If your table consists of reasonable people, with shared ownership of the elements of the game and a pretty consolidated view of the general direction of the fiction, you don't need stakes. You can just go with victor narration and mutter the GMs mantra 'In Gamers We Trust'

If you have a fucknugget at the table, or have players who are more interested in making their character SHINE, then you need stakes.

You could, of course, have a halfway house. Say a system where, when there is a conflict, there are some 'keywords' that have to be included in the resolution - to at least guide the narration. So, in OA, you might have a cop trying to chase a murderer down an American style fire escape and the keywords might be 'Fire Escape','Rain','Shadows' and 'Dumpster'. These are set by the players before the conflict is resolved and the winner gets to spin the tale using those words.

(Actually if you don't use that, I might well use it for something else ... hmmmm!)

ANYWAY, the other issue is with limitations on the level of 'victory' that is narrated. So in traditional games the GM is the boundary keeper of the game and is trusted to stop people crossing that boundary. In a stakes game, the agreement of the stake setting acts as the boundary keeper. What would be your stakeless boundary keeper? Whats to have the winner of, say, an air guitar conflict suddenly narrating his opponent dead due to nut allergy!!

I think its very do-able, it just might need a little bit of round about thought.

Neil

Take the King's shilling at http://www.omnihedron.co.uk/dutyandhonour/

The thing you have to watch is that you're not just rolling dice to see who speaks and then they can say < anything >.

I like the method is Sorcerer and its children:
A free-and-clear so everyone is clear about what the conflict at hand actually is, and what everyone is trying to do (I punch him, I dodge, I grab the diamond...)
A conflict roll
The winner "winning" and the losers "losing" based on what everyone was trying to do and what everyone rolled.

In complicated conflicts you will have a heirarchy of winning and losing. (You can see this in 3:16 where at each step someone can cancel downward but they can't affect the people level with them or above them).

So, for OA I think you're almost there with how you talk about it. The "stakes" thing is all about the free-and-clear becoming a succession of escalations and fixing of outcomes that have a binary behaviour. The "narrator" then has no narration to do, we just accept that pre-narrated Stake A, B or C happens.

The worry I would have is that you have one person speaking at the end and how they are constrained by what everyone was doing?

Gregor said: The thing you have to watch is that you're not just rolling dice to see who speaks and then they can say < anything >.

Neil continues: You know, that was what I was trying to say, but in about a fifth of the words?!

Can I just ask how free-and-clear discussion, creating binary outcomes, with determined outcomes is any different from say, working out how you attack a monster, rolling a d20 to see whether you hit and then doing your damage until it dies?

Does the free-and-clear include the 'new stuff coming out of left field narrative' that I associate with shared narrative?

Neil

Take the King's shilling at http://www.omnihedron.co.uk/dutyandhonour/

Whether you kill the monster doesn't actually tell you if you got what you wanted.

You might have wanted to impress your friends with your bravery. Or intimidate it into submission with your superior skills.

Free and clear gets us all on the same page about why we're even rolling the dice.

-Matt

Realms Publishing

Solipsist ended up with a halfway house (which I need to be more clear about in the next version of the text). You have a half-stake phase, where you say in what direction you intend to change reality, but then when you've done it you get to describe it actually happening, and throw in extra details that you hadn't outlined before (and other people get to contribute too, though you have veto), so long as they don't violate the guidelines you set up before the change.

I'm not 100% happy with that all the time though, maybe because I've not been explicit enough about it. There is a lot to be said for the simplicity of

* try to do something (e.g. punch someone)
* roll
* describe how you did it

That gives narrative freedom with constraints, but without the pause that can be involved in explicit stake setting (which can be really annoying)

http://www.solipsist-rpg.com/ - http://cubicle-7.com/starblazer/

To be clear, I think...

The "stakes" thing is all about the free-and-clear becoming a succession of escalations and fixing of outcomes that have a binary behaviour. The "narrator" then has no narration to do, we just accept that pre-narrated Stake A, B or C happens."

...is the "bad Stakes setting" that sprung up and everyone gnashed their teeth about and fought over.

