A Penny for Your Thoughts


The goal of designing rules is to change social contract.

When I design a set of rules, I'm trying to change the way that people relate to one another, within the confines of the game. I'm trying to force, trick, or provoke them into treating one another in particular, possibly unnatural ways. I'm screwin' around with their working creative relationships.

Beyond apportioning credibility, rules create permission and expectation. Permission and expectation are the real building blocks of social contract; cunningly designed rules have access to human interactions at a deep level.
I think this is worthy of its own topic.

We all pretty much know what we need in order to have a functional arrangement of permission and expectation. We need checks and balances, we need transparency and accountability, we need to deal well with our failures and build on our successes. We need common ground and common goals; we need tools and mutual support.

Those are the goals of RPG design, then. And those are why writing up a setting and slapping some "physics of the game world" mechanics on it is not the real thing.
It may work well to think of RPG rules as strong or weak, flexible or brittle: a strong RPG draws the players into its particular play, where a weak one allows them to play however comes naturally. A flexible RPG can survive or redirect a broad range of social dynamics, where a brittle one requires a particular social dynamic to be in place, or the game crashes.

On 1-20-05, Matt wrote:

Seems to me that in many cases the "unfinished" aspects of a roleplaying game are tossed into a refuse pile of power apportioned almost completely to the "GM." If one person has all the power, and if the group is accustomed to that lopsided distribution, it's a shortcut to making the game flexible.

Now equally distributing that power AND succeeding in a flexible game, that's something to see.

On 1-20-05, Ben Lehman wrote:

Vincent -- are you on my topic? Hell yes! Am I on your topic?

What I was really trying to say is this: each group brings its own social situation, all the past games that every player has played, and a whole bunch of other stuff to the table of every role-playing game. I mean, like, the fact that two players in Polaris used to get in fights about girls all the time -- that's way more important than "who bids first in conflict?" Most designers expect this, hence all the shortcuts (games that don't describe "GM" and "Player" roles are essentially taking a big chunk of system straight from D&D). No wonder gamers are big rule and setting hackers! They're already bringing half the system to the table already!

Your description of strong/weak and flexible/brittle is really a natural extension of this. But, I would add, I don't think that strong and flexible rules are all that is useful. Polaris makes heavy use of brittle rules and, while that may be to the game's detriment, I think it helps shape the play experience a lot.

Can I actually revise your statement "The goal of designed system is to reshape social contract," where we understand system in the broad sense I was using it in that thread? Because I think we're way beyond "roll vs. TN" here. The setting of Dogs in the Vineyard, say, is all about rearranging what we can play about, in terms of morality and religion. The TV color of PTA is all about rearranging our distance from the game. The tragic color of Polaris is all about making it okay to lose. These are all about permission, not necessarily credibility directly.


On 1-20-05, Chris wrote:

Hi Vincent-

On point! I think its interesting because a quote that comes to mind very often when I think about social contract is something Clinton said about Inspectres- something along the lines of -"Of course you can give players control and power- who would want to ruin the game?" which also went with the idea that you're playing with a socially viable group.

It's interesting because the "brittleness" of Inspectres is that it expects social contract to be the primary restraint for people jumping up and doing crazy shit in play. Compared to Universalis where the system gives a legitimate process for social pressure of the group to counteract any given individual from going too crazy.

So now- here's a question- what determines flexibility/brittleness? I have a suspicion that it has everything to do with what kind of options a game allows for different approportioning of credibility.

For example- HQ can be played straight Illusionist to full blown Narrativist(with or without director stance). Riddle of Steel doesn't do Sim very well at all(at least if you're following the rules as written). Universalis would be highly flexible, because through the option of making gimmicks one could "build" a traditional GM/player set up within it.

And- in most cases- is this what people appeal to in incoherent games? Because credibility isn't clearly divided that people say the game is "flexible" when in fact it might just be failing to address critical points of play?

On 1-20-05, Vincent wrote:

Ben: Yep.

In fact let me take the "preexisting" and "already" out of brittle. That way I can say that Dogs is a strong+brittle game, in that it requires a very particular dynamic to work, but then also makes that dynamic all but certain.

And in that case, strong+brittle and weak+flexible are every bit as good as strong+flexible. Weak+brittle, only the group that game's designed for can play it - but for that group, it's as good as any other.

