Circular Cruises/Token Trading

From Eccentric Flower

«Circular Cruises


This is the second of three "gaming rants" which I wrote, completely by coincidence, in March of three successive years. The other two are On User-Directed Narratives and Five Conflicts.

Token Trading, Colored Locks, and Pixel Hunts

12 March 2001


There must be something about March that makes my fancy turn lightly to thoughts of game design. Almost exactly a year ago, I wrote "On User-Directed Narratives."

In that essay, I said "[UDNs are] seen as having a niche audience ... which is of course self-fulfilling prophecy."

PC Gamer magazine, in its April 2001 issue, has a feature entitled "The Adventure Continues." The purpose of the article is to convince readers that the adventure game genre is not moribund. I'm unconvinced.

The article notes that The Longest Journey, an adventure game which sold very well in Europe (it was originally written in Norway), and which was reviewed enthusiastically by all the applicable computer-game magazines, sold only 351 copies in the United States in 2000.

That number is probably premature; the game wasn't even reviewed until late in the year. When I purchased it I had to go to six stores to find it. Two days ago - more than a month after I had finished the game - I happened to check one of the software stores where it hadn't been, and they had several copies. So it may be that the word is spreading slowly.

But that doesn't change the point, which is that an adventure game must apparently sell slowly by word of mouth, as opposed to the latest-and-greatest shooting/combat game, which may sell many hundreds of thousands of preordered copies long before the game even arrives in stores.

Some of this is promotion. Funcom, the U.S. distributor for The Longest Journey, has never placed an advertisement for the game in any medium I've seen. They may not have the money, of course, but even monoliths like Sierra rarely advertise (or, for that matter, make) traditional-style adventure games any longer. (My use of the qualifier "traditional" will become clear eventually.) LucasArts is the last company making and advertising such games on a regular basis ... and their latest Monkey Island game has not yet met sales expectations as of this writing.

So how can PC Gamer tell us all this bad news, talk of all the luminaries who have left the field in disgust - such as Gabriel Knight's Jane Jensen, who is now apparently writing bad novels for a living - and yet still assure us that the adventure game is alive and well?

Easy. They do it by changing the definitions.


The problem with the phrase "adventure game" is that it's vague. Historical, yes (it comes from the text game Adventure, the mother of the genre), but vague. All games are adventures, or at least they had better be.

What game magazines and fans such as myself mean when they say "adventure games" are usually token-trading games. In a token-trading game, you obtain key items - tokens - which must then be given to the correct receivers to proceed with the game. Usually these are chained, so that your prize for figuring out that a given token goes to a given receiver is ... another token, which you can then give to someone else for another token, until finally you get the token which wins the game for you.

Tokens can be multiple - that is, you have to give several different objects to a particular receiver, usually in the right sequence. They can be reusable - you get the token back after doing the right thing with it, because you'll also need it somewhere else. They don't even have to be physical - a token can be an item of information, a snippet of key dialogue.

Token-trading is virtually the only way to successfully write an "adventure game." There is a long-standing tradition - and I approve of it - whereby an adventure game (which is supposed to be about brains, not reflexes) does not have a ticking clock. Nothing happens until you make it happen - if you wander around aimlessly trying to figure out all the places you can possibly use a given token (and you will, at some point, alas) - you will not be penalized.

Token-trading, though, is precisely the part of adventure gaming which has a bad reputation. There are two reasons for this. Both are avoidable.

In "On UDN," I said:

Infocom games and other similar games by companies which used to specialize in them (Sierra, LucasArts) have fallen into disrepute as "give the tuba to the ostrich" games - where you are reduced to random transactions or object-hunting to keep the storyline moving: "I know I have to give the farmer something in order to get the eggs I need, but I have no idea what he wants. I'll just offer him everything in my inventory and see which one he accepts." This is a sign of bad game-writing - it doesn't have to be that way. More importantly, it shows that no one was giving serious thought to the story. If the farmer were a real character, if his motivations and your surrogate's had been well-described, you probably wouldn't need to guess what the farmer wants.

