This is a project that I and some folks over at the GiantITP boards have been working on for a while now, but while mechanical feedback is both sought after and welcome, what really intrigues me are a few ideas for freeform plot construction, based on the discussions I, David Berg, Gareth(contracycle) and many others mulled over recently.

David and I had some fundamental disagreements on how this form of gameplay should develop (which we resolved privately,) but I think that Ursula le Guin's Earthea setting offers particular potential for some of these ideas to develop.
No. 1:
Earthsea is divided into many small-to-large islands, seperated by ocean, which provide discrete limits for bite-sized locales, along with associated ethnocultural groups or political factions.  The sea itself provides particularly rapid, but risky, transit between such locales.
No. 2:
The Earthsea setting has an inbuilt kismet/karma system, with the notion of being 'bound to the consequences of one's actions' inherent to the Equilibirum- moreover, the more powerful a character becomes, the fewer effective choices they have.
No. 3:
Earthsea is already a rich and detailed setting, large enough to be interesting but small enough to be manageable, so that the DM needn't worry so much about conjuring up background details without contradiction.
No 4:
While the notion of rigorously enforced cause and effect is absolutely crucial to the setting, most conflicts in the story are dealt with through travel, diplomacy, subterfuge and, of course, magic, rather than straightforward combat.  All these channels have a potential for very unexpected outcomes, which may be useful for the GM.

As far as mechanics go, I have some basic rules worked out for spell composition, attributes and 'qualities', if anyone's interested o hear them.  But the single most important goals are as follows:

1.  No explicit railroad plot.
2.  A method for introducing player characters gradually and establishing their qualities in an 'organic' fashion.
3.  The ability to generate an endgame scenario (by gradually narrowing options for the players.)

I'm actually hoping for a system that's largely narrativist, (in the sense that any complicated or exacting 'realistic' rule sets would only be referenced between sessions- any checks that have to made on the spot should be relatively quick to perform.)  Balancing in the gamist sense would be useful but not vital (there's a significant gulf in power between wizards and everyone else regardless.)

Anyways.  This is just to lay out the basic premise and objectives of the system.  I'll be supplying more details later, so if you're curious on any particular point (and/or don't feel like wading through the referenced threads,) just ask.

Well, since the plot generation guidelines are what most interest me, I'll recap on the essential guidelines that I use from the 'player-directed-exploration' thread first.

Players want:

*  A set of challenges to overcome.
*  A believable and fleshed-out setting, discoverable to varying degrees through exploration.
*  Freedom to do as they choose, combined with significance to their actions.

The GM wants:

*  To be able to construct a world in accordance with his/her interests.
*  Minimal workload, granting a reasonable balance between preparation and improvisation.
*  To be able to provide material for the players' explorations.

The world is composed of Elements, which include locales, interests, and events:

A Locale is simply a physical location that the player characters can explore.  If it's possible to travel from locale A to locale B, it should- all else equal- be possible to travel back from B to A.

An Interest is a faction or personality with some vested stake in proceedings within a particular locale, or several of them.

An Event is tied to to given interests and locale, and can be any significant happenstance.  (Even rather slow, large-scale changes are considered Events for such purposes.)

Level of Immediacy (LoI) describes how likely it is that the player characters can explore a given element in the near future.  Particulary obscure or remote elements have low LoI, particularly close or obvious elements have high LoI.

Level of Persistance (LoP) describes how difficult it would realistically be for the player characters to alter or affect a given element.  Particularly large-scale elements representing powerful, deterministic forces will typically have high LoP, small-scale elements dependant on chance or weakly supported will typically have low LoP.

At the end of each session, the players give the GM an idea of where they want to go and what they want to explore.  The GM will then do some preparation between that session and the next for such purposes.  Simple enough.  The players may also specify long-term goals for their characters as they are established.

The GM may use incremental detail to rough out events, locales or interests as he likes, but the degree of detail he invests in such preparation depends on LoI and LoP.  Events, locales, and interests with high levels of Immediacy and Persistance should get most attention, those with low levels should get correspondingly less.  As the players move around and impact the world in however small a fashion, different elements should be roughed out in accordingly greater detail, and certain events may need to be dropped entirely (as they are no longer plausible.)

The GM may not script elements which actively interfere with the players' goals or explorations after such goals/explorations have been specified, except as a consequence of the players' choices- and the players must have 'fair warning' of potential adverse consequences first.  Elements which obstruct goals or exploration in this fashion are known as Constraints.  Note that trapping the players between a rock and hard place in terms of available choice is, in itself, considered a constraint.

There are two additional 'mechanics' for fleshing out the story which I'd like to touch on.  The first is that, while the GM is establishing details about the world and how it works, the players can establish details about their characters in a similar fashion.  If the player wants to decide his character has great strength, for instance, then he/she can introduce this quality during a given session (there is a cost/rarity buy point mechanic for such purposes)- provided that this does not contradict previously established facts about that character.  Long-term goals for the characters are used to establish plausible motives for the party to band together and work toward common objectives.

The other situation where the GM may seize direct control over the plot is in instances where the players' luck just runs out.  The most obvious example is player death ...because it can't happen.  The game contains no resurrection mechanics- at all- so the worst that can normally happen to a player character is that he/she becomes Stricken- mortally wounded, exhausted/comatose, etc. etc.  To maintain plausibility in such cases, the GM may be obliged to script some form of fortuitous rescue circumstance for the players' benefit.  (This is tied to the notion that random events in Earthsea aren't random events- they reflect an underlying pattern.  Not hugely original, but there you have it...)

I realise it all seems a little vague, so I'll be giving some more details on the game mechanics thus far in another day or two.  Any thoughts?

Alfryd, can you post or PM me about the magic rules your using? I'd be interested to see those.

The magic mechanics use a seed system, where you put spells together from various subcomponents, and calculate their cost in terms of power and skill accordingly.  There's a more detailed example here.
Spells would be composed or researched between sessions (there is a provision for making up spells on the fly, but it's much riskier, to avoid in-game slowdown.)

The attribute system is based off 9 guages, as illustrated here.  These guages are arranged on a grid, with opposite attributes conflicted- passion/wisdom, action/concentration, etc.- and adjacent attributes related, (such as endurance between wisdom and action, or concentration between trust, logic and muse.  I'm hoping to build in suitable levelling mechanics on this basis.)
My plan so far is that levelling of a given attribute accounts for most character progression.  Skills (such as Barter, Skirmish, Navigation etc.) are based off either one or two attributes, and can only be advanced through trained qualities (similar to feats in D&D,) giving a relatively limited check bonus (up to +10.)

Aside from this, characters are defined by taking innate or background qualities, which are similar to merits/flaws.  The main difference is that many qualities modify eachother's cost and rarity (for example, Great Strength is rare among those of Small Stature.)  Rarity exists to place a limit on how outlandish you can make a given character, without penalising effectiveness.  Anyhoo, it's nothing too radical.

The basic check mechanic is something I'm still unsure about.  At present, you effectively roll 3d20 + skill/attribute bonus against a given difficulty rating, comparing each die result seperately.
3 'hits' = critical success
2 'hits' = partial success
1 'hit' = partial failure
no 'hit' = critical failure

So I don't have to worry much about margin of success.  The difficulty rating would either be fixed by circumstance, 10 + opposed skill bonus for defensive checks, or maybe 'take the middle value' from 3 rolls of d20.  There would also be mechanics for assisted checks.
I'm still worried it may be a little too intensive for something narrativist, but there you are.  I have a set of basic combat rules roughed out, since they're the easiest, if you'd like an example of skill use.

One area that I would definitely like to touch on for relation to storyline is long-term training and Lore skills.  Most skills (and attributes) are improved through practice (the XP varies with check result,) but Lore skills can only be improved by long-term training or tuition outside of active adventuring.  Of course, taking months or years off to read up on herb lore offords many opportunities for plot hooks to emerge, which hands influence to the GM.  On the flip side, a successful check on Lore skills (such as History, or Contacts) allows the player to establish facts about the world independantly of the GM, which hands influence to the player.  I'd love to hear some suggestions about mechanising/arbitrating that kind of interaction.

Hi Alfryd,

I'm a little fuzzy on what all the facets you've presented are supposed to add up to.  I get the impression that play is supposed to produce fun adventure stories where the characters gain power and acheive their desires, with some sort of satisfying ending assured.  Is that about right?

I dig the idea that there is an ending.  I'm wondering what your criteria would be for a "good" one.  Something where the characters achieve a dramatic victory in a tough situation?  And die tragically in the process?  Or live happily ever after?

Feel free to ignore any of the questions below if I'm asking things hopelessly out of context.  Some of these might be useful food for your thought while others will simply be clueing me in -- I dunno which are which.

Alfryd wrote:
Earthsea is divided into many small-to-large islands, seperated by ocean, which provide discrete limits for bite-sized locales, along with associated ethnocultural groups or political factions.  The sea itself provides particularly rapid, but risky, transit between such locales.

Alfryd wrote:
The Earthsea setting has an inbuilt kismet/karma system, with the notion of being 'bound to the consequences of one's actions' inherent to the Equilibirum-

Alfryd wrote:
moreover, the more powerful a character becomes, the fewer effective choices they have.

Alfryd wrote:
Earthsea is already a rich and detailed setting, large enough to be interesting but small enough to be manageable, so that the DM needn't worry so much about conjuring up background details without contradiction.

Alfryd wrote:
The world is composed of Elements, which include locales, interests, and events:

Alfryd wrote:
The GM may use incremental detail to rough out events, locales or interests as he likes, but the degree of detail he invests in such preparation depends on LoI and LoP.  Events, locales, and interests with high levels of Immediacy and Persistance should get most attention

Alfryd wrote:
The GM may not script elements which actively interfere with the players' goals or explorations after such goals/explorations have been specified,

Alfryd wrote: except as a consequence of the players' choices- and the players must have 'fair warning' of potential adverse consequences first.

Alfryd wrote:
There are two additional 'mechanics' for fleshing out the story which I'd like to touch on.  The first is that, while the GM is establishing details about the world and how it works, the players can establish details about their characters in a similar fashion.

Alfryd wrote:
Long-term goals for the characters are used to establish plausible motives for the party to band together and work toward common objectives.

Alfryd wrote:
the worst that can normally happen to a player character is that he/she becomes Stricken- mortally wounded, exhausted/comatose, etc. etc.  To maintain plausibility in such cases, the GM may be obliged to script some form of fortuitous rescue circumstance for the players' benefit.  (This is tied to the notion that random events in Earthsea aren't random events- they reflect an underlying pattern.  Not hugely original, but there you have it...)

Thanks David, the feedback is much appreciated.  I'm sorry I can't give more concrete answers at present, but I may have something more fleshed out later.

I'm a little fuzzy on what all the facets you've presented are supposed to add up to.  I get the impression that play is supposed to produce fun adventure stories where the characters gain power and achieve their desires, with some sort of satisfying ending assured.  Is that about right?

I dig the idea that there is an ending.  I'm wondering what your criteria would be for a "good" one.  Something where the characters achieve a dramatic victory in a tough situation?  And die tragically in the process?  Or live happily ever after?

This could either be awesome or awful, depending on what the players hope to get out of the game.

Is the expectation that anyone running this game has read the Earthsea books?

What sorts of actions are thus incentivized/disincentivized?  There's some kind of "changing your mind" that's punished, right?  What kind?

I'm assuming that this warning is supposed to come via transparency of in-game situations and not through out-of-game GM-player communication, right?

How do you see "rapid, risky" transit playing out?  Is this something you want the players to use on a whim?  Is it something they should use as a last resort, and struggle to survive?  Should there be major resource costs for travel?

I'm curious about how much of this you intend to tell players, and about whether you plan to apply the "non-random-events" metaphysic in other circumstances.  If you don't, this will feel contrived (though possibly to a quite acceptable degree, e.g. if all you care about is plausibility).

I've had bad results with letting players pick individual goals first and then trying to ensure reasons for them to stick together.  Obviously it can be made to work, but you might wanna instruct the GM on how you think that can best be made to happen.  (This may already be obvious given other systems you have in place; I dunno.)

An alternative is to come up with some sort of group purpose first and then come up with individual goals.

Alfryd, do you have the rights to use the Earthsea setting? I mean, as far as publishing, it seems like those are going to be pretty expensive.

Oh, heavens, no.  This is just for experimental purposes.  If the system turns out to work well enough, I/we might give some thought to publishing.  Do you have any comments on/suggestions for the magic system?

Alfryd,
I recently bought Burning Wheel and am reading through it.  The guy who designed it claims a lot of inspiration from LeGuin, and my first impression is that the game seems capable of a lot of the stuff you want (as far as I understand it).  Just thought I'd mention it. 
I'm brainstorming to see if I have anything further to add here.  I definitely have more questions, but I'm not sure if they'd serve much purpose.

I'll certainly take a look.  Thank you David.

Hi there

I am an enormous fan of Earthsea (as so many are), from the time I went on a week long Earthsea summer camp as a child to the present day, so I am enormously excited about any sort of game of Earthsea.

But making games to capture the essentials of books like this is, in my experience, a very tricky thing, because you are after the spirit and not just the action. It would be easy to write up Earthsea as a setting for (say) GURPS or D&D, but it would do nothing to capture the essential feel of the books, any more than the Earthsea mini-series did.

I know you are already some way down the line with mechanics, but what I would like to see is a statement of what essential facets of the books you are trying to capture, and then how the mechanics tie up with those. That is especially true of the two biggies in my book, magic and the sea.

The Sea

For the sea I like the way you are thinking of using the sea as more than a means of travel. Sure trading ships go everywhere in Earthsea, but in the books the act of crossing the sea is always with a purpose and a symbolism, and is really nothing much to do with getting somewhere, and nothing at all to do with how. I'd be inclined to have (off the top of my head) something which said you can always take a ship to where you want to go, but have some sort of action of fate in play when people are at sea, like

• You get where you said you would go, except ...
• The GM may spend some resource to try and invoke the sea to obstruct or deflect you
• You can also spend some sort of resource to have the sea take you where you *should* go

So for example Ged sets sail with no destination in mind but to hunt the Shadow, the rules say he *will* find the Shadow, unless the GM spends a resource to have a storm, or pirates, or another island, or whatever deflect or obstruct him, or the player spends a resource to have the sea take him where he *should* go, in this case to the island of the old woman and the old man with the ring.

Magic

For magic, it's tricky, but I have to admit that though I am a big fan of Ars Magica's logical and modular spell system I don't like the thought of one for Earthsea. You don't see Earthsea wizards memorising spells, and it's not part of the spirit of the action for them to sit around constructing them, and if it isn't right for the characters why should the players have to do it? I'd rather see a set of skills related to, say, Naming, Summoning, Transformation and the like, which you roll against to do magic just like any other skill, just as a wise-woman or witch might have a few small spells to go along with her mundane skills which you would not want to have to quantify.

In Earthsea everything is magic, and also nothing is, because magic is essential to the way the world works, as you note with the Equilibrium. Making the 'avert' gesture, and naming Yevaud are the same sort of thing, just on different scales, so having a system of magic that would be easy for any character to have a little bit of would make sense to me, while some parts were reserved for specific lores.

So you might have skills like

• Healing
• Summoning
• Transformation
• Naming
• Weatherworking
• Lore of Peln
• Lore of the Terranon

And so on, each with a description of the sorts of magic that can be attempted with it, and who can learn it. That way all attempts to heal, whether by a wise-woman or the Master Herbal, would use the same skill of Healing, and common Windworkers could learn Weatherworking, but only a Mage of Roke would learn Transformation and so on.

Death

I love the idea of no accidental death! No one dies in the books without a reason that has nothing to do with physical circumstances, which is a great idea. Players should be allowed to have characters (their's or NPCs) die, though, to achieve some sort of plot resolution. Perhaps they gain some sort of resource (the same they spend at sea?) when people stories are resolved, whether by death or otherwise? So you can gain a token (or whatever) when you deal the death blow to the slaver lord, or help the young woman find love, or nurse the old woman on her death bed and guide her to the wall ... and then spend it to have the sea take you where you need to go?

Lots of ideas ... hope they are not too objectionable! :)

Much as I appreciate the feedback, to be honest I haven't been working on this project for a while now.  I consider your suggestions valuable but I might not be able to make use of them in the near future.

(Just a quick note: I wrote LeGuin two or three years back, asking whether she'd considered letting someone create/publish an Earthsea RPG. She was very negative to the idea of games based on books, and said that even though the creators of the Earthsea TV film/series had the license to create games based on Earthsea, she hoped they wouldn't use them. I've written a storytelling/role-playing game that uses the books as an example - it's called Archipelago).