Jack of all trades vs the one-trick pony

Cast your mind back to the start of the year, and a discussion we had about how digital interactive fiction is breaking out of the gamebook ghetto by using maps, comics, animations and audio instead of prose. Or as well as prose, anyway. Among many good points he made in the comments, Emanuil Tomov said this:

"There's a useful tension, even a moral one, in reconciling minmaxing with possible unsalutary effects this could have on the narrative. Imagine a system where you can either spend or hoard XP, spending being linked to short-term benefits, hoarding being linked to 'advancement' through a richer background for the character, learning skills through interesting narrative that deepens your understanding of the character, forming certain bonds with powerful allies, etc. XP literally represents your experience in the world. Now imagine a PC who spends all their XP short-term and they're really, really, really good in a pinch; but they're completely flat, a competent, one-trick bit player in a story where they could've been much more. It's an interesting trade-off both on the minmaxing and the narrative front.

This point about the value of versatility reminded me of a section in Game Architecture and Design, which I co-wrote with Andrew Rollings to get the horrors of working in the trenches at Eidos out of my system.


The book is twelve years old now, and games have sure moved on, so don't feel you have to run out and buy it. By the time I was working at Elixir Studios, only a few years on from the Eidos of the late '90s, software development for games had been completely revolutionized, and that made my job as designer a pure pleasure. But I digress; we were talking about versatility...

Versatility in gameplay
A useful rule-of-thumb for anticipating gameplay is to ask what is the best and worst thing about each of the player’s options. For instance:

• This maneuver does the most damage, but it's the slowest
• This maneuver is the fastest, but it leaves me defenseless
• This maneuver gives the best defense, but it does little damage

And then there's a unique kind of choice:

• This maneuver is never the best or the worst, but it's the most versatile

So a useful question to ask yourself when designing a weapon or strategy for your game is "When, if ever, is this the best option for the player?" Most choices that you put into the game should be the best in some way. And one of these can be the choice that works only moderately well, but in many different ways: the jack of all trades option.

The more unpredictable the game environment, the bigger the payoff for having versatility of choice. Beginners in particular will benefit from versatile options in a game, as it means there's something they can do while working their way up the learning curve. But versatile options are handy for expert players too. When fighting an expert opponent, you must expect the unexpected, and choosing the versatile maneuver or unit may buy time to put together a more considered response.

One obvious kind of versatility is speed. The fast moving character or unit can quickly go where it's needed. So, normally, you won't want the fastest units to also be the best in other ways.

Also, the value of a fast-moving unit depends on the game environment. On the battlefields of the 14th century, a knight was deemed to be worth 100 foot soldiers. That wasn't because knights were each individually as tough as 100 men, but rather because, in a terrain of hedgerows, ditches, ploughed fields and heathland, the knight had more chance of being at the right place at the right time.

There are many other ways to make an option versatile. If a beam weapon can be used to mine asteroids as well as to destroy incoming nuclear missiles, then that versatility can make up for a disadvantage elsewhere. Of course, if there is no compensating disadvantage, there's no interesting choice. Be careful not to make the versatile choice dominant over all others. Also, be aware that the versatility of a choice may not be obvious even to you as designer. In the last chapter, we saw how the designer of the fantasy game Arena hadn't originally anticipated the way players might use the fireball spells.

You can measure versatility by looking at the switching costs in the game. This is how much it costs a player to change his mind about the strategy he's using. An example in an espionage game might be if you recruit a spy and later realize you need an assassin instead. The switching cost is however much you wasted on the wrong character, assuming for the sake of argument that the spy is not usable elsewhere. So, say that both cost $1 million. When deciding which to buy, at first you'd think, "If I buy the spy and I need the assassin, I'll end up paying $2 million. If I choose right, it costs me just the $1 million. On the other hand, suppose I buy both now. I only need one, so I'll have definitely wasted $1 million."

Now suppose there is another character, the ninja, who can function as either spy or assassin. How much should the ninja cost? It depends how unpredictable the game is. In this example, if the game were completely predictable, the player would know in advance which character to recruit and so versatility is of no value - the ninja should cost $1 million just like the others. In a completely unpredictable environment, the average cost would be $1.5 million ($1 million if I choose right, $2 million if I choose wrong), which is what a good gambler would pay you for the ninja. Since the truth will lie between those extremes, the versatile unit should cost more than $1 million but less than $1.5 million.

