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

[AlViking]

What? There exists the remote possibility that this is a crazy person tackling me because I'm there and not because I punched out. I am unable to say absolutely that the punch influenced him, because there isn't a 100% chance that it did.

Because being connected =/= caused. You are trying to establish cause where there isn't any.

Naw. I punch a passing stranger instead. Did I cause their deaths? Nope! The fire did. I have no obligation to risk my life to save them.

I treat circumstances the same way 5e treats ability check. I roll if there is doubt as to the outcome. Sometimes there will be an auto yes or auto no. In the case of the cook(or other NPC living there), there's a very, very small chance someone is awake in the middle of the night when the PCs show up, but it's not zero, so a roll would be involved.

It didn't. The farrier was always there, because the town would have had one for hundreds of years. I just forgot to add it.

The player's request didn't cause the farrier to appear. The only thing it caused was me to consider whether one would have been there or not.

Adding details that don't contradict established lore is par for the course It's also vastly different from "A guard shows up because you failed a check."

 

[AlViking]

Are your parents rolling skill checks here? If not, don't be ridiculous. We're talking about causal events in the context of the game.

I am so not going to ask my parents about how well they performed.

 

[Maxperson]

I am so not going to ask my parents about how well they performed.

Whoever you are talking to has me blocked, and now I'm sooooooooo curious about the context of that statement.

 

[hawkeyefan]

Not really, in that downstream connections can often only be observed in hindsight; and we're playing in the here-and-now.

Sure, punching someone in the eye might cause a bystander to try tackling me, and when that attempted tackle goes wrong he sails past me into traffic and gets hit by a bus, and when the bus hits him its sudden stop causes a standing passenger on the bus to fly forward into the windshield, and so on and so on as one thing leads to another. The passenger whose face became one with the bus windshield doesn't know any of this, however; the connections are as yet invisible to him and might forever stay that way.

But all those things could be attributed to the punch in the eye.

If the guy who hot the windshield is in recovery after plastic surgery and someone says “what happened” do you think he’s gonna leave out the fight that caused the accident?

If he said “some idiot punched someone, and then another guy tried to jump the puncher, but he missed and fell in front of the bus” you’d correct him and say “oh no, the puncher had nothing to do with it!”

In game, this all happens one sequential step at a time without foreknowledge of any downstream consequences, because they haven't happened yet. Nobody knows the bystander will get hit by a bus until we resolve his tackle attempt and he wildly misses. Nobody knows a passenger on the bus will get hurt until we game-mechanically resolve what happens to the passengers when the bus makes a very sudden and unexpected stop. Etc.

It doesn’t have to, though. Again, I understand your preference… but there’s no need for us to break down every conflict into as many individual tasks as possible.

In the house-on-fire example, at the moment of my failing the lock-pick attempt nobody knows the people inside will later die because we haven't yet resolved any other attempts to save them.

Why would the GM not know? What if that’s it… this is the only attempt? It’s now or never?

 

[AlViking]

Whoever you are talking to has me blocked, and now I'm sooooooooo curious about the context of that statement.

"Are your parents rolling skill checks here? If not, don't be ridiculous. We're talking about causal events in the context of the game."

 

[hawkeyefan]

The only reason the guard approached was because you failed the test. Had you succeeded the guard would not have approached.

How do you know that?

Because it isn't the same situation. The cook wasn't always in the kitchen, or she would be there even if the roll is successful. The fact that she is there if it fails and not there if it succeeds makes it very different.

First… you’re making the same mistake as @AlViking above. You’re assuming you know what would have happened on a successful check. You don’t.

Second… it’s the same situation. A player did something and that prompted the GM to introduce an NPC to the situation. In one case, it was a question about an NPC, in the other, it is a die roll.

For some reason, you’re able to adapt the fiction to the farrier so that he existed all along, but not for the cook that she was awake or alerted in some way by the attempt to pick the lock.

Because a successful roll would prompt the same thing if she was there all along. That's how, "there all along" works.

But the farrier wasn’t there all along either. He was made to exist through the players’ request.

All kinds of things in play only get established in the moment as needed. It’s perfectly normal for this to happen.

