, , , ,

When commenting on my post about development cycles, Hans Jagerblitzen mentioned "the lesson to learn from Incarna". But the lesson he meant and the lesson I think needs to be learned from it are considerably different. It’s a post I wanted to write for quite a while, it’s been sitting as a draft on my drive for most of the week, I guess it’s about time.

Today, I shall tell you why, do I believe, Incarna failed, and why shelving it so completely as it did get shelved is a bad idea. (I’m told all work on post-Incarna development, including that neat sleeper site exploration idea, has now been halted, please correct me if that is wrong.) But to get to my arguments for this, I will need to start from afar.

As you might have heard, a rare neurological condition exists called ‘prosopagnosia’, the inability to recognise faces. The human brain actually has a designated region and neural structures hardwired to do face recognition, which are developed fairly early on, along with other visual processing circuits.

The ability to recognise faces is extremely important, just as the consequences of not having it are.

As I mentioned before, people have evolved to live in the world created as a product of society, full of virtual entities, between people, and other people are the most interesting objects in it. At least, they’re the only ones that actually matter. It naturally follows that telling people apart is the very first thing you have to learn doing to live in such a world.1 This ability has some interesting properties:

  1. It is naturally tailored to members of your own race, while members of other races will often cause the detection algorithm to fail until specific training is acquired. While the Western people will often think of Asian people as "indistinguishable", so will the Asians themselves think of Western people.2
  2. It is far easier, and far more natural, to think of something as ‘not human’, and therefore, outside the moral horizon, if it doesn’t have a recognisable face. This is one of the reasons why in fiction, the Evil Overlord’s minions will almost universally come with face-concealing helmets, and sometimes even the Evil Overlord himself will forgo the expressiveness of a face in favour of being more menacing.
  3. Just a face is often sufficient to anthropomorphize something that very much isn’t human, barring the effects of the uncanny valley.

It’s the key part of the subsystem that lets you tell ‘us’ from ‘everything else that might look similar’, a survival mechanism very important for primitive hominids, shaping certain properties of human culture to this day. It is quite notable that just about every virtual world ever deliberately constructed, through natural limitations, inflicts people who use it with a certain level of prosopagnosia, forcing them to replace the recognition of faces with recognition of other features, like people with developmental prosopagnosia learn to do. It is also notable that as technology evolved, world constructors did everything in their power to bring the faces back.3 From forums, which these days include avatar pictures that users can set as a standard feature, to selecting the faces of your game avatar from a set of predefined faces, to well-developed avatar constructor of Second Life, which evolved to be it’s whole point, to Eve’s Carbon, which is just about the best avatar constructor on the market today. MMOs do their best to supplement the low resolution at which faces will be visible in the actual game with other distinguishing features, but actually having a face goes almost without saying, and the more technology evolves, the more facial detail you typically get, often to the exclusion of other detail. Faces will at times have more polygons devoted to them than the rest of the body.4

It’s done for a reason, of course. People know that what they’re dealing with inside the virtual world is other people, and them having a face makes it easier to keep that in mind and internalize. But for this to work, the face has to be there. I remember reading Reddit/r/Eve a few years ago, and someone’s aptly worded question I saw: "What kind of spaceship should I be?" While it is often repeated in Eve, "You are not your ship. You are not even your pod!" the player instinctively treats the ship itself as the only body they actually get, see, and experience. Just the same, the other players, with their static avatar images that take some clicking to even get to, or only show up as tiny squares smaller than a postage stamp, are not readily seen as human, but more like inanimate obstacles. This hampers integration into the universe — it certainly does not completely stop it, as there’s no end to things a brain can eventually imagine,5 but it requires other neural circuits to take the job over, something players did not necessarily sign up for.

It is no wonder, therefore, that Incarna, out to put a human face on that senseless name you see on screen,6 was seen as a required vehicle to attract new players to Eve. It would certainly help, too, though it would probably never be as decisive as CCP expected. So why did it fail so hard?

