Failing Forward

[pemerton]

Looking at this example, something like this could certainly arise in one of my games. But it wouldn't be the result of a roll so much as a result of me thinking through the NPCs and events that are ongoing. In my game, Wassal's disposition on this matter (and whether he is angry about orcs and blaming the PCs) would be something I establish for myself before the PC even attempts to deal with him

<snip>

It sounds like your approach is almost opposite of mine (not in a bad way, just procedurally you seem to be taking the roll results to help establish that sort of thing). So it sounds almost like a gift wrapped scenario. Neither you nor the players have any clear sense of what is going on with Wassal until that roll, and then the roll determines some of the contents. So it isn't just about generating a result (i.e. he finds Wassan). It is about generating some fiction around Wassan as well. Is this correct?

This is correct. The fiction around Wassal is generated in response to the roll.

One way to look at it is this: by making a Circles check, the player is taking a gamble. If the player wins, he gets to make it true that the local captain of the tribesmen is a friendly former associate who will help the PCs out. If the player loses, I get to narrate something instead. "No one turns up" is a legitimate narration, but flagged in the GM advice as also the most boring option. "The enmity clause" is the more interesting option permitted to the GM - you meet the NPC you wanted to, but s/he is not disposed to help but rather to hinder. The GM has to narrate the fiction around that, but I hope you can see from my example that this fiction is not just spun out of nowhere but built around prior backstory and events of play.

When the PC failed his "check circles" roll, was that something where he basically said "I am going to check my circles", made a roll, then you narrated that he was captured, or was the capture something that was played out at the ground level.

The capture plays out at ground level - I narrate that the PCs are surrounded by evidently hostile tribesmen, and then there is a bit of back-and-forth between the PC mage and Wassal, in which some of the relevant backstory (eg the identity of the Desert Fox, Wassal's anger at orcs being brought into the desert) comes out. The capture is then a formality, in the sense that the players can tell that their PCs are no match for the tribesmen, and so when Wassal commands them to come with him back to his oasis camp, they comply.

Moving to a higher level of metagame, one reason the players are relatively happy to allow themselves to be captured is because they know that in this system, capture isn't the end of things but just another springboard to something or other. Upthread I quoted [MENTION=99817]chaochou[/MENTION] (hi, chaochou!) saying that failure is not penalised that heavily in these "fail forward"-type games. This is an instance of that truth being manifested in play. The failure is a real failure, but the players know that it won't be a block to their PCs doing stuff - it's just that the stuff they do (in this case, try to bargain with Wassal and persuade him of the truth about the orcs) is not the stuff they hoped to be doing (leading the tribesmen on a desert rescue mission somewhat in the spirit of Lawrence of Arabia).

 

[pemerton]

if I'm running Tunnels and Trolls failing just means failing. And laughing and pointing, obviously.

T&T is a game which, even moreso than Moldvay Basic, is not built around "fail forward" techniques.

Of other games from that general period, I think RQ and Traveller are interesting. In many ways the sort of large scale operatic grandeur they seem to want to generate would benefit from fail forward narration, but that would undermine the ruthless austerity and integrity of their ultra-sim resolution systems.

This is part of my issue with those systems, and also with Rolemaster - the promise of grandeur too often becomes a practicality that verges on the tedious. Though in the case of RM, because I spent so many years GMing it, I developed various workarounds. These days, I find that BW gives the same sort of grit and detail in PC building and resolution as RM, but without the need for the workarounds.

 

[I'm A Banana]

I generally like fail forward, though I'm not sure if you're describing what I'm used to considering "fail forward".

Specifically, "fail forward" is not, as I understand it, "Succeed, but at cost," as FATE games often put it. It is "Fail, but there is a pretty clear path to try something else." And, as such it isn't so much a mechanic, as a bit of advice for the GM to not have all progress in an adventure blocked by a failure.

For example - Say the PCs are exploring a tomb, hunting the BBEG, who is in his secret lair, behind a super-secret door. The players go through the dungeon, search for for secret doors, but they botch the roll, and fail to find it.

In "standard" play, this is basically a blocking issue. The PCs cannot continue forward unless they find that door. There's no clear path to moving forward. The PCs don't even really know where they failed, as they don't know for sure there was a door to begin with. All they know is they were told the BBEG was here, and they didn't find him. Oh, well...

