The Forge Archives

General Forge Forums => First Thoughts => Topic started by: MikeF on May 24, 2010, 08:23:31 AM

Title: Limiting GM fiat with a token system
Post by: MikeF on May 24, 2010, 08:23:31 AM
I've been thinking a lot recently about GM fiat, and ways of limiting GM fiat without affecting their ability to introduce interesting challenges for players.

These were the simple rules I came up with:

#1: The GM starts and ends scenes.
#2: The GM is responsible for describing the actions and reactions of all NPCs and setting elements.
#3: The PC is responsible for describing the actions and reactions of their PCs.
#4: The GM can introduce new NPCs and setting elements at any point in a scene.
#5: All PC actions succeed unless the GM declares a Challenge.
#6: To declare something a Challenge the GM has to spend a number of Challenge Tokens [a finite resource]
#7: The Player has to roll higher than the number of Tokens [detailed dice mechanic still to be fleshed out]
#8: On a successful roll the Challenge is overcome and the PC's actions occur as described.
#9: If the roll fails the Player can spend Stunts [Character special moves / abilities  / gear /etc] for a reroll
#10: If the roll still fails the GM describes how the action failed. They can impose Consequences on the PC [taken from a list, each of which has specific effects on dice rolls in future Challenges].

The key part for me is #5: All PC actions succeed unless the GM declares a Challenge. "Say yes or spend a token", essentially. I think this opens up the game for the GM to be as inventive as he likes, without either overpowering the players or forcing them to act in particular ways. I'm hoping that the way the players respond to any incidents / NPCs / story elements will allow the GM to only invest Challenge Tokens in those conflicts that the players seem most interested in / respond to well. Does anyone agree, or see any major traps or flaws in this?

Michael.
Title: Re: Limiting GM fiat with a token system
Post by: Vulpinoid on May 24, 2010, 08:45:50 AM
I'm toying with something like this at the moment with a project I'm working on...

QuoteGM sets a difficulty for their session, this is defined by a number of secrets.

The total number of secrets is defined through a combination of the number of players, the time limit for the session and the complexity of the story.

Easy Sessions: 3 secrets per player
Moderate Sessions: 5 secrets per player
Hard Sessions: 7 secrets per player

Short Session (up to an hour or so): No extra secrets
Typical Session (from one to two hours): +5 to +10 secrets
Long Session (two to three hours): +15 to +20 secrets
Marathon Session (longer than three hours): +25 secrets
It is recommended that longer or more complicated stories be divided over two or more sessions.

Obstacles
These cost the GM a single secrets to introduce into the story. They force a player to make a roll before they are able to progress in the story. Obstacles could be enemies, traps, or simply changes in circumstance. Anything that needs to be immediately addressed and could cause issues for one or more characters counts as an obstacle. By spending an additional secret, an obstacle gains a positive trait. Perhaps an enemy is now "Armed", or maybe they are "Protected" by a bodyguard; perhaps a trap is "Hidden" and needs to be detected before it can be avoided. The trait provides an extra die for the obstacle, and in many cases it must be eliminated before the obstacle may be eliminated.

Twists
Any time a player wants to pursue their own goal, the GM may spend a secret to add an extra die to their hand, and a potential degree of sacrifice. The GM must describe the complication to the situation and what negative trait might arise as a result of this. 

The whole idea is that the GM spends secrets to throw obstacles in the way of the characters, or spends secrets to increase the difficulty of obstacles faced.

If this is confusing (because it references game elements that aren't explained)...I'll post up the whole thing shortly.

V
Title: Re: Limiting GM fiat with a token system
Post by: MikeF on May 24, 2010, 10:07:31 AM
Hi V,

Yes, that looks very similar. My system is, I think, less rigid than yours in terms of what 'Challenge Tokens' represent - they're an abstract difficulty rather than representing things like whether or not an enemy is 'armed' or is 'protected' - but essentially we're doing the same thing from the GM's side.