The "free and clear" in Sorcerer (et al.) is just what any functional GM in a functional game will do. Asking "What are you doing?" When you reply, it's taken that you haven't actually done anything yet. We resolve it using dice, cards or numbers and then fictionally we see what happened.

For example, "I stab him" or "I make her love me". We know what you are trying to do, but you can't say you've done it or not, or what happened until the dice are rolled. You might make her love you but only by taking a horrible stab wound from her brother first. Or maybe she says she loves you then he stabs you. Or maybe he doesn't stab you and she doesn't love you. Or she loves you and you remain unstabbed.

Once the dice are rolled all the "actions/reactions" resolve in the fiction. Don't waste time doing it all before the dice are rolled. And don't give someone the right to monologue afterwards where it's unrelated to what the system dictated would happen.

I think Solipsist is OK, actually. It says: (i) you've reached a point where you have to change reality to get what you want -- the crucial thing is ... what do you want?, (ii) you bring in fictional elements from your Obsessions/Limitations as the resolution (no dice in Solipsist, but this is where the dice would be), (iii) someone narrates the outcome and the new reality based on failing, succeeding perfectly or succeeding too well. That final part of Solipsist cannot just be a binary outcome of some pre-narrated from point (i) of the process, since it now includes all the stuff from point (ii) and the metric of point (iii) (what flavour of success is it?).

Andy, what's a typical conflict, here? Could you give an example of meddling with the lives of mortals? And how do conflicts fit into the game as a whole?

I think the question will end up being whether stakes are right for this game, rather than about the role of stakes in general.

(I'll lay money they're not right for the game, because stakes are rubbish, but anyway).

Graham

Ok Graham, I'll I'll throw out a couple of conflicts that cropped up in the playtests so far and see where they stand.

A demon has a little girl by the throat. The angel declares: I want to kill the demon. The demon declares: I want to snap the girl's throat.

The angel tries to rally the spirits of a doubting angel by quoting scripture and verse at him. The angel declares: I want to rally him. The GM declares: he wants to be left alone with his misery.

Two angels are in a fight with a Fallen. The Fallen has a sword to one of their throats. The angel declares: I'm trying to break the sword. The Fallen declares: I'm trying to scar the angel's face.

So these are the sorts of conflicts the game deals with, at their simplest level. There are several nuances that I won't go into in detail, such as having to justify getting involved is a test on the part of the angel or part of the plan (otherwise you really are meddling, which is bad), and the ability to surrender or seize the right to narrate a conflict, even if you haven't won.

Well, how about we pick on one.

A demon has a little girl by the throat. The angel declares: I want to kill the demon. The demon declares: I want to snap the girl's throat.

How do you see that playing out between the players at the table?

Lead me through what you want to happen in the fiction. That might help. What sorts of outcomes with narrations do you wanty to have?

I remember reading some great Mountain Witch clarity from Tim Kleinert. Going by that this might not even be a conflict. The Demon hasn't opposed the Angel and the Angel isn't trying to save the girl. So, the question is in which order things happen. If the Demon dies first that saves the girl?

But that's probably not quite right here. Is the Angel actually trying to stop the Demon killing the girl? Is the Demon opposing being killed itself? Or is it toast however this goes down?

Anyway, in your fuzzy ideal. How does it play out? Who speaks and how are they constrained in the narrative as you talked about in the first post? Are they just adding colour? Or is ity much more than that?

Ok, let's give it a go. Here's how I envision it playing out in my fuzzy imaginary game.

Ok, so there's the GM & a player playing Afriel, the protector of children.

GM: You round the corner and at the end of a darkened alley is a teenage boy, track marks up his arms and a vacant look in his eyes.

Afriel: A demon...

GM: He turns as he sees you approach, and you can see he clutches a little girl in his arms, his claws digging into her skin. What do you do?

Afriel: Screw subtlety, I'm the god-damned protector of children! I flick out a blade from under my coat and advance menacingly on the demon, shouting passages and verses at him as I do so.