In "fail forward" the PCs fail to find the secret door. Oops! So, shortly, a minion comes up from the area of the dungeon they have cleared, that should be empty. If they are smart enough to not kill the minion outright, the minion may be a source of information on where the BBEG is. The PCs still have a chance to find the enemy, even though they failed the basic way. Perhaps this will be a bit harder, or more complicated, as their guide is untrustworthy, or perhaps not.

In "succeed, but at cost," you find the door alright - just as the hairy troll steps out through it! In order to use that door, roll for initiative!

We might say that, "Succeed, but at cost" is one way to get a fail forward, but it is not the only way.

By this description, my only significant issue with "fail forward" (maybe aside from how jargon-y it is) is that it is ultimately a way through a bottleneck.

A well-designed game session should strive to avoid the bottleneck in the first place.

Like, your example presupposes that the inevitable end state is to fight the BBEG. But it should be OK in D&D game to not have to fight the BBEG. If you fail to find the secret door, maybe you DON'T fight the BBEG, maybe you DON'T thwart his plans, maybe you go back to town in defeat and tomorrow have to deal with his invading forces.

Utter and complete failure, or ignoring the path entirely, should be an option, and it should be enjoyable, too.

Or to use a D&D metaphor: if an adventure is a dungeon, it should have multiple entries, multiple exits, and it should never be the only dungeon you can reach from town. Fighting a particular encounter should never be something that is assumed to happen.

 

[pemerton]

my only significant issue with "fail forward" (maybe aside from how jargon-y it is) is that it is ultimately a way through a bottleneck.

A well-designed game session should strive to avoid the bottleneck in the first place.

<snip>

Utter and complete failure, or ignoring the path entirely, should be an option, and it should be enjoyable, too.

"Fail forward" is not primarily a way through a bottleneck. The whole idea of a "bottleneck", or of a session designed to avoid bottlenecks, suggests the type of prescripting of adventures that "fail forward" is an alternative to.

The discussion over the past half-dozen or so pages, dealing with some actual play examples as well as [MENTION=6696971]Manbearcat[/MENTION]'s toy example of the pudding quest should give you a sense of what "fail forward" is for. Primarily, it is about maintaining narrative momentum.

 

[Umbran]

A well-designed game session should strive to avoid the bottleneck in the first place.

This has already been discussed some, but to remind folks:

1) The GM's best laid plans do not survive contact with players. You can't account for everything they may try beforehand, and you can't keep them from painting themselves into a corner.

2) Any plan that requires perfection on the GM's part is a bad plan. GM's are human, and make mistakes. There *will* be flaws in your adventure designs. The GM should have tools to deal with design flaws in situ.

3) As I mentioned upthread - this is often used in systems where what a D&D-only player would call a "well designed session" does not exist. There may be no detailed map down to the 5' square level, with every secret door labelled and every trap with a well-known CR and method of deactivation, monster and NPC detailed out in full stat blocks, and powers and spells carefully chosen, and placed on aforementioned map. Rules engines like FATE and Cortex+ take as a base posit that some of the content of the adventure will be built out of these complication bits. In a more improvisational adventure, you can't design out bottlenecks before play - instead, you use rules systems that disperse bottlenecks as they develop. And we are talkign about bottlenecks to *action*, not necessarily to a prescripted goal.

Like, your example presupposes that the inevitable end state is to fight the BBEG. But it should be OK in D&D game to not have to fight the BBEG. If you fail to find the secret door, maybe you DON'T fight the BBEG, maybe you DON'T thwart his plans, maybe you go back to town in defeat and tomorrow have to deal with his invading forces.

You have missed the several times over where we have mentioned that it isn't really a predestined end we are aiming at in general. You're resurrecting a boogeyman. The *players* have a goal.

Utter and complete failure, or ignoring the path entirely, should be an option, and it should be enjoyable, too.

You see, that last bit, about it being enjoyable, isn't generally true, or in any way ensured by the D&D rules, or many other systems. The issue at hand isn't even the failure, it is the result of failure - stalling without meaningful choices to make. Failing forward is, in essence, making sure the player has meaningful choices after failing.

Or to use a D&D metaphor: if an adventure is a dungeon, it should have multiple entries, multiple exits, and it should never be the only dungeon you can reach from town. Fighting a particular encounter should never be something that is assumed to happen.

Nobody is assuming that a particular encounter should happen. Get the railroad boogeyman out of your head, *please*.

Also note that, if your party has Inigo Montoya in it, going off to another dungeon instead of chasing Count Rugen is *not* going to be an option. Sorry, just not happening.