This is getting outside the scope of real things / fictional things / people, though, you're right. [Which is why I moved it here.]

On 1-20-05, Ben Lehman wrote:

Oh, you're using strong/weak and brittle/flexible to describe whole game systems? I was using them to describe specific rules. Huh.

This actually reminds me of something that I was thinking about earlier -- RPG players are usually talking about several things at once when they talk about rules heavy vs. rules light. I was thinking that it would be useful to seperate out the weight of the ruleset from what you call the strength of the rules set -- Most gamers thing that heavy/strong and light/weak are the only possibilities but a game like, say, MLWM is very much light/strong because, while the rules are not complex, absolutely apply in all circumstances.

I don't think and anyone has thought about your "flexible/brittle" binary before.


P.S. Referenced Forge Thread

On 1-20-05, Matt wrote:

By particular dynamic in part we're talking about the ability of the players to express satisfaction or lack thereof to one another, right?

To my great sadness, I haven't had the opportunity to play Dogs yet. But it seems that a GM could abuse some power by pushing conflicts upon players rather than allow them to seek it out. Like, you know, bad guys show up in the night with axes. When fallout can end the game, that's going to require some amount of group dynamic. Primetime Adventures, in contrast, has specific constraints over the producer's ability to manipulate conflicts. Is that part of what you're talking about with strong/weak flexible/brittle?

On 1-20-05, Emily Care wrote:

I suspect I won't be the first to reply to this, but here goes.

Matt Wrote:
But it seems that a GM could abuse some power by pushing conflicts upon players rather than allow them to seek it out. Like, you know, bad guys show up in the night with axes. When fallout can end the game, that's going to require some amount of group dynamic.
That's actually just fine in Dogs. Because whether the ax swinging bad guys could end the game or not gets determined by the players in conjunction with the GM, not by the GM alone. The GM is actually freed up by this--you don't have to pull your punches to safeguard character script immunity. It gets negotiated in play with each conflict, and if your players don't want to give you that, you don't get it.

Primetime Adventures, in contrast, has specific constraints over the producer's ability to manipulate conflicts.
What specifically do you mean? Dudget and fan mail? That gives all the players influence, but the parameters of any conflict are set the same way they are in Dogs: by discussion of "what's at stake". At least, that's as I recall.

On 1-20-05, Ben Lehman wrote:

Emily -- what Matt is talking about is that the Producer in PTA cannot arbitrarily set difficulty numbers -- they have to pay to make life tough for the PCs (this is brilliant! brilliant!) In Dogs, I could as a GM say "Ten guys with axes attack you. I roll 50d10. You all die." You wouldn't, but you'd be within your rights as the GM. (This is something that has always bothered me about Dogs -- it seems to me that the GM can still mandate success and failure by the number of dice that they pick up.)

It occurs to me that the auto-generation of NPCs may serve to limit this. I realize that I have been guilty of thinking of that thing as a optional tool, not a required restriction.


On 1-20-05, Matt wrote:

Ben Says: "it seems to me that the GM can still mandate success and failure by the number of dice that they pick up"

Yes, that's exactly what I'm getting at. You have to rely on the players' ability to both recognize that and to act upon it if it happens. and the trouble isn't really the egregious acts like your example. It's the slight slips over that line, wherever it is.

On 1-20-05, Vincent wrote:

Ain't a single thing in Dogs that's an optional tool. You knew I was going to say that! Frickin' game designers, we think we know better than you.

That said, "ten guys with axes" is a large number of dice; that's a conflict you can't win by yourself. But see here for some subtleties about being killed in your bed.

Other non-optional things - play the town, don't have an outcome in mind, things like that - will keep the GM from mandating success or failure. Except when it's genuinely called for, which it occasionally will be.

I cop to my bad designs. I'll stand by this one as good.

On 1-20-05, Ninja Hunter J wrote:

Ben, I always really dig what you write, but you're flat-out wrong here.
You said:

"Ten guys with axes attack you. I roll 50d10. You all die." You wouldn't, but you'd be within your rights as the GM. (This is something that has always bothered me about Dogs -- it seems to me that the GM can still mandate success and failure by the number of dice that they pick up.)"

First off, the players could Give before the guys with axes chopped them to bits. Then they could launch a follow-up conflict where they set the town on fire and shoot everyone coming out.