Versatility is more prized in an uncertain environment. No multiplayer game is completely predictable, since you can never know what the other player(s) will do. Even in a relatively predictable game, some levels are more uncertain than others. All of which makes the choice between specialization or versatility an interesting one because it all depends on the circumstances.

Designing from the bottom up

More tips on game design today culled from my and Andrew Rollings's book Game Architecture and Design. (Sure, it was published a whole fifteen years ago and game development has moved on - but hey, some people still get advice from the Bible and the Koran.)

If you throw a ball and take many high-speed photographs of its flight, you'll see that the trajectory the ball took is a parabola. But the ball didn't follow that path because gravity told it to move in a parabola. A parabola is just a symbolic concept in the analytical domain of mathematics, and the universe doesn't know anything about mathematics or analysis or symbols. These are human concepts. In reality, there are just a bunch of physical processes, each of which deals only with the processes and circumstances just before and just after it. So, the ball is at one position, and gravity tells the ball's velocity to change, and the ball's velocity tells its position to change.

This is the opposite approach to that taken in most software applications. There, processing power is at a premium, so the sooner you can go to symbolic constructs, the better. The tradeoff is that software can crash when your symbolic "shortcut" misses something that the one-step-at-a-time approach would have taken in its stride.

Researchers in artificial life have identified an analogous problem:

"The classical AI approach has been criticized because the symbols and symbol structures on which planning and decision making are based are not grounded in the real world. The problem is that unequivocally decoding sensory data into a symbol and turning a command without error into its intended action may be unsolvable."
—Luc Steels, "The Artificial Life Roots of Artificial Intelligence" in Artificial Life (MIT Press, 1997)

Here is an example: suppose you are putting a monster into your new Frankenstein adventure game, and the idea is that it will jump out of its vat when the player enters the laboratory. Instead of putting in a lot of complicated AI to do with detecting humans and having the goal of wanting to kill them, you just choose the shortcut of placing a trigger tile inside laboratory door. When the player steps on the trigger, the monster will appear and attack.

Okay so far, but what if the player manages to get onto the tower roof, jumps down, and, by some fluke, manages to land safely on the balcony of the laboratory? Now he can explore the lab, get all the power-ups, and read the journal about the monster (an entry that is supposed to be poignant if he's just fought and killed it, but is meaningless otherwise). Only when the player goes to leave via the door does the monster climb out of its vat and growl, "You shall not steal my master's secrets!"

When cutscenes were pre-rendered, ones of the purposes of tying events to a trigger point like that is that you could be sure of where the player’s character would be standing (and the state of the laboratory, in this example) when the cutscene began. Nowadays, it is more likely that the cutscene will be generated in the game engine. The cutscene thus becomes an example of “machinema” which can be slightly different every time, depending on the game state at the time it is triggered.

The use of the trigger point illustrates symbolic design. The designer assumes there is only one way for players to enter, and that’s via the door. The alternative nonsymbolic, or equation-free, approach would recognize that the true trigger event is the monster’s awareness of intruders in the laboratory. Whatever way the player enters the lab — even if by teleportation — the game still responds appropriately.

Discussing Deux Ex at the GDCE conference in London in 2002, designer Harvey Smith of Ion Storm cited how nonsymbolic design is changing games. In testing a maze level (the walls of which were set high enough that the player couldn’t jump them) the developers discovered an ingenious way to escape the maze. A player could fix a limpet mine to the wall and use this as a stepping stone to jump out of the maze. Harvey Smith pointed out that old-style designers might have regarded this as a bug, but in fact it was an extra opportunity that enriched the gameplay.

“We need to reward the goal,” he concluded, “and not the method the player uses to achieve the goal.”

In the past, the nonsymbolic, step-by-step approach was not practical. The processing capability wasn't available to deal with that and graphics too. Hence design used a symbolic approach and the idea of one correct solution to every problem became ingrained. But now much of the graphics work is done by the video card, and computers are doubling in power every 18 months or so. At last, it is starting to be possible to create "uncrashable" games by avoiding the need to design using symbolic shortcuts.

