Volume 2 Number 4 April/May 1989

Contents

Pac-Man = Zork
Chris Crawford

Culture, Idealism, and the Real World
Jim Gasperini

Publishing Contracts: A Few Notes
Mark Lewis Baldwin

How To Expedite Conversions
Gregg Tavares

Great Expectations
Gordon Walton and Karen Hunter

Once Upon a Time
Bill Hopkins

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 1989.

_________________________________________________________________________________________________________

Pac-Man = Zork
Chris Crawford

Gametrees
Interaction is the essence of the computer game.  Meaningful interaction requires that the computer respond to the actions of the player in a meaningful way.  Since the player’s actions are expressions of his decision-making, we can characterize the player’s movement through the game as a series of branch points forming a path through a tree of possible actions:

I have previously called this a “storynet”; for this article I shall call it a “gametree”.  Compare this structure with that of a story:

We note that the story’s structure is linear, a notion borne out by the often geometric terminology of storymaking.  We speak of storylines or plotlines, a twist in the storyline, or the story taking an unexpected turn.  A game’s structure is more complex.  Note further that a single playing of a game constitutes a traversal of the gametree and is structurally identical to a story:

 Thus, a game might be called a metastory, or perhaps a story-generator.  The number of stories it can generate depends on the bushiness of the tree structure.  A scrawny tree structure can generate only a few distinct stories, and the resultant game can therefore provide only a few decent replays.  A thick, bushy tree provides a huge array of interesting choices to the player and can generate an abundance of stories and many happy replays.

The goal of the designer, then, must be to create the largest and bushiest possible tree structure.  How can this be achieved?

Hard Wiring
The first approach is to directly specify each of the nodes in the tree.  This hardwired approach to game design suffers from a profound limitation: the designer can never create enough nodes.  Let’s walk through the numbers.

Suppose that you create a tree structure with 1,000 hard-wired nodes.  Each node specifies the situation in which the player finds himself, the graphics, animation, and sound associated with that situation, the options available to the player, and the consequences of each option expressed in terms of the identities of the situations into which each chosen option would propel the player.  For purposes of simplicity, let us assume first that each node provides the player with exactly three options, and that there is no degeneracy in the tree structure, that is, no two options anywhere in the tree lead to the same consequent node.  Given these assumptions, then the tree will have a total of six layers in it, with 729 final outcomes.  Thus, the game will be only six turns long, and there are only 729 possible outcomes.  Compare this with Tic-Tac-Toe, a game capable of generating about 500 possible outcomes.  Chess can generate something like 10**126 possible outcomes, if my memory serves me.  Obviously the hardwired approach does not produce rich games.

Maze plus state variables
There is another way to create a large and bushy gametree.  First, we create a maze through which the player moves.  The player’s spatial path through the maze then corresponds to his gamepath through the gametree.

This scheme creates a gigantic tree from a small maze.  The number of distinct paths through a maze increases very rapidly with the size of the maze.  There is, however, a fatal flaw in the pure maze scheme: the gametree is degenerate.  That is, a single junction in the maze may correspond to thousands of nodes in the gametree, but the nodes are indistinguishable.  The player coming to that junction for the tenth time faces exactly the same choices that he faced on the first pass through the junction.  If we collapse all the identical portions of the gametree together, it shrinks to microscopic proportions, with just one node for each junction.

There is salvation for the scheme: we endow the maze with state variables that the player acts on as he travels.  The state variables might be changeable by the simple act of travel, or perhaps by more complex behavior on the player’s part, or perhaps by factors outside of the player’s control.  If these state variables change often enough, then each time the player comes to a junction, the state variables will have taken new values, and the node in the gametree corresponding to this junction will be distinguishable from all the other nodes corresponding to this junction.

Examples: Pac-Man and Zork
Pac-Man uses this technique.  There is a simple maze with a few dozen nodes.  There are four groups of state variables: the status (eaten or uneaten) of the regular food-dots; the position of the four ghosts; the status (used or unused) of the power pills; and the state of the player (empowered or not empowered).  These state variables insure that the player’s options at a given location take on new meanings each time he encounters that location.  Thus, suppose that our hero comes to a junction and finds Blinky the ghost just to his left.  At this junction, under these circumstances (state variables), the player would want to move to the right.  Suppose, however, that at a later time he returns to this junction and finds Pinky the ghost just below him.  Under these circumstances he would want to move up — unless, of course, the player is empowered, in which case, he would want to move down. A single maze junction can generate zillions of gametree nodes, all because of the state variables.

Zork uses exactly the same design strategy.  It too has a maze through which the player moves.  The maze by itself would be uninteresting, but the state variables associated with the maze enable it to generate a truly interesting gametree.  When the player enters a given room in Zork, the range of possibilities available to him depends on the state variables, most of which are simple Boolean variables.  If the player has set the appropriate state variable TRUE (by reading the book, breaking the window, lighting the candle, or whatever), then the room will offer him options otherwise unavailable.

There is an interesting weakness in the traditional text adventure exemplified by Zork.  Only a very few of the state variables will be meaningful in any particular room.  Having the book in hand may save your life in one room, but will have no effect whatever in most of the others.  Thus, the number of distinct nodes in the gametree associated with these rooms is limited, and the gametree is correspondingly emaciated.  This suggests that game designers should take care that many of the state variables be significant at each juncture in the game. 

Refutatio
One could argue that the bushiness of the gametree is derived only from the state variables, and that the maze is therefore unnecessary.  I will not defend the claim that mazes are a necessary component of all or even most games.  I will, however, argue that the maze in games of such architectures does serve a useful purpose.  To wit, a maze provides a simple way to express a large number of state variables.  One could, after all, declare some random collection of, say, a thousand bits to represent the state variables for a game.  These 128 bytes could take 10**300 different states, a rich enough gametree for most players.  Mathematically, then, a small amount of effort put into the state variables could generate a huge gametree.

