Defining games (the cheapo way)

Many books on game design have a chapter, usually early on, that wrestles with putting a definition to the term “game”.  Since that’s something for which we all have a pretty intuitive sense, it’s surprising how broadly our definitions of it can diverge.  Try it!  You’ll find it’s pretty difficult to come up with that ideal string of words that are true for everything we call a game, but which also clearly exclude those things that aren’t games.  For example, you might say:

  • “A game is a fun activity.”  Hmm, well I’ve been to parties that were fun activities but that weren’t games.  I’ve also played some games that weren’t fun.  When I was in fourth grade all of the boys in my class would spend recess simulating pro-wrestling matches, which I personally found to be just plain painful.  But I’d have to admit that the shortage of fun didn’t stop it from being a game.
  • “A game is a rules-based form of play.”  It’s certainly true that all games have rules, no argument there.  But so do computer programming languages, highways, and sessions of Congress.
  • “A game is a frivolous diversion from the real world.”  No, that can’t be right.  Militaries stage games to simulate conditions of war, which is about as far as you can get from a frivolous pursuit.  A blackjack table is a game, but since the players are putting up real money it can have very tangible impact in the real world.

I think the difficulty stems from impulse to tackle the problem using a straighforward Webster’s / OED approach, which only works until you find one example to the contrary.  I vow never to try to do that.  Instead, it’s a little easier to describe the characteristics that, taken together, comprise a gameplay experience (sort of a cheapo approach).  In the past, I’ve found some success with these three characteristics of all games:

  1. Static objectives. One or more explicit, measurable conditions that all players are trying to reach.
  2. Environmental constraints. The things and places that enable play.  Think of cards, dice, checkerboards, and football fields.  These set hard limits on what people can do: a deck of cards only has four aces, no matter how much you might need a fifth one.
  3. Formal constraints. AKA, the rules.  These are the intangible limits on what people can do.  There’s nothing that keeps the players following these constraints, except for the fact that they all agreed they would.

That’s it.  Those three things are true of any game under the sun.  Also, anything where those three characteristics are present must necessarily be a game.  You’ll notice that makes it a pretty expansive way of thinking about games, and the characteristics could easily encompass things we wouldn’t normally identify as games.  Education, financial planning, and even work would be caught in a net that wide.  That’s by design!  I really believe that many mundane, everyday experiences can be understood as games, even if we’re not used to thinking of them that way.  And in turn, they can benefit from the elements of design that make games compelling and enjoyable.

Jesse Schell (who wrote a fantastic book called “The Art of Game Design”) gave a very future-looking presentation at the DICE Summit in Las Vegas last week, where he suggested integrating game design into the littlest things people do every day.  Brushing your teeth.  Eating breakfast cereal.  Riding the bus.  Reading a book.  He suggests that all of these things can be detected through sensors and engineered to earn you points, achievements, or tax credits.  Absurd?  You bet.  Schell’s deliberately overshooting the mark to invite us to stretch our imaginations beyond the traditional, limiting definitions of “game”.  Somewhere short of remote toothbrush surveillance is a much more compelling way to do Quicken, Outlook, or Craigslist.

If Schell’s proposals seem absurd, it’s more because we’re unaccustomed to them than because of any real-world barriers to actually bringing them to fruition.

Hunting for coupons at Old Navy

Game experiences are slowly creeping into regular old user interfaces.  A great example is OldNavyWeekly.com.  On the surface level it serves as a normal circular, announcing big sales on the hip clothes.  But it also prominently prompts users to “Click around to find hidden in-store coupons”, playing as a kind of easter egg hunt.

In some cases you need to drag items from one outfit to another, like a pair of shoes or a heart imprint.  In others you need to click a character that appears briefly onscreen, like a baby chick who periodically runs out of a basket (and it’s actually really hard to catch the little bastard).  For each find, you’re rewarded with a coupon to print and bring to the store — but you can only keep one, so in each case you need to decide whether it’s a better deal than the one you already have.

Now of course this goes against the must fundamental, ingrained tenets of usable design, like making sure that users can easily find the things they want to find.  So can this really be a good thing?  I’d suggest that there may actually be a few ways that a game such as this one can be helpful to the retailer:

It couples the coupons with a sense of achievement. You had to invest effort and ingenuity to get that coupon, dammit, and that investment won’t be fulfilled until you make use of it.  If you don’t, then the time you spent working on it could only be seen as time wasted.  The more difficult the challenge, the greater the sense of obligation.