Comparing Nonsymbolic And Symbolic Design

In the original Warcraft, peasants collected gold by entering a gold mine and bringing sacks back to your town hall. At the start of the game it was always worth spawning peasants because the more peasants you had, the greater your revenue stream. However, there came a point when the peasants started to get in each others' way. Adding more peasants would then lead to traffic jams as the peasants encountered each other on the streets of the town and would have to back up to let others get past. The situation was alleviated if you planned your town with wide streets. Additionally, it was not a good idea to place your town hall too close to the gold mine — giving a little more space also helped avoid traffic congestion.

Now, an economist could derive an equation to describe the flow of gold to the town hall. The factors would be the number of peasants, the placement density of the town buildings and the distance from the town hall to the mine. We can imagine that it would be a pretty complex equation. The point is that the designers of Warcraft never needed any such equation. They simply programmed in the basic rules and behaviors and the economic simulation emerged directly from those.

Contrast this with a game like Caesar 2, which used underlying equations to create a simulation of an ancient Roman city. This approach is less satisfying because the player is not directly viewing the reasons for success and failure. Instead, when playing a game like Caesar 2 (or any simulation of its type) you are trying to build an abstract match to the game's underlying equations in your head. The simulated economy and the gameplay are less visible, lessening the sense of immersion.

Rock, paper, scissors - and bots

Back from the arena of politics today to something much saner, calmer and more logical: game development. (Sure I'm serious. You'd rather crash-land on a white dwarf than a neutron star, right?) This is a fictionalized case study from my book Game Architecture and Design, co-authored with Andrew Rollings.

Some of the case studies in GAD read like horror stories, and the scariest part is they were based closely on real projects we'd worked on or had first-hand accounts of. I'm happy to report that the game development process has got a lot better in the decade and a half since we wrote the book.

Here's one case study that was entirely invented, and rather than being a cautionary tale of team-dysfunctional disaster (we're perilously close to politics again there) it was intended to illustrate intransitive game systems. That's Rock Paper Scissors in plain English, or strictly speaking Scissors Paper Stone, as we called it in the English Home Counties where I grew up. You'll appreciate that there's a spherical cow aspect to simplifying a discussion for teaching purposes, but the game design points made are interesting.

Picture by FangtheTyphoon used under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License

Peter, the lead designer of Warbots, is discussing some gameplay ideas with Charles, the project manager. Charles had requested a circular unit relationship, similar to Warcraft 2, and Peter is sketching out some possibilities.

"Say we have three main combat units," he suggests. "The Drillerbots are fast. Frighteningly fast. They can do lots of little attacks in quick succession that take the slower Hammerbots apart. They make this horrible sound as they do it, too... Nyeeowmm."

"I know, like my root canal work!" laughs Charles. "Ouch."

"Then you've got the Juggerbots. They're slow too, even slower than the Hammerbots, but they're massively armored. The Drillerbot can't even scratch them – sparks everywhere, but no damage. Meanwhile, the Juggerbot is pulling the Drillerbot apart like a crab taking a leisurely snack"

"And the Hammerbots? They can get through the Juggerbots' armor?"

"Crack 'em like nuts," grins Peter. "Only it's not so one-sided as those other matches. The Hammerbot wins, but it's a brutal slugfest with scrap metal flying everywhere."

"Good," says Charles. "So there's no best unit. Have you thought about what the player can do to change the odds?"

"I'm going to talk to Nick about that. He's still working on the physics system, which will affect a lot of things besides combat. It could be that the Drillerbots use laser drills, and they might get less effective as they move away from the player's power pylons, or maybe they weaken over time and then have to recharge."

Charles nods. "As long as it's an automatic recharge, like the units that use energy in StarCraft. We don't want players having to fiddle about sending energy to 'bots from their resource stocks."

"Agreed. Also, we could dispense with the mining 'bots. Drillerbots can mine as well as fight, so they have a versatility advantage."

Charles thinks about it. "No, keep the mining 'bots, but make them cheaper and have the Drillerbots better at blasting the ore but not collecting it. It'll give the player some interesting choices. And what about Hammerbots versus Juggerbots?"

"Underwater - under liquid methane, I should say - the Hammerbot's attacks will probably pack less punch. The Juggerbot fights with these kind of vice-grip pincers, which won't be affected so much."

"It all sounds fine," says Charles. As a longtime Warcraft fan, he is quite satisfied. "So much for combat. What about other factors? Visibility range, say. Can you get that to fit a cyclical pattern, too?"

