Failing Forward

[iserith]

I just edited this in to the last post. Dang it for timing

I also think you're looking at the goal incorrectly. The goal of getting to the top safely is a two part goal. Get to the top AND safely. Failing at one part of that goal doesn't mean that getting to the top still isn't a success at the other part.

I'm not reading progress too literally. I'm reading progress for what it means. Moving towards your goal, but not getting there. If you have moved all the way, it's not progress. That's completion.

Again, if my goal is to get X and I end up getting X plus an unwanted Y or Z, then I've made progress in getting X, but Y or Z is the setback and failure. I didn't get to the goal I wanted, but made progress toward it. And since we've gone back and forth on this point several times already to no avail, it's safe to say we won't come to agreement and can move on.

 

[grendel111111]

That is what's happening. As I said, it's just stake-setting.

Really it is a question of where you are setting the stakes. In the climbing example you are setting the stakes as get to the top and have all your gear and ready to get going right away and having no other set backs. Anything less than they you would view as being a failure. And so you group all that under one "heading". Am I correct in this reading of what you are saying?

Having set those stakes you decide what failure will look like (in this case they make it but the lose an important item). All this is then put down to a climb roll because that was the "action" that was being taken by the character.

For me in the same situation it might well go the same most of the way. I "need" them to get to the top of the mountain and it seems climbing is the only possible way, so they have to succeed at climbing. So what are they risking to make a roll worth taking and not just state that they make it to the top? The thing that is at risk is that they might lose an important item. OK, that works fine. How do I know if they lose an important item during the climb? (For me) This is the point where the applicable skill or ability is determined. What are the stakes (Clearly not the climbing, but rather the not dropping anything) and what skill or abilities best reflect this stake? So I would come up with a skill roll reflective of the actual thing that is going to change between a success or a fail. Did you properly secure everything before setting off seems like it would give a better indication of if you would lose something, rather than a climbing roll (given that the climb part will succeed independent of what he player rolls). So a survival roll (Int or Wis) modified by if the characters had taken time to secure their packs, any character traits that would effect the roll etc. As securing packs and preparedness are all part of survival.
I think this would give a more accurate skill to consequence connection.
Why do you feel that a climb check would better reflect if someone drops something or not?

 

[iserith]

Really it is a question of where you are setting the stakes. In the climbing example you are setting the stakes as get to the top and have all your gear and ready to get going right away and having no other set backs. Anything less than they you would view as being a failure. And so you group all that under one "heading". Am I correct in this reading of what you are saying?

Having set those stakes you decide what failure will look like (in this case they make it but the lose an important item). All this is then put down to a climb roll because that was the "action" that was being taken by the character.

For me in the same situation it might well go the same most of the way. I "need" them to get to the top of the mountain and it seems climbing is the only possible way, so they have to succeed at climbing. So what are they risking to make a roll worth taking and not just state that they make it to the top? The thing that is at risk is that they might lose an important item. OK, that works fine. How do I know if they lose an important item during the climb? (For me) This is the point where the applicable skill or ability is determined. What are the stakes (Clearly not the climbing, but rather the not dropping anything) and what skill or abilities best reflect this stake? So I would come up with a skill roll reflective of the actual thing that is going to change between a success or a fail. Did you properly secure everything before setting off seems like it would give a better indication of if you would lose something, rather than a climbing roll (given that the climb part will succeed independent of what he player rolls). So a survival roll (Int or Wis) modified by if the characters had taken time to secure their packs, any character traits that would effect the roll etc. As securing packs and preparedness are all part of survival.
I think this would give a more accurate skill to consequence connection.
Why do you feel that a climb check would better reflect if someone drops something or not?

I don't have any strong preference as to what check is used to resolve the uncertainty.

 

[grendel111111]

I don't have any strong preference as to what check is used to resolve the uncertainty.

So you would be happy if we used (Cha) intimidation to scare the mountain into letting us climb without dropping anything?

 

[iserith]

So you would be happy if we used (Cha) intimidation to scare the mountain into letting us climb without dropping anything?

Assuming a regular old mountain, the approach (scare the mountain via intimidation) to this goal (climb the mountain without dropping anything) does not have an uncertain outcome, thus there would be no check.

 

[grendel111111]

Assuming a regular old mountain, the approach (scare the mountain via intimidation) to this goal (climb the mountain without dropping anything) does not have an uncertain outcome, thus there would be no check.

So the check does matter and for you at least there needs to be some connection between the thing being affected and the check used.
Can you see that for some people, they might like more of that connection than you prefer?
They might like the consequence to relate more directly to the skill being used than you like?
(Both approaches being totally valid)
Can you see why for some people the disconnect between a climb roll leading to dropping something is a leap too far for them?
And why they might prefer a consequence that for them more meshed with failing to use the skill they were trying to use?

 

[iserith]

So the check does matter

The goal and approach offered by the player in the context of the fictional situation matters more to me. That is what the GM is judging. If I call for a "climb check" and the player says that it sounds more like a "survival check," I go with what the player wants, (edit: provided the player is acting in good faith). I'm determining whether there is uncertainty and what the stakes for success and failure are. The check just resolves that uncertainty and allows me to narrate the result. It's not important enough to me to debate which skill applies better.

 

[innerdude]

But here we get back to the same argument as with Mt Pudding: can you in any way reach the top (i.e. succeed) on a failed roll in either system? If yes, then "fail" is probably the wrong word to be using.