Even in a game as good as The Longest Journey, there were periods where I was reduced to wandering - slowly and painfully - through all the accessible locations in the game and trying every single human or object which I could interact with in any way, running through my inventory of tokens each time, in a desperate trial-and-error attempt to find a receiver willing to take a token. It quickly becomes obvious when you have reached a point in an adventure game where the game is waiting for a token to be received before any other parts of the game will "unlock." This not only ruins the illusion that you're in an open-ended story, but if you don't have a clue what to do next, the process gets very frustrating very quickly.

The Longest Journey plays fairer than most. When every other character in the game tells you that you need to go have a word with Cortez, it's plain as day where you're being steered. But what if you don't want to talk to the other characters? What if you don't want to walk through each option on each character's "dialogue tree" of possible subjects to discuss, because doing so is hugely tedious and nine-tenths of the time will result in their giving you snippets of dialogue you've already heard?

Clearly the thing to do is to have a successful token-drop give you a piece of information which hints at the next token-drop. That is, when you successfully get the eggs from the farmer, he also tells you that the widow down the road is his most regular egg customer.

But some people will say this makes the game too easy. And it may also make the game too linear. For ideally you will have several unclaimed tokens floating around at any time, several puzzles that the player can work on, choosing which order to work them in. This means that the player may not need a hint about the eggs - the player has already figured that out, but could really use a hint about what to do with the steak knife (a puzzle the farmer has nothing to do with).

Or, conversely, you could just build a separate hint system into the game.


The other problem with such games is something that originally wasn't a problem. Once upon a time, adventure games were all-text. Gradually this was considered obsolete; as the other kinds of games became increasingly graphics-dependent, the people who wrote adventure games realized they'd better put some pictures in or their games wouldn't be sexy enough to sell. (You may be able to detect that I'm less than thrilled about this. Then again, I still read paper books, so obviously I'm a Luddite.)

In the text-adventure world, you never had to worry about what the "live" objects were in a room or location - that is, the things you could manipulate, pick up, or otherwise interact with - because the game would happily list them for you. It would also list all the possible exits from a location, so you never had to worry that there was some room/location you were missing entirely simply because you didn't realize where the doors were.

But manipulating graphic representations of objects, combined with the small image area on even a large monitor, resulted in what magazines and players refer to disparagingly as a "pixel hunt." That is: Your character walks into a room where you see a lab table, cluttered with beakers and test tubes and all sorts of bubbling apparati. Only one test tube is vital, and it is the only live object on the screen - assuming you can find it. The live or "hot" area is about twenty by twenty pixels or less, because a test tube is rather small, and what you eventually end up doing is sweeping your cursor over every object, every inch of table surface, maybe over the entire room ... watching for the change in the cursor that says you are hovering over (or have momentarily passed over) a hot spot.

This is even more tedious than walking through everyone's dialogue tree for the hundredth time.

One solution - and some games have done this - is to have a key you press down which makes all the live objects in the room flash at you or otherwise indicate their presence. Another solution is to have the live objects indicated more clearly - for example, have the live test tube be visually isolated on the table, an empty space around it, or maybe it's the only chemical that's glowing, or some such. But the latter may not be as effective as it should be; some of us have been so thoroughly trained by years of bad game design that we'll probably sweep the room anyway, out of habit.

And of course we could just go back to all-text games. Yes, I'm cranky, but at least I'm not the only zealot in town. Scott Adams, a pioneer of the field (no relation to Mr. Dilbert), has gone back to putting out the same sort of all-text adventures that made him famous in the days of the TRS-80. He sells them on his website, direct to the public ... which of course is the only way he possibly could; no game distributor would give an all-text game the time of day.


Once upon a time there was the world of adventure games, and there was the world of reflex games, and that was that. You played Adventureland or you played Lode Runner; you played Zork or you played Pitfall; there was very little in between. (I've picked from several different types of hardware systems and years there, just to make my point.) There were a few strategic games - war games, and the occasional galactic conquest game. There were a handful of very early role-playing games (RPGs), which were usually combat games in thin disguise. The first-person shooter was barely extant; the hardware wasn't available to do the format justice yet.

Then, eventually, the field shook out like this: Adventure games, shooters, RPGs, and turn-based strategy (which includes wargames and empire-building games like Civilization). The arcade-style reflex game - in the Lode Runner or Donkey Kong sense - was so throughly eclipsed by the first-person or third-person shooter that it is only now beginning its comeback.

Finally Blizzard and others came up with the "real-time strategy" game, which combines elements of fighting games, RPGs, and strategy/resource management. I don't play these because I don't like games where the enemy can outflank you while you're trying to figure out what to do next; I hate the sense that the clock is ticking. But I recognize that these games were something brilliant and new - a fusion of ideas that had honestly never been done before.

(Not even most first-person shooters are real-time, although they try to conceal it; you can usually wander around all day until you walk into the key location that activates the next group of bad guys. It's easier to write an event-driven game - where the computer waits for a key event to happen - than it is to write one where all sorts of things take place with every tick of the clock.)

But it's interesting to note that practically the only element not used in Warcraft and Starcraft is token-trading. Granted, it didn't really belong in those games, with their focus on resource and troop management ... but it's also absent from Blizzard's excellent Diablo games, where it would have been quite suitable and might have added to an already excellent playing experience.

The point is, combining token-trading aspects with other game types was, until very recently, unthinkable - a way to guarantee your game would be harshly criticized by reviewers and players who "didn't want to deal with any of that puzzle-solving stuff."

This shouldn't be taken to imply that people who play shooters and other reflex games prefer their games to be free of brain activity. For one thing, it isn't always true. Certainly, planning the best way to ambush a building or deploy your troops is a mental exercise, although a different kind. That's the core problem; conventional wisdom says that the people who play reflex games and the people who play adventure games are looking for wholly different mental thrills.

There is another adventure game tradition which says that you can't die, or do something that would make it impossible to finish the game. The former is permissible as a way of showing you've made a poor choice, but with saved games taking so long to reload these days, this can really be frustrating to a player who is working through all the alternatives, trial-and-error style. On the other hand, if they're having to do that, it probably means the game was badly designed anyway. As for the latter, if they reach a point where they can't proceed because of game mechanics - say, if they gave away an object they need and have no way to get it back - it is always an indicator of poor design.

This means that, in theory, you will always be able to finish an adventure game eventually, with brainpower and perseverance. Yes, it can be frustrating to be unable to solve a puzzle, but it is even more frustrating when you're playing a shooting/combat game and you reach a point where you can't continue, through utterly no fault of your own, because your reflexes just aren't good enough. Shooting games have "cheat codes," used to beef up your proxy's survival ability or lower the opponents'. Adventure games don't have or need cheat codes. They have "walkthroughs," which give the answers to puzzles. This is a difference in religion.


But how much of that religious gap is real and how much is perceived? One problem is that the token-trading elements had already appeared in shooters and RPGs, in a very denatured form ... and thereby given the token-trading genre an even worse reputation in the eyes of the reflex crowd.

In this denatured form of token-trading, no attempt is made to conceal where the token belongs - or only a token attempt (pardon the phrase) is made at making a puzzle of it. "Hmm, here I have a stone shaped like a spider ... and over there is a door with a spider-shaped depression in it exactly the size of the stone ... wonder what the stone could possibly be used for?" (The Tomb Raider games are infamous for this latter sort of thing.)

A half-hearted "puzzle" like this is tedious and insulting. You might as well not even pretend you're hiding the function of the tokens, if you're not planning to do the job right. Diablo II has a few tokens in the course of the game, but it knows its goals and sticks to them; in each case you are told well in advance what token you're searching for, and once you get it, you are told where and how to use it. The challenge in that game is getting TO the token - fighting your way through whatever is guarding it - not figuring out how to use it.

The problem is that, with rare exceptions, this sort of approach might as well be left out of the game. In a game like Diablo II, which has an extended-quest storyline, the few key objects are used effectively; in a game like Quake, which is all about kill or be killed, they do indeed get in the way. But that seems to be because they are grafted in poorly or misused ... not because the genre of game is intrinsically mis-suited to these elements. In short, it has a bad reputation because it has rarely been done correctly and people have learned to expect the wrong things.

One example of doing the job right was Descent and its sequels. An otherwise pure reflex game, almost all levels had three keys, and the pattern was almost always the same; first you would find the blue key, which would open up a new section of the map; somewhere in that section, inacessible until you had found the blue key, was the yellow key, which would open up another new section; somewhere in that section was the red key; and the red key opened up the last section of the map, wherein could be found the reactor you needed to blow up to end the level.

By quickly teaching its players the rules of the game, and getting them in the habit of what to expect, it was doubly effective when a map broke one of more of these rules - like having no yellow key, or having the blue key get carried around the map randomly by a wayward robot - and Parallax had the good sense to only mess with the rules often enough to keep everyone on their guard.

That's why I call this sort of arrangement a "colored locks" puzzle, as a reminder of one of the few games which managed the trick, and also as a testament to an old arcade game called Pulsar (1981), which is the earliest I remember seeing a variation on this concept.


All of this essay so far has been devoted to one of two themes: How Things Can Go Wrong or What Has Gone Before. Now to talk about the present.

As I said, PC Gamer has painted a rosy future for the adventure game, but they have done so by changing the definitions. Don't blame them for this, though; they are merely reflecting the state of the industry.

The "traditional adventure" game is probably dead. I believe it is dead. The industry, already conservative in this respect, is going to completely misinterpret the poor sales of The Longest Journey and that will be the final nail in the coffin.

But, as game designer Tim Schafer says, bits and pieces of the corpse keep showing up all over the place. In the last twelve months I have played several games which don't just blur the traditional boundaries I've outlined, they obliterate them. These are the sort of games that PC Gamer is crowing about:

System Shock 2 had the storyline of an adventure game, the inventory management and character advancement of an RPG, and the combat and first-person perspective of a shooter. It did not have token-trading, but it did have specific tasks for the player to complete in a specified order.

Deus Ex is another shooter with a strong story and RPG-like character development and inventory management. It has some token-trading, in disguise (the tokens are almost always verbal or conceptual - information counts for a lot in this game).

Soulbringer is probably a reflex game first and foremost, but it has many RPG elements and some token-trading elements. In fact, it's hard to say what Soulbringer is, exactly. The game has gotten some bad press because of its design flaws, but most of the reviews seem to have a common theme: If you get past the bad interface, the game is strangely compelling.

I think it's compelling because it is a wholly new breed, and I hope that someone else will come along and take the same idea and do something better with it. It is the game which, even more than the other two, addresses the problem of What To Do While You're Questing.

You see, the big problem with games like The Longest Journey is that you spend a lot of time wandering around, trying out tokens and conversations ("and what good is a book," Alice thought, "without tokens or conversations?"), trying to unlock what happens next. And then you do, and you get an enormous information dump. And then you go into a long dull period of wandering again. Traditional adventure games have a pacing problem.

But in Soulbringer, you can always go fight a battle somewhere or help clear the bandits out of the wilderness while you're working on the solution to your current quest - sometimes fighting the bandits IS the solution - and there is generally always mayhem to be had somewhere. The game does have a few slack periods where it seems like you've done everything you can do, but not many. And there seems to always be a new, tiny revelation just around the corner. The game doesn't force its revelations down your throat either; many of them are in supplemental reading material that you may pick up, or you may not; you can play the game without ever reading any of it, or you can read it and get that much more backstory.

The problem with Soulbringer - beyond its many interface problems, I mean - is that there are generous combat sequences, and finishing some of the quests means surviving those sequences, which means that people who want to know how the story comes out but have horrid reflexes or a poor mouse hand will eventually grind to a halt in frustration. And Soulbringer has no cheat codes that I've been able to find. (I reached this point of frustration - an enemy that I was unable to beat in combat, and unable to proceed without beating. I went into the game's files and hacked the event script to get past this little problem. There are some advantages to doing this sort of thing for a living.)

What we need is a hybrid that has the combat and other entertainments, but doesn't depend on them. That is, all quests should be solvable by brains alone; fighting and treasure-hunting is useful for making a little money or getting your character's stats up, but not an utter requirement. Maybe puzzles could be required to have a "stealth" solution in addition to the "brute force" solution. (This is a feature I love about Deus Ex, where each goal usually has several approaches; I tend to favor the lurk-in-the-corners sniper rifle approach, as opposed to a direct assault, but the other is possible in the right hands as well.)

Of course, this sort of approach would annoy the people who just want to run around and shoot things. But they can do what I do: When I get in that mood I play Quake.

Copyright © March 2001. All rights reserved.

Personal tools
eccentric flower