The Forge Archives

Quote from: On Pale Fire's 'Ygg take 3' thread in Indie Game Design, Gareth (contracycle)I propose we reconsider a setting thread - an active discussion of manipulating and protagonising setting, and how and to what end such an idea can be used. I think several people have been groping toward a concept which walt has neatly packaged for us.

What I packaged (it remains to be seen how neatly) was this:

Quote from: On that thread WaltWhat I'm imagining, in a vague way so far, is a system of metagame rules akin to a complete narrativist game system that is executed in secret and entirely by the gamemaster. This system uses its hero dice or whatever to regulate the GM's (or setting designer's) use of rule exceptions in such a way as to increase their effectiveness while minimizing the adverse effects, such as by keeping the GM on the right side of the fine line between "protagonizing the setting" and arbitrariness. For example, the system could give free-form attributes to parts of the setting ("The Old Forest is Semi-Sentient, Pathless, and Energy-Sapping") and rules exceptions occurring there would each have to tie into an attribute, perhaps with a die pool roll to determine a cost for the exception, paid for out of a limited currency whose starting value would represent the overall "wonderfulness" of that place. A player's direction-finding magic spell is countered by the GM drawing upon the "Pathless" and "Energy-Sapping" attributes of the forest, and if the GM pays the cost the dice determine, the player's magic fails. In other words, let's consider "protagonizing" setting and situation using the same metagame design techniques that have already proven effective at protagonizing characters.

Just a few comments to start:

1. "Protagonising setting" makes a nice catch phrase that captures the general thrust of what I'm thinking here. But it's not to be taken as a technically precise term. Arguments of the form "Protagonists do X, Setting can never do X, therefore what you're talking about is invalid" would point out only that the term is being loosely, perhaps metaphorically, applied.

2. My example of the old forest was poorly chosen, because the attribute "semi-sentient" is a red herring. If the forest is in any way sentient, then treating it as a character is obvious. The whole point is that setting could be treated this way even though it's not sentient.

3. Gareth pointed out:

QuoteIncidentally, it was mentioned in the Dramatica article - about how the setting like "driving storm" takes over the confrontational role filled by the supoprting cast.

My idea actually derives from a related and long-running idea in the Narrative Intelligence community (people using AI approaches to interactive storytelling on computers). This is that setting elements could be treated as quasi-characters whether they're filling in for antagonists in the story or not. The basic idea is that in conventional authoring, for example, a storm that affects the plot is fully controlled by the author's intentions. Having a storm created by a weather model is a poor alternative. But fully modeling an author's intentions for everything in a story is too complex and even if feasible, might be too rigid to admit interactivity. So, as a compromise, model the storm as a character instead; that is, as an autonomous agent with its own intentions (mimicking, but generalizing, what an author's intentions would typically be) and just enough awareness of what's going on in the story to be able to act upon those intentions. The idea of "everything in the story is a character" is generally considered a really good idea in NI, but has never actually been tried. (A not at all uncommon situation in NI, where actual development of ideas into works is glacially slow and inordinately expensive.)

4. This is all related to Illusionism. Exactly how, I haven't sorted out yet. But on the original thread Ron suggested a discussion of system design for supporting effective Illusionism. I think this might be it -- even though the final result might not necessarily remain Illusionistic.

5. Although "protagonising setting" is the more easily grasped (and the more directly applicable to Pale Fire's questions on the old thread), I think there might ultimately be more meat in "protagonising situation."

The most interesting analysis of Icelandic sagas I've ever seen was in a book called "Feud in the Icelandic Saga" by Jesse Byock. Byock's argument is that the sagas are predominantly organized around the feud (the conflict). In essence (this is my own paraphrase), the feud is the living breathing main character. The human characters and their relationship maps are just the growth medium that sustains it.

Without going to quite that extreme of a viewpoint, could we, for example, make use of attributes and resource pools and fortune-in-the-middle resolution for the dynamic development and handling of situations?

6. The big practical question in all of this is that protagonisation appears to be a limited-sum resource. If you "protagonise" setting or situation what's the effect on your ability to protagonise protagonists? Would this approach just lead back into a morass of "balance" issues?

- Walt

Hey Walt,

I think you're onto something potentially productive. But I think the notion of "protagonizing setting" is misconceived. A story has room for very few protagonists. My inclination would be to reserve those positions for the player characters. But protagonists need antagonists (to deliver adversity), and significant foils, and other support characters. The notion of "characterizing setting" (rather than "protagonizing setting") could be fruitful in this regard. There doesn't, in my mind, seem to be any reason why a storm or a fire or a forest couldn't be a characterized antagonist, or a best friend to a character.

You might take a look at the http://www.dramatica.com/theory/theory_book/dtb.html">Dramatica Theory Book for some ideas in the abstract about the functioning of support characters in stories. The idea of handling setting elements as support characters is interesting, virgin territory in roleplaying.

There was a http://indie-rpgs.com/forum/viewtopic.php?t=1498">Forge discussion of Dramatica a few months ago.

Paul

Well, I'm all for settings being given dice and so forth.  I see no reason that a forest could not have a rating in "tangled underbrush" for example, used as a resistance value for character movement abilities and the like.

Situations I see as being more tricky - certainly *I* see certain types of situations, crises if you will, having their own momentum which compels their resolution and frames its form.  But I can also see many arguments about what constitutes a situation, from various characters perspectives, or how "valuable" they are in the abstract.

Well said, Paul.

Walt, this idea is used quite literally in the game Primeval. When telling stories of conquering "inanimate" obstacles like mountains, etc, you actually get points for describig them as antagonists. So, as my hero scales Mount Vesetna, I would describe it's irritation at my attempt and how it hurls boulders down upon me, and spewes lava at me sputtering in rage. And how in a last final frustrated attempt to kill me it commits suicide by splitting apart in an attempt to throw me free of it's slopes.

Mike

OK, shot in the dark: Only characters get protagonised - setting frames the manner in which the protagonism happens.  The objective of a setting object or agent is to interact with the characters to protagonise them in an appropriate manner.  So one approach might be - if you have a place the characters are going to, you try to find an aspect of the place to reveal to that character which has relevance to them.  This gives the setting relevance, perhaps utility, in the players mind and hence encourages engagement.

In broader terms, I guess setting/situation are where the players get a problem to solve, however distributed the authorial and directorial rights to do so.  I would think that in selecting you are already conceptualising some ways in which it will be experienced by the characters, or by what dramatica calls the obatacle characters if they correspond to NPC's you have in your head.

Settings also have a property of being recurring, and so if people travel through ther haunted wood a lot, you might want a lot of ways to address characters being spooked - else it becomes mundane and "you travel from point X to Y".  I guess random encounters might have been an attempt to "give a scene" to the landscape.  

Dramatica is quite interesting, contains a lot of classification.  It has some interesting things to say about trepresenting the "we" and "them" viewpoints, though.  For one thing, I'd say that if a wood was haunted, reports of its hauntedness by the wider community would be needed to verify this and maintain the sense of wonder, of danger.  IOW, scenes must be built to show off the setting every so often and to reinforce what the character/players are here for.

I think one interesting point is that dramatica distinguishes between character and player - player as actual embodiment, as a character, and character-as-archetype.  Sometimes, the antagonist or contagonist or whichever might not be present in person, and that role fulfilled by the setting.  In that regard it might be possible to use dramatica's classifications to design a set of scenes by which the setting protagonises different characters.
2p

Great comments, folks. Going in order:

Paul, you're right about the terminology. I deliberately went overboard to make a point, and I should back off it. I like "characterizing" setting, situation, etc. for the double meaning of "treating like a character" as well as "describing." Of course, most people seeing that usage are going to assume the second meaning: "Sure I characterize my setting, see, I have a map right here." But I can live with that.

The bigger terminology problem is, what do we call the bad thing that happens when an emotionally signficant antagonist, supporting character, setting, or situation is resolved or dealt with in play in a way that retroactively diminishes its emotional significance? The bad thing that Theseus' use of the clew of thread did not do to the Labyrinth. The bad thing that Theseus's actions would have done to the Labyrinth if he'd gone in and mapped the sucker instead. "Decharacterization" doesn't convey it. And "deprotagonization" comes so temptingly close. After all, the diminishment of the element's emotional significance is important because, and only because, it ultimately diminishes the characters' protagonism. Theseus wouldn't be fully Theseus if the Labyrinth were mappable (even though the way he did deal with it was easier to execute than mapping it would have been; it's an emotional, not a practical, issue).

Since protagonizing the player-characters is one of my goals (which are different from Pale Fire's, I believe), I would tend to draw the line in such a way as to allow NPCs and setting to be involved in exploring questions related to the Premise but not able to resolve them. As NPC examples, both Galadriel and Boromir make decisions that illuminate and explore, but do not resolve, the Premise decision that Frodo is faced with. If in play, I as GM role-played Galadriel's and Boromir's crucial scenes, then yes I'd be telling stories to the players but I don't believe I'd be deprotagonizing their characters in so doing. As setting examples, consider Chandler's LA or Lawrence of Arabia's desert or the aforementioned Labyrinth or the Titanic. These settings are tied to Premise metaphorically, so closely that reacting to the setting is exploring the Premise. Limiting the player-characters' powers to diminish the significance of these settings through their decisions and actions is appropriate even if that does impinge somewhat on the characters' direct protagonism.

(Side note: how can discussion on The Forge explore story in so many dimensions and so extensively, yet hardly ever mention metaphor? It's uncanny, like a 1000-page book on photography that somehow manages to never mention light.)

Contracycle, I agree that situations are more tricky. I'm looking at this primarily as a technique for vanilla Narrativist play that's still largely GM-centric. So arguments around the table about what constitutes a sitation shouldn't be a problem. But for this to be a helpful technique, on a mechanical level the system would have to address determining when a character is "in" a situation (it's much easier to tell when a character's "in" a setting -- I think), and what operational effects being "in" a given situation would have. Both of these require the characterization of the situation to be largely in terms of its importance to the characters -- so we seem to pretty much agree on where the trickiness lies.

Mike, thanks for the Primeval example. What I'm visualizing is similar but presented very differently, in the context of a more conventional game structure. The GM is thinking of the mountain the same way as your hero is: as it being annoyed by the hero and using its powers to try to thwart him. But the description conveyed to the player doesn't necessarily anthropomorphize the mountain at all. The system behind the scenes is just a way for a GM, who might have no natural instinct for such things or who might have quaint ideas about preserving "fairness," to decide how much the mountain throws at the hero, when it's "defeated," and how that "defeat" should be narrated without dewhateverifying the mountain (and thereby deprotagonizing the hero) in the process.

Finally, cc (again), all good points. I'm not keen on the recurring setting unless there's a thematic purpose. Coming back to the haunted wood and being spooked yet again would need to have a cumulative or changing meaning of some kind to be worthwhile; merely being spooked in a new different way wouldn't do it for me. In practice I try to minimize recurring settings, which is dodging the problem. A system for characterizing setting might lead to a whole different approach (characters, after all, can be interesting when encountered repeatedly, even if they aren't changing, even more so if they are). A starting point might be, what does a setting learn? (Places don't learn, but characters do, and in stories the one reflects the other.)

- Walt

Thanks for pursuing this thread Walt. It's very interesting and I'm keen to see what will evolve from it.

Quote from: wfreitag
Finally, cc (again), all good points. I'm not keen on the recurring setting unless there's a thematic purpose. Coming back to the haunted wood and being spooked yet again would need to have a cumulative or changing meaning of some kind to be worthwhile; merely being spooked in a new different way wouldn't do it for me. In practice I try to minimize recurring settings, which is dodging the problem. A system for characterizing setting might lead to a whole different approach (characters, after all, can be interesting when encountered repeatedly, even if they aren't changing, even more so if they are). A starting point might be, what does a setting learn? (Places don't learn, but characters do, and in stories the one reflects the other.)

This makes me think about the Perilious Forest or whatever it's called in the stories about King Arthur. There are some interesting characteristics such as:
* You can go through it and not find a thing
* You can find something, but if you don't explore and return later it might not be there
* Once you have explored something it might disappear or suddenly become mundane.

These are tricks for the stories, because they cover the "what-ifs" people always want to know.

Maybe they are relevant in this case. I could easily see the Haunted Forest "demystified" (and rendered poweless) once the heroes has thorughly conquered it. At the same time, although the heroes (or protagonists) have learned how to become immune to its effects, the same cannot be granted to others. So the heroes become guides into this strange world where they themselves are immune to it effects, UNLESS they try to go beyond what they earlier have achieved (the taoist mystic living alone in the temple haunted by ghosts are not disturbed by them, but they still prey on wanderers. It is not until the taoist decides to help stamp out the ghosts that again the ghosts and the haunted temple forest turns against him).

So, I don't know if it's necessary to let the setting "learn". A reasonable model might be to allow a setting to have a certain limit (dice pool or whatever you envision) to stop a hero. Once used up, the setting cannot try to "attack" the hero any more  because he is now "immune" to it's effects. However, if the hero later tries to "demystify" the setting to help others withstand its effect, the hero may again be a target for the characterized setting's "attacks".

(Doesn't have to be a hero of course, it could just as easily be a bad guy or an NPC as well as a PC)

So wait a minute - we've just recently had a marvellous description of a PC who was a tree, right?  And Dramaticas interesting if not necessarily correct methodology allows for a dfistinction between player and character in their own terms, and for the substitiion of setting for character in terms of player.  So - is there any reason we can't just declare the setting to be an NPC?

Most of our "character description" tools are based around hominids, with rare exceptions.  But the looser systems like story engine, while making that assumption, don't mechanically require the subject to be human(oid).  Does a character even need to be an organism?  There is precedent in many rules for swarm monsters that are treated collectively; surely with a suitably broad description method we could write character sheets for elements of the setting and actualy use them in play.

For example, HeroWars character description is carried out at least in part by drawing keywords from a 100 word paragraph.  There seems to me no reason that a 100 word paragraph about a settlement could not also be mined for keywords and that setting entity attributed with appropriarte values.  Especially in the case of social structures, there are a number of parallels with biological organisms, but even ecological habitats can exhibit some of these properties.  A mountain range could be given attributes, and this, for example, used as a counter value against the likes of earthquake magic or whatever.  The landscape would consist of systemitised objects which are capable of interaction.  Tihs would also provide a natural vehicle for locating more exotic elements such as animist spirits, or for rating fortifications and so forth.  Again, precedent lies in some of the city supplements which attribute social and economic ratings to neighbourhoods.

Gareth, I absolutely agree. This is exactly along the lines I was thinking when I wrote "...using the same metagame design techniques that have already proven effective at protagonizing characters."

Systems like The Pool that are already free-form in nature need no new rules to do this. (Perhaps just some in-play precedent, or perhaps not even that.) Hero Wars is a good example of a game where the existing tools could be modified just slightly to permit setting-description effects. A setting descriptor mechanism would also fit very easily into Donjon Krawl. I don't know enough detail about Sorceror to know whether places can have their own Demons. But in any event, one might also decide that these games just don't need it. Their focus is elsewhere and the ability to character-ize setting might not advance their goals very much.

But I look at Yggdrasil or, for that matter, AD&D and it appears to me that these types of mechanism for character-izing setting would kick ass in that context. Exploration of setting is one of their main premises, and yet they have no way within the conventional sim rules to convey the character of a setting. The descriptions are filled with fantasy adjectives: steep, forbidding, gloomy, ancient, rugged, misty, arid, rocky, vast, stormy, verdant, blasted, icebound, bewildering, broken, bottomless, craggy, splintered, ruined, starlit, eerie, oppressive, trackless, relaxing, broken, torrential, chilling, forgotten -- and no mechanism for making any of these adjectives meaningful in play. Instead, setting descriptions are filled with often poorly play-tested special rules effects ("any character stopping to rest in the peaceful shade by the pool will fall asleep for 5-20 turns") that have no connection to whatever the players are trying to do at the time and so end up being just annoyances. Locations with a magical quality might, at best, have a list of player-character spells and item effects that the place interferes with. For the most part, setting is characterized (but not character-ized) by what sorts of monsters live there! It just doesn't work, which is the problem Pale Fire started out with. That's where metagame-level setting character-ization tools, whether used overtly or illusionistically, could make a difference.

- Walt

Quote from: contracycleSo - is there any reason we can't just declare the setting to be an NPC?

For example, HeroWars character description is carried out at least in part by drawing keywords from a 100 word paragraph.  There seems to me no reason that a 100 word paragraph about a settlement could not also be mined for keywords and that setting entity attributed with appropriarte values.  Especially in the case of social structures, there are a number of parallels with biological organisms, but even ecological habitats can exhibit some of these properties.  A mountain range could be given attributes, and this, for example, used as a counter value against the likes of earthquake magic or whatever.  The landscape would consist of systemitised objects which are capable of interaction.  This would also provide a natural vehicle for locating more exotic elements such as animist spirits, or for rating fortifications and so forth.  Again, precedent lies in some of the city supplements which attribute social and economic ratings to neighbourhoods.

My Accord (renamed from Zero System after suggestion from reader) generic game system encourages and supports this process. The written description can come from any source, either player imagination, game or other setting fiction, historical records, mythology or legends, or even tourist brochures! :) Basically the descriptors embedded in a description can be used as game system values and so used to oppose PCs. So a "mountain with steep cliffs", requires climbing skill (to cross those cliffs), and is one significant step more difficult than a "mountain with cliffs", which is again more one significant step more difficult than a "mountain" (which doesn't require climbing skill to cross).

Quote from: wfreitagThe descriptions are filled with fantasy adjectives: steep, forbidding, gloomy, ancient, rugged, misty, arid, rocky, vast, stormy, verdant, blasted, icebound, bewildering, broken, bottomless, craggy, splintered, ruined, starlit, eerie, oppressive, trackless, relaxing, broken, torrential, chilling, forgotten -- and no mechanism for making any of these adjectives meaningful in play. Instead, setting descriptions are filled with often poorly play-tested special rules effects ("any character stopping to rest in the peaceful shade by the pool will fall asleep for 5-20 turns") that have no connection to whatever the players are trying to do at the time and so end up being just annoyances.

I think you're absolutely right, but if we step down from the abstract concept to tangible game mechanics... how the heck does one implement this well?

Your initial suggestions led me to think you were suggesting a system  where the setting is given a (limited) number of freely variable challenges which it might be able to try to stop the characters with, but that's only one of the ways the setting should be empowered.

It would make sense to issue a challenge rating to the different ways the setting could try to resist the characters, a little like skills for characters.

Then somehow there has to be a limit to the amount of challenges a setting can apply to the characters. We can think of this as the "hit points" :) of the setting.

But the applications should go beyond that of challenges, things like "cannot be affected by magic" is not really a challenge, but should be covered by the same rules.

Aside from challenges and immunities, what more characteristics does a setting possess?

In Herowars, a complex opposition (don't know if it's the real word in English) is always a kind of conflict ...

Let's say a character with climbing 5M (which means a score of 25) wants to climb a steep mountain, difficulty 10M (a score of 30).
That means :
the character will have 25 action points (kind of life points), the mountain will have 30 action points, and they'll try to eat each other's action points, by betting APs and opposing dice rolls (the mountain seems to be better than the hero, but he's got Hero Points, and certainly some magical help).

I remember a scenario in which the character struggled against an earthquake on a mountain ; the "conflit" was an hour long, with a series of risky bets, critical rolls, prayers ... just 4 heroes trying to escape falling rocks and other dangers, helped with a gamist+narrativist system (gamist because you really feel like you're playing poker, narrativist because the simple and universal system emphasises creativity into describing what happens).