GM: He's not one to let you get that close, so he's going to hold the girl between you and him. Test?

Afriel: I'm going to test my belief in god, as I want to trust in the almighty and put my full faith behind my words long enough to make him recoil. Then I gut the fucker.

*both roll dice, with slight infernal interference (more 6s than 1s rolled)*

GM: Ok, you win. What happens? Remember the interference.

Afriel: Each word I speak is like a hammer blow to the demon, and he physically recoils from me, dropping the girl to one side long enough for me to get close and drive my blade into his belly, twisting it slightly so he screams before he dies. A slight sulphurous rasp escapes his lips as he does "you're too late."

GM: Nice touch. You can remove a token from the plan too - presumably an adversary token, considering that narration.

Afriel: Yup. It was one of my parables too, so I get to complete that and get some faith back.

The interference reminds me of Inspectres: that table which, depending on how well you roll, tells you how to narrate the success. As long as you've got that sort of guidance, I think it'll be fine.

Number of 6s minus number of 1s rolled:

Up to 3: Slight infernal interference. Narrate a general success, with a bit of shit going on.

Over 5: Major infernal interference. Narrate a success, which is overwhelmed by the huge amounts of shit going on.

Again, that sort of thing works fine in InSpectres, so I think it'd work here.

Stakes don't seem what you're looking for to do this.

Graham

Agreed, I didn't see any stakes in that, more you had a situation of conflict and the intent is decided in the narration after the roll is made

http://www.solipsist-rpg.com/ - http://cubicle-7.com/starblazer/

Well there are no stakes because I carefully wrote the example without any! But do you think, given the example, there'd be any problems with narration after the roll? Is it open to abuse?

Or, given enough guidance, will this actually work?

I'm 100% for narration after the roll, as long as the intent (not the stakes) before the roll were clear enough. It's how most games I play work, and it resists abuse 99% of the time (and the 1% is usually due to the intent not being clear enough, i.e. it's an honest mistake).

I think if you are worried about abuse you need to look at the scope of what the roll accomplishes. This sort of roll is the mainstay of most task resolution systems, and doesn't get abused because everyone knows the scope of each roll is usually small (i.e. yes, you can say you grab the rope with your agility test, but no, you can't then say you also tie up a whole squad of troopers). Its where one roll resolves a huge amount of action that you get issues, and start to think about stakes.

In Solipsist for example I require players to describe the sort of change they are going to try before they test to see if they do it, simply because the scope of what they can do with the change is too huge, I can't simply leave it all for post-test narration, because they could do anything at all. If your scope is smaller you shouldn't need that.

http://www.solipsist-rpg.com/ - http://cubicle-7.com/starblazer/

Intent is rarely clear. Ever.

That's why stakes, goals, scene types and so on are really useful. They get you to actual character intent, not an assumed task. Don't get hung up on how one game handles this, "Stakes" is Dogs' way of getting to intent.

The key is getting to why the hell we're going to the system to resolve this and what is it we're resolving.

Andrew's example contains explicit goals for both parties, "Prevent closing in" and "make him recoil". So we know what the conflict is about and why we're rolling. Is that statement of goals part of play, or is it an ad-hoc improvisation by the players?

-Matt

Realms Publishing

It's going to be part of the procedure. Second to that (as you can see above) is the player explaining exactly which belief he's using and how he's being tested. So he first explains his intent, then he explains how he's executing that. Then, after dice are rolled, the winner gets to describe what happens, in the framework already set out. I'd like it to be as clear cut as that.

But do you think, given the example, there'd be any problems with narration after the roll? Is it open to abuse?

It'd work fine, I think.

That's what I meant about InSpectres. In InSpectres, you've got a model of a game where letting people narrate their success, and putting conditions on how they narrate their success, works fine. If it works in InSpectres, I think it will work for you.

You could put lots of "Don't be a dick" rules in, if you wanted to. But I'm pretty sure you don't need them.

Graham

Right, Andy, very cool.

Afriel: I'm going to test my belief in god, as I want to trust in the almighty and put my full faith behind my words long enough to make him recoil. Then I gut the fucker.

