Failing Forward

[iserith]

I might argue that it isn't, actually. At least not from a player's perspective. The character probably seeks that, in-character, but the player seeks complications and difficulties that they can then overcome (or fail to overcome) to show off the personality traits and fantastic abilities of their character.

Climbing the ravine on the way up Mt. Pudding is the complication and difficulty to overcome in this example.

The player wants to roll dice and do math to defeat challenges. An easy climb to the top, while it might be what the character has in mind, probably isn't satisfying for a player (too easy, too uninteresting).

Let's not confuse the example again as I did upthread. The ravine is likely one of a number of challenges on the way to the peak of Mt. Pudding. I would add that rolling dice isn't typically to the player's benefit in a game where the GM decides on success, failure, or uncertainty. (The latter case is when we roll dice or otherwise resolve with some mechanic.) Hoping to get lucky with the dice isn't a good plan. Striving for outright success is better.

The old example of failing a social interaction so that it degenerates into combat is handy here - when the incentive is toward combat (either directly, through things like XP, or indirectly, through robust systems that allow the player to control dynamic interactions), degenerating into combat isn't a punishment, it's more fun for the player (even though the character might've sought to avoid that).

Difficulty itself isn't a disincentive, it's something you WANT as a player.

At least until it becomes so difficult that your goal can't be realized.

But the examples of Fail Forward so far don't seem to comfortably accommodate a goal that can't be realized, so that level of difficulty would also seem to be off the table, if you're applying the design logic of Fail Forward.

I think some more issues are getting muddled here. Of course difficulty is desirable - it's part of what makes a challenge satisfying. And being desirous of interesting success and failure conditions is good and I share that (again, that's just stake-setting which is all fail-forward is!), but as some have stated, they don't care if sometimes things turn out not to be so fun as long as the net fun over the long haul is positive.

 

[iserith]

I've given the one way that I can see that the rod could be a part of the climb check. A pack that isn't secured well or damaged is something else entirely. It could break and the rod is lost due to bad luck, and it could even happen as a result of the failed climb check, but there would still be the failed climb check to contend with. The loss would be a separate result that was triggered by the climbing failure, not the actual climbing failure. The climbing failure would be a failure to advance in the climb, a loss of progress, or even falling.

I understand what you are saying, but I don't agree with it. A failure to climb is a failure to climb in some way and nothing else. The violence of that failure could result in the loss of items, including the rod, but that loss is a separate cost that is only indirectly tied to the failed climb check.

What you appear to be saying is as another poster essentially said upthread: The stakes can only be this one thing. But again, that's a subjective call by you or possibly by the game system being utilized. Those that employ fail-forward can choose to either have the character fall into the ravine as a result of the failed check or drop the divining rod (or lots of other reasonable outcomes). We use whichever one is most interesting at the time. I see no value in limiting myself to a single set of stakes as it might not always lead to fun.

 

[Maxperson]

What you appear to be saying is as another poster essentially said upthread: The stakes can only be this one thing. But again, that's a subjective call by you or possibly by the game system being utilized. Those that employ fail-forward can choose to either have the character fall into the ravine as a result of the failed check or drop the divining rod (or lots of other reasonable outcomes). We use whichever one is most interesting at the time. I see no value in limiting myself to a single set of stakes as it might not always lead to fun.

I totally agree that you can change the stakes from climbing to something else. You should absolutely do what is best for your game. For my game, though, a climb check is going to check climbing and not a rod in a pack. A roll checking whether a rod stays in a pack is not a roll that is checking climbing. Circumstances around the failed climb check may also cause the loss of the rod, but it will be caused directly by the failed climb check.

 

[iserith]

I totally agree that you can change the stakes from climbing to something else. You should absolutely do what is best for your game. For my game, though, a climb check is going to check climbing and not a rod in a pack. A roll checking whether a rod stays in a pack is not a roll that is checking climbing. Circumstances around the failed climb check may also cause the loss of the rod, but it will be caused directly by the failed climb check.

May I ask which game you play the most?

 

[Janx]

I totally agree that you can change the stakes from climbing to something else. You should absolutely do what is best for your game. For my game, though, a climb check is going to check climbing and not a rod in a pack. A roll checking whether a rod stays in a pack is not a roll that is checking climbing. Circumstances around the failed climb check may also cause the loss of the rod, but it will be caused directly by the failed climb check.

what would ever trigger a roll checking whether a rod stays in a pack then while climbing another 10' section then?

Because as worded, in your game, nothing interesting outside of the strict parameters of a check could trigger something unexpected like that.

