I wasn't disagreeing about D&D, just expanding on some things I had been thinking about.
<snip>
Here's the different between the guard's reactions and the widget. The reaction or consequence to a social action has an immense range of possibilities. The GM has a lot of leeway to respond appropriately to the player's actions in a bunch of different ways, and thus has a wide range of options vis a vis narrating both success and failure. With the un-found widget, it's either there or it's not. The only option the GM really has is to move it by sleight of hand, or continue to frustrate the players. I'm not judging one of those options as better than another either, sometimes player frustration is a good thing. The range of possibilities that result from an action like I ask the guards if there is way I can get into the castle is pretty manifestly different from I search the maguffin for the widget.
The way that the D&D rules handle that is still very different from some other games though, and different GMing styles and play styles also change the stakes a lot.
I've put the two quotes side-by-side because I'm not 10% sure if, in the bit I've bolded, you're talking about D&D in particular or RPGing in general.
For the conventional way that D&D is adjudicated, I think that what you say is true, because - in the conventional way that D&D is adjudicated - the GM responds to the
We search , , , action declaration by consulting his/her notes and answering either yes or no. If the GM doesn't have any notes, then the conventional approach is to extrapolate in imagination from what has been noted, and again on that basis to answer either yes or no.
Of course the same reliance on actual notes, or on "virtual notes" established by way of extrapolation, is also possible in the social context. Gygax's DMG, for instance, contemplates that the GM might have a note written about dungeon occupants that
they always attack. And some monsters have similar notes in their descriptions - eg from memory, AD&D kobolds always attack gnomes. But I agree that, in general, it is less common for the outcomes of social interaction (compared to searchings for widgets) to be resolved in D&D by reliance on or extrapolation from pre-authored notes.
But if one is talking about RPGing in general, then I don't think that the range of possibilities is different in each case. A check made to resolve searching might (if successful) result in the thing being found. In a "success with complication" system
or a "fail foward" system, it might result in finding the thing sought but (eg) not quite in the nick of time, or in a damage condition, or . . . . On a failed check, there is the possibility of simply not finding it, of finding something else instead that is not desired or is a sign of trouble, of having the search interrupted before the thing is found (so maye the widget is in the place being searched, if ony the interruption can b dealt with), etc.
As I said, we may be ad idem on this. I've made this further post about it mostly because it is directly relevant to the discussion about BitD adjudication.
I don't see how having a reduced frequency of unmitigated success can be described as "competence."
The fact that the PCs are rarely going to succeed without screwing something up in some way is a different thing, and that seems pretty clearly to be the case.
It just seems to me that some of the sorts of complications that have been mentioned seem like the sorts of things the PCs should have learned about as part of their planning.
your criminal-as-a-lifestyle character either overreached his competency or didn't adequately plan for something. Even your example with the opponent in the knife fight kinda bespeaks something other than full competency (slash, not stab, hold the knife so it won't come out of hour hand, go for the quick kill).
If the fighter's failure to one-shot the beholder directly caused him damage of some sort, or resulted in his being prone at the feet of the beholder's ogre minions, that'd be a fairer comparison.
I've bolded what I think is quite wrong.
I'll explain why, using a D&D example.
Some people play 3E D&D using a "players roll all the combat dice" variant. I think it might even have been expressly spelled out in the DMG, or perhaps in a WotC supplement. In one version of this variant (there are varioius mathematically-equivalent ways of setting it up), instead of
the GM rolling to see if a monster or NPC hits a PC, by making a d20 roll, adding the monster's/NPC's to hit bonus, and then comparing that to the PC's AC, the player rolls a d20, adds his/her PC's AC, and if the result is less than the monster's NPC's to hit bonus + 22 then the PC is hit by the monster/NPC.
I have never heard anyone suggest that, if a player in a D&D game fails a defence roll, and hence suffers damage from a monster/NPC, this shows that the PC is
not competent or
has screwed something up. A failed defence roll doesn't mean that, in the fiction, the PC
directly caused him-/herself damge eg by hurling herself onto an enemy's blade. It means that, in the fiction, in that particular exchange of blows, the PC failed to so dominate the exchange that the monster/NPC was completely unable to wear the PC down.
Now in a system where the players roll all the dice, you could combine the attack and defence roll into one roll, resolved on a table (as a traditional wargame would) or via a spread of probabilities. This would not change the fiction. It would just reduce the number of dice rolled. In D&D, and moving beyond the realm of combat resolution, 4e skill challengs are something of an example of this: the players roll all the dice, and failure on a given check may indicate that the PC did something wrong, or that some other actor (a monster/NPC, the environment, etc) did something which hindered or even set back the PCs' efforts. The overall question of whether the PCs succeed or fail in what they are trying to achieve is not known until
all the checks have been made.
In BitD, or other PbtA systems that use the "7 to 9" result, the move that the GM makes when the "partial success"/"success with complication" result occurs may reflect that the PC did something less than perfectly, or that some other actor did something which hindered or even set back the PC's efforts, or simply that the circumstances were not as propitious as the PC had hoped when the conflict started unfolding. The notion that, because the player is rolling the dice, it must be the case that
the PC directly caused some problem for him-/herself is completely without foundation. And a GM who adjudicated a PbtA game in that fashion would make the game suck
even worse then a GM in a players-roll-all-the-combat-dice 3E game who narrated every failed defence check as
Sorry, Redgar, you've impaled yourself on the hapless orc's spears once again. And that latter would itself suck pretty badly.
As far as the suggestion, specific to BitD as a heist-oriented game, that
some of the sorts of complications that have been mentioned seem like the sorts of things the PCs should have learned about as part of their planning, I'm pretty sure that BitD has "flashback" and other planning-oriented mechanics that deal precisely with this issue - allowing the players to spend resources to reduce what might otherwise be the severity and impact of GM-narrated consequences.
Equally, however, a heist-oriented game in which
nothing ever developed in an unforeseen way would probably suck. Think of Sean Connery in The Great Train Robbery - he hadn't factored in that the soot from the smoke would soil his clothes as he walked along the top of the carriages, necessitating him borrowing Donald Sutherland's jacket which in turn results in him being apprehended.