The mathematical richness of this scheme is meaningless if it cannot be communicated to the player, and indeed no player will tolerate a screen packed with a thousand 1s and 0s, with no apparent relationships between them.  The perfect example of this point is the flight simulator, capable of communicating a large number of state variables (airspeed, altitude, pitch, yaw, roll, and the first time derivatives of these) in a direct sensory format.

A maze provides this service in a very simple-minded fashion.  A maze is a device for structuring spatial movement.  It reduces the infinite number of possible movements available to a player on an open plane to a handful of options.  This simplification clarifies the player’s options and makes the game easier to play, even as it gives visual form to the state variables. 

_________________________________________________________________________________________________________

Culture, Idealism, and the Real World
Jim Gasperini
[Jim is senior designer and story guru with Trans Fiction Systems.  His game Hidden Agenda was recently published by Springboard Software.]

Recent articles in the JCGD have made interesting predictions about the future of the interactive medium. Hidden Agenda (HA), our recently published political simulation game, affords an opportunity to test some of these predictions.

First, a brief description of the game. HA puts the player in the role of the Presidente of a simulated Central American country. Aspects of “Chimerica” have been drawn from the recent histories of Nicaragua, El Salvador, and several other countries in the region. An unpopular dictator has just been overthrown. It is now up to you to decide what course the country will take. 

All action takes the form of encounters with characters representing political parties, professions, classes, and other nations. These characters make proposals about what you should do, and tell you directly if they approve or disapprove of your decisions. They may demonstrate outside the Palace if they lose patience with your policies, or attempt to throw you out of office. Players may also read excerpts from domestic and international newspapers commenting on their decisions, or review charts of social and economic indicators. Finally, an entry from an encyclopedia of the future evaluates your performance as Presidente. 

Characters
Chris Crawford, David Mullich, David Graves and others have argued in the JCGD that “the key to the future of computer games” is the creation of interesting, engaging characters. In HA, all action comes in the form of conversations with characters. To make character interaction dominate the game, we have taken the extreme step of entirely dispensing with both the manipulation of objects and the exploration/conquest of geography.

As Graves points out, most works of interactive fiction center on the physical manipulation of an object economy. Usually these games are otherwise devoid of any interactive media, which means that “you cannot talk with (the characters) because there is nothing to talk about.” In HA, we have filled this void with abstract topics of discussion such as land reform, control of the army, shortages of food, and insurgency. We have been aided in our design work by the fact that we are attempting to model a real-world culture.

Culture and Fantasy
A game’s characters will remain one-dimensional artifices unless its designer also creates a believable culture for them to inhabit. One problem with some efforts to create artificial personalities is that they also create artificial cultures, typically planets with social systems concocted to fit the needs of the game. When you create artificial characters in artificial worlds, they must inevitably act according to artificial motivations and have artificial needs and concerns. You’re asking the player to swallow a whole lot of artificial ingredients at once. The game can quickly become too abstract. Even if it is well-designed and fun to play once begun, it may be difficult to convince many players to take the first step.

I agree with Jeff Johannigman when he suggests that the weak popular reaction to such games as M.U.L.E and Trust & Betrayal may be due to their “lack of the kind of fantasy that most people dream of participating in.” But I disagree with his suggestion that the only kinds of games that will succeed are those that might conceivably be turned into a ride at Disneyland. 

There are many intriguing roles that have little to do with Disney fantasies. In Life and Death, published by The Software Toolworks, for example, you take on the role of an abdominal surgeon. Hidden Agenda asks you to imagine that you are someone in the position of a Daniel Ortega. Personally, I would love to see what the Disney designers came up with if they decided to build a ride called “Latin-American Socialist Utopia” for the Land of Tomorrow. The day such a ride opens, however, I promise to buy a permanent ticket on the Bozo bus. 

Stressing Our Advantages
We must be careful about analogies to other forms of communication. Just because we began by borrowing fantasy forms from other media doesn’t mean we are stuck with them forever. The interactive medium offers a radical redefinition of ’protagonist,’ which gives us a fresh approach to material hard to handle with other forms of narrative.  

For example, the writers and directors of plays and films (and to a lesser extent stories and novels) must consider the extent to which an audience will identify with their central characters. Unfortunately, the need to focus on the personal stories of such characters can sometimes warp the approach the work takes to its larger themes, and even deflect it from the original artistic impulse of its creators. 

A case in point is the Oliver Stone film “Salvador.” I have no doubt that Stone sincerely wanted his film to grapple with the conflicts that have produced the war in El Salvador and the US involvement there. To do so, he chose to tell the story of a hard- drinking American journalist of the Gonzo school who wanders down to El Salvador in the middle of the horrific violence of the early 1980’s. The chaos of the protagonist’s personal life mirrors the chaotic conditions around him.

Stone’s decision to focus on such a character is sound commercial movie-making. His audience is more likely to identify with someone from their own culture, no matter how dissolute, than with a Latin protagonist with alien thoughts and motivation. By presenting the character as a devil-may-care drunk who manages to offend just about everybody he meets by speaking the unvarnished truth, Stone sets him up as a mouthpiece for The Truth About What Is Going On, No Matter How Crass. We are therefore prepared to believe the set-piece speeches he delivers later on, criticizing US involvement. His raw recriminations seem more believable than the carefully polished lies told by propagandists from the US Embassy and the Salvadoran government, which we then see blindly parroted by the mainstream US media.

The problem is that the device takes over the story. What really IS going on in El Salvador recedes into the background, while the personal problems of this ultimately not terribly interesting character become paramount. El Salvador, its people and all their conflicts become merely a setting, a place of chaotic, unexplained and apparently indiscriminate violence where Americans are in over their heads and would be wise not to probe too deeply. 

