Friday, May 23, 2014

Procedure vs Consensus in Mechanic Assignment

+Brendan S posted a brilliant article about Proceduralism.  I'm going to talk about it, so if you haven't read it, you should probably read that first.

Basically, what (I think) Brendan's is talking about is how task resolution is handled.  (Although, my essay may be a little flawed, since I think we're talking about two slightly different things.)

In a procedural mechanic, a player says "I want to do X." and the success/failure/effects of that attempt are resolved via procedure listed in the book.

The opposite of a procedural mechanic, then, is when a player says "I want to do X." and the success/failure/effects of that attempt are resolved via player consensus.

An example of a mechanic = "If 2d6 is higher than the monster's morale, the monster will flee".  If this mechanic came from the rulebook, it was arrived at via procedure. If this mechanic was something your group decided (or DM) invented five minutes ago, then it was arrived at via consensus.

Consensus and procedure are two different ways of getting a mechanic to resolve some action. 

I. What I Mean by Player Consensus

 Consensus is what we use when there is no listed procedure.

Sometimes we used consensus happens because the listed procedure is inapplicable.  ("Look, I just don't think that the grappling rules were meant to handle wrestling 28 squirrels.")

Sometimes we use consensus because parts of the game world are undefined.  ("Unobtanium is a metal.  All metals are conductive, right?")

Sometimes the listed procedure sucks, or it isn't appropriate for the kind of game that we want to have.

DM fiat is a type of consensus.

After all, DM is a democratically elected position.  He wields no more power than what the players agree to give him. If your DM bungles his authority, impeach 'em.

The whole table might discuss the conductivity of unobtanium, giving special authority to the materials engineer in the group, and then arrive on a ruling together (This is open discussion).  Or the DM might just say that unobtanium is a superconductor at room tempurature, duh.  (This is DM fiat).  Or more likely, a mix of fiat and discussion will occur, as the DM asks the engineer what he thinks before spitting out a ruling.  And in the end, everyone at the table is (hopefully) satisfied, because consensus.

When we talk about procedure vs consensus, we are not necessarily talking about DMs.  You can have a DM-less game run by strict procedural mechanics (40K) or by group consensus (Microscope).  And you can have a DM'ed game according to either method.

II. What Consensus is Not

Consensus doesn't mean creative, storytelling freedom.  You can have storytelling freedom in both models.  The difference is lies in mechanical freedom.

I can elaborate.

Nearly every game system has a mechanism for attacking an enemy and then determining how much closer the enemy is to being KO'd.  Those examples are trivial.

But let's look at how consensus and procedure-based mechanics handle players who are trying to do something really weird.

Consensus-based example
P: "I want to feign dying and, while falling over, knock the trashcan down the shaft, where it will hopefully smash the glass roof of the reactor core."
DM: "Hmm.  Let me think about out how likely the thugs are to see through your fake dying.  I also need to think about out if a trashcan filled with unobtanium is heavy enough to break the glass (I have no idea how much that stuff weighs).  Then the thugs can take their turn as normal."
P: "Okay."

Procedure-based example
P: "I want to feign dying and, while falling over, knock the trashcan down the shaft, where it will hopefully smash the glass roof of the reactor core."
DM: "Hmm.  That's not one of the Gambits listed in this Environment Scene.  So just roll all of your Daring dice without any modifiers to see if you can do it.  Then the thugs will take their turn as normal."
P: "Okay."

The choice between consensus and procedure is not (necessarily) a choice between freedom to attempt whatever you want or not.

But in the last example, the procedure-based mechanics were broad enough to allow wide interpretations of player's actions.  That's not always the case.  4th edition D&D, for example, had combat abilities that were only defined in a combat context.  So even though the cleric could fire lasers at-will, all day long, he couldn't use them to destroy a door.  This is an example of a procedure-based system with very clear boundaries, and doesn't tolerate actions outside those boundaries.

III. The Interesting Part of This Essay

So what we're actually comparing here is how task resolution mechanics are arrived at: whether they are pulled from the book or reached by agreement (discussion, fiat, or a mixture).