Does this mean that your rules - either explicitly or implicitly - take the same approach as mine for the PCs, in that a PC action is guaranteed to succeed unless it's actively opposed by the GM (using 'Tokens' in my game, 'Secrets' in yours)? Or are there situations where the Player still has to make a roll, even if the GM hasn't spent anything? Are there things the PC can't do when there are no Obstacles in play? (These are the bits I'm wrestling with at the moment).

Michael.
Title: Re: Limiting GM fiat with a token system
Post by: PeterBB on May 24, 2010, 04:28:19 PM
What stops a player from saying: "I hack into the US Government's laser satellite system and fry the arch-villain," forcing the GM to spend a lot of tokens to stop this obviously ludicrous action, and making the rest of the challenges necessarily easier? Sometimes a player wants to do something that has little chance of success, and rewarding that behavior with easier future challenges seems counter-intuitive.
Title: Re: Limiting GM fiat with a token system
Post by: MikeF on May 24, 2010, 05:54:01 PM
Peter,
QuoteWhat stops a player from saying: "I hack into the US Government's laser satellite system and fry the arch-villain," forcing the GM to spend a lot of tokens to stop this obviously ludicrous action, and making the rest of the challenges necessarily easier? Sometimes a player wants to do something that has little chance of success, and rewarding that behavior with easier future challenges seems counter-intuitive.

Yes, absolutely the million dollar question. I think the only answer is that there have to be some additional constraints on what a PC can do.My main idea for how to do this is through the use of goals.

Goals:
# Each PC has a specific goal.
# They can only achieve this goal after progressing through a set number of Steps.
# Each scene counts as a Step towards the goal.
So you could have it that the PC's goal is to defeat the arch-villain. They can meet and wrestle with him repeatedly throughout the game, but they can only defeat him (by killing him, say, or by overthrowing his criminal empire, or by unmasking him) after a certain number of scenes.
That means that a PC could, if he chooses, decide right at the start of the game to hack into the US Government computer network, and pinpoint exactly where the arch-villain lives - but he couldn't actually achieve his goal straight away. The Player would have to allow the GM to introduce other challenges / incidents / obstacles before he got the chance for the big denouement scene.

I think there are also a couple of additional things which would help to prevent the sort of situation you describe:

Playing in character:
# Reward Players for having their PCs act in character.
I like the idea of character 'keys'. Have them regenerate the use of Stunts, or gain dice, or win points every time they act in character. Have them lose dice every time they act out of character.

Shared group vision:
# Ensure the players have a shared vision of the type of game it is - the scale of PC's powers and the aesthetic of the game world. Any two players (or one player and the GM) think that something is impossible or too implausible they can veto it.
Title: Re: Limiting GM fiat with a token system
Post by: Callan S. on May 24, 2010, 07:40:37 PM
Hi Mike,

Nice to see discussion explicitly at the 'The GM decides' level rather than in 'it's all determined by the characters environment' missinformation level.

Quote#10: If the roll still fails the GM describes how the action failed. They can impose Consequences on the PC [taken from a list, each of which has specific effects on dice rolls in future Challenges].
Taking it those effects on latter die rolls will be penalties, this would form a death spiral system. Eventually you'll fail, and this sets off a spiral toward more failure, which sets off more, etc.

Not that that couldn't be a useful. It certainly means the game, by and large, has an ending.

Quote# They can only achieve this goal after progressing through a set number of Steps.
It might be interesting if there is some option to roll for a chance (like a 5% chance) of skipping X number of steps required. So sometimes the ending occurs at more of a surprise moment, rather than "oh, were at 8 steps now...okay, do the denouement now" which is a bit cut and dried.

