D&D 4E The Best Thing from 4E

[pemerton]

what would you both think of a setup whereby the "room with a clue" might be visited before the "evil lair" based on a random roll of a die? Say there was a straight 50:50 chance that the room full of incriminating documents might be visited or might not

It would not be very appealing to me, in any case, since it removes player agency.

<snip>

A game is a series of meaningful choices. You may not always know the full ramifications of each choice, but they should always matter.

What is opaque to me is what the meaningfulness of the choice consists in.

If all the players know is that there is a left path and a right path, but do not know what lies down one path rather than another, what agency are they exercising? How is it any different from rolling a die?

This is not to deny that the choice matters, in the sense that the left choice results in outcome A, and the right choice results in outcome B. But then [MENTION=27160]Balesir[/MENTION]'s die roll would matter in exactly the same way. Either way is a random determination of what challenge or situation the PCs are confronted with.

The only thing I can see in favour of player choice over random roll is the aesthetics of going left rather than going right. Which seems to me, as I said a long way upthread, to be mere colour.

 

[pemerton]

That's the thing, then. Stakes, like story, naturally occur on their own. I don't need to think about them, or address them in any way.

I think its OK if the DM starts attaching some more weight to this guy, if the players want to engage that story element. In that case perhaps it does turn out that the stranger is the lieutenant and has a chance to assume his fallen master's mantle.

"Story" has at least two meanings in the RPG context.

One meaning is "a sequence of fictional events". In this sense, any RPG campaign will produce a story.

Another meaning is "a dramatically satisfying sequence of fictional events". In this, sense, many RPG campaigns will not produce stories. It is a significant challenge in RPG design to invent RPG mechanics and techniques that will allow a group of people, through RPG play, to reliably produce a story in this sense although no single person is authoring it (ie there is no "force" in the sense that [MENTION=6696971]Manbearcat[/MENTION] has talked about upthread).

Stakes will naturally occur, in the sense that most RPG sessions will oblige the PCs to make choices where objectives can be furthered or undermined. But there are plenty of episodes of RPG play where the players don't know what is at stake (eg the ever-popular "the NPC who hired you is really a traitor" motif). And there are also questions about who decides what is at stake: players or GM.

If the PCs are making a blind choice to go left or right, and that's all there is to it, then there's no practical difference whether it was pre-determined or determined on-the-spot by a die roll.

Agreed.

Players are going to disagree about which way to go, or someone is going to run off

<snip>

Maybe they'll get halfway down a corridor and then change their mind. You never know.

On its own, this doesn't seem to add anything to random determination. If what the players are arguing about or responding to are reasons to go one way or another, that's a different matter, which takes us to . . .

there will be an explosion which blocks one of the passages.

Right. From my point of view, the interesting question is who gets to decide that there is a risk of explosion.

If the GM decides this unilaterally, it is a GM-driven game. (It may or may not be illusionistic.) If the GM decides this on the basis of player signals about stakes for their PCs, it is a player-driven game.

 

[pemerton]

The DM has no way of knowing if or when the PCs will encounter this stranger

I don't understand how this can be true.

The GM knows where the PCs are (geographically, temporally). And the GM decides where the stranger is (geographically, temporally). So the GM can very easily decide that the PCs encounter the stranger by saying (for instance) "As you walk down the left-hand corridor, you see a stranger stepping out of what looks like a library."

The GM can choose to randomise whether or not the PCs meet the stranger - for instance, by writing a note that "the stranger is in place X at future time T" before s/he knows where the PCs will be at time T. Or by putting the stranger on a random encounter table.

But that is a choice. It is not forced on the GM (ie it's not the case that s/he has no way of choosing otherwise). There may be reasons for making such a choice, but I don't think they've really been spelled out in this thread.

I'm not sure what you mean by 'frame' in this instance. Is that like railroading?

No. The GM narrates the situations in which the PCs find themselves. "Framing" means being deliberate about this: narrating situations which will make the players make choices that relate to the stakes they have signalled they want to be put on the line.

In the example I'm imagining, they might overhear this NPC at the tavern, complaining about her boss or some other sort of vague grievance. Maybe they approach, and start asking questions, in the way that adventurers do.

These two sentences can carry a wide range of meanings, depending on other assumptions about play.

I've played in games where the GM would use this sort of device to steer the players onto some particular path: it's a species of the genera "plot hook". The players are expected to follow the GM's leads in this respect. From my point of view, it's a type of railroad. It's also heavily metagamed, in the sense that "the way adventurers do" is a thin ingame veneer over a metagame expectation that the players will follow the GM's lead.

In my game, if I tell the PCs there is an NPC grumbling about her boss, they will assume (correctly) that the NPC's complaints relate in some fashion to some dramatic concern that has been expressly or implicitly flagged, because they know that I run a game in which I respond to the players' cues. The metagame imperative runs in the opposite direction (roughly: the players "hook" the GM, rather than vice versa).

In a truly metagame-free game, the players would be free to just ignore the grumbling NPC (much as, in the real world, I ignore more than 90% of the strangers I hear complaining in cafes, on the train, etc, and would never think to stick my nose into their personal business). But of course that is not very conducive to RPG play.

More likely, they ignore her as irrelevant, only to recognize her much later on when she does her thing.

This is exactly the sort of play I try to avoid. As a general rule, it is not dramatically satisfying.

In any case, though, the secret documents aren't necessarily important. Maybe the party never went into town, so they don't recognize anyone on the list. Maybe nobody can read the language it's written it. It is what it is. If they find it, and it helps them, then great. If not, then don't worry about it.

And I have much the same response to this: what was the point of this episode?

If the players choose to have their PCs explore the library, only to find a bunch of stuff that either their PCs can't read, or that tells them nothing of use, and then head down the right-hand path only to find that the detour via the library meant the prisoners were sacrificed, I would find that an extremely unsatisfying episode of play. It seems that the players are being punished for making a choice that looked reasonable, and that they had no way of knowing would be pointless and self-defeating. It seems pretty like a microcosmic version of "rocks fall".

I want the DM to set things up in whatever way makes most sense to the DM, without particular consideration for how the players feel about it.

That's why I call it a GM-driven game. If the GM then pretends that the players' choices make a difference to things, that is illusionism.

 

[EzekielRaiden]

That's why I call it a GM-driven game. If the GM then pretends that the players' choices make a difference to things, that is illusionism.

That seems like a nice, simple definition of the distinction. That is, illusionism is a proper subset of GM-driven games, because illusionism must be GM-driven, but with a GM that is forthright about things being GM-driven, a GM-driven game does not need to be illusionist. A player-driven game is thus automatically non-illusionist because it is not GM-driven.

That said, I can still see where some people might feel a little confused, because in all of these, the DM is still deciding what actually does happen in the world. I don't see a problem with it, but the logic for the other side is sensible (even if I disagree with it). That is, one could say that the DM is always the party "actually" responsible for what occurs in the world, in a way analogous to the existentialist proposition that individuals are "actually" responsible for valuing something, rather than the thing itself possessing value.

I'd simply respond that while the DM always exerts control, that is not the same as always being responsible for the result. If the DM's choices are made specifically because they are predicated on what the PCs have done and what the players want, what does it matter that the DM is free to not do that if she chooses? She hasn't chosen to. She is responsible for the choice to make the players "responsible" for the story, but that is not the same thing as she herself being responsible for the story.

 

[The Crimson Binome]

"Story" has at least two meanings in the RPG context.

One meaning is "a sequence of fictional events". In this sense, any RPG campaign will produce a story.

Another meaning is "a dramatically satisfying sequence of fictional events". In this, sense, many RPG campaigns will not produce stories.

That seems like an unnecessarily harsh criticism. People can disagree whether a given story was good or bad, but I don't think anyone can say that any story was so bad that it isn't even a story.

Right. From my point of view, the interesting question is who gets to decide that there is a risk of explosion.

If the GM decides this unilaterally, it is a GM-driven game. (It may or may not be illusionistic.) If the GM decides this on the basis of player signals about stakes for their PCs, it is a player-driven game.

Okay. That seems like a weird definition, though. In my view, it is entirely the job of the DM to design the evil lairs and their structures and the likelihood that a passage might collapse if someone gets too careless with the Evocations. I mean, that's just part of the job description. It's the players who drive the action, though, and the DM kind of just follows them around and describes what happens. The DM describes the hallway that could potentially collapse under the right circumstances, but it's usually left to the players to actually cast the spell to make it happen.

Unless you mean dramatically speaking, under the principle that villains act and heroes react, in which case the DM is the one driving the game because the DM controls the Big Bad.

 

[JamesonCourage]

I think it's interesting if you find a 50% chance as a result of some skill check significantly different from a simple, independent 50% chance, but what I had in mind was a simple, independent 50% chance.

Okay, we'll proceed with that in mind for now.

So, the suggestion is that the GM's intent is a critical element, here? Not merely the fact of what is done, but the intent behind it? How are the players to have any inkling of GM intent?

A critical element, yes. And the players don't even need to have an inkling one way or another. Why would that matter in a discussion of illusionsim or railroading? Whether or not players are aware of those techniques are completely independent of them being used.

Just for the record, what I had in mind really didn't feature the "fork in the way". You could think of it as a sequence of rooms where the order of entry was random. In other words, the first door leads to one room or the other, and beyond that another door leads to the second room.

Okay, also noted for this discussion.

An interesting sequel to the comparison might be, as well as your view on rolling for which room is through the first door - the clue room or the fight room - would it make a difference if the order of rooms was diced for in advance, before the session, as opposed to when the players announced that the characters were going through the (first) door?

I think it really only makes a difference if the GM then goes on to change it. But in your example we don't even have a choice by the players (going left vs right). So, illusionism will be harder to produce from your example. The GM may or may not be railroading, depending on how we got to this point. More context is needed.

An uninformed decision by someone is thus identical, to me, to a random selection.

I guess I don't really believe in actual "random selection" really happening, even if players are uninformed. They may not know which way is better (since they have no knowledge either way), but they're probably basing their decision on something. Lucky number, a perceived pattern, guessing how the inhabitant's psychology works, a gut instinct, etc.

So, for the purposes of this exercise, please exclude the "skill check for a helpful skew" possibility.

Alrighty.

How can the players have any notion of their choice genuinely mattering if they have no information on which to base the decision?

I guess they don't necessarily. But by being presented with a choice, they probably assume that the choice matters. Why pause to address it otherwise?

The only thing they might possibly perceive is that they are making a genuine gamble - which remains true if a die is rolled for the result. There is, incidentally, no reason I can see why the players should not roll the die (except a purely aesthetic one if they wish to pretend that there is some sort of "reality" to the game world beyond the GM - i.e. a Simulationist agenda).

I don't think the players or the GM rolling the die in this scenario matters, either.

I don't see what is the difference between this and example 2? Can you explain further?

That should have been "the GM has notes" and not "the GM has no notes." Sorry about that. I was worried I'd mess up while going out the door (with the copy-pasta modification), and apparently I did. My fault.

As a further explanation of my own position, I really don't see the point of the fork unless it has some value in player decision making.

In my mind, making a decision, even without a way to discern the different, is still something that I don't want to rob the players of. In this unlikely scenario (the fork with no indications and no information for the players), which way they pick will affect the rest of the campaign. Why would I want to change that? I want to see what happens when we play it out!

As Robin Laws says in Gumshoe - what is interesting about clues isn't finding them, it's what you do with them when you have them. In a sense, this might be the real dichotomy between ROLLplaying (rolling to find the clues) and ROLEplaying (the decisions made once the clues are revealed).

I like both in my game, because both can lead to really, really interesting scenarios. But in my RPG, I don't run anything that looks like dungeons (as far as I can tell), so I also have a suspicion that my campaign is wildly different from most tables. It's more about unfolding events: wars, politics, diseases, dynasties, uprisings, disasters, etc. The players react and involve themselves in the events as they unfold, or struggle to avoid them (or the consequences of them, at least).

Heck, like I've said previously, 90% or more of my game isn't combat, even if the players like to build combat-capable characters (when my RPG in no way demands any combat proficiency); right now, the PCs include a polearm warrior, a fire mage, and a dark knight (kind of a TK/fire knight). Of course, they can all do stuff outside of combat (the warrior is a capable physician, surgeon, diplomat, courtier, and knight-level noble, for example).

Anyway, as a sandbox-oriented GM, not finding a clue might mean they never get something. They never know that a plot is underway, giving it greater odds of happening (I don't really run things as "only the PCs can stop this event!"). But that's okay, because the plot going through and the king being assassinated can be really interesting. They weren't directly involved, but now they get to deal with the social ramifications of such a change in power. How will the new king act? How will other nations react? The noble houses? Bandits?

And this might come about because they chose right instead of left. And I'm cool with that. That makes for an interesting game for me, as GM, because I have no idea what's happening, and get to discover things with them. I don't want to do the heavy lifting of setting a narrative. I'd rather explore it with the players (as much as I'm able to). All of that make sense?

 

[The Crimson Binome]

The GM knows where the PCs are (geographically, temporally). And the GM decides where the stranger is (geographically, temporally). So the GM can very easily decide that the PCs encounter the stranger by saying (for instance) "As you walk down the left-hand corridor, you see a stranger stepping out of what looks like a library."

By the code of conduct to which I adhere, as a DM, I will not do this. I'm not going to meta-game, to guarantee that convenient coincidences actually happen. I will put that NPC where it makes sense for that NPC to be, and if the players find her, then great. If they don't find her, then so be it. It's not my place to force the encounter. The players must be in charge of their own fates.

No. The GM narrates the situations in which the PCs find themselves. "Framing" means being deliberate about this: narrating situations which will make the players make choices that relate to the stakes they have signalled they want to be put on the line.

That sounds like meta-gaming to me. It's the DM controlling where the PCs end up, which is a huge violation of player agency. The only time where I would consider this acceptable would be in the pre-game. You can open with whatever scene you want, because that's part of the premise, but once the game actually starts it is no longer an option.

I've played in games where the GM would use this sort of device to steer the players onto some particular path: it's a species of the genera "plot hook". The players are expected to follow the GM's leads in this respect. From my point of view, it's a type of railroad. It's also heavily metagamed, in the sense that "the way adventurers do" is a thin ingame veneer over a metagame expectation that the players will follow the GM's lead.

It's a plot hook, possibly, but it's only a type of railroad if the players are forced to engage with it. In that sense, it's more like a marked trail, that they're free to stray from at any point. Or it could amount to local color, and not lead to anything very interesting. If the players choose to pick up on it, then it should be because they're playing the types of PCs who like helping random strangers at the bar, and not because the players expect the DM to have contrived this situation for their benefit.

This is exactly the sort of play I try to avoid. As a general rule, it is not dramatically satisfying.

Really? I would find that quite satisfying, for all of those previously-ignored clues to suddenly click into place.

And I have much the same response to this: what was the point of this episode?

If the players choose to have their PCs explore the library, only to find a bunch of stuff that either their PCs can't read, or that tells them nothing of use, and then head down the right-hand path only to find that the detour via the library meant the prisoners were sacrificed, I would find that an extremely unsatisfying episode of play. It seems that the players are being punished for making a choice that looked reasonable, and that they had no way of knowing would be pointless and self-defeating. It seems pretty like a microcosmic version of "rocks fall".

That's on you, though. You shouldn't see it as a punishment, because you should trust that the DM is neutral on the matter. The DM never told you that the sacrifices would wait for you. The DM never told you what was or was-not in the library. You gambled that you had the time, and you gambled that you might find something useful, and you lost.

If you do see it as a punishment, then that feeling would be based on your belief that the DM is supposed to contrive dramatically satisfying scenarios for you. You should probably talk that over with the DM, before you join the game. Although, to briefly return to the topic at hand, you might take that as a given for a 4E game, but not necessarily for any other edition - if you're going into the game with that sort of goal, then you might as well play the edition that supports it.

That's why I call it a GM-driven game. If the GM then pretends that the players' choices make a difference to things, that is illusionism.

And that still seems misleading, to me. All of the choices are in the hands of the players, and only their outcomes are uncertain. If the DM is doing a reasonable job of describing the world, then the players should feel ownership over their decisions.

 

[Manbearcat]

That said, I can still see where some people might feel a little confused, because in all of these, the DM is still deciding what actually does happen in the world. I don't see a problem with it, but the logic for the other side is sensible (even if I disagree with it). That is, one could say that the DM is always the party "actually" responsible for what occurs in the world, in a way analogous to the existentialist proposition that individuals are "actually" responsible for valuing something, rather than the thing itself possessing value.

I'd simply respond that while the DM always exerts control, that is not the same as always being responsible for the result. If the DM's choices are made specifically because they are predicated on what the PCs have done and what the players want, what does it matter that the DM is free to not do that if she chooses? She hasn't chosen to. She is responsible for the choice to make the players "responsible" for the story, but that is not the same thing as she herself being responsible for the story.

Just using your post here as a point of further analysis as you're trying to penetrate where some confusion may persist (and defuse it). Gonna pop these back in here again right quick.

The Black Curtain: The effects of a variety of Techniques a GM may employ to keep his use of Force hidden from the other participants in the game, such that they are at least somewhat under the impression that their characters' significant decisions are under their control. See Illusionism, Force.

Force: The Technique of control over characters' thematically-significant decisions by anyone who is not the character's player. When Force is applied in a manner which disrupts the Social Contract, the result is Railroading.

Illusionism: A family of Techniques in which a GM, usually in the interests of story creation, exerts Force over player-character decisions, in which he or she has authority over resolution-outcomes, and in which the players do not necessarily recognize these features.

I'm also going to include Calvinball because it is relevant when employed by the GM (especially covertly): A potentially-dysfunctional Technique of Hard Core Gamist play, characterized by making up the rules of a game as it is played, especially in the immediate context of advantaging oneself and disadvantaging one's opponents.

Let us consider this conflict again.

1) The player of Saerie has made a thematically-significant decision; "I want to locate civilization so I can beseech them to take on these orphaned children."

2) 4e has transparent resolution mechanics and play procedures which are meant to be deployed to organically and authentically derive the outcome of this player-invoked-conflict; The Skill Challenge.

3) The only time that it would not be Force for a GM to say "there is no civilization to locate", thus exerting control over the player's thematically-significant decisions, is if this doesn't (somehow) violate the social contract. I can think of about 1 scenario (perhaps there are more); you are playing in a world utterly bereft of civilization and the players (characters are irrelevant) are aware of this. Off the top of my head, I can think of no such world in genre fiction (not even any form of the darkest post-apoc from Dying Earth to Mad Max to 28 Days Later to The Road)...but hey, if that was stipulated overtly to the players and they have buy-in, have at it.

4) In any other situation, it is the GMs job to (a) confirm the stakes with the players, (b) frame the conflict, (c) play the adversity while (d) consulting the resolution mechanics with the players. Through that play procedure, the table will (e) derive the resolution-outcome without Force ("find out what happens" or "create story").

So why would I want to covertly appropriate the player's agency (Force being control over thematically-significant decisions and Illusionism being its covert usage) here? Historically, this happens for one of (or a combination) a few reasons:

* I think my own ideas on story creation are better than this player's ideas (either for them or for the table as a collective); eg GM's metaplot.

* I'm running a prescribed module or adventure path, this declaration and/or its prospective result is "off the grid", and I'm insufficiently equipped to let the trajectory of play move beyond that prescription.

* I'm an adversarial GM who perceives play in a "me vs them" paradigm and any conflict won by the players is points scored by the opposition (see "Calvinball" above).

That is all I can think of. There might be others.

So how would I accomplish the covert appropriation of the player's agency (Force being control over thematically-significant decisions and Illusionism being its covert usage) in 4e?

Well, it would have to be extraordinarily subtle. In effect, this is subalpine mountain exploration conflict. The player has built their PC such that they are "weak" on the CHA social skills (not Insight). "Weak" in this case meaning 50 % chance at the medium DC. The overwhelming majority of DCs in a Skill Challenge are medium (and it becomes moreso as you move up the complexity due to advantages...and you also get more augments with # SS scaling 1:1 with complexity). Could I somehow turn the whole friggin thing into a social challenge that tightly funnels play toward her social skills? I don't see how that would be remotely possible. Even if it was, it would be about as subtle as a flying brick to the face. How else? Manipulation of target DCs? They are right there, in your face; 11 > 15 > 23 for level 6. 6 of them should be medium and 2 high. Again, about as subtle as a flying brick to the face.

Overt Force, Calvinball goal-post shifting/rule changing, and covert Force (illusionism) is just extraordinarily difficult in 4e. I mean, I guess if you have players that are utterly uninvested and don't know the rule system at all...then...sure. But if that is someone's litmus test for "how vulnerable system n is to Force", that is some low-hanging fruit.

 

[pemerton]

in all of these, the DM is still deciding what actually does happen in the world.

I think that addressing this issue is quite fundamental to distinguishing playstyles.

I want to talk about in relation to framing conflicts/narrating situation, and then in relation to action resolution.

In relation to the framing of conflicts, here are two examples from actual play:

* GMing a thieves-only Greyhawk campaign, I have the PCs discover that the Lady Mayoress of Critwall is a member of a Chemosh-worshipping cult (the session was over 25 years ago, so I'm hazy on the details);

* GMing my current 4e campaign, the PCs discover Orcus cultists working as kitchen hands for the Baron of Adakmi.​

From these descriptions, you can't tell whether the game is player-driven or GM-driven. To learn that, you need additional information.

In the first episode, the reason that I, as GM, introduce that story element is because I've been reading my Dragonlance Adventures hardback, I like some of the flavour around Chemosh (roughly, the Orcus of the DL world), and so I introduce a Chemosh cult into the game. That's a GM-driven moment of play. I can't remember much about the Mayoress, but my best recollection is that she was only an element of the game because I had introduced her.

In the second episode, the reason that I, as GM, introduce the story element is because half the PCs in the campaign are Raven Queen cultist and hence diametrically opposed to Orcus. The importance of the PCs' relationship with the Baron was also something that the players had been pushing, as part of their emergence into paragon tier in the context of the town's political scene. That's a player-driven moment of play: although the GM is the one actually doing the job of framing the conflict, I am following the players' leads. I'm not inviting them to follow my leads.

I think any game in which the authority for framing conflicts and narrating transition resides with the GM (which is pretty much all mainstream RPGs, "indie" or otherwise) individual episodes of play may involve elements of both player and GM direction, and from episode to episode the balance may shift back and forth from one to the other. One reason for this is simply that the GM, in framing conflicts, has to introduce fictional content, and if s/he does that job well the players may find it interesting and end up building on it - thus generating a back-and-forth between GM and player contributions to the fiction as the game unfolds.

Turning now to action resolution. In some RPGs, the outcomes of action resolution are expressly binding on the GM. (I think this is the default for D&D combat, but not so much for D&D non-combat outside of 4e's skill challenges).

And in some RPGs, (i) part of action declaration is establishing the player's intent for his/her PC, and (ii) if the player succeeds on the check, then that intent is realised. (Burning Wheel is an instance of this, and I think D&D combat resolution is meant to work like this, at least as written.)

In this case, it is not the GM who decides what happens. It is the player. And that's another important aspect of a player-driven game: if the players can introduce content into the shared fiction (via actin resolution which is binding on the GM) then they can shape and constrain the story elements that are available to the GM in framing future conflicts.

Of course, if the GM were free to just introduce new story elements at will, then there would be no such shaping or constraint - but this just goes back to the first issue I addressed in this post, of where the material for framing conflicts comes from. If the GM is obliged (whether formally or informally) to have regard primarily to the player's signals in relation to introducing story elements in the framing of conflicts, then s/he is not free just to introduce new material in order to circumvent the outcomes of action resolution.

I've given some practical examples of this upthread, from my 4e game: with Torog dead, the Underdark is erupting in chaos; with Lolth dead, her webs are not longer there to hold the cosmos together. I am not free just to introduce new material into the fiction (eg new gods, or powerful NPCs, or whatever) to circumvent these consequences of action resolution.

I'll finish with a slightly orthogonal point: this also illustrates how 4e can be both a combat-heavy game and a story-oriented game. If the NPCs and entities with whom the PCs are fighting play meaningful roles in the fiction, such that defeating them in combat matters to the state of the fiction, then there is no tension between the game being one in which combat between the PCs and these entities is prominent, and the game being one in which the fictional stakes are what drives the campaign onward.

