Author: Josh Fox

I recently bought a copy of Kagematsu through the ever-wonderful Bundle of Holding (highly recommended if you haven’t come across it). Kagematsu is a game about the attempts by the women of a Japanese village to woo a wandering ronin in the hopes he will save their village from a looming threat.

I haven’t played the game (yet) but reading it has highlighted an interesting issue that I’d like to talk about here. The issue is: mechanics that ostensibly do nothing, but actually exert an important psychological effect.

Here’s a flow-chart I made showing how Kagematsu is played, from a mechanical perspective.

The solid lines and boxes represent game events and the flow of time. The dotted lines and boxes represent game stats and the flow of mechanical causation.

There’s a couple of things  I want to highlight here.

The first concerns Pity. In Kagematsu, in every scene a villager tries to elicit an affection from Kagematsu (the ronin); this is carried out through the mechanics shown schematically above. At the end of the scene, regardless of the mechanical outcome up to this point, Kagematsu’s player must decide whether to allocate the villager a point of Love or a point of Pity. That is represented on the diagram by the dotted arrows from “End scene” to “Love” and “Pity”.

Notice that while there are a couple of dotted arrows from Love to other bits of the diagram (Love is important in the game; it improves your chances of winning affections in future scenes, the Kagematsu uses it to confront the threat at the end of the game, and also, though this isn’t shown on the diagram, it shows who gets to go off into the sunset with Kagematsu if he defeats the threat), there are no dotted arrows from Pity to other bits of the diagram. What this means is that mechanically, Pity does not do anything.

But Pity is an important part of the game, because of the influence it has on player psychology. If we didn’t have Pity, the choice would be: award the villager a point of Love, or don’t. It seems pretty clear that this would lead to a lot more Love being given out, simply because the alternative is to do nothing. By introducing Pity, even though it has no mechanical effect at all, we give the Kagematsu a real choice – do you love this woman more than you pity her? It also makes the choice somewhat less controversial, since while one might be peeved at not receiving Love, one is probably more likely to accept that one deserves Pity.

The second point I want to highlight concerns the Shadow Track. Every time anyone rolls a 6 during a scene, that 6 is placed on the villager’s Shadow Track. If three 6s are placed on the Shadow Track, the scene is interrupted by the looming threat. The villager describes how the threat breaks the scene up. The villager does not receive her affection; indeed, it is as if she had never attempted to gain it.

The Shadow Track does have a mechanical impact, in that it nixes the last affection attempt and ends the scene prematurely. But the overall effect is to slow the game down a bit, rather than to actively push it in any particular direction. So while it’s less empty-seeming than Pity, it is fairly weak mechanically speaking.

But the constant risk that the threat will muscle in on a scene, and the occasional reminder that the threat is present, have an important psychological impact. They reinforce one of the central themes of the game and boost atmosphere. They remind everyone what’s at stake.

My point is, both of these are examples of good game design. On paper they look like mechanical dead-ends, failing to influence the key game outcomes much if at all. My initial reaction on reading them was to think the designer had made a mistake. But their psychological impact is important. I shall try to bear this in mind for my own design work.

GNS theory cuts roleplaying creative agendas up into Gamism, Narrativism and Simulationism. The first two of these get more play, and greater respect, than the last, in my opinion. Yet I will argue this is because Simulationism has been misnamed; and in fact many important roleplaying innovations have been in a so-called simulationist space.

Quick disclaimer here: I’m going to talk about simulationism as described by Ron Edwards, but this is not really a theoretical article, and even though it tries to point the reader into a different view of what simulationism is, it isn’t any kind of attempted takedown of Edwards’s theory or other such shenanigans. Indeed it is rather selective in quoting Edwards, which would be a cardinal sin in a theory essay, but I hope I can get away with this in the above context.

The man who codified the concepts of GNS, Ron Edwards, talks about simulationism in his essay GNS and other matters:
Simulationism “is expressed by enhancing one or more of the listed elements [Character, Setting, Situation, System, Color]; in other words, Simulationism heightens and focuses Exploration as the priority of play.”

He talks about simulationism as focusing on exploration, a concept that is important to all roleplaying, but assumes pre-eminence and becomes an end-in-itself in simulationist play. Yet the term “exploration” is misleading. It implies a pre-existing reality which we together explore; or perhaps it is intended to imply a single fiction, probably created by one person, the GM. A more neutral term is creation, because that is in reality that which is being explored is simultaneously being created, whether by the GM or by the players, individually or collectively.

Similarly “simulationism” suggests an attempt to replicate some ideal – a realistic game world, a particular genre’s conventions, a well-realised character. But play that focuses on the act of creation (as opposed to exploration) need not be about simulating anything; it is often about the creation of imaginative, evocative content which might or might not relate to an ideal. Creativism would be a more accurate and comprehensive term for this type of play.

If we were going to do definitions, creativism might be defined as:
Creativism “is expressed by enhancing one or more of the listed elements [Character, Setting, Situation, System, Color]; in other words, Creativism heightens  and focuses the creation and development of the shared fiction as the priority of play.”

Now I’m not seeking to pick an argument with Ron or anyone else here. Instead, I’ll pause to note that “creativism” already probably sounds much more appealing to many roleplayers than “simulationism”, which is resonant with computer games and wargames – popular with some, but not with all. “Simulationism” hints at rules and dry technicality, while “creativism” points to imagination and shared endeavour. I’d also like to talk a bit about where that reframed concept leads us.

For one, it leads us to focus on questions about who creates the content of the game. Much of indie roleplaying design has been concerned with handing over creative authority, or at any rate extending it, from the GM to the rest of the players. Indeed, many indie games have no such distinction. This has some implications for the pursuit of drama (narrativism) and challenge (gamism) but its biggest impact is on the act of creation itself. It enables more people to take part in the act of creation, and thus plays into a creativist agenda.

We can go further. Creativist concerns lead us to look at how the act of creation is regulated. Drama points, games structured into acts, Dogs in the Vineyard’s traits, and many more, serve to regulate the flow of content creation. They serve other purposes too, of course; but one major impact is to shift creative power and constrain the creative act. This leads to the well-known paradox that one can sometimes be most creative when one does not have a free hand to create anything one wishes to.

I’m spending a lot of time thinking about these issues at present. And indeed, I’ve been teasingly referred to as a simulationist. Maybe I am – but these days I feel like I’m more of a creativist.

So, further to my last article. I have been thinking about this a bit. I think that many of the criteria I set out are, fundamentally, compatible with each other. But I darkly suspect that my first and second criteria (Drama; and Colour and Impact) may not be entirely compatible with my third (Tactical Depth). This is because tactics implies detail and precision on positioning (whether spatial, temporal or otherwise within the fictional space) and, importantly, time to carefully consider combat decisions. Shall I attack this opponent, or that one? Shall I use this combat move, or this other one? All of this makes the fight more interesting from a strategy/gaming perspective, but crushes any sense of atmosphere and pace.

On a related note, I have been doing a bit of thinking about how fight scenes are portrayed in books and movies, and how this differs from the way it typically works in RPGs. One big thing that I notice is how <i>bitty</i> RPG combat is. It’s all “your turn, now my turn, now Bob’s turn” and nobody gets to build up a flow. A really dramatic fight scene in a book or movie is more likely to focus on a single character or small knot of fighting for a comparatively extended period, like a paragraph or two, and so we’re on the edge of our seats as that fight develops and we wonder who will live and who will die. We never build up that sense of anticipation in an RPG because when Bob is down to his last hit point we have to wait for everyone else to take their turn before we find out what happens to him. This is another area where incorporating this insight into an RPG system would tend to push us away from a tactics-focused system, because if we’re focused on one small part of the fight scene for a longer period, there’s less chance for other combatants to make tactical choices to break off what they’re doing elsewhere in the scene to intervene.

So, question for any system designer: which of these do you most care about? Drama or tactics? It isn’t like they are totally incompatible; you can have a sort of “summing up” phase after all the gubbins of tactical decision-making have been sorted to bring back the rich description of the action, or you can blend a kind of light-weight tactical system in with an otherwise more freeform affair. But there is a limit to this, and trade-offs to be made. I think there has been a lot of work in the first space (heavy focus on tactics, with description sort of crowbarred in), but less in the second – combat systems (as opposed to generic systems, remember) focused on drama, with less focus on tactics.

Whenever I think about operating in that second space, I start to get worried about descending into the generic. What I mean by that is: combat starts to feel like it doesn’t matter what decisions you are taking, as they are all mechanically the same. Does it matter whether I’m trying to kill this person or KO them, capture them, drive them off? It feels like it should. But in order to keep things simple and pacy, I find myself starting to design out those distinctions. I end up with “roll the dice, if you succeed impose a condition – give it a name, move on”.

I really want my fight scenes to feel dramatic. Grinding through a tactical battle scene can be fun – I enjoy war games, after all – but I’d like to be able to breathe life into fight scenes so they really feel edge-of-the-seat.

What games have you played that gave you a real sense of the excitement of a fight scene?

So, if it wasn’t apparent already, I’m a system geek. I obsess about rule systems for RPGs. And none more so than the rules for combat.

Combat rules can be anything from “the players and/or GM describe what happens based on what they think would be cool and dramatic” to a sort of hyper-detailed miniature wargame. No matter. There are a list of things that I want to some degree or other from any combat system. Now, I think there’s a separate discussion to be had about whether some of these might be incompatible with each other; that’s for another day.

1. Drama. The whole point of having a separate combat system (as opposed to just using the bog-standard conflict resolution system for your game) is to deliver the suspense and excitement that a fight scene in any book or movie can provide. Will the hero live or die? Unlike most books and films, in an RPG that might be a question that could have more than one answer. The system needs to deliver that sense of drama and risk.

2. Colour and impact. A combat system needs to promote description that is rich and exciting. I want to feel like I’m in a fight. You might think any combat system can do this; it’s just down to whether the players and/or GM can manage to describe things in a sufficiently interesting way. Probably true to an extent – but there’s no doubt that a combat system can kill this off by encouraging a mechanistic “I hit him he hits me back” type dynamic.

3. Tactical depth. Being able to make interesting tactical decisions in combat and have them affect the outcome is an important part of why you bother having a combat system at all. You can do this (and many systems have!) through hyper-detailed battlemaps and oodles of options for special moves, feats and whatnot, or you can do this through a system-light approach where the GM and/or players assess the effectiveness of the tactics chosen and reflect this in the outcome described. The latter poses a risk, of course – I make what I consider a sound tactical judgement and the GM decides it wouldn’t work, for their own inscrutable reasons. Meanwhile the more “crunchy” approach ensures that there’s an agreed framework for making tactical decisions, but risks requiring unmanageable mechanical complexity in order to support a range of possible strategies. There are other ways of introducing tactics, through a kind of system-focused approach (like Dogs in the Vineyard’s see-and-raise approach), but these often lead to a focus on the dice at the expense of the fiction.

4. Realism. Controversial, this. It isn’t strictly necessary, but many of my gripes with commercial systems stem from a perception that they aren’t realistic in some way. Take D&D3e’s rule that smaller characters get a bonus to their hit rolls and a bonus to their armour class. Right: so in any given fight, the smaller fella is going to prevail, on average. Right? Let’s try putting the world’s best featherweight boxer in the ring with a mediocre heavyweight and see what happens, shall we? Of course, realism doesn’t mean you have to be hyper-detailed; there’s an argument that the players and GM are the best arbiters of what they consider realistic, so their own judgement is the best tool for ensuring a combat that has verisimilitude for them, without the need for reality-simulating mechanics.

5. Quick, smooth, easy to understand and reference. This is a basic requirement for any rule system. Systems which address tactical depth and realism best are often so mechanically detailed that they can be impossible for anyone without a maths degree to understand, take forever to look up and require five minutes just to count up the number of dice you have to roll. Needless to say, I do not consider this to be a good thing.

6. Supports a range of outcomes. Far too many systems are all about one party or the other being killed. Yet the majority of fights in real life are fought to subdue, intimidate, prove a point, force the defender to run away, and so on. For example, most RPG combat systems use rules for knockout and grappling which make these approaches very challenging in character (low chance of success) and also hard to handle out of character (mechanically complex). Again, not a good thing. I want it to be as easy for my cop character to cuff a perp as it is for my sniper to blow away a terrorist. Or whatever.

Phew! It’s no wonder I spend so much time thinking about this stuff – those are some challenging criteria. Anything else you look for in a combat system that I’ve missed out?

I was prompted by a post over on Department V to go furtling through some old Forge articles, and I stumbled upon this bit of text tucked away behind some musings about coherence.

“In most Narrativist designs, Premise is based on one of the following models.

A pre-play developed setting, in which case the characters develop into protagonists in the setting’s conflicts over time. Examples include Castle Falkenstein and Hero Wars. Pre-play developed characters (protagonists), in which case the setting develops into a suitable framework for them over time. Examples include Sorcerer, Everway, Zero (in an interesting way), Cyberpunk 1st edition, Orkworld, and The Whispering Vault.

I have observed that when people bring a Narrativist approach to Vampire, Legend of the Five Rings, or other game systems which include both detailed pre-play character creation and a detailed, conflict-rich settting, they must discard one or the other in order to play enjoyably.”