Beyond that, I think there's a bit of a gamer culture out there that's more interested in that the imagined space thingie has some sort of integrity against being violated by 'ludicrous' things. See here (http://www.indie-rpgs.com/forum/index.php?topic=17334.msg188019#msg188019), and a quote from it:
QuoteA great deal of the aesthetic power of Simulationist play, as I see it (and I mean that literally), lies in (a) adding to or developing that package, and (b) enjoying its resiliency against potential violation.
...
Always remember the (b)! Without it, (a) is merely the chassis for any Creative Agenda.

Alot of people will treat B as being the foundation of all roleplay, ever - rather than B being just one type of roleplay. But they'll advise you B is vital otherwise it's not roleplay at all and a disaster.

But feel free as a designer to not have B at all, is all I can say. Alot of people will push and shove the idea you must have B, but I say feel free to make your own choice on that.
Title: Re: Limiting GM fiat with a token system
Post by: Vulpinoid on May 24, 2010, 07:55:18 PM
I'll give my perspective on the question.

But first some counter questions...

Why would a player say that they wanted to "hack into the US Government's laser satellite system and fry the arch-villain"? Does this make sense in the context of the story? Does it seem realistic within the setting? Does it make a good game?

I'm a big fan of the idea "Say Yes or Roll the Dice", I've been running my games this way since the 1990s (long before the phrase was coined if I understand correctly, and certainly before it gained popularity in Indie Gaming crowds). But if you open up an opportunity like this to a character, then you're already making some conscious decisions about where the game is going.

Designing a game mechanism to handle a situation like this really needs to address what the game is about...you'll hear that a lot around here.

The game design I'm working on (the one I referenced earlier in the thread) works with Otherkind Dice (http://www.lumpley.com/archive/148.html) (in a system similar to John Harper's Ghost/Echo (http://www.onesevendesign.com/ghostecho/)), but with some stricter framework in place for story development.

Any time a player wants their character to do something that makes sense for their character to easily do, it just happens...any time they want to take a risk and do something special (either gaining an advantage or confronting the story) they roll otherkind dice...stating an objective and picking a stake relative to the situation at hand. If they want to gain a bonus trait for use later in the game, they roll the dice, if they want to buy off a penalty trait (or remove a penalty from someone else), they roll the dice. There's always a chance of something good happening, but there's always a chance that something bad might happen too.

If it's important to the storyline, that's where the GM's secrets come into play, and the GM may up the difficulty according to the tokens they spend against the character.

I've incorporated into the system something similar to the "steps" mentioned. Where my system basically states that once the GM has spent a certain number of secrets, a story objective is revealed. This story objective may be a special item useful at the story's finale, it might be a new ally, or possibly a new piece of knowledge. Character don't need to successfully obtain these objectives along the way, but possessing them will certainly make things easier at the showdown.

I believe there should be multiple ways to achieve any scenario, and the twists and turns along the way can be just as important as the climax.  
Title: Re: Limiting GM fiat with a token system
Post by: greyorm on May 24, 2010, 08:40:55 PM
I went the same route when designing ORX, regarding giving the GM a limit on their input (though I handled it slightly differently, though I like your idea on just "saying yes" better), and came up against the same question about how to stop the players from rolling one die and getting everything they want. I approached it by talking about conflicts, scales of action, and setting stakes in the form of announcing short-term goals that needed to be overcome.

I don't think the solution is really mechanical. If a player sits down at the start of the game and announces something as broad as "I hack into the US Government's laser satellite system and fry the arch-villain", then all you need to ask is, "Ok, how?" And build smaller conflicts leading up to the goal from there.
Title: Re: Limiting GM fiat with a token system
Post by: Callan S. on May 25, 2010, 04:51:49 PM
Well, that is mechanical - as in a mechanical procedure can be extracted from the behaviour
1. The GM has the option of deciding the stated action is 'too big', which means the action is invalid for resolution. The GM instead asks for smaller actions. Repeat this rule until an action the GM deems small enough is found for resolution.

The identification of mechanics like this is what rpg theory needs more of.

