D&D General [rant]The conservatism of D&D fans is exhausting.

[hawkeyefan]

Right, this is what I described. Is my understanding accurate, then?

It was unclear to me if you were expecting a second action or not, that's why I clarified. But yes... a hard move should follow a soft move. So... I'd establish that someone's in the kitchen, on the 7-9 the character would get inside, but alert the cook. Then I'd ask what they do... and the hard move or not would depend on that action and roll.

If this is what you meant, then yeah... I'd say your understanding is accurate.

 

[Maxperson]

How big of an area are we talking about? Is it like a wilderness hex? Or more specific than that?

I mean... how can you predict with that level of accuracy where the PCs will actually physically be in that forest? How do you know to place the trolls on that exact spot?

Or is it more a case of the PCs are in this big area and so are the trolls, so they will encounter one another?

You aren't reading to understand, just to respond. Try reading to understand. If your party is going through the forest, wandering monster rolls happen at specific intervals. Let's just say 2pm since I don't want to look it up. If the roll indicates an encounter, it will be where at the location in the forest you are at at 2pm, not anywhere else. My rolling it in advance is only so that I can make the encounter more enjoyable if a wandering monster is indicated. The party is still where they would be at when those trolls wander through, not anywhere else.

It has nothing to do with the size of the area, or any sort of predictions. Nothing is quantum for the players.

Okay, I somehow missed that. I suppose you made the situation such that the Fail Forward was more absurd.

I did not. I used a scenario that made sense, not one designed to make fail forward more absurd. If you feel that it is somehow more absurd because of a perfectly sensible wizard tower set-up, then the issue isn't with the set-up. Perhaps it's just that fail forward and prep don't mix well.

Meanwhile, I have similarly fleshed out the scenario in a way that makes it work. Weird.

Not so weird. I'm used to folks trying to craft things so that their way prevails.

No, just nonsensical and the like!

It does sometimes end up being "nonsensical and the like", but not because the style is inherently bad, but rather because it doesn't play well with prep, and when you mix them, nonsensical things can happen. You noted as much with the perfectly reasonable wizard tower set-up and fail forward.

 

[pemerton]

Second, it's the middle of the night when the cook would be in a deep sleep, yet the roll teleports the cook to the kitchen awake and ready to spot the party, because pick lock fail. It's not as if being in the kitchen during sleep hours is some routine thing for the cook. There's almost no chance of the cook being there.

This goes to @Campbell's point. Someone who doesn't like "fail forward" gives an example of clumsy GMing - narrating the cook in the kitchen with no foreshadowing/soft move/other rationale, and when it makes little sense due to it being the middle of the night - and this is supposed to show that "fail forward" can't work!

As I've posted, this is like me "proving" that dungeon crawling can't be interesting or fun by adducing a bad, boring dungeon I wrote when I was 12.

 

[Micah Sweet]

And none of the comments made here since pointing out why the example you sited was incomplete or how it could be made to work have done anything? I mean, maybe they haven't changed your mind, but you can at least onboard the information and stop describing things as "unconnected" and "random" and so on.

And "random-ass stuff" wouldn't really be any better. Know why? Because... again... a cook in a kitchen isn't really random.




Again, those need not be unrelated. If you're picking a lock in the real world, what are some of the possible consequences?

I would think that they would include any number of the following:

• failing to open the lock
• damaging the lock in some way
• damaging the lock pick in some way
• being discovered while picking the lock

These are all things that when someone decides to pick a lock, they're going to be worried about in some capacity.

None of those consequences, if they're realized, are unrelated to the lock picking.



In some games, the players make all the rolls. The ones I've had the most experience with recently are Blades in the Dark, Stonetop, and Spire. Each of them allows for degrees of success depending on the outcome of a roll. Critical Failure, Failure, Success With Complication, Success, Critical Success.

So depending on the outcome of the roll, the GM has NPCs act accordingly. Sometimes this is dictated in specific ways. When you "Clash" in Stonetop, for example, you roll 2d6 and add your Strength stat (ranging from -1 to +3). If you roll a 6 or less, you're taking your enemy's damage. On a 7-9, you hit your enemy, but they also hit you. On a 10+ you hit them, and you can either avoid their attack or choose to do some extra damage.

In Spire, you roll a dice pool of d10s based on applicable stats and situational factors. You'll roll anywhere from 1 to 4 dice, and keep the highest roll. If it's a 1, you suffer your enemy's attack and take twice the normal damage. On a 2 to 5 you fail and suffer your enemy's attack and take damage. On a 6-7 you succeed and inflict damage to the enemy, but also take theirs. On an 8-9 you succeed, inflicting damage and taking none. On a 10 you inflict double damage.

In these kinds of games, it is expected and required that a dice roll does more than mitigate just the task of the PC. The game won't function if that's not the case since the GM never rolls. This is why some folks, notably @Campbell , have cited how poorly the thief/cook situation works as an example. It assumes so much about the way the scenario is constructed and presented.



Personally, I'd have hinted at the presence of the cook (or someone) beyond the door before the roll was made. Part of making Fail Forward and similar techniques work is telegraphing things... establishing stakes before the roll is made. A flickering firelight seen beneath the door, the sound of footsteps, or someone humming while they work... anything like that, and then the cook is just "offstage" but ready to be introduced.



Again, you're failing to acknowledge everything that has been said about that example since it was first shared.



Why not? According to what game?

For me, they easily can be... even if I'm playing D&D.

