Tagline: This is the book which accompanies quite a few games – the odds and ends which got excised from the primary rulebook, padded out with a handful of NPCs.

In the few short years since its creation and demise at the hands of poor business acumen Feng Shui has established itself as one of the classics of the industry. Unfortunately it was marred by a mixed bag of supplements – some were very good, but others were drawn from the bottom of barrel. Back for Seconds is, in my opinion, the worst of the lot.

To understand the atrocity which this product is you have to understand from whence it was conceived: First, when Robin D. Laws finished the core rulebook some of the material he produced had to be excised due to space considerations – the in-depth description of Operation Killdeer and also several of the character creation Types. Second, the setting for Feng Shui was loosely based on the elements found in the collectible card game Shadowfist — included in the Shadowfist game were several cards detailing specific characters and some cards detailing feng shui sites.

Welcome to Back for Seconds, a bastard child of supplement creation. With a foundational structure of the dozen pages or so of Laws material which didn’t make the final cut of the rulebook, Jose Garcia (the “executive producer” of the Feng Shui line of products) hired a hodgepodge of creators to write out the characters and locations from the Shadowfist cardgame. John Tynes (the product’s “developer”) describes the result best in his introduction as “Central Casting” – but I can’t say I’m impressed with the results.

(None of this, by the way, should be taken as a condemnation of the creative personnel involved in this project.—with names like Bruce Baugh, Greg Stolze, and John Tynes working on it there’s no way you could sanely say this was true. But the product was conceived as a hodgepodge and the result is precisely that: A hodgepodge of questionable usefulness.)

Insofar as it is a hodgepodge, Back for the Seconds is probably the best product of this variety I have yet encountered. The NPCs are generally engaging and interesting (even if they are largely based on one sentence “catch-phrases” from the Shadowfist cards). The types which were cut from the main rulebook provide valuable options to the character creation system. The locations are interesting, if of limited versatility. But the overall product can’t overcome the basic flaw of its conceptual base. It simply fails, in my opinion, to be a valuable sourcebook. About the only thing which would make this worth your money would be the Robin D. Laws material (which is why this receives a 3 and not a 2 rating in substance), otherwise you’d be just as well off spending your money on renting a few Hong Kong action flicks and ripping them off. You’d have more fun.

Style: 3
Substance: 3

Writers: Bruce Baugh, Dave van Domelen, David Eber, Jose Garcia, Steven Kenson, Bob Kruger, Robin D. Laws, Andy Lucas, John Tynes, Greg Stolze
Publisher: Daedalus Entertainment, Inc.
Price: $14.00
Page Count: 78
ISBN: 1-888335-02-5

Originally Posted: 1999/04/13

Feng Shui was one of two games that got me back into roleplaying games. In the mid-’90s I had drifted away from roleplaying games: I had lost my regular groups, other interests were beginning to consume my time, and I had also lost easy access to the online communities that had seen me through previous dry spells. But in the summer of ’97, I was living in Mankato, MN with my dad.

About midway through the summer, I read a blurb in the Comic Buyer’s Guide talking about TSR, Inc. being essentially bankrupt… Wait. What?!  (I guess that explains why I haven’t seen an issue of Dragon Magazine in months.) In an effort to figure out what was going on, I ended up figuring out how I could access some of my old haunts on Usenet and Fidonet. The result was that I got plugged back into gaming news.

I don’t remember now whether it was Feng Shui or Heavy Gear that intrigued me enough that I looked up a local hobby store and biked across town to check it out. This little hobby store only had 3 shelves of incredibly unorganized RPG material, but by some miracle they had these games: One I had come looking for; the other captivated me. I ended up buying both and spent most of the rest of the summer biking back for more.

(And it was an incredibly exhausting bike trip: Mankato is built around and over a cliff facing a river. My dad’s apartment building was located, almost improbably, halfway up this cliff: So no matter where I went, I was either biking up the cliff to get there or up the cliff to get home.)