 

[Maxperson]

First… you’re making the same mistake as @AlViking above. You’re assuming you know what would have happened on a successful check. You don’t.

Alright. I'll bite. How often do things go badly with a success?

Second… it’s the same situation. A player did something and that prompted the GM to introduce an NPC to the situation. In one case, it was a question about an NPC, in the other, it is a die roll.

It's not at all the same thing unless the cook would be there regardless of success or failure, and in the same mental state.

But the farrier wasn’t there all along either. He was made to exist through the players’ request.

No. If I had thought about him when I made the town, he would have been there. He wasn't made to exist, I just thought about him and gave him a location when they asked.

Being made to exist through a player's request would be if I knew he wasn't there, and I made him because a player asked me to. That brings him into existence at the player's request.

All kinds of things in play only get established in the moment as needed. It’s perfectly normal for this to happen.

I know this. How it happens is what is at issue here.

 

[Enrahim]

Tricks of the Trade​

When you pick locks or pockets or disable traps, roll+DEX.
✴ On a 10+, you do it, no problem.
✴ On a 7–9, you still do it, but the GM will offer you two options between suspicion, danger, or cost.

This move does not permit the Thief to declare what is on the other side of the safe, so the example is already negated before we even begin, as I have previously said, just in generic terms.

Thank you for your extended example! I think I now see where communication went wrong. I followed the reply thread quite a bit up and see that at some point @pemerton at least we're talking a bit about systems with intent based resolution where success guarantees intent. The Desert Rose problem was formulated in this context. It appear to me like you saw this desert rose example and interpreted that as somehow as a critique against narrative games in general. You also likely saw the related discussion about the distinction between task and intent and didn't realize this wasn't about how to describe outcomes of a check, but rather how to determine the outcome of a check.

I think you have now solidly demonstrated that DW (at least in the pick lock case) do not have this system property. In other words this criterion do not apply:

Also you were yourself introducing "picking a lock to get at the documents inside the safe" as a valid example for where this technique could come at play - ensuring that the docments will be still there on a full success. So pick your favorite game where this example would be a valid enforceable play.

And as such this is not a game that presents the problem @Lanefan was struggling with.
There do exist systems with this property though. I think Lanefan's observations related to how implementing such a system is detrimental to hide and seek play is not something that can be easily dismissed, the way it appeared you were doing it

 

[hawkeyefan]

Alright. I'll bite. How often do things go badly with a success?

I didn't say things go badly with a success.

I said that it's possible the cook is still there. She may be in the kitchen, unaware that anyone has just broken in. She may be near the kitchen, in her quarters sleeping, unperturbed by the break in attempt.

It's not at all the same thing unless the cook would be there regardless of success or failure, and in the same mental state.

But that's the thing... only one thing happens. The lockpick attempt either fails or succeeds, the cook is either there or not. We don't know what the situation would be if the opposite happened because the opposite didn't happen.

I'll add the caveat that the example as presented on the blog implies this... but as has been pointed out, that example is pretty much doo-doo as presented.

@AlViking is making the same mistake with his take on @pemerton 's example about the singing attracting the guard. He wants to know if the guard would have shown up if the roll for the singing had been a success. Who can say? That's not what happened, so we don't know how it would have gone.

No. If I had thought about him when I made the town, he would have been there. He wasn't made to exist, I just thought about him and gave him a location when they asked.

Being made to exist through a player's request would be if I knew he wasn't there, and I made him because a player asked me to. That brings him into existence at the player's request.

But no one claimed to know the cook wasn't there. A kitchen in a noble's manor certainly implies a cook... just as much as a town implies a farrier (maybe even more so?). She's not "made to exist" either... she's "always there", just as the farrier is.

 

[EzekielRaiden]

Thank you for your extended example! I think I now see where communication went wrong. I followed the reply thread quite a bit up and see that at some point @pemerton at least we're talking a bit about systems with intent based resolution where success guarantees intent. The Desert Rose problem was formulated in this context. It appear to me like you saw this desert rose example and interpreted that as somehow as a critique against narrative games in general. You also likely saw the related discussion about the distinction between task and intent and didn't realize this wasn't about how to describe outcomes of a check, but rather how to determine the outcome of a check.

