Situational Awareness

Regarding the display and representation of information, Ben Shneiderman's famous mantra "(o)verview first, zoom and filter, then details-on-demand" (OFZaFtD-o-D) may seem unambiguous.   Most of us likely intuit that we consume the global structure of a scene first then move onto its details.  Yet it worthwhile to remember that it wasn't until the digital age that information layering became a practical option in many cases - recall the visually stunning chrome dials of automobiles of the days of yore. 

However, the OFZaFtD-o-D is an incomplete answer.  The follow-on question becomes what is the right organizing global structure to use?   The virtual world interface provides us with an opportunity for an interesting thought experiment...

By and large the disagreements in design of virtual world user interfaces (e.g. to emphasize a Head's-Up-Display or not) center on the best way to communicate to the player an awareness of her situation in that world.  Few would likely quarrel with the need.  The arguments start when trying to decide what is the right global situational structure from which to hang the rest of the game world representation.  Desired is a form that reflects the priority of the player (and implicitly the world/game design).

In some ways  recent discussion on the role of the avatar (Dead Metaphor Walking) and the Non-Linearity of Chat (chat display) are proxies for a different point:  what is the best way to represent the the MMOG universe to a player (OFZaFtD-o-D)  given the limited resources of screen real-estate, player interaction and memory?    In the Cockpit of your Avatar  these trade-offs were suggested using a more theatrical turn:  if you can't fly your avatar, you surely can't play it.

Recently P&V posted "Visualizing the overall structure of a tennis match," reviving a topic from a 1997 article [1.].  The graphic is worth a thousand words, take a look at it there.  The goal is simple enough - as  a tennis game has a hierarchical structure to its play that is not immediately clear to noobs, why not invent a graphic that communicates it more effectively?  In this case a TreeMap.

Now, there are a number of easy arguments one can throw at this were it the display of a hypothetical eTennis.  Starting with a.) how immersive would this be  b.) wouldn't a strategic view  transform a tactical game into a different game; c.)...

Yet, stepping back.   I wonder if in this age of rapidly approaching ubiquity in game world  access (e.g. "Towards Situation-aware Cross-platform Ubi-Game" [2.]) when even our good friend Alice Taylor speculates (AULA presentation) whether Google Earth is going MMO, there may come the need to freshly reevaluate how to design the virtual world experience (interface and all) to effectively communicate an efficient and engaging situational awareness.

New paradigm anyone?

-----------------------------------------------

[1.]  Liqun Jin; Banks, D.G. (1997): TennisViewer: a browser for competition trees, IEEE Computer Graphics and Applications, 17 (4), 63 - 65

[2.] JungHyun Han, Hoh Peter In, Jong-Sik Woo, "Towards Situation-Aware Cross-Platform Ubi-Game Development," apsec, pp. 734-735,  11th Asia-Pacific Software Engineering Conference (APSEC'04),  2004.

Comments

[Bart Stewart]

1.

Great question, Nate.

I started programming back in the early 80s when "structured programming" was still going to save programmers from themselves. One of the components of this model was "top-down programming."

Through experience I learned that neither top-down programming nor its bottom-up predecessor worked well for me. Both approaches missed important things. So I worked out my own approach to design, which I naively called "inside-out" design. Rather than starting either at the top or the bottom, I would start from the middle, seeing the whole system as a skeleton of core systems branching out into reusable subsystems.

Another way of putting it is that I learned to "flatten" the map of systems. Looking at either the top or the bottom hid necessary information -- by exposing systems from the middle outward, I could see all the key system relationships at once. (A key here for me was encountering John Gall's remarkable book Systemantics, which for its exploration of both theory and practice remains one of the most useful works on systems design I've ever read.)

So is OFZaFtD-o-D like top-down systems design? Or is it more like inside-out design?

Later on I discovered that although I personally was more comfortable with this systems-oriented approach to design, there were some people who still preferred top-down design... but there were even more people for whom bottom-up is simply the only way they are capable of thinking.

In the group of developers I manage, for example, there's one who regularly misses things because she "chunks" systems. Her code usually accomplishes what the customer really needs, and fits well with the rest of the system, and gets done in a timely fashion. But there's almost always some little piece that doesn't work, causing me to have to budget more time for testing her code than that of other developers.

But then there's the other developer who is constitutionally incapable of thinking in any way other than details-first. He is superb at collecting requirements, and at insuring that no requirement is missed in implementation... but his stuff takes forever to write, it often duplicates existing subroutines, and it doesn't always integrate well with existing code. Worst of all, it sometimes has to be substantially rewritten later because it didn't question what the customer asked for versus what they actually needed. So I have to allocate time to providing high-level organization for this developer's work.

