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

[Faolyn]

It seems to me that there's at least 2 lines of argument against a fail forward model, which I'm categorizing as "the argument from naturalism" and "the argument from gameplay."

I'm frankly less attached to the argument from naturalism, at best I think it can serve to make the game state more understandable to the player and avoid negotiation. The idea is that the situation should unfold according to prior inputs and not be causal to a new board state of the result of an action is reasonable just the status quo prevailing. I agree largely with what you're laying out here, that nothing happening is nothing happening as a rebuttal; I don't think naturalism is it's own defense. There should be a gameplay purpose and the only reason I think you'd want a status quo result is to ensure it's understandable to the player. That is, a player can't necessarily map consequences from "I try to pick the lock" to a fail forward result like "the guards hear you and open the door to see what's happening" unless they're explicitly told ahead of time this will happen.

Which brings us to the argument from gameplay, as that immediately creates the grounds for negotiation of either that specific consequence, or for the player to propose a different action to get a different consequence. Negotiation is the death of gameplay, because it subsumes all other mechanics. It always has the potential to be more effective or to overturn any other kind of mechanical interaction, so if it's allowed to creep in, it becomes the gameplay loop immediately.

The negotiation issue aside, fail forward runs the risk of removing tactical agency. If you make players commit to actions without knowable consequences, they lose much ability to discriminate between risks in the first place and it's not necessarily clear that they should prefer picking the lock to having down to the door in the first place.

Perhaps ironically, the bolded bit is actually the most realistic and plausible way to go. In real life, we can make educated guesses by drawing on our past experiences. But in RPGs, especially D&D, GMs are usually strongly encouraged to shake things up because having similar things happen too often is boring and leads to players getting complacent. In one particularly obvious example, it's why there's so many different varieties of monsters in D&D. Is this slime mold monster the one that dies when its exposed to fire, or the one that aggressively grows and becomes more dangerous? That dragon is green and brown; is it evil, and what does it breathe? Ditto for just about everything else: new interpretations of standard races and adversaries, innovative ways to build your adventure locales, new and different types of traps, and other new twists on old tropes.

In other words, unless the GM has a very curated, and above all short list of things the players may encounter, there's not going to be much of a way for a typical party to know the consequences. So "making the players commit to actions without knowable consequences" is, in fact, logical. They've never been in this situation before; how would they know the consequences?

However, in most games, either (a) the players are allowed to make knowledge rolls ahead of time or during the encounter or (b) in games that don't rely on skill checks, the GM is allowed or even encouraged to give the players knowledge based on information written on their sheets or backgrounds. "I have the 'Arcane Botanist' trait. What do I know about that plant-monster?" And that means the players aren't going to go in blind nearly as much as you might think--even if they've never been in this situation before, they probably some knowledge that can help them.

 

[Faolyn]

I see that you find these bits tedious. But phrasing it as if that's objectively true does not help your case.

The decision about how to spend time was considered unimportant and worth skipping until it came up. Not everyone considers it unimportant in that way.

In that case, you don't skip it, if everyone is cool with it. But what do you do if one PC spends an hour collecting herbs while the rest are twiddling their thumbs?

 

[Faolyn]

I find nothing happens to be an interesting result. Where does that leave us?

How is it interesting when literally nothing happens?

And by nothing, I mean nothing. It doesn't cause time to pass or a clock to advance. It doesn't cause the expenditure of resources. It doesn't provide a clue of any sort (no dogs to not bark). It doesn't bring any new things to take action against. It's just... nothing happens. Everyone just continues to stand there.

So seriously, how is that interesting to you?

 

[The Firebird]

How is it interesting when literally nothing happens?

And by nothing, I mean nothing. It doesn't cause time to pass or a clock to advance. It doesn't cause the expenditure of resources. It doesn't provide a clue of any sort (no dogs to not bark). It doesn't bring any new things to take action against. It's just... nothing happens. Everyone just continues to stand there.

So seriously, how is that interesting to you?

It adds to the feeling that the world is a living, breathing place and exists beyond the PCs.

 

[The Firebird]

One of my favorite rpg situations were when the players spent about 1 hour trying to figure out what was up with an obvious trap at the bottom of a staircase (holes in the ceiling, and a mechanism that turned the stairs into a slide). Finally they sort of gave up and decided to just enter the area. Nothing happened. (the trap had been triggered long ago, without any automatic reset) Maybe it is wrong to call that the most "interesting" result, but I think it might easily be the most amusing/memorable result possible.

