That's clearer, thanks. So the issue isn't so much dice rolling as who gets to frame checks?
Eh,
kinda, in a way, but not really.
That makes it sound like I'm staking out procedural turf that I don't want players to encroach. Ideally, what I want in a game, whether playing or DMing, is an interplay between the players and the DM in which the DM presents the world, and the players engage with that world, through the DM, as much as possible remaining "in character" -- that is, making their decisions through the eyes of their character. Taken to an extreme, this is ultimately an exercise in improv. It's summed up by this description of play in the 5e rules:
5e Rules said:
1. The DM describes the environment. The DM tells the players where their adventurers are and what’s around them, presenting the basic scope of options that present themselves (how many doors lead out of a room, what’s on a table, who’s in the tavern, and so on).
2. The players describe what they want to do. Sometimes one player speaks for the whole party, saying, “We’ll take the east door,” for example. Other times, different adventurers do different things: one adventurer might search a treasure chest while a second examines an esoteric symbol engraved on a wall and a third keeps watch for monsters. The players don’t need to take turns, but the DM listens to every player and decides how to resolve those actions.
Sometimes, resolving a task is easy. If an adventurer wants to walk across a room and open a door, the DM might just say that the door opens and describe what lies beyond. But the door might be locked, the floor might hide a deadly trap, or some other circumstance might make it challenging for an adventurer to complete a task. In those cases, the DM decides what happens, often relying on the roll of a die to determine the results of an action.
3. The DM narrates the results of the adventurers’ actions. Describing the results often leads to another decision point, which brings the flow of the game right back to step 1.
Now, within that framework, it's understood that for certain situations (some dictated by the rules, others left to the DM's discretion), randomizers are used to simulate swings in probability, and to create outcomes that I do not know beforehand, in order to give the game a more "live" feel, reduce mental workload on the part of the DM, and to take certain outcomes out of the DM's hands, so that the game doesn't consist solely of the "convince the DM" type dynamic that Tony Vargas described below. If improv is one end of the spectrum, and a rules-bound game such as Monopoly is the other, we're closer to the improv side, but with using the rules and mechanics to
facilitate that kind of play.
A player who unilaterally makes a check, even when doing so with an "in character" description of their action, is breaking away from the above dynamic. They are not engaging with the world through the DM, they are engaging with the world through the mechanics. What that means is that the world is now dependent on, and to an extent limited to, those mechanics.
Now, let me add the caveat that it is of course entirely possible for a DM to ignore that engagement with the mechanics, and focus on the accompanying description. I did this to an extent with the B/X example, saying at one point, "You don't have to roll. A quick glance reveals there are clear tracks." In practice, though, I do not favor this. There is perforce what Neal Stephenson calls "metaphor shear" when that happens. If the player rolls low, and the DM sets the roll aside, there is typically no problem, because the favorable outcome tends to compensate for the metaphor shear. (Of course, if this happens too often, players may feel that the DM is fudging in their favor.) If the player rolls high, then a negative outcome accentuates the metaphor shear. Overlaying all this is the possibility, nay, the likelihood, that the player is explicitly engaging with the mechanics because that's the kind of game they want. As DM, I feel I should oblige them, though I feel the game as a whole has been to some degree limited, and my handle on the world (
not control) has been compromised.
To give another example, I was recently running a game that involved a social interaction. The players were trying to get information about a particular person from an NPC who had been seen with him. The NPC they were talking to was improvised on the spot (as will happen), so I had some idea in my head of how he would react to certain lines of questioning, but nothing concrete. After initially talking with the character reasonably, one player suddenly had his character go ballistic, viciously threatening the NPC, capping off his tirade with a unilateral dice roll, exulting when it came up quite high.