This takes Peter by surprise, "Of course not - " he starts to say. Then, after a moment's thought, "Hey, maybe we could, at that! Suppose A has the longest radar range but the shortest sight range. B has the longest sight range, just a bit under A's radar range, but has no radar itself. C has a medium sight range and again no radar, but also is invisible to radar."

He writes on a pad, plugging in some numbers. "Yes. A spots B before being spotted itself. Same with B to C and C to A. Hey, that's nifty! I never would have thought you could get an intransitive relationship with visibility ranges."

"See, I'm too dumb to know what's not possible," laughs Charles. "But enough of this algebra stuff. Which units are we talking about here, A and B and C?"

Peter smiles as another realization dawns on him. "This has a big impact on gameplay. Say we have the Jugger spot the Driller, and so on, so that each 'bot can see the 'bot he's best equipped to take out. That'll make for a fast-moving aggressive game. The predator can always see his prey before he is seen himself. But, if we try it the other way around, so the Hammer sees that Driller coming, the Driller can avoid the Jugger, and so on... then you have a slower, more considered kind of cat-and-mouse game."

"Where the cats are made of titanium steel and the mice weigh eight tons apiece, sure." Sensing they could be one up on his favorite strategy game, Charles looks a little bit like a cat with cream himself right now. "Anyway, which way is better?"

"Better?" Peter looks up from his notepad. He couldn't look more astonished if Charles had just asked him to calculate the value of the truth-beauty equation. "I can't say which is better. It depends on the kind of game you want. We could even arrange it so that it sometimes works one way and sometimes the other. Making it vary depending on weather or the day/night cycle is one option. More interesting is if radar is messed up when there are lots of scrap robot chassises lying about, when metal buildings are nearby, that kind of thing."

"I see," says Charles. "So 'bots might see their targets, pounce, there's lots of vicious infight¬ing, but the more the bodies pile up the more difficult it gets to spot your victims...”

"And the easier to spot the 'bots to avoid. Yes. But even that's only part of it. We also have to settle on the attribute balance between visibility range and combat ability. Because, if the cost of building new 'bots is very high, winning every combat becomes critically important. If they're cheap, it doesn't matter so much. And then there's the question of not just whether a 'bot sees another, but the ability of each 'bot to close the gap or to get away."

Charles nods. "Okay, you're right. I was hoping for an easy answer for once, but there's only so far you can get with hand-waving and whiteboards, isn't there?"

“I'll get Nick to knock up a testbed. You'll be amazed what a couple days of tweaking can achieve."

With a wild surmise

Here is another excerpt from Game Architecture and Design. This one isn't really a might-have-been game, it was just intended as an example of a one-page pitch. Years later, I extended it into a full outline for a book-plus-game project I was trying to get off the ground at Dorling Kindersley. But that's a whole other story...

A old man in a tunic is being guided by a boy. The old man’s eyes are closed. We’re in extreme close-up: we see the boy’s fingers holding the man’s gnarled hand, a close-up on scuffed sandals, the old man’s cane tapping the packed earth floor as he goes. The background is in darkness but we get the sense of a huge interior space, perhaps a hall.

They reach the centre of the hall. The old man steadies himself, leans on his stick. The boy scurries off. The man stands there, head bowed, gathering his energy. His back straightens. He lifts his head.

We’re close on his face. Think of Richard Harris. The old man takes a breath and starts to speak. He has a voice of surprising power:

“Fury.”

He opens his eyes. They are milky white, clouded over. Blind. But he sees with his mind’s eye.

“Goddess, fill my lungs with breath. Give me the words to tell of Achilles’ fury. Murderous and doomed, it was a fury that cost the Achaeans so many men, and sent brave souls into the underworld leaving bloody limbs for dogs and birds to pick apart. Thus was the will of Zeus…”

As the old man speaks, our view switches behind him in a wider shot and as the camera rises up we see a hundreds of men seated at long wooden tables, lamplight picking out details in the gloom of this huge hall. They listen in utter silence as we cut to –

A beach in dazzling sunlight. In close-up as crystal-clear water gently laps the shore and then the prow of a longboat drives into the sand. Shouts and a clamor of jangling war-gear. A sandaled foot jumps down from the boat, leaving a strong imprint as the warrior strides ashore and we pan up to see –

Achilles, greatest of the Greek heroes, standing tall against the brilliant blue sky. There is a proud smile on his face as he surveys the hinterland. He has no need of armor; the gods have made him invulnerable to harm. He has just arrived and already he is eager for battle.