I like this example a lot. And I agree it is much more interesting than "failure with a complication = the trap does have a reset mechanism", or the like.

In that case, you don't skip it, if everyone is cool with it. But what do you do if one PC spends an hour collecting herbs while the rest are twiddling their thumbs?

You don't do that. If herbalism plays a role in the game ideally there is a table or system or whatever for it. It could be the player can make the rolls and resolve the results themselves so it is quite fast.

 

[Lanefan]

And in every other situation? Or do your players only go to places that haven't been disturbed in ages.

They go all over the place; I was merely trying to posit an example of a situation where "nothing happens" is the most plausible result.

Which you're not actually playing in. I don't think anyone signs up for a fantasy game looking to emulate every aspect of the real world.

You're missing the point. The players are doing something they consider interesting, whether its raiding a dungeon or opening a shop.

Having it possibly fail because of things outside their ability to influence, like fluctuations in the economy or a distant war cutting off their supplies, would be boring, and quite frankly a jerk move because as the GM you control those things.

Boring perhaps, but by no means a jerk move.

Having it fail because of things they can at least attempt to influence--an evil competitor who is using underhanded means to destroy them, monster attacks, a curse on the land--that's interesting.

Depending on the fiction already established, inflicting oen of these on the PCs' attempts might be much more - and much more obvious - of a jerk move than your previous examples.

That's the same as having progress stop because the players can't get past a point that you, the GM, are controlling, by setting difficulties or penalties to rolls or whatever your system uses, or because you think the players should be smart enough to think of going about it a particular way. If they're going to fail at it, it should be in an interesting way that they can at least attempt to influence, not just GM fiat.

Because this is a game, their lives should be interesting. Maybe they aren't going to be making changes to the world at large, but they are making changes in their immediate lives.

II'm not as concerned about wrapping the narrative around the PCs as you seem to be. The PCs do what they do (I can only assume by the fact they're doing it that the players find it interesting) and I-as-DM react to that in ways that more or less make sense with established fiction, perhaps informed by a random roll to allow for something bizarre to happen on rare occasions.

That has nothing to do with the individual's own story. Unless the point of the game is that the players are nothing but powerless pawns who can't even control their own lives--and I can't imagine how that's even remotely fun.

They can control their own lives as much as we can in the real world plus what their fantastical abilities allow.

In your games. Maybe in a lot of games, but at my table, unless they're running in an "official" world (the Realms, Ravenloft, etc.), a new setting is made for each new game.

Me too, but when I make a new setting I expect it to last a while. The current one's gone 17+ years and has seen a few hundred PCs come and go (a lot of whom were what I call "one-hit wonders"; low-level types who didn't survive past their first or second field combat), and has also seen over a dozen players come and go (and come back, in some cases). It's a persistent setting that has thus far outlasted all its original players (though one has now returned after a ten-year absence) and almost all its original characters (one of the founding group of PCs is in deep retirement somewhere).

 

[AlViking]

OK, so why did you ask me about it? And then yell at me when I answered your question?

First, I disagree with your assessment based on my admittedly limited understanding of the definitions. Saying that you've agreed to follow the DM's campaign is not the same as abdicating control of your character. Of course I also haven't seen a definition of stance from the GM's perspective.

Second I wasn't asking for you to just repeat what you said. I don't care what the forge says because it seems to be quite biased and with an obvious agenda based on what I've seen.

But if you've paid much attention you've paid you'd know that a lot of people don't care much for the Forge. Unless it reinforces your preferences.

 

[Lanefan]

I'm going to quote from the new Daggerheart game here, which uses Rolling With Hope to indicate a "good" thing happening (whether or not you succeed) and Rolling With Fear to indicate a "bad" thing happening (again, whether or not you succeed); I bolded those bits:


So with your example, on a failed roll, you climb the cliff, but not nearly in time to save your friend; they don't necessarily die (unless that's a likely result), but they're definitely harmed in some way. On a success, you save them, but whatever was menacing your friend is still there. On a really good success, you are capable of hurting or maybe even driving off whatever was menacing your friend (maybe it's a free attack, or a surprise round). On a really bad failure, you take some damage in your mad scramble up the cliffside.

