Contents
Editorial: It Ain’t Art
Chris Crawford
Editor Chris Crawford
Subscriptions The Journal of Computer Game Design is published six times a year. To subscribe to The Journal, send a check or money order for $30 to:
The Journal of Computer Game Design
5251 Sierra Road
San Jose, CA 95132
Submissions Material for this Journal is solicited from the readership. Articles should address artistic or technical aspects of computer game design at a level suitable for professionals in the industry. Reviews of games are not published by this Journal. All articles must be submitted electronically, either on Macintosh disk , through MCI Mail (my username is CCRAWFORD), through the JCGD BBS, or via direct modem. No payments are made for articles. Authors are hereby notified that their submissions may be reprinted in Computer Gaming World.
Back Issues Back issues of the Journal are available. Volume 1 may be purchased only in its entirety; the price is $30. Individual numbers from Volume 2 cost $5 apiece.
Copyright The contents of this Journal are copyright © Chris Crawford 1991.
_________________________________________________________________________________________________________
Editorial: Patents, Poobahs, and Peasants
Chris Crawford
It seems that a number of big entertainment software publishers have begun to patent game technologies. So far, there are only a few patents pending, but we’re just seeing the beginning of a trend that will surely grow.
Why? First, as development costs have risen software publishers have begun to move development in-house where they can exercise tighter fiscal and editorial control. This has in turn required the creation of expensive in-house studios, with sound specialists, graphics artists, animators, producers and lots of programmers. Publishers naturally want to protect their substantial investment in such in-house studios and the technology they create.
Add to this the fact that big publishers already have a lawyer on staff and budget a significant amount of money for legal matters. A six-person development shop might balk at the expense of patenting its technology, but to a big publisher, this is just part of the cost of doing business. Moreover, there are economies of scale involved in patenting a company’s technology; once a company learns the basic procedures necessary to facilitate a patent application, it becomes cheaper to apply for more patents.
Thus, our industry has passed an important milestone in its history. We have become big enough and rich enough to start patenting everything we do. That’s good, right? Wrong! The patent process is peculiarly biased in a way that can only work to the detriment of the industry as a whole. Our industry is in a transitional phase during which patents perform an industry-wide disservice.
Some Background
Let me review some basics about patents. The intent of patent law is to encourage technological innovation by offering legal protection to the fruits of that innovation. Technology costs money to create, and no businessman will invest in research if the technology that his research yields can be sanpped up by his competitors. The government therefore offers the businessman a temporary monopoly on his technology.
The process, however, is cumbersome and expensive. The cheapest patent will cost about $5,000 to obtain. A complex patent or one in a highly competitive field can cost much more.
Why so expensive? Let’s think in terms of an imaginary intellectual territory, a land of ideas and technologies. A patent stakes out a portion of this terrain as your personal property. Nobody can trespass inside it without paying you rent. Your goal in applying for the patent is to stake out the largest chunk of territory the government will let you have. The more territory you control, the better the chance that somebody will roll a 12 and hit your Boardwalk with a hotel on it. Of course, everybody who’s obtained a patent before you has attempted to do the same, so you have to pay for a patent search, a survey of all pertinent patents to see what neighboring territory has already been staked out. The more complex or lucrative a field, the more existing plots there will be; it can be quite a thicket, and an expensive proposition to chart.
An important concept in patent law is "reduction to practice." You cannot patent an idea, only the actual method by which it is reduced to practice. For example, you could not patent the idea of collision detection; you would instead patent a particular scheme for collision detection.
Another important concept is "prior art." You cannot patent technology that everybody has been using for years. When you apply for a patent, you are required to notify the Patent Office of possible competing technologies that will help define prior art. One of the best ways to fight a patent is to show that it uses prior art and should never have been granted in the first place.
The Problem
This brings me to the reason why patents are bad for our industry. It used to be an industry of small-time operators, individual freelancers and tiny development groups. Most of the prior art was created by these tiny groups. But these tiny groups don’t have lawyers and are probably unaware of the patent efforts underway. They are legally voiceless. As far as the Patent Office is concerned, they don’t exist.