Also saying "Ok, how?" would actually be breaking the rule of 'say yes or roll' as it's neither saying yes or rolling.
Title: Re: Limiting GM fiat with a token system
Post by: StevenS on May 25, 2010, 08:06:46 PM
I like the "Say yes or roll the dice" motto, but I am not using it for the game I'm in the process of designing (Free Your Mind, the RPG of Psychedelic Conspiracy) because I want a bit more GM-control; instead, I'm using something more like this:

If it's minor, just say yes.
If it's not-so-minor, look at the appropriate character stat, if there is one, and judge on that basis -- Flow for the sort of assertions that go "There's an X there, and I grab it", because a high flow means someone's well in tune with the Universe. Coherence for "I ask a person to go along with X" because high-Coherence people are the sorts owho can sweep others along in their wake. Karma as a modifier for either of those.
If it's really big, or there's a clear opposition involved, or the player *insists*, even if their Flow/Coherence/Karma says no -- then go for the dice. ;)

As to the problem of "I make up a ludicrous challenge to burn off GM tokens so that later challenges are easier" -- that is, I think, a self-correcting problem, with a GM and players who are aware of it.

First, IIRC, the original proposal mentioned that a failed result would involve consequences -- and if those consequences are dire enough, well, people will stop trying to pull that same stunt. One obvious "consequence" would be a restoral of a certain number of tokens, so that the effort, if failed, would not have anything like the intended "clear the GM's ability to challenge us in one fell swoop" effect.

Secondly, and this sis omething I bring in again in the WiP -- the players can often be the best restraint upon another player's behavior. Appealing to them about the rationality of an action is likely to bring about a better result (at a guess) then simply declaring it "out" -- because the other players also have a stake in getting to do stuff.

Now, this obviously will not work with players whose whole stake is "winning", because they won't care which of them does -- but I don't think the design mechanism you're putting in place is really all that appropriate for a strongly Gamist group; it seems too much designed to reduce GM-impact, rather than making them a real and proper opponent.



As always, my $0.02 is worth what you paid for it. ;)
Title: Re: Limiting GM fiat with a token system
Post by: PeterBB on May 25, 2010, 09:05:00 PM
I proposed the "hacking government lasers" example as an extreme case, but I think it illustrates a general problem. Under this system, future conflicts are harder if the early ones are easier. In practice, this means that you are penalized for coming up with creative ways to make things easier, and rewarded for doing things in an unnecessarily difficult way. I don't think you can get around that without altering it so much that it's a different system.
Title: Re: Limiting GM fiat with a token system
Post by: MikeF on May 26, 2010, 04:16:21 AM
Hi Peter
QuoteI proposed the "hacking government lasers" example as an extreme case, but I think it illustrates a general problem. Under this system, future conflicts are harder if the early ones are easier. In practice, this means that you are penalized for coming up with creative ways to make things easier, and rewarded for doing things in an unnecessarily difficult way. I don't think you can get around that without altering it so much that it's a different system.

Maybe I'm misunderstanding your point, but I don't see it this way. Why does the mechanical difficulty of a challenge mean that overcoming it is more or less creative? The system doesn't automatically penalise - or reward - you for coming up with creative solutions to problems. That's up to the GM, who can arbitrarily make your chances of succeeding more or less likely using tokens. You could have your PC take the most circuitously "difficult" approach to a problem within the fiction and still have the GM make it "easy" by assigning a low number of tokens. So how you approach the problem and whether or not you succeed seem to me to be entirely independent of each other, mechanically speaking.

