In Target Numbers, or the last bastion of GM fiat and the earlier Difficulty/Success Needed as it relates to GM Fiat, there has been some mismatch in reactions and focus between people. I participated in the thread started by David Bapst; no one really came to grips with his topic. Instead of joining the new thread with another wandering viewpoint, I'm starting this one to discuss a specific angle.

I see the question of difficulty as tied into why we are using a resolution mechanic. Unless we understand and accept both the reason for using dice checks and the motive for setting numbers, we will find the system arbitrary and unrewarding.

I suggest that frustration with the details of task resolution might be because the task itself doesn't need resolving or the wrong factors are being used. Unless physical strength has some relevance, setting the strength of a character will be arbitrary. If it is worth recording, it should be possible to find out what its role in the game is. If you record meaningless traits and pointlessly resolve tasks you don't care about, any distribution of those parts of the system will still be unsatisfying. When the traits and difficulties help the game achieve your goals, the system shouldn't be painful.

Mixing these issues with who does what, fairness, and so on makes it unclear if my perception of the issues matches what David or Mikael is trying to address. If you know what needs to be done, and why, who should do it ought to be fairly easy to figure.

John

The way I see it, you have target numbers, dice resolution or some other skill-modified random outcome generation process for cases where:

1) skill, ability and conditions do not immediately and obviously generate automatic success or failure.

and

2) player ability makes a difference in the chance of a favorable outcome, and most importantly, one player can make more of a difference than another because of some ability, trait or skill.

and

3) the outcome is important to the continuing narrative, ranging from "how did the audience with the queen go?" to "can I avoid getting mauled by the brain-eating zombies?".

For instance as "no target number/roll needed", 1) is the "I cut his throat while he is unconscious" rule. You don't need to roll, skill is not really important. It's like walking and chewing gum. 2) is "do I get chosen for a random search at the airport?". Unless I am overtly suspicious, it is a random chance that is not influenced by my skills or attributes. There may be a "target number" and a roll, but it applies equally to all characters. The GM could just as easily decide it does or doesn't happen as part of the narrative. And 3) is  "what's the quality of the soup at Chez Swill tonight?" Unless the stuff is toxic, it is just a background detail that the GM can make up and doesn't need a "random restaurant food quality table" for.

For the best use of target numbers and dice rolls, all three of the requirements should be met. That is, it is important(3), the player can make a difference(2), and the outcome is up in the air(1). There is more dramatic tension in a die roll than in simply having the best qualified player do the action and assuming the GM will make it all work out in the end. The dice are fickle, and so is life. Sometimes you don't save the princess from the dragon, or Darth Vader does lop your arm off or Jabba does freeze your pal in carbonite.

As to how you set/determine the target numbers...that's a matter of what you want for "system feel". However you do it, it needs to generate results that are consistent with the nature of the gameworld and the "feel" GM's and players expect in that particular world or mileu (which is not necessarily a "realistic" result). "Film noir detective rpg" may have a completely different way of doing things than "over the top giant mecha rpg", even though they both have dice and target numbers.

My 2 cents worth,

Greg Porter
BTRC

This seems relevant, but it's possible that I'm missing your point in some fairly profound way, so apologies if this is off topic…

It's never seemed to me that changing target numbers or subtracting dice or increasing the DC or what-have-you has much of an effect at all.

This isn't to say that these things don't change the nature of the SiS in some subtle probabalistic way. Rather, they don't seem to change my (or anyone else's) gaming experience in any meaningful sense. It's no more difficult for me to add a smaller integer to the roll off a d20 than it is for me to add a larger integer; it's no harder for me to roll two d10 rather than twenty. More importantly, the fact that I'm rolling fewer dice doesn't make things more difficult for my character—the GM does that, out of game, in the often-magical, sometimes-random, and frequently-silent process of figuring out what TN or modifier is appropriate. Traditionally, the relevant features of the world (or my character's opposition, in the general case) are only narrated to the degree the GM feels like narrating them, be it in describing what my character overcomes in success, or what overcomes her in failure. Obviously, various systems switch up who narrates what a bit, but even against a lot of the most hippy Narrativist systems, my basic problem still stands: If you say something is more difficult, then shouldn't it, y'know, be more difficult in some way? Not merely less likely to go in your character's favor, but actually consuming of more narrative time and effort.