No amount of direction or assistance from me has changed these behaviors. These are simply the ways their minds work, and the best I can do is find tasks for them that fit their styles.

All of which is an excessively long preface to my agreeing enthusiastically with the observation that "[d]esired is a form that reflects the priority of the player (and implicitly the world/game design)." This fits precisely what I've come to believe about MMORPG design, which is that in too many cases it's not player-centric enough. Or, more accurately, that features and interfaces are designed for only one type of person, that being the detail-oriented kind of person. If you're not that kind of person, sorry, you don't get a game.

Many people think in terms of concrete, external details. These folks want interfaces that let them interact directly and reliably with tangible elements of the game. They like the "radar"; they like having the numbers of the RPG exposed; they like seeing information about a potential target (either floating over its head or detailed in a pop-up window, or both). They like things that can be held, owned, possessed -- weapons, armor, tools, clothing, houses, furnishings, decorations -- and they want to be able to control the appearance, positioning, usage, and stats of all these things. Oh, and they want all of that Right Now. Like the jet fighter pilot, they value constant situational awareness. Clearly most MMORPGs are designed to satisfy people with this worldview.

But not all gamers see the world that way. Some of them naturally have top-down minds. They prefer not to see the nuts-and-bolts details; to expose that information breaks the illusion of being in a coherent, dynamic world-story. These gamers prefer surprise over certainty; they enjoy realizing how systems fit within systems to create a coherent whole; they like stories-within-stories that unfold over time; they want "ecological" AI rather than farmable loot bags; they prefer reading layerable maps over following magic lines; and by a large margin they want tools to build their own objects and stories instead of just taking whatever they're given. (If it's hard to provide examples of top-down game features and interfaces, that's because few MMORPGs offer those kinds of things!)

So where are the virtual worlds that consider both of these ways of looking at the (game)world to be valid? What MMORPGs support both top-down and bottom-up thinking in the design of both gameplay features and user interfaces?

Is such a thing even possible?

--Bart

[monkeysan]

2.

Instead of designing the means for 'situational awareness' into virtual world clients, the future is clearly to let the user build her own avenues for situational awareness, a la WoW (sorry to bring up WoW, but this is one of several areas in which they have trounced all comers).

This is really the only way to proceed as virtual worlds get increasingly open-ended, functionally demanding, diverse in possible user roles, socioeconomically complicated etc.

The new mantra for VW interface design should be built around providing users toolkits and standardized ways to hook into game data rather than designing the interfaces for them.

[Andy Havens]

3.

(Where's that damned soapbox... it's around here... oww! CRAP... toe... friggin...)

We always search for GOD. GOD = Goal. Oriented. Design.

A VW/MMO with as much flexibility and as many tools and meta-tools and user buildability as the real world would be... uh... the real world. Plus wings and furries. Which might be fun, but would it be a "game" per se?

Here's my thought experiment:

Let us imagine that at some point in the future we WILL be able to reproduce any virtual reality in total and complete faithfulness through various technical interfaces. That it is 100% realistic, that it encompasses all senses. That it is entirely possible to do ANYTHING in this space, among as many players simultaneously as possible as is wanted by the designer(s), with no limits on objects, users, internal space, numbers of inputs, etc. In essence, any/all game/world spaces a la the Metaverse, but total and complete, virtuality without end, amen. You can taste it, swim it, hump it, sleep it, etc. No real physical changes occur when you exit "the box" and turn it off... but while you're plugged in, the programming can make it seem like anything is happening to you. Absolutely anything....

1. Marshall McLuhan argued that before the time of the written word, we lived in an "aural" space rather than a visual one. That we were much less linear. That text has made us "hyper-specialists." That we are now reliant on our sense of sight, and that, in that distant past, our sense of sound was much more important. In the "perfect VW," could we have a "hyper aural" interface? One that is totally without visuals? Why would you? What would it accomplish? What would we learn?

2. Web sites were originally created pretty much by programming geeks; at least for the first few years. Thus, the tech and content was dominated by geeks. Even when the tools came "down" in complexity, the content was still mediated by geeks. OK... so there was this sphincter, and it only allowed so much content to get through. Then blogs came along, and now everyone can get their word on the Web much, much easier. The barrier to entry is just that much lower. What will game VW/game design look like on that "perfect VW" day when designing in that space is as easy as blogging? When artists and children can say, "This is my idea for a world?"

3. What things do we do now that aren't games that might be "game-y?" The idea of making games out of all the Sim Crap that's been Simmed over the past decade would have, at one time, been thought ludicrous. But the possibilities of using those energies to maybe hook up actual problem solving (the distributed SETI project come to mind) to hook up what *I* think of gaming with what *you* think of as drudgery... what does that interface look like?