For an explanation of where these reviews came from and why you can no longer find them at RPGNet, click here.

Go to Part 1

Now that we’ve discussed incomplete game structures, let me go fully the wrong way about it and explain what I mean by a “complete game structure”.

Most importantly, what I do not mean by a complete game structure is that the players will never be able to take actions outside the structure. As I’ve been saying all along, roleplaying games are functionally open-ended: Unlike a boardgame or card game, players are always free to propose whatever course of action strikes their fancy.

Instead, what we’re talking about is a structure which can theoretically provide a complete experience. If the players choose to stick to a complete game structure, that structure will never deliver the game to a place where no structure exists.

Practical experience at the game table plays a role here, and I’m also expecting a dose of common sense: It’s possible to argue that Traveller presents a “complete game structure” as long as the PCs never leave their ships. But, of course, that’s not what actually happens at the game table, is it? What happens in actual practice is that the Traveller scenario structure delivers a ship to a starport, the PCs get off their ship, and… the scenario structure has delivered them to an unstructured place.

In practice, complete game structures require vertically integrated structures and closed resolution loops. They will also usually feature complete transitions from macro-level to micro-level game structures.

For example, consider the basic mystery scenario structure: You start in a location and you search for clues. The clues take you to another location, where you search for clues (which will, of course, take you to another location).

That’s a closed resolution loop: You can follow that structure forever.

Earlier, we talked about the smooth transition from dungeoncrawling to combat in D&D: You’re in a dungeon room, you pick an exit, and you go to another dungeon room (there’s your closed resolution loop). But in that room are a bunch of monsters, so you switch to the combat system (which is another closed resolution loop) until you defeat the monsters. Once the combat is finished, you swap back to the dungeoncrawl structure, pick an exit, and go to the next room (which might also have monsters in it, in which case you swap back into combat).

That’s a vertically integrated structure: The dungeoncrawl structure provides a specific trigger (“there are monsters here”) which transitions you into the combat structure; and the combat structure provides a specific trigger (“all the monsters are dead and the room is empty of interest”) which transitions you back into the dungeoncrawl structure.

(As a thought experiment, imagine that you were using a dungeoncrawl structure but you had no combat system. Can you see what happens when the dungeoncrawl delivers you to a room full of orcs? The structure is incomplete. Now, imagine that you’re using a dungeoncrawl structure but instead of a combat system the rules had a well-developed mechanical structure for resolving riddle contests. That’s a very different dungeon, isn’t it?)

It’s also possible to extend the chain of macro-level to micro-level transitions. An easy demonstration of this is a hex keyed with a dungeon complex: The hexcrawl structure triggers the discovery of the dungeon; entering the dungeon transitions you to the dungeoncrawl structure; and meeting some monsters in the dungeon transitions you to the combat structure.

I occasionally think of this as an inverted pyramid, but it’s not necessarily a tidy one. (For example, hex keys can also easily trigger you straight into combat.)

FEATURES OF A COMPLETE GAME STRUCTURE

Let’s take a moment to look at the basic features a complete game structure requires. By matter of necessity, these will be somewhat preliminary conclusions: There aren’t many complete game structures in roleplaying games, so the body of data is sparse.

First, each part of the game structure requires a clear and specific trigger. (For example, combat is triggered when someone wants to make an attack and initiative is rolled.) Vertically integrating your structure requires that the macro-structure will deliver the triggers necessary for the micro-structures. (For example, the dungeoncrawl structure delivers the trigger for combat when you enter a room filled with hostile monsters.)

Second, you need a default scenario hook. This is actually just the trigger for the macro-structure which contains the micro-structures and it doesn’t need to be terribly complex: For example, the default scenario hook for a dungeoncrawl is “the dungeon entrance is in front of you”.

Third, you need a default goal. This is what motivates the players to engage the scenario hook. For example, the PCs go into the dungeon (i.e., accept the hook presented by the dungeon entrance) because they want to find treasure.

(It should be noted that the default scenario hook and goal are often the first things to get swapped out when you’re adapting a game structure into actual play. For example, the PCs might be going into this dungeon – which is actually a fortress – because their friend has been kidnapped by the slavers inside.)

Fourth, you need a default action: Picking an exit in the dungeoncrawl; picking a direction in a hexcrawl; looking for clues in a mystery; attacking an enemy during combat; etc.

Fifth, taking the default action should return you (either directly or after a sequence of actions) to a point where you can take that default action again. This is the point where you close the resolution loop and complete the game structure.

EFFECTIVE GAME STRUCTURES

There are also a few things which aren’t necessarily required for a complete game structure, but which are probably a good idea.

First, the structure should be flexible. Flexibility for the players means the ability to make choices which are not constrained by the structure. Flexibility for the GM means the ability to include a broad and creative array of content into the “containers” of the structure.

(These constitute the “structure, not a straitjacket” and “flexibility within the form” principles we talked about before.)

Second, the default action should usually allow the players to make meaningful decisions. Roleplaying games are primarily about making choices and structures which don’t allow for those choices to be made will, in my opinion, struggle. Furthermore, a structure which doesn’t include points at which players are free to make decisions is probably going to strongly inhibit their opportunities to make choices which are not constrained by the structure (see above).

(As a counterexample, consider a hypothetical piratecrawl structure in which the players choose a direction to sail and then the GM randomly determines what type of ship they run into: This has a great deal of flexibility for the GM, since he’s basically free to include anything which fits into a ship-shaped container. But you’ll notice that the players’ decision of sailing direction is meaningless, which will result in less than satisfactory play.)

Third, assuming that we’re talking about RPGs and not STGs, the players’ decisions within the game structure should be associated and in-character.

(For pretty much all the same reasons that the mechanics of an RPG should be associated.)

Fourth, the reward structures of the game should probably be tied into the default actions and default goals of the game structure.

(The easy example here, as I’ve mentioned before, is D&D: The default dungeoncrawl / combat game structure rewards you with XP for the default action of fighting monsters and with GP (and XP) for the default goal of treasure-hunting.)

Fifth, the game structure should be fun. This, of course, can be a somewhat nebulous quality and will also vary from one group to the next. But the important point here is that a complete game structure will act as the foundation for your scenario or campaign: If that foundation is fun, then everything you and your players build on top of it will be fun. If it is simply workmanlike bookkeeping, on the other hand, you may still end up having fun if the stuff you build on top of it is awesome enough, but it’ll be less of a sure thing.

Sixth, the game structure should be easy to prep. Part of this is providing a clear structure for prep (and I think you’ll find that clearly defined game structures will simultaneously provide straight-forward guidance on what you need to prep). But it’s also about limiting the amount of prep necessary for play to begin. (If your structure requires the GM to prep 500 pages of material in order to use it, it’s not going to see a lot of use in the real world.) And all of this, of course, can be helped by giving the GM tools which assist with prep (like the old dungeon- and hex-stocking tables that used to be a mainstay of D&D).

I was also going to add “easy to run” to this list, but I suspect that this is a quality which actually arises naturally out of properly constructed game structures. But if you’re in need of a reminder, check out Part 8 of this series.

Go to Part 12: Using Scenario Structures

As a brief tangent to our discussion of game structures, how could we go about plugging the “hole” in Traveller’s scenario structure?

Since Traveller’s starmaps are hex-based, we could start by rifling through the pockets of the hexcrawl and snagging the concept of “when the PCs enter the hex, trigger the hex’s keyed content”. But what we’ll quickly realize is that there’s a reason Traveller doesn’t do this already: Planets are really, really big.