This is interesting to me. It makes a kind of sense: if you set up an immutable (or at least, relatively fixed) setting and an immutable (or at least, relatively fixed) set of characters at the start of a game, chances are good that these are not going to work well together. The players don’t know all of what was in the GM’s mind when writing the setting, the GM doesn’t know all of what was in the player’s minds when writing their characters, and if everyone insists on staying faithful to what they pre-decided, chances are you’re going to get some friction.

In fact, I have observed this in many games. The GM writes an awesome, detailed setting that they just can’t wait to set the characters lose in. The players read a light summary of the setting, this triggers a cool idea for a character and they go wild writing up a history for them. All too often, one or t’other ends up feeling their vision is being compromised, or that what they have created doesn’t quite “fit” with the rest.

Certainly from a narrativist point of view it seems relatively high risk – is this going to create interesting issues to resolve in play?

It seems to me that the design of Apocalypse World very much plays on this observation. The players create their characters and then, collaboratively, seed the world. The GM adaptively brings the world to life and introduces elements of conflict, reacting to the characters the players have created. The exhortation to the GM not to plan anything out seems like it must have at least partly had this thought in mind.

Anyway, interesting. I’m pretty sure a lot of my campaign design has totally broken the above advice. I’m not saying this has ruined my campaigns, far from it. But I and my players have certainly had to be ready to adapt things over time to avoid disappointment.

So, Dan Maruschak recently posted to Story Games (the G+ community, not the forum; which you would think they were the same thing, and they are – except they aren’t) about the frequently expressed view that too many/too complicated rules are bad in a roleplaying game. Now, his post had a point all of its own, which I shall ignore because I want to talk about something else. Take that, rules!

Anyway. In discussion on the said post, I arrived at the view that there were two types of “rule”, which I shall here call structure and mechanics. Why is this relevant? I shall tell you if you would accompany me to the next paragraph…

Glad you could join me! The point I was responding to in making the above distinction was that sometimes, rules make roleplaying easier. Take a simple example. Fiasco has almost no in-scene rules. It essentially leaves the job of running scenes completely unconstrained – sure, one person sets the scene while another bunch of people decide the broad outcome (or vice versa) but everything else is down to whatever you collectively want to do. And the thing is, that works for some people, but for others it leaves them lacking direction and unsure when they should jump in. You have to develop the kind of culture that improv groups make use of all the time, and developing that culture can be challenging.

In contrast, Fiasco makes generating the overall scenario for the game much easier by providing a basic setting and a bunch of simple rules for generating story elements. You take turns, and nobody is in doubt about what they can and can’t do during this stage of the game.

So, weirdly, the most rules-heavy bit of Fiasco is in some ways the easiest and smoothest part of the game. All those rules didn’t get in the way after all!

…which brings me back to my point about structure and mechanics. See, I think Fiasco’s set up phase is not really a “rule” as traditionally conceived in roleplaying games. This is a bit of a vague concept which I’m having trouble articulating, but what I call a mechanic – the traditional RPG rule – is a very well-defined procedure for taking a well-defined input and generating a well-defined output. “When you are hit by a short sword, roll d6 and subtract it from your hit points.” “You can take two half actions or one full round action every combat round.” …that kind of thing.

In contrast, the Fiasco set up isn’t really like that. It’s all “before you start the game you should create some elements to use in play”. Now, I’m contradicting myself here slightly (did I mention I’m having trouble articulating this?), because the element generation tables have all the hallmarks of what I’m calling a mechanic, and the rules about how you arrange relationships and other elements around the table look like that  too. But the overall effect is merely to guide play towards a relatively ill-defined form: a structure, if you will. Similarly, Fiasco’s two-act structure and its token-based scene resolution are designed not truly to constrain play but to provide a framework on which to hang your story. Likewise, defining roles (is there a GM? What do they do? If there isn’t, how does that work?) is more about setting a framework rather than fixed procedures. This is all what I call structure, and although it kinda fits in the category of rules, it serves a radically different function.

Now apropos of Dan’s discussion, I’m not saying that structure is good while mechanics are bad. But it seems to me that roleplaying games have historically had a tendency to major on mechanics and leave structure to the GM to work out. And, furthermore, they have tended historically to err on the side of too much mechanic (for some people’s tastes) but very rarely got even close to too much structure. Even Fiasco, which is quite a structured game by RPG standards, is in my view not structured enough.

So in principle: more rules is neither good nor bad. But in practice, more mechanics is often going to turn out to be too much, while more structure is very unlikely to be too much. That may not stay true, if RPGs continue to develop and diversify, but even post-indie revolution it’s still the case for most games , in my opinion.