Now, this posed several problems for me. For one, it caught me entirely off-guard. Had there been no dice roll, I could have digested his role-play, and formulated a response. In hindsight, with detached perspective, I can say I probably would have called for an Intimidation roll. But in doing so, I would have already been formulating possible results based on the outcome of the roll. And since it was some good role-play, total failure would not have been on the table, even if he'd rolled low. But in the heat of the moment, my mind went blank. Not only did I have to react to the role-play, but I also had to think, on the spot, of a response that more or less reflected his high roll. Ultimately, I fiddled with my dice a bit, buying some time, and eventually came up with a response that, I think, worked for everybody and the game continued. But, personally, as participant in that game, man, did that suck. Metaphor shear up the wazoo for me. Instead of working through the interaction organically, I was put to a decision, my options limited.
Looked at from an opposite point of view, it would be typically thought to be bad DMing to do the same to a player, which is generally why bluff, persuasion, and intimidation checks aren't usually used against a PC, except perhaps in the rare case of player-initiated contexts. Or at least, that's how I see it. I like skill checks as a tool to resolve certain situations, not as weapons for the PCs and DM to wield against each other.
I think the "agency" aspect to this arena of play is fairly important - there is an element of the players wanting to be able to make a move (in this case, a mechanical move at the "meta" level) that obliges the GM to dispense some backstory. (Or, in the case of, say, an Athletics check, that obliges the GM to allow that the fictional situation has changed in the relevant way.)
Indeed. That desire to operate in the "meta" level is the distinction that I (and I think @
Sacrosanct) were trying to make.
I also think there is a degree of "training", which is linked to certain play assumptions. For instance, to the extent that there is an implicit understanding (among some players, perhaps promulgated by some game texts) that the mechanics are the "physics" of the gameworld, then you may have players who think they need to work out the physics in order for the fiction to change; or GMs (not you, obviously) who insist on the roll as part of working out the physics. (This came up in the big "fail forward" thread, for instance, where some posters took it for granted that climbing a mountain, or trying to navigate, must involve a check against Climbing skill, or Survival skill, etc.)
This is also an excellent insight, I think. Thinking back, it seems to fit my experience. Especially as a B/X player and DM, since the mechanics there are most certainly
not physics simulators.
Personally, and probably more controversially, I also think some of the explanation is the lack of clarity around the rationale for various dice rolls in classic D&D, which can then lead to some of these rolling procedures taking on almost a cargo-cultish status (in the sense of: players are going through a process, or at least what they take to be a process, without that process actually having any sort of appropriate causal connections or larger context that renders it sensible).
Well, certainly the ultimate roots of that kind of play go back to the Thief class, and the steady encroachment since then that if you don't have a particular skill, you
can't do that particular thing. This was reinforced with NWPs in late 1e and 2e, and brought to full flower with 3e. That said, I'm not exactly sure if that's germane to the issue I'm discussing, in that while this shows an evolution of the skill system in D&D, it is my impression that even through 3e skill systems were represented as a resolution system used at the DMs discretion, rather than features for players to use unilaterally. 4e Essentials
did introduce "skill powers", but I suspect that was merely a sign of where things were going, rather than a cause in and of itself. I could be wrong. I never played 3e, so I don't know if how it was presented or commonly played might have contributed to this trend, other than, as you note above, foreground mechanics as player-side tools of agency as well as physics simulators.
For instance: why does classic D&D require a roll at all to hear noise behind doors? Is this primarily modelling the thickness of doors? Is it about rationing information in order to enhance the play experience? And if the answer is the second, and we then read that back into the fiction by positing thick doors (which also helps explain the STR check needed to open them), why does a thief have to make a roll to hear noise at a window (which is typically not thick at all)?
The answers to these questions aren't really made clear in the Moldvay rulebook, and even less so in Gygax's DMG, and to be honest I don't think they're that much clearer today. The rules talk about checks to hit in combat, and checks to notice hidden things, as if there's no interesting difference between them other than the stat used; whereas the former is primarily about making a move in the fiction, while the latter is about increasing the backstory available to the players, which I think are clearly quite different things that might be governed by quite different considerations (as your quotes from Mike Carr seem to suggest).
To be frank, I think they are there merely to model probability, and I think this is made somewhat clear in Gygax's DMG, which opens with a discussion of probability and bell curves. That said, I don't think it had much of an influential effect, as from OD&D through 3e there was a constant movement toward having mechanics that "model" something concrete, rather than just mere probability. The example
par excellence for this is the attack roll. In OD&D (and default Expert D&D), the attack roll merely represents the probable chance of doing some damage within 1 minute/10 second round of combat. Within that 1 minute or 10 seconds, the participants were expected to be busy with multiple attacks, multiple hits, multiple misses, multiple dodges, multiple parries. This was made explicitly clear in OD&D, AD&D, and XD&D. And yet...first came variable weapon damage, then weapon speed factors, then 6 second rounds, then attacks of opportunity, etc., etc., until now, even with 5e paying some lip service to the idea of multiple undescribed actions within a turn, the game
feels like 1 attack roll = 1 swing. To the point that martial characters get Extra Attacks.
I agree that more-or-less standard/ordinary RPGing works best when the GM has a clear authority over framing. But I'm not surprised that there are some players who push back against this, because the boundary between GM framing and GM force/adversarialism can be a fairly murky one.
When I re-read Gygax's DMG, one of the things that strikes me are the references (not all that many, but enough that I notice them) to the GM sticking first-and-foremost to what is written down in the dungeon key. To me, this seems central to ensuring that the game is a fair puzzle-solving enterprise for the players, rather than something closer to Calvin-ball.
Anti-scripting, "No Myth"-ish games of the sort I tend to enjoy also have all sorts of formal and informal devices to regulate the way the GM frames things and introduces adversity (MHRP - the Doom Pool; BW - GM's obligation to frame by reference to player/PC flags; 4e - much more informal, but the players tend to have enough depth of resources, espeically at paragon and up, that the GM only has to notice the encounter-building guidelines out of the corner of his/her eye to have a sense of what will or won't be fair to toss into the mix).
But there is a lot of D&D play that has drifted away from the constraints of Gygaxian dungeon crawling without adopting other sorts of devices for helping to make play fair. (Luke Crane suggests that this break down of fairness happens even with the move from Moldvay Basic to Marsh/Cook Expert, because the context for wilderness play and wilderness mapping doesn't give the players enough opportunity to construct the information they need, unlike the much more constrained environment of dungeon mapping.)
The default solution - that probably had its heyday in the 2nd ed AD&D era but which seems to find at least some support in 5e - is to put all this back on the GM. That's a big load for one participant to carry, even if well-intentioned, and does put pressure on the framing-vs-adversarialism distinction.
Well, I don't necessarily agree with Luke Crane on the Expert Set. I mean, I agree that Moldvay Basic puts forth a clean, open-but-with-structures game that when you play in that style, creates a certain visceral experience, well described in his posts on his game (and then emulated with Torchbearer); and that Expert does not offer that kind of play. But I feel that perhaps Crane does not extend Expert the same open mind that he does to Basic, in that he seems to view Expert through the lens of Basic. I think Cook/Marsh Expert is entirely as fair as Basic when played in the style suggested. A discussion for another time, perhaps.
We've discussed this before, but I think a key weapon in the arsenal of a D&D DM to avoid adversarialism is random rolls. I think this was in full bloom with 1e and Expert, but began to be downplayed with 2e, as the game moved from "DM as Referee" to "DM as World Builder and Storyteller" (and also from "adventure module" to "campaign setting". Happily, while 5e has moved back to putting more responsibility on the DMs shoulders, they have lightened that burden in two ways. One is the increased reliance on campaign-length adventures, such as HotDQ, PotA, OoA, etc. These provide new and casual DMs with the structure and advice with which they can run the game. The other is to lighten the homebrewing DM's burden by a great many random tables.
But all this is far afield of the topic Sacrosanct was posting about, I think.