It commands greater attention. Games require active, participative engagement in the experience.  Since anything onscreen could be a trigger, the user has to pay more careful attention everything.  Eyeballs are good, but attentive eyeballs are much more likely to respond to ads.  The game also also encourages repeated visits as it’s redone each week.

It encourages free peer-to-peer advertising. You also have the option to gift one coupon to a friend via Facebook.  That’s great for Old Navy, because it comes with an implicit endorsement from a trusted friend.  If it was worth sending, then the person receiving it must read it as saying “This is a great deal, you should check it out”.

It invites users to think of themselves differently. Web users are often cast like the audience of TV or magazines, who use or consume information at the end of its journey and after it’s fully formed.  People playing a game, on the other hand, join in making the experience.  This invites users to think of themselves as belonging to in-group, with a role to play as a part of the Old Navy brand.

I’m going to try to make contact with the site’s designers, to ask them about the intention underlying the game approach and how well it’s worked for them.  If you have any questions you’d like to me to ask them, please feel free to add comments to this post.

Interview with Stone Librande, Lead Designer at Maxis

I recently spoke with Stone Librande, who has worked as a designer on games including “Spore” and “Diablo III”.  Stone also leads an annual design workshop at the Game Developers’ Conference and teaches a college course in game design.  We discussed game design process, including a method of paper prototyping that UX designers will find both familiar in concept and new in execution.

Q: Tell me a bit about your background.

A: Before I went into gaming I was actually doing a lot of work in user interface design.  I commercialized a technology that parameterized artwork and allowed users to quickly sift through thousands of drawings just by pulling sliders mapped to different characteristics.  We found a lot of video game applications for it.  Then I took a job managing a Web design team at a company called MPlayer, which was a social gaming network that was a little bit ahead of its time.  But looking all the way back to my childhood, game design was always something I was interested in.  Eventually I worked my way into Blizzard and from there on to Maxis to work on Spore.

Q: How was Spore’s game experience created?

A: Well there were really two pieces to that.  First, there was a high-level description from Will Wright.  In one case, we were asked to make a game about cells swimming in a drop of water.  Then there’s the bottom-up design of the game mechanics.  An important consideration in the cell game was creating the right balance of risk and reward.  In any game you don’t want it to either be too hard (which would become frustrating) or too easy (which would make the game boring).  But everyone’s different and we wanted Spore to have broad appeal to both casual players and hardcore gamers.  The question is: How do you make an experience to fit many different tastes?

One way we approached that was by giving players opportunities to outfit their cell creatures with different pieces as they evolve.  Novice players can finish the whole cell experience with just the basic creature design.  You can get by while taking very modest risks, but you also won’t reap great rewards from it.  But for hardcore players, there’s an opportunity to really dig into the game by experimenting with the effects of different pieces.  They’re invited to take a lot more risks, and they put themselves in more danger of failing.  Since the traits they pick up in the cell game effect the later stages, those players who take on a greater challenge can also put themselves at an advantage and realize a greater reward.

Q: How do you guide players’ behavior in games?

A: A lot of those ideas you learned in Psych 101 like reinforcement schedules are fundamental to game design.  People are subject to the same behavioral influences as pigeons and rats.  You can influence the players’ behavior by attaching a meaningful reward to the actions you want them to take.  For example, say you’re designing a card game and you want players to try to collect three 3’s.  You could force them to do that by making it the winning condition — there’s your reward.  Or you can make people pursue that same goal less aggressively by saying that three 3’s are worth 3 points, while all other collections of cards are worth one point.

The most powerful reward you can give a player is a social reward.  Intrinsic rewards are nice, but adding in a social component exploits people’s basic competitive nature.  If someone else has something that you don’t have, you’ll work really hard to obtain it.  There’s also a element of inclusion, of being part of an in-group that’s tied together by the game experience. 
 
Q: You gave a presentation at last year’s Game Developers’ Conference about paper prototyping.  Tell me about how your method works.

A: First of all, the paper prototype is not a representation of the actual game, and it’s not intended to be.  That’s not the purpose.  Instead, the point is to ask and answer one simple question about the game you’re working on.  Second, it should be something that you can experiment with and iterate very quickly. 

Game board with pieces

This is the prototype game used to design one aspect of Spore's cell game. Click to view full size.

So for Spore’s cell game, a key design question was figuring out the various creature parts that would be available to the player, and how they balance against one another.  So I put together a board game version on paper.  I wrote up a large list of parts and their abilities, going big at first so we could test a lot of scenarios and then scale it back.  Players would assemble a unique cell creature using different combinations of eyes, mouths, graspers and tails. The cell pieces have different game abilities. For instance, tails allow the cell to move forward and rotate. During the game, each cell would either attempt to eat the most green food tokens (herbivore victory) or to attack and kill the opposing cell (carnivore victory).