Fact: Most TTRPGs use a mixture of both procedure and consensus.

Procedural mechanics come out of the rulebook, and they are applied like boilerplate to all applicable instances.  If your system uses d20 + attack bonus vs AC to resolve the accuracy of an attack, you will use exactly that system whenever someone swings a sword at someone else.  Consensus is reached at the table.

When consensus is useful.

Consensus is useful for flexibly generating appropriate mechanics.  Yes, obviously games that allow for all sorts of rulings will be very flexible, story-wise.  If you get turned into an orca and want to know if you can throw a dead penguin onto the lich's boat, an appropriately talented DM will be able to give you a % chance that seems fair to (hopefully) everyone.  This is a trivial conclusion.

What I mean by "appropriate mechanics" refers back to the story-flexible but mechanically-rigid example in section II.  Remember the "roll all of your Daring dice" line?  You can have a system that allows for flexible storytelling (you can do anything you want) but still has a strict procedure for determining success (no matter what you attempt, it's going to involve rolling Daring dice against the same DC).  Because that system achieves flexibility+proceduralism via dissociated mechanics.

But what I mean here is that consensus allows the table to device mechanics that are both story-flexible and mechanically-appropriate.  For example, if my PC was mind-controlling an pangolin with my love, I should be able to say to my DM "I should be able to use Charisma for the pangolin's Will save, instead of the pangolin's wisdom, since it's my love that's motivating it more than anything else." And the DM might say yes.

Because this is the freedom of consensus-arrived mechanics.  The DM can assign combustion temperatures to gnomes.  The DM can invent a way to resolve climbing attempts that involves all of your stats (except Cha).  And the DM can do this on the fly.  It might not be elegant, but it resolves the action in a way that uses the numbers on your character sheet in an appropriate manner.

Consensus-based mechanics are useful for resolving actions that are too nuanced or complex for codified rules.  This is why social interaction gets defined rules less often than combat (and when social interaction does get well-codified rules, there are many complaints).

Social interaction is the big one here (many hard-to-define factors interacting in a very complex way), but there are plenty of other examples.  If you have a dragon that's just been evolved into a cerebral meta-dragon from the future and you are trying to seduce it by dancing in binary, most systems will not be able to generate a procedure for resolving that action.  It's probably going to involve consensus somewhere in there.

When procedure is useful.

Procedure-based mechanics are useful when the players can't be depended on to reach a good consensus.  Rules for player vs player duels should be strictly codified, for example, so that there won't be any whining afterward that the rulings were unfair.  This is also the case for things that could be argued about forever, such as penalties from sleep deprivation.

Procedure-based mechanics are useful when the resolution absolutely has to be fair, such as when the stakes are high.  When the stakes are high, players are uncomfortable leaving things in the hand of the DM (or in group discussion, whatever).  If my PC attempts something and dies, I will be a lot more satisfied if I know that my PC died as a result of an unlucky roll than something the DM pulled out of his ass.

This is why combat is D&D is so strictly procedural.  Roll initiative, attack rolls, damage rolls, save vs sleep.  In combat, there is little room to discuss the circumstance, because the entire circumstance is already fully described.

Correspondingly, it might also be very difficult to practice ventriloquism while drinking a glass of water, but because the stakes for ventriloquism are so low, there is little motivation to codify rules for it.  Even in a ventriloquism-heavy campaign, you are unlikely to see anyone develop rules for ventriloquism unless the stakes for ventriloquism are high.

People are also fond of saying that procedural mechanics give meaning to a game.  Because when a DM announces "okay, it's been a while, and your torch is about to go out" it feels (and is) arbitrary.  Similarly, your DM announcing that torches last for 37 minutes isn't much better.  It makes the victories feel cheaper when you know the DM could just change the rules in order to give you an edge.  And it makes the losses feel unfair for the same reason.  This is what Gary is talking about when he says "YOU CANNOT HAVE A MEANINGFUL CAMPAIGN IF STRICT TIME RECORDS ARE NOT KEPT."