If we think of a dungeon room or a wilderness hex as a “container” which holds keyed content, it’s pretty easy to recognize that a planet-sized container completely dwarfs the scale of any single encounter or location. Take an encounter and stick it in a dungeon room; it fills the room. Take the same encounter and put it in a wilderness hex; it gets a little bit lonely but there’s still enough to quench our thirst. Stick it on a planet, on the other hand, and the glass still looks completely empty.

And this isn’t just a matter of aesthetics. It impacts the GM’s ability to logically transition the scene: If the PCs are traveling through a wilderness, it’s relatively easy to find a transition from the macro-level of wilderness travel to the micro-level of the keyed encounter. You just say something like, “As you pass through the forest, you see a bunch of orcs.” But this method doesn’t work quite as well if the PCs are approaching a planet.

Perhaps more importantly, this also affects the ability for players to make meaningful decisions within the context of the scenario structure. If the GM of a hexcrawl says, “You see a castle perched upon a rocky crag of blackened stone.” The players can say, “Okay, that looks scary. I think we’ll just skirt it as far to the west as possible.” And they can do that. But if they were to say, in this hypothetical game structure for spacecrawling, “Okay, I don’t want to deal with this settlement of murderous cannibals. Let’s lift-off and land somewhere else on the planet.” Suddenly they’ve exited the purview of the scenario structure.

One logical leap from this conclusion is that we simply need to key more content to the planet. Maybe, for example, we could just map the entire planet and key it as a hexcrawl: Thus, just like a hexcrawl contains dungeoncrawls, our spacecrawl would contain hexcrawls.

But hexmapping and keying an entire planet in any sort of meaningful or interesting detail? That’s an absurd amount of work. It’s obviously completely impractical.

HAILING FREQUENCIES OPEN

So maybe, instead of that, we let our logic take us the other way. Maybe we accept our limitations and implement a structure where each planet is keyed with a “hail”: When an intrepid band of interstellar scouts enters orbit, they’ll receive the hail.

This basically handles the problem of scale by embracing the old pulp SF trope of “every planet is a village”. The result is less Traveller: Firefly and more Traveller: Star Trek. But as long as the players are onboard with the fact that they have to either engage with Oxymyx and Krako or bypass the planet (i.e., they can’t just decide to head over to the other side of the planet and talk to a different set of gangsters), you’ll probably be in pretty good shape.

Unfortunately, the structure starts to breakdown once you’re dealing with civilized planets (instead of exploration). But, of course, this is also true for traditional hexcrawls (which provide no guidance for PCs once they pass through the gates of Greyhawk).

ABOARD THE KING ARTHUR

Unsurprisingly, I’m not the first person to start thinking this way about Traveller. Back in 1981, FASA produced Action Aboard: Adventures on the King Richard. In addition to detailing the ISCV King Richard and providing two sample adventures, this module included a dozen or so “Outline Adventures”, all of which constitute primitive scenario structures into which any number of specific scenarios could be poured. For example, the Hijack:

I. HIJACK

The forcible capture of the vessel and its contents may come from either inside or outside the ship.

A) Internal

A hijack from inside assumes that the hijackers and their equipment are all aboard at the beginning of the scenario.

1) Players as crew – the players may or may not be issued weapons from the ship’s locker, depending upon the level of surprise.

2) Players as passengers – The players will not be given weapons and, unless they have social standing A-F, they will be abandoned at any habitable planet (roll 1-4 on  one die) or shot, and/or dumped into a vacuum (roll 5 or 6 on one die). If they have high social ranking, they will be ransomed.

And so forth.

Primitive? Yes. Still incomplete? Sure. But you can see how this begins to provide a basic structure into which a few stats can be poured to give you something that can be used in play. In combination with the other “Outline Adventures”, Action Aboard begins to give you enough structure to run an entire campaign based around the voyages of the ISCV King Arthur.

Go to Part 1

Consider, for a moment, the explicit game structure presented in the original Traveller.

(1)   Create a subsector. For this, the game gives guidelines for creating a starmap; populating that starmap with planets; determining the population, law level, and technological level of those planets; determining travel zones and trade routes; and so forth.

