Is Second Life's User Interface Cursed by Knowledge?

Among my 2008 predictions for Second Life was a somewhat bearish one: "Active monthly user base passes 750,000-- but doesn't pass 1,000,000." I should explain that more.  Because while this would still make SL among the most successful, fully 3D online worlds on the Internet, it's decidedly less optimistic than most of us would have guessed six months ago.  That was when the growth rate was still regularly moving in an upward arc; however, it's reached a plateau since then. 

Why?  Besides frequent downtimes (which are largely back-end based, and thus relatively easy to improve with server/physics/coding upgrades) everyone agrees that the crucial problem is the complex and confusing user interface.  (Pictured above in all its intimidating, rococo glory.) In plain terms, it's too damn difficult and time-consuming for a new user to figure out what's worth doing in Second Life, or how do to it.  Despite a constant stream of new accounts (12K-20K sign-ups a day, by Tateru's count), nearly all who try the world out, leave. The Lindens and third party developers like the Electric Sheep are creating new versions of the user interface, and they will surely improve matters somewhat-- but in my view, will only get so far.

Because it's not just that most new users don't understand how to enjoy Second Life; it's more fundamental than that.  The real problem is that Second Life developers don't know how to show most new users how to enjoy it.  This has to do, in my opinion, with "the curse of knowledge", a concept featured in a great New York Times article from last week.  This passage in particular made the back of my hair stand up:

“I HAVE a DVD remote control with 52 buttons on it, and every one of them is there because some engineer along the line knew how to use that button and believed I would want to use it, too... People who design products are experts cursed by their knowledge, and they can’t imagine what it’s like to be as ignorant as the rest of us.”

This also describes the Second Life user-interface, in its excess of buttons and its curse of knowledge.  Leading to the paradox of experience we have now:

To understand why SL is valuable, you generally need to struggle through a four hour+ learning curve, to even get minimally proficient at using it-- how to get around, build, discover genuine quality content, meet a handful of helpful veterans who can guide you, and so on.  Once you get past that hurdle, however, it gets increasingly difficult to imagine what it was like for you in the beginning.

So how do you fix the Second Life user interface?  In all honesty, I probably can't say.  Then again, the Lindens can't say.  Metaverse developers can't say.  Longtime Residents can't say.  And if you read this blog on a regular basis, you likely can't say, either.  Thanks to the curse of knowledge, the very people who know Second Life most are also the least qualified to introduce it to a mass audience.  (This is probably why the open source initiative and heads-up displays have failed to improve user retention-- most of the improvements and features to come out of them are made not for new users, but for established Residents.)

If I'm right, the curse of knowledge actually leads to a double paradox.  The only way the Second Life user interface will be substantially improved is when a designer with no hands-on experience with Second Life leads a program to rebuild it from scratch.  But then, if they're not already experienced with the world, why would a designer bother to invest the time and resources to do that?  Considering the Lindens' insular and technocratic corporate culture, it's unlikely they'd hire and sufficiently empower the kind of "zero-gravity thinker" mentioned in the Times piece.  Considering the immediate demands of their clients, it's unlikely that third party metaverse development companies would hire one any time soon, either.  (And all of them are largely comprised of longtime SL users, anyway.)  In the end, it may require many months or even years of slow growth until Second Life has a large enough audience where creating an intuitive Second Life viewer becomes an obvious venture.

Last year, I asked Philip Linden about the user interface, and he offered me an interesting analogy.  In the beginning, he observed, MP3 players were unwieldy and difficult to use.  It took some time for the early adopters to prove that a portable MP3 player was a viable technology, and only then did Apple come along and create the iPod which launched it into mass acceptance.  The same thing would eventually happen, he believed, with Second Life.

The thing is, even before the iPod, it was immediately easy to understand the appeal of a credit card-sized media player capable of holding tens of hours of CD-quality music.  How long, do you think, will the value of a 3D, shared, dynamic development platform become as readily obvious?

Comments

[Casius Masala]

As a user interface designer and a Second Life user I have given this situation a great deal of thought.

I would love to test some theories I have about storing more information how to do things in Second Life within the game itself. In the physical world we store information on how things work within the world - for example, we indicate how to open a door by putting the handle on the door. However, in second life my clothes are in "folders". When I select an object to stretch it, it has white blocks and not directions stretch arrows - the list goes on. For the environment to "make sense" as a "world" it needs to use the sense making approach that we use in the world.

Right now the model is some kind of Microsoft/Apple/Adobe software hybrid

The UI also needs to be more flexible. One moment I am dancing, another I am shopping, another I am building. Showing the correct level of information density at the right time and allowing the user to create their own set of custom tool pallets would be very helpful.

Of course, just adding the TP "back Button" was a huge step forward for Electric Sheep.

[CyFishy Traveler]

This is, honestly, one of the main reasons I've been crafting my little Newbie Site (clicking my name should get you there, if I do it right.) I want to make a simple resource for people who are just starting out so they can shed the noobness more quickly and get into the swim of things. There are lots of blogs and tutorials out there, but I didn't see much in the way of a unified place to explain the simple things. So I'm making one. (And, yes, I need to get more done on it--my next Tips For Noobs page I have planned is "Your Friend, the Search Function.")

[Eris]

I think it's more fundamental than a confusing UI - we're missing the most important point.

Comparisons are often made between SL and the web's journey to acceptance. There's one big difference tho' - most people USE the web, they don't BUILD the web. There's a huge cavernous difference between, say, Firefox and Dreamweaver - they're both web-related but fulfil 2 very different briefs.

The basic problem with SL's interface is that currently it's BOTH Firefox AND Dreamweaver, it's for browsing AND building. Altho' hopefully some build/creative features will always be available to the casual user of SL (with a full and expanding suite of creative features available to the committed resident!) we have to face the reality that most people want to wander/browse/shop SL, not help build it.

We need a client that allows them to do that, with great Search features and a constantly updating window on what's happening right now and how to get to it. We need a Firefox/Safari/Opera for SL and we need a better Dreamweaver - but they should probably be two separate things.

[Rusalka Writer]

The one thing that continues to drive me mad about SL is Search. Honestly, I thought it was terrible before they "fixed" it. I'm pretty familiar with SL and get frustrated enough when I can't Search and find something I know is out there. Anything I look for in the Classifieds is answered with a skin shop, or porn animations, or whatever. How horrible it must be for a newbie! How can they ever be expected to explore SL? Find products they want? Forget it. I hope Google does open a world to rival SL. I'd join in a second.

[Alberik]

I rezzed in June 2007. I'm prolly fast drifting into being too knowledgeable to contribute much. My orientation process consisted of being griefed repeatedly, getting no information from anyone on the island, overhearing one new entrant being told to fuck off by an alleged mentor, having my clothes changed involuntarily about a dozen times. Eventually someone told me to find somewhere interesting and teleport there and I was gone. That left me angry, confused and standing around in a new world without even the most basic information and my underwear on the outside of my jeans.

The way you enter Second Life is humiliating, confusing, and repulsive. Accounts in SL media by other newbies are similar, often worse. Orientation needs serious policing and is something of which every single Linden should be deeply ashamed.

The UI is a huge problem, but so is the orientation process itself.

[Domchi Underwood]

The first thought that crossed my mind when I saw SL demo as a noob (and I do have some background in UI design) was "what a terrible user interface, look at all those buttons".

But after five minutes of using it I actually found it pretty intuitive and easy to use, and as I used it more and more I actually like how I find more and more interesting options in it - although you do need a video tutorial from time to time to grasp the interface depth fully.

So in the end, I don't think the client UI is bad as everybody is saying it is. It's pretty good actually.

The one thing it needs, I think, is not intuitively obvious. It needs a bit of a face lift. That is, it needs to get rid of the ugly blue buttons, and brownish dialog background... the interface is not bad, it's the colors that stink. And I really mean it, the colors are really really bad. You don't notice it consciously, but the UI colors are perhaps the single most repelling point of the newbie experience. I mean, look at the active tab. Violet? Who thought of that?! If the UI colors had more style, the interface would tempt noobs to explore it.

Seriously.

[tooter claxton]

Here's an old page with good tips on 'creating passionate users' in communities. It's hilarious but it also makes sense. And it has the best infographics ever.

http://headrush.typepad.com/creating_passionate_users/

[followmeimthe piedpiper]

Thinking hard about this my preferences would be to leave things much as they are. It would be a shame to lose functionality in the pursuit of tidyness.
At the moment the viewing screen is pretty clear, just two narrow toolbars at the top and bottom plus maybe the camera controls.This is how it should be. Pop up windows for features are ok and only there when needed.
OK, we can mess around with the colour scheme etc but the only real moan I have is the lack of an undo button when building- unless I've missed it, lol

[Rheta Shan]

As a complete noob to interface design, but unwilling user and victim of so many interface concepts and implementations (starting from the small ecosystem of remotes that seems to have evolved in my living room, and ending with my PC OS), I would like to remark that the user interface, to me at least, is far more than the window system and « skin » used by an app. I don't think the G(raphical) in GUI is redundant : the UI as I understand it is the way the functionality of a complex system is exposed to the user (me !) for interaction.

The problem with SL'S interface is less its graphical part (yes, the buttons are ugly, but they work well ; yes, the graphical design is butt ugly, but it works according to accepted GUI conventions, i.e., uses main and context menus, the usual widgets, supports copy and paste, drag and drop — etc.). The main problem is that too much functionality is buried in unexpected parts of the interface : mouse tricks (alt+strg+shift click and pan does, oh and there's advanced walk in mouselook, and...), byzantine menus (but you can tear them off, look), modifiable and context dependent keyboard shortcuts (oh, so Enter shows the chat line, and might close it again, if you set it that way. And arrow keys work even in chat, or not, according to settings... uhm. And did you know space makes you take little steps, and breaks a fall... unless you have the chat line open, that is), additional HUD elements (setting the camera control HUD on by default would ease newbie frustration incredibly), object context menus (give calling card springs to mind ; why is such a useful function buried in the deepest evel of the pie menu of AVs ?) — I could go on. Plus a lot of things are implicit, you have to work them out for yourself even if you have read what sparse documentation there is and tried to get the point (I only cottoned to alt-clicking outside of building and modding when I finally understood that it does not only lock the camera on something, but, by implication, also locks the AV's gaze on it, thus being the key to eye to eye contact in conversation).

The problem of SL's interface is that SL itself is an incredibly powerful, lavishly stocked toolbox for virtual world activities — there is literally little you cannot do —, but that half of the screwdrivers are with the jackhammers, the powertools are buried in a stuck drawer, and the pliers are hidden in boxes labeled « old screws » (guys, don't scold me for the simile, I've never seen a toolbox more closely than at arm's length :) ). No measure of GUI glitz will help that.

[Negko deVinna]

I totally agree with Eris above.
Why is the client an application when it could be an environment of several apps?
want to search and visit? Good -> open the browser
want to chat? Great -> Launch the chat tool
Want to build and create? Perfect -> Start the creative app

Let's do it the mac way and start from what people want to do, not from what developers can make :)

[Ignatius Onomatopoeia]

The points made above are convincing, though as a LL mentor who works an orientation island regularly, I can tell you that noobs' expectations are either too high (b/c much of OI is silly) or too impatient to "get to the real SL."

I don't play online games. What sorts of learning curve/mentorship exists in WoW or others? From the old days of paper-and-pencil gaming from Avalon Hill (I'm that ancient) an approach that worked well was sequential rules. One could play a game's early scenarios with a few rules and features and, as one moved to later scenarios, add complexity.

LL is putting resources and thinking into the volunteer mentor program, but Orientation Island is still not good. Help Island is better; there is more to do there (such as its sandbox or photo booth) that do not seem like a bunch of trivial, yet seemingly mandatory, tasks. Yet Help Island demonstrates some important skills.

Let me add one plea--from a person who could not design a user interface were his life on the line.

For this resident, one issue that plagues SL, and will continue to plague it, is the lack of out-of-world INV backup. For even the moderately experienced user, that first purchase is a big deal. Then, to cite an example from yesterday, LL tells us "um, don't rez anything now while we fix things."

I know, I know: how would we do this for no-copy items? There must be a way that would protect creators' IP. But if LL wants SL to be a business platform, they'd best consider this issue, given the buggy state of inventory and Linden balances.

[Cyn Vandeverre]

13 months ago, I started SL, through one of the standard orientation islands. My experience with computer games was nearly none; MYST and its progeny only.

I didn't find it difficult at all, except for two things:

1: there was someone there at the rez point shoving everyone off the cliff, so I AR'd them.

2: I didn't learn about mouselook or how to change the camera/focus, so there were many signs on the main grid that I never saw -- anything above two meters or so. When I build, these days, I make sure there are signs that can be seen without camera shenanigans.

Now, thinking back about this, I couldn't say why other people don't have the easy time that I did. I think I spent about 2 hours on orientation/help island (whatever it was called) before TP'ing to the main grid. I read everything and did the exercises. The interface didn't bother me.

[larryr]

hire design experts to make the products, not blog "experts"...;)

Fixing the SL interface for mass usage is not hard. Wanting to admit that others not "trained" in the "programming" tech myths of SV are.:)

Very few tech driven companies ever figure this out.

usaully its too late. The times arcticle sounds like its quoting Don Norman again, since he was repeated on CBS 20 min, saying the same words. He represents a group of professionals that spend careers in user design issues for product design. So many of them are not "invited" into the Tech software development process in general. But this is an old story, one ive "lived" for 25 years.;)

[T_S_Kimball]

One thing that may help is to have OI run 'locally' (and therefore privately) before logging in. That would give the user time to figure out the controls without interruptions (something I was able to do only because SL was not getting swarms of logins back in July 2004 :-).