The same thing happens in the Rafael Lima play “El Salvador,” the Joan Didion book “Salvador,” and in fact most of the recent popular treatments of Central and South American conflicts.  With rare exceptions Central America and its people become background, a wild place given over to what appears to be an irrational Latino propensity for violence. The real story focuses on the more civilized liberal gringos who steam upriver and confront the heart of darkness. The horror! 

In HA we use the unique strengths of the interactive medium to approach the subject differently. In this case that means subordinating the particular story of the central character to the larger dramatic conflicts of a highly polarized political system. Problems of audience identification are swept aside. You are yourself the protagonist, and therefore must begin to think like an actor, not like an audience. 

The challenge is to imagine yourself into the role of Presidente, and test your own assumptions about a political situation you may only know from the casual reading of newspaper accounts. Tackling an area of continuous contemporary concern, the game offers some players a relatively painless way of informing themselves further about something they may feel they ought to understand better than they do. At the same time, it offers more informed players an amusing way to test their understanding of the conflicts behind the headlines.

By comparing the various points of view continually available on any issue, the player should soon come to see that when Central Americans fight each other, they do so for reasons that have deep historical roots. The strength of the characters in HA derives in part from the fact that they represent real people in a real culture. Their emotional reactions to your decisions are grounded in the impact similar real-world decisions actually have on many lives.

This is not the only way to give interactive characters significance, but it is one that will have wide application. Let me go out on a limb here and say that, at least for a time, the strongest characters in interactive storytelling will be those drawn from a specific real-world cultural setting.

Pathways
Along with a radical redefinition of ’protagonist,’ the interactive medium suggests a radical redefinition of ’plot.’ HA represents a true instance of what Chris Crawford has called a ’storynet.’

As a sometime practitioner of the art of adventure game design (Star Trek: The Promethean Prophecy), I have a higher opinion of its aesthetic worth than do many contributors to the JCGD. I agree, though, that a game that builds its borders with puzzles uses an essentially linear plot. The challenge to the player comes in overcoming what I call the “resistant plot,” which may involve taking various false turns and detours along the way. But no matter how many detours there may be, once the puzzles are solved the plot is revealed in all its naked linearity. Often (though not always) such a plot turns out to be less than the sum of its parts. At any rate, most players will have no desire to go through the game again. 

Such a game demands a particular kind of audience. The adventure aesthetic essentially appeals to a methodical, scientific mind-set that enjoys testing innumerable hypotheses about what command will elicit what reaction, what spell will open what door. The fact that many of the hypotheses come to nothing is simply part of the challenge to those who enjoy this way of thinking. But most people, at least most adults, would prefer not to have to bang their heads against imaginary walls.

HA offers a truly non-linear plot. There are no obstacles to completing the game. (Though many people will interpret being overthrown in a coup d’etat as an obstacle, it is in fact simply another kind of completion). You play through once, appointing certain people to your cabinet and making a certain set of choices. Then you play again, changing things. Some things will happen the same way. Unavoidably, other things will happen differently. The game is likely to look very different when played with a left-wing cabinet, for example, than it will if you follow a centrist or right-wing philosophy. 

Here we get to a key point:  it is in the comparison of different pathways chosen through the game that it becomes most meaningful. This is true interactive replayability.  Such a structure only works, however, if the player is convinced to care about the characters, to participate in the imaginary culture and to feel accountable for the influence he or she has had upon it.  Playtesters have told us of the sense of responsibility they feel as Presidente of Chimerica. Some become quite upset, for example, if they fail to protect the head of the Mothers of the Disappeared from assassination at the hands of death squads. 

This is a lot easier to pull off if there is already some inherent significance and consistency to the world being modelled. Since my bias is toward real-world information, that suits me fine. At Trans Fiction Systems, however, we are also currently working on other kinds of stories using the same techniques. 

Winning
In most games, the only way different pathways through are compared is in terms of “did this set of choices lead to my winning the game” or “did it help me win most efficiently.” David Mullich calls for “progress toward expanding the game’s ending...beyond a single screen display announcing the results of the final conflict.” 

In HA, whether you are voted out of office, overthrown in a coup, or finish out three years as Presidente, you’re given a look at the “Verdict of History.” This is an excerpt from an encyclopedia looking back on your accomplishments as Presidente from the perspective of the year 2076. 

It is largely up to the player to decide if he or she won. It’s arrogant enough for us to presume to present the complex conflicts of another culture in the form of a computer game. For us to baldly assert that certain political philosophies are right and others wrong would be much too much. Thus, we present a text compilation that you must interpret for yourself.

One player may be dismayed that his policies failed to bring down the Infant Mortality Rate. Another might not care a hoot about health statistics, looking only to discussion of her economic programs in deciding if she has or has not won. Though many people will likely feel that they “win” if they are elected to stay in power and that they “lose” if they are voted out, the Verdict may make them wonder if such a conclusion is always warranted. 

Why we don’t get no respect
At an EA developer conference a while back, Trip Hawkins posed a challenge:  “design a game your father might want to play.” Well, my father was an attorney deeply concerned with the issues of the day. His favorite TV shows were “Meet the Press” and “Face the Nation.” At the time, I was already deeply into Hidden Agenda, but Trip’s remark made clear what I am trying to do. I’m trying to broaden our audience, attempting to include people like my father, people who would rather read “The Fate of the World” than “The Lord of the Rings.”

I know a number of people for whom Balance of Power is the only game they have ever purchased. They dismiss the rest of the genre as simple-minded adolescent fantasy. Personally, I like Tolkien, and I like SF, but there are limits to how many times we can milk the same great cliches. SF and fantasy forms have had to struggle to be treated seriously in the world of literature. Should we be surprised if they are accorded even less respect when they appear in the form of games? 

We need to lure in people who have never played a computer game, who are busy with other concerns and have to be convinced that such a thing might be worthwhile. Many adults, like myself, once read fantasy because they found the real world (or at least their own little corner of it) rather dull. Now they see the world as a complex and fascinating place, and prefer entertainments that at least pretend to help them understand a little more of it. 