Fairness can also be had if the DM arrives at a number by fiat but then (a) tells the player the odds and (b) gives them a chance to decide (e.g. "You have a 35% chance to pull the wings off the giant beetle if you try.  Do you still want to do that?").  Fairness, but not meaningfulness.  Does that make sense?

IV. Caveats

This is not a strict dichotomy.  Halfway between consensus and procedure there is still a way of arriving at a mechanic, e.g. adapting a procedure to a similar task.

Automatic success/failure is still derived via procedure or consensus.  We use mechanics to turn the numbers on our character sheet into a certain probability of success.  Sometimes that probability is 0% or 100%.  Whether that 0/100% was arrived at by book-derived method or DM fiat doesn't matter.

Some people are simply more comfortable with one method of resolution than the other.  Some people don't like to see the DM (or the other players) inventing rules whole cloth while the game is underway.  How can they be trusted to be impartial when they all have vested interests?  (Trust and authority will be discussed in the last section).  These people will express a preference for procedure.

Other people don't rules.  They just don't.  They like feeling like the game is organic and reactive.  They enjoy arguing for advantages, or discussing how loud the music has to be before it will affect an arrow's accuracy.  These people will express a preference for consensus.

Neither method is necessarily faster than the other (but if I had to pick one, I'd say that procedural can be faster).  Procedural play can be slowed down by looking up rules (but those eventually get learned. . .)  Consensus play can be slowed down by discussing rulings ("do onions make beholders cry?") or by enumerating options ("is there any way I can get the ogre to stand closer to the fire?").

Yes, it's very easy to point at a game like 4th edition and say that the combat takes so long because the game is very procedural.  But that's not true.  Huge HP pools on bad guys is not an innate characteristic of procedural games.  Calculating many fiddly modifiers is not an innate characteristic, either.  When it's a player's turn in 4e, they can quickly see how many options they have and calculate exactly the effect that their choices will have.  Both of those things allow for fast decision-making.

Compare that to the average OSR game, where a ton of time is spent figuring out your options and trying to figure out the consequences of each.  (Personally, I think that's the fun part, but there are people who differ.)

V. Implications for Game Design

Different games use different amounts of consensus and procedure in their design.  Moreover, they use consensus and procedure in different parts of the same game.  Some games have procedural social interaction and consensual combat.  Others are the inverse.  So what does it mean when we make a certain mode of gameplay procedural or consensus-based?  What gameplay can we expect?  What kind of players will that attract?

Look back at section III.

When a game designer chooses to make a gameplay mode procedural, they are usually saying "this is a part of the game where the stakes are high".  Combat is pretty much always procedural, with lots of mechanics for hitting and caveats for attempting trips, for example.  But things like tracking rations and torches aren't always.

So when a game omits or marginalizes rules for how long a torch burns, this is what they're really saying:

  • You'll never have high stakes (life and death) come from the torchy parts of the game.  (The torchy parts of the game will not be the ones that kill you.)
  • Your sense of victory or defeat will not come from the torchy parts of the game.  (The torchy parts of the game don't need to be meaningful to be enjoyable.)
  • The game does not lie within the torchy parts.
This is fair.  Most people have fun worrying about whether or not the troll will kill them on its next hit.  Not everyone has fun worrying about whether or not they'll all die next turn when their torch goes out.  (Fucking grues!)

So when you're designing your heartbreaker, think about where the game is.  Is the game just the combats, while everything else just flavor and setting?  Or is the game combat + exploration?  Do you really need so many rules for tracking how long rations last in different types of hexes?

See also: Encumbrance, Tracking Food, Tracking Light, Travel Speeds, Crafting, Sanity

So the response to "YOU CANNOT HAVE A MEANINGFUL CAMPAIGN IF STRICT TIME RECORDS ARE NOT KEPT" should be "You can still have a meaningful campaign without strict time records, but the meaningfulness will not extend to tracking food, light, or travel speeds, since those have become more arbitrary."  