Chart of creatures

Output from the prototype, used in designing the actual game.

We ended up with 12 parts that were given away over the course of the cell game’s five stages.  We also defined the other creatures you’d encounter in each of those stages, ranging from harmless to more difficult as the player progressed through the game.  That output was what made it into the final game.

Q: Why do this on paper, when you could model thousands of different scenarios in one go using a computer?

A: I run a workshop teaching this technique at the Game Developer’s Conference, and computers aren’t even allowed into the session.  Building  prototypes with paper fosters team interaction.  As people work on it, they’ll start role-playing and getting into the characters of the game.  They also develop a shared vocabulary for discussing elements of the game.  If you did it with computers, everyone would just be working on their own and you wouldn’t get that kind of interaction.

Q: What works best prototyped on paper?

A: You can’t represent the full gameplay experience, that’s just not practical.  A video game like Spore has a lot of physics and math, and that just can’t be done on paper.  Input controllers like mice or keyboards are also really difficult to simulate.  Anything that’s too complex would just be misery to test.  Similarly, if a user interface designer were prototyping the front end for a database, you could show what the form elements and buttons look like but you couldn’t simulate the return of actual data.  That’s just too complicated to do.

That said, when you really abstract a design problem there’s a lot that you can pull into a non-electronic prototype.  In my workshop, I do an exercise where I have people build prototypes of existing video games.  A few years ago one team decided to try doing Rock Band, and I was really skeptical that it would work.  Surprisingly, they came up with a game that captured Rock Band’s core mechanics.  There were five players, one of whom had a shuffled deck of colored index cards.  He would throw out the cards in sequence, and all of the other players had to dig through their own cards and throw down matching colors.  When you matched the pattern, the moderator would give you coins.  If you missed, he would take coins away.  Players could support one another by throwing coins to band members who were missing their beats.  Even though there was no music and there were no plastic instruments, the game really captured the Rock Band feel.

This is a really amazing method.  Thanks so much for taking the time to talk, Stone.

Interview with Mike Ambinder of Valve Software

Valve Software has designed top-selling games including Left 4 Dead, Half-Life, and Team Fortress.  I recently spoke with Mike Ambinder, PhD, the company’s full-time experimental psychologist, to discuss the professional practices that ensure high-quality game experiences.

Q: What’s your role at Valve?
A: My job is to apply knowledge and methodologies from psychology to game design.  That means performing statistical analyses, developing playtesting methodologies, conducting  design experiments, a little bit of interface design, and investigating alternative hardware among other things.

Q: How can psychology guide game design?

A: Well for example, in the Left 4 Dead series there are several predetermined locations in the game called “drop points” where health items or weapons will spontaneously appear.  To decide what’s dropped, where, and when we considered reward and reinforcement schedules, which are elements of behavioral psychology.  You can put things on a fixed schedule so that they’ll appear at regular intervals.  This makes the gameplay experience more predictable, and there can be real value in that.  Or you can use a variable schedule so that you don’t know what’s going to show up or when it’ll pop in.  Variable schedules can create a higher rate of engagement in the game and make the experience more enjoyable as uncertainty of occurrence can increase arousal.  A large component of the gameplay in the Left 4 Dead series is the use of these variable reinforcement schedules.

Q: How is testing integrated into the design process?
A: We’re constantly playtesting.  Our philosophy is to playtest as much as possible, and to start it as soon as we have a playable prototype.  Of course our designers are experienced and generally make good decisions about gameplay, but we don’t want to just assume we’ve got it right.  Game designs are hypotheses, and every instance of play is an experiment.

Q: What’s your standard testing method?

A: We use a variety of methods, but the most favored is direct observation of real players working their way through the game.  I’m not a fan of the think-aloud protocol, in part because the constant prompting detracts from the gameplay experience and can introduce inadvertent bias, and in part because people can be really bad at explaining why they do what they do.  Better to just sit back, watch, say nothing, and try to understand the player’s actions.  So quiet, direct observation is our preferred method, but we combine that with player Q&As, surveys, quantitative metrics, eyetracking, and design experiments, and we’re investigating methods of measuring the player’s emotional experience during gameplay.

Q: How can eyetracking help to inform game design?
A: Generally, you want to eliminate frequent long eye movements because they lead to fatigue.  For example, if the area map is in the bottom right corner of the display and your progress through that map is shown in the upper left, you’ll see the player’s eyes transiting the screen a lot.  The proximity compatibility principle tells us that things which are mentally proximal should also be physically proximal, and eyetracking can tell us which things are mentally proximal.  By arranging related information together, you can reduce fatigue and make the interface more efficient to use.