(And, not shockingly, that's how Logos plays. But that is off-topic, so I shall refrain.)

I do sometimes wonder if I'm missing something here. Maybe someone can enlighten me.

I can only speak for myself in this, and then all I can say is that the conflicts/tasks being resolved appeared to be important; erroneous measurements of character effectiveness and meaningless tasks are easily done away with (mostly by playing games that deal with shit that doesn't matter).

To some degree you are right, however... I've been an Illusionist/Participationist for a long time. One some level, the difficulties stopped mattering because they never really impacted play in the first place. It didn't matter if the gnome failed his Stealth check now or later, it didn't matter if the Water Aspect failed to steal the gem or not. My story was still going to cause the things that needed to happen to take place in that scene.

From such a perspective, I am trying to realize a method of deciding difficulties fairly. In Gamism, this is important but easily solved (see Rune), it never really matters in Simulationism, and in Narr, so far I can only look to PTA and a few other RPGs as forcing me to be fair.

John,

I don't have much more to say than, "I agree".

*****

Black Iris Dancer,

Many systems do make something that is more difficult require more narration time - it just isn't obvious how.  Anything that operates as an extended contest, such as depleting your opponent's hit points in D&D, is an example.

Black wrote: If you say something is more difficult, then shouldn't it, y'know, be more difficult in some way? Not merely less likely to go in your character's favor, but actually consuming of more narrative time and effort.

John wrote: I see the question of difficulty as tied into why we are using a resolution mechanic. Unless we understand and accept both the reason for using dice checks and the motive for setting numbers, we will find the system arbitrary and unrewarding.

I suggest that frustration with the details of task resolution might be because the task itself doesn't need resolving or the wrong factors are being used. Unless physical strength has some relevance, setting the strength of a character will be arbitrary. If it is worth recording, it should be possible to find out what its role in the game is. If you record meaningless traits and pointlessly resolve tasks you don't care about, any distribution of those parts of the system will still be unsatisfying. When the traits and difficulties help the game achieve your goals, the system shouldn't be painful.

Mixing these issues with who does what, fairness, and so on makes it unclear if my perception of the issues matches what David or Mikael is trying to address. If you know what needs to be done, and why, who should do it ought to be fairly easy to figure.

Elliot,

Excellent recap. I felt that the discussion of target numbers would be better served by taking a step back to see a bigger picture as you describe.

John

It's important to note that setting the target number is not about the difficulty of the task -- it's about the possibility of success or failure.  The 'difficulty of the task' is a consideration of the fictional elements in the SIS -- the in-game character's in-game chances based on in-game circumstances and his in-game abilities.  The GM may use the target number in order to communicate the difficulty of the task as she sees it, or she may use the target number as a means to control what happens in the game.  The target number can be an act of communication, or it can be an act of control.

I'm going out on a limb here, but I'm going to say that, ideally, the target number is used as both communication and control.

The communication part is essential -- because the GM is a participant in the game just as much as the other players, she needs to be able to communicate her vision of the SIS.  This is something that all participants in the game do constantly; some games just apportion different means of communication to different players (the GM sets difficulty, the players can determine which stats they are using, for instance).  Setting the target number is communicating about the SIS, whether it is about the concrete elements in the SIS that contextually make the task difficult or simple, or whether it is a genre trope that makes such tasks likely or unlikely.  When the GM does this, she helps reconcile her understanding with the other players' understandings.

