After taking a BIG break from my hobbies (board games, pen & paper games, wargames etc), my interest in game design seeped back into my life rather naturally.  I was looking at the terrific artwork in my monochrome B1 In Search of the Unknown (Carr, 1979), but felt disappointed that playing D&D doesn't actually feel anything like the storytelling the art portrays.  Later, I was reflecting on the peculiarities of a Tolkien adventure narrative (specifically, in The Lord of the Rings and The Hobbit) vs the gameplay of a D&D (1974) adventure.

So I casually set about creating the infrastructure of a good, flexible, storyteller game that is a good fit for Middle-Earth (the setting of the aforementioned Tolkien stories).  My design goals are flexible, light gameplay that encourages the particularities of Tolkien's conception of adventure while remaining easy for the referee to control and direct.  What do you think?  Feel free to add ideas for game mechanics.

Experience
In these stories, fights are to be avoided, and the point of the quest is not to get rich (even in The Hobbit).  The adventure is a long journey (There and Back Again), with different episodes to detail each step along the way.  Experience will be granted only for travelling to a new destination (fighting monsters may be inevitable, but it is not particularly rewarded).  The mission then is to use your resources to get to the next location and make it one step closer to the Lonely Mountain (or Mount Doom, or Carn Dûm, or the Undying Lands etc).  Every time you advance in experience, you may add 1 to any attribute.

Heroes
Rather than providing different game mechanics to help the party succeed, each hero will work similarly and it will be the fellowship and support their clever players' can offer that will make each character unique.  Each hero has five simple attributes (courage, wisdom, fellowship, agility and might) that are determined by rolling 1d8+6 for each in turn (resulting in scores 7-14).  Any hero can be a human (who gain +1 to their greatest attribute), but you must "roll into" other races by having appropriate above average and below average attributes: Dwarves (Might 11+, Agility 10-), Hobbits (Courage 11+, Might 10-) and Elves (Wisdom 11+, Fellowship 10-).

Core Mechanic
When you enter a setting (see below), each attribute provides a pool of dice (every roll uses 8-siders) you use to advance to the next setting (when the pools refresh).  So if you have 14 Might, you will have this many Might dice when you arrive at the fortress of Barad-dûr and you have only this many dice for Might-based tests to advance to the next setting (Mount Doom, perhaps).  To take a test, the referee will describe the situation as best your hero would understand it, you pick however many dice from the appropriate attribute pool and describe your action and roll the dice.  Each die that equals/exceeds the target number (based on how difficult the action would be, between 2 and 8) is a successful die.  The trick is, the player does not know what the target number is, and must choose which dice to keep and which to "drop" (based on his best guess at how difficult the referee made the action sound).  If he keeps only successful dice, each die indicates how far along in the action the hero got (so it is always at least partly successful).  If he got greedy and kept failed dice, the roll is a fumble - not only do none of his successful dice count, but something bad likely happens.  Spent dice are lost, although each roll refreshes a single die to the pool (so you never have fewer than 1 die in the pool).  Most rolls are closed, but on open rolls, each natural "8" explodes, allowing you to throw another die into the roll.  When the referee rolls tests for his NPC's or monsters (because he effectively can't fumble), his target numbers should be a point or two higher (as he sees fit).

For example, say you wanted to scout the orc encampment for a weakness or a way around.  You declare your intention, pick up a handful of your Agility dice (let's say 5 of 12) and roll the dice (coming up with 8, 7, 5, 3 and 2).  From the referee's description, you guess the orc camp is fairly alert, and keep only the 8, 7 and 5.  The referee declares that these each beat the threshold (in his mind, he had assigned a difficulty of maybe 4) and that you make it over the stream (first success), past the patrols (second success) and up a hill to see a break in the picket the orcs have failed to observe.  If the referee had chosen a difficulty of 6, the roll would have fumbled and the hero may have been captured by the orcs.

Settings are kind of like extended scenes - if you arrive at a ruined fortress, you could camp at the edge of the forest one day, sneak around the ruined tower another, attempt the hedge maze and then fight a troll in the bailey.  It is one long, drawn out scene as far as character resources (attribute pools) matter, and will usually present a number of challenges (mental, physical and social) to advance beyond.

Combat
Combat uses simple tests.  Actions are declared, movement and maneuvers are imagined and attacks are made in order of weapon length (initially) and speed (in subsequent rounds).  One successful Agility die will inflict a hit on an enemy (against a difficulty relevant your opponents skill and battlefield conditions, whatever the referee dreams up).  Successful Might dice (against a difficulty relevant to the opponents armour) indicates the degree of damage inflicted (daggers can only test 1 Might die here, while bows and spears can test 2, longswords 3, axes 4 and great weapons 5).  It's important to note that the referee can make up difficulties as he sees fit, and these can (and should) change from round to round (bodies are always in motion and situations are constantly changing, there is no reason to have a fixed defense number for orcs or dragons) - although communicating the narrative to the players is critical to give them a chance to interpret which dice might fail or succeed.

Each degree of wound causes the defender to lose 1d8 from all attribute pools simultaneously (to a minimum of 1).  Being reduced to 1 in every pool results in an incapacitated hero.  Three wounds and a hero perishes.  Wounds should have their location described and noted, as the referee may want to consider this in later difficulty thresholds, or whether a character with unhealed wounds should refresh his attribute pools entirely in the next setting.  Foes fight either as groups (like a mob of goblins) or large, single monsters (like a dragon, troll or giant spider) - either way they have a single pool (maybe 20, 30 or 40 dice) which they can draw from to act.  Groups should lose members to wounds whenever the referee sees fit (a wound that causes only a few dice to drop is better described as a group member being wounded).  As they use up their dice to act (and lose them from wounds), the groups morale fades until wounds drop them below 1 die and the last member(s) fall or flee.

Different maneuvers can always be attempted using different rolls (as the referee sees fit).  Perhaps you can assault an enemy (each successful hit allows you to make a subsequent attack, unless the opponent chooses to give ground) or lock blades and push him back (opposed might roll, with fumble indicated someone falling).  Maybe you can parry instead of attack, adding each successful agility die to the number of successes the enemy needs to hit you (which by default is only 1).  Perhaps a monster induces fear (the Balrog might cause fear 5), which drains every attribute pool a set number (minus every successful courage die).  Obviously, there is a lot of flexibility.

Magic
To fit Middle-Earth, magic should be very rare.  Magic items, or skill in sorcery, use the same mechanic - a separate pool of dice that refresh every setting, which can be either added to attribute rolls (if the spell is appropriately described) or rolled as an attribute in its own right to accomplish some action.  As noted in this article, Middle-Earth magic tends to enhance and work on already existent things, rather than conjure unnatural effects.  Again, each successful die indicates how far along the spell got.

So anyway, that's most of what I came up with so far.  Is this a workable basis for a game?

1. Certainly it sounds like workable start to a fun game. I'm especially intrigued by the idea that experience comes solely from seeing new places.

