Failing Forward: yay or nay ?

[silva]

A concept I was recently introduced to (through Apocalypse World) is failing forward mechanics. For those who never heard of it, it makes basically two things:

1) Momentum. Keeping things moving in the game, instead of the classical pass/fail resolution where the "fail" simply stops the game in its tracks (possibly making the player roll again).
2) Making rolls interesting, as each time the dice hit the table it can lead to new and unexpected situations.

Eg: a thief is trying to picklock a door. In a