Q: And how can you measure the emotional experience of gameplay?

A: This is still early on, but we’re looking at biometric methods like EEGs which measure brainwaves, and EMGs which measure the electrical activity of muscles.  But there are questions of their cost and efficacy.  They’re also both very intrusive methods, requiring either a cap that’s wired into a machine or electrodes attached to the face.  In testing you want to mimic the home experience as much as possible, and EEGs and EMGs both make it feel more like a lab environment.  But new technologies are emerging that could change that.  Remote detection of facial expression seems promising; these systems produce data along the lines of an EMG but only use a camera to measure muscle activity in the face.

Emotion can be viewed as a vector and measured along two scales: magnitude and valence.  Magnitude describes the intensity of the emotion, while valence describes its quality (either positive or negative).  You can measure the magnitude pretty reliably using something like heart rate, but understanding the valence is the tricky part.  How do you know if that intense emotional response is good or bad?  Of course you could just ask, but again that’s not a preferred method because people don’t describe their own experiences reliably and you’re introducing bias into the response.  Context is a better basis.  If someone is getting killed repeatedly, you can assume that they’re experiencing a negative emotion.  However, to validate we’d love to have a system which quantifies valence in real time.

Once we can measure these qualities reliably, we can start asking what the ideal emotional experience should look like over the course of the player’s interaction with the game.  Maybe that would be something like a pattern of peaks and valleys that steadily rises over time, as opposed to a prolonged burst of emotion that’s experienced all at once.  That seems like a plausible theory, but we won’t know until we’ve measured it.

Q: What are some of the design elements that you’ve found make better player experiences?
A: I can suggest a few things.  First, the player needs to be able to understand the experience.  If you die, you need to understand why you died.  If you reach a decision point, you need to understand what the implications are of taking path A or path B.  The designer needs to provide a sensible environment.

Variety is also really important.  Don’t give people the same monsters again and again, or force them to traverse the same levels over and over. There are obvious counterpoints to this, and the constructs of the game may dictate a lack of variety, so it’s not a hard and fast rule (none of these are), but it is something we try and emphasize.  The Left 4 Dead series is a great example, because you’re always interacting with a new set of players with different skill levels and different tactics, and that will completely change the dynamic of the game.  It doesn’t play the same way twice.

Third, you want to provide people with a feeling of continuous advancement.  People prize rewards if they increase in perceived value.  They want to feel that the required level of skill builds gradually as the game progresses.

Finally, have the player make interesting choices.  Which weapon should I choose?  Which armor should I take?  If these decisions don’t involve meaningful tradeoffs, then you’re probably not creating an enjoyable experience.

Q: How do you foster collaboration in multiplayer games?
A: Left 4 Dead is really designed to force players to cooperate.  If you go out on your own, for example, you’ll get incapacitated very quickly.  The game doesn’t prevent you from doing that — it’s a choice you can exercise, but it’s inevitably a losing strategy.  If you have other players near you then you can collectively put up a stronger fight, and when you fall then they can easily revive you.

Testing helped us improve collaboration in Left 4 Dead as well.  In the original design, the thinking was that players would build awareness of each others’ locations just through verbal cues, speaking to one another through a headset.  But it turned out that in the midst of gameplay that doesn’t work well.  When a teammate fell and needed to be revived, the other players had a difficult time finding him or her.  They needed another cue, so we introduced glowing outlines that appear around your teammates’ bodies, and which are visible through walls.  We found that really increased the players’ situational awareness, facilitated cooperation, and created a better gameplay experience.

Q: What kinds of quantitative metrics do you use to inform design?
A: We work with tons of data.  We can track any variable available in the game.  We’ll take information about where people die in each level, then overlay it on an image of the level to show whether people are dying in the right places, and in the right numbers.  We can examine the growth in players’ skill levels over time by any of various measures, depending upon the needs of the game’s design.  That may be a fairly coarse metric such as the ratio of kills to deaths, who gets the most kills, who stays alive the longest, and so on.  Or you can apply several measured in combination to satisfy a very precise definition of the ideal skill level, such as players who have a moderately high rate of kills but who win a lot and stay alive for a very long time.

I really appreciate your time.  I’d wish you luck, but with these kinds of practices it really doesn’t sound like you need it.