The control part has historically been overpowered, and is presently being toned down a lot.  That the GM has power in the game is almost a taboo to be stricken down wherever it occurs.  This is silly.  The GM has power to determine what happens in the game just as the other players (should) have power to control outcomes in the game.  Where control is absolute (no, you can't try that, it won't possibly work) there can be issues, but in the case of setting target numbers, the control is only partial.  The GM can make it highly unlikely, but the GM cannot make it impossible (in most systems).

Now, the game can also give players elements of control that allow them to challenge or modify the GM's input.  A easy example would be most World of Darkness games, where spending quintessence/glamour/willpower/snot/whatever can lower the target number or increase the dice being rolled against against the target number.  This is also partial control -- the players can make it more likely, but rarely can they make it a sure thing.  Ideally, the acts of player control can also double as acts of communication about the player's understanding of the SIS -- that is, because Task X is so hard, their character puts in extra effort, or the character can do better than normal because Task X must be completed so they can save their girlfriend, or whatever.  As with setting the target number, this can be concrete in-game things (spending magic power) or can be founded on genre tropes (Drama Dice in 7th Sea).  Whatever 'it' is, the important thing is that it communicates and adds to the group's understanding of the SIS.  This parallels the GM's dual acts of control and communication when setting the target number in the first place.

Railroading and Illusionism and other curse-words come into play when the GM is not both controlling and communicating.  If the target number is set with regards to forcing events one way or the other, and no rationale is ever explained, then the GM has not communicated anything, she's just exerted control.  There is no reconciliation of what everyone is imagining -- while the other players are thinking, 'that door must have been really stuck,' the GM is thinking 'whew, that was close'.  When it starts becoming obvious that the GM is not actually communicating, this quickly turns into 'arbitrary and unrewarding' because, quite suddenly, the players haven't been sharing anything with the GM; they have in fact been deceived by the GM that there was something common that they shared.

John wrote: I see the question of difficulty as tied into why we are using a resolution mechanic. Unless we understand and accept both the reason for using dice checks and the motive for setting numbers, we will find the system arbitrary and unrewarding.

Hi, Paganini.

Actually, I think your comments sort of miss the point the same way that Greg's sort of do. IMHO the real point is that resolution mechanics can "be" any of several things which are quite different and thus serve quite different priorities. I wouldn't set my RPing experiences up as typical or exhaustive, and I'm hesitant to even frame my experiences and tastes within GNS. That said, if you take board gaming as analogous to Gamism, I'd say that "resolution mechanics as selectors" is too general to be useful for that mode. The issue is what *kind* of selector? *Does* it represent something in the SIS? And/or does it provide an opportunity to make strategic choices? And/or does it act as a sort of spur to creativity? (The last might be comparing to throwing coins for the I Ching, or creative methods such as cutting up newspaper articles and drawing words at random to create poetry.)

(Oops. Gotta truncate this but hopefully I've made my point. If it's not obvious the first of the three options would represent, in technique if not in terms of CA, simulationism; the second gamism; the third narrativism.)

Hey Ewilen,

You're right, "resolution mechanics are selectors" isn't really enough information to tell us anything about specific design. I just think that (A) realizing that resolution mechanics are indeed selectors is a prerequisite for really effective design and that (B) not very many people actually think about resolution mechanics in this way.

Also, I think that specific resolution mechanics have a much greater effect on the actual play experience than just bumping you towards a particular CA.

Paganini --

Are resolution mechanics really selectors?  I mean, are they really, actually deterministic, or are they more inspirational?  How many games tell the GM to fudge die rolls 'when necessary' (the definition of when it is necessary being highly variable)?  How many games really dictate what a die roll means and does not leave the bulk of the articulation of the system's result in the hands of the GM and other players?

Example: I want to jump over the canyon.  I roll dice, I get a 'fail' result.  Does that mean I fall to my doom?  Does that mean I hesitate and don't jump?  If I hit the ground, do I die, or can I hit the water and survive, or clutch onto some branches?  Can I roll dice again to grab those branches, or is my character lost?

The essential element -- and the deterministic element -- is the player articulation of what the die result means, not the actual die roll.  To my mind, the die roll doesn't actually determine or select anything -- it just inspires the players to articulate what the results mean to their SiS and how it validates their efforts (or they reject the inspiration, throw it out, and reroll).  To that end, John is right -- when the players don't need any inspiration for what happens next, when they know what is going to happen or they know what 'should' happen based on genre or story conventions, they don't need to roll dice.

(Uh oh, I'm cross posting with Joshua. This was started before I looked at his post.)

Hi, Paganini,

Well, we may be having a bit of a semantic disconnect here. To me, calling them "selectors" is either too general as I alluded above, or if we decline to go beyond the generalization, I get a sense of arbitrariness--meaning they just act as a sort of magic eight ball to tell us what happens next based on the game-state (which may include certain declarations of what someone is trying to insert into the SIS). At that very moment, the resolution mechanic is certainly a selector. But at the approach to the mechanic, when someone is thinking about doing something which could invoke the mechanic, or someone is deciding whether the mechanic ought to be invoked, and how, the mechanic can be more, and generally is.

Anyway, I'd like to add that when the resolution mechanic itself neither represents something in the SIS, nor provides an opportunity to make strategic choices, nor acts as a spur to creativity, it's at best an illusionistic technique. In highly illusionistic play, where the GM successfully hides the instances where he uses Force techniques from the players, the simple invocation of a mechanic *could* have these effects on an emotional or psychological level. But if the players are aware of a precise instance where the GM and/or the System is using resolution arbitrarily, the resolution is highly unlikely to do any of those things.

Joshua,

I think sometimes they're determinative ("deterministic" confuses me when we're talking about something that often involves rolling dice). Sometimes they aren't. It gets into how Stakes are determined in the Intent/Initiation/Execution/Effect part of the System. Sometimes the Stakes come first, sometimes they come last--as with GM fudging or a ruling to allow additional "saving rolls".

Joshua,

Yes, they really are selectors. I think I may be focusing on a higher level here than you. (That is, not higher level as in "more advanced" but as in "further up the food chain of stuff that happens.") You're talking about how mechanics may be misued, poorly designed, or how things that look like mechanics can do not-mechanics things. (As Eliot says, give an illusion of choice to forcing techniques.) I'm looking at the System (as in the Lumpley Principle) layer of things - "what mechanics are for."

I understand your underlying point - no, of course mechanics are not deterministic; stuff only happens in game *if we (the players) decide that stuff happens.* The mechanics simply facilitate the negotiation process - getting from point A - a bunch of different people who want to imagine something - to point B - a bunch of people who have agreed to imagine something specific. We use mechanics to decide what happens because we've agreed to beforehand. We think that the mechanics are more fun than just arguing about it, or the mechanics help us create a particular kind of play experience that we desire.

Um, just want to correct my last post and say that I should have written "instant", not "instance". Essentially, I've read threads where people have said they're quite happy to know that their GM is faking die rolls, using Moving Clues, etc. some of the time. They just don't want to have it shoved in their faces, and they don't want it to happen all the time. So when the GM starts rattling dice with an evil gleam in his eye, it can enhance their sense of risk, believability, whatever, even though they don't know for sure if the roll really means anything. Personally I'm not sure I'd enjoy this--I think I'd prefer to have the GM be more up-front about when he's guiding play.

(Mike Holmes provides some examples--search for his name in this thread for his experience in a game set in Middle Earth.)

And again, here is a plea to not use "deterministic" to mean "determinative".

Paganini wrote: Yes, they really are selectors... I'm looking at the System (as in the Lumpley Principle) layer of things - "what mechanics are for." ... We use mechanics to decide what happens because we've agreed to beforehand. We think that the mechanics are more fun than just arguing about it, or the mechanics help us create a particular kind of play experience that we desire.

Well now we're getting into a slippery area in theory since the Lumpley principle can be interpreted descriptively as saying that System is everything that goes into establishing credibility and setting "facts" in the SIS. Which could conceivably swallow up the players as part of System.

Maybe that would be useful in some analyses. However, Joshua, I think you're right to argue that it isn't very useful here. So I agree that even an established mechanic doesn't really do anything unless and until the participants decide what its inputs are (and these inputs can be established elements of the SIS as well as player declarations and metagame abstractions), how to convert these inputs into the "stuff" the the mechanic literally operates on (if necessary), and how to take the output of the mechanic and convert it back. But the System can regulate all of this, as you say, by establishing who gets to do the selection and conversion (interpretation), when the selection and conversion happen (FitM vs. FATE, for example), and limitations (establishing Stakes or providing parameters/guidelines for interpreting inputs and outputs).

The System isn't going to say that Glory skewers the Hierophant, but it might say that Glory has to define both the success (priest-kabob) and failure (whiff!) before rolling dice; it could equally leave interpretation until after the roll but at the risk of making the resolution mechanic meaningless. To recover the meaning you might need some strong paramaters. And of course a resolution mechanic can be far more than a die roll--it could entail the optional addition of resources to fix bad rolls, or raising stakes with rerolls, exotic mechanics like interpreting pictures on a card, etc.

Joshua wrote: If we want to raise the stakes up into Lumpley territory, the 'System' outputs not just 'success/fail' or perhaps degree of success/fail but also who gets to articulate that success or fail.  What the System does not do, and will never do, is articulate its own output.  The System does not describe the sword swing, the double entendre, the navigation error, or the like.  It may dictate it, but it does not describe it; it may set boundaries for the ensuing description, but it does not do the description.  At least for my purposes, I don't find this to be a determinative selection process.  Otherwise, one crit fail narrated by fellow player Joe is the same as another crit fail narrated by Joe.

Okay, horses held. I still think there's something slippery there if someone wants to pursue it, but I don't think it's germane to this thread.

Here's what I think we have so far, in a nutshell: resolution mechanics are (or ought to be) selectors, in the sense that to be useful they must be distinguishing possible alternatives from impossible ones and then assisting in selecting among genuinely-possible, genuinely-distinct alternatives. But how they select (in all kinds of ways) has a big effect on the interactions between the players and the game.

Are we leaving anything out, John?

I'm going to be nitpicky, Eliot.

Here's what I think we have so far: resolution mechanics are delimiters, in that they must be useful in distinguishing acceptable alternatives from unacceptable ones, and then assisting the players in selecting and articulating among the likely, distinct, and interesting alternatives.

Needless to say, how the system delimits alternatives has a big effect on the interactions between the players and the game.

I say delimit rather than select because the mechanic does not determine the development of the SiS and story, it only offers a narrower range of possibilities, from which the players choose.
Because not all systems are realistic, mechanics do not necessarily select possible and impossible alternatives.  The operation of the mechanic does, however, make some alternatives unacceptable -- the strongest example of this is the common 'throw the result out if you don't like it' rule.  Alternately, once you get that critical success and everybody around the table sees it, it's unacceptable to narrate into the SIS that you screwed up.  That said, I'm not sure if I like acceptable/unacceptable, either.
And lastly, it's is always and fundamentally the players who do the selection of what happens, and the articulation of how the 'what happens' plays out in the SIS.

But all of this happens after the mechanic is used.  This is all a matter of outputs, and while having an eye for the end product is important when talking about the inputs (ie setting the target number), we may be straying a little off-topic.  The rights on who gets to articulate the system's results and how that validates the players' interests are fascinating to me, but belong in another thread.

Paganini wrote: A mechanic on its own is useless. It has to have some made-up stuff to select *from.* The mechanics never make stuff up themselves. They can never tell the players what to make up. All they do is help the players decide from some possible made-up stuff which made-up stoff to actually use.