* * * 

The game is Troy. It’s a wargame - but not like any wargame that has gone before. Players will take the role of various legendary heroes like Achilles, Odysseus and Ajax. Each hero is accompanied by a war-band of non-player characters whose morale and fighting ability will reflect how well the player is doing.

This is an epic story. It’s also elegiac. The world is never again going to see heroes of this caliber. They are favored by the gods, are almost gods themselves. Men like the blind poet, Homer, will sing of their exploits for thousands of years. But the war is destined to end in the deaths of most of the heroes and the destruction and pillaging of the beautiful city of Troy. It is a tragic time. The end of an era.

We intend that the game will reflect this. Just as the original Iliad poem interweaves themes of rage and melancholy, glory and waste, this is the first videogame that will convey both the excitement and the tragedy of war.

How to achieve this? We said before that the actions of the player heroes affect the rank-&-file non-player soldiers. (Think of Dynasty Warriors, for example, but with more varied AI among the soldiers.) A war-band that has lost its hero will start to degenerate into a rabble. They will lose their tight formations and rigid discipline. They will start to skulk away from the hard fighting. When an enemy is struck down, instead of moving onto the next foe, they’ll wait around to loot the body. After the death of honor, these men will become the carrion dogs that Homer spoke of in the intro.

Hence, as the game progresses and heroes fall on both sides, the war moves from being a glorious adventure to a dirty, desperate, vicious struggle for survival. The gods, who are initially willing to provide aid, advice and magical weapons, withdraw from involvement as more heroes die. Even the graphics depict the dying of the light. To begin with, bright colors evoke a time of glory. As the heroes are whittled down, the images become more gritty, the colors flatter and desaturated. It is as if a Technicolor epic like Spartacus were gradually turning into Saving Private Ryan.

Grimm... it ain't no fairy tale

No, not Grimm the TV show (which I still haven't seen; opinions, anyone?). This is another of those might-have-been game concepts that I worked up in outline form while working at Eidos in the late '90s. Here it is exactly as I wrote it back then. I referenced it in the first edition of my book Game Architecture and Design (co-authored by Andrew Rollings) but other than that it hasn't seen the light of day in fifteen years...

Grimm is set in a fantasy world derived from the most ancient folktales. Players are clans and will sometimes struggle as rivals, but often must ally to counter threats to the whole realm.

How Grimm differs from standard CRPGs is that quests and interactions are oriented to the community rather than single individuals. In that sense the game is like a cross between Might & Magic and SimCity or Warcraft. You can use your most powerful characters, the Heroes, as a standard adventuring team if you like, but that leaves your clan vulnerable while they're away. Offense and defense must be balanced, and as your clan gains wealth and power your Heroes will become stronger and more skillful.

Each clan comprises:

• Four Heroes: The Lord, the Lady, the Priest and the Wizard
• Up to twelve Elders: initially with undefined powers
• Up to twenty-four Clansmen: the generic "citizens" of the clan

The Heroes are the most powerful characters, each representing an archetypal power:

   HERO         POWER                 RESPONISIBILITY IN CLAN

    Lord            battle                        military fortifications

    Lady            wisdom                    agriculture, trade

    Priest           spirituality                 health, protection from evil

    Wizard         secret knowledge     magical research

The Elders start out with no special responsibility. By assigning an Elder as lieutenant to one of the four Heroes, you cause him to "borrow" some of his master's power. An Elder assigned to the Lord becomes a Knight; if assigned to the Priest he becomes an Acolyte; if the Lady, a Minister. Once an Elder has one of these titles, he becomes a lower level equivalent of the Hero he serves with about one-third of their power.

The Wizard's lieutenants are special. Each gets a special title. Instead of being one-third as powerful as the Hero as in other cases, the Wizard's lieutenants are as powerful as their master but only in one area of magic. The Illusionist can conjure mirages, the Summoner can create temporary servants, the Elementalist can hurl bolts of fire, wind and rain.

Building up the clan

The ordinary Clansmen are not directly controlled by the player. Rather, their AI just responds to the power structure of the ruling council (Heroes plus Elders). The Clansmens' activities will reflect the priorities you've set in the council. For instance, if you assign six Elders to the Lady then half the Clansmen will concern themselves with trade and harvesting.