Looking for Haiti Relief Games

Two weeks ago Farmville made a special white corn crop available for real money, and passed the proceeds on to Haiti relief efforts.  In a news release, Zynga reports that they raised over $1 million.  I think it’s a really amazing example of a game interface used to meet a real-world need.  I also expect it’s fleeting, so if you know of any other online games that are doing something similar, please do let me know through a comment to this blog post.

Applying UX skills to game design

Nathan Verrill of Natron Baxter Applied Gaming provides a wonderful in-depth explanation of how he designed Signtific Lab, a game that facilitates prolific high-quality brainstorming around a central question.  While the project was wholly a game, Nathan’s descriptions of the process steps and deliverables will sound very familiar to any UX designer: mind maps, wireframes, design comps, user testing, analytics, and so on.  His background is in the design of conventional user experiences, and those same core competencies lent themselves well to the design of a successful game.

While they currently reside in different industrial families, user experience design and game design share a common parent in human-computer interaction.  To the extent they differ, there are opportunities for cross-discipline learning.  To the extent that they’re similar, expertise and skills transfer well from one practice to another.  My own experiences applying UX skills to game design provide examples of both.

I designed my first game in 2002, when Unisys started an annual tradition of sending e-cards with embedded games to clients, employees, and partners.  Each year, the in-house Web team would design and develop an original game, taking it from concept to delivery.  Our first idea was for a miniature golf game that fit in with the company’s sponsorship of professional golf tournaments.  While we were excited about the opportunity, none of us had made video games before.  So we applied the same methods and skills that we used in the design of websites, simply because they were the only ways we knew to approach any design problem.

We started by conducting ethnographic research at a miniature golf course.  Now I realize that last sentence reads like it’s meant to be facetious, but this was actually an indispensable step in understanding what makes the real-life game interesting, exciting, frustrating, funny, social, competitive, and worthwhile.  For example, we discovered that the courses were often designed to tempt people who overestimate their own proficiency to attempt difficult putts which, if missed, put the ball much farther away from the hole.  This in turn creates a social dynamic that can reverse the fortunes of beginners who play it safe, and skilled golfers who take greater risks.

Portion of the game wireframe, showing the putting interface

Portion of the game wireframe, showing the putting interface

From there I designed a short wireframe, available here as a PDF.  In some ways this was a traditional document, showing the core functionality while saying very little about the game’s appearance.  But in other ways it was very different.  The document focused on small interactions, as we were developing every interface element from the ground up instead of relying on ready made widgets like those baked into Web browsers.  These were presented as atomic pieces that could be assembled to build a course, much like a pattern library.  I also experimented with ways to show motion over time, and the effects of objects moving relative to one another.

Golf game wireframe, showing some of the obstacles

Wireframe of some of the obstacles players had to face

The finished game, which you can play here, had its strengths and weaknesses.  The visual presentation was fantastic and the level design was really good (owing to the efforts of Todd Horning and Mike Rosario), but it had some important usability and learnability problems (precisely the things that I should have been on top of).  I think the core mistake was in describing the interface elements as individual pieces without showing how they should be put together.  At the time, I reasoned that game design needed broad creative latitude and that the traditional prescriptive wireframe would have been too limiting.  But it turned out that the way the pieces hang together, as with a conventional user interface, is really critical the experience of the game.  For subsequent games in the holiday series my documents actually started to look more and more like Web wireframes.

Do you have examples of games you’ve designed using conventional UX methods?  If so, I’d love to hear from you!

Interview: Luis Von Ahn, creator, Games With a Purpose

In 2003 Luis Von Ahn introduced The ESP Game, which challenged two players working online to independently pick the same words to describe a picture.  But The ESP Game was also designed with a covert purpose: to improve search technology and the accessibility of the Web by gathering metadata about untagged internet images.  Impressed by the game, Google picked it up and renamed it Google Image Labeler.

Dr. Von Ahn, a professor at Carnegie Mellon University and recipient of the Macarthur Fellowship, has since built out a collection of similar “Games with a Purpose”.  I spoke with him recently to discuss the theory behind his work and his vision for how it can change the way we approach the design of user interfaces.

Q: So what are “Games with a Purpose”?
A:
To the player a GWAP is for all purposes a game, but as a side effect of play it’s designed to produce useful work.

Q: Why build something like Google Image Labeler as a game?  Why not just show people a picture and ask them to submit tags for it?
A:
Well, because nobody would do it.  There has to be motivation for doing work.  There are a few ways you can provide that.  You can pay people for work, and that’s effective but it’s also expensive.  Then there are motivations that drive people to contribute to something like Wikipedia, perhaps because they believe it’s a worthwhile thing or because they like the feeling that they played a personal part in building it.  But that model has failed when people have tried to apply it in other contexts, so it’s not a reliable motivator.  Then there are things that people do because they enjoy them.  So with GWAPs, instead of paying people with money you pay them with entertainment.