That's not to say the GM should be acting entirely independently of the player's creativity - I would hope that they would use their tokens wisely to pace the story, and invest the most tokens in those bits of the story that the players seem most interested in and respond to most. But fundamentally that's a GM technique, it's not forced by the mechanics (though I guess you could tweak the rules so that GM's received their tokens in installments, to force the pacing a little: so the GM gets, say, 20 tokens to use over the first three scenes, 20 tokens for the next three, and then fifty tokens to use in the climactic endgame).
Title: Re: Limiting GM fiat with a token system
Post by: MikeF on May 26, 2010, 05:03:31 AM
Steven,

QuoteIf it's minor, just say yes. If it's not-so-minor, look at the appropriate character stat, if there is one, and judge on that basis -- Flow for the sort of assertions that go "There's an X there, and I grab it", because a high flow means someone's well in tune with the Universe. Coherence for "I ask a person to go along with X" because high-Coherence people are the sorts owho can sweep others along in their wake. Karma as a modifier for either of those. If it's really big, or there's a clear opposition involved, or the player *insists*, even if their Flow/Coherence/Karma says no -- then go for the dice. ;)

I think that's fine as far as it goes, but the problem I have with this type of system is that I think you're still potentially leaving wiggle-room for GM fiat - because as GM I can, without cost, force the PCs to react in a particular way that requires them to roll.

So say the PCs are trying to escape Pirate Island, and as GM I don't want them to - I have a brilliant idea for a plot twist that needs them to still be on the Island. So the PCs get in a ship, and I have it attacked by a giant squid. The players have to roll, but they manage to beat the squid, so I have a second squid attack them (bigger than the last one). The players roll and overcome that one too (they're lucky players) so now I have a storm drive the ship back to shore. And if they beat that then I'll upgrade it to a typhoon, and then a hurricane, and then I'll have the Kraken awake, or the ghost ships of Atlantis rise up from the deep and force those players to go back to Pirate Island. And if all of that still fails I'll have their bloody compass break and they can sail round in a circle. As GM I can keep on escalating until the Players are forced to go to the place I want them to go.

It's an extreme and perhaps absurd example, but I think it reflect the fact that GMs railroad PCs by simply forcing them into an increasingly narrow set of options. What I'm *trying* to come up with is a system that allows the GM to still do all of the above if he wants, but he will be constrained by having to invest some of his finite resources in order to do it.
Title: Re: Limiting GM fiat with a token system
Post by: Callan S. on May 26, 2010, 07:46:37 PM
QuoteSo say the PCs are trying to escape Pirate Island, and as GM I don't want them to - I have a brilliant idea for a plot twist that needs them to still be on the Island. So the PCs get in a ship, and I have it attacked by a giant squid. The players have to roll, but they manage to beat the squid, so I have a second squid attack them
"But a good GM wouldn't do that!/Only a dick GM would do that!"
Sorry, just my warped sense of humour, parroting the all too common behaviour of sanctioning someone for actually playing the system while also giving up on mechanical system design. No, I'm not saying that, just joking!
QuoteIt's an extreme and perhaps absurd example, but I think it reflect the fact that GMs railroad PCs by simply forcing them into an increasingly narrow set of options. What I'm *trying* to come up with is a system that allows the GM to still do all of the above if he wants, but he will be constrained by having to invest some of his finite resources in order to do it.
If I understand you, I agree that if a GM has unlimited resources, then whatever you do is at his whim. Even if he doesn't railroad you and you go the way you want, it's because at a whim he decided not to deploy his unlimited resources. In a way your only ever doing what he wants when he has unlimited resources/your always railroaded into what he wants (I think that's why people grope at 'oh, he'd be a dick to do that' as some sort of social currency, but it involves actually calling people names/threatening to call names as part of regular gameplay...*brrrr*).

I think finite resources will make a significant difference on the matter. Just writing a supportive post on the matter! Hope you don't mind!
Title: Re: Limiting GM fiat with a token system
Post by: PeterBB on May 27, 2010, 03:36:35 AM
Quote from: MikeF on May 26, 2010, 04:16:21 AM

Maybe I'm misunderstanding your point, but I don't see it this way. Why does the mechanical difficulty of a challenge mean that overcoming it is more or less creative? The system doesn't automatically penalise - or reward - you for coming up with creative solutions to problems. That's up to the GM, who can arbitrarily make your chances of succeeding more or less likely using tokens. You could have your PC take the most circuitously "difficult" approach to a problem within the fiction and still have the GM make it "easy" by assigning a low number of tokens. So how you approach the problem and whether or not you succeed seem to me to be entirely independent of each other, mechanically speaking.