(2)   Own a starship. The game offers several options by which the PCs can own, lease, or otherwise operate a starship.

(3)   Interstellar movement. Providing mechanics for determining how far and how fast PCs can move between planetary systems.

(4)   Trade and revenue. Finally, explicit guidelines on how revenue can be earned by carrying cargo, passengers, and the like.

When you boil it down, this is Firefly: The Roleplaying Game.

But what this scenario structure notably lacks is any support for play below the interplanetary level. Traveller recognizes this lack and works to patch the hole with the concept of the Patron:

The key to adventure in Traveller is the patron. When a band of adventurers meets an appropriate patron, they have a person who can give them direction in their activities, and who can reward them for success. The patron is the single most important NPC there can be. (Book 3: Worlds and Adventure)

Basically, the patron serves as a default method for delivering adventure seeds to the PCs. And Traveller integrates the patron into its larger game structure by triggering patrons through its random encounter system. (So, basically, every time the PCs fly into a starport there’s a chance they’ll be contacted by someone with a special job.)

Of course, this still leaves the vast gulf of what the game structure for the actual mission itself is. But it’s not as if Traveller is alone in having such gaps in its scenario structure. In fact, virtually all RPGs have such gaps. (And, at a micro-level, virtually all RPG mechanics constitute incomplete game structures, as our example of the Duchess of Canterlocke demonstrated.)

MECHANICS WITHOUT STRUCTURE

I’ve long maintained that RPGs naturally gravitate towards their mechanics. For example, when I added counter-intelligence mechanics to D&D, counter-intelligence suddenly became a significant part of my campaign. When I added usable encumbrance mechanics to my OD&D campaign, encumbrance-based gameplay immediately followed.

What’s even more true, however, is that RPG gameplay naturally gravitates towards structure. Because they’re open-ended, of course, RPGs are not bound to their structure (like a boardgame is) and good game structures in RPGs won’t act as straitjackets, but clear game structures nevertheless attract players and GMs alike.

Or, to put it another way: If you’re in a dungeon, at some point you’re almost certainly going to start dungeoncrawling.

An interesting corollary of this is that mechanics which aren’t (a) required by a well-defined game structure or (b) enhancing a well-defined game structure are often ignored.

As a result of this corollary, you’ll often find popular game systems ringed with a large number of rules which nobody uses. Many of these rules seem to accumulate from a vestigial urge for simulationism.

This is particularly true with specialized supplements. For example, the D20 market is crammed full with supplements that aimed to provide the “Definitive Guide to Ships”. The little simulationist urge says: “There are ships in fantasyland. So we must need rules for ships.”

But once you have them, what do you do with them?

A typical seafaring supplement, for example, usually included all kinds of rules for varying the speed at which a group traveled: The ship they’re using, wind speed, crew experience, navigation checks, weather conditions, tidal drifts, and so forth. But unless you’re using a scenario structure in which travel time matters – and in a modern era of railroaded scenarios, it generally doesn’t – all of these rules are pretty much irrelevant. Oh, having some guidelines for how long it takes to get from Point A to Point B is nice, but anything involving a lot of calculation which varies from one day to the next is basically useless chaff.

That’s how you get books full of feats that nobody takes, spells that are rarely cast, rules for ramming that nobody bothers with, and so forth. Without a supporting scenario structure, this stuff just flounders: It occasionally gets toyed with, but it rarely gets used.

But imagine for a moment that someone took the time to design a fully-integrated scenario structure for sea-based play that, for example, made playing a pirate or a privateer just as much fun as dungeoncrawling or solving mysteries. You could build entire campaigns around this structure, or just slot it in as appropriate. Maybe you could even go ahead and publish a full campaign that people can just pick up and play. Suddenly all those rules for handling crew morale and ship-to-ship combat are being used.

And if your new scenario structure were really successful, suddenly you’d have opened up a whole new market for support products.

THE VALUE OF PARTIAL STRUCTURES

The “gravitational effects” of clearly defined game structures may also help to explain why even partial scenario structures have often proved monstrously successful in the RPG industry.

For example, consider Shadowrun and Paranoia. Neither game features a comprehensive scenario structure, but they both have default methods for delivering scenario hooks which also tend to lend a common shape to their scenario concepts. And you can see evidence for the efficacy of these techniques in the number of “traitorous Mr. Johnson” stories you see from Shadowrun and the number of “briefing room horror stories” you see from Paranoia.

In fact, even partial scenario structures seem to be very effective at providing a commonality of experience which can draw a player base together into more meaningful communities. (This is particularly true for default scenario hooks.) Such communities provide a strong network effect which further strengthens the game.

This commonality of experience also makes it possible to produce supplements which target that common ground. 76 Patrons for Traveller or Mr. Johnson’s Little Black Book for Shadowrun are obvious examples given the context of our current discussion, but it’s particularly true when it comes to adventure modules: You can trivially produce an adventure module for D&D which could theoretically be plugged into 90% or more of the current campaigns being run in the system. On the other hand, it would be essentially impossible to produce a Heavy Gear module for which that would be true.

Of course, having a viable market for those kinds of adventure products makes it easier for a publisher to produce adventure products. And having adventure products available makes it easier for new GMs to start playing the game. Which further enhances the network externality of the game.

Go to Part 11: Complete Game Structures

Go to Part 1

As we wrap up our discussion of the hexcrawl game structure, I thought it might be interesting to take a few moments to revisit a few archaic game structures that have been abandoned by the hobby.

Until recently, of course, the hexcrawl itself could be described as an archaic structure. By 1989 there were a few vestigial hex maps cropping up in products, but none of them were actually designed for hexcrawl play. The 2^nd Edition of AD&D removed hexcrawling procedures from the rulebooks entirely. It wasn’t until Necromancer Games brought the Wilderlands back into print and Ben Robbins’ West Marches campaign went viral that people started to rediscover the lost art of the hexcrawl.

The original game structure for hexcrawls described in OD&D is actually quite distinct from the hexcrawl procedures I described earlier (which were largely innovated by Judges Guild and than adapted into AD&D). The bulk of wilderness adventures in OD&D focused on castles:

Castles: As stated, the ponds [on the Outdoor Survival gameboard] indicate Castles. The inhabitants of these strongholds are determined at random. Occupants of these castles will venture out if a party of adventurers passes nearby. If passing over the castle hex there is a 50% chance (die 1-3) that they will come out, if one hex away there is a 33-1/3% chance (die 1-2), and if two hexes away there is only a 16-2/3% chance (die 1). If the party is on the castle hex and hails the castle, the occupants will come forth if the party is not obviously very strong and warlike. Patriarchs are always Lawful, and Evil High Priests are always Chaotic. All other castle inhabitants will be either hostile to the adventurers (die 1-3) or neutral (4-6).

This is followed by a random table for determining the occupant of a castle and their guards/retainers. Then specific procedures for each type of inhabitant are provided: Fighters will “demand a jousting match with all passersby of like class”, Magic-Users will “send passerby after treasure by Geas if they are not hostile, with the Magic-User taking at least half of all treasure so gained”, Clerics will “require passersby to give a tithe (10%) of all their money and jewels” and so forth.

A vestigial remnant of this structure survived all the way into the Rule Cyclopedia (where tables were still being presented to randomly determine the attitude of castle occupants), but I’m guessing it’s been essentially nonexistent in actual play since 1980 at the latest. And, unlike JG-style hexcrawling, I doubt these “castle occupant rides forth” game structures are likely to enjoy a significant renaissance any time soon.

(Although, on the other hand, there are some potentially interesting things lurking in there: First, the implied setting – in which civilization is so sparsely populated that feudal lords will ride out to meet travelers who pass anywhere within a dozen leagues of their walls – is a fascinating one. Second, note how the structure provides default scenario hooks: Like the treasure maps seeded randomly into OD&D hordes, the quests proffered by feudal lords can spontaneously transform the aimless exploration of the hexcrawl into specific direction. But I digress.)

The other major procedure for wilderness play in OD&D was the idea of “clearing” a hex of monsters (which was a prerequisite for establishing a barony or stronghold):

The player/character moves a force to the hex, the referee rolls a die to determine if there is a monster encountered, and if there is one the player/character’s force must remove it. If no monster is encountered the hex is already cleared. Territory up to 20 miles distant from a stronghold may be kept clear of monsters once cleared – the inhabitation of the stronghold being considered as sufficient to maintain the monster-free status.

This basic structure was greatly expanded in AD&D. (For example, it added a precise system for determining when and how monsters return to an area, along with all kinds of modifiers – like placing skulls and carcasses as warning signs.) And then it, too, faded away.

In fact, if you look back at the early history of gaming you’ll find these kinds of explicit game structures all over the place: The system for artifact use in Gamma World, the mercantile models of Traveller, the posse and tracking system for Boot Hill (which, in the 2^nd Edition, could be tied into a larger competitive structure of lawmen vs. outlaws), the status points of En Garde, and so forth.

What you had, in short, were a bunch of wargamers who were very familiar with creating specific game structures heavily laden with the details of world simulation: After all, they’d been using precisely defined game structures to model historical battlefields for years. Roleplaying games, however, cracked open the game world, and the game designers were applying game structures to a much wider “world simulation”.

Over time, for a variety of reasons, these explicit game structures became more and more simulationist in nature. As the focus shifted from structures that were fun to play to structures that were accurate “models of the game world”, however, the structures became rather boring and (as the details of the simulation became fetishized) often too complicated to use in actual play to any great effect.

What followed next (and this all happened over the course of only a few short years) was almost inevitable: The explicit game structures became vestigial and then, with the advent of universal systems, disappeared from rulebooks entirely. (With the notable exception, of course, of highly structured combat systems.)

But — and this is important to understand! — the game structures didn’t actually disappear from gaming! They are, after all, essential for play. Instead, a handful of the most popular structures became treated as a sort of common knowledge: Everybody “knew them”, so game designers didn’t bother explaining them. (Although, in truth, very few people really thought about them at all.)

In short, the general sense that playing an RPG consisted of nothing more than “the players tell me what they want to do and then we resolve it” settled over the industry. But, as we’ve seen, this is completely false. What happened in actual practice was that GMs would use a random grab-bag of unexamined techniques that they had collected from people they played with, published adventures, and the occasional unique insight. And this, of course, resulted in a lot of frustrating play.

The lack of explicit game structures in the rulebooks – particularly explicit scenario structures – also helped to make roleplaying games a lot less accessible to people who hadn’t played them before.

Imagine that you took the rules for the boardgame Arkham Horror and you stripped out all the rules about explicit turn sequencing (i.e., the scenario structure of the game): Instead, you’re left with some rules about how to make skill checks; how far you can move around the board each turn; how to fight monsters; how you can move through gates and close them behind you; how you can fight Ancient Ones if they wake up; and so forth. But… what do you do with all those rules?

And that, in short, is what virtually every RPG for the past thirty years except for D&D has looked like to newcomers. (And it is somewhat worrisome to note that every iteration of Dungeons & Dragons since 1983 has reduced the amount of explicit game structure in the core rulebooks until, finally, 4^th Edition eliminated dungeoncrawling procedures almost entirely.)

But as we look back at the archaic era of explicit game structures, I think there is something else of import to note that may help to explain why they went way: With the exception of the ‘crawls and a few others, almost all of these game structures were fundamentally incomplete.

Go to Part 10: Incomplete Game Structures