So, as a GM, are you going to leave Inigo frustrated that he can't get through the locked door, or are you going to make it that Fezzik is just near enough to hear Inigo's calls for aid - and let Rugen set up an ambush as the complication?

 

[I'm A Banana]

"Fail forward" is not primarily a way through a bottleneck. The whole idea of a "bottleneck", or of a session designed to avoid bottlenecks, suggests the type of prescripting of adventures that "fail forward" is an alternative to.

The discussion over the past half-dozen or so pages, dealing with some actual play examples as well as [MENTION=6696971]Manbearcat[/MENTION]'s toy example of the pudding quest should give you a sense of what "fail forward" is for. Primarily, it is about maintaining narrative momentum.

Yeah, the jargon-y problem is being difficult to lock down what it is or what it is for.

But the Mt. Pudding example doesn't really dispel my concerns over the idea as a player or as a DM. That example posits that the intent of retrieving the pudding is something that is not really changing. Thus, it is related to my description of Umbran's "finding the secret door to fight the BBEG" - a "bottleneck" in that play can functionally proceed in only one direction (or be bereft of interesting choices/stop while we wait for someone to make a check/etc.). Though events happen on the way, the action is driven inexorably toward the pudding/BBEG, and this is accepted by all players as basically the ride you're on.

For my enjoyment, it is better to be able to be able to raise the question: what happens if I don't get the pudding/fight the BBEG? What possible actions are capable of potentially changing my intent, to use Manbearcat's verbiage? What would make Bob not want the Pudding, or make the Pudding forever unavailable to Bob, and how would Bob react?

I like these questions because they produce interesting gameplay scenarios about character motivations - what do I want, what am I willing to do to get it, what happens if I can't get it - and leave the ultimate arc of the narrative in question (is this going to be a story where the hero does something heroic or a story where the hero fails to do something heroic?). Every challenge becomes a decision point - do I undertake this risk, or do I do something else? Do I want the pudding that badly? Less "How do I get the pudding?" and more "Do I even want to get the pudding?"

It's an old acting trick - what is your motivation and how is this scene building to it? The pudding isn't important, but the reason my character wants the pudding is critical. "Fail forward" seems a bit more concerned with the Adventure to Get The Pudding or the Quest to Slay the Evil Thing than it is with The Story of Bob (who might like pudding and hate evil), which weakens it as a role-playing tool, IMXP. At least if Umbran and Manbearcat present it fairly.

You're resurrecting a boogeyman

I might be using the term "bottleneck" in a slightly different way than you may be fearing. I've no fear of Fail Forward, no crusade against it. There's no value judgement placed on "bottleneck", it's simply an attempt to describe a player's meaningful options in a given scene.

If the scene can lead in only one real direction (toward the pudding, toward the BBEG), that's a bottleneck/story funnel/railroad/arrow/directional movement/queue/roller coaster/flowing to the sea/waterfall/whatever-you-want-to-call-it. The end of the scene is: Bob is making progress toward the pudding. The end of the scene is never: Bob makes healthier choices about his diet.

From your description and Manbearcat's description, that seems to be a core feature of Fail Forward - Bob is never just UNABLE to move toward the pudding (which would kind of paralyze play). Instead, Bob needs to decide HOW to move toward the pudding.

I see the same problem - Bob's unable to move toward the pudding - and rather than solving it by allowing Bob to move toward the pudding (and deciding how), I'd prefer to solve it by making Bob's other options (to move toward the broccoli, to search for a secret door, to hang out at the inn, etc.) as equally appealing as the pudding. Even to the extent of having many other options, and no possible way to actually perform them all in the time allotted.

Because to me, it's an interesting decision when your character has to question their goals, often more interesting than a character who just has hurdles in place of achieving their goals.

I'm a big nerd, so my go-to is a project management metaphor. Fail Forward seems to be analogous to a waterfall model: a cascade of cause and effect that all leads to the ultimate goal of Your Character's Goal (the end product). The crevasse scene hands off to the ridge scene which transitions into the peak scene. I prefer a bit more of a spiral model or an agile model, with flexible goals (Do I really want the pudding?), risk assessment (what could go wrong in pursuit of pudding? What if I never get the pudding?), and iteration (can I get the pudding this way? Maybe that way? How did my last approach fail?). The end product here might not entirely be what we set out to create, but it is the product that arose from that creative process. The crevasse and the ridge and the peak are all there to charge into if you're ready, but you might fail any or all of them and have to go back and contemplate your pudding.