Just a thought on this, hopefully not too far off on a tangent --- A game's mechanics can definitely lean one way or the other to point out to players and GMs how binary or fluid "failure" should be.

For example, I don't recall anywhere in the D&D 3.5 PHB or DMG there being references to non-binary failure. As in, "One failed check may not mean 'total failure,' it may merely mean a complication has arisen," etc. etc.

Well, as a consequence, one of the things that happens when binary pass/fail is the default mode of resolution is players start looking for ways to guarantee sucess when making checks. When hard failure is failure, you tend to find ways to alleviate that problem.

As a result, in 3.x the only reliable indicator to a player that a character will succeed at a task is the bonus number associated to the relevant skill or check---the natural result being that players power-gamed the heck out of the skill system to increase the bonus numbers (further exacerbated by the fact that D&D has never really supported a "bell curve" model for probability).

To lead away from this kind of circular "system building," your system either needs to incorporate broader levels of competence across skills (this is Savage Worlds' approach, and I think 4e's as well), directly incorporate more granular levels of success/failure in resolution (Savage Worlds also does this to a small degree), or influence GMs to build granular success and failure into scene frames with codified scene-resolution mechanics (the obvious example being skill challenges in 4e).

Anyway, my real point is that "fail forward" mechanics are a "looped in" process that may have a number of general side effects on system development.

To @Lanefan's point, how binary is that "failure" to climb the mountain?

Interestingly, 4e seems to at least unconsciously address this with skill challenges ---- Reading between the lines, the 4e skill challenge mechanics seem to be a clue to GMs that they should be very careful about using binary pass/fail situations for framed scenes, and do something more nuanced.

 

[Maxperson]

Again, if my goal is to get X and I end up getting X plus an unwanted Y or Z, then I've made progress in getting X, but Y or Z is the setback and failure. I didn't get to the goal I wanted, but made progress toward it. And since we've gone back and forth on this point several times already to no avail, it's safe to say we won't come to agreement and can move on.

There are two goals in climbing. The primary goal of getting to the top, and the secondary goal of avoiding bad things. It's not one unified goal. I'm okay with moving on, though

 

[Lanefan]

Just a thought on this, hopefully not too far off on a tangent --- A game's mechanics can definitely lean one way or the other to point out to players and GMs how binary or fluid "failure" should be.

For example, I don't recall anywhere in the D&D 3.5 PHB or DMG there being references to non-binary failure. As in, "One failed check may not mean 'total failure,' it may merely mean a complication has arisen," etc. etc.

It's not in 1e either that I know of. Can't speak for 2e.

Well, as a consequence, one of the things that happens when binary pass/fail is the default mode of resolution is players start looking for ways to guarantee sucess when making checks. When hard failure is failure, you tend to find ways to alleviate that problem.

As a result, in 3.x the only reliable indicator to a player that a character will succeed at a task is the bonus number associated to the relevant skill or check---the natural result being that players power-gamed the heck out of the skill system to increase the bonus numbers (further exacerbated by the fact that D&D has never really supported a "bell curve" model for probability).

Interesting point, and quite right I think. That said, a DM always has or had the ability to narrate the failure in myriad different ways, not all of which involve bad things happening to the PC but none of which involve actually succeeding...which is what I'm trying to preserve here.

To lead away from this kind of circular "system building," your system either needs to incorporate broader levels of competence across skills (this is Savage Worlds' approach, and I think 4e's as well), directly incorporate more granular levels of success/failure in resolution (Savage Worlds also does this to a small degree), or influence GMs to build granular success and failure into scene frames with codified scene-resolution mechanics (the obvious example being skill challenges in 4e).

Anyway, my real point is that "fail forward" mechanics are a "looped in" process that may have a number of general side effects on system development.

To @Lanefan's point, how binary is that "failure" to climb the mountain?

Interestingly, 4e seems to at least unconsciously address this with skill challenges ---- Reading between the lines, the 4e skill challenge mechanics seem to be a clue to GMs that they should be very careful about using binary pass/fail situations for framed scenes, and do something more nuanced.

To me the only "binary" bit is that a fail means you don't get to the top. Everything else that may have happened as a result of the fail can be determined either by DM narration, dice rolls, player narration, or some combination of all those things and others.

Let's use another example, one where something like 4e's skill challenge mechanic would seem to work quite well: finding your way through a trackless forest. Here, a binary pass-fail would give two possible results: you get where you're going, or you're lost. A more flexible pass-fail (where "pass" still means you get where you're going) could have a fail mean one or more of:
- you find something else of interest instead which diverts your attention
- you don't reach your intended destination but have found a hilltop with a view and can see which way you need to go
- you encounter someone or something in the forest that you maybe didn't want to
- you encounter someone or something in the forest that you ultimately did want to
- your party somehow got split up and now can't find each other, never mind the destination
- you're going the right way but you took too long and darkness fell

All of these except the last one provide either more story options or a new challenge for the party, or both; and advising DMs to keep this sort of thing in mind is very worthwhile. That said, none of them sees you get where you were originally going...which is my point.

Another way of looking at it is that in the Mt Pudding example the DM moves the goalposts when narrating the failure: the roll is to succeed at the goal of climbing the mountain but the narration says you failed at holding on to your gear; and these ain't the same thing.

Lanefan