Second off, and this is important, what's at stake is not "Do they kill all the Dogs?" it's "Do they stop the Dogs from finding out that the Minister eats children and loads his guns with their teeth?"

So, in this example, they Dogs would be saying, "holy manure (no swearing!), the whole town's come out to chop us up, we'll figure out what's wrong after running away and regrouping."

At that point, maybe they took a little fallout before realizing that they should bail, but the fallout is no more likely to be that they're dead than if they were standing in a street at high noon in a showdown.

Remember the example in the book, "The dude hits you in the head with an axe while you're asleep!" and it comes out OK?

On 1-20-05, Ninja Hunter J wrote:

(the crap? I didn't double submit! Look out for more PHP weirdness, V!)

On 1-20-05, Ninja Hunter J wrote:

... oh, and is 'illusionist' jargon that I haven't heard before, or is it what I was talking about the other day, where you deliberately find yourself in the world, just doing what you're doing because you want to see it happen?

On 1-20-05, Vincent wrote:

"Illusionist" means that the GM is the sole author, but uses covert techniques to maintain the illusion that the other players are contributing. Contrast with "participationist," where the GM is the sole author, and everybody knows and is cool with it.

On 1-20-05, Chris wrote:

It's also interesting to note how giving one person full authority over the shared imaginary elements has allowed for a good deal of social contract abuse- like the town of axe murderers example. It is only "justified" by the rules in that the GM is given authority to produce those situations in the shared imaginative space.

Its also neat to note how Trollbabe explicitly allows players to sidestep stupid stuff like that by establishing conflicts themselves("Oh, the townspeople all show up with axes to kill us? Ok, the conflict is whether I kill some of them before they come to their senses or after?")

All that said, even the strongest design only holds up if people are actually following the techniques in the text(check many instances of Riddle of Steel and Sorcerer where folks "don't get it")

On 1-20-05, Matt wrote:

I suppose I could thoroughly read the game before I make vague accusations about it, but reading is for chumps.

If I'm understanding right, there's a Trollbabe-like ability to withdraw from a conflict between the seeing and raising and stuff?

On 1-20-05, Ben Lehman wrote:

J -- Glad to hear that you like what I write sometimes :-) I'm totally cool with being wrong right now.

My point was about Dogs in dysfunctional situations -- as I understood it, the GM has the authority to throw as many dice at you as she cares to, excepting very specific circumstances. Vincent has told me I'm wrong about this, 'cause the GM is limited to the people she got from the batch NPC generation system. So I'm wrong.

But it occurs to me -- even if that was the case, it wouldn't really be a problem. Dogs is, as Vincent says, strong and brittle. It breaks if people aren't going to buy into it. It breaks if anyone tries to break the system during the game, especially the GM. PTA, in the gripping hand, is a little more flexible -- it doesn't break even when there is a person trying to break the system, as long as everyone isn't trying to break the system. Still breaks if someone doesn't buy into the schtick, though.

I'm going to be all hippy non-judgemental for a second and say that this isn't a bad thing or a good thing. Having some conditions under which your game simply fails to operate is, honestly, necessary. Yet clearly a little flexibility is nice.


P.S. Dogs and PTA were my two favorite games I got from GenCon. And now, I'm geeking out about both of them at once. I'm in hog heaven.

On 1-20-05, Vincent wrote:

Matt: Well, no - it depends what the conflict is. If I say "10 guys with axes, what's at stake is: do they kill you?" then that's how it is. You're in that conflict until you win or lose, and if you lose, they "kill" you.

I put "kill" in quotes because there's an intentional loophole when you're killed by stakes, not by Fallout. Being killed by stakes is the same as being probably-killed by Fallout. If you like, when I say "what's at stake is: do they kill you?" what I'm really saying is "what's at stake is: do they chop at you for a while and then leave you for dead?" The odds of being chopped at for a while and then left for dead, yet surviving somehow, are pretty good. Read the Forge post I linked to, chump.

If the conflict is something else, like "10 guys with axes, what's at stake is: do you get the dirt on the minister?" then it's exactly as Ninja J says. You're like "10 guys with axes? Screw it. We'll get the dirt on the minister some other way."