Idealism
Brenda Laurel’s experience has taught her that “learning doesn’t sell.” Well, HA is unabashedly a game that you can learn something from. It’s also fun to play. Am I the idealist she describes, popping up once again with naive “haunting fantasies of interactive experiences that have the power to change minds and hearts?” We’ll know soon enough. 

I believe that in the history of narrative the development of forms based on computer technology may ultimately prove more significant than the development of forms based on celluloid and videotape. Interactive forms will become one of the most important means of cultural expression in our society in the twenty-first century. 

Interactivity will only come into its own, however, when we begin to apply these powerful tools to subject matter worthy of their sophistication. If we truly care about the future of our medium, we must learn to care about the rest of the world. 

_________________________________________________________________________________________________________

Salary Survey Time!
Yes, it’s time for our annual salary survey.  Every year the JCGD carries out a salary survey of its readership and publishes the results.  What has changed in the last twelve months?  Has our financial return become better or worse?  Here’s our chance to find out!My approach this year is less structured than last year.  I will not provide a postcard for you to fill out.  I ask you to do a little more work.You should not send your information unless your primary income-producing activity involves the creation, design, or programming of computer games.  In other words, if you have a day job and work on games nights and weekends, don’t count yourself; this would bias the results.  OK, so you consider yourself to be a full-time computer game design professional.  Figure out your net income for 1988.  That would be the total amount of money that you earned in 1988, minus any money that you passed on to other computer game design professionals.  Thus, if you lead a team of computer game design professionals, and all the income for that team flows through you, count only the money that you kept for yourself.  What I want here is your net personal income.  In a change from previous years, I am asking you to provide your net income instead of your gross income; this means that you should deduct your business expenses.  This should make it easier to directly compare the incomes of salaried people with those of freelancers. OK, so you have a number in mind, representing your net personal income from computer game activities only.  Write that number down on a piece of paper or postcard and send it to:Dave Menconi357 Spring Valley LaneMilpitas, CA  95035Dave will take your number and write it down with all the others.  I will receive only Dave’s tabulation.  Since Dave doesn’t have access to the mailing list, and I don’t have access to the original letters, there is no way that your income can be linked to you personally; your privacy is protected by this scheme.I very much hope that you will take the time to respond to this salary survey.  It is one of the most useful services that the Journal provides, but its value is dependent on the willingness of the subscribers to share their information.  Please, take a moment, estimate your income (or just look it up on your tax return for 1988) and send it in to us.  I’ll  be printing the results in the next issue of the Journal.

_________________________________________________________________________________________________________

Publishing Contracts—A Few Notes
Mark Lewis Baldwin
[Mark is the author of Empire for the IBM PC.]

One of the big mysteries in the computer entertainment world is the publishing contract — the little paper that lets a stranger meddle with our great work of art in exchange for a few green pieces of paper.  Now I’m not a lawyer, so I can’t advise you about the legalities of a contract.  However I am a game designer and sometimes help other game designers in their negotiations.  I have therefore build up a list of items that I look for in a contract with a publisher.  The list in not all inclusive, but it may help the next time you negotiate a new contract with your publisher.  Also, remember, there are an infinite number of solutions to the problem of arranging publication.

A contract is an agreement made jointly between yourself and the publisher.  That means that you have just as much responsibility in designing the contract as does the publisher.  Don’t accept their boiler plate contract as gospel.  The publisher’s version of the contract is biased in favor of — guess who — the publisher.  Look at it, massage it, modify it, rewrite it, or even write your own.

The contract that you are negotiating is much more than evidence to be dusted off in the event of litigation.  If the contract is prepared properly and covers all the bases, you should never find yourself in court.  That’s because the real purpose of a contract is to spell out each party’s rights and obligations.  If each such right and obligation is clearly laid out, discussed and decided on during contract negotiations, there should be no misunderstandings.  All parties will be satisfied and you will be  able to talk to them again about your next  magnum opus.

When you originally start negotiating with the publisher, your primary concern is the royalty.  Normally, this is a percentage of either the wholesale or suggested retail price.  US publishers tend to base royalties on wholesale while European publisher (as well as most book publishers) tend to base royalties on retail prices.  Each has an advantage and disadvantage.  By tying your royalty to the wholesale price, you are connected much more directly to the publisher’s business dealings.  For example, one of my  publishers recently negotiated a higher whole-sale rate to the distributer.  This directly increased the royalty amount I received per copy.  On the other hand, if the publisher decides to sub-lease production at a lower rate, you could be burned.  If your royalty rate is tied to the retail price, neither change would bother you.

Both schemes fail when your product is at the end of its market life.  When the publisher decides to dump the product at a reduced price, your royalties plummet.  You might want to  consider a minimum fixed price per copy.  But remember that the publisher might in that case choose to drop the product completely, without even clearing out the old merchandise.  You may make more at bargain basement royalties than no royalties at all.

Sub-leases and wholesale royalty rates add new problems.  A publisher has a certain cost to manufacture your game.  If the publisher sub-licenses the product at the difference of the production cost, it is removing that cost and could conceivably be increasing its profit at your loss (since you are not receiving royalty on the portion of the wholesale cost allocated to production).  This may seem confusing but it is very real.  My solution has been to have a second royalty tier for sub-licensed products.

What royalty rates are acceptable?  I’ve seen wholesale rates as low as 3% and as high as 35%.  The norm seems to be between 7.5% and 25%.  As time goes on, I expect (hope) to see the low end disappear and settle down to between 15% and 30%.   Depending on the publisher’s distribution, wholesale price will be  anywhere from 25% to 50% of retail price. Be aware of these numbers when negotiating with the publisher.  Also, get a very clean definition of what ’wholesale’ means to the publisher.  This could be monies owed upon shipment; monies received from previous shipments; monies received less taxes, shipping, returns, etc. or something else equally strange.  Obviously this definition is very important, or you may end up owing the publisher for each copy shipped!  Tying royalty to suggested retail sounds better and better, doesn’t it?

