« Please Welcome Our New Ethnomethodological Overlords(Or, Bob Moore Joins the Author Roll) | Main | Size Zounts »

Jul 20, 2006



Great analysis Bob, and totally on target. There's so much rich engaging material, certainly in the social sphere, that virtual environments often don't tap at all. I mean, we've been passionately engaged in this stuff for the last 30,000 years (culture that is) - why shouldn't we try to bring some of those dynamics online in structured ways?

And certainly too, the online technology offers enhanced communication and information opportunities- I think that's half the allure of the virtual world, the ability to communicate and or gather information by means we don't normally have. Remember life before wikipedia?

And glad to see you're on the masthead now, so to speak! Look forward to staying in touch with your ongoing work.


So the root of the issue is that you're using a word that has some specific history and context within the games industry, and you'd like to use it the way you've always used it in other contexts?

I'm going to say no, you can't expect expect to enter a new field and expect everyone in it to understand the subtleties of your usage of word, no matter how reasonable it seems to you, unless you want to spend the rest of your time explaining that to every new person you meet, and attempt to change the meaning of the word to match your own.

If you want to mean "believable", then that's the word you should use - The quest for believability is quite different to the quest for realism.

In general, yes, I think "realism" is considered too hard, and a waste of limited resources (like programming time, CPU and bandwidth), since the word is (as far as I see anyway) usually understood to imply the whole spectrum of features, without discrimination. I suspect that you'd do far better discussing features of real life that might be worth translating into the game design for specific purposes, rather than accidentally implying you want to make everything more like real life without any particular reason.

So, in summary, don't talk about making games like real life, talk about learning from real life to improve games (the translation of concepts, and the utility of them is absolutely key).


So when do you think the R-word is appropriate in virtual world design? When is it not?

The easy (and largely correct, IMO) answer is: invoke realism or believability when it helps gameply; avoid it when it doesn't. (I'm going to use believability because it's less freighted with negative connotations, as Daniel points out.)

The more difficult answer requires asking what "helps gameplay"? If the game is about killing monsters and getting gold, a lot of social, narrative, object, and other forms of believability are going to be given a light treatment. If social relationships, object interactions, and/or narrative are more important, then you need to add more to the game in these areas -- perhaps at the expense of a combat system with fewer details and/or less believability.

You provide some interesting examples of social interactions or context for social interaction that might improve realism/believability, but I'm not sure I agree with them. For example, I don't believe a world needs to be "realistic" in its depiction to benefit from more believable social interactions. WoW is more visually stylized than EQII, but I think it could benefit from some of the same social devices EQII uses (and that those are still at the shallow end of the pool).

Similarly, I'm not sure There's word-by-word display of text (or SL's "I'm typing" animation) would afford much better social interaction in a 3D game. Almost anything is better than having the avatars standing around like zombies though, so this is a good area for research!

You mention that our lack of proprioreception and peripheral vision in-game means that gestures and facial expressions are of less use than in the physical world. I'm not sure this is so; it may well be that this is a user interface limitation in existing games, not a limitation inherent in the medium.

Overall, I think this is a fascinating topic. I have to say I was a bit dismayed this year at E3 to see so many fantasy games attempting to have "realistic" depictions -- Vanguard, Aion, Titan Quest, etc. Too often this effort at visual realism is misplaced, resulting in visually rich but emotionally flat views of the game -- no sense of composition, stagecraft, etc, as these things have been abolished in favor of the illusory goal of "more realism." Contrast this to WoW's stylized views that often look like sets in a movie or play, complete with unlikely but highly satisfying lighting and atmospheric effects.

I think perhaps that physical "realism" is the thing designers reach for when they have no sense of how to create physical setting or drama, and, paradoxically, social believability is avoided by the same designers when they have no sense of how to embed that within their world. This gives us worlds that are ever more detailed to the eye, and yet ever more unsatisfying in their gameplay.


Can I go off on a tangent here?

I have made a habit of being extremely conscious about using the distinctions virtual and physical to describe our multi-layered existence, rather than virtual and real, though I have to constantly correct even myself! (RL and IRL are just far too tempting as short-hand) The idea that virtual communication, relationships, etc. are somehow less real simply because they occur in a different modality has always seemed silly to me. I guess I have to take an existential approach: it's real if we say it is, and not real if we think it's not. Or is that a cop-out? Maybe I've been too influenced by Philip K. Dick, but it's very clear to me that there are lots of different versions of reality... (including the simply varied versions we occupy as individuals within any social milieu). To imply on any level that there is only one reality, or a right reality, is an extremely limited view of things. But that's my little soapbox.