The clanhouse is a walled enclosure (like a motte-and-bailey castle) which constitutes the player's "city". Clansmen do not leave the vicinity of the clanhouse under any circumstances.

Maintaining the clan

Each Hero's area of responsibility is magically tied to him/her, the way the land was supposedly tied to the High King in Celtic times. This means that when the Lord is absent from the clan, the walls will no longer be maintained and will gradually crumble, the Clansmen assigned to sentry duty will become less alert and capable, etc.

Quests

Quests must be undertaken by a group of no more than three characters. These could be three Heroes, but note the drawback mentioned above (your clan will decline in their absence). More often you'll send one Hero with the power most suited to the quest and a couple of Elders with different powers to back him up. Therefore the first phase of any quest is usually to find out what's in store, so you can pick your adventuring team to meet the likely challenges.

Examples of quests:

The gods demand that a player clears and blesses a path through the Forest of Thorns. The player knows that the Priest will be needed for the blessing. The Lady would be useful for scouting a path, but he cannot spare her so he sends a Minister in her place. The third member of the team is a Knight to deal with bandits rumoured to live in the forest.

Following on from the previous quest, when the path is blessed a mysterious stranger comes through the forest and bestows a broken sword on the clan. He departs without saying anything, but the Priest learns from prayer that the stranger was the Herald of the Sea Goddess. Presumably if the clan can locate the other half of the sword they can forge it into a useful magic weapon - but what if that other part resides with another clan who won't readily give it up?

A marauding band of skeletal warriors appear at the edge of the map, attacking any who pass nearby. Trade suffers, and Clansmen are reluctant to go out to the fields. If nothing is done, the skeletons start to build a tower. Players must act to stop them before a necromancer moves in and starts calling new soldiers from out of the graveyards.

The Trickster God steals the sun, causing continual darkness to descend on the world. The crops and cattle are dying as eternal winter sets in. Someone must find where the sun has been hidden and restore it to its rightful place in the sky.

Economy

Wealth (in the form of gold) is gained by prospecting and trade. Gold is used for a one-off payment to build and repair structures, as in most strategy games. It can also be gained in the form of ransoms, as gifts from the King, or by selling magic items to passing NPC merchants.

Buildings

Each Hero has his or her principal domain: the Lord has the Barracks, the Lady the Hall, the Priest the Church, and the Wizard the Tower. These structures broadcast a continual supply of power to the Hero. Upgrading the structure (with gold) increases its "power supply".

There are also other structures that provide special functions - eg, the Watchtower from which balloons can be sent aloft to spy out the land. Some of these peripheral structures are specials that you can only build once you've got a certain item; for instance an ancient codex allowing a clan to build a Naphtha Turret.

Birth and death

New clansmen are born some time after one is killed or promoted, the time depending on the power of the Lady. If an Elder is killed, a clansman can be promoted by the Heroes.

If a Hero is killed, he awakens the Afterlife and must journey back to the temporal plane. In his absence, the clanhouse declines in the area he's responsible for, so the death of a Hero is a serious business and you must try to get him back from the Afterlife as quickly as possible.

A journey through the Afterlife is always fraught with peril but it is an opportunity for the character to gain experience. Also you may meet up with slain Heroes from other clans, and forming an alliance may be the best policy.

A typical game
A campaign starts with just the four Heroes, four Elders and four Clansmen. The way you assign the Elders will decide the activities of the Clansmen. One Elder to each Hero leads to a well-balanced development of the clanhouse, but some levels may require you to specialize. (For instance, a clanhouse set on haunted plains might make early concentration on religious affairs a priority.)

According to your policy decisions and any gold or items found, the clanhouse will grow and upgrade. Typically, upgrading of a building requires only time and manpower. Upgrades at a building (eg, the Plough to increase farm productivity) then become available but cost gold to carry out. And special quest items are needed to construct special buildings.

New Elders are usually encountered on quests, or may arrive as rewards between levels. (“The High King is pleased and sends you his cousin to serve as Elder.”) More Clansman gather to serve you as your Clanhouse’s renown and experience spread.

It’s envisaged that the view would be 3D or isometric, as in Diablo. Control during a level is via the usual interface for realtime strategy games. Between levels, you would have a strategic interface allowing you to reassign Elders, send campaign messages, etc.