 

[Janx]

Right, but when light bulb #1 blew out, Edison had no chance of completing the light bulb with that failed check. He failed. The same goes for the climb check. If you fail, you should have no chance to succeed in the climb with that failed check. Like Edison, you have to try again in a different way. Failure for #2 at making a light bulb wasn't his screwdriver falling out of his pocket and the light bulb working.

I think there's part of your misunderstanding. Fail Forward most likely does not mean on attempt #2, he fails the roll and the light bulb works anyway.

It more likely means that the GM has something interesting happen that makes it possible for attempt #3 to be attempted (note my wording does not imply guaranteed success, only the opportunity to try again or try something else).


I made a sword rack this weekend. Quick project, it had some failures, but we worked around them. Used a nailgun. Had 3 nails go awry. We snipped them off as it was easier than backing them out. Planned out the top rails at the wrong spot, and to move and renail them once we realized I designed it wrong. I succeeded in my goal, despite individual failures.

Now as this was real life, there was no GM. I simply worked around the failures as I encountered them. A GM with a situation where the players are still "working the problem" and moving forward (forward being any direction from the North Pole) doesn't need FF. FF is likely a better tool for when the game is going to stall out or get boring (those 20 climb checks on Mount PudMore...).

 

[Imaro]

If your character is not known for his fumbling fingers or to be an "incompetent butterfingers," as you say, then it may not follow from the fiction that you drop your divining rod and thus would not be a good choice of narration for the GM. You're arguing against something nobody is stating even though you seem to understand that the GM's narration must follow from the fiction (bolded in your quote).

But anything not defined prior is up for grabs in the fiction...correct? If so I'm not arguing against "something nobody is stating"... especially since you made no reference to the character being a klutz before trying to climb a mountain peak. My issue is that as the DM you've taken the liberty to create a narrative (whatever that narrative may be) that may not gel with the image I have of my character based on a roll that had nothing to do with the consequences you narrated.

The consequences are a part of the stakes which are set prior to the check and based on the fiction up to that point. No agency is lost here. You are still acting freely and impacting the fictional world by your action. I explained this clearly upthread. If you are moving around searching for a trap and you fail the appropriate check, then it is reasonable that you step on a pressure plate. (You found the trap, but...") If your stated goal and approach clearly did not include moving about to search, then stepping on the pressure plate after a failed check would not follow in the fiction and would be thus unreasonable.

So you tell the players what the fail forward consequences will be before every check? [MENTION=42582]pemerton[/MENTION] said he depended on the trust of his players and I can understand that but this seems kind of clunky... Also, if so... how does this gel with the "on the fly decision making" that characterizes fail forward? which was referenced earlier in the discussion?

As to whether you throwing me into a trap because I failed a search check for it or not is "reasonable"... well that's a matter of opinion. I don't think it's reasonable unless I take specific actions that warrant me springing it... what if the reason I missed it was because I was nowhere near it while searching the room? You as DM have eliminated that possibility and chosen the one you want...based on what exactly? Because you wish for the story to progress in a certain direction or way that you find interesting?

It's not railroading or "railroady" by any definition of the term as I understand it.

What's the definition of the term as you understand it?

Depending on the goals of play of the game being played, perhaps you should be - and what the players find fun and interesting, too. Or what contributes to the creation of an exciting, memorable story. Or whatever.

It's the "whatever" that bothers me.

If you don't know, you can ask. As I've mentioned, offering the stakes prior to the roll makes sure everyone is on the same page and allows for the occasional renegotiation from a player such as in the case of:


Whereupon you could say, "Sure, that's sounds good. Roll for it."

So now we're having debates and negotiations around every failed roll because we don't want the game to slow down or be stalled... that seems kind of counter-productive to the overall goal... all IMO of course.

 

[iserith]

But anything not defined prior is up for grabs in the fiction...correct?

Correct.

If so I'm not arguing against "something nobody is stating"...

You are, if you are arguing that anyone advocating fail forward as a technique is suggesting that the GM narrate a result that doesn't follow in the fiction. It seems like you did exactly that.

...especially since you made no reference to the character being a klutz before trying to climb a mountain peak. My issue is that as the DM you've taken the liberty to create a narrative (whatever that narrative may be) that may not gel with the image I have of my character based on a roll that had nothing to do with the consequences you narrated.

You don't have to be a klutz or an incompetent butterfingers, as you say, to drop something. What I'm saying is that if it has been established that your character is specifically not those things - perhaps you're a rogue that can juggle daggers and is fastidious about securing his gear - then dropping the rod may not follow in the fiction. Some other cost or complication might apply better.

So you tell the players what the fail forward consequences will be before every check?

More often than not, but not always. Sometimes the stakes are self-evident based on the context up to that point.