Another thing to realize is that the publisher may wish to have an escalating royalty rate depending on the number of units sold.  This helps the publisher cover the initial publication cost.  It also gives you a bonus if your game turns out to be a hit. 

When do you receive the monies owed to you by the publisher?  What you’d like is 30 days after the calendar month when the debt is accrued.  But I have seen 45 days, 90 days and even yearly royalty cycles.  Can you afford to wait that long while the publisher uses your money?

What will the publisher pay you for income from conversions?  50% of the original rate seems to be common, but this can also be negotiated based on how easy/complex the conversion is.  I insist on a clause that gives me first option to execute the conversions myself.  This way, if I do the conversion (or at least control it, I have converters working for me), I can insure the artistic content and quality of the conversion.  Otherwise, I’m at the mercy of the publisher.  Furthermore, I like to design my code with conversions in mind.  If I do the conversion as well, the cost should be a great deal less than the 50% I would have to pay the publisher otherwise. 

In return for all this, you give the publisher an exclusive license to market your software.  The license may also include rights to conversions and derivative works.  The license may include sub-licenses or not.  It may be constrained only to a specific country like the USA.  Make sure these items are spelled  out.  Also, specify who owns the motion picture rights (you laugh!).  You may also be required to provide documentation (or at least, draft documentation) and maybe a secondary program like a functional demo.  You should also be responsible for a certain amount of maintenance and bug updates.  I would be suspicious of a publisher who did not want maintenance, since it would imply to me that the publisher does not support his customers.

Another item that the publisher may desire is first option and first right of refusal.  First option means that if you write another game, the publisher has first chance to bid on the product.  First right of refusal means that before you sign any contract with another publisher for a new project, your current publisher must be offered a chance to meet the same terms of the contract.  The reason a publisher would want these clauses is that he has invested time and money in establishing your name and product and would therefore like to keep you in the family.  I don’t particularly have a problem with first option because I want the publisher to feel that I will be working with him for the long haul, not just for one product.  However, I feel that first right  of refusal is a dangerous clause.  The publisher could effectively damage any negotiating position you may have with another publisher since your original publisher can always come back with another bid.  And most publishers don’t like to negotiate knowing that the entire effort could be negated by another publisher.  If your publisher insist on a first right of refusal, you better feel that you will be happy with the publisher for the long haul, and even then, make them pay through the nose for the privilege.

Other items you may want to consider in the contract include:  

 - do you have the right to audit the publisher’s records with respect to sales of your product?

- who gets the copyright and who is responsible for maintenance of the legal aspects of the copyright?

- make sure all the industry words/terms are specifically defined and are unambiguous.

- what responsibilities does the publisher have in marketing (I require them to spend x dollars advertising my product).

- how long does the publisher have to release the product once you deliver the final product to the publisher?

- what veto rights if any do you have on packaging, marketing or conversions?

- do you receive any complimentary copies of the final product, and if so, how many?

- who is responsible for law suits for damages caused by the product (don’t laugh) and enforcement of the copyright.  Who bears the cost and penalties?

- who is responsible for any cost involving recalls and updates? 

- will you be receiving advances on your royalties?  If so, are they refundable to the publisher if your actual royalties to not match the advance?

- what to do about both your trade secrets and the publishers trade secrets.

- how is the contract terminated?  What are the responsibilities of each party upon termination?

- does the publisher get a copy of the source code?  What can he do if you refuse to (or cannot) maintain the product?

- will the publisher provide you with any design/code assistance or assets?

-  how do you handle disagreements involving the contract.   You really don’t want to go to court about every minor misunderstanding.  It’s expensive and boring.  Arbitration is normally an accepted procedure if all else fails.  Make sure you specify who will do the arbitration (like the American Arbitration Association) and where (preferably your home town/state).

_________________________________________________________________________________________________________

How to Expedite Conversions
Gregg Tavares

[Gregg has worked at or for M.U.S.E, MicroProse, Designer Software, AtariSoft and is presently working at Cinemaware.]

After reading the recent article entitled “The Conversion Malaise” I felt compelled to explain why I think many conversions are not up to snuff.  I have personally converted three products to different computers and I have participated in the conversion of at least seven others.   I have conversed with many other people who have converted programs.   With such experience I feel qualified to write this article explaining the conversion process, its problems and steps to improve it.

It has been my experience that converters and conversion houses will always do the minimum amount of work required to get the product finished.  Once a contract has been signed it is to the converter’s advantage to get the product finished as quickly as possible.  This is because most converters are paid a flat fee.  It has also been my experience that nobody likes to do conversions (at least nobody I ever met).  For most programmers it is just a quick way to earn some money.  Moreover, most programmers who created the original products don’t want to look at their own code once they are finished with it.  They want to move on to something new and exciting.

Here are my suggestions to make conversions less of a pain and to get more quality out of them:

Developers
Write your code cleanly!  Separate the machine-dependent from the machine-independent.  There is rarely a reason that this cannot be done and even when it cannot you should at least explicitly comment the area of code in question and explain what the code is attempting to accomplish so that a converter can understand it and code around it.  One of the most tightly coded games I’ve worked on was Gunship, yet the coding was so clean that it was easily converted to other machines. 

A specific example might help.  When the program needs a keypress it should call a function.  The converter needs only to replace that function rather than searching through the entire program to find every case in which you access the keyboard.  

Similarly, when you draw to the screen, try to set aside special routines to do the actual drawing.  Then document what gets passed or what globals are looked at and how they are used.  If you are using special graphics objects like hardware sprites, document the locations of your position and shape variables. The same goes for scrolling. Document where in the code the values of the scroll variables are valid and represent what is currently displayed on the screen.  Call routines for sound that could be easily replaced by other sound routines.  Document all your variables explicitly!  (i.e., This variable is used for this ... during this part of the program it is used for this ... during that part of the program it is used for this ... It is only valid after calling that routine.)