I think you have now solidly demonstrated that DW (at least in the pick lock case) do not have this system property. In other words this criterion do not apply:

And as such this is not a game that presents the problem @Lanefan was struggling with.
There do exist systems with this property though. I think Lanefan's observations related to how implementing such a system is detrimental to hide and seek play is not something that can be easily dismissed, the way it appeared you were doing it

Well, more my point was that this is...a pretty easily foreseeable problem. As in, essentially every designer of a narrative game is going to see this possibility many miles away. Long before any books hit the table, designers will already know that absolute total freedom where ANY not-explicitly-nixed intent is 100% perfectly acceptable always forever...is going to lead to bad results.

So harping on this "problem" is a little bit like someone harping on the idea that a class-based game will forbid players from playing archetypes that don't have a specific class that fits them. Of course that's correct, the designers DEFINITELY already know that, and the design will attempt to respond to this. (Or, I guess, they can just say "it's not a bug, it's a feature", but I don't know of any games that do that for this task-intent resolution process.)

Hence, while this is theoretically a problem for some possible, not-yet-realized game...it's functionally never a problem in any actual games because they will be designed in such a way that, like DW, they won't permit egregiously stupid inputs. If egregiously stupid inputs are prevented, then the question becomes moot, in an admittedly uninteresting way, because the answer is simply "you aren't allowed to do that".

Again I refer back to the "how much damage does a harmful action do to an enemy?" question I raised earlier. There are many, many different ways any given game could answer that question; indeed, most games will answer that question differently, and will often have many narrower, more-specific answers, depending on the specific context of the situation. E.g., the amount of damage a regular ol' attack roll does is, if not fixed, then at least mostly so, barring particularly special circumstances (like a fancy magic item, or a temporary buff spell)--but the amount of damage a cantrip does will vary depending on the level of the character casting it. Hence, it simply is not possible to provide a clean, simple answer to the question, "In a D&D-like TTRPG, how much damage would be dealt after a successful attack roll?" You would need to know MANY additional pieces of information--whether it was a weapon or a spell, whether it was a magical weapon (or enhanced by a magical implement/focus/etc.), what the level of the character is, what the character's stats are, whether the creature has relevant features like resistance or weakness, etc., etc., etc.

But all of that does not, in even the slightest degree, mean that, as you previously asked...

So you are saying the group need to engage in basic game design in order to play this game?

Because no, the players in D&D 5e do not need to engage in basic game design in order to play the game. But an answer that would be correct for 5e almost surely will be incorrect, to at least some degree, for 4e or 3e or 2e or 1e, let alone systems that aren't even D&D but are in the same wheelhouse. Likewise, the players of DW or Burning Wheel or Masks or what-have-you do not "need to engage in basic game design in order to play [that] game", but we cannot--even in principle--give you an answer that would be correct and fully specific, yet applicable to every one of these games.

Each game addresses the process in different ways, but all of them have the same general requirements, namely that:

• Only reasonably well-established fiction qualifies as the basis for a valid intent
• It is a conversation between players and GM which determines what has been established and the degree to which it has been so
• Players have some ability to establish new fiction, but that ability is often subject to rules and requirements
• Both GM and players alike need to make only reasonable extrapolations--hence, as DW puts it, your move must "follow from the fiction" (on the GM side)/"you have to do it to do it" (on the player side)

In other words, effectively every game has in place things which simply make the described example non-applicable. I have to say "effectively every game" rather than just flat "every game", because cannot promise that NO game ever in the history of time has ever made this kind of profoundly stupid, bad, harmful game design choice. It's certainly possible, people do stupid things all the time. But I can assure you that, if it ever did happen, it would be an extraordinarily unusual example. Again, the equivalent of a D&D-alike game which for whatever unfathomable reason just...never told the players how much damage their attack rolls would do. As I'm sure you'd agree, it is theoretically possible for a D&D-clone type game to do that--but it would be so profoundly foolish on the designers' part, the odds of this actually happening are effectively zero.