Q: How much can you accomplish by playing games?
A:
On average, Americans spend 1 hour every day playing videogames.  That’s over 100 billion humanhours a year.   That’s a humongous opportunity, considering that it only took 7 million humanhours to construct the entire Empire State Building.  And consider too that while people are spending all that time playing games they’re using their brains.  If you could turn all gameplay into useful work, people would be amazingly productive.

Q: If people are just playing around, then how do you know that the results are of good quality?
A:
There are a couple of tricks to that.  First, you can correlate one player’s results with those of other unrelated players.  For example, in the ESP game the same image will be shown to multiple players who are asked to submit tags describing it.  Since those players have never met and never had the opportunity to interact, if more than one person gives the exact same answer then it’s much more likely to be a reliable tag for the image.  Second, you can give players questions for which you already know all possible correct outputs, to see if they’re answering honestly.  If their responses fall outside of the set of correct outputs, then you can flag them as suspicious and ignore the rest of their responses.

Q: Since you started promoting Games with a Purpose, do you feel that the use of GWAPs has progressed as you’d envisioned in the broader community of design practitioners?
A: Yes and no.  I think it has been catalytic to what is today called “crowdsourcing”, which didn’t even have a name when we started.  But games haven’t gotten to the point where I’d like them to be.  Ultimately I’d like to see all work turned into a game (I don’t see why it couldn’t be), but we’re not there yet.  That’s probably because it’s very, very hard to design a good game.  Once you add in the constraint of the game producing useful work, then  it becomes even harder.  The potential’s there, but I think designers are just starting to figure out how exploit it.

Q: So how do you go about designing games?
A:
Well first we just think about them.  We think about them a lot.  Then we build a prototype using just paper and pencil, and start testing it like hell.  You really can’t tell whether a game will be fun or not until you test it.  And if you find that it is fun then you build a simple live version and test that, revise it, and so on.  And even then there’s no guarantee that it people will enjoy it.  Of the games that you complete, you’ll find that some are much more fun than others.

Q: Can you talk a bit about fun?
A:
Actually I’m not sure how to define the word “fun”.  What really matters is whether or not people play the game.  It’s a strange paradox that people will often play a game that they don’t even find enjoyable.   So I prefer to sidestep philosophical questions about whether or not people are really having fun, and focus on what we can measure.

This is a spectacular direction for user experience design.  Thanks so much for your time!

Webcast: Extending Game Design to Business Applications

I love games and think they have a lot to teach us about user interface design.  This July, I’ll be webcasting a revised and updated version of my talk on extending the gaming experience to conventional UI’s through Rosenfeld Media and Smart Experience as a part of their Future Practice series.  I debuted it last year at the IA Summit and EuroIA, and got a lot of great feedback from people who found it fun, interesting, and useful.

Details are available on the webcast’s description page.  You can watch it individually, or project it in a conference room for as many people as you like on the one registration.  If you decide you’re interested, be sure to use the code FERRARAWBNR to get a 15% discount.

If you’re interested in hearing something different, insightful, and kind of neat I think you’ll find this one fits the bill!

Letting crows do the work for you

Let’s say you wanted to gather all the loose change within a 5-mile radius, but you lack the time to do it yourself (which seems likely).  Why not enlist crows to do the scavenging for you?

That’s exactly what Josh Klein has done.  For his master’s thesis at NYU Klein invented a device that, through operant conditioning, trains crows to gather coins and drop them into a slot.  The device then disburses peanuts as a reward, like a vending machine.  It’s a fantastically efficient arrangement, since he only needs to keep the machine stocked and the crows take care of the rest.

C. Brachyrhynchos, the American crow

C. brachyrhynchos, the American crow

This is formally described as “synanthropy”, adapting animals to work within human environments.  But I think it’s an awesome demonstration of an even larger strategy: getting an organism to do a useful job while pursuing an unrelated goal.  The crow doesn’t share our interest in money, it’s just trying to get some peanuts.

Luis Von Ahn applies the same strategy to solve difficult computational problems, using human beings as crows.   A researcher at Carnegie Mellon, he’s developed a series of “Games with a Purpose” that surreptitiously gather useful metadata while engaging players in 2-person online computer games.  As with Klein’s crows, the players are pursuing their own goal — in this case to have a bit of fun.  But through the games’ design, they create a byproduct that has real value.  All of the games attach human meaning to data that machines can’t read.