The latter choice allows my character to achieve discrete goals, but also lets them shift goals or abandon goals or realize that a certain goal just isn't going to work for them due to the limitations of the campaign as it plays out. It also keeps the interesting decisions focused around which goals you pursue, which helps display a character's personality and motivation - what's really important?

Those tend to be more interesting questions to me than "How do you get to the big bad?" or "How do you get the pudding?", by and large.

 

[Manbearcat]

In my BW game, the PCs had to spend a week crossing the Bright Desert, from an oasis protected by a good naga to the ruined tower that had once been the redoubt of the PC mage, in the foothills of the Abor-Alz.

I used the standard Fort checks to avoid tax to Fort, and when the Orientation check failed they (i) had to make an extra Fort check, and (ii) found the first pool at the edge of the desert already fouled by a dark elven adversary. A subsequent failed attempt to track down the elf meant that, when they got to the tower, he had also had time to dump rocks into its well.

I always seem to move towards NPC adversaries rather than nature as an adversary.

I've read through all of your recaps of your BW sessions (I believe). If you were able to keep an ocean voyage and a perilous journey across a desert interesting enough for your players to enjoy the time at the table, I suspect you're not giving yourself enough credit!

Nonetheless, given your attraction to romantic themes and your background/career in philosophy, it seems pretty intuitive to me that arid deserts and angry seas would inevitably give way to NPCs as primary antagonists!

 

[Lanefan]

If my goal is to climb the mountain unscathed and I arrive at the top of said mountain having lost something, then I've failed to achieve my goal.

Who said anything about unscathed? The goal is to climb the mountain. Scathed or not is merely a degree of success or a degree of failure depending which way the dice go.

As well, a failed check doesn't necessarily mean outright failure of the goal. It depends entirely on the game system. In D&D 5e, for example, the Basic Rules tell us that rolling below the given DC for an ability check is "...a failure, which means the character or monster makes no progress toward the objective or makes progress combined with a setback determined by the DM." This is D&D 5e's "fail forward," and it's right in the Basic Rules (page 58). Thus "a failed skill check roll is telling you" that climbing the mountain but losing one's divining rod is perfectly fine according to the rules of the game.

"Makes progress" does not equal succeed. So you initally fail at the climb, and the DM determines that halfway up you lost your divining rod...so far this is just fine. But you're still only halfway up, at the point you lost the rod (in my view the action stops here to give the player-as-character a chance to choose what to do next) meaning you have not succeeded in the climb and thus the narrative agrees with the dice. If the player-as-character chooses to continue the climb this would prompt another check to see if she makes it the rest of the way up; and if she instead chooses to climb back down and try to recover the rod you're into a different check, and so on.

Lan-"mmmm...pudding"-efan

 

[Bedrockgames]

This is correct. The fiction around Wassal is generated in response to the roll.

One way to look at it is this: by making a Circles check, the player is taking a gamble. If the player wins, he gets to make it true that the local captain of the tribesmen is a friendly former associate who will help the PCs out. If the player loses, I get to narrate something instead. "No one turns up" is a legitimate narration, but flagged in the GM advice as also the most boring option. "The enmity clause" is the more interesting option permitted to the GM - you meet the NPC you wanted to, but s/he is not disposed to help but rather to hinder. The GM has to narrate the fiction around that, but I hope you can see from my example that this fiction is not just spun out of nowhere but built around prior backstory and events of play.

The capture plays out at ground level - I narrate that the PCs are surrounded by evidently hostile tribesmen, and then there is a bit of back-and-forth between the PC mage and Wassal, in which some of the relevant backstory (eg the identity of the Desert Fox, Wassal's anger at orcs being brought into the desert) comes out. The capture is then a formality, in the sense that the players can tell that their PCs are no match for the tribesmen, and so when Wassal commands them to come with him back to his oasis camp, they comply.

Moving to a higher level of metagame, one reason the players are relatively happy to allow themselves to be captured is because they know that in this system, capture isn't the end of things but just another springboard to something or other. Upthread I quoted [MENTION=99817]chaochou[/MENTION] (hi, chaochou!) saying that failure is not penalised that heavily in these "fail forward"-type games. This is an instance of that truth being manifested in play. The failure is a real failure, but the players know that it won't be a block to their PCs doing stuff - it's just that the stuff they do (in this case, try to bargain with Wassal and persuade him of the truth about the orcs) is not the stuff they hoped to be doing (leading the tribesmen on a desert rescue mission somewhat in the spirit of Lawrence of Arabia).