Here's a test.  Killing a dragon is obviously a meaningful feat (the players made lots of choices, rolled lots of dice, they could have died).  In your game, do you want crossing the subterranean desert to be a similar challenge, with lots of choices, dice rolling, and significant risk of death?  I mean, the DM could just say "You guys enter the the Umbral Desert.  It takes you 3 weeks, and you almost die from thirst, but you found an oasis.  Now you're out of the desert."  Do you really want to spend a session managing unruly camels and counting loaves of drow bread?  Because your ruleset will attract the kind of people who's ideas of fun mesh with yours.

I understand the appeal of striving for completeness, or wanting to emulate a certain type of game, but according to my calculations, you don't want rules for things that aren't high stakes.

My calculations:

procedures --> high-stakes --> tension --> fun via meaningfulness

(That's not 100% accurate, because fun comes from more places than just tension, but whatever.  This is meant to fuel discussion.)

VI. Implications for Dungeon Masters

The last section applies to you, too.

If your game uses sanity rules, but your PCs have no chance to go insane OR if the penalties for insanity are minor/easy to remove, stop using sanity rules.  There's little reason to keep using them.  It's just a bookkeeping burden.  Just arbitrate sanity when it comes up.  Ad hoc it.

Likewise, if a party is doing short journeys into a shallow dungeon and there is no chance that they'll be down there long enough to run out of their torches, stop tracking.

If your party has 8 bags of holding, drop encumbrance.

The mechanics for these things are better off being generated via consensus.  They can be more flexible and applicable that way.

(This section isn't 100% true either, because maybe you have fun knowing that your movement rates (300 lbs on a light arabian horse across a steppe prairie) is as historically accurate as possible.  You get the point.  Just think about what it means before you include detailed travel speed rules to your game.)

VII. Criticism

But Arnold, codified rules and pre-established procedures insulate me from bad DMs!  If we don't have well-established rules for a wide variety of situations and actions printed in the book, I could end up playing in a game that is way too hard, way too easy, or worse.  Don't we need to maximize procedure?

Why yes, Mr. Straw.  The more the rules cover every possible contingency, the less impact the DM has on the game.  This does insulate you from bad DMs (and good DMs with terrible hangovers).

If the procedural sections of the game had 100% coverage, the only task left to the DM would be to track unrevealed knowledge (where are the ninjas hidden), do funny voices, and possibly handle enemy AI.  You could practically DM yourself!

But at that point, you're merely playing a laggy video game with poor voice acting occasional glitches.  (Good AI though).  Perhaps you are only there for the social aspect, but are you sure you wouldn't be happier playing a video game instead?

But I digress.  Let's talk about how we arrive at mechanics.

You talk about insulating yourself from bad DMs, but you are also insulating yourself from good ones.
No TTRPG has procedures that cover 100% of all possible circumstances.  Sooner or later you're going to end up going off the edge of the map.  A good DM can craft a mechanic that is tailor-fit to your attempted action.  Are you trying to wrestle laser elementals inside your dreams?  Your DM can come up with a mechanism that is customized to the situation, so that your character sheet can be applied to the situation in the best possible way.

Games that depend on procedural mechanics (4e, GURPS) have an incentive to "keep you on the map".  They strive to minimize consensus-derived mechanics and DM fiat.  GURPS does this by having rules for everything.  4e does this by limiting your actions.

Also, do you really want to rely on procedural mechanics for everything?  The big one is social mechanics.

It's not fair to say that procedural = rollplaying = "I roll Diplomacy" and consensus = roleplaying = 3 minutes of silly elf voices.  Remember that we are talking about how the mechanics are arrived at, not what the mechanics are.  You could have a game where no one wants to roleplay and then the DM decides (a consensus method) that this situation is best decided with a Diplomacy roll.  Inversely, you could have a whole bunch of roleplaying, and then the DM consults his book and announces (a procedural method) that party has succeeded since they said the three magic words "princess", "love", and "please".