His games are:

Games with a Purpose

Games with a Purpose

  • The ESP game, where two players are shown the same image and asked to suggest tags describing it, then awarded points when they match up.  When more than one person has independently chosen the same tag, that implies it’s a reliable descriptor of the image.  Search engines can then use those tags to return more relevant image results.
  • Tag a Tune plays a musical track for both players.  The players need to figure out whether they’re listening to the same track or different tracks by sending each other descriptions of what they hear.  The value of the game is that in the process they’ll provide useful tags like “piano”, “airy”, and “female vocals” to indicate qualities that would otherwise be indecipherable to a computer.
  • Verbosity is a password-like game that gives one player a word to describe to the other player, who has to guess it.  The game provides a number of canned relationships that can be used to describe the word, like “It is a type of ___”, or “it is the opposite of ___”, or “it looks like ___”.  For example, if the word were “cough” then the clues might be “it is a type of physical reaction” and “it is used to clear your airway”.  This game assigns meaningful ontology relationships (known as “triples”) to words, which can be used to deepen computer understanding of human language.
  • Squigl takes the images and tags created in the ESP game and asks players to trace the portion of the image in which the tagged object appears.  So for a picture of a woman walking her dog that’s tagged “leash”, both players would be tracing a similar area of the picture.  Points are awarded depending upon how well the areas overlap.  There are several possible utilities for the data gathered, from deciding what proportion of an image is relevant to a tag to informing image-reading applications.
  • Matchin simply shows both players two pictures and asks them to pick the one they like best.  Each consecutive time they pick the same one, they’re awarded an increasing pool of points.  This game provides data along the lines of Flickr’s “interestingness” function, adding human subjectivity to digital images.

Von Ahn compares the design of these games to the design of an algorithm.  Each one ensures that its data output is correct, and has a certain level of efficiency associated with it.  Like Klein’s device, they also take advantage of operant conditioning by awarding players points on the site’s leaderboard.  It’s a simple currency and less tangible than peanuts, but effective because it’s promoted as a measure of prestige.  Profiles even allow players to use the site as a matchmaker, pairing people who see the world in similar ways.

There are plenty of other examples of people acting as crows on the Web:

  • Search logs of the most commonly submitted queries provide Web designers with a prioritized list of the things users expect to find on the site, expressed in their own words.
  • Flickr users feed image-processing applications when they tag their images for personal use.
  • Foreign language translations of documents are used to inform probabilistic translation applications.

In all of these cases, the people doing the work have an objective that’s unrelated to the way in which someone else makes use of their work.

User-distributed work is emerging as a central component of Web 2.0, but some jobs are too laborious, too tedious, or too unfulfilling for people to pursue them on their own merits.  Learning how to use crows will be an important part of governing the role of the human being in future applications.

In Defense of Documentation

At recent conferences and in discussion groups, I’ve found it’s become increasingly popular to deride design documentation as time-wasting, inscrutable, and pedantic.  There’s a growing drive to jump directly into development, and manage design on the fly.  Few people are providing cogent arguments to the contrary.  As an IA who prides himself on the quality his documents, I feel the need to address the role of the spec in the age of Agile.

Bad documentation is real, and must die

There’s no doubt that there are some absolutely terrible documentation practices out there.  If you’re writing several dozen pages of design strategy in plain prose, you shouldn’t be surprised when no one bothers to read it.  A core measure of a development effort’s success is its efficiency, and documents that slow down the process should be hastily ditched.  It makes no sense to compound the time lost in writing poor documentation with even more time spent reading it.

Good documentation is also real

Good documents are completed quickly, express ideas efficiently, expedite development, build trust among team members, and allow people to achieve things that would otherwise be impossible.  People enjoy reading and discussing them, and teams of designers, developers, writers, and clients work as a whole to review and iterate them.

An analogy

Two years ago my wife and I decided to upgrade to an HDTV.  The problem was that we were going from a 24” tube set to a 46” flat panel, and the new set wouldn’t fit our existing furniture.  Even worse, the money we were spending on the TV precluded us from buying new stuff.  So I decided that I’d build a TV stand myself.

"Wireframe" of our TV stand

"Wireframe" of our TV stand

Now you need to understand that I’m not the sort of person you might describe as “handy”.  I’d never built so much as a birdhouse, and my high school didn’t even have a wood shop.  I had absolutely no idea how to construct a physical object.  Furthermore, my wife strictly forbade the use of any saws or sharp objects for fear I might cut off something important.