Network gaming

Played on a local network, the game can be run as a regular multi-player RPG or one player can use the built-in editor to become Games Master and design the Quest Challenge. Using a set number of points he must "buy" adversaries, hide clues, ward quest objectives, etc. The players then attempt to complete the quest, the winner earning points to spend on designing his own Quest Challenge.

The One King

The land is ruled by an NPC King who is sufficiently powerful that he would be a tough adversary for all the players' clans combined. The King will set tasks for clans (usually a contest like "Bring me the antlers of the questing beast") and success earns royal favour that the player can use later. Favour can also be gained just by Heroes spending time at court, but as noted above this entails a penalty in the form of clanhouse deterioration.

Online gaming

The problem with MUDs (“Multi-User Dungeons”) is that they usually turn into an ego trip for power-gamers. Players who have been in the game a while get so powerful that newcomers are put off - they just have no way of matching the old hands in power.

Grimm gets around this by making a clan's power mainly a matter of versatility rather than brute strength. You might collect a hundred magic items, all with different powers, but each character can only take say two items on an adventure. So the long-term player will usually have more items to choose from, but he's not impossibly more powerful than the newcomer.

In addition, the One King imposes laws that all must obey. For example:

• No-one may attack allies without giving one minutes' notice of breaking off the alliance.
• A group that includes a Priest may not kill an enemy who's surrendered.
• No hostilities are allowed within the Royal Parks.
• A Lady on her own must never be attacked.

These laws provide a regulatory structure that ensures the game never degenerates into a free-for-all. Players can still plot, scheme and cheat one another, but the need to observe certain rules of engagement enhances the fun.

Games from a parallel universe

A decade is a long time in gaming. I look back at my book Game Architecture and Design, co-written by Andrew Rollings, and it's clear how different today's triple-A games are. (Technologically, that is.)

GAD is a big book. If you're in the interrogation business and waterboarding isn't getting results any more, consider investing in a copy. A couple of clouts around the lugholes with this meaty tome and those terrorists will be singing "The Star-Spangled Banner" with at least as much enthusiasm as Roseanne Barr.

One problem Andrew and I had was that we needed to discuss the concept and design stage of games without referencing real games whose creators would not have appreciated our putting words in their mouths. Where time didn't permit us to interview those creators, we got around it by making up games that could stand as generic examples of their genre. Andrew and I both being physicists, this came as naturally to us as starting with the case of a spherical cow.

It worked perhaps a little too well. For a couple of years after that, I'd have people asking me in job interviews how they could get hold of games that we'd actually only dreamt up to make a specific point. Then, looking at GAD the other day, I realized that some of those examples do make for quite tantalizing game concepts. And, since we quite often end up talking about games or gamebooks that might have been, here's one. This is just as it appeared in the GAD chapter on "Look and Feel" back in 2004, by the way, so you'll have to make allowance for the ten-year-old comparisons:

GAZE
The world of Gaze is La Vista, a single vast city that is technologically advanced (electric cars, computers, surveillance satellites) but socially conservative.

Introduction 
Our first view is a gray, unchanging surface that is moving like a featureless landscape below us. Then, catching sight of an observation port, we are able to take in the shape and size of what we're looking at. We realize the gray material is the skin of a dirigible, which moves slowly away like a weightless ocean liner to reveal...

…the retro-futuristic cityscape of La Vista. This is the city of the future as imagined in the 1930s and 1940s: vast office blocks, streets like canyons, gleaming skyscrapers of concrete and glass catching the sun. It's bright, clinical, and overwhelming.

Our viewpoint descends through wisps of cloud around the highest buildings. Recall the futurist architecture of the Third Reich, Fritz Lang's Metropolis, the Empire State Building, The Hudsucker Proxy. The quality of the light is hazy; the daylight turned to brass close to street level by the fine dust of those swept-clean city streets. Sleek cars like huge cryo-capsules whoosh down tarmac avenues on silent tires. Looking along the block, the avenue goes on and on until lost in the distance, unchanging like a reflection in a pair of blue mirrors.

The crowds on the streets are uniformly dressed: the men in dark suits, the women in gray or white dresses. This is not a world like ours with a dozen different fashions and colors. And that means that the occasional splash of color on a hoarding or in a window display is all the more striking.