2. I notice that all of your examples invoke Middle Earth or Tolkienesque except combat. Can you describe potential Tolkienesque combat using your ideas?

Jerry

Some questions:

If humans get +1 to a stat,  what is the advantage of "rolling into" a different race?

The conflict resolution mechanic is interesting. However, who determines how many success dice are needed to fully succeed at a task? Is it completely up to the GM, so that they can always make it take one more success dice to achieve my goal than I roll?

The guessing what the target number is interesting to me, but sounds built to halt the players in their tracks too easily. Not sure if there couldn't be a narrowing of the range, eg high or low difficulty indicating 5+ or 4-, or a set range of 2 or 3 difficulties max. Or perhaps a mechanism to reduce failed dice from fumbles to just negative successes.

I have a weak spot exploration game, but unfortunately these types of games can be very hard to poll off.

I think that you make the common mistake of not having a clear idea of what kind of stories you want to be toled in this game. Or if you have a clear idea, it does not really show in the system you present.

In these stories, fights are to be avoided, and the point of the quest is not to get rich (even in The Hobbit)

Why not have each setting consist of 1 to 8 obstacles?
When the players reach a setting, the GM can inform them, "There are three obstacles to overcome to get through Mirkwood."

If the players succeed at a test, then they move forward. If they fail, then an additional obstacle can be added to the setting or the GM can choose to let the group overcome the obstacle but the character is "damaged" -- reduce a pool in the NEXT setting.

With some guidelines like these, the players will have a better idea how to manage their resources.

Also, it doesn't seem like you need a separate combat system. If you really want a simple, fluid system, just use your basic resolution mechanic for combat. Perhaps offer one tweak: rather than setting an arbitrary difficulty, the GM rolls for your opponents (maybe even in the open).

-L

Good comments.  I do give a lot of control to the referee ("GM"), mostly so he can control pacing and excitement.  Back when I used to run D&D, a player would declare some action and I'd usually either 1) let them do it automatically or 2) ask them to roll any shaped die they wanted.  Only after seeing the die result (higher is better), I'd decide if they make it.  I think I'm taking some notes from how I ran D&D for this game as well.

Certainly the referee may make up the difficulty threshold after he sees the dice rolled (although this should not be obvious to the players, who might feel it unfair).  3's are a good target number for easy tests (that the referee wants to describe in many stages, like scouting the orc encampment above).  7's are a good target number for hard tests (and for tests that the referee wants to either pass or fail without degrees of success, like picking a lock).  The referee, of course, is not for or against the heroes, and should be open to how they want to defeat a setting (even though there was no epic battle with the Balrog, the "cheap" way the Fellowship avoided him made for great storytelling).  Usually an action will only require 1 success - the referee should make it fairly obvious when more are required (and he should set the difficulty much lower to allow for it), saying something like "well to find a weakness in the orc picket line, you will have to first enter their territory, then evade the patrols, and then sneak undetected onto the hill top".  Partial success is always success (he won't ever get the fumble result, getting caught, unless he fumbles), so that two successes will give the hero some information he can work with, if not the complete story.

To best capture a Tolkien-style adventure narrative, the final destination and mission will be clear to the heroes from day one.  This is always very far away, however, and getting their is the adventure.  In The Hobbit, the episodes the characters face in their pilgrimage are largely unrelated to the ultimate mission.  In The Lord of the Rings, each new setting brings more and more foreshadowing of the ultimate mission.  I think both of these styles would work fine in this game, and which is best for you depends on the game style your players enjoy.  Here's a sample adventure path:

The party forms in one of the last surviving settlements of refugees in Arthedain, down on the coast of Harlindon (see map).  Arthedain and the neighboring kingdoms have become lawless and empty regions, where the last, wandering minions of the Witch-king of Angmar are fought not by armies, but by remnant adventurers.  They are given a task by a Gandalf-like figure (perhaps Gandalf himself, perhaps one of the unnamed Istari), to reclaim the ancient home of the Elf lords in the Grey Mountains (if there are several elves in the party, for instance) and drive out the vile denizens led by the lieutenant (aka the Witch-king) of (dormant) Sauron.  By doing this, the men of Arthedain will once more be able to build a safe kingdom for themselves, and the recent Hobbit migrants will be able to settle their new Shire in peace (some hooks for other players).

Setting out, the party will have to cross the Blue Mountains (perhaps a goblin warren introduces problems), deal with the elves at Tower Hills (or Grey Havens, whichever they think will be more hospitable).  From here, the party has a choice - traverse the Shire (a place currently scoured and terrorised by the agents of the Ringwraiths since Arnor's fall), or sneak through Annuminas (an ancient city, long since abandoned to ruin, and rumoured to be haunted by ghosts).  Then it is onward to Fornost, the last capital of Arthedain (likely fallen at this point, but the referee may want it still to be hanging on, or under siege) and into the North Downs (a deserted and dangerous range of hills that will take wisdom and courage to traverse, and perhaps a fight with a hill troll).  From there, the party must enter the Forodwaith (a frozen desert) to cross the realm of the Witch-king of Angmar.  Perhaps they are captured out in the open and taken to Carn Dûm (the capital of their enemy), and must escape westward towards their ultimate destination in the Grey Mountains (sneaking past Mt. Gundabad in the process).

Each of these settings might take up a single session, and after each session the heroes would advance one level, so that they would each be atleast level 9 by the time they arrive at the Grey Mountains.  If we adopt the adventure style of The Hobbit, the referee should throw even more encounters in-between (every time with a new creature, either hostile or not, like the Mirkwood Spiders or Beorn), just to broaden the fantasy world with each new discovery.

Evan wrote:  In The Hobbit, the episodes the characters face in their pilgrimage are largely unrelated to the ultimate mission.  

Experience
Gained only for travelling to new areas (there and back again for the best reward)
Rewarded upon arriving at a destination (a chance to relax)
Every advance you gain a single point to any attribute

Combat
Push the enemy back (causing them to fall)
Assault (if attack hits, you can make a subsequent attack etc, fallback interrupts)
Morale (flee or stand and drool, test when stamina lost, casualties taken etc)
Stamina (ability to keep in the fight)
Ward (your hits don't do damage, but force enemy back from push or assault)
No hit points (a good damage role indicates a heavily damaged leg etc)
Three bad wounds and you are dead
Initiative of strikes is determined by length (initially) and speed (thereafter)
Fear, unless overcome by a courage roll, reduces all pools by a given number

Attributes
Each attribute (courage, wisdom, fellowship, agility, might) is 1d8+6
Dwarf (Might 11+, Agility 10-), Hobbit (Courage 11+, Might 10-), Elf (Wisdom 11+, Fellowship 10-), Human (raise best attribute by 1).  If a player wants to play a demi-human, she may set any offending scores to the requirement (no greater) and play that race.