Comment your code!  It doesn’t matter if you comment it during the project or after it as long as you comment it well — well enough that another programmer could understand exactly what each routine is trying to do without looking at the code itself.   At the very least, comment each routine, what it is trying to do, what variables and inputs it uses, and what variables it changes and results it produces.  This is most important for the machine-dependent portions of your program.  Some of the machine-independent parts may be treated as black boxes, (i.e., something that should be converted instruction for instruction without understanding) but it is still better to comment that too. 

In short, write your code and comment it so that a converter can quickly convert the game logic as a black box and then replace the graphics, sound and I/O routines with machine-specific code.  

Sit down with someone who knows the target machine and design the conversion in detail.  Look at the capabilities of the target machine and decide what should be added, improved, or cut from your original design.  Remember, machines like the Amiga, Mac and IIgs use digitized sounds.  Would super sounds help create a better atmosphere in your game?  These sounds also require lots of memory.  Do you have room for them?  Machines like the IBM and Apple II have lousy sound capabilities.  Does your design require sound?  Can you change the design so it doesn’t?  

Can the target machine handle the original design?  A 3-D flight simulator from a 16-bit machine may not even be possible on an 8-bit machine.  What compromises must be made?  Decide how you would do the game on the target machine if you were to do the conversion yourself.  Would you use sprites for this effect?  Would you use a display interrupt/copper list/raster interrupt to get another? Discuss your opinions with the converter in detail and find out what is possible and what is not.

Publishers
Make programming comments and conversion design a requirement for final advance payment or first royalty payment and get it checked out by the converter.  If you don’t require conversion design you will not get it and your conversions will end up taking two to three times longer!  Decide whether or not you think the artwork for the product should be done in house, by the converter or by a third party.  In my experience it is best to do the artwork in house as it is the one part of the product that can be evaluated on a daily basis and modified if necessary.  Converters are in it for the money and not for promotion of the Arts.  Most developers should be able to convert the product without the final artwork.  As long as some approximation of the artwork is available (like artwork ported from the original game or even just colored rectangles the approximate size of the final artwork.)  Then, near the end of the project, the final art can be installed into the product.

Converters
Make sure to know what you are getting into!  Take a look at the product and ask how you would implement it on the target machine.  Find out who will create the artwork, sound and copy protection, if any.  Examine the source code and determine how the program works, where the pitfalls might be, how understandable the code is. Look especially at the quality of the comments, remembering that quantity does not necessarily denote quality.  

Above all, NEVER ASSUME ANYTHING!   In the last product I converted both I and the Publisher assumed too many things and now we both regret the experience.  In preparing my bid, I examined only one of the 25 source code files.  The one I looked at was well commented.  The 24 I didn’t look at had almost zero comments.  They supplied me with some artwork that I assumed was in the correct proportions for the game and I also assumed it was the artwork they wanted me to use.  It was not.  I assumed that all the shapes for the animation in the game was represented in the artwork  I was missing 92 of 182 shapes!  I told them on my contract that I would “Enhance Selection Graphics”. They assumed that meant I’d redraw the graphics for the entire game!  I assumed that this program was written with easily replaceable graphics routines.  Wrong again.  

All this has served to make the product late, making the publisher very unhappy.  It also obligated me to do much more work then I had ever intended.  So I am very unhappy as I finish this product for the lowest financial return on time invested that I’ve ever gotten.

Conclusions
I’m sure that most of these ideas are obvious and have been said before but why then do so few people follow them? Remember that if everyone would follow these simple suggestions conversions would be much more likely to turn out better and faster.  Don’t be penny wise and pound foolish.  Two weeks spent commenting and generally making the product conversion ready will save months in the conversion process.  

_________________________________________________________________________________________________________

Great Expectations
Or, How to Make a Living in Game Development

Gordon Walton & Karen Hunter of Digital Illusions, Inc.

Introduction
In this article we will look at the major advantages and pitfalls of being a self-employed game developer.  Only by surviving the process can you see your work make it out into the marketplace!

Getting Started
First you need to decide if you are up to the challenges of working for yourself.  It is going to require that you possess an ego the size of Muhammed Ali’s, the patience of Job (not Steve Jobs!) and the perseverance of a pit bull.  You will need programming expertise on at least one of the major microcomputers.   You will also need adequate development equipment — hardware and software.

You may think that only superman (or super-woman) can be a game developer.  Actually, while superpowers are desirable, you mainly just have to be willing to go the extra distance to build the product of your dreams, even though the problems may appear insurmountable.

Preparing a Proposal
Why do you want to be a game developer?  Do you have a idea for a game of your own, or would you just like to implement other peoples’ games on your favorite machine?  We will cover both possibilities below.

Programming Other Peoples’ Games
This is the easier course, and we recommend it to anyone who wants to break into the business.  Having a track record of completing a product, even a translation, will increase your credibility immensely when dealing with publishers.  We suggest that you contact some of the major and minor publishers, letting them know about your abilities and your desire to do a conversion for them as a stepping stone to doing an original product.  Once you find a publisher who has a product that you have an interest in converting, ask them to send you the information you will need to make a bid on the translation.  Note that before you see this information you will have to sign a non-disclosure from the publisher.  Prepare your bid proposal carefully, giving yourself adequate time to achieve each milestone.  Don’t bid too low — you will need to support  not only your ongoing expenses but also the extra expenses caused by self-employment.

Your Own Game
For some this course is better, particularly if you really believe in your game idea.  Unfortunately, unless the idea is both novel and obviously a winner (a rare combination), you have a lot of selling to do!  Prepare two proposals, the first being a one- to three- page summary of the primary thrust of the game, its key features and benefits, and any graphics that will help you sell the concept.  You will use this proposal under non-disclosure with a publisher to see if they have any interest in your game.  The second pro-posal is a full specification of your design, including an outline of the entire flow of the game, sketches (or storyboards) of all graphical effects in the game, how long each development milestone will take, how much funding is required, and may even include the logic that will run the game.  The more detailed and well-defined this proposal is, the better your chance of getting a contract and producing a winning game to a schedule.  Do not let a publisher examine a copy of this second proposal, except in your presence, without having a signed contract.