On 1-20-05, Emily Care wrote:

So if we're talking about the permission aspect here (ie what right does the GM have to hose the players, or what right does the player have to sidestep or reverse a conflict), how are the expectations for what actually happens set in these games? Or are they doing both at once?

On 1-20-05, Matt wrote:

"The odds of being chopped at for a while and then left for dead, yet surviving somehow, are pretty good. Read the Forge post I linked to, chump."

Oh yeah? Them's fighting words. I escalate. Dodge this, sucka!

Emily: yes, that's in tune with what I'm being all fussy about. Who in Dogs decides that the stakes are "do I get chopped up and left for dead?" Do I get any say in that? Or do I play Dogs accepting the possibility that at any moment my guy could lose a testicle or something? Even if I survive the ordeal of being hacked up, it's really likely that I'll suffer lasting harm, no? So if we keep playing and like every night there's a new crazy librarian trying to axe me, and I'm down to seven fingers and missing an ear, it might feel a little deprotagonizing to me.

On 1-20-05, Ninja Hunter J wrote:

The stakes bit is something I've been thinking a lot about recently because it's how I've been phrasing PTA conflicts, too (with more or less decent results), and I've decided that the stakes is "Do they kill you?" sucks in most any situation. If that's what it's about, if it's really about killing you, if it's personal, some sort of vendetta thing, then it's great.

BUT! If it's not, if what's at stake is, "do they keep you from getting the dirt on the Minister," or "Do they make Cathedral City leave us alone once and for all," then all sorts of story can happen.

It depends on the level you want to do things, but as V said, there's a loophole when it's personal. It inherently adds to the story. Think about Yojimbo. The character's totally badass in a fight, but the entire gang jumps him, beats the snot, tar, and sense outta him, and then the story gets all the awesomer for it because now he has to do other stuff to accomplish his goals.

And if you haven't seen Yojimbo, you are not my son.

... until you do.

... uh, and even then, probably not, but for different reasons.

On 1-20-05, Ninja Hunter J wrote:

Matt, it's important to remember who writes down the nature of the fallout. I'm giving these testicle-losing examples beacause that's what my favorite Dog is like. He's making up for being a Mountain Person, he feels this guilt, so he gets the crap knocked out of him deliberately.

Your fallout can just as easily be "These folks don't deserve savin' no more - 12d4" and you stay in the saddle.

On 1-20-05, Vincent wrote:

"Even if I survive the ordeal of being hacked up, it's really likely that I'll suffer lasting harm, no?"

As it turns out: no.

Check this action.

Me: 10 guys! With axes! Stakes: do they kill you?
You: Dangit with the axes again. I give. They kill me.
Ben: When they're good and gone, I rush to his aid.
Me: Stakes: do you save his life? I roll 4d8+4d10 with no escalation!
Ben: I roll some dice, plus Traits, plus I escalate. I win!

You live without taking any lasting harm a'tall.

Anybody can say what's at stake, and everybody has to agree to it - but I'm well aware, "everybody has to agree to it" is subject to social pressure. The safeguards live elsewhere: in followup conflicts (especially this mandated last-chance one), in the How to GM rules, and in the Fallout.

Because taking Fallout is good for you, as player. Having your character lose a fight gives you more dice, thus more say in how things go, not less. Plus you get to decide what your Fallout is, not me, no matter who chose the stakes.

Swear to God, to GM Dogs abusively you have to go far, far out of your way. The rules will fight you every step, and you'll get sick of it and play something else before you beat them.

On 1-20-05, Emily Care wrote:

Loud and clear, Matt.

So if we keep playing and like every night there's a new crazy librarian trying to axe me, and I'm down to seven fingers and missing an ear, it might feel a little deprotagonizing to me.
Yeah, this brings up another place where the lines get drawn--who gets to say whether the flock of murderous, axe-wielding librarians gets to wake you up with a cold-steel scalp treatment? I know this cool game where everyone takes turns framing scenes....

(But jumping jehosephats! How'd the poor, innocent librarians get mixed up in this jumble? Most lib's I know would be more likely to kill you with paper cuts than axes. Maybe it's that calm exterior--underneath they're seething.)

Oh! And wasn't I surprised to learn that in PtA the scene framing rules work thusly:

1). Player requests a scene (gives Focus, Agenda & Location)
2). GM sets up the scene.