How does this relate to Bob's point? We can have reality without necessarily having realism. Or I think so, anyway. And maybe one of the major limitations in virtual worlds is attempting to achieve realism in an explicit way. I'm more interested in honoring the reality of the experiences there, regardless of what metaphors or interaction devices are used, and whether they mimic the physical or not. My sense is that for what we lose in a lack of physical interaction, signals, etc. might be gained in other ways. But maybe we just haven't figured that out yet.


Mike said:

For example, I don't believe a world needs to be "realistic" in its depiction to benefit from more believable social interactions.

Yeah, that's what I mean, except kind of in reverse. I don't think a world has to be 'realistic' to have 'real' social interactions, and I think some of the most profound interactions occur within contexts that have very little physical realism, like when we type words in a chat window. It's kind of weird when you think about it; it has nothing at all to do with any metaphors for interaction in the physical world, yet it functions spectacularly as social glue.


Realism is for when you don't want players to have to think about something because the player has already internalised it. Realism means gravity takes things down, breath shows in cold climates, you get wet when you run through a stream, when you turn left the world turns right. It's things you don't notice, because you expect them. If you do notice them, then that means the world isn't working like the real world works.

Designers have to tread a line between things that are at odds with reality for reality-related reasons ("hey, that guy walked through a wall - maybe it isn't a wall at all but some kind of illusion?") and ones that are there for mere pragmatism ("if people couldn't walk through each other then you'd hardly be able to move at all").

Virtual worlds have their own physics. This maps onto real-world physics at the most basic levels, because otherwise it's too hard for players to play - they have to relearn what they've been biologically hardwired just to believe about reality. There can be additional physics - magic, for example - but these have to be consistent or again, players will be distracted too often. This means you occasionally get people complaining along the lines of "the fireball should have destroyed the map he was carrying but it didn't - it's unrealistic!" (as if fireballs are perfectly natural).

Many of the gameplay-related issues concerning the "realness" or otherwise of a virtual world are due to inconsitencies or false promises, when the players extrapolate from the don't-even-have-to-think-about-it level (rain comes from above) to the reasoned-beliefs-about-the-consequences level (rain should make fire elementals hurt).

We had this in textual worlds, of course, although the visual element of graphical worlds does make it easier for players to notice when the virtua world doesn't perform as the real world promises.



Welcome, Bob! I can't improve on the previous responses, but some additional possibilities:

1. Within your "Narrative" category, it might be worth highlighting that this is (apparently) about the game world exclusive of player characters. As I read your comments, narrative is about the expression of the world-story through non-player artifacts: artwork, sound, mob behavior, and verbal communications either from mobs to characters (as conversation) or from developers to players (as framing text).

What about the communication of world-story by players to each other? Is that what "player-generated content" is mostly about?

2. It seems to me that a couple of the subcategories in your Narrative category might be important enough to be broken out on their own: plausibility-of-behavior and plausibility-of-meaning.

Plausibility-of-behavior encompasses a lot of what people think of as mob AI. How does an animal act when it's injured or when its local food source has been consumed? What does an NPC do when a nearby faction member is attacked? More interestingly, what behaviors can an NPC express to advance its own interests? I suspect that some percentage of players who self-report as interested in "realism" have this plausibility-of-behavior in mind. They want mobs to react in plausible ways in response to stimuli.

Plausibility-of-meaning is the next level up for players who are interested in narrative -- it's about NPCs communicating why they behave the way they do. When a player interacts with a particular NPC repeatedly, and discovers that the NPC has three and only three possible forms of interaction (e.g., Stand Around, Give Quest, Run From Attacker), they may rightly conclude that the NPC is just an automaton, that although it has the form of a person, it doesn't possess the intentionality of a person. Narrative-oriented players will decry this effect as "unrealistic" -- even though NPC predictability benefits gameplay -- because they know that real people are capable of a much wider range of behaviors and expect this capability from actors that look like people.

3. Some graphical worlds provide a Level Of Detail (LOD) slider -- what if there were a Level of Realism slider?

Players who say they want high levels of realism could pull the LOR slider to the right... but what effects would that have?

Are there any players who would choose to pull the LOR slider all the way to the left? Why? And what would that do?



There's no denying that "realism" should be applied only as far as it makes a game fun. However, many times, the abstractions we use to represent real-world events are divergent enough to lead to very different complex systems.

If we want a "realistic" Tolkien-clone of a fantasy world, we might envision combat represented as a complex set of abstract rules, and it makes for an engaging game. The tactics that emerge in that game are very different from the tactics we'd use in reality, but that's fine, since it's fun, and enjoyable.