This completely fails (sorry) to address my concerns about failure being turned into success.

Remember, the task at hand that's being rolled for is to climb the cliff. That's it. Any concerns about anything at the top of the cliff have to wait to see if you even get there.

The bolded outright turns a failure into a success - you failed, but still climbed the cliff - and that to me is a complete non-starter. Fail means fail.

The other as-yet-unmentioned piece about concatenating multiple actions into one declaration is that those multiple actions may require testing vastly different character skills and-or abilities. Climbing the cliff to save your friend is a perfect example: you need some sort of climbing or athletics ability to climb the cliff and some other skill or ability (could be healing, could be combat, could be whatever, depending what the in-fiction situation is up there) to save your friend.

Squeezing both these tests into one roll isn't right.

 

[Pedantic]

Perhaps ironically, the bolded bit is actually the most realistic and plausible way to go. In real life, we can make educated guesses by drawing on our past experiences. But in RPGs, especially D&D, GMs are usually strongly encouraged to shake things up because having similar things happen too often is boring and leads to players getting complacent. In one particularly obvious example, it's why there's so many different varieties of monsters in D&D. Is this slime mold monster the one that dies when its exposed to fire, or the one that aggressively grows and becomes more dangerous? That dragon is green and brown; is it evil, and what does it breathe? Ditto for just about everything else: new interpretations of standard races and adversaries, innovative ways to build your adventure locales, new and different types of traps, and other new twists on old tropes.

In other words, unless the GM has a very curated, and above all short list of things the players may encounter, there's not going to be much of a way for a typical party to know the consequences. So "making the players commit to actions without knowable consequences" is, in fact, logical. They've never been in this situation before; how would they know the consequences?

However, in most games, either (a) the players are allowed to make knowledge rolls ahead of time or during the encounter or (b) in games that don't rely on skill checks, the GM is allowed or even encouraged to give the players knowledge based on information written on their sheets or backgrounds. "I have the 'Arcane Botanist' trait. What do I know about that plant-monster?" And that means the players aren't going to go in blind nearly as much as you might think--even if they've never been in this situation before, they probably some knowledge that can help them.

Hold up, you're conflating two different things here. Novel stuff, like encounters, creatures, situations, traps, all that, is not the same as an unknown result to an action declaration. I'm all about all about knowledge checks, perception all that (and have argued before they should generally be treated as defenses and used in a passive process) and think players should generally be given or have routes to get information about stuff, but that's orthogonal to my point.

My point is that player action declarations should have knowable impacts on the board state for all possible outcome cases; the lock will be picked and open or not based on the outcome of my declared action, there are no other possible outcomes that are inherent to my action declaration itself. Other elements of the established gamestate that I could in some way assay might affect what happens next. If there are guards on the other side of the door, and hearing them playing dice is contingent on my investment in Perception, or perhaps something like burning a charge on a Ring of X-ray Vision, I might have access to that information and change my action declaration to mitigate the risk, probably bringing in whatever the stealth rules are, or finding another way around, or perhaps I don't successfully employ whatever mechanic governs knowing that, and we're in consequence town.

But if the possibility of guards reacting is contingent entirely on my declaring an action in the first place, I've lost a whole avenue of gameplay. I can't have gotten the information ahead of time, and I can't have picked a different action.

To put it more pithily, I'm saying that the tools of interaction should produce knowable outcomes; the objects being interacted with should contain the potential for unknowns. My problem with moving the unknown up to the mechanisms of interaction is that it's flattening; I want to play a game where picking the right action declarations matters.

 

[Lanefan]

But it’s not really a retcon. Retcons change something that’s been established, or somehow provide new context that changes how we perceive what we knew.

Skipping past a travel scene means we’ve not established what happened during the journey. Which leaves us free to determine if anything meaningful or interesting happened or not.

The moment you narrate they've arrived safely at their destination you've also just established that nothing of any significant danger occurred during that journey, meaning that any retconning they want to do is now under the umbrella of that safety you already established.

This becomes highly - extremely highly, in fact - relevant if what they want to retcon might otherwise have any degree of risk attached; unless you're willing to overturn previously-established fiction, which I think we all pretty much agree is poor form.