This skews the patent process, making it easy for big companies to grab up patents for work that they really did not originate. To extend the intellectual territory analogy, imagine the scruffy and legally naive Gold Rush 49ers being kicked off their claims by the slick corporations from back east whose lawyers know how to deal with the Claims Office.
These are not idle paranoid ravings; I understand that two patent applications have already been filed by entertainment software companies for technologies used in game software. Doubtless more are underway. If the trend continues, you can look forward to the day when you receive a certified letter from an attorney informing you that your latest product violates somebody’s patent, and demanding that you remove it from the market or pay a license fee.
Did you know that the basic technique for displaying a cursor over a bitmapped screen display has been patented? We all know how to handle the problem: blit the cursor bitmap with the background, exclusive-orring it with a mask. It’s a standard technique -- but it is patented and several hardware manufacturers have paid stiff license fees to the patent holder rather than fight the patent in court.
A Proposal
At this point, the more cynical among us might be tempted to shrug our shoulders and declare this to be yet another example of how the rich get richer, the powerful oppress the weak, good guys finish last, etc. etc. But such is not the case. We CAN do something about it.
Recall the role of prior art. The defendant in a patent-infringement case can prevail if he can demonstrate the existence of prior art. That’s where we come in.
Suppose, for example, that MoneyBags Software, Inc, has slapped Joe GameDesigner with a patent-infringement claim because they hold a patent on collision detection, and Joe’s software uses the same collision detection method.
Suppose that Bob Bystander developed the same collision detection method back in 1984, long before MoneyBags filed its patent application. If Bob could be made aware of Joe’s predicament, and came forward to document his prior work, then the tables would be turned. MoneyBags’ patent would be jeopardized or nullified. Realizing this, MoneyBags would back down.
Consider now that there are hundreds of us, and that collectively we have worked on thousands of projects over the last decade. Think of all the prior art we collectively know of. Now consider this: do you think that there is anybody anywhere in the industry who could invent a new games technology that has not in some way been anticipated by somebody with prior art? Not bloody likely, is it?
Now for another consideration: most of the relevant factors that affect patent cases are anticipatory in nature. Patent litigation is appallingly expensive. It is seldom worthwhile to pursue patent cases in court. Which means that the MoneyBags of this world might well be deterred from making overly ambitious patent applications if they can be made to realize that, first, there is a high probability that prior art exists for the claims they’d like to make; and second, that there are people out there willing to come forward with the patent-busting information.
For now, there is no pressing need for any overt action on our part. All the major publishers read the Journal and they are smart enough to figure out the implications of all this. If anybody ever does receive a patent-infringement demand letter, let them forward a copy to me. I’ll put out the word here in the Journal. Then, presumably, some good citizen will step forward ready to document prior art. If the strategy is executed properly, MoneyBags Software will be driven back snarling but defeated, with little more expense than that caused by a few letters flying between lawyers.
And if MoneyBags’ patent is fair and deserved, nobody will step forward to document prior art because none exists. Everybody wins (well, everybody except the lawyers, but they don’t count, do they?)
In a few years, I suspect that this problem will go away. Most technologies will be developed inside publisher studios, and the patent wars will be something of a fair fight between giants. For now, patent wars are about as one-sided as Europeans versus Native Americans; soon, we little guys will all be extinct and the problem will go away. Isn’t that sweet?
_________________________________________________________________________________________________________
How to Build a World
Chris Crawford
Global Dilemma (my title was Guns & Butter) was not one of my better games. It had some severe problems integrating economics with the military portion of the game. But it did contain a number of elements of which I am proud. One of these was the continent-generating algorithm. World-building routines have been part of a number of games: Seven Cities of Gold, Starflight, Empire, and others sported world builders. The basic task is simple to state: generate a unique terrain map from a random seed, insuring that the map contains interesting yet functional terrain. Achieving it, however, can be quite a challenge.
For Guns & Butter, I wanted to make a world-builder that would go beyond current technology, a world builder that would support complex terrain types yet look good and not concoct impossible terrain. In this article, I will describe the world builder that I designed.
