is(n’t).  Agents need to move around a world, sure; and showing crowds of agents walking purposefully about makes for a great visual demo.  But to be interesting — or even more, meaningful — they need to do a lot more than that.

Another way we describe what we do is that we’re doing “heart and head” AI as opposed to “spinal” AI.  I believe that to be meaningful to players, or to human users in general, AIs will have to have robust personalities, bona fide emotions (and the ability to understand your emotions), memories, relationships, the ability to learn from their experience or from what others tell them, and of course the ability to (in some form) talk about what they’re wanting, thinking, and doing.  That’s a long list, but we’ve tackled most of these and are well on our way with the others.

‘Relationships’ is somewhere in the middle of that long list, and is a key aspect that shouldn’t be overlooked.  AIs need to be situated in a “social context” (the next step beyond a “social network”) just as humans are.  This is true in both game and non-game applications.  I get very excited whenever I think of the potential for socially plausible AIs that have relationships with those around them participating in social networks/contexts.  Whether this is Facebook or in an MMO, this adds hugely to the broad appeal, potential utility, and as-yet unknown forms of interaction.  I can’t wait to see where this takes us.

All I know is, the future of AI isn’t better pathfinding or strafing.  The future of AI is social — it’s us.

A new blog!  How unusual!  How exciting!

Okay, what’s this all about?

In this blog I’ll be writing about artificial intelligence, game design, virtual worlds, massively multiplayer online games, social games, and a variety of usually related subjects.  I welcome topical, respectful discussion.

My background: I run Online Alchemy, a small company in Austin, Texas, specializing in online social games, particularly those with a virtual world component.  We have been working for six years on advanced social AI for games and simulations (I’ll talk about that in another post), including working for several years with the Defense Advanced Research Projects Agency (DARPA). We have also created a unique social networking technology and have a variety of games in development.

Read on for more details of my past (well, not that many).

Before founding Online Alchemy in 2002, I worked for Electronic Arts for three years as a Senior Game Designer, leading game designs for several projects including SimCity Online, The Sims 2, and Ultima Online.  Prior to that I co-founded and was the Chief Creative Officer for The Big Network, an early family-oriented social networking company.  There I designed and produced several Java-based games, and in 1998 designed our primary product, MyPlace.   Before The Big Network I was one of the co-founders of Archetype Interactive (started in 1994, acquired in 1996), where I was the lead designer on Meridian 59, the first 3D massively multiplayer online game (MMOG).

Before getting into games I worked in software engineering, primarily in developing 2D and 3D CAD systems, and in user interface design and user-centered design for medical and scientific visualization systems.  In the early 1990s I designed the user interface for the first 3D visualization system used in the operating room by neurosurgeons.

I have a BS in cogntive science and have done graduate work in AI (expert systems, neural networks, genetic algorithms, etc.)

Oh, I live in Austin, Texas, with my wife, the last of our six kids (the rest are married or in college), three dogs, and a sugar glider.