GM: Ok, I win...

Oh, What happens in this case? And what input does the player have on it? (I guess this is hwen they can spend bennies to grab the win back or something?)

I'm just wondering how far you feel the GM can go before letting the player get another conflict roll, if you see what I mean?

Now that is the question - how far can the GM go?

The GM stated his own intent, but perhaps that was more of a reaction, stating that the demon wasn't going to let Afriel get close. So I haven't explicitly laid out what the demon intends to do, or what contraints he has.

As the angel lost the test, he also loses a point of faith as well as the plan token he was attempting to remove (the GM gets it instead). So, within the constraints of the narrative and the game, it's a considerable setback for the character. So I could narrate it as quite a bad loss, especially as there was infernal interference - I could certainly narrate the demon killing the girl, or lashing out and hurting the angel.

The player could mitigate the failure by spending a point of faith to seize the narration instead. He'd still have to narrate the loss, but wouldn't have to go quite as far - so the demon might escape with the girl, rather than kill her.

As for repeating a test - you can't, unless one or two elements of the test change first, and you use a different belief. So you could have another go with another belief and change the location, or change the method, or change the protagonists. So to get another crack at the test, the angel would have to give chase as the demon escapes (changing the location), employ another method (bargaining or threatening the demon) or even get another angel to attempt the same thing.

So, the consequence of the failed roll is taken care of by the mechanics laready, so the narration is just informed by that?

So the limit on going too far is that the mechanical consequences don't tie up with a "too-far narration" right?

So, maybe a list of example "weights" to successes and failures?

And the clairty above about following up one test with another one (one or two elements have to change from the original test).

That seems solid to me. And the player can always frame the roll? Or does the GM get to knock back something? Can the player just keep rolling on the same thing? Or do the outcomes differ from what you roll on, which encourages people to take on different rolls in different circumstances?

There are mechanical consequences for failure, yes, although they can be interpreted fairly broadly. Failing a test shakes an angel's faith, causing him to lose faith or increase doubt, which could be interpreted in a variety of ways. Failing a test also causes the GM to take the plan token instead of the angel, which is a setback to the angel's progress, and again could be interpreted in various ways. I'll make sure both failure and success have plenty of examples for each.

How do you mean by framing the roll? If you mean, gets to explain how he is being tested, then yes - if he can't justify the test, he has to meddle instead, which is where he increases his doubt (as he's going against the Plan). The GM shouldn't be knocking back tests willy-nilly, but equally should be prepared to make a stand and say "actually, you're grasping at straws there - take the hit or don't roll the dice."

Assuming that the player kept changing his belief, the location, the method and so on, he could keep rolling on the same test over and over again... up to a point. As each test is fought over a plan token, which are finite, eventually these will run out. If they run out and the GM has a big pile in front of him, then the players will have lost the Chapter. So there is a finite limit anyway, although I'm wondering about adding the additional stipulation that once you've made a test for the same goal with each of your (3) beliefs, that's it. No more. This Chapter at least.

>>There are mechanical consequences ... for each.<<

Sound perfect to me and the system for resolving rolls is done. The limit is the mechanical stuff and the plan token. All good. You just find something in the fiction to contest that the GM and Player have got a Plan Token to win out of it.

>>How do you mean ... or don't roll the dice."<<

I was meaning who picks what stats we're using. So, at the moment it's the player with the GM having veto on it if sounds bollocks. Right?

>>Assuming that the player ... This Chapter at least.<<

Yeah, I wonder if that would be an idea. The Plan Tokens are limited, so at some point the "pool" refreshes, but I'm wary of players always using the same roll and min-maxing to just roll one combo each and every time.

(How I got round this in 3:16 was by making the two stats have different outcomes and so there are times you will want to use one rather than the other. In other cases I flat out stipulated which one it was to be, e.g. Dominance just is a test of NFA, etc.)

It sounds to me you are more solid in your thinking. (Do Angels ever compete? Or is it strictly turn-based?)

What about we go through another example? And see if the new procedure works?