And it's quiet. The cars are electric and make very little noise. The people hurry to work without a word. In our opening shot from high above the street, the first sound you hear is just the lawnmower hum of the dirigible's rotors.

What feature of all this is startling? We see it as the camera spirals down, taking a leisurely view of the streets and the people and then turning towards the center of the city as it reaches ground level. We're now looking into the burnished bronze glare of the sun. What we didn't see before was a massive statue that towers above the buildings, matching the highest skyscraper. At first it might evoke a resonance with the Statue of Liberty, but then we see the spiked crown, the balance, and the blindfold. This is not Liberty. It's Justice.

Main playing screen
Gaze is an action-adventure game and the main screen is a third-person view like in Enter the Matrix or Max Payne.

Something we must decide: Does the view ever cut, or is it a continuous tracking shot throughout? Grim Fandango and Dark Earth use the cut and all shots are static, allowing pre-rendered backdrops. This favors adventure games with strong storylines, because you can use the cut to create suspense: a sudden high angle with the hero far below, a shot from behind as a door opens, etc.

In such games as Tomb Raider and Enter the Matrix, the story matters less. Action is more important and so a smooth tracking shot is sustained throughout. Where every action counts, the player doesn't want to keep switching views.

The graphics engine will determine if the number of characters on screen would be an issue. It would be nice to be able to at least hint at the heaving mass of humanity filling the streets during the rush hour, so as to make more of the utterly deserted streets during the rest of the day. Obviously, the first-person viewpoint always has the advantage that it's one less character on screen. In any case, Gaze is a game of suspense and nail-biting tension rather than in-your-face bloodbath action, so, in fight sequences, we'd expect only a few opponents to be on the screen at any one time.

Our thinking on this has been that we'll probably go for a continuous third-person tracking shot most of the time, as per Max Payne, with very occasional cuts or pre-scripted camera movements at key dramatic moments.

Overview screen
Our original impression had been that, between encounter areas, we'd switch to a 2D map of the city on which you'd click to go to a new location.

The problem with that is that it's not immersive. It would be far better to have a seamless way of moving between the two views. The ideal would be to pull back from the hero in the close-up main view, and keep moving up and away until you had a high-angle shot of the city with the hero now a tiny figure down on the streets. Not quite as ideal, but almost as good, would be to start pulling out and then cut to the high-angle shot.

Interface 
Obviously we'll want to keep the screen as uncluttered as possible. We'd prefer to avoid having a status bar. Instead we'll show injuries on the hero himself (a torn shirt, scratches, bruises, and so on) and by the way he's moving (bad injuries cause a limp, he hunches down nursing his arm, and so on).

Selecting items from your inventory takes you to an extreme close-up of the hero pulling items from inside his jacket, while the full range of items in the inventory is shown across the bottom of the screen. (This view will be more immersive than switching to just a clinical scroll-through item list.) You pick items using either arrow keys to get him to pull out one item after another or with function keys tagged to the full inventory of items shown at the bottom of the screen. You can reorganize items in the inventory so you'll have at hand those items you'll need in a hurry.

We need to decide how to handle items that are dropped. We could say you can deposit items only at a storage locker, say. Otherwise, it's possible to get a very cluttered screen with far too many objects on it. Another way is to have a generic "dropped object" graphic and you discover what the object(s) is/are only when you pick up that graphic.

Characters 
For character style, think of those chunkily drawn private eyes in big suits that you get in comic strips from the 1940s. Bob Kane, creator of Batman, seems to have been the main influence on that style.

What we’re envisaging is that most people in La Vista (the city) seem heavy set and move a little stiffly. Their chiseled features evoke a robotic impression. Don't make too much of all this, though; be subtle. Just because they're conformists doesn't mean they have to stumble around like sleepwalkers!

But our hero, Bracken, is a free spirit, not a cog in the machine, and the graceful way he moves tells us that. The same is true of Gaze, the mysterious woman for whom the hero is searching and who seems to hold the key to changing this stagnated world. When we get to animating the characters, it will help to think of Bracken as a hawk: proud, swift, capable of both fierce concentration and ferocious activity. Gaze is a tropical fish: fragile, languid, ethereally beautiful. The incidental characters share a kind of late ‘40s uniformity so, to mark a contrast with that, imagine the protagonists cast as silent-movie era stars: Rudolph Valentino for Bracken, Louise Brooks for Gaze.