For a very simple reason: It didn’t do what it needed to do.

An avatar in an online avatar-based environment is essentially a puppet, used to express your own state7 and supplement the verbal communication. While the verbal communication typically goes through relatively unhindered, barring intonation, nonverbal channels are missing and the capabilities of the avatar are employed to provide substitutes. Give people even the simplest avatars, and they will use them, evolving a symbolic language for the purpose — even with the most primitive avatar environments like Virtual Places people would use their relative positioning to mean things. Second Life has a system for animating the entire avatar body, but only a predefined set of avatar facial expressions — but very quickly, it was discovered that by combining those expressions, new ones are possible, which were quickly used to supplement the nonverbal language. Animation for avatars is one of the biggest industries in Second Life.8 But notice that this only works when avatars interact with other avatars — that is, people interact with other people through avatars.

Captain’s Quarters in Eve provides no opportunity to do that in any way. Your avatar puppet will forever remain visible only to you, and you can’t even invite someone into your quarters for tea with crumpets.9 Instead of supplementing interaction with new channels, it introduced a hindrance to anyone who’s playstyle depended on being able to efficiently manipulate large numbers of game objects — i.e. most veterans, particularly people who multibox — and ushered in a sense of loneliness even more pervasive than previously. Now "you" are human, but the rest of the universe bizarrely aren’t. There are no emoticons, even rudimentary, and your avatar will forever retain the blank stony facial expression, which probably doesn’t even match the clever smile you have on it’s photograph that is your actual avatar as far as the rest of the gameworld is concerned.

That was obviously not what everyone had been expecting from Incarna, which has been talked about and hyped up ever since I initially subscribed back in 2006 — it’s been on the books even longer than that. And while there’s a lot to discuss about introducing first person shooter elements into Eve,10 the key reason to have avatars in the first place — interaction through them — is still not there.

In short, Incarna failed not because it wasn’t about spaceships, but because it did not add better opportunities for interaction, essentially did not exist. Still doesn’t. At the same time, it raised numerous worries about "golden ammo" with the Noble Exchange, and seriously damaged the hopes that other, nagging problems, would one day be attended to, as future development plans were announced. It was released before it could be considered usable or useful, to much fanfare it did not actually deserve, in a large part, because of the expansion cycle prodding it on.

What did Incarna need to be successful? Not the "establishments" talked so much about, nor shooting people in corridors, no, all that could wait and there’s a lot of discussion to be had about whether they’re even needed. Not even more detailed Captain’s Quarters, which it did get eventually for some reason. Just one, very simple thing. Decontaminating the atmosphere of the station for capsuleer consumption and actually letting people in or out through this door with the red button. Even if it were just one extra person per room. Even with no emoticons, people would evolve a body language to use one way or another.

It may or may not have been technically feasible, of course. The avatar model of Carbon is actually very detailed, and video cards struggle to render even a single one, let alone a hundred that you could expect to see in a space station. I’m not a good judge of the programming intricacies this would involve, having not messed with programming 3D engines seriously since the days raycasting was new. But it was what was required, and without that the whole expansion appeared a complete and utter waste of time, money and manpower, and locking people into using this waste as the basic day-to-day interface was an insult.

Captain Obvious would say that players don’t like to be insulted, and let’s leave it at that.

As it stands now, Incarna has become a rudiment, a nagging appendix on the bowels of Eve, which can no longer be completely excised (because that’s the only avatar generation system the players have) and can no longer be actually developed, because all the people who could do that are occupied with other things, now seen as more pressing. It would be unfair to say that they aren’t pressing, but Incarna remains a monument to failure. Monuments to failure are a bad thing in general, if only because they cement the failure as such and prevent salvaging what good could come of it.