But... we also want "grand warfare" with seige gear and castles and all that sort. We KNOW what "reality" we want to see here, too... but too often, the tactics that lead to success in the personal combat scale to a very, very different kind of large-scale battle.

Seige weapons might never have a use, or may have uses very different from a developer's intention. With PC carrying capacities, there may be no need for "supply lines" and therefore, no value to infiltrating the rear. Fortifications may offer no benefit or even be a liability to the defender. Players might get more benefit out of hopping around like a bunny instead of using cover and concealment.

Now, the "less realistic" situation is fine- if the developer prepared for it. It might be tough to fit into the "narrative" of the world, but it can be a fun game.

However, we've seen plenty of instances in the past where where developers appear to have based their expectations on REALITY and not on the way the game works.


"However, we've seen plenty of instances in the past where where developers appear to have based their expectations on REALITY and not on the way the game works." -Chas

Exactly. One frustrating example would be the Battlefield series, where the developers clearly assumed people would play war as they were supposed to, with bells and whistles going full out like a war movie.

What happens is that players run around like headless chickens, spamming random radio commands, shooting their team mates, doing nothing to attack or defend in any coherent fashion, speculating in raising their personal kill score but little else.

In reality, gamers might not want to cooperate with each other to achieve goals, or even to experience the game as the developers designed it for.

Thus you get the bunnyhopping etc.

I'm not even talking about realism, it's more a matter of "does this game, this fantasy, this experience play out somehow like I would expect it to do?"

When I play a realistically LOOKING wargame, I sort of expect the players to act like in a real war, taking cover from fire, ducking, sticking together and so on.. Can't help feeling disappointed when I log on and see a bunch of complete loonies running around.

That disappointment is about broken expectations, not about realism.


Ron said:
There's so much rich engaging material, certainly in the social sphere, that virtual environments often don't tap at all.

Absolutely! I hope game companies continue to explore that social sphere instead of simply continuing to refine D&D-style cooperative combat systems.

Mike said:
WoW is more visually stylized than EQII, but I think it could benefit from some of the same social devices EQII uses...

WoW could definitely benefit from increased sophistication, if not realism, in its social interaction system. (Please let me turn my head independently of my shoulders so I can look at my party mates!) THERE is an example of another visually stylized world but it has probably the most realistic social interaction system out there today (except for its chat groups).

Lisa said:
The idea that virtual communication, relationships, etc. are somehow less real simply because they occur in a different modality has always seemed silly to me.

Yet another important use of the term "real."

Richard said:
It's things you don't notice, because you expect them. If you do notice them, then that means the world isn't working like the real world works.

I believe it was Ari Rapkin of Industrial Light & Magic (who did cloth and flesh simulation for digital Yoda) whom I heard say, "I've succeeded if the audience DOESN'T notice my work."

Ethnomethodology and conversation analysis can be helpful here in that they are concerned with exposing and understanding the "seen-but-unnoticed" features of social interaction.

Bart said:
What about the communication of world-story by players to each other?

When players display an interest in narrative consistency, which often involves realistic elements, we call it "role playing" or "RP."

I suspect that some percentage of players who self-report as interested in "realism" have this plausibility-of-behavior in mind. They want mobs to react in plausible ways in response to stimuli.

NPCs who don't seek shelter when it starts to rain are not very plausible. I think this is a very exciting frontier for design!

Chas said:
With PC carrying capacities, there may be no need for "supply lines" and therefore, no value to infiltrating the rear.

Similarly I think the reason player houses go mostly unused is that characters don't need to sleep or eat, and they can carry a garage's worth of junk with them wherever they go.


When setting up for a 3-day marathon pen-and-paper RPG, one of my oldest gaming buddies would always ask the GM, "Are we playing down to the dirt, bugs and food?" Some GMs (not me) like to model the whole kit-n-kaboodle, and if you forget to specify, "I lay down on top of my bedroll," will hit you with some kind of Wild Dillbertian Ants. If you forget to say, "We stop around sunset and eat X% of our supplies as the evening meal," they'll eventually say, "Your party has not eaten in 2 days... you all collapse from hunger."


It's a game. I don't want to itch and roll against my "Scratcing Skill." I don't want to have to stop and really sleep, eat, poop, etc. When was the last time you read a book, saw a movie, etc. where it went into any mundane details about that kind of "realism?"

But... of course... the description of the heroine's eyes? The way the sun felt on the back of the good guy's neck as he waited in the street for his nemesis to arrive? The precise details of the weapons load-out for the new class of star ship being built by the Empire of the Far Reaches... Totally vital details.