I can't imagine not having the players just intro the scene. Does anybody really play that way? (No, I'm not biased here at all, no way :) But why not go the whole nine yards?

On 1-20-05, Vincent wrote:

Yes! Let's grill Matt instead of me for a little!

I meant to say: Taking Blows is good for you as player. Losing conflicts doesn't give you Fallout, of course; taking blows does.

On 1-20-05, Ben Lehman wrote:

Vincent: Your example is hilarious. I seem to be really giggly today, apparently.

I, for one, totally and utterly concede the point. Although I just noticed that, if you really want to kill players, you need the stakes to be something other than "Do you die?" Like, if they care about saving Sister Patience from temptation, and the stakes are "Does Sister Patience sleep with that no good out-of-towner," and it goes directly to guns, you're much more likely to see a dead PC. In the hands of a *skilled* killer GM, the conflicts are going to be about things that the PCs care about and involve gunfire.

And that is so totally awesome. I have no words.

Emily: I play PTA that way. At least, the last time I played it, it was with a bunch of people who were a little new to this whole "scene framing" thing so I, as producer, worked together with them to build up the scenes. I imagine we could have shed that if more had been played, but the collaborative aspect was nice. (The game was a one shot. Show was called War Stories: News from the Front. It kicked ass.)


On 1-20-05, Vincent wrote:

Now here's me praising Matt instead of grilling him, god dammit:

Before PTA, a one shot called War Stories: News from the Front was out of reach. Unreal, that game.

On 1-20-05, Vincent wrote:

I think it's important to note that it's the safeguards that get the players to grant permission to one another. I've played lots of games where as the GM I had to pull my punches, because my rights outstripped the permission granted to me.

This is the responsible-GM flip side of Matt's fear.

On 1-20-05, Emily Care wrote:

Ben: That makes so much sense. I'd look at it as training wheels, and let 'em fly on their own once they cut their teeth on it.

So, back to grilling Vincent:

In Dogs, how exactly is it communicated to the players that they have the right to pull a Ben and bring their fellow dog back to life? It's in the safeguards, of course, but what's their cue to know that that kind of power is within the bounds? (Ain't got no copy of the game at work, unfortunately, so I'll be lazy & ask instead.)

I've played lots of games where as the GM I had to pull my punches, because my rights outstripped the permission granted to me.
And doesn't that just say it all.

On 1-20-05, Ben Lehman wrote:

Here's some grilling of no one a'tall, just a thought about permission and rights.

Rights > Permission -> Dysfunctional play, either killer or pulling of punches.

Permission > Rights -> Rules drift to the point where permission = rights, or just tossing out the rules. In general, milquetoast rules.

Personally, I think that the ideal balance might be setting everyone's rights just slightly greater than their permissions, allowing people the push the edges. But I like edge-play in my games.


On 1-20-05, Eric Finley wrote:

Wow, that's come a long from flexibility and brittleness. Not trying to redirect, neat stuff, but I am a little curious how the permission thing meshes with the flexible/brittle, strong/weak dissection. Tentatively, looking at a quick overview of examples, it looks to me like there's a link between "grants heavy & explicit permissions" and "strong system" in the above sense. Trouble is, my gut says this may be a case of correlation not equating to causation.

And as an aside in reference to a comment about five up... Ben, if you really want to kill *players* (sic) then we're talking some kind of kpfs: the LARP thing here, and I'm not sure we want to know about it...

On 1-20-05, Vincent wrote:

Eric: oh - well a strong RPG is one where the game designer decides what expectations and permissions you'll have and grant when you play, and a weak game is one where your group does. A brittle game is one where only a certain set of permissions and expectations will work, and a flexible game is one where you can successfully play with any of a wide range of permissions and expectations.

On 1-20-05, Eric Finley wrote:

That's a good way of putting it. Permissions and expectations as elements of the CA, basically.

Note that having thought about it I suspect that "rights" as Ben is using it may be a red herring. If rights don't match up with permissions, it's simply the case that the rules fail to acknowledge the Lumpley principle. The permissions are what really exist.

On 1-20-05, Kip Manley wrote:

(You know, I go to the trouble of printing out everything over the past couple weeks and the Forge threads too, in the hopes of catching up, and then you post again and in one day there's, like, 30-some-odd comments; oy.)

On 1-22-05, Vincent wrote:

Jump in when and where you can, Kip!