Specifying a world
I began by determining that my world would consist of a single continent. My game did not need naval forces and I thought it best to avoid the complication of maritime rules. Next, I defined the size of the rectangle into which the continent would fit. The continent itself would be composed of fundamental units called provinces. Each province would be an irregular region with a single city as provincial capital. Countries would be collections of provinces. Straight roads would connect adjacent provinces, but not all contiguous pairs of provinces would be connected by roads.
To lay down the continent, I simply laid down provincial capitals randomly. My random number generator randomly selected points in the available space, the only restriction being that no new point could be placed within 64 pixels of an already existing provincial capital. This prevented overcrowding of provincial capitals. I laid down 64 provincial capitals. These were the skeleton of the nascent continent.
With the capitals in place, my next task was to form the provincial boundaries. The boundaries had to divide up the existing territory among the various provinces, and then assign nicely meandering boundaries. After all, provinces that look like polygons would never do. I decided to tackle the problem in two steps: first, to assign polygons that divided the territory, then "meanderize" the polygon sides.
Drawing the polygons
The task of making a clean set of polygons out of a set of points is quite tricky. After I had solved the problem, I discovered that it is addressed in Sedgewick Algorithms, 2nd Edition, pp 408-410, although his discussion is even more hand-waving than this essay.
The problem can be simplified by casting it into a different form. For the set of cities, create a set of spokes connecting the cities such that 1) all cities are connected by spokes; 2) no two spokes cross each other; 3) every pair of cities that could be connected with a spoke is connected with a spoke and 4) the total length of all spokes is minimized. I’ll show you how these spokes can be used to make provinces later. For now, I shall concentrate on building the spoke set. Consider the following diagram:
I shall create the spokeset for City 1. I first create a list of the nearest cities, ranked in order of their distance from City 1. In this case, the list has cities 3, 6, 2, 5, 4 and 7. Starting at the top of the list, I begin adding cities to the spoke set of the City 1, but two factors could exclude a city from the spoke set. In this case, Cities 4 is masked from inclusion by City 3. The spokes are too close together and the subsequent border will be too seriously deformed. Therefore I must reject City 4. The obvious way to do this is to calculate spoke angles and look for spoke angles that are too close together. As it happens, the use of trigonometry in this calculation is prohibitively slow, so I used another, more complex system that runs faster. This other system is too messy for me to explain here. It used analytic geometry to calculate whether any of a set of lines intersected.
Building a province from spokes
The result of all this calculating is a set of spokes for each city on the map. Those spokes break the space of the continent up into triangles. This is important, as quadrilaterals will cause this algorithm to fail. Note that condition 3) of the above set of requirements for spokes insures that all the shapes formed by the spokes are triangles. To form a province around a city, we create two sets of midpoints: the midpoints of the spokes themselves, and the midpoints of the triangles. We connect all these midpoints to make a province:
Randomizing the province boundaries
We now have a set of polygonal provinces. The borders are straight lines, which looks much too artificial. We need a reliable way to make those borders wiggle realistically. Moreover, there’s a nasty restriction: we have to have the same wiggling on both sides of the border. That is, when we create each province’s final shape, the zigs and zags of its border must match the corresponding zigs and zags of its neighbor’s border. This, it turns out, is a major problem.
My solution to these requirements was to create a deviation table 1,024 elements long and stock it with random numbers between -7 and +7. The algorithm begins at one vertex of the polygon and prepares to walk down the edge of the polygon. First it selects the starting table index from the deviation table that it will use. This selection is based on the coordinates of the vertex and the direction (clockwise or counterclockwise) in which the polygon is being drawn. This guarantees that the same deviation table entries will be accessed regardless of which side of the border is being drawn. This is what insures that we get matching borders for the two adjacent provinces. This trick also insures that different sections of the deviation table are used to draw different borders, so there will be no telltale pattern to the deviations.
The algorithm takes many small steps down the polygon edge towards the next vertex, at each step wiggling the border in the x and y dimensions by the amount given in the deviation table. The result of all this is a beautifully irregular province.
Adding mountains, forests, and deserts
My final task was to add terrain features: mountain ranges, forests, and deserts. I decided that I wanted these terrain features to act as apparent blocks to roads connecting provinces. You will recall that all adjacent provinces are connected by spokes. Well, not all spokes became roads. Some spokes are left empty. The spaces created by these non-road spokes are the spaces that I wanted to fill with terrain features.
To define these terrain regions, my algorithm traces a complex polygon by laying down a right-handed path. It first grabs some non-road spoke. It then lays down an initial point one-quarter of the way down the non-road spoke.
If the spoke on its right is not a road spoke, it draws a line to the one-quarter point on that spoke. It continues this process until it encounters a road spoke. When that happens, it draws a line to the midpoint of the province polygon edge emanating from the road spoke. It continues drawing to the midpoint of the line connecting the city in the opposite province to the end of province polygon edge which it had just bisected. The process described in this paragraph is iterated until the algorithm returns to its starting point.
The result is a polygonal region occupying the non-road areas between provinces. Depending on whether it is a forest, a mountain range, or a desert, I randomly scatter the appropriate icons inside the region. Although they are created in random order, they are drawn from top to bottom so that the base of one mountain or tree does not obscure the top of a lower one.
Results
A typical end result is shown below. As you can see, the continent is quite pretty. It takes about five minutes to build this continent on a Mac Plus.
_________________________________________________________________________________________________________
Letter
Greg Costikyan
[A Response to “It Ain’t Art” in the October 1991 issue]
The notion that art is elitist and its appreciation restricted to those with noble souls is historically recent. The spirit masks, pottery, vases, sculpture, and painting which we now view as historic objet d’art were largely created by craftsmen matter-of-factedly attempting to please clients. The works of Shakespeare appeal as much to the bawdy humor of the lower classes as the interest in intrigue and character of the upper. The novels of Dickens were written for Victorian serial magazines, the TV-series equivalents of the day, and their length is a direct function of their popularity: an unpopular serial would be wrapped up quickly, while a lengthy one would go on for many issues.
The spread of the notion that art is elitist is, in fact, one of the great artistic tragedies of the modern era. A hundred years ago, poetry had a mass following, the best poets of the age were folk heroes, and poems commonly appeared in the daily papers; now, poetry is an historical lacuna, maintaining a pallid half-existence in the pages of small-circulation literary magazines and thin volumes found in the more pretentious bookstores. “Literature” has similarly retreated; the writers best-regarded by the literary establishment are generally not read. New music in the classical tradition, too, retreated into cacophonic obscurity for years, reemerging into public view only through the more symphonic works of Philip Glass and Steve Reich.
Why the obscurity of such elitist art? Precisely because it is elitist. Artists bought the bullshit that “art” is inherently “uncommercial,” and that to satisfy an audience was ipso facto proof of inferior talent.
The truth is that any object or work that is crafted with intelligence, care, creativity, and a desire to communicate is art. It may be bad art or good art; it may succeed or fail; but it is art. Art is merely the application of craft with imagination and sensitivity.
Game design is an artform. The fact that most games are designed for purely commercial motives does not change this fact. “Entertainment value” and “artistic value” are not opposites; they lie on orthogonal axes.
Indeed, we need to promote the idea that game design is art. All too often, designers make game decisions for crass reasons: the producer wanted it, this feature was popular in game X, the little droogs seem to like this crap. If designers realized that they might well be judged on the intrinsic merits of their work, rather than purely on the sales, they might be tempted to spend more care, apply the intelligence and craft they do possess with more diligence. Your editorial basically says, “Don’t worry, be happy; design crap, it’s okay, it’s only entertainment.”
I remain, your humble and obedient servant,
Greg Costikyan
Hmm, apparently the bitterness that I wrote into that editorial did not show itself well. Take your ‘Don’t worry’ paraphrase of my editorial and say it with a sarcastic tone of voice and you’ll be much closer to my intended meaning. - Editor
_________________________________________________________________________________________________________
Advertisement
Top designers reveal how to design and sell videogames. 60 minute VHS casette.
Send $39.95 to Sligo Video, Dept CD Post Office Box 35433, Los Angeles CA 90035
_________________________________________________________________________________________________________
Another Letter
Doug Clapp
My wife is mildly amused, but not really interested. My friends--of an age with me which, unfortunately, is around 40 — don’t get it.
My thirteen year old son, however, thinks I’m both hero and genius.
The reason? I’m madly trying to decide between a Super Nintendo and a Sega Genesis. And an IBM ’386-33 with a Super VGA monitor.
The reason? Because I want to play some games. Some really good, really fun games.
My wife, I’m guessing, is amused because the house is already littered with computers: Macs, my big black NeXT cube, and jeez we’ve already got a Nintendo!
Game playing, for me, isn’t a constant. It’s a raging desire that comes in, hard, every few months. After working and reading and playing with the kids and making supper and making money and doing dishes for week after week —well, you just want to play some games, you know? Kill some monsters, big time. Fly a spaceship. Best a villain. Power up.
So I’m thinking. And it’s tough. The Super Nintendo has hardware scaling, rotation and zoom. That’s cool and should make for some neat games. But they’re not here yet, and won’t be until later in 1992.
The Genesis has great games now. And there’s a CD-ROM that should appear in a month or so. That’ll open the way for more great stuff. (I hear there are already amazing games for the Genesis CD-ROM in Japan.)
The Genesis is also about $50 bucks cheaper, but who cares if you’re also thinking about the ’386, which—let’s face it—will end up at about $2,700 no matter how many times you page through Computer Shopper.
Which is really only a preface to what I want to say.
Here’s what I want to say: God damn Apple for doing this to me!
Really. It’s 1984. The Macintosh is released. The price? $2,400.
$2,400!
Other prices were discussed: $1,995, $1,600, $1,200 and $995.
Hertzfeld wanted $995. All the True Believers wanted $995.
But it didn’t happen. $2,400.
Did it work? Was $2,400 the “right price?”
Yes and absolutely no. Yes, because the profits from Macintosh made Apple one of the largest and most profitable computer companies. Oh my but the money poured in, year after year. (Not as much as Steve planned and hoped for, but — hey! — a ton of money, okay?)
But absolutely no. Because now, in 1991, Macintosh has a 10% market share. IBM compatibles have the other 90%.
You can have the money now, or the market share in time. Apple took the money.
And now, as I wander through my neighborhood “computer superstore,” there’s a little section of Macintosh games and a long, long aisle of IBM games.
That’s part of it. The other part is this: While none of use were paying attention, the Apple II just...sorta...went away. These days, Apple Computer, Inc. makes Macintoshes.
The Apple II. The blessed machine. Color, remember? And great, great games: arcade, strategy, role-playing, educational. Games for little kids, too. Fun stuff. Great stuff.
Well, that’s for hobbyists, after all. The real market is business, right? And don’t worry, 1-2-3 will be here any day. And sure, we’ll make it easy to read DOS disks.
Well shame on you Apple! Because the market that was (oh: admit it!) “beneath your corporate attention” is now about to skewer the Macintosh. Right through the old, rather piddling, market share.
You’re about to be lanced by Nintendo and Genesis and NEC and Neo Geo. And IBM clones with Super VGA monitors.
Because that’s where the great games are. My son awaits Dragon Warrior III. My four-year-old daughter loves the Sesame Street games. And me? I just wanna kill some monsters, turn the sound up really loud, blow some stuff up.
There are no good monsters on Macintosh. At $995, the monsters — I believe this — would be here. But they’re not.
So another consumer will soon vote. I’ve got the money. Somebody will take it to the bank. This, you should know, is good America money. Same stuff, worth just as much as money from any corporation, any business, any yuppie suit.
I’m leaning toward Genesis. And I hear that Joe Montana Football II is even better than the original.
_________________________________________________________________________________________________________
Low-Interactivity Entertainment Products
Chris Crawford
One of the common complaints against computer games is that they are too demanding of the player. A great game like Secret Weapons of the Luftwaffe comes with a 224-page manual, and you’d better read the manual before you attempt to play the game. The big hits from Sierra all require dozens and dozens of hours of trial and error as the player attempts to crack the game’s puzzles, many of which are fiendishly obscure. My own work is just as demanding: one of my games sported a title screen that warned: “Persons who play this game without reading the manual are wasting their time.”
The problem with all this is that most people aren’t willing to invest that much time in a computer game. They want to be entertained without having to jump through hoops. They don’t want to be forced to work for their fun. This has been one of the reasons why computer games have failed to reach the mass-market.
Hence the temptation to concoct a game that doesn’t require much effort of the player. But what do we mean by “effort”? In the early days it was felt that effort was synonymous with reading documentation. The rule at Atari in the early days was that a game’s rules had to fit onto the small flyer that accompanied the cartridge. If the rules wouldn’t fit, the game was too demanding for most players and had to be redesigned. Nowadays we aren’t so dogmatic; we have seen some very successful games that have extensive documentation. Besides, the games that require little documentation all seem to be skill-and-action games.
The second factor that many associate with effort is interactivity. If the player has to work too hard to play the game, perhaps it is because the game asks him to make too many decisions. If we could reduce the interactivity, we could then ease up on the player’s workload. Perhaps this would solve the problem.
It is tempting to contemplate the possibility of eliminating interactivity altogether. After all, who anointed interactivity as the ultimate goal of all games? Why shouldn’t a sufficiently clever designer be able to create a game with no interactivity whatsoever?
This idea, so appealling at first glance, quickly collapses under scrutiny. If there is to be no interactivity whatsoever, then what will the player do? He could, of course, type on the keyboard or manipulate the mouse, but in a zero-interactivity game the computer would be precluded from responding to the player’s actions. The player is reduced to watching the multimedia presentation that the game designer creates. While this certainly reduces the player’s workload, it also reduces the personal computer to one hell of an expensive VCR — and not a very good one at that.
No, a zero-interactivity game is truly a contradiction in terms. It is indistinguishable from a movie or a novel. But there remains another possibility: the low-interactivity game. The idea here is to create a game that doesn’t demand much of the player, but he still gets to make some input every now and then. The player’s choices are simple and do not require much in the way of thinking. Think of a low-interactivity game as a kind of movie with occasional branchpoints. The player spends the bulk of his time watching and listening, and only rarely is he required to do something. If only we could figure out what that “something” is.
Some History
This is not a new idea. It has been around since the early days of computer games. I remember people at Atari talking about games for Joe SixPack. The marketing people at Atari often requested a game for couch potatoes. A great deal of hot air was vented on the subject, and there were a number of projects initiated, but nothing ever seemed to come to fruition.
In the first big boom of computer games, from 1981 to 1984, a number of low-interactivity games were attempted. One of these was Alien Gardens, published by Epyx. You were a kind of alien bee flitting through a garden of alien flowers, trying to pollenize them. It was a very low-key game, definitely low in the interactivity department. Unfortunately, it made no sense and ranks as one of the great turkeys of computer game history.
1985 saw another low-interactivity product: Little Computer People from Activision. This odd product created a small family on your screen, moving around their dollhouse in the course of their daily activities. You the player watched them. The product attracted much attention from the press but it was not, I believe, much of a commercial success. And it spawned no imitators or descendents.
Epyx roared back in 1988 with more low-interactivity products: its line of VCR games, released with much hype and excitement. Realizing the clumsiness inherant in the serial format of a videotape, the designers rightly limited interaction to the bare minimum, focussing most of their attention on providing interesting footage for the player who would occasionally fast forward or rewind. Now, here was the ultimate couch potato game. You didn’t need a computer to play it and you didn’t have to do much work. All you did was sit back and watch the tape and occasionally push a button. Sounds great, right? It sounded great to a number of publishers, who frantically put together their own VCR products. Yet, despite their obvious advantages and some expensive marketing campaigns, VCR games bombed. They were a total disaster. Mindscape shipped one product and cancelled the second one, even though it was ready to ship, because the first game had failed so completely.
Another experiment in this direction was the CinemaWare line of games. These games were strong on spectacle and weak on interaction. The marketing thrust of the CinemaWare line was that these games were just like movies, except that you could play with them. Most of the design effort was put into making lots of pretty pictures and animated sequences. The gameplay itself was weak. The first line in the series, Defender of the Crown, created quite a sensation and sold very well. But after that, it seemed to be all downhill. CinemaWare went bust early this year.
There’s a lesson here: low-interactivity games don’t sell. They sound like a great idea, such a great idea that people keep going back and doing them over and over. And, in pure Darwinian fashion, the companies that have cast their lot in with low-interactivity games have suffered extinction. Epyx, Activision, CinemaWare, and Mindscape have all been reduced to ashes. But the survivors seem unable to learn from their competitors’ failures; proposals for low-interactivity games keep popping up like some time-hopping Sisyphusian dodo bird bent on repeating its extinction in as many eras as possible.
Why?
Why have low-interactivity games been such a dismal failure? One would think that there should be some small fragment of a market for them. Why is the historical experience so decisively negative in defiance of common sense?
There are two answers, I think. The first is that the available hardware is not up to the task. We have not yet hit the right combination of ingredients to build good low-interactivity games. The VCR gives lots of imagery, but its access times are so slow that even low-interactive games suffer. The computer itself simply cannot generate or maintain images of enough variety and quality to entertain the player by themselves. This arguement suggests that optical media games might solve the problem. They offer faster access times than videotape, yet much greater image capacity than the computer. Whether the combination will be fast enough and visually rich enough, we cannot yet say.
The second answer is more pessimistic. I have long maintained that interactivity is the essence of the gaming experience, and that the quality of the interaction determines the quality of the game. If this be true, then the very notion of low-interactivity games is intrinsically wrongheaded and such products will inevitably fail.
One way of expressing this line of reasoning is to start with high-interactivity games and then move towards lower interactivity. What do we gain and lose as we move in this direction? As we lose interactivity, we reduce the total quantity of decision-making that the player must perform. This reduces his workload. It also reduces his ability to creatively influence the outcome of the game. In other words, as we reduce the interactivity of the game (its gameplay), the player’s degree of participation in the outcome is diminished and he therefore becomes less involved in the outcome, which becomes more pre-determined.
But there’s a catch: the player’s workload is not proportional to the quantity of decision-making. Decision-making consists of two parts: a laborious process of learning the basic parameters for making the decision, and a faster process of applying those parameters. The player must go through the first process whether he makes one decision or a hundred. Thus, his workload is equal to a fixed quantity (learning the rules) plus a variable quantity (playing the game.)
An example might help here. Suppose I present you with two games. The first is a truly minimal-interactivity game. You will be asked to make one decision during the entire game. It is a murder mystery game the denoument of which places you in a room with the six main suspects and a gun. You must decide whom to shoot. That’s exactly one decision, about as low-interactivity as you can get. Yet you will likely ask a great many questions before making your decision. Can I shoot more than once? Can somebody else shoot me? May I choose not to shoot anybody? (True gamesters will note that the problems are trivially solved by playing the game several times, experimenting with each of these options in turn. While entirely possible, this flies in the face of the stated intent of the low-interactivity game.) Note that these questions are really questions about the rules of the game. You will have a considerable workload just learning the context for your single decision, and inasmuch as the outcome of the game rides on your single decision, you had damn well better learn the rules thoroughly.
The second game is a more conventional game with many decisions. Once again you will have the workload of learning the rules of the game, and in addition to that you will have the workload of making all those decisions. Yet the workload of learning the rules is most likely the more substantial of the two. In other words, if you end up making a hundred decisions during the course of this game, your total workload will not be 100 times greater than your workload with the minimal-interactivity game. It might not even be twice as great.
Thus, as we move from the higher-interactivity game to the minimal-interactivity game, two factors are reduced: the player’s workload and his ability to influence the outcome of the game. But — and this is the key point — the latter falls faster than the former. Reducing interactivity gains us only small benefits in terms of reducing workload, but costs us heavily in terms of the player’s ability to creatively influence the outcome of the game.
This, I think, is the real reason why low-interactivity games have been such failures. Diminishing the interactivity just makes the game less fun faster than it makes the game easier. What we gain in terms of reduced workload we more than make up for in terms of diminished fun.
An Interesting Exception
There is a small group of low-interactivity games that have been undeniably successful. These are the games produced by Cyan (Manhole, Cosmic Osmo, et al) and Amanda Goodenough (Inigo Gets Out, Your Faithful Camel, et al) They are low-interactivity games, really more like vaguely linear stories with some buttons to press. They have been moderately successful. What is striking is that all of these products are designed for young children. It appears that our industry’s Darwinian methods have at last found a suitable habitat for this otherwise less-than-fittest species of game.
Why is it that low-interactivity products are successful with young children when they don’t seem to work with older players? I think that the answer can be found by asking another: why don’t high-interactivity products work with young children? Try foisting SimCity or Robotron or Ultima VII on a six year old, if you’re willing to risk accusations of child abuse. The poor kid will be overwhelmed by such games. He just doesn’t have the perspicacity to handle such a game. What’s left for him but the low-interactivity games?
The Cyan games and the AmandaStories games show that low-interactivity games can be produced, and they give us a good idea of how a low-interactivity game works. They also demonstrate that such products do not work well in the larger world. They offer both positive and negative lessons.
Wierd Ideas
Lastly, there are the blue-sky concepts for low-interactivity games. Most of these center on some form of storytelling. In one approach, the computer tells the player a story, with the player somehow providing cues that the computer uses to adjust the story to suit the player’s interest. For example, if the computer mentions an encounter with a beautiful girl, and the player so indicates, the computer could proceed to describe a sexual liason. If the player is female, it might tell of a friendship developing between the two.
The problem with this lies in the nature of the cues provided by the player. Exactly how does the player communicate his desires to the computer? If we use a series of predetermined branchpoints, then the game has reverted to a conventional adventure game, and the player still must learn the language of expression for the adventure. Proponents of such schemes often fall back on deliberately vague formulations. The computer will “sense” the player’s mood, they claim. I find it difficult to imagine just how this sensing will take place, and how the computer will interpret whatever it senses.
A variation on this scheme makes reference to the manner in which a performing artist senses the mood of his audience and adjusts his performance accordingly. This, it is asserted, constitutes an advanced form of low-interactivity that could be harnessed for new types of games. The problem lies in the input and processing required to accomplish this. The performing artist is analyzing fine shades of voice intonation and subtle nuance of facial expressions. This type of processing is way beyond anything we can process on a personal computer, even assuming that we could equip our games with microphones and television cameras to provide the input.
There is a second and more powerful arguement against such schemes. Even if we could implement them, they would still be inappropriate. The performing artist who adjusts his work in response to the audience’s feedback does so on a very gross average of the audience feedback. Some people will be screaming “Faster!” as others are yelling “Slower!” The artist can’t do both, so he responds to the majority. What’s more important is the fact that the audience understands this. We can’t all have our way, so we accept the situation. But when I am the only user of a computer game, I am completely justified in expecting that I can have my way. I expect the computer to respond to my wishes. If the computer fails to understand my wishes or is incapable of executing my desires, then I will be dissatisfied. So if I grunt or laugh or scowl or drum my fingers and the computer fails to get my message, then the product will have failed.
Conclusions
The historical evidence is quite clear: despite many attempts, low-interactive games have never been successful except as products for young children. We’ve been talking about this concept for more than ten years now and can’t seem to make it work. There is a slim possibility that we can make viable low-interactivity games using optical media. As yet, this is speculative.
The concept of low-interactivity entertainment is a ghost that we will never exorcise from this industry. The concept just keeps popping up like an annual flu bug. Some naive fool will come forward with “this great new idea that nobody has ever thought of before.” As I discussed, the concept seems sound on first examination, so people will probably give it credence. Who knows, some credulous publisher might be persuaded to part with development dollars to explore the idea.
_________________________________________________________________________________________________________
...and speaking of low interactivity
We seem to be having some difficulties interacting with the Journal lately. Once upon a time the Journal was a kind of slow-motion interactive operation, with readers submitting articles and this editor printing them. Of late, however, the Journal has become a more unidirectional endeavor, with this editor doing almost all the talking and the readership doing little more than reading. Now, as much as this arrangement suits the exalted opinion this editor holds of his own intellectual prowess, the voice of laziness shouts louder than that of vanity. I really can’t afford to write the whole damn issue; I lose enough money on this operation as it is. So please, get me some articles I can print!!!
_________________________________________________________________________________________________________