So I decided to create a physical object in the same way I would create any virtual one: I drew up a wireframe in Visio.  Before picking up a hammer or even buying the wood, I first mapped out the whole vision on paper.  I shared early top-level designs with my wife, we talked them through, and I modified the plans until we agreed on the approach, and then I filled in all the low-level details.

The completed TV stand

The completed TV stand

I’m pretty pleased with the result (see the picture at right).  It’s not exactly Herman Miller, and it weighs about 50,000 pounds, but it meets our needs precisely and cost me just under $100 to make.  And this example illustrates a few key advantages to a good documentation process over going straight to development:

  • Faster completion time. The documentation itself only took two days to develop, and since everything was planned out beforehand there was no wasted development time.  I got the project done in six weeks, much faster than I could have without the documentation.
  • Control of design. There’s a terrific symmetry between the design on paper and the finished product.  This was very important, because a lot of factors had to be balanced in the design: ensuring structural integrity, fitting all of the components in without making the stand too tall, and creating aesthetic beauty.  Jumping straight in, I wouldn’t have been able to manage that balance.
  • Planning around restrictions. It’s hard to build something out of wood when you can’t use a saw.  Lowe’s could cut the wood for me, but they could only do straight cuts at 90° angles to the board.  This was a built-in hardware limitation – not a deal-killer, but it required some creative planning.  Accounting for restrictions in the documentation keeps them from becoming show-stoppers later on.
  • Lower risk. Of course I could have jumped straight into the build, but that would have been much more risky.  I would have wasted a lot of lumber.  It could have come out looking terrible.  I could have discovered halfway through that there was a much better way of doing it.  Designing it on paper first was harmless, and maximized my chances of success later on.

Hmm, now what could that have to do with UI design?

By now this connection should be abundantly obvious.  In fact, from the most general point of view of process there is no useful distinction between building a TV stand and building a website.  Both require design and development efforts.  Both involve costs and risks.  Both produce an end product that needs to satisfy many needs equitably.

Quality documentation allows people to achieve things that are otherwise impossible.  The human brain can only do so much on its own – we need to extend its capabilities by processing things on paper, with our hands, and through conversation.  The task of designing software is often so complex that you can’t even grasp the full problem until you’ve written enough of it down.

But perhaps most critically, a good document provides for a foundation of trust to be built between team members.  Developers shouldn’t have to do design – it’s a distraction from their responsibilities, and exposes them to accountability for decisions that should have been handled by someone who’s paid to make them.  If the developers can trust that all foreseeable design decisions have been anticipated and built into the documentation, they’ll stick to it zealously.

What about Agile?

“Agile” is the name given to a process of rapid iterative prototyping and efficient communication.  I think it’s a terrific thing, because it allows concepts to be tried out and massaged into good shape in short order.  I employed a bit of Agile while building my TV stand – using scrap wood to figure out to make dowel joints and apply a stain, and grabbing whatever was within reach to get a job done.

But Agile has spawned a certain animosity toward even very high quality, efficiently produced, time-saving documentation.  I don’t think that’s Agile’s intention.  I think it rightfully seeks to eliminate time wasted producing documents that no one will bother reading, but perhaps its mantra has been repeated so often that it’s grown beyond its original meaning.  Played out to its logical extreme, ditching documentation would lead to disaster on all but the simplest projects.

So what should we do?

We need to reestablish the role of documentation in the product development lifecycle.  To do that, we first need to come to grips with the fact that what people have received in the past has been found wanting.  So we need to make a few key changes:

  • Crank it out. We need to be able to turn documents around faster than developers can build to them, or else we’ll become a bottleneck.  Our documents should be done in the simplest form possible, with no excess ornamentation or verbosity.  Ugly is efficient.
  • Communicate efficiently. We do after all work in usability, so our top priority should be to keep the needs of the document’s consumers in mind.  Integrate images with text effectively, speak clearly, and edit out anything that doesn’t solve a problem.
  • Know your medium. We don’t need to be coders, but if we understand how HTML, CSS, scripting, and databases work we’ll be able to take advantage of the strengths of these technologies.  Marshall McLuhan was right – the medium is the message.
  • Make people’s jobs easier. Bad documentation adds to other people’s workloads.  Good documentation lightens them.  Everyone will be happy to have something that reduces the total number of decisions they need to make.  Our documents should take control of everything that will affect the usability of the system.
  • Produce gold. People need to believe you’ll design something that will make everyone involved look good, and bring them to a better outcome than they could hope for without us there.  Get everyone’s input into the document as it’s developed, and iterate it repeatedly to build confidence in the finished product.