Failing Forward

[Umbran]

The player failed, and the narrative moved forward. In fact so far forward he made it to the end.

With respect, I find this to be a reductionist, non-constructive way to engage with the topic.

Yes, if you are being a super-duper-stickler on the absolute meaning of words, you can view "the narrative moves forward to its immediate end" as being a *technically* accurate statement.

But how about looking at language as people actually use it? If the character is dead, the narrative no longer moves forward, really. To claim that is still "forwards" is kind of silly.

 

[Maxperson]

This is where a number of us disagree. It all depends on what you're considering in play for the climb check. The extreme narrow view is that the climb check determines solely whether you succeed and move forward up the wall/cliff, and failure means you fall. In other words, the roll is specifically about a task during that situation, and nothing else. The reality is this doesn't allow even a simple you didn't fall but you also didn't move forward.

Others feel that it can encompass more than that. Using a degrees of success/failure allows other things to happen, such as pulling some loose rocks free that starts a small rockslide and causes minor damage. You don't fall, but still suffer consequences and don't move forward. Or your piton pulls loose, and again you don't fall, but don't move forward. Or your pack gets caught on rocks and roots and you must free yourself.

Failing a climb roll doesn't mean that you have to fall. It can also mean you just fail to climb. My view isn't climb or fall, but simply that any result should directly relate to the actual climb. A rod falling out of the pack is an addition to the actual climb, so it's indirectly related and can happen in my game, but only as a result of a different roll or someone rolling a couple of 1's in a row. Two 1's causes things to happen in addition to the task at hand that are still related, even if indirectly.

The rod falling out of your pack is a step beyond that in general concept in that it's looking at the situation rather than the specific skill or task. So instead of failing to simply move up the wall/cliff, or falling, you have a setback that may require some action on your part to resolve.

In a non fail forward approach, many DMs might have had something like the rod fall out of the pack if they rolled a 1 on their climb check. Or maybe it falls out as you tumble down the mountainside, and it's stuck up on a ledge and you must retrieve it before you can continue.

The point is that it's a widening of what the check covers. The wider net the check throws, the more options you have as a DM, and the fewer additional checks you'll need.

I get it. The type of fail forward that involves things not directly related to the check just doesn't work for me or my group.

 

[Ilbranteloth]

With respect, I find this to be a reductionist, non-constructive way to engage with the topic.

Yes, if you are being a super-duper-stickler on the absolute meaning of words, you can view "the narrative moves forward to its immediate end" as being a *technically* accurate statement.

But how about looking at language as people actually use it? If the character is dead, the narrative no longer moves forward, really. To claim that is still "forwards" is kind of silly.

I was just being facetious. But the intended meaning of the term is not obvious, and a good number of people think that it means 'instead of failing outright, they still move closer to their goal' or in other words, that failure isn't an option any more.

Ilbranteloth

 

[Ilbranteloth]

Failing a climb roll doesn't mean that you have to fall. It can also mean you just fail to climb. My view isn't climb or fall, but simply that any result should directly relate to the actual climb.

Which I did mention as well. It's just a lot of GMs don't recognize that option.

A rod falling out of the pack is an addition to the actual climb, so it's indirectly related and can happen in my game, but only as a result of a different roll or someone rolling a couple of 1's in a row. Two 1's causes things to happen in addition to the task at hand that are still related, even if indirectly.

I get it. The type of fail forward that involves things not directly related to the check just doesn't work for me or my group.

Yep, and I think that's perfectly valid. And frankly, my actual preference is closer to yours.

The bottom line, is that I think there is a continuum of possible rulings. From a very strict binary resolution, to a very broad approach encompassing the entire circumstance.

While I don't feel that 'fail forward' is the right term, I do think that quantifying a general approach that highlights what makes good exploration encounters/events, but not tying it to a specific system that also identifies the sort of 'degrees' of that continuum so a GM can find what approach they like.

You like incorporating the Fate system in your game. I'm not a fan of that myself, but I'm not sure I'd go all the way to a climb check where you drop the rod. Instead, if the possibility of dropping the rod is a bit beyond what I would consider in the climb check myself. But I would consider the possibility that they might lose it if they fail their climb check severely enough, but not enough to fall, although they'd probably have to make a Dexterity save or something.

This discussion is all very helpful to me, because it is helping identify the boundaries I have.

Ilbranteloth

 

[GMMichael]

Is this thread failing forward?

More often than "fail forward," I like to use something like "auto-success," which means there's no need to make a Climb check if the PC has a decent chance of pulling it off. Unless, as GM, I've planned for something interesting to happen on or near the cliff, the climbing happens without fail.