With that stepping stone, from there the UI can be introduced in a more expanded manner, much like application tutorials today. This has some technical limitations - primarily the need to run these simulator(s) locally - but my understanding is that OI's 'cached data' is downloaded as a permanent part of the client anyway.

The above post regarding the 'mashup' feel of the UI has a valid point. It should be standardized in *some* format - It don't matter which one, but 'icon' bars (that change based on context) used by applications like PaintShop Pro would be a good example. Admittedly, M$ did something right there. ;)

--TSK

[Faerie]

..."setting the camera control HUD on by default would ease newbie frustration incredibly"...

In letters of fire 20m high, tell the Lindens that:

SETTING THE CAMERA CONTROL HUD ON BY DEFAULT WOULD EASE NEWBIE FRUSTRALTION INCREDIBLY.

So simple.

So important.

[Patric]

You point out that it's next-to-impossible for an outsider to show up and be embraced by Linden Lab long enough to start from scratch with an easier-to-use interface. But what if this outsider was in the form of a competing world--like a Google version of a virtual universe? Then the Lindens would notice a comparable world, and its different improvements, with fresh eyes. And perhaps LL would 'borrow' some of these improvements, consciously or unconsciously, in its next release. And likewise, the Google Universe would evolve side-by-side with other universes, and steadily improve its own 'look'. I think the only way things will get better, as in all of nature, is with competing technologies.

[Nexii Malthus]

Indeed the user interface is a hurdle but it should not mean that LL should start immediately stripping out all the features. Instead a seperate viewer should be created designed to be simplified as just an effective browser and nothing more. Just as Eris mentioned before about creating the firefox and dreamweaver seperate equivilants.

The problem with the introduction of another virtual world is that seperation would be a dangerous thing at such an early stage and would require heavy cooperation between LL and the competing party to create the protocols and standards to let users freely move between the worlds, otherwise the seperation between virtual worlds would cause cultures to divide and the great awesome advantage of virtual worlds disappear for a long time in a clash of idiocy and ignorance.

[Rheta Shan]

I just wanted to let you know that I have expanded on my comment above in this post on my blog.

[annetheresa]

Here's a new user's perspective: I first logged on to SL last spring, thinking I'd like to find out what all the hype was about; but I became intimidated in Orientation when I couldn't figure out how to drop the torch I'd been instructed to pick up, and didn't fancy wandering around looking like an idiot! Plus I kept getting propositioned by other avatars (that's another problem that someone should be thinking about solving; nothing is more off-putting to a newbie, let me tell you). I didn't have the time or energy to go back into SL and try to figure things out till just recently. But now, the program doesn't work; I guess my laptop's specs became outdated in the interim.

Grrr! It's all too complicated, and I'm not gonna buy a new graphics card just to be frustrated further. But I'd like to have learned it.

[Tina]

I first logged into Second Life in September of '08, And I have to completely agree with Eris, if it was not for meeting some key people I would have been completely intimidated by search alone, forget the interface. Felt like I was dropped in the middle of nowhere. Trying to look for a 'green dot'(or person) on the map for them to explain what in the world this thing was. I personally came for rl reasons(college) and they really didn't do an awesome job of explaining what sl really was. And totally didn't prepare you for EVERYTHING sl is. So yes, the interface is not that great, but more than that the overall first impression of sl is very very important. I would be careful of who I'd recommend to sl if I didn't know I would be right there to guide them around the porn images, insanely detailed skin shops and other things that would completely scare newbies away. I don't know how it used to be on sl(hear its much better now) But I'm looking at what is here now, and it needs a lot more work. That's coming from a still newb person.

[rocco]

The SL user interface is absolutely disgusting. Its like being presented with the entire contents of your hard drive in a random order. I tried to use SL for a couple of hours and gave up in total frustration. I tried shopping and couldnt even put on an outfit. (Apparently you have to go to a special place before you can open the bag which contains your newly bought outfit (huh?!!wtf?))None of the navigation menus makes any sense at all and all the help topics are completely irrelevent. Why are 3d virtual worlds always designed by morons with no design abilities whatsoever? Whats wrong with a big tab saying "outfit" and another saying "friends". Why does the inventory have to be in stupid folders that if you expand stretches to half a kilometre long? The names of any items are like 45 characters long - whose dumb idea was that? SL will never be a mainstream application until entry level users can use it intuitively. Virtual worlds should be scooping up users on the same scale as facebook but the lack of design ability by the creators of 3d virtual worlds remains their single greatest weakness. Any noob casually browsing SL will just shake their heads in bemusement at how anyone could possibly think this was a good way to navigate around an app. Creating an intuitive interface is not that hard, but SL is trying to use the interface from the space shuttle to do something as simple as ride a bike. Needs taking back to basics and reworked from scratch.