It is fair to say that procedural methods don't leave any room for DM interpretation.  Without any DM interpretation, you are bound by the results of the mechanic presented in the book, without any bonuses or penalties that the DM might decide are appropriate (or the whole table can decide, if you trust them).  Which is a pity, really, since roleplaying is something that is so easy to do at the table and can yield such good common-sense results.  Common-sense results that are usually so much more satisfying than whatever the procedural mechanic yields.

Plus, by my prior criteria in section III, social interaction is a good fit for a DM fiat system since the stakes are rarely as high as life or death.  Sure, you can offend the goblin king with your roleplay.  But the DM won't say "You offended the goblin king.  His guards kill you.  Give me your character sheet.".  The DM will say "Roll for initiative" and then you'll be moving into combat, the gameplay mode that everyone wants to be procedural.

But Arnold, flexibility and freedom is the one thing that makes D&D so magical!  It's the one thing that tabletop games can so better than video games and it's so rewarding!  Shouldn't we be maximizing that?

Yes, Mr. Straw, the flexibility and freedom to attempt literally anything you can think of is the great strength of D&D, but that doesn't mean we should maximize it.  We certainly want lots of it, but if we need procedural mechanics in our game to legitimize our victories and losses.

You might be cool with your DM saying "Okay, hmm.  You'll have to roll under half your Str to pull the wings off the giant beetle.  Actually, roll under 2/3 your Strength--the beetle is pretty old.", but there are people who want a mechanics that feels more legitimate.  Surely you've heard the new school gamer's lament of "If the DM is deciding the difficulty on the fly, then you only win when he wants you to win, and you lose when he wants you to lose."  That's what they're talking about.  Meaningfulness.

The sweet spot is different for each person.  For example, a wargamer can play 40K built on procedural mechanics and have a great time.  (And you can have great roleplay in a 40K game, too.  Have you ever thought about how orks would trash talk the eldar?)

But Arnold, aren't procedural and consensus this just two sides of the same coin?  One is decided by 5 guys at a table at Wizards of the Coast, the other is decided by 5 guys at a table in your grandma's basement.  What's the difference, really?

Well, Mr. Straw, it has to do with trust and authority.  Here, I'll write a section for it.

VIII. Trust and Authority

It's a good question.  What makes a book's authors so much more trustworthy than your Friendly Local DM?  In a nutshell:

  • We trust the book writers to be more experienced, and have spent more time thinking about it.
  • We trust the book writers to be impartial, unlike our DM who might fudge the dice for his friends as readily as go all Killer DM on those assholes who show up at his house every Saturday and drink all of his Mountain Dew.
Neither of these are necessarily true.  I've read some professionally published books with some shitty mechanics.  (Seriously, what was up with the "roll a die to figure out what die you use" stuff at the beginning of Carcosa?)  Likewise, there are companies that might intend a module to be fluffy-and-approachable or Tomb-of-Horrors-fuck-you, depending on what they're going for.  Aaaaand, there are a great deal of DMs that pride themselves at being as fair and impartial as possible (myself included).

I suspect that a lot of this is psychological.  If a jumping spider just bit you on the nose and killed you, which would you rather hear?

"Sorry, bro, but it says right here on page 24 that if you fill the pit with pterodactyls they will start moshing, and their wing buffets will kill you unless you save vs death."


"Sorry, bro, but you filled the mosh pit with pterodactyls.  Of course they're going to start moshing.  I figure their wings weigh about 100 lbs each.  I think a save or die was appropriate in this case."

I also suspect that a lot of this depends on the DM.  

Some DMs are very good at devising fair and appropriate mechanics on the spot.  Others are not so much.  Other DMs don't treat their players objectively (like the pieces of meat that they are) or perhaps they don't seem to do so.

Can we extract any advice from this line of thougtht?

DMs: Consider how you measure up.  Are you good at improvising mechanics on the spot?  Do you find that the book's mechanics over-applied, inflexible, or poorly designed?  Do you like adjudicating ad hoc mechanics?  Are you well-suited for the system you are using?

Players: Consider your DM.  How much do you trust their ability to invent fair mechanics quickly?  How impartial do you think they are?  Does he kill your PC if you neglect to bring him food?