None of which has anything to do with reality/realism.

Good and great art, whether highly "realistic" or extremely interpretive (non-representational, fuzzy, modern, post-modern, Dada-esque, cubist, "un"-realistic... whatever), is effective because it works. Richard and Lisa both make the same point, essentially:

Realism is NOT reality.

That is especially true in art. Is an impressionist painting "less real" than a photograph? The painting itself, not what it depicts. Even if the feelings/thoughts engendered by highly non-representational work like that of Jackson Pollock are widely divergent... the work itself is quite real. Many non-art "things" in reality are equally as controversial, eh?

I think that trending towards "realistic = real" is very dangerous hyperespecially (word?) in games and virtual world building. It is all about the game, after all. If it works, it works, if it doesn't, it doesn't.

It's strange... I hate the "I'm typing, I'm typing, look at me I'm typing" animation in SL. I disable it whenever possible. Why? Because, for me, it looks insane to the point of blithering idiocy. It's a connection between my avie and me that I don't want. I (me, player) am typing. My avie is TALKING. Often accompanied by gestures, facial expressions, etc. And there's no keyboard. And they're not really typing; just waggling the fingers.

On the other hand, in standard chat, it's nice to see the "John is typing a message to you" message. Let's you know that something is happening. In that context, it makes sense.

Yes, it's... odd... when I can blow up a crate with my mind powers but not sit on it. But it's within the bounds of "the game is about blowing things up, not sitting." I can ignore sitting in game (which I do all day in RL) for the joy of blowing things up with my mind powers.

What are the joys of the game? Does bringing something of the "real" into the game increase them? That is the question asked of all art.

Essentially, should the game rhyme?


It seems like CoX is addressing some of these issues of believability more with the latest update--NPCs, for instance, have a wider variety of postures now; you'll find them sitting and lying around instead of standing in little clusters, although there isn't really any weather to influence their behavior (beautiful sunsets, though).

CoX also provides a very limited visual representation of certain instrumental gameplay actions, namely picking up missions: in CoV, where you can get them out of a newspaper, a character who is looking for one will be shown reading a paper; a character who "calls" a contact is shown on a little phone, I believe. The former action, however, is also available as a general emote, making it kind of ambiguous, and to my knowledge there is no animation to show that someone is taking stock of their inventory, a more common and lengthier process that you would imagine would have more ramifications for teammates if you're doing it in their presence.

Someone else noted that the inability to use facial expressions, etc. well in games is a limitation of the current technology, which I suppose it is, but even if you could roll the camera view from far-off to up-close (which you can, to some extent, now), it seems like a lot of trouble to get information that, most of the time, probably wouldn't be made available anyway--until we have electrodes mapping the player's own facial expression, anyway. CoX offers a range of emotes that you almost never see used in gameplay, although specifically social events involve them more. So there, I guess it comes back to a question of what players will actually use when it's available. The basic positioning of bodies is easy to display and see and informative to my gameplay (are my teammates following me?) but facial expressions don't really matter for my purposes anyway. Although I guess it might make teamplay more satisfying if people who did stupid things really looked contrite afterward...


The reason facial expressions and other emotes aren't used in existing MMOs is because they're not relevant to gameplay. Imagine for example that online games were only about casting spells from a distance. Games might include sword-swinging animations, but no one uses them because the gameplay isn't about fighting things up close (difficult to imagine, I know :) ). But if a new game provided gameplay where hand-to-hand combat was relevant, suddenly those animations would become a lot more meaningful.

In the same way, as long as we imagine MMOs being emotionally distant, facial expressions and the like won't matter much. But if we can create new kinds of gameplay where someone's facial expression is as significant as a cast spell or combat move, two things happen. First, the game becomes a lot more emotionally relevant (and thus inherently more meaningful), and second the game becomes accessible to a much broader range of people -- those who are perhaps more familiar with the meaning of a smile or grimace than an armor class or critical hit.


I don't think we'll see social cues and mechanisms (be they facial expressions, or something else) getting developed more when they become relevant to some part of "gameplay".

I think we'll see them develop further when game developers grow up past our current focus/obsession on gameplay and game mechanics being what games have to be focused on. Granted, many of them can and should. But a world that's primarily about socializing, not about killing dragons, is going to be the first one to hit mass market size audiences, mark my words. A lot more people want to use ICQ, AIM, MSN Messenger, Yahoo, etc. than want to play World of Warcraft.

When we grow past thinking games MUST be centered around gameplay, we can make things like Habitat, Alphaworld, Furcadia, Second Life, etc.


There have been multiple worlds primarily focused on socializing, including the ones you mentioned and others. None have taken off with the mass market (Habbo Hotel is one possible exception, but it's difficult to tell). The most commercially successful, Second Life, continues to sit at about 1% of WoW's current active user base, and others have been spectacular failures (e.g., The Sims Online). Given that, I wonder what leads you to think that such socially focused worlds will take off in the mass market? Also, none of these have, to my knowledge, made extensive use of facial expressions or other interactional social cues -- do you see this changing?

IMO, to be commerically successful, to be able to build significant social network effects (which weave the communities on which all virtual worlds depend), a virtual world must be about something other than the socializing it hopes to engender. In the physical world we see plenty of examples of this: churches, bars, gyms, bridge clubs, schools, etc., all have significant social aspects, but all have a primary focus on something else -- meanwhile "dating clubs" lag behind, continually dealing with the semi-forced nature of the social interaction because there's nothing else to focus on.

This is why I believe having "something to do" (i.e. gameplay) that is fun and relevant to social interaction is going to lead to both deeper and broader appeal in the game. We'll see.


@Mike: Remind me, then, of why the heck every race has different dance animations in WoW?


Just a guess, but I think it happened because they let the artists loose one night when everyone else on the team was out to karaoke. Or maybe because even artists have a sense of humor.


I would say Habbo Hotel is far more commercially successful than Second Life. They're one of the few MMORPGs to get a userbase up into the millions, and I believe they've been profitable for a while as well. Second Life is only just about to turn profitable, and has only just recently passed Furcadia in number of users (they estimate based on number of people who logged in during the last 90 days, we base ours on the last 30 days, so their estimates are higher). Habbo also powers www.cokestudios.com, no idea how much Coca Cola pays them for that, but I've seen it advertised at movie theatres. I'd also say a lot of Neopets is socially oriented, though they don't have a massive "world" per se - not yet, anyway. Gaia Online is a real up and comer too, and they've recently added a walkaround world with customizable homes.

Anyway, I don't feel the social worlds are competing so much with WoW, Everquest, etc. I think in a lot of ways it makes more sense to compare them with services like MySpace, Skype, ICQ, AIM, YouTube, etc. Some of those are already reaching mass market audiences, unlike WoW. As soon as somebody focuses on providing one (or several, or all) of those types of services, using an online-world type interface, I think the potential there is huge. Why wouldn't an already-popular type of service thrive if it were made even easier to use and with a lot more interaction with fellow users? Has to be done right though, therein lies the rub. If it's MORE difficult to do the basic things they like about those other services, people won't go for it. I do think "something to do" as a core for people to gather around, have something to talk about, etc. is crucial. But think bars, dance clubs, pool parlors, movie theaters - gaming doesn't have to be the only "something to do" that people will gather around. Nor is it even the biggest one.

Do I think online worlds will add computer animated facial expressions? Not too likely in my view, for the same reason typed-text chat is going to be hugely overshadowed by voice chat. (Dollar-wise, telephone service and cellphone service already blow away all forms of text chat in annual revenues. Also blows away videogames, televsion, movies, etc.) You can EASILY express your mood with your real voice and your real face. The same way you are already in real life, all the time. And believe me, when young people (or older people) fall in love online, they want to see photos and webcam video of their sweetie a lot more than they want to see some pixelated representation - even if we give the person control buttons to make that avatar wink, nod, smile, frown, etc. No comparison. First generation pioneers who loved the imagination and anonymity and roleplayability of text are leery of voice chat and probably even more so of video. But kids growing up on the internet love 'em - and I have to admit I do too. And I never found the fact that I saw my buddies around the table in real live D&D games looking and sounding distinctly non-elf-like to be any kind of obstacle to roleplaying. I was just happy to see and hear my friends. :)


I use facial-expression and gesture commands all the time: instead of saying "yes," I often *nod*; instead of saying "hi," I often *wave*; instead of saying "just kidding," I often *wink*; instead of saying "I don't know," I often *shrug*; instead of saying "thanks," I sometimes *smile*; instead of saying "gratz," I often *cheer*. To me this makes the conversation and the avatars more interesting.

The problem in my view is that players orient more to the text emotes that accompany each of these commands - e.g., "Bob winks at you" or "Bob shrugs at you" - than to the avatar animations themselves. Avatar animations (especially facial ones) are easy for their intended recipients to miss. The text emotes are more noticeable and therefore more reliable as an interactional resource; however, their use renders the avatars somewhat irrelevant. I want a gesture/facial expression system with noticeable and reliable avatar animations and NO text emotes describing those animations.

The comments to this entry are closed.