Getting Your First Contract
Once you have found a publisher who likes your proposal, you must negotiate the contract.  We have found this to be the most odious and time-consuming part of game development.  All publishers have standard development contracts that contain lots of verbiage that protects them, and very little to protect you.  The first rule is: do not accept any contract you don’t like or understand.  Never accept exclusivity or “first right of refusal” clauses, as they will definitely limit your future business prospects.  Accept late penalties only if you can afford to pay them.  Ask a contract attorney to explain each portion of the contract you don’t fully understand.  Use the attorney’s expertise to come up with alternative wording that will protect your interests.  If the publisher wants your game, they will change portions of their contract.  At the same time you must be sensitive to the areas where the publisher will have little or no flexibility.  Be prepared to be patient or willing to settle on a contract you don’t like.  Our worst case took 2.5 years of negotiation to reach contract signature.  The topic of the game under negotiation changed three times during the course of negotiation! 

Doing the Work
This is the easiest part for most game developers.  Game developers tend to be technology and game play oriented, and this part of the deal is right up our alley.  The main pitfall most developers fall into is that they are eternal optimists and drastically under-estimate the time required to complete the product when making the initial proposal.  A detailed design document, combined with conservative time estimates is the only way to avoid this common problem.  Another problem is the very human tendency to put off until tomorrow something you could do today.  The only way to survive and prosper in game development is always do something today that gets you closer to shipping the game.  Your consistent goal should be to get as far ahead of schedule as possible.  After all, the sooner you finish the product, the faster you get paid.

Getting Paid
Next to negotiating contracts, this is the second most despicable part of game development.  We will talk about two different types of payments, milestone development payments and royalty payments.

Milestone Payments
If you are being paid advances or fees based on milestones the only way to ensure getting paid in a timely manner is to have well-defined milestones in your contract.  Loosely defined milestones give the publisher the ability to ask for more than what was intended before you get your milestone payment.  At the same time, you as a developer must not try to pass off an incomplete milestone as being complete.

Royalty Payments
Some publishers pay royalties on time, but many do not.  The only way to ensure that you get your royalties as prescribed by your contract is to add a penalty clause for late royalty payments.  Every publisher will assure you that there is never a problem, but my experience dictates the opposite.  The other important aspect of royalties is for you to be very clear on what amount you get paid on.  Different publishers compute the gross amount on which they pay royalties in many different ways.  There have been rumors of some very creative accounting practices that had the net effect of reducing authors’ royalties, so be careful.

The Taxman
So you have gotten and done the work, and even gotten paid!  Now comes the really hard part: doing all the administrative work so you don’t become an unwitting tax evader.

If you are an independent contractor (i.e. you do not have an employer taking out taxes during the year), you must file a Form 1040-ES (with a payment enclosed) on April 15, July 15, October 15, and January 15 for the present tax year.  You must discipline yourself to do your own withholding, and planning ahead is the key.  As you get paid (whatever form this takes: milestone payments, advances, royalties, etc.), you should put 20-25% in a separate account.  If you do this, then when it is time to file your quarterly estimated taxes, you will have the money.  The total estimated taxes paid (after the January payment), should equal 100% of the previous year’s tax liability or 90% of the present year’s estimated tax liability.  Keeping your records organized is very important so that you do not underpay.  The penalties for not paying (or under-paying) estimated taxes are stiff, as noted on form 2210 from the IRS.  When filing your Form 1040, you will have to fill out a Schedule SE to determine the amount of self-employment tax you will have to pay.  The self-employment tax is equivalent to social security taxes.

The total taxes you must pay as an independent contractor are more than you would pay as an employee since your employer would normally pay half of your social security tax.  To offset the higher taxes, you can deduct all business expenses including your computer hardware, software and accessories.  The expenses are put on Schedule C, and the equipment is depreciated on Form 4562.

This extra tax work and planning is one of the primary problems in being self-employed, so beware!  If you already file form 1040 instead of 1040EZ or 1040A, and/or already are self-employed, this may not be as big a problem for you.

Summary
This article should have given you an overview of the challenges and potential pitfalls of being an independent developer.  Our hope is that you don’t have to learn as many painful and expensive lessons as we have during the last four years.  Good Luck! 

_________________________________________________________________________________________________________

Once Upon a Time
The Untapped Potential of the Computer as Story-Teller

Bill Hopkins

[Bill is a computer game designer, author of several games for the Air Video in-flight game system and co-founder of Free Association Development, a computer game design house based in Toronto.  Bill can be contacted through non-electronic mail at P.O. Box 291, Station O, Toronto, Ontario, Canada, M4A 3N3]

Copyright 1989 by Bill Hopkins

Several articles in this journal have remarked on the newness of the computer as an entertainment medium and how we must take the time to recognize its special properties.  In “Similarities with Other Media” (JGCD, Volume 1, Number 5) Chris Crawford warned designers of the dangers of being drawn in by the computer’s apparent similarities with film, and producing “computerized” movies that are neither good films nor neither good computer entertainment.  In “Computer Games /Computer Entertainment”  (JCGD, volume 1, Number 7) Stephen Linhart suggested that in addition to games it would be possible to develop a new form of computer entertainment that would appeal to a wider audience, while still exploiting the interactive properties of the computer medium.   In “Learning from Fiction (JGCD Volume 2, Number 3), Greg Costikyan discussed the relationships of games and narrative fiction, suggesting how game designers might successfully employ some of the techniques of fiction writers while avoiding the pitfalls of complete stylistic imitation.  In ”A Techno-Wimp Revisits ‘Scribes/Authors’” (JGCD Volume 2, Number 3), Chet Day discussed his excitement over the potential of the computer as a storytelling medium and his hopes for “the first new genre of literature in many years.”  All of these authors strain to point out the importance of recognizing the computer as a separate and valid medium for entertainment.  While most of these articles are primarily concerned with creating better computer games, they are all exploring the properties of the computer as an unique artistic medium.  The next step is to consider what other forms of entertainment this new medium can support.  I believe that it is possible to develop a theory for a new form of computer entertainment  by examining the most basic elements underlying the more popular forms, and determining how the entertainment of a game differs from that of a book or a film.