Everyone: Does your ruleset have procedural gaps (situations not covered by the procedures in the rulebook) that leave you flummoxed while you try to come up with a mechanic?  Maybe try a more comprehensive system.  (Pathfinder is very comprehensive.)

Everyone: Is the inverse true?  Are there any rules/mechanics that come up again and again that doesn't seem to be quite suitable for what their intended to do?  (These grappling rules really aren't suited for squirrels, guys.)  If the mechanic describes a life-and-death situation (grappling with squirrels) consider replacing it, or creating a more specific mechanic for your needs.  If the mechanic describes a trivial situation (petting squirrels) consider removing it and just letting the DM rule it on a case-by-case basis, or replacing the mechanic with a simpler one.

Is proceduralism the best word?  Procedure is almost a synonym for method or mechanic.  Precedent vs Arbitrated?  Established mechanics vs Invented mechanics?  What about Orthodox mechanics vs Discovered mechanics.  I like that one.  Professionally-derived mechanics vs Locally-derived mechanics?


  1. Wow. I hate this adjective, but: epic post.

    A comment on Section V. Implications for Game Design:

    The Brendan S post that sparked this talks a bit about indie/story games and their broad but highly dissociated mechanics. I'm not sure that the idea of a dissociated mechanic is relevant in the context of games that allow narrative control stances other than strict Actor style, but in any case, I've been thinking a lot about exactly what you say here: that the procedures provided in a game should be a function of WHERE THE GAME IS. So Cyberpunk 2020 has detailed procedures for computer hacking, but in some other game it's all "roll your Hacking skill vs. difficulty 15". This should, in a well-defined game, line up in some way with character choices: so if your four PC options are driver, magician, hacker, soldier and accountant, I'm going to want some pretty detailed mechanical procedures for financial auditing in there along with my rules for combat, magic, driving and computer hacking. It's surprising how often this turns out not to be the case.

    The worst match of mechanics to game world I can think of would be Kult, where the only description we get of the world is of spacetime-bending cannibal Boddhisattvas wandering through the extraplanar graveyard where God is buried, and the mechanics are all about how many rounds a submachine gun can fire per round and the market value of a first aid kit.

    Anyway, this is getting long, but I think my point is that the indie approach of using broad mechanics seems to me like the only option for a game of indeterminate focus: if you want to have a game of social maneuvering AND swordfighting AND polar exploration AND courtroom drama, then you're going to need something like the "roll Courage vs. the other guy's Determination" approach set out above. But preferably with a bit more procedure and texture, like, say, the "Bring Down The Pain" mechanic from The Shadow of Yesterday. This also allows setting-specific stuff to be added to provide further texture, while keeping it within the same consistent core mechanic.

    1. Like you say, Broad Strokes is the easiest (and maybe the most elegant) way to have a universal (i.e. unfocused) game. Because the mechanics are dissociated, they are applicable to everything (and still procedural).

      But another way to do a universal system is to do what GURPs did, and just make a shit-ton of rules for all possible situation. Procedural mechanics cover every possible situation through sheer volume of mechanics.

      And yet another way to do a universal system is just to do it OSR-style, and use DM rulings whenever the players wander outside the magisterium of the rulebook. Unlike the two previous paragraphs, this is a consensus-based system.

      Which one you a player wants is determined by their expectations.

      A player who wants mechanics to be more appropriate to the intended action will be dissatisfied with the Broad Strokes style storygame (Torchbearer). He'll say "Why is it the same exact roll even when I'm using a different tool and a different approach?"

      A player who don't want to familiarize themselves with the vast mechanics available in GURPS will be unhappy with it. (This isn't a good example, because GURPS has dissociated mechanics, too.) And the more rigid a system is, the less the edge cases tend to make sense. He'll say "This system is designed for small-scale combat, so it's no wonder that the mechanics break down when you grapple 28 squirrels. See also: Cleave + Bag of Rats."

      A player who doesn't trust their DM to be impartial AND create fair mechanics on the spot will be unhappy with the OSR model, where the rules are sparse and the rulings are encouraged. It won't feel meaningful to them. He'll say, "When I lose it's because the DM wants me to lose. When I win it's because the DM wants me to win. And if the DM sucks at making good mechanics, the success of a task is pretty much random, regardless of my stats or tactics."

      The nice thing is that these aren't absolutes. Systems exist in all intermediate degrees along the consensus vs procedure axis (and along the dissociated vs diagetic axis as well).

      A group can pick whatever ruleset best suits them. And if they trust themselves to edit it, they can houserule it for an ideal fit.

  2. This, THIS, is the sort of post that becomes the basis of entire schools of game design. Brilliant. Game design post of the year. Amen.

  3. Well, I think most of this is pretty obvious stuff but perhaps worth wading through for the key point(IMO): rules that the players know about and enforce insulate players from the DM without regard to the quality of that DM.

    "Consensus" is a very bad choice of word, though. "Inventive" or "Ad-hoc" might be better as the distinction is really between printed and at-the-table rulings. Consensus implies things that are not correct - for example, that "procedural" rulings from the books are impossible for a group to reject.

    From the player PoV, all rulings come in the first instance from the DM. It's only as the players become more experienced that they really start to grasp that some are NOT based on the contents of the rule books. But, since the DM decides which to use and which not to use the difference is really quite arbitrary. And many players do not (and, IMO should not have to) even think in terms of rules, but only in terms of what actions their character is attempting. Does the DM look up a table or invent a ruling? The player ideally should not know nor care - the DM is a "Chinese Box" in many ways.

    Once players start to question the DM's decisions, the group is in trouble and that can happen with any ruling, either "why did you give me only a 5% chance of success" or "why did you decide to enforce that stupid rule about cat allergies?" Again, the source of the ruling makes no real difference to the players who object to it.

    One response to losing faith in the DM is to bind the DM more closely with rules which supposedly empowers the players. But the binding applies to the good decisions as much as the bad and, ha ha!, to the players as much as the DM. The sense of security some players derive from this is balanced by the fact that, with a decent DM, they are less able to play their characters. They're getting a "guaranteed minimum" quality but they are also losing the "possible maximum" at the other end.

  4. The procedures I am talking about are broader than only task resolution. In many games (or parts of games), the kinds of questions that you can ask are constrained. Consider, for example, Epidiah Ravachol's swords & sorcery game Swords Without Master. This is not a set of rules for resolving PC actions based on player desires. Instead, it is a series of phases that play moves through with "rogues" (PCs) taking action under the aegis of a dramatic tone (essentially, major or minor, like in music). Consider:

    Each phase has its own rules for when the dice can be picked back up and passed on to another player. Until that time, the player who rolled the dice has the narration. Attend to what this player is saying.

    This is a procedure which structures what can be said, not how to decide fictional outcomes.

    One might object that this sort of thing is a different kind of activity than a "traditional" RPG that defines PCs based on a set of numbers or measures and allows players to then manipulate fictional reality based on a series of operations using those numbers, but little bits of the above kind of procedure creep into D&D as well occasionally, like in Jeff's "party like it's 999" carousing rules, which retroactively define what happened when a PC was out partying too hard (and in the process removes the option for problem solving). Result 10: Beaten and robbed. Lose all your personal effects and reduced to half hit points. What do you mean beaten and robbed? There were not even any attack rolls involved and my PC has plate +3 and a frostbrand sword! Yes, but we entered a procedure where those details are not relevant.

    And, of course, combat is procedural, though in a different way.

    I do think there's definitely something to this set of relationships:

    procedures --> high-stakes --> tension

    Procedures from outside of the referee allow deferral to an external authority, and thus a degree of impartiality around something that might otherwise be difficult to resolve with a meaningful chance of failure (starving, getting lost, etc).

    I have something to say about consensus, that that will require more time for me to collect my thoughts.

    1. These are all good points.

      I think part of the problem with talking about RPGs is that we don't have a good vocabulary for it. The field is not mature.