Core Mechanic (Maybe)
Each hero has a pool of dice for each attribute equal to that attribute.  This pool can be used each setting to accomplish different things, and each use refreshes a single point (no matter how many were spent).  Players select which dice to keep and then the referee indicates the threshold.  If the player selected dice under the threshold, the attempt fumbles, otherwise the number of dice kept indicates the degree of success (how far the hero got).  Any natural 8 allows another die to be added to the roll.  Magic works similarly - the magician describes the spell and the dice indicate how far along it got (and perhaps if it even succeeded beyond its intention).  Wounds may indicate that you do not get your entire pool refreshed when you enter a new setting.  How long wounds take to heal (and what it takes) is entirely up to the referee.

The referee doesn't (normally) fumble, but his dice also do not (normally) explode (also, his thresholds should generally be a point or two tougher).

Agility tests to hit someone (normally need 1 success, unless they parry with an agility test).  Might tests to hurt someone (daggers are max 1 die from pool, bows and spears 2, longswords 3 and axes 4).  Armour (descriptive, not from a list) provides referee with ideas for threshold (apply weapon vs. armour if you like).  It's important to note that thresholds never stay the same from one strike to the next - the situation is always in motion and the referee is free to decide new thresholds whenever a roll is made.

A light wound kills minor foes and causes heroes to lose 1d8 dice from all attribute pools (down to 1).  Midi and more severe wounds cause the hero to lose 2d8 (or more) dice from all pools.  If the hero is down to 1 in every pool by a wound, she is incapacitated (perhaps she can crawl and speak, but not fight).  Three wounds (or being brought to 1 in every pool) will kill you, whichever happens second.  Wounds can also increase thresholds, as the referee sees fit.

Noclue wrote:
Some questions:

If humans get +1 to a stat,  what is the advantage of "rolling into" a different race?

The conflict resolution mechanic is interesting. However, who determines how many success dice are needed to fully succeed at a task? Is it completely up to the GM, so that they can always make it take one more success dice to achieve my goal than I roll?

chance.thirteen wrote: The guessing what the target number is interesting to me, but sounds built to halt the players in their tracks too easily. Not sure if there couldn't be a narrowing of the range, eg high or low difficulty indicating 5+ or 4-, or a set range of 2 or 3 difficulties max. Or perhaps a mechanism to reduce failed dice from fumbles to just negative successes.

Luke wrote: Why not have each setting consist of 1 to 8 obstacles?
When the players reach a setting, the GM can inform them, "There are three obstacles to overcome to get through Mirkwood."

If the players succeed at a test, then they move forward. If they fail, then an additional obstacle can be added to the setting or the GM can choose to let the group overcome the obstacle but the character is "damaged" -- reduce a pool in the NEXT setting.

With some guidelines like these, the players will have a better idea how to manage their resources.

Also, it doesn't seem like you need a separate combat system. If you really want a simple, fluid system, just use your basic resolution mechanic for combat. Perhaps offer one tweak: rather than setting an arbitrary difficulty, the GM rolls for your opponents (maybe even in the open).

Luke wrote: This is not true at all. Each episode in The Hobbit present a challenge through which the character demonstrates his worthiness for the greater task at hand or by which we come to learn something about the character or in which the character gains a piece of lore that will be needed further down the line in the story.

Whoops, ignore the above post, I copied my original notes instead of half the stuff I wanted to post.  Also, I'm not allowed to report my own posts, so I can't get it deleted.  Can someone else report it?  Thanks.  Here is my original post:

jerry wrote: 2. I notice that all of your examples invoke Middle Earth or Tolkienesque except combat. Can you describe potential Tolkienesque combat using your ideas?

Noclue wrote: Some questions:

If humans get +1 to a stat,  what is the advantage of "rolling into" a different race?

The conflict resolution mechanic is interesting. However, who determines how many success dice are needed to fully succeed at a task? Is it completely up to the GM, so that they can always make it take one more success dice to achieve my goal than I roll?

chance.thirteen wrote: The guessing what the target number is interesting to me, but sounds built to halt the players in their tracks too easily. Not sure if there couldn't be a narrowing of the range, eg high or low difficulty indicating 5+ or 4-, or a set range of 2 or 3 difficulties max. Or perhaps a mechanism to reduce failed dice from fumbles to just negative successes.

Luke wrote: Why not have each setting consist of 1 to 8 obstacles?
When the players reach a setting, the GM can inform them, "There are three obstacles to overcome to get through Mirkwood."

If the players succeed at a test, then they move forward. If they fail, then an additional obstacle can be added to the setting or the GM can choose to let the group overcome the obstacle but the character is "damaged" -- reduce a pool in the NEXT setting.

With some guidelines like these, the players will have a better idea how to manage their resources.

Also, it doesn't seem like you need a separate combat system. If you really want a simple, fluid system, just use your basic resolution mechanic for combat. Perhaps offer one tweak: rather than setting an arbitrary difficulty, the GM rolls for your opponents (maybe even in the open).

Luke wrote: This is not true at all. Each episode in The Hobbit present a challenge through which the character demonstrates his worthiness for the greater task at hand or by which we come to learn something about the character or in which the character gains a piece of lore that will be needed further down the line in the story.

You combat system is broken and has nothing to do with your premise of exploration or simulation of Tolkien's work. You're already getting bogged down in making an esoteric system balanced and interesting. It doesn't serve the design. Drop it.

-L

I absolutely love the potential in the "push as hard as you like, but if you try to push too hard too fast, you will fail" resolution mechanic. That said, I sense some difficulties. :)

7's are a good target number for hard tests (and for tests that the referee wants to either pass or fail without degrees of success, like picking a lock).

Ah interesting.  I do get the gist of reply #10, but the word "broken" seems to stick out in my mind (as it could mean a lot of things, and I'm not sure which meaning is intended).  My design goal and premise is to come up with something that works like stories do - outcome based, yet descriptive (whether in exploration, battle, social encounters etc).  There are no doubt many different approaches to this that would work excellently; I do suspect mine may be one way.

I personally completely eschew game balance (in favour of raw referee control, indeed "game balance" is a dirty word to me), and am not sure how my system would be seen as particularly balanced.  Certainly the referee is not bound to any rule in the game, and the different pools are just a tool to control pace and spotlight.

I do think a system should be interesting, however, and suggest a certain framework to play (a mantra in indie RPG's, I believe).  Groups of foes operating with a pool of dice seems interesting to me.  Heroes taking wounds that reduce their ability to manage later obstacles is interesting, to me.  A difference between "to hit" and "to damage" is interesting to me (imagine fighting Shelob - the boney limbs are easier to hit but harder to wound, the eyes are harder to hit but easier to wound).  Descriptive and narrative difficulty thresholds are interesting to me, and "overreaching" into a fumble is interesting to me.  All of these things seem to encourage the implied friction ("physics"?) in the novels decently (not great, but pretty workable).

However; the idea of weapon length, speed and damage is admittedly a throwback to AD&D, and I agree the game could do without those.  I never played AD&D much, but it's always intrigued me.