They can be in D&D, if you decided that skills are different than D&D presents them. Sounds like a houserule. Nothing wrong with that, but that's what it is IMO.

 

[AnotherGuy]

What I don't want is a complication after every single failed check. It would feel artificial to meme.

That's fair. I'm not advocating for rules change just exploring in what situation a Fail Forward could work.

 

[Faolyn]

Your stat and things on your sheet are not what people are talking about. It's looking into a window or sending a familiar in to do some scouting or listening at the door. Stats and abilities will always apply no matter how careful you try to be.

And as I've said that, you can do all that as well.

If you look into a window or send a familiar in or whatever and determine the room is empty, then the GM isn't going to poof someone into existence. Not a GM with any sense of continuity, at any rate.

 

[pemerton]

Well, no. You could listen at the door or try to get the work schedule ahead of time or ask (or coerce) the night watchman if anyone came in at an odd hour or send a druid to scout in advance or scry or smell if anything is cooking...in short there are a lot of choices you could have made to get that information.

But in the lock picking case, none of these would reveal the cooks presence because the cook wasn't present.

This last claim seems to me to be false. Or rather, to be true only if the GM is bad - that is, if in response to the info-gathering moves the GM says <this stuff>, and then when the time comes to follow through with a hard(er) move, the GM ignores all <this stuff> and instead introduces the cook like John Harper's ninjas from the ceiling.

Yes, if the GM is bad the fiction won't make sense. I'm not sure what light this sheds on "fail forward" as a technique, though.

There are bears in the forest, whether or not you happen to see one is quite random but they exist whether you see them or not. The cook only exists because of a failed check.

Specifically, Dremmer's cook Pattycakes was only there because of a 7 to 9 on Acting Under Fire. On a 10+, no Pattycakes.

Again, these claims seem to me to be false. They seem to equate meeting the cook with the cook existing. Whereas the cook is no different from the bears.

There is no explicit connection and that is the point. The world exists independently of the players.

In the failed die roll case, the encounter exists only because the players failed. It is entirely dependent on the players.

I'm confused. Quantum was being applied to the cook (or Pattycakes), not the technique. I guess it could apply to the kitchen too. The thief observes the kitchen, and only then is Pattycakes there or not there. Hence, quantum.

In the fiction, there is no "quantum". The bear exists; and it meets the PCs. The cook exists; and they meet the PCs.

In the real world, the GM narrates a meeting with the bear because of a roll. (The encounter check.) In the real world, the GM narrates a meeting with the cook because of a roll. (The player's failed roll.)

And a bit more on "quantum" beings, encounter rolls, and players' failed rolls:

Not the way I've seen it. There's a limit to how many of any given creature is available to run into on the move. Or at least IMO there should be.

So there is a limit? Or there should be a limit?

I don't think Moldvay Basic talks about any limits. Gygax's example dungeon in the DMG has a boutique wandering monster table: some of the monsters are attributed to particular locations (eg skeletons patrolling from Area 27) but it doesn't say anything about limits.

I think that's true for some older mods, but it's not universal, and I prefer ones where it is not the case. That said, even if there is an endless supply, the existence or nonexistence is still fixed--occurring with a fixed probability--rather than depending on an unrelated skill roll.

the cook isn't rolled on a wandering monster table, it's appearing in response to a completely unconnected pick lock roll.

Unrelated or "unconnected* only on the premise that the point of the roll is not to find out whether or not the attempted action goes well or poorly. If the point of the roll is to find out whether or not the attempted actions goes well or poorly, than the attempted action leading to an unhappy encounter is related. As that is a way of the attempt going poorly.

Which brings me to:

Here, meaningful decision making. If they exist beforehand than a different decision can bypass an obstacle. If they only exist after the roll, then my choices weren't so important -- the roll was.

As I said "...and decided that the most fictionally appropriate complication...", you had to come up with a complication because the rules said you had to do so. That's what I wouldn't care for, it would feel artificial to me as GM and player.

"Fail forward" as a technique - the avoidance of narration of "nothing happens" - is not designed for puzzle-solving or exploratory play, where the players try and overcome obstacles presented by the GM in order to achieve some goal which lies beyond those obstacles. If that is the sort of RPGing you want, you therefore won't use "fail forward".

 

[pemerton]

this was part of my problem with narrative games. The chances of success don't really vary based on approach--as long as you roll well, almost any approach will do--so it became a game of "how can I justify to the DM that I'm using Hunt here?"

This is pretty different from my experience.

 

[pemerton]

Now, BW maybe works a bit in the opposite direction, but then I don't think BW players are justifying anything to anyone, they're just doing stuff.

If a player wants their PC to hunt, their PC can go hunting. What exactly are they doing? What's their intent? Then we set an obstacle and resolve it.

If you aren't hunting, you won't be rolling your hunting skill. You'll be rolling whatever other skill maps to the task that you're attempting.

 

[pemerton]

In a fail-forward scenario, the consequence for failing forward is necessarily one the GM would consider plausible, so it is natural that the GM would be prepared to explain why the consequence is plausible.

What you wouldn't see in a fail-forward scenario is a situation like the one here, where @AlViking provides an example from a website, than complains that the situation isn't plausible, then when examples of why it might be plausible are raised, searches for reasons why the situation isn't plausible.

Right. The "refutation" of fail forward by presenting examples of it being done badly - fully-staffed kitchens in the middle of the night; cooks introduced although all prior fiction suggests there is no cook in the house; etc - is bizarre.