That's not to say the GM should be acting entirely independently of the player's creativity - I would hope that they would use their tokens wisely to pace the story, and invest the most tokens in those bits of the story that the players seem most interested in and respond to most. But fundamentally that's a GM technique, it's not forced by the mechanics (though I guess you could tweak the rules so that GM's received their tokens in installments, to force the pacing a little: so the GM gets, say, 20 tokens to use over the first three scenes, 20 tokens for the next three, and then fifty tokens to use in the climactic endgame).

Hmm, maybe I see where the disconnect is. I was assuming that "obstacle level" was supposed to fairly closely map "perceived difficulty level". Is that not true?
Title: Re: Limiting GM fiat with a token system
Post by: Eero Tuovinen on May 27, 2010, 05:18:47 AM
Fastlane and Primetime Adventures, alongside the aforementioned Orx, implement systems like this. Universalis, too, in fact. The way out of the mechanical interaction discussed here in all cases is to make the GM's pool of resources dynamic, which breaks the conflict pattern of easy-followed-by-difficult and vice versa. Rather, in Fastlane the GM is free within certain flexible constraints to choose which conflicts will be easy and which will be hard, and he may expect to have a constant flow of resources that'll allow him to continue doing so indefinitely. The overall resource level might mean that "easy" at some time is 1 chip's worth while at other time it's 3 chips, but that's what makes it interesting, really.

Of course it's still possible to game the system as a player by introducing frivolous conflicts, which is why all of these games in actual fact are based on the more fundamental cooperative interaction: the group has a common creative agenda, which is fulfilled by appreciating the joint fiction. This basic requirement means that those games do not have to worry about players constantly declaring frivolous conflicts as a way to leach the GM of resources, as doing so would be counterproductive for the joint goals of play. Individual game texts will differ on whether the management of the common creative agenda should happen by informal concensus or whether the GM should have some sort of veto to simply block the occasional brainfart, but the common theme in each case is that inappropriate conflicts (for the group's fictional sensibilities) are not taken through the resource-consuming conflict system because group concensus filters them first and only allows the good stuff to get mechanical impact. This is not so different from any other conflict resolution system, generally all roleplaying games rely on the players wanting to uphold the integrity of the fiction, even if some of those games will achieve this by appointing one player as the chairman.

Anyway, it seems to me that checking out one or more of the above titles might be educational in this line of design inquiry. Those are all quite interesting games, I think.
Title: Re: Limiting GM fiat with a token system
Post by: MikeF on May 27, 2010, 05:13:32 PM
Callan,
QuoteIf I understand you, I agree that if a GM has unlimited resources, then whatever you do is at his whim.
Thanks, yes - what you said!

Peter,
QuoteHmm, maybe I see where the disconnect is. I was assuming that "obstacle level" was supposed to fairly closely map "perceived difficulty level". Is that not true?
That's it. The 'obstacle level' is entirely at the player level, not a reflection of some kind of objective difficulty for the characters within the game world.

Eero,
QuoteFastlane and Primetime Adventures, alongside the aforementioned Orx, implement systems like this. Universalis, too, in fact.
Thanks. PTA was a direct inspiration, as was the 'Token Effort' system of In Spaaaaace... I'll try and read around Fastlane and Orx too - useful tips!
Title: Re: Limiting GM fiat with a token system
Post by: PeterBB on May 29, 2010, 03:53:03 AM
Cool, makes sense. I keep forgetting how different my preferred CA is from a lot of people here. So, instead of criticizing I'm now just curious. How do you imagine a GM deciding difficulty levels?: