D&D 5E Roleplaying in D&D 5E: It’s How You Play the Game

[clearstream]

There's nothing about D&D that precludes stakes being made clear. I suspect at many tables they are not, but those are probably not tables using stakes/narrative-based "say 'yes' or roll the dice".

In the context of 4e play, if it's not clear to me what a player thinks is at stake in a situation, I speak to them to clarify it.

Marked differences between what, or whom?

If the player sees the situation one way - and hence declares such-and-such an action for their PC - and the GM sees the situation another way, I'm not sure why it is essential that the GM's conception should prevail.

I take it that you're describing play that is more about learning the GM's conception of the fiction (eg the GM has already decided whether or not Marlinspike Hall is patrolled by dogs). So the fictional position of the PCs includes stuff that the players don't know about because it hasn't yet been revealed to them by the GM.

If I'm wrong, then I'm even less clear about what you're saying than I thought!

Thank you, I feel you and @Oofta explained the dividing line far better than I did.

In a game where players are explorers rather than authors of the world, "say 'no', and don't roll" is at times the right response to cases that still fall within genuine participation. That is because there may be things unknown to players, but known to DM. The stakes - whether narrative or simulation - are sometimes not fully clear to the players. Even so, players can make choices that adjust them: changing a "no" to a roll, a roll to a better DC or "yes", a "yes" to a roll, etc. This is the normal mode of D&D.

In a game where authorship is shared, mysteries are equally unknown to the group and as you say, why prefer the DM? Even so, the group must have something in mind that answers the question: "say 'yes', or roll?" It may be they cannot see that a declaration is valid, so as a group they might say "no" or as you put it require expansion to get to "yes" or "roll". In any case, this is not the normal mode of D&D.

In both these modes, I do not see the stakes as unable to be impinged upon by player choices. To me, some of the most interesting player or group choices are those that change the stakes in play. Including worsening them! Is it that you require players to commit to the stakes as articulated, or can they position themselves in the fiction in a way that could impinge upon the stakes? Maybe that is a more fundamental division?

As to whether D&D precludes the second mode. As written it does mechanically. An example being ability checks where players in most cases aren't empowered to call for one or set its terms. And as written it does narratively, casting DM firmly as storyteller. What is gained is much as @Oofta describes. Characters live in the world, without knowing everything about that world.

 

[Ovinomancer]

Thank you, I feel you and @Oofta explained the dividing line far better than I did.

In a game where players are explorers rather than authors of the world, "say 'no', and don't roll" is at times the right response to cases that still fall within genuine participation. That is because there may be things unknown to players, but known to DM. The stakes - whether narrative or simulation - are sometimes not fully clear to the players. Even so, players can make choices that adjust them: changing a "no" to a roll, a roll to a better DC or "yes", a "yes" to a roll, etc. This is the normal mode of D&D.

In a game where authorship is shared, mysteries are equally unknown to the group and as you say, why prefer the DM? Even so, the group must have something in mind that answers the question: "say 'yes', or roll?" It may be they cannot see that a declaration is valid, so as a group they might say "no" or as you put it require expansion to get to "yes" or "roll". In any case, this is not the normal mode of D&D.

In both these modes, I do not see the stakes as unable to be impinged upon by player choices. To me, some of the most interesting player or group choices are those that change the stakes in play. Including worsening them! Is it that you require players to commit to the stakes as articulated, or can they position themselves in the fiction in a way that could impinge upon the stakes? Maybe that is a more fundamental division?

As to whether D&D precludes the second mode. As written it does mechanically. An example being ability checks where players in most cases aren't empowered to call for one or set its terms. And as written it does narratively, casting DM firmly as storyteller. What is gained is much as @Oofta describes. Characters live in the world, without knowing everything about that world.

This is... not quite right. There appears to be a strange assumption that in games that feature "say yes or roll the dice" that the players have the ability to know things about the world at all times. This is, in fact, opposite to the intent of play. The intent of play here is for everyone at the table to discover things about the world at the same time. Let's take the nearly canonical example of "say yes or roll the dice," the secret door. PCs are in a situation where, for whatever reason (canonically being chased by guards), have decided to look for a secret door. The GM, in this approach, doesn't know if there's a secret door here as much as the players do not know if there's a secret door here. If there's no narrative stakes, then the GM can "say yes" and there's a secret door. This would be because there's nothing at stake here -- the existence of this door is pure flavor and not anything of any heft in play. If this isn't true, and the canonical example say it's not due to being chased by guards (a clear narrative stake), then the dice are used. A success here would mean that there is, indeed, a secret door. A failure would mean that the GM can establish whatever fiction they want -- they may establish that there is indeed a secret door, but opening it reveals that the guards know that too and have used it to outflank the PCs and are now coming out of said door. Or that there isn't a door, and now the guards have arrived. Or there is a door, but it creates a completely different danger to the guards, thus stacking up the challenge. At no point here do the players know more about the setting or world than a D&D player would, they just have the understanding that success means that their intent follows and cannot be blocked by the GM consulting notes or their preferences and just saying "nope, no door."

The idea that their isn't ambiguity or mystery involved in approaches that use "say yes or roll the dice" is deeply flawed. I mean, I'm currently playing in an PbtA game (that generally features a "say yes or roll the dice" approach) that focuses on mysteries! (The game is The Between.) So, this cannot be an actual feature of this kind of approach to play and is actually counter to the entire conception.

All that said, the mostly right part of the above is that the standard approaches to D&D are very much that the play is finding out what the GM's conception of the fiction is. This isn't a bad thing at all -- it can be quite fun. It can also be something that people can dislike, as @pemerton does. It's just a feature, and you can like or dislike the feature according to your preference. How it works isn't changed by that preference, though.

 

[Krachek]

The Secret Door case is a good sample to explore the various play style used at a DnD table.
From the gritty realism to snowflake fantasy the secret door case can be solved in numerous ways. Is pretty hard to evaluate a solution without considering the play style involve.

 

[Ovinomancer]

The Secret Door case is a good sample to explore the various play style used at a DnD table.
From the gritty realism to snowflake fantasy the secret door case can be solved in numerous ways. Is pretty hard to evaluate a solution without considering the play style involve.

I don't really think that "say yes or roll the dice" is compatible with D&D at all, with the noted exception of 4e. @pemerton tagged me on this awhile back in this thread, but I was taking a break for the holidays (and may continue to do so), so I'll expound a moment using 5e. SYORTD requires that there be a useful and consistent mechanic on the roll the dice side. This is absolutely present in D&D, but only in combat. In fact, combat is probably the closest to SYORTD that 5e gets -- it's almost always roll the dice but the resolution mechanic here is consistent and useful to resolve combat for the most part. Very few instances exist where the GM is expected to deploy "no" to an action declaration. So, here, mostly because the roll the dice aspect is robust, the approach works.

The moment you get away from this, though, it breaks down and the system starts fighting you. This is because 5e places the entire burden of how the mechanics works on the GM's judgement of the fiction. Even if you allow for No Myth, it's still the GM's job to determine the DC of a challenge, and whether or not a single check is sufficient to resolve a conflict. This means that the difficulty of a challenge is arbitrary and fluid and not consistent and further that the expected length of a challenge is similarly so. You can establish rules and principled to lock these in, sure, but now you're adding layers to the game that go directly against it's design intent. Still, this would be fine except that you now have to deal with the player side of the game, and that means how classes are built, how skill bonuses work, and, in general, how the entire character build section is aimed at providing player-side tools against arbitrary GM DC setting that, when the GM becomes constrained, suddenly erupt into player-sided overwhelming of the systems. If I'm fixing DCs on the GM side to make them consistent enough that "roll the dice" isn't an arbitrary exploration of the GM's conception of the fiction, then things like bardic inspiration, expertise, guidance, and even just dis/advantage suddenly become outsized and disruptive. So, now to do this in 5e I have to address this as well and now we're very far away from the designed game and into what I like to call "a different game altogether." I struggle to understand how so many claim to be playing 5e with heavily modified games such that they're actually different games, but this seems to be a thing that's done.

And, all of that above goes to another huge point about why SYORTD doesn't really work in most D&D, not just 5e, and that's how the characters are actually challenged. D&D is built on an attrition model that's extended over multiple consecutive challenges (exception 4e, which was balanced to the encounter). If I want to put pressure on a D&D PC, I have to do this consistently over a period of time, and that time period is pretty uniform within each edition of play. This means that the GM must be able to control pacing in the game if this pressure is meant to come to bear, and if you're controlling pacing you're already outside of SYORTD. The games that do best with SYORTD are ones that allow for short-scale high pressure play that isn't contingent on pacing. BW, for instance, can go from zero to 60 very quickly, and PbtA is entirely structured on the snowballing effect of success at cost and failures being the majority of rolls. These systems, through their very structure, create pressure during play without having to do anything outside of play to introduce it. D&D lacks any real ability to do this. 4e allows for SYORTD because it's pacing mechanisms outside of an encounter/skill challenge are largely weak (with only healing surges being used) so that pacing isn't a thing that the GM needs to pay attention to much at all. You can significantly challenge 4e PCs every single encounter without worry and, in fact, the game works best when you do this. Thus, no need for the GM to control pacing and deal with how that directly confronts SYORTD as an approach.

 

[clearstream]

This is... not quite right. There appears to be a strange assumption that in games that feature "say yes or roll the dice" that the players have the ability to know things about the world at all times.

I was thinking about that as I wrote. My thought is that players know everything there is to know about the world at all times, but the world can come to contain new things.

I think this is epistemologically sound as I might have a justified true belief at one time that there are three children in my family, and later when our fourth baby arrives the world contains a different fact i.e. that there are four children in my family. At both times, I correctly knew the number of children in my family.

This is, in fact, opposite to the intent of play. The intent of play here is for everyone at the table to discover things about the world at the same time. Let's take the nearly canonical example of "say yes or roll the dice," the secret door. PCs are in a situation where, for whatever reason (canonically being chased by guards), have decided to look for a secret door. The GM, in this approach, doesn't know if there's a secret door here as much as the players do not know if there's a secret door here. If there's no narrative stakes, then the GM can "say yes" and there's a secret door. This would be because there's nothing at stake here -- the existence of this door is pure flavor and not anything of any heft in play. If this isn't true, and the canonical example say it's not due to being chased by guards (a clear narrative stake), then the dice are used. A success here would mean that there is, indeed, a secret door. A failure would mean that the GM can establish whatever fiction they want -- they may establish that there is indeed a secret door, but opening it reveals that the guards know that too and have used it to outflank the PCs and are now coming out of said door. Or that there isn't a door, and now the guards have arrived. Or there is a door, but it creates a completely different danger to the guards, thus stacking up the challenge. At no point here do the players know more about the setting or world than a D&D player would, they just have the understanding that success means that their intent follows and cannot be blocked by the GM consulting notes or their preferences and just saying "nope, no door."

Great example: it makes the case very clear. So I would be suggesting that when the party passed through here earlier, there existed a collection of facts that at that time did not include a secret door. Players were in possession of those facts. Later, a new fact is added.

It seems that we might prefer to say that at no time do the participants know the limit to the number of facts about the room. There is no limit on those facts. Obviously this still contrasts well with one participant being in possession of extant facts the others lack.

The latter case can still turn out to contain yet further facts: unknown to all. That happens a lot in open-world games.

The idea that their isn't ambiguity or mystery involved in approaches that use "say yes or roll the dice" is deeply flawed. I mean, I'm currently playing in an PbtA game (that generally features a "say yes or roll the dice" approach) that focuses on mysteries! (The game is The Between.) So, this cannot be an actual feature of this kind of approach to play and is actually counter to the entire conception.

While firmly agreeing with your later point about styles of play, I personally enjoy established mysteries over stochastic ones, although I certainly lean into facts that are established stochastically. That might be analogous to procedural worlds, that tend to contain a number of hand-crafted sites as typically those can have refined features. Against that, play in the moment can be very exciting the other way.

All that said, the mostly right part of the above is that the standard approaches to D&D are very much that the play is finding out what the GM's conception of the fiction is. This isn't a bad thing at all -- it can be quite fun. It can also be something that people can dislike, as @pemerton does. It's just a feature, and you can like or dislike the feature according to your preference. How it works isn't changed by that preference, though.

Well put: I strongly agree.

 

[Ovinomancer]

I was thinking about that as I wrote. My thought is that players know everything there is to know about the world at all times, but the world can come to contain new things.

I think this is epistemologically sound as I might have a justified true belief at one time that there are three children in my family, and later when our fourth baby arrives the world contains a different fact i.e. that there are four children in my family. At both times, I correctly knew the number of children in my family.

Perhaps, and please forgive, you had an illegitimate child with a previous lover and had no idea of the child's existence? You would then not be correct, and not know if this turned out to be true. However, if it turned out to be not true, then you would have been correct. In SYORTD approaches, the veracity of your belief you had X children might be called into question, and a roll made, and a new truth discovered -- meaning it was wrong at all times prior. Your construction is that everything is entirely fixed, only new additions make changes. This isn't it at all -- what's known is known to the same degree we might be able to know something, but that leaves vast space for things we just don't know yet. This is fundamentally the same between SYORTD and other approaches, the difference is that no one at the table knows what might be found out with SYORTD while with the GM centered approaches, the GM knows (or decides and so knows before the players do).

Great example: it makes the case very clear. So I would be suggesting that when the party passed through here earlier, there existed a collection of facts that at that time did not include a secret door. Players were in possession of those facts. Later, a new fact is added.

If the door was there, the door was there. It's not a new fact, it's a discovered truth -- the fact is that the door was always there. This isn't different from normal D&D play -- if the players go through a dungeon corridor, they have a set of facts. It might turn out a later search reveals a secret door there, and the difference is that the GM put it there and hid this fact from the players whereas in SYORTD, no one knew until the dice say so (or the GM says yes, in which case the door is just flavor).

It seems that we might prefer to say that at no time do the participants know the limit to the number of facts about the room. There is no limit on those facts. Obviously this still contrasts well with one participant being in possession of extant facts the others lack.

The latter case can still turn out to contain yet further facts: unknown to all. That happens a lot in open-world games.

I think you're trying to conceive this in a framework were everything is resolved at any given time, but this is not a sound conception. It's not that things are thus until an event arrives that changes it, it's that things are as they are, and we imperfectly perceive them. This is a key part of a roleplaying game, in my opinion, the navigation of the fiction with imperfect perception towards discovery. What's eligible for discovery is one major distinction between games. (Frex, D&D doesn't feature discovery of character as a part of play, this is left entirely to the player to imagine on their own because perception of your character is never imperfect from within the game, whereas other games put character at risk to enable discovery of character within the game instead of as a meta exercise.) Another, and the one we're talking about here, is how a game operationalize and instantiates both the fiction and how discovery of that fiction operates. 5e and a PbtA game do this differently -- the fiction is instantiated in a different manner (created by the GM vs created through play, respectively) and how that fiction is discovered is also different (taking actions to get the GM to describe the fiction vs taking action that create the fiction for everyone at the same time, respectively). Neither is better, except for preference, they are just different.

While firmly agreeing with your later point about styles of play, I personally enjoy established mysteries over stochastic ones, although I certainly lean into facts that are established stochastically. That might be analogous to procedural worlds, that tend to contain a number of hand-crafted sites as typically those can have refined features. Against that, play in the moment can be very exciting the other way.

You misunderstand. The mysteries in The Between are not stochasticly generated ones, but actual mysteries, that are investigated and solved. The manner of investigation differs, in that successful actions can generate clues, which are statements of facts but cannot ever conclusively answer one of the important questions (like what is this threat, where does this threat come from/lair, what does this threat want, what is this threat's weakness) but instead provides a detail that goes towards it. The players collect these clues, and then can put them together in a way that makes sense to them and answers one of the questions with these in hand (they can also choose to disregard a clue if it doesn't fit their conceptions). The number of clues used goes as a bonus to a special roll that, if successful means that the player's conception of how these clues fit together is true and they can now act using this truth. On a hit (success with complication), the answer is true, but carries an additional problem/complication. On a miss, the GM can be as mean as they like, but it's going to cause a serious problem for the players -- all of their earned clues are now spent and they have to continue investigation plus the threat will react hostilely. This isn't really a stochiastic mystery -- in that it isn't random or a creation of die rolls -- but rather a structured mystery and investigation that involves lots of danger, the need to investigate, but also pairs well with SYORTD by providing a framework but not preconceived fiction to be followed.

Well put: I strongly agree.

 

[clearstream]

Perhaps, and please forgive, you had an illegitimate child with a previous lover and had no idea of the child's existence? You would then not be correct, and not know if this turned out to be true. However, if it turned out to be not true, then you would have been correct. In SYORTD approaches, the veracity of your belief you had X children might be called into question, and a roll made, and a new truth discovered -- meaning it was wrong at all times prior. Your construction is that everything is entirely fixed, only new additions make changes. This isn't it at all -- what's known is known to the same degree we might be able to know something, but that leaves vast space for things we just don't know yet.

It seems then that the facts must be considered modally. It is currently true that in some (not all) possible worlds there is a secret door. The roll establishes which possible world we turn out to inhabit. A "yes" asserts we are in that world. It's not clear why a "no" would not also be useful seeing as it just establishes that we are not in that world? I suppose it may be seen as stifling rather than encouraging.

[EDIT Notice "modally" not "morally". Stupid auto-correct. Additionally, a "yes" or "no" could assert something is necessarily true or false: I leave that out as I don't believe it harms the argument... it's interesting to contemplate, however.]

This is fundamentally the same between SYORTD and other approaches, the difference is that no one at the table knows what might be found out with SYORTD while with the GM centered approaches, the GM knows (or decides and so knows before the players do).

I'm most interested in developing a satisfying blend of DM-decides and Players-invoke. The notion is that players gain narrative fiats via mechanical fiats. Likely extending what can be invoked to tools and perhaps skills. DM remains storyteller, but I wonder if there is a way to put players in possession of some mysteries? About their characters, say?

This is a key part of a roleplaying game, in my opinion, the navigation of the fiction with imperfect perception towards discovery.

That's a part I find tremendously satisfying. Surprises, in particular.

What's eligible for discovery is one major distinction between games. (Frex, D&D doesn't feature discovery of character as a part of play, this is left entirely to the player to imagine on their own because perception of your character is never imperfect from within the game, whereas other games put character at risk to enable discovery of character within the game instead of as a meta exercise.)

This perhaps could change. How might character be put at risk to enable discovery of character, in D&D?

 

[Ovinomancer]

It seems then that the facts must be considered modally. It is currently true that in some (not all) possible worlds there is a secret door. The roll establishes which possible world we turn out to inhabit. A "yes" asserts we are in that world. It's not clear why a "no" would not also be useful seeing as it just establishes that we are not in that world? I suppose it may be seen as stifling rather than encouraging.

[EDIT Notice "modally" not "morally". Stupid auto-correct. Additionally, a "yes" or "no" could assert something is necessarily true or false: I leave that out as I don't believe it harms the argument... it's interesting to contemplate, however.]

Again, I'm not sure that you're situated correctly on this. SYORTD can very easily result in a 'no', but only if that's the result on the dice. Then we all find out together what 'no' looks like. What doesn't happen is the GM unilaterally deciding that the fiction is in this shape and so says no. It's an approach difference -- a different mode of discovery from the GM establishing facts and the players discovering them. Here everyone discovers at the same time. Again, this is just different, not better or even desirable. I like both modes of play, but I'd never want to run a B/X style resource management skilled-play dungeon crawl with SYORTD on the table -- it's not suitable for those goals of play.

As for possible worlds, I disagree -- this is a construct that's not necessary and can be misleading about what's going on in play by introducing a philosophical conceit that isn't needed. It can very easily be one world, and we're just discovering it together. There isn't one possible world where there's a secret door and another where there is not, but rather just this world, and we find out if it contains a secret door together. I mean, to use your construct the normal D&D method is that there's multiple worlds but the GM picks which one applies right now and the players do things to find out which world the GM has selected for this moment. The functional difference here isn't a difference in overall conception of the fiction, but rather what means is used to construct/discover it. Trad D&D uses "the GM decides" while the other uses the dice.

I'm most interested in developing a satisfying blend of DM-decides and Players-invoke. The notion is that players gain narrative fiats via mechanical fiats. Likely extending what can be invoked to tools and perhaps skills. DM remains storyteller, but I wonder if there is a way to put players in possession of some mysteries? About their characters, say?

You already have a ton of this in D&D -- spells and class abilities. Button pushes that result in discrete sets of fiction authoring from the players. The issue really resides in that there's no way you can combine an approach that values and places the GM as storyteller with an approach that devalues and displaces the GM as storyteller. The intention of the types of play are different. System matters.

That's a part I find tremendously satisfying. Surprises, in particular.

I am continually surprised as a player in the Blades game I'm in and in The Between game I'm in. Staggeringly so. Having another person conceive of the surprise first doesn't mean that it makes for better surprises, or that surprises cannot happen otherwise (and not just stochastically, either).

This perhaps could change. How might character be put at risk to enable discovery of character, in D&D?

I think this is absolutely unwanted by the vast majority of D&D players because it requires placing character -- note not THE character, but character -- at risk, and the results can be having to play characters that aren't as you might wish to envision them. This isn't what D&D is really about.

 

[Krachek]

I don't really think that "say yes or roll the dice" is compatible with D&D at all, with the noted exception of 4e. @pemerton tagged me on this awhile back in this thread, but I was taking a break for the holidays (and may continue to do so), so I'll expound a moment using 5e. SYORTD requires that there be a useful and consistent mechanic on the roll the dice side. This is absolutely present in D&D, but only in combat. In fact, combat is probably the closest to SYORTD that 5e gets -- it's almost always roll the dice but the resolution mechanic here is consistent and useful to resolve combat for the most part. Very few instances exist where the GM is expected to deploy "no" to an action declaration. So, here, mostly because the roll the dice aspect is robust, the approach works.

The moment you get away from this, though, it breaks down and the system starts fighting you. This is because 5e places the entire burden of how the mechanics works on the GM's judgement of the fiction. Even if you allow for No Myth, it's still the GM's job to determine the DC of a challenge, and whether or not a single check is sufficient to resolve a conflict. This means that the difficulty of a challenge is arbitrary and fluid and not consistent and further that the expected length of a challenge is similarly so. You can establish rules and principled to lock these in, sure, but now you're adding layers to the game that go directly against it's design intent. Still, this would be fine except that you now have to deal with the player side of the game, and that means how classes are built, how skill bonuses work, and, in general, how the entire character build section is aimed at providing player-side tools against arbitrary GM DC setting that, when the GM becomes constrained, suddenly erupt into player-sided overwhelming of the systems. If I'm fixing DCs on the GM side to make them consistent enough that "roll the dice" isn't an arbitrary exploration of the GM's conception of the fiction, then things like bardic inspiration, expertise, guidance, and even just dis/advantage suddenly become outsized and disruptive. So, now to do this in 5e I have to address this as well and now we're very far away from the designed game and into what I like to call "a different game altogether." I struggle to understand how so many claim to be playing 5e with heavily modified games such that they're actually different games, but this seems to be a thing that's done.

And, all of that above goes to another huge point about why SYORTD doesn't really work in most D&D, not just 5e, and that's how the characters are actually challenged. D&D is built on an attrition model that's extended over multiple consecutive challenges (exception 4e, which was balanced to the encounter). If I want to put pressure on a D&D PC, I have to do this consistently over a period of time, and that time period is pretty uniform within each edition of play. This means that the GM must be able to control pacing in the game if this pressure is meant to come to bear, and if you're controlling pacing you're already outside of SYORTD. The games that do best with SYORTD are ones that allow for short-scale high pressure play that isn't contingent on pacing. BW, for instance, can go from zero to 60 very quickly, and PbtA is entirely structured on the snowballing effect of success at cost and failures being the majority of rolls. These systems, through their very structure, create pressure during play without having to do anything outside of play to introduce it. D&D lacks any real ability to do this. 4e allows for SYORTD because it's pacing mechanisms outside of an encounter/skill challenge are largely weak (with only healing surges being used) so that pacing isn't a thing that the GM needs to pay attention to much at all. You can significantly challenge 4e PCs every single encounter without worry and, in fact, the game works best when you do this. Thus, no need for the GM to control pacing and deal with how that directly confronts SYORTD as an approach.

You realize that you make the emphazing of your favorite play style.

 

[Ovinomancer]

You realize that you make the emphazing of your favorite play style.

Which is that, do you think? I mean, I play and run 5e, and enjoy it. I'm also running an Alien's game, which is similar to D&D in that it's the GM that determines the fiction. I also enjoy playing in and running Blades in the Dark and some PbtA games. Looking forward to some Dogs in the Vineyard and Torchbearer as well. Oh, and I'm looking at some Stars Without Number as a potential game to run. So, which is my favorite style, again?