Okay I think I am much clearer on how things work in your games and how Fail Forward functions in them. Thanks for sharing the campaign info. It is a different style but the Lawrence of Arabia vibe is something I like (I did a whole middle east campaign at one point).

This may be too long for your tastes, but I have logs of my most recent campaigns here if you are interested getting a sense of how I run things. They are not play by play, more just me getting the info down so I have it preserved (and I often gloss over individual rolls in the retelling). Still you may get a sense of what kinds of developments are likely to crop up when I am running; as well as a sense of the flow. This is the weekly game and on the right side bar I have the sessions listed under The Demon Moon Cult and the Secret of Je Valley respectively: http://thebedrockblog.blogspot.com

 

[Manbearcat]

Just a quick post. [MENTION=85555]Bedrockgames[/MENTION], there is a place for falls, death, and loss in games that feature/systematize Fail Forward is a/the primary technique for play propulsion. Just upthread I sblocked the majority portion of a conflict (you can reference it for context if you need/care to). Here is the (not sblocked) end of that conflict:

GM (Me):
The load off the goblin brother immediately invigorates him as his other hand firmly grasps the sled. Your heroic efforts and seeing his brother have to firm hand-holds on the sled instills further strength in him to survive.

The sled very, very slowly rises as the weight is still immense. Your fingers are growing so very weary. There is little chance that you can just hold on like this for the time it will take for the sled to rise to the top.

Otthor's player:
With my extra weight off the sled, I know it will rise more quickly. If I fall, so be it.

Defy Danger (Str)
1, 1 - 1 = 1

Mark 1 xp

My strength is gone. Before my hands let go of their own volition, I shout to Saerie in elven. "Fear not for me. Carry on. I will find you. May gentle breezes guide you and sweet waters comfort you, my friend."

I let go.

GM (Me):
As you descend into the darkness, Saerie appears in the crack of light above with a dangling rope for the goblins. The dog crests the top and the whole of the sleds begin to ascend rapidly.

You plunge into icy water and are carried in a hard current over unforgiving rocks.

- You take b[2d8] from the freezing cold and the rocks, no armor applies.

Otthor's player:
8, 3. I take 8 damage.

GM (Me):
Somewhere in the freezing darkness, amidst the endless tumbling, consciousness was lost. When you open your eyes, you're immediately greeted by utter darkness, the sounds of running water and a heavy, intermittent drip on your face. You seem to have washed up on a shore and your arms have clung to a rock without your commanding them to do so. Your legs are aching as they are still in the cold waters. Despite the darkness, your innate, warrior-honed sense of spatial perception tells you that you are in a place with an exceedingly high ceiling. There is a distinct lack of air circulation and the temperature in this place is much warmer than the surface, especially without the biting wind. Nonetheless, you are cold...freezing to death, in fact. You know that if you do not find warmth soon, you will freeze to death...

What are you doing?

I'll stop there. The players Undertook a Perilous Journey to transit the frozen tundra between the ruins of World's End Bluff and Earthmaw, a hobgoblin redoubt that serves as the primary trading outpost in the highlands territories. There are several reasons that they wanted to go to Earthmaw:

1) To find the refugee families (if any) from World's End Bluff.
2) To Resupply...they were perilously low on Rations (with many followers/companions mouths to feed!), Adventuring Gear [a catch-all, abstract resource that gives bonuses/grants an outright "yes" when you expend a use], and Ammo.
3) To Parley with the hobgoblin king and hopefully gain audience with his Blizzard Dragon patron (who is the alpha of the entire highland realm).

Otthor's fall through the crevasse into the icy underground river would have outright killed most mere mortal men and women. But he is a PC with a level of plot immunity (PC HPs). So he survives...and washes up in Earthmaw's basement garbage dump where a Darkmantle and a Roper skulk in the deep frozen dark to guard the back entrance if the hobgoblin redoubt and serve double duty as sanitation workers. If you read the sblocked text in the post above, you'll note Saerie lost her coinpurse in the desperation of trying to get herself, her companions, and her sled out of the mess of falling through collapse of the thin ice over the crevasse.

So, they arrive at Earthmaw...just in a fashion that diminishes their resources and their capacity to get what they want (2 and 3 are made much more difficult because of the failures accrued in that scene.). The whole of it is the kind of thing that is spontaneously generated during play. In the same way that a Spout Lore check (like a Wise in BW) spontaneously generated facts about Earthmaw and the hobgoblin's Blizzard Dragon patron.