[MENTION=42582]pemerton[/MENTION] said he depended on the trust of his players and I can understand that but this seems kind of clunky... Also, if so... how does this gel with the "on the fly decision making" that characterizes fail forward? which was referenced earlier in the discussion?

I'm not exactly sure what you mean by the "on-the-fly decision making." The GM is making a decision on the spot as to the uncertainty of the player's goal and approach and coming up with the stakes (ideally in my view) prior to the roll. Is that what you mean?

As to whether you throwing me into a trap because I failed a search check for it or not is "reasonable"... well that's a matter of opinion. I don't think it's reasonable unless I take specific actions that warrant me springing it...

Which is my position. I believe reasonable specificity should be garnered by the GM from the players when it comes to stating a goal and approach. "I search for traps" isn't adequate. By what means? Where? How much time do you spend on it? etc.

What if the reason I missed it was because I was nowhere near it while searching the room?

If you're nowhere near the thing you're searching for, you likely fail outright, no roll. The outcome is not uncertain.

You as DM have eliminated that possibility and chosen the one you want...based on what exactly? Because you wish for the story to progress in a certain direction or way that you find interesting?

You, the player, have eliminated the possibility of finding the trap because you were not in the fictional position to find it.

What's the definition of the term as you understand it?

Control of a player's decisions by someone other than the player (typically the GM) in a way that violates the social contract. You decided to search for the trap or climb the ravine. The GM gets to say how that turns out, sometimes using mechanics and dice to determine an outcome. The GM didn't decide that you searched for traps or climbed the ravine.

It's the "whatever" that bothers me.

It shouldn't because "whatever" is "whatever the goals of play for that game are."

So now we're having debates and negotiations around every failed roll because we don't want the game to slow down or be stalled... that seems kind of counter-productive to the overall goal... all IMO of course.

I think you're conflating slowing down the unfolding narrative with slowing down gameplay. Which is not to say it slows down the game play at all. It's not a matter of debate when people are acting in good faith, just simple page-setting.

 

[innerdude]

To me, the concept of "fail forward" could be more easily summarized as, "The GM should act in good faith on behalf of his players and his players' characters."

I use fail forward concepts regularly simply because it more approximates "real life" scenarios than not. There's very, very few scenarios in real life that have exactly ONE path, ONE solution, or ONE understood outcome. We are constantly making decisions that incorporate a variety of variables, with information that we both explicitly and implicitly know. This is in line with the idea that characters in an RPG know much, much more about their implied game world than the players do.

The question becomes, at what point does "fail forward" stop, and the final outcome of an attempted action, or set of actions, become a binary pass/fail? Because I also agree that when failure is not a possible outcome, that it lowers the tension, stakes, and drama in a game.

First, I generally tend to shut off "fail forward" when opposed checks/rolls are being made. If what a character is doing directly affects another character in the game world, I'm generally apt to have the consequence simply be what it is.

I also generally turn off "fail forward" when the stakes are either very low or very high. When the characters have multiple other options available to them, a failed lockpick check is just a straight failure---the party simply has to bash the door down instead of going around. The consequence of having failed to pick means they have to draw attention to themselves by bashing down the door, or find some other means.

The "high stakes" disabling of "fail forward" is pretty obvious --- when the stakes are high enough, I don't want to devalue the decisions that have been made leading up to that point. This might be stuff like heroic self-sacrifice of a character, a high-stakes negotiation with far-reaching consequences, or combat.

In all of this, however, the real idea is to act in good faith with your players. In most cases in RPGs, success is more interesting than failure---but it's the failures that make the successes rewarding.

 

[Neonchameleon]

Right, but when light bulb #1 blew out, Edison had no chance of completing the light bulb with that failed check. He failed. The same goes for the climb check. If you fail, you should have no chance to succeed in the climb with that failed check. Like Edison, you have to try again in a different way.

Except that a lot of those tries will have been iterative. Bulb #67 has failed. Try again. Exactly the same design but using slightly thicker glass. Or even Bulb #650 actually provides light. It's just not going to be economical. Exactly the same design using a different metal for the filament.

And then there was "Dropped bulb 77. Same again for bulb 78." And "Dropped bulb 112. Same again for bulb 113." Repeat every 50 to 100 bulbs where you were trying the same thing exactly the same way. Plus "Glassmaker screwed up this batch. Bulbs 161-180 all cancelled. Same again for 181-200."

You normally try to change one single factor at a time. And even a failed check to invent the light bulb will have brought Edison ever so slightly closer. So "Can't try again the same way" is an artificial gamist limitation that makes for a much better game (dozens of tiny iterations is boring) but doesn't match real life.