It's all about the goals. Myst had an interface like no other game before it. As did Black & White. As did Populous. Most sports games have interfaces that are much different than those presented in MMOs. SL doesn't have a HUD, per se, unless you choose to turn it on, and most players I talk with do not, except for very specific activities. Map goes on, map goes off. The chat window is the only tool that mars the view.

What do you want to do? What do you want the game to do? Asking about an "efficient and engaging situational awareness" depends so much on the "situation" and what you're trying to be "efficient" at... right? Not killing anything today? No need for an enemy hit-point meter. Or, if there's a way to see that data in their visage, then there's no need. Not going to ever die yourself? No need for a health meter for yourself. No magic? No manna bar.

I'm of course being a bit simplistic with those examples... but up above there, I went way out of my way to be way out. The point is... what is the GOAL of any game or VW? The goals should dictate the design. The design of the interface, the rules, the size of shards, the pricing, allowance of RMT, social aspects, etc.

You wanna play in the dark? Then you turn off the lights.

[monkeysan]

4.

I agree with Andy, but I'd substitute GOALS for POSSIBILITIES. I think the future of VWs will make it much harder to advance the art of virtual world construction if the focus is on 'the goal' as opposed to recognizing that the most interesting and engaging play arises from users exploration of what a world allows rather than what that world's creators direct them toward.

[nate combs]

5.

monkeysan>
Instead of designing the means for 'situational awareness' into virtual world clients, the future is clearly to let the user build her own avenues for situational awareness, a la WoW

-----------------

To a limited point yes. But to my view - WoW add-ons (for e.g.) explore the space of a fairly limited and well-established genre interface pattern: 3d view + couple of camera views, chat box, group meters, personal (character) meters, a tray of verbs, minimap...

The reason why add-ons cannot lead a broad solution here is hinted by Bart: we're still talking about a designed system, albeit one perhaps conceptualized differently by the different participants.

There can be many customized/individualized views into the "model" but ultimately the model is constraining of the surface choices. A model can be flexible, but its flexibility is designed.

[monkeysan]

6.

nate: "There can be many customized/individualized views into the "model" but ultimately the model is constraining of the surface choices. A model can be flexible, but its flexibility is designed."

no question.

[monkeysan]

7.

I'd only add that the fact a model is designed does not entail anything about whether it constrains users to achieving GOALS (unless it's FIFA). And while the model constrains the surface of choices, as you say, it is not sufficient to describe the entire surface of choices. Questions of interpretation, usage, etc. contribute as well.

[Andy Havens]

8.

I don't mean "goals" (hopefully yer bein' funny) in that limited sense... and I don't mean the goals of just the players.

And possibilities are fine... but if you only define possibilities, then "failure" will always be on your radar as something that can happen with a shrug, a toss of a cigarette butt and, "let's move on to the next set of possibilities."

Goal definition is incredibly important in good marketing (my line of work) because if you don't know where you want to go, you don't get there except by accident. I'm not suggesting that we constrain what users can do necessarily... unless constraint be one of the goals or sub-goals necessary to accomplish the overal strategy of the game/publisher/universe.

Where so many projects (of any type) fall apart, is that they have divergent or conflicting goals from the get-go. "I want to be in good physical shape," is one goal. "I want to eat nothing but fried foods and lots of 'em," is another. Those two goals CAN live together... but they take a lot of work not to be mutually exclusive.

The same holds true in games. We've all learned over the course of having played hundreds of games, that certain game mechanics don't work well for certain types of games. Is it good to challenge this history? Hell yes. But is it good to throw it out or ignore it? No. If we know that people really, really are irritated when you take away the option to reverse the pitch control in an FPS if they want to... why force them into doing it another way? If we know that a radar map that spins with the facing of the character is preferred 7-to-1 to one that is fixed... why even bother to do it the other way?

What we're talking about here might be termed "the poetry of gaming." In written poetry -- good poetry, anyway -- you are required to have a reason for every word choice, rhyme, meter and line break. The point is to try to drive specific feelings, thoughts or chain of ideas. It's never random, although it may seem that way. A great poem is like a very precise model of an inner, ideal moment. Poets spend hours (and hours...) fighter themselves and each other over word choices designed to ellicit a particular response; designed to match and achieve the *goals* of their pieces.

The same can be said of great games; every design choice that does not further the "goals" of the game is a problem and should be edited. I almost always turn off in-game music. Why? Because I like music. Most game music ads nothing to the experience. Adding nothing, I find that it actually subtracts from the experience.

So rather than just adding features, design elements, music, art, story, explosions, particles -- and, yes, even possibilities -- to a game/vw, I'd first ask, "What is it that this does for the experience? How does it bring the user/player closer to the Final Place where he/she says, 'If you take this from me, I will bite you?'"