If anyone actually wants my advice, which I admit is unlikely, but who knows, here it is: Keep the plans for sleeper station exploration with avatars where they are. Keep the ideas of establishments and the first person shooter angle where they are, too, there will be a time for them yet. But do enable the Incarna code to do what it needs to do: allow a new channel for social interaction. Open that door with the red button, make a common room on stations with bland furniture and potted plants, and maybe a window to the outside space, add an emoticon that lets you smile. That’s it. That will be enough for the next year or two while you deal with other pressing issues, if anyone says ‘get back to our spaceships!’ well, tell them you’re stopping at that until further notice.

The results will be subtle. It will not fix 0.0, it will not revitalise lowsec, it will not radically increase the number of new signups, it will not fix any of the numerous other problems of Eve.11 But it will do one important thing:

It will make it easier for people to treat each other as people. From that alone, many other beneficial things will follow.

Oh, and it will also get all those people wishing for it off your back for a while. Failing that, at least turn off the Captain’s Quarters as the default new player experience environment and stop teasing new players with something they will never get — failed expectations are not good for player retention either.

  1. For quite a while I suspected I have developmental prosopagnosia myself, because talking to people through a keyboard seemed so much more natural than face to face. But testing showed my ability to identify faces is actually slightly better than average. What I do have instead is developmental social-emotional agnosia, that is, my ability to recognise the emotions of people from their facial expression is impaired. People don’t normally notice, because I have developed the ability to extract this information from text and speech patterns, which has the amusing side effects of detecting lies effectively and intuitively identifying text authorship, including detecting alts. You lose some, you win some.

  2. Similarly, Japanese people will typically require special training to tell between spoken "r" and "l", because their language does not treat them as different. The idea of a seven-colored rainbow is likewise not universal across humanity, (In fact, it apparently started with Newton — and even Newton supposedly decided on seven because he thought of an analogy to the 7-note musical scale) and there’s a promising avenue of research into how cultural development shapes the formation of certain neural structures and vice versa. "Peach is a fruit, not a colour!" Well, if you can’t see it, what’s different about you is not your eyes.

  3. The only exception would be certain environments which bypass the visual recognition entirely and go straight for text processing.

  4. Actually, take a look at the Penfield homunculus.

  5. "Virtual reality", as I explained before, actually happens in your head, not before your eyes or other sensory organs. To this day, the most immersive constructed environments I have ever experienced have been on the other end of a telnet connection, coming in as a field of text, 80 characters wide, 25 lines tall.

  6. As a side note, what’s it with player names these days? Sure, they need to be unique for many technical and social reasons, but dispensing with people’s attempts to be witty, "someguy344" is not a name people pick because they like it, it’s because they’re trained to think that the name they actually want isn’t available. Because it so often isn’t. Second Life used to have a clever system that would give you a list of last names to pick from, while letting you make up the first name yourself, but it was lost on many users for the same reason. A nicer solution is probably in order, but I can’t say I have any bright ideas at the moment.

  7. Or the state you wish to express. That’s an important point from roleplaying research. If you ever played Star Control 2, you might remember the Utwig race, who claimed that they became more socially advanced by instituting the universal practice of wearing expressive and strictly codified masks, which portrayed the things they consciously wanted to express, rather than what might be inferred from their instinctual body language, making the message clearer than it would otherwise be. That, essentially, is the most important social benefit of roleplaying in a nutshell, I kid you not.

  8. Closely overlapping with sex industry. But I digress…

  9. Interestingly, Star Trek Online had at least parts of that right off the bat. You could invite someone to your ship’s bridge, use a variety of emoticon-like expressions to supplement your communication, and much of the game involved running around space stations and bumping into people, even though there was very little call to treat them as anything other than background scenery. Let’s not forget the away missions too. I hear this capability evolved a lot since I had last been there, I suppose I should dust off my old account and go check.

  10. Which many people have been expecting, and which I think are quite extraneous, but — it’s up for discussion and I’m not going to be arguing about them here, or at least, now. What I’m arguing for is short, simple, and in my view of a sociologist, important beyond anything else about Incarna.

  11. People will need some time to evolve the conventions of using the avatars as a new non-verbal language component. That is two months or so before anything interesting happens.