So, with the basic ideas of the system we talked about, how would you represent the fight at Balin's tomb?

I appreciate all the input a great deal!

@Guy:

Excellent stuff.  I think, Guy, you are coming up with the same problem Noclue had in Reply #2.  I intended from the start that the referee must spell out what 1 success would mean, what 2 would mean etc.  The threshold (the target number for each die) is the only thing that the referee and the situation merely imply.  So to "scout out the orc encampment", the referee would say "two or three successes would be needed to get the most information" or that "one success will infiltrate you into their territory, two will get you past the patrols, three will get you to the vantage point you want".

The difficulty threshold shouldn't be artificially lower for actions that imply degrees of success.  The hero scouting the camp happens to be an elf or a ranger, who does this stuff naturally, and all he really does is move undetected when the orcs aren't particularly looking for him.  That is why the threshold is low.  If the hero declares he wants to dress up like an orc to infiltrate their camp, convince their captain that he is actually the king of the orcs, and then demand they give him and his hobbit friends safe passage, the threshold would NOT be low at all, and multiple successes would be near impossible (and rightly so).

All I meant by suggesting a "degree of success" type action is that you don't need to make 100 different rolls to accomplish what is essentially one extended task.  Roll the dice once and let 'em ride.  Burning Wheel has a similar rule, I believe.

I think it's a fair assessment that I'm still figuring out how players will see these rules from their side.  With Frodo's situation (that the player wouldn't assume his blade could keep better dice than Boromir's), I think we're forgetting that the referee almost always must communicate the changing situation to the players (unless it is something that simply happens TO the heroes, like Frodo's chain shirt).  Thus, on Frodo's turn (after Boromir's blade clattered to the ground), the referee must say something like "Sting flashes violently for a moment, and hums in your hand as it points towards the big, ugly troll!".  The excitement is then when the player picks up on this - should he trust the referee?  Go for it!  The referee has indicated you have a once in a lifetime chance!

Evan wrote:
I personally completely eschew game balance (in favour of raw referee control, indeed "game balance" is a dirty word to me), and am not sure how my system would be seen as particularly balanced.  Certainly the referee is not bound to any rule in the game, and the different pools are just a tool to control pace and spotlight.

I do think a system should be interesting, however, and suggest a certain framework to play (a mantra in indie RPG's, I believe).  Groups of foes operating with a pool of dice seems interesting to me.  Heroes taking wounds that reduce their ability to manage later obstacles is interesting, to me.  A difference between "to hit" and "to damage" is interesting to me (imagine fighting Shelob - the boney limbs are easier to hit but harder to wound, the eyes are harder to hit but easier to wound).  Descriptive and narrative difficulty thresholds are interesting to me, and "overreaching" into a fumble is interesting to me.  All of these things seem to encourage the implied friction ("physics"?) in the novels decently (not great, but pretty workable).

Evan wrote: "My design goals are flexible, light gameplay that encourages the particularities of Tolkien's conception of adventure while remaining easy for the referee to control and direct."

Hehe, ok I'm going to think on your stuff and not answer it right away.  I will describe the battle in Chapter 5 (at Balin's tomb) though, as food for thought:

After reflecting on the last words of the doomed dwarven expedition, drumbeats are heard from the depth.  Either the players decide to make a stand here to delay the enemy (and count as an obstacle towards beating the "Moria" setting), or the referee suggests it and the players agree it would be fun.  Before the enemy arrives, the party bars the door against the enemy (generally preparing adequate defenses might require a Wisdom roll, or the referee might simply give it to the players for free if they are clever and descriptive - it depends on your group's style).

A strange, massive, scaly creature appears behind the door and bashes at it, poking his limbs through.  Boromir wants to make a might roll to scare the thing off; the referee says it may take a couple successes to make the troll disappear for the battle.  One success means the creature is warded for now, but will come back later in the battle.  Two successes and it will only come back after the battle.  Three successes and they won't have to deal with it again.  The referee describes it's thick scales again, and answers a question or two Boromir had, and he picks up 4 or 5 Might dice and rolls, describing how he swings to cleave a limb from the beast.  He reaches a bit too far and keeps some bad dice, resulting in the troll getting angry and Boromir dropping his notched sword.

The referee feels it would have been pretty cool to scare off the troll, and the players seem a little disappointed, so he throws them a bone, indicated to Frodo that Sting seems to want this creature's blood.  One last chance, the courageous hobbit flings himself at the monster for the same action, keeping two successful dice (I would have even let him roll Courage instead of Might to accomplish this action).  With a cheer, the troll limps off for the remainder of the battle.  But didn't the referee say he might be back afterwards...?

The party's defensive preparations worked - it repelled the troll, their foes best asset.  Now the doors fly open and combat begins.  The referee describes countless orcs on the other side of the door, and jots down "30 dice remaining" in his notes.  The referee clearly describes the orcs armour and various weaponry to the players, by quantity AND quality.  The order of combat begins initially with the best reach.  Some of the orcs have bows (100 foot reach), so he rolls a couple 2-dice missile attacks (with high thresholds for not having proper time to sight their targets) which clatter against the back of the room (no successes).  A couple orcs also have spears (6 foot reach), so together they make one 3-dice attack against Gimli (who wanted to be in the front), which hits.  Spears have a damage of 2, which means generally only two dice can be put into the Might test to injure Gimli.  The dice come up too low (remember, foes have slightly higher thresholds, since they cannot fumble) to pierce the dwarf's armour.  The referee notes that he used 9 dice, and the orc swarm has 21 remaining.

As the remaining orcs have scimitars (basically, short swords), the party goes next before the orcs can attack again (ties in reach or speed go to the heroes).  They hack and slay, dropping 13 dice off the orc swarm, reducing them to 8.  The referee narrates that 13 orcs have fallen to the combined attacks of the fellowship, but countless remain.  Finally, the orcs with scimitars get their turn.  Attacks are made on Strider, Boromir and Sam totalling 6 dice, but only Sam is hit.  The referee puts his last two dice into a roll to wound Sam, and comes up with one success, dropping him 2 (from a 1d8) dice from every attribute pool simultaneously.  With a scrape on his head, Sam will live, but now the orcs are out of dice, and they flee in dismay.

The referee didn't see this battle being over in one turn.  He wants this battle to be memorable and hard fought, and wants to reveal something interesting about Frodo's chain shirt as well.  He announces that a fierce orc chieftain arrives to rally the orcs, and describes as he dodges by Strider and Boromir to stab Frodo with a spear.  Stepping out of the normal combat mechanics, he simply rolls an impressive handful of dice to wound Frodo.  He rolls well, but nothing better than a 7, and declares Frodo miraculously unharmed after Strider cleaves the chieftain's head and the remaining foes again flee in dismay.  The players wonder at why Frodo's threshold was so high, and Gandalf is quoted as exclaiming that there must be "more to the hobbit than meets the eye".

Battle is over, Boromir opts to try to force the door closed behind and they flee to the next obstacle, while Gandalf opts to deal with the troll everyone had conveniently forgotten...

Alternatively, if the referee considers the mithril shirt a magic item, he could have given Frodo a small dice pool for the magic item to resist the spear point.  The threshold would be pretty low (the orc chieftain was strong, but his weapon wasn't magical or particularly well made), and each success reduces a degree from the wound.

To resolve the battle WITHOUT using any rules specific to combat, I suppose I'd tell the players that they would each have to roll attributes appropriate to the tactic they employ, and that we'd total all their successes to see if the foe was surmounted.  Legolas might use his archery (rolling some Agility dice), the hobbits might attempt various daring (if brash) attacks (rolling Courage), Strider may sing songs to encourage everyone (rolling Fellowship) and Boromir and Gimli might use their Might to fight.  Any fumbles would indicate a hero getting wounded (not necessarily the hero who fumbled), or something else bad happening.  The degree that they succeed would indicate how much more resistance they could expect from the Moria setting before advancing beyond it.

To me, this certainly works fine, but lacks some of the flavour I guess?  Combat in these books is actually kind of gruesome, not overly abstract (more gruesome than the recent Peter Jackson films, certainly!).

I can play that combat as you described in detail in just about any traditional RPG system. Why? Because the results don't actually result from use of the system. The results, as you describe them, are born from arbitrary decisions as made by the GM.

-L

To be fair, I love arbitrary decisions made by the "GM" :D

The second to last campaign I ran was Star Wars "Saga Edition" (a d20 game that resembled Dungeons & Dragons 3rd Edition).  That game heavily implied that the "GM" must play by the rules and indeed all the players knew all the rules.  When I tried to break a rule, I was called out on it by the players.  When I ran an adventure, several of the players made it obvious to me that they had read the adventure beforehand, and didn't care that they were being entirely obvious about it.  To put it bluntly, I'd rather slam my dick in a door than play that campaign again.

The last campaign I ran was Basic D&D using Castle Zagyg as a setting.  Everytime we sat down at the table, the players all knew I could drop a ball of lightning down on them and kill their characters for no reason.  They knew that there were no rules for diseases, but I could roll a bunch of arbitrary dice and give their characters the clap.  The players only knew about a quarter of the game's rules (even though I bought each of them their own book).  They fucking loved it :D

They trusted me to make the story interesting, to be entirely fair yet challenge them, and to secretly root for their success.  They trusted me to surprise them and captivate them.  They trusted me to bounce off their ideas and let them take the story where they wanted to take it.

Luke wrote: Stop, you're hurting my brain. If the referee is not bound to any rules in the game, why do we have a game text? Why don't we just have a referee whom we all plead to, intimidate, wheedle and perform for? DO NOT ANSWER THIS. THINK ABOUT IT.

Luke wrote: I can play that combat as you described in detail in just about any traditional RPG system. Why? Because the results don't actually result from use of the system. The results, as you describe them, are born from arbitrary decisions as made by the GM.

What I am trying to figure out here is, why bother? You have spend weeks working out the best set of rules you can dream up but if the last stage in the process is, "The GM then decides what happens" then all of your clever mechanics are irrelevant. Your players might as well ask you to interpret their success from some pig knuckles thrown in a bowl or the sound they make when they crunch some Doritos.

I like some of your ideas and you clearly have a good grasp on the source material but you are undermining it all with GM fiat.

Hi brian, I got your exact sentiment in a PM from someone else just now!  I think I've badly misrepresented myself, perhaps.

On an inconsequential note, I actually came up with this game in about 45 minutes.  It just happened about a week ago.  I haven't developed this idea at all, except for the conversations in this thread, and I certainly haven't playtested it!  I'm just fooling around with a quick and silly idea for a game.

Here is the response I gave in that PM, which may help me be more clear (or may engender a conversation that clarifies to me where I have become confused).

How each game approaches roleplaying is unique and different, but the suggestion is a gaming philosophy, not a strict modus operandi.  I think it's useful to shed the "unhelpful associations" that roleplaying has accrued from its original, core meaning.  The definition of the word roleplaying does not entail a pendantic texbook-style procedure.  At the core of the word "roleplaying" is a "play philosophy" that sends you in a new direction and opens your mind to a new method, while you adopt and personalize the new territory.

No one is arguing that game text doesn't contribute anything if the referee is allowed to do what he wants to personalise the play.  Take it to the extreme, would anyone argue that game text adds everything, and players and referee add very little?

Evan, the problem with GM Fiat is not that the GM is necessarily a tyrant, it is that it makes player choices meaningless and rolling dice merely suggestions to the GM. Please, please, please lose any notions of "fudging" dice to keep the story interesting or just allowing the GM to handwave based on his/her arbitrary reading of tea leaves. It does not make the narrative exciting and tense, it just makes the narrative annoying and a total waste of my time as a player. The GM might as well just tell me a story.

If you want to have the game focussed more on narrative input and less on tactical dice mechanics, that's cool. Allow the players in on all this narrating. Let them make meaningful choices and add meaningfully to the shared fiction.

The amount of GM fiat I'd allow is the degree of GM fiat I described in the above example of the fight for Balin's tomb.  IE the referee would pick 30 dice worth of orcs, set difficulty thresholds, occasionally decide to give the players a second chance (at scaring the troll) if it meant giving other characters some spotlight time, and sporadically throwing a new problem in their way (the sudden appearance of the orc chieftain).  How do these things make player decisions meaningless?

The game is a product of referee input, player input and dice.  No game is different.  And dice are just as arbitrary as referees or tea leaves can ever be.

Ok, I think I get it, and I may have been misunderstood.

Luke said my game was broken (which to me means one PC is more powerful than the other).  If a game is not broken, I understand it to have "game balance".  Luke also said I was trying to establish game balance (yet this is a narrativist game, by Ron Edwards' definition, and accordingly I don't give a crap about game balance).

I overreacted, and said something I probably shouldn't have said on these forums: the referee is not bound by any rules in the game.

Over on Dragonsfoot, this would be a perfectly acceptable sentence, because all it implies over there is that, on rare occasion, if the game mechanics do not permit the story to go in a way everyone at the table wants it to go, the GM is allowed to bend the rules.  99.9% of the time, GM fiat meets (and is thus tempered by) random chance, aka dice.

Luke has indicated that a true indie-RPG funnels narration through the mechanics.  If I intended GM fiat to play a huge role in the game, then it is true; the system for resolving narration mechanically in the game would be meaningless.  What I actually intend for this game is for drama (Ron Edwards' term for unregulated narration) to meet fortune (his term for dice and chance) squarely, and for these things to temper each other.  So YES, the orc happens to be clad in black mail from head to foot (and this fact is entirely GM fiat), but this has definite connection to in-game numbers.  The GM clearly communicates these values (without explicitly stating them) to allow the player both access to controlling the game mechanics (in which dice he knows he should keep) while remaining an obscuring veil that doesn't reduce the battle into a numbers game (well I know I have a 2 in 8 chance to hit, so I should use no more or fewer than 4 dice to hit).  If the GM wants to encourage a climactic ending to a long campaign by secretly lowering the (somewhat arbitrary) difficulty he had previously envisioned, in order to allow the hero to chop the ring finger from Sauron, then he could do that (even if the rules don't allow it).  Again, this should happen less than 1% of the time, and shouldn't be obvious.

After all, that's sort of what GM's do.  GM's want to participate too, and they want to make the game awesome for everyone.  They set up the basic framework of the plot, and if they bend the rules on rare occasion to give the players more decisions, they shouldn't be detested and spat upon.

  I am with you on the whole GM thing.  As a GM, I know there is that one time when the dice just don't create a good story.  And seriously, in a rules-heavy game like D&D, no one sticks to EVERY rule.  With each group there is a little bit of fudging.  I do agree though that with a more rules-lite game this becomes less of an issue as there becomes less need to fudge.
 
  However, I'm not posting to discuss this issue.  I'd just like to say your game looks really cool.  To fire your imagination as you fired mine, let me share a random cool idea that popped into my head while reading, take it or leave it.  In Tolkien, but LOTR especially, there are some 'dice pools' that can't be refreshed by just any place.  Frodo's wound from the Ringwraiths couldn't have been healed in Hobbiton.  Perhaps Boromir's inevitable betrayal was put off for a little while by the contemplation of a view with a glorious sunset. 
  Pools that 'carry over' despite going to new areas could make an interesting twist in storytelling, whether it's a deadly wound or the timer to an emotional bomb (in LOTR, Boromir's obsession with saving Gondor) that's about to go off.

Great stuff, Noah.  I think your idea mixes perfectly with Luke's idea of "some damage continuing to the next setting".  The Boromir thing definitely enhances this concept, so thanks!  I had been thinking of a way to encourage dissention, greed and cowardice at critical moments, but you've come to the same idea from a different angle.

I was actually about to post that my previous post was a cop out and a bunch of bullshit.  I use GM fiat all the time.  It depends entirely on your group whether this is acceptable form of narration (and making a game that fits your specific group's taste is the entire ethos of indie-RPG's).  I realised though that I very rarely use GM fiat to overrule the dice, and mostly use it to resolve an encounter WITHOUT dice.  In my Castle Zagyg game, the party came across three boggibears dragging a large, bloody sack through the woods.  The boggibears fixed them with blank stares as the freaked out PC's gave circled around them to continue on their way.  Once out of sight, the PC's heard arguing in a horrible language, then heavy feet beating towards their location.  The PC ran screaming out of the woods.  Everyone present told me it was one of the best encounters they had ever had in a game.  Did I use game mechanics to narrate whether the PC's beat the boogeymen in a speed test?  No, I just let them get away, and they remembered that encounter that much more sweetly.

Noclue, what you said about letting the players in on all this narrating has been stuck in my mind.  I'd love to hear how you might do it, since group narration is one of my favourite things.  If it didn't clash with the previously discussed mechanics, and fit a fantasy literature feel, I'd definitely ratify it.

Evan wrote:
The amount of GM fiat I'd allow is the degree of GM fiat I described in the above example of the fight for Balin's tomb.  IE the referee would pick 30 dice worth of orcs, set difficulty thresholds, occasionally decide to give the players a second chance (at scaring the troll) if it meant giving other characters some spotlight time, and sporadically throwing a new problem in their way (the sudden appearance of the orc chieftain).  How do these things make player decisions meaningless?

The game is a product of referee input, player input and dice.  No game is different.  And dice are just as arbitrary as referees or tea leaves can ever be.

Back when I used to run D&D, a player would declare some action and I'd usually either 1) let them do it automatically or 2) ask them to roll any shaped die they wanted.  Only after seeing the die result (higher is better), I'd decide if they make it.  I think I'm taking some notes from how I ran D&D for this game as well.

Certainly the referee may make up the difficulty threshold after he sees the dice rolled (although this should not be obvious to the players, who might feel it unfair). 

The referee doesn't (normally) fumble, but his dice also do not (normally) explode (also, his thresholds should generally be a point or two tougher).

…the referee is free to decide new thresholds whenever a roll is made.

Since the referee creates the thresholds, he must fairly describe the situation to give the players an idea of what they may need.  He can also use this to "fudge" die rolls, accepting rolls that didn't meet his secret threshold.  The idea is to control pace and keep everything exciting and tense.

Certainly the referee is not bound to any rule in the game, and the different pools are just a tool to control pace and spotlight.

To be fair, I love arbitrary decisions made by the "GM" :D

Evan wrote:
The last campaign I ran was Basic D&D using Castle Zagyg as a setting.  Everytime we sat down at the table, the players all knew I could drop a ball of lightning down on them and kill their characters for no reason.  They knew that there were no rules for diseases, but I could roll a bunch of arbitrary dice and give their characters the clap.  The players only knew about a quarter of the game's rules (even though I bought each of them their own book).  They fucking loved it :D

They trusted me to make the story interesting, to be entirely fair yet challenge them, and to secretly root for their success.  They trusted me to surprise them and captivate them.  They trusted me to bounce off their ideas and let them take the story where they wanted to take it.

Well GM fiat isn't really inherit in the game mechanics I was describing, it is just how I tend to run games.  If the GM hasn't made up his mind yet when the player rolls the dice, it's OK if he makes a final decision on the difficulty threshold after seeing the dice.  That's way more "well, that seems like it should make it" or "wow, no, not even close!" than GM fiat.

GM fiat is taking narration and resolution away from the game mechanics and into his hands.  There's no ulterior motive to seeing dice and interpreting them.

You're absolutely right that I'd never just kill characters off for no reason (and expect the players to like it).  I just wanted to illustrate that the players KNEW GM fiat happened on occasion and they still ate the game up (I just IM'd a couple of them and they said GM fiat made the game better).  That's actually why I remembered the boggibears reference above.

And there's totally player choice, because I allow player fiat as often as I allow GM fiat.  It's basically called "resolving stuff without crunchy game mechanics" and it works well.

Evan wrote:
Noclue, what you said about letting the players in on all this narrating has been stuck in my mind.  I'd love to hear how you might do it, since group narration is one of my favourite things.  If it didn't clash with the previously discussed mechanics, and fit a fantasy literature feel, I'd definitely ratify it.

Evan wrote:
To resolve the battle WITHOUT using any rules specific to combat, I suppose I'd tell the players that they would each have to roll attributes appropriate to the tactic they employ, and that we'd total all their successes to see if the foe was surmounted.  Legolas might use his archery (rolling some Agility dice), the hobbits might attempt various daring (if brash) attacks (rolling Courage), Strider may sing songs to encourage everyone (rolling Fellowship) and Boromir and Gimli might use their Might to fight.  Any fumbles would indicate a hero getting wounded (not necessarily the hero who fumbled), or something else bad happening.  The degree that they succeed would indicate how much more resistance they could expect from the Moria setting before advancing beyond it.