The angel tries to rally the spirits of a doubting angel by quoting scripture and verse at him. The angel declares: I want to rally him. The GM declares: he wants to be left alone with his misery.

The angel tries to rally the spirits of a doubting angel by quoting scripture and verse at him. The angel declares: I want to rally him. The GM declares: he wants to be left alone with his misery.

....

GM: Uriel sits in his high-backed chair. In front of him on the table is an envelope, ripped apart. On the floor around him are the burnt remains of a letter and 13 empty cans of Tennents Super Strength lager. Wisps of smoke pool in eddies in the ceiling.

Afriel: Go and grab the envelope. Who is it to?

GM: It's addressed to Uriel, and you see it's in Morgana's handwriting. Uriel doesn't even stir as you grab the envelope from the table in front of him. His eyelids are heavy and he is wallowing in his own misery.

Afriel: I grab him by the shoulders and try to rally him. I quote scripture and shake him out of his depression!

GM: That's a test. He wants to wallow in his own misery. What are you rolling?

Afriel: Blah and Blah.

GM: OK!

*both roll dice, GM wins more 6s than 1s*

...

I'm a little lost on this conversation, but it sounds like we're talking ourselves back to stakes (X happens or Y happens). Andy, I thought what you had before was fine and workable, and that the worry about "How far can people go in narration?" isn't actually that much of a worry. Must go and catch a flight.

Graham

Andy, is all that we actually need to know (before rolling) is what the PC wants to do. If it fails then something else will happen and that's informed by how the dice go and what mechanical changes happen (and that's based on what the player rolls, right?)

I'm not sure we are talking ourselves back to stakes. Gregor's example, as mine, simply has the player stating what they're doing (and, to a lesser extent, what they're trying to do).

Let me build on your example Gregor, taking things back a step.

GM: Ok, what do you guys do next?

Afriel: I want to get rid of a couple of those intervention tokens. I know an angel who might know something.

GM: I'll frame the scene, unless you have something else in mind?

GM: Uriel sits in his high-backed chair. In front of him on the table is an envelope, ripped apart. On the floor around him are the burnt remains of a letter and 13 empty cans of Tennents Super Strength lager. Wisps of smoke pool in eddies in the ceiling.

Afriel: Go and grab the envelope. Who is it to?

GM: It's addressed to Uriel, and you see it's in Morgana's handwriting. Uriel doesn't even stir as you grab the envelope from the table in front of him. His eyelids are heavy and he is wallowing in his own misery.

Afriel: I grab him by the shoulders and try to rally him so he'll talk. I quote scripture to try to shake him out of his depression!

GM: He wants to wallow in his own misery and has no interest in talking to you. That'll be a test.

Afriel: Ok, well I'm going to use my Belief in the Word, obviously. I'm going to let the power of the Word do the talking this time round.

*rolls dice, GM wins, more 6s than 1s*

GM: Ok, you lose a point of faith and that intervention token goes into my pile - your faith is shaken a little and you find yourself no nearer to solving the mystery than you were before. Unless you want to spend any faith, I'll narrate.

Afriel: No, you go for it. I'm nearly out here. In fact, I'll increase my doubt instead of losing any faith.

GM: Your words seem to ring hollow and untrue. Uriel looks up at you as you speak to him, and there's a somewhat malevolent twinkle in his eyes. "You see it too, don't you? Just how profoundly empty that scripture is. Once we're done here, there'll be nothing left you know, just the taste of ashes and tears." He offers you a cigarette and lights it with the tip of his finger.

See the difference there? There's a tiny bit more procedural set-up - ie the players decide which type of token they're going after in the scene (intervention, mystery or adversary); there's the stage where the player explains how he's being tested (not just what he's doing, but why and how); and then we get the consequences and the narration arising from failure.

I really like that, and it little post there is a nice bulleted list of how to do it. Great!

As an aside, those examples rock! Really make me want to play :)

http://www.solipsist-rpg.com/ - http://cubicle-7.com/starblazer/

That's great to hear David - both situations were taken from the first playtest game, so they're not entirely pie in the sky!