Games are currently the only form of computer entertainment available to the public.  While computers certainly provide a wonderful venue for games of all types, and many of the computer games of the last few years have vastly expanded our ideas of what a game can be, the potential of the computer as a medium for popular entertainment has hardly been touched.  No matter how clever, no matter how detailed, a game remains a game — a form of entertainment enjoyed less often and by fewer people than entertainment in the forms of movies, television, plays and books.  What separates a game from these other forms is its interactive nature.  Movies et al. require the audience’s attention and nothing more.  Of course the finest works in all of these forms will provoke emotional responses and focus the audience’s attention on the issues that underlie the story presented, but no active response on the part of the audience is necessary.  The creator of this type of entertainment seeks only to titillate and perhaps educate;  the audience seeks only to experience.  

Games, on the other hand, require the player to react in specified ways under all sorts of restrictions or rules.  Games are fun because they are interactive, challenging (even difficult), occasionally frustrating and frequently surprising.  They stimulate the players’ minds and emotions in an aggressive way by presenting them with a specific task and some sort of competition.  Games are unique as an entertainment form because they threaten and challenge their audience.  Games are driven by the players, allowing (forcing?) the players to create their own enter-tainment within the framework of the game’s design.

In contrast, the major entertainment forms entertain by telling a story.  Now, some may say that the more complex computer adventure games are telling stories in their own way, but if they are, something crucial is missing.  The best stories survive multiple re-tellings, but few people will bother to play a computer adventure more than once.  Why?  Because a computer  adventure is at heart a game, a game of linked obstacles that are difficult to solve only the first time they are encountered.  The storyline exists only to give the puzzles of the game some sort of framework that shapes the player’s experience and enhances his enjoyment.  Moreover, the story-line of a computer adventure does not significantly develop plot or character, the central elements that allow the enjoyment of a story even though one has heard it a hundred times before.

So where does all this lead?   It points to a form of expression and entertainment that is currently unexplored in the computer medium.  Why can’t a computer tell a story as well as a book or a film, and in its own unique way?  Every medium can tell certain stories better than any other; everyone has had the disappointing experience of seeing a film based on a favorite book that fails utterly to tell the same story in a convincing fashion.  Computers offer the chance to develop a new form of story telling, one that can tell stories in a way unlike any other medium.  

The first attempts to tell stories on computers have focused on their interactive power, the very element that makes a computer such an excellent vehicle for games.  But computers are capable of more than just interacting with the user;  they are also able to control the way in which a user interacts with them and to modify the presentation of data in a controlled way, according to the author’s instructions.  This ability to change and adapt a story is perhaps where a computer’s story telling ability truly lies.  When a film or a book has been edited for the last time and is released to the public, it can tell one story and only one story in precisely the same way.  On the positive side, this gives the work’s creator a great deal of control over the response of the audience and allows him to present a particular message and encourage an interpretation of the meaning of the story he is telling.  If the work is rich enough there will be those who will want to experience the story over and over again, discovering new levels of meaning and new beauty each time.

A computer story need not have such limitations.  The author could create multiple plot lines that are nearly parallel, each subtly modifying the experience but presenting the audience with a story that creates roughly the same emotional response.  Or perhaps the author would choose to incorporate vastly different points of view of the same story, presenting the perspective of the hero or the villain, depending on the actions of the audience.    Unlike a game however, the sum total of possible events would be under the control of the author who would have carefully determined beforehand the order of the central events and what sort of treatment those events would be given.  The audience would experience the story as they would a book or film, enjoying the artful telling of a tale in a masterful way that fully exploits the properties of the medium.  A story told in this fashion (if told artfully enough) would bear re-telling because its entertainment value would lie in the specific events and characters of the story, as well as the overall atmosphere created by the author’s style of story telling.

While it still remains for us to discover the techniques and conventions that will allow an author to tell his story on a computer, the computer would seem full of promise as a story telling medium.  A computer can present the carefully crafted lines of text found in books and combine them with the visual imagery, sound effects and musical score of a film.  All of these elements can then be controlled  by the computer to present a vast array of different but consistent experiences of the same story.  The great themes of human experience, both comic and tragic, can be explored in new ways and with new intensity using this unique medium.


A  Rebuttal from the Editor
In two years of editing the Journal, I have never rebutted an article, but my editor’s abhorrence for a textual vacuum demanded some kind of filler. 

My first objection falls on the final paragraph. It seems to see the computer as little more than a multimedia machine, a film projector with text.   There is no intrinsic reason why text cannot be included in conventional film.  The fact that it is rarely used suggests that there is a good artistic reason why text and image do not work well together.The final paragraph also waxes eloquent on the vast array of possibilities available to the computer artist.  Even using a compact disk, the data capacity available for assembling this vast array of possibilities falls far short of that available to a film director.  Just what is the computer bringing to this party?

I question the value of multiple, nearly parallel plot lines, and multiple simultaneous points of view.  If we can’t tell a story well through a single sequence of points of view, how will adding more points of view make it better?  All we manage to achieve is boredom as the audience slogs through umpteen different versions of the same event, umteen-1 of which are too distant to be compelling.

I deny the value of a fixed, hard-wired story on a computer.  I argue that the story must be soft-wired, i.e., generated at runtime from storymaking algorithms responding to player actions.

Chris.

_________________________________________________________________________________________________________