To clarify my view on GM fiat. I have no problem with GM fiat that does not render player choices as irrelevant. No game is going to be completely devoid of fiat, nor do I think that is a worthy goal. Empowering players to contribute meaningfully in a collaborative activity does seem to be a worthy goal.

define fiat, no, really.

http://upload.wikimedia.org/wikipedia/commons/7/7b/Maserati_Gran_Turismo_red2.jpg

7VII7 wrote:
define fiat, no, really.

Evan,

I love the idea of guessing at the number the GM selected, but the game would be way more interesting (not to mention way more of a game) if the GM were bound to the decision he made early on.

You want to sneak into the orc camp. I decide that's a difficulty 4. I write this down on a piece of scrap paper, turn it over, and put it in front of the table. There's the difficulty, sitting right there, and no one knows. Now it's a fair game.

Even better, I have a bunch of tiles, say four 2s, four 3s [...] and one 8. I have resources now I have to spend secretly.

Alternatively, I can set the difficulties how I want (still putting it face down in front of everyone), but now greater difficulty means greater reward.

--

Don't use any of these specific suggestions, but it's very easy to make this more of a game, and more fun. You have a really cool idea here with the guessing of the GM's target number. I would hate to see it used on something where the fact that the GM doesn't ever have to abide by the rules, and thus, in my opinion, wasted.

Robert wrote:
Evan,

I love the idea of guessing at the number the GM selected, but the game would be way more interesting (not to mention way more of a game) if the GM were bound to the decision he made early on.

You want to sneak into the orc camp. I decide that's a difficulty 4. I write this down on a piece of scrap paper, turn it over, and put it in front of the table. There's the difficulty, sitting right there, and no one knows. Now it's a fair game.

Even better, I have a bunch of tiles, say four 2s, four 3s [...] and one 8. I have resources now I have to spend secretly.

Alternatively, I can set the difficulties how I want (still putting it face down in front of everyone), but now greater difficulty means greater reward.

--

Don't use any of these specific suggestions, but it's very easy to make this more of a game, and more fun. You have a really cool idea here with the guessing of the GM's target number. I would hate to see it used on something where the fact that the GM doesn't ever have to abide by the rules, and thus, in my opinion, wasted.

Evan wrote:
Well GM fiat isn't really inherit in the game mechanics I was describing, it is just how I tend to run games.  If the GM hasn't made up his mind yet when the player rolls the dice, it's OK if he makes a final decision on the difficulty threshold after seeing the dice.  That's way more "well, that seems like it should make it" or "wow, no, not even close!" than GM fiat.

Evan wrote:
GM fiat is taking narration and resolution away from the game mechanics and into his hands.  There's no ulterior motive to seeing dice and interpreting them.
...
And there's totally player choice, because I allow player fiat as often as I allow GM fiat.  It's basically called "resolving stuff without crunchy game mechanics" and it works well.

Evan,
Don't attribute false qualities to my words. I never said anything about "true indie." I'm talking about basic, basic, basic roleplaying game design here. Elements that are part of every game, traditional or otherwise.

You're being defensive and you're misinterpreting terms in order to build a huffy little moat around your outmoded ideas. To wit, the definition of GM Fiat is not "any decision that the GM makes." GM Fiat involves decisions by the GM that break, ignore or obviate textual rules. These decisions are made in service of the GM's story -- not the story being created collectively by all players' interactions with the game.

A game is a system in which participants make meaningful decisions, receive feedback and derive a consistent outcome. There are always meta decisions made by players as they play a game. The most basic meta decision is the assent to play the game by the rules. But assuming or allowing a player to make a meta decision that invalidates the system, invalidates all of the other participants' decisions. Why play if one player can just say, "No, it happens like this" over all of the other players. Oh really? Then why don't you just tell me how it turns out? I'll listen and sip my lemonade.

You posted here looking for feedback on a game idea. It's not a bad game idea. It needs to be tightened up. In the course of getting feedback, you seem to be saying that you don't care much about criticism or challenges to your assumptions. Well, Dragonsfoot may be a bunch of "that's great!" softies, but this community is built on constructive criticism and the challenge of ideas. If you continue to participate here, then expect this cycle to continue.

And to return to my criticism -- if your game about Tolkien's fiction requires GM fiat in order to recreate Frodo's miraculous survival in Balin's tomb, then the game is broken. This is a crucial and exciting moment in the stories that can be recreated in a game situation without putting all of the power in one player's hand and expecting him to sort out the details and make it interesting.

Personally, if I'm going to put all that power into one person's hand -- if I'm going to rely on one person to make the fight at Balin's tomb exciting -- it's going to be Tolkien. He's a better storyteller than any GM. Why shouldn't I simply reread the work?

You can and should make a mechanic that allows the GM to increase the threat level -- tossing in or recycling dangerous foes, for example -- and a resource that the players can manipulate in order to get the better of the situation -- Frodo revealing his shirt. In the context of the game, Frodo's player might never have even HAD a mithril shirt up to that point, but perhaps by spending hero points at that moment to save his butt, the player invests Frodo with a new quality and reveals something special about the character. Can you see the difference between my example and yours? I'm relying on all interactions with the system -- GM: Orc Chieftain! Boo! Spear! Agh!; Frodo Player: Argh! Spend Points! Save my butt! I have a mithril shirt to protect me! -- to create interesting moments that ultimately woven together into a story.

Since the game is about exploration, perhaps exploring your character and revealing new elements about your character is just as vital and mechanical as exploring new settings.

-Luke

Difficulty values being a GM resource is definitely a cool idea, Robert, and something I've been thinking about since I read your post this morning.  I would love to integrate it.  I do have one slight hang-up, though...

If the GM gave himself a finite number of difficulties (for a single setting), would it become more difficult for the GM to direct the play later in that setting?  For instance, what if he accidentally made it way too hard, or way too easy?  The heroes in Tolkien stories rarely flat-out fail; they usually get to push onward.  What if a setting had too many difficult numbers, and the players' dice just can't make it happen.  I don't want to put the GM in a situation where he has to fudge everything so they can advance to the next stage.

I agree though that difficulties should be basically set in stone.  Although I said earlier that difficulties should shift around a bit, I was thinking more +/- 1 point, not a significant change at all (and certainly not to take the narrative AWAY from player control, but to bring it more TOWARDS player control, while still remaining the group's idea of fun; appropriate levels of challenge, risk, reward and victory).  The biggest issue, one of the very most important things I am after, is that I want to describe difficulty values through narrative.  I want mechanics and random number generators to be working behind the scene, but the part the players mostly interact with is words and choices.

I agree with Noclue though, difficulty as a spendable resource looks like it would be great.

brianbloodaxe wrote: But with the system described above having the GM look at the dice and then set a threshold is equivalent to saying, "Do I want them to succeed really well or fumble really badly?" To my mind that is the worst decision to leave to fiat.

brianbloodaxe wrote: Please don't get me wrong, I like what you are doing here. I just think that you are trying to achieve two contradictory goals with one game.

Luke wrote: You can and should make a mechanic that allows the GM to increase the threat level -- tossing in or recycling dangerous foes, for example -- and a resource that the players can manipulate in order to get the better of the situation -- Frodo revealing his shirt. In the context of the game, Frodo's player might never have even HAD a mithril shirt up to that point, but perhaps by spending hero points at that moment to save his butt, the player invests Frodo with a new quality and reveals something special about the character. Can you see the difference between my example and yours? I'm relying on all interactions with the system -- GM: Orc Chieftain! Boo! Spear! Agh!; Frodo Player: Argh! Spend Points! Save my butt! I have a mithril shirt to protect me! -- to create interesting moments that ultimately woven together into a story.

Since the game is about exploration, perhaps exploring your character and revealing new elements about your character is just as vital and mechanical as exploring new settings.

Evan, I'm very glad you're reconsidering.

If you don't want horrible failures to be frequent, or easily possible, construct the system such that they aren't and it isn't. Or, steal from Luke and have as an up-front part of the conflict system that the GM is free to take a failure and turn it into success with dire consequences or special conditions.

I repeat my earlier statment, define fiat, I can't help anybody IF I DON'T KNOW WHAT THE HELL YOU'RE TALKING ABOUT.

It would certainly help me if you screamed your response in all caps, too.

Well excuse me for getting pissed off when I ask a simple and valid question that probbably has a simple answer that would take mere moments to answer and somebody decides to try and be funny by posting a link that has absolutely nothing to do with the question.

Evan wrote: GM fiat is taking narration and resolution away from the game mechanics and into his hands.

Ok, that's all I wanted, thank you. Now I haven't been paying that much attention to this thread so I don't know if it's been suggested before or not but purpose you should implement a mechanic such that failure isn't so much actually failure but things not going the way you want them to. I dunno, does that make any sense at all?

7VII7 wrote:
Now I haven't been paying that much attention to this thread...

You can have GM fiat with breaking or obviating the rules.  "Is there a fire extinguisher in the hall" is an example of the sort of question players ask, and which is answered by GM fiat.  That is inevitable unless absolutely everything is specified.  GM fiat which does overpower the rules is an application of that general facility which, I agree, undermines a a sizable chunk of the point of play.

I also don;t think that its safe to assume that Tolkien would necessarily be a better storyteller than any given GM.  Sure, I'll thats likely, but its not impossible to have a GM withb a real talent for storytelling, who uses the medium with skill to conjure powerful and images in the minds of the players.  And in fact, I prefer that sort of thing, which is why I'm basically uninterested in systems that feature significant player "contribution" to story.

John Wick makes an interesting point about the the similarity between the GM and the shaman by reference to heroquesting; available on YouTube.

All of which is to say: it is not quite as cut and dried as system-based contribution contrasted with systemless fiat.

Evan wrote: I suspect we have different definitions (which is why I said early on that I didn't know what you meant by game balance and broken).  I see at least the word "broken" defined in your recent post, now.  I suspect we have different views on GM fiat too, which is fine.  I've seen it put to use in involving players more directly in the narration.  I've seen it used to deny players access to narration (by obviating the rules by which they normally have access to the narrative).  I don't argue that the second method is acceptable at the table, and whether the first method is acceptable really depends on your group.

There are, of course, meta-rules on interaction beyond the written rules of the game.  If the players and the GM put their heads together and ignore the rules for a moment for a noble cause (to twist the story in an interesting way, perhaps), then they are all still playing the game.  We are now just talking about some unwritten club rules that (let's assume) everyone understands and accepts, that engender good play and run parallel to the mechanical determination the written rules of the game offer.

Part of the problem is that you have all these ideas in your head, and we only see the ones you write down. This is a summary of what I'm seeing in your examples (mainly the initial post and the Balin's tomb example) of how play progresses:

The adventure progresses through destinations. Each destination is a progression through obstacles. The destinations are set ahead of time by the GM. The GM is in control of describing the destination and its reaction to their choices; at certain points in each destination, the players can choose (sometimes at the GM's suggestion, but it's still the player's choice) that they would like this part of the GM's description to become one of the obstacles they need to pass through to progress through this destination. At this point, the characters are in danger of bad things happening to them, such as death or injury in combat-oriented obstacles.

Obstacles must present a minimum level of difficulty and/or danger. If it turns out that the GM fails to give the obstacle the needed difficulty/danger, the obstacle's difficulty/danger will need to be increased. (Or it won't count as an obstacle? Aside: could this be done with player buy-in just as the obstacle itself was, rather than by the GM secretly changing the encounter?).

Is that (minus the aside) a fair summary of how the Balin's tomb fight came to be part of the game, and why it was escalated?

I'm a huge Tolkien fan/nerd and love this system. I would love to use a similar mechanic for other games. I don't like the combat system but it looks like your discarding that. I like the idea of not getting experience from combat but what about getting it from "self-discovery" type stuff or maybe getting some bonus from lore scraps or from working together as a fellowship or something like that. Somethig like what we see hapening in the books. Also for combat. Maybe some kind of mechanical effect of description. Something to encourage describing stuff in a Tolkien-esque way (maybe not as long-winded but still) Same thing goes for skill challenges. I know this is all rough but I forgot most of what I was gonna say while reading your stuff. By the way I think this base mechanic (minus the fellowship stuff) would work great for an Amber game.

7VII7, I think that's not a bad way at all to think of fumbles.  Flat-out failure is bad way to handle narrative RPG's, so your idea is right on the money, I think.

Contracycle, I think you're right about everything you said.

Luke, thanks for the link; I've heard good things about V. Baker and I'm looking forward to checking that out.

jerry wrote: Part of the problem is that you have all these ideas in your head, and we only see the ones you write down. This is a summary of what I'm seeing in your examples (mainly the initial post and the Balin's tomb example) of how play progresses:

The adventure progresses through destinations. Each destination is a progression through obstacles. The destinations are set ahead of time by the GM. The GM is in control of describing the destination and its reaction to their choices; at certain points in each destination, the players can choose (sometimes at the GM's suggestion, but it's still the player's choice) that they would like this part of the GM's description to become one of the obstacles they need to pass through to progress through this destination. At this point, the characters are in danger of bad things happening to them, such as death or injury in combat-oriented obstacles.

jerry wrote: Obstacles must present a minimum level of difficulty and/or danger. If it turns out that the GM fails to give the obstacle the needed difficulty/danger, the obstacle's difficulty/danger will need to be increased. (Or it won't count as an obstacle? Aside: could this be done with player buy-in just as the obstacle itself was, rather than by the GM secretly changing the encounter?).

Is that (minus the aside) a fair summary of how the Balin's tomb fight came to be part of the game, and why it was escalated?