(In my view, the main reason that classic hack-and-slash games are both perceived as, and experienced as, shallow is not because of the preponderance of combat, but rather because of the lack of meaningful stakes. Filler combats are the enemy of story, as well as an enemy of player-driven play.)

 

[pemerton]

That seems like an unnecessarily harsh criticism. People can disagree whether a given story was good or bad, but I don't think anyone can say that any story was so bad that it isn't even a story.

I'll give an instance from actual play - the first session of Moldvay Basic that I GMed.

(A long time ago. So only the barest outline is recalled.)

A party of 12 or so characters was rolled up. They entered the dungeon I'd designed, which had a 5th level (or thereabouts - able to use the 3rd level spells mentioned in the rulebook) magic-user at the bottom of it. Some wandering monsters were encountered, a few rooms explored, the wizard confronted. In the end one of the characters - a halfling named Gloin Baggins - survived.

There is a sequence of events there, but no story. Nothing of value to any of the characters was put under pressure except their survival. The treasure they hunted was of nothing but instrumental importance to them.

That is not dramatically satisfying, though at the time it was a sufficient pleasing episode of play that we kept at it.

For me, as a participant in episodes of RPGing (and as someone who chooses to spend my leisure time RPGing rather than doing other things with my friends), there is no comparison between that play experience and (say) the 20 seconds of tension when I waited to see whether the invoker/wizard would channel the souls to the Raven Queen or allow Vecna to take them: that was the culmination of foreshadowing by both me as GM (poking away with Vecna motifs, for so much of the campaign, including (just to give one memorable example) the PC suffering damage when he handled the Sword of Kas at low paragon, which the player correctly deduced was because the Sword could sense the PC's allegiance to Vecna) and by the player (implanting the Eye of Vecna into the imp), and came at the culmination of a quest that the PC Questing Knight had chosen for himself (to wreck the Soul Abattoir) - which added the frisson of potential intra-party conflict.

I'm not saying that I could publish the second story and retire on the proceeds!, but I think it is in a completely different ballpark from the first adventure, and it was the adoption of certain GMing techniques that made it possible.

In my view, it is entirely the job of the DM to design the evil lairs and their structures and the likelihood that a passage might collapse if someone gets too careless with the Evocations. I mean, that's just part of the job description.

<snip>

The DM describes the hallway that could potentially collapse under the right circumstances, but it's usually left to the players to actually cast the spell to make it happen.

In this Burning Wheel session, why did I (as GM) decide that the NPC wizard live in a tower? Because one of the PCs had the instinct "Always cast Falconskin if I fall", and so I (as GM) wanted to put some of the action into a high place where a fall might take place!

By giving that PC that instinct, the player had shown that the risk of falling was something that he wanted in the game.

Generalising a bit, as GM I get to decide whether or not the gameworld contains towers, or passages that are vulnerable to collapse, etc. What should guide those decisions? In what I am calling a player-driven game, the choices made by the players, which signal what it is about their PCs that they want the game to force them to put on the line, provide the GM with guidance.

It's the players who drive the action, though, and the DM kind of just follows them around and describes what happens.

I think by "drive the action" you mean "make action declarations for the PCs". It doesn't sound as if they have a big influence on determining the content of the shared fiction.

For instance, if the GM decides who all the mysterious strangers are, where all the clues are and what they are clues to, what the treasure is and where/how it is hidden, without reference to player signals (whether implicit or explicit) then it seems to me that the shape of the action has been decided mainly by the GM.

Unless you mean dramatically speaking, under the principle that villains act and heroes react, in which case the DM is the one driving the game because the DM controls the Big Bad.

That could be a part of it, but isn't the core of what I'm talking about.

Who gets to choose the nature of the "Big Bad"? In my reply to [MENTION=6790260]EzekielRaiden[/MENTION], above this post, I gave two actual play examples about this. One illustrated GM-driven, the other player-driven.

In a player-driven game of the sort I'm talking about, the players - through their choices in PC build, in backstory authorship, in the course of play - play the central role in deciding who is the Big Bad. Thus, in my 4e game, I'm not the one who decided that Torog, Orcus, Lolth and (ultimately) Vecna would be nemeses of the PCs. The players chose that.