 

[Ilbranteloth]

Is this thread failing forward?

More often than "fail forward," I like to use something like "auto-success," which means there's no need to make a Climb check if the PC has a decent chance of pulling it off. Unless, as GM, I've planned for something interesting to happen on or near the cliff, the climbing happens without fail.

That's a good point, and I think it's actually part of the discussion. As is the take 10/take 20 concept, or variations like what I use. If something is within their ability, then 'failure' is often an indicator of time. My go to example is trying to pick a lock before the guards arrive. If they have a high enough skill, then they'll eventually succeed. So the difference between the roll and the DC is how many rounds it take (they don't know how many), and they have to deal with the potentially returning guards while they are working on the lock.

The reality is, knowing when auto-success, vs take-x, vs 'fail forward' vs multiple checks is all related and all valid at different times.

Ilbranteloth

 

[nomotog]

One idea I have thought about would be to include a stress system as a fail forward system. It's tricky to think of a punishment for every failed action, even trickier if you want the two to be related. The idea is that failing a roll generates stress and when stress gets to much, ... well some kind of bad thing happens. It's an idea I took from maid RPG. It's nice because stress is so vague. It can be logically attached to any action.

 

[pemerton]

Sure. That's bad luck, though. It's also possible to get frostbite while climbing a mountain, and it's possible to start an avalanche while climbing a mountain, sprain an ankle while climbing a mountain and so on. A failed climb check involves climbing or the failure to do so. I occasionally call for a fate roll in my game, and sometimes bad luck or good luck happens in addition the results of whatever else is going on. If the PC fumbles that fate roll while climbing, bad luck happens. What doesn't happen is for it to happen instead of a failed climb check.

<snip>

I also have skill fumbles, and if you roll multiple 1's in a row, bad things directly connected to the event at hand start to happen.

Failing to keep things in order and/or being sloppy has nothing directly to do with a climb check. It's bad luck that would cause it to fall.

It isn't that it's unrealistic to drop a rod while climbing. It's that it's not realistic for the climb check to be the reason. Climb checks check climbing and that's it.

<snip>

In and of themselves, no. They aren't unrealistic. What makes it unrealistic or not is how those things are brought into play. I failed a climb check so it started raining and slowed my climb down is not a realistic result of a climb check, even though rain is realistic. Rain has nothing to do with a climb check.

To my mind, the distinctions being drawn here have nothing to do with realism. They are about preferred stakes for certain game checks.

In the real world, a skilled climber is:

* Less likely to drop/lose things (better packing, more careful in avoiding snags, etc);

* Less likely to have handholds break under his/her weight (better judgment, better/more careful testing of holds, etc);

* Less likely to get frostbite (better at wearing protective gear, better at judging when fingers are getting too cold, etc).\;

* Etc​

Of course suffering any of these things is bad luck, but skilled climbers make their own luck. Putting it all into a non-skill-dependent "fate roll" or skill fumble system does not seem, to me at least, to encourage or reflect realism about skill with climbing.

What it does do is confine the stakes of a climbing check to one very narrow question: does the character go up or down? Personally I don't think that's the only interesting question (or, always or even often, the most interesting question) to ask about a fiction in which a character is trying to achieve his/her goal by climbing up a mountain.

If you've read the thread, fail forward can have nothing directly to do with the roll at hand. Like the rod falling out of the backpack as the result of a failed climb check.

People like different levels of connectedness.

I'm not 100% sure what "connectedness" means here, but if it's a reiteration of "realism" then it's out of place as far as the ascent of Mt Pudding is concerned. There is no more or less "connection" between climbing skill and not falling, climbing skill and not having a handhold break, climbing skill and not dropping or snagging some gear, climbing skill and not suffering frostbite. All these things are connected to how skilled one is as a climber.

Examples of "non-connection" have been given upthread - eg there's no connection between skill in Scavenging and a PC's brother having been evil even before possession by a balrog - but Maxperson's complaints aren't directed at those examples.

I suspect that a DM that needs to hold so tightly to the story, so as to insist there is only one way to do things, is either a very new DM or wouldn't be comfortable giving up so much control to be willing to try fail forward anyway.

"Holding tight to the story" is a red herring. In [MENTION=6696971]Manbearcat[/MENTION]'s example, no one is holding tight to a story. The question is about what can be at stake when a check is declared and resolved. One might even say that insisting it must always and only be one thing - does the character go up or go down - is a form of holding tight to the story!

Another reason some of us don't see the same need for fail forward is because we are not having these "problems" that it is trying to fix.

As I've posted repeatedly upthread, the "problem" for which "fail forward" is a fix is that of achieving a dramatic story via RPGing without the GM preauthoring a story.

If you don't have that problem - either because dramatic story isn't a high priority for you in RPGing or because you don't mind GM pre-authorship (and historically, D&D play has tended to fall into one or the other of these camps) - then you probably won't be interested in "fail forward" as a technique.

But this has nothing to do with what is or isn't "realistic" as stakes for a Climbing check.

 

[pemerton]

I prefer detailed DMs with really deep campaign worlds that I can discover. I want those worlds to be built like a clock so that the interconnections are already there and are not waiting for me a player to invent them.

In the games that first overtly talked about "fail forward" techniques - eg Burning Wheel, HeroWars, etc - the player doesn't invent the world either. Control over backstory - and particularly over consequences of failure - is in the hands of the GM. But the backstory in "fail forward" games is not authored primarily in advance.

Having the backstory already authored, so that the players discover it like the workings of a clock, would be one example of the pre-authorship that "fail forward" as a technique is intended to avoid.

I would also add: worlds that are authored in response to player action declarations can also be very deep. If you look back at the actual play examples I have given above, I don't think they imply a "shallow" campaign world.

I want a game where the players as their characters confront challenges presented to them by the DM. They do everything in their power to achieve their goals

This is also true of the games I run. The GM authors the challenges. The players, via their PCs, confront them and do everything in their power to achieve those goals. Sometimes they succeed; sometimes they fail.

When they fail, new challenges result. (See some actual play examples upthread.)

For a good general discussion of this aspect of "fail forward" play, see the Eero Tuovinen blog I've linked to already upthread.

In a "fail forward" game, however, the parameters of the challenges - ie what backstory is constraining the possibility of success - is not spelled out in advance. So the players can't, for instance, reduce the chance of failure to (near-)zero by exploiting the fiction. In this sort of play, the drama of confronting challenges is prioritised over the logistics of overcoming them. (Contrast Gygaxian D&D, which reverses those priorities.)

Those who want a consistent world

Who doesn't?

The idea that "fail forward" undermines consistency is another red herring. To go back to Mt Pudding, there is nothing inconsistent about a world in which climbers sometimes lose important gear down ravines.

One style actively involves the DM and players teaming at the metagame level to ensure that the story outcome is interesting and exciting. So players will happily throw their characters into trouble spots via some metagame construct even if the character would never want such a thing.

I agree with [MENTION=6696971]Manbearcat[/MENTION] that you are drawing a false dichotomy here.

I really think it would be helpful if you engaged with some of the actual play examples that have been posted upthread, Then you could talk about how actual games are actually being played rather than how you imagine them being played.

In my game where the PCs searched the ruined tower for the nickel-silver mace, and instead found black arrows apparently forged by the mage PC's brother before his possession by a balrog, there was no "teaming at the metagame level". The players were just playing their PCs. What is different from the style that you seem to prefer is that I, the GM, authored some new campaign backstory as a result of the failed Scavenging check, so as to put the fiction into a situation which (i) was not what the PCs (and players) had wanted it to be, and (ii) forced the players to make new, hard choices.

That is "fail forward" in action.

The DM knows the capabilities of the party (and players).

That's not generally true for me as a GM, both in the sense that I don't memorise the players' character sheets - so they might have skills, equipment etc I don't know about or have forgotten about - and in the sense that I certainly can't extrapolate from everything on those sheets to every feasible action declaration the players might make for their PCs.

What I do know is the PCs' goals. (Which, in [MENTION=6698278]Emerikol[/MENTION]'s terms, have been "meta" authored by the players so as to lead to action - a bit like Gygaxian D&D depending on the PCs being hungry for wealth and fame.) These are what I refer to in framing challenges. It's up to the players to work out what their PCs can, and want to, do in response to them. (In 4e I also know the PCs' levels. BW, though, doesn't have the same sort of scaling system that 4e has.)

 

[Mustrum_Ridcully]

If you've read the thread, fail forward can have nothing directly to do with the roll at hand. Like the rod falling out of the backpack as the result of a failed climb check.

That is also an important aspect to me.

What does a failed climb check mean? Sure, you could have listed in the ksill description: "A failure always means you fall back x feet" or whatever. But it's not usually a good idea to try to be that precise in a rulebook if you cover a lot of different situations. Maybe falling during a climb is not actually the problem you are trying to avoid, for example. It might also be considerably too narrow - if you're carrying something or someone with you, you might want to know if you keep it, for example.

So interpreting a failed climb check as recieving a complication could be falling down on a lower level, or it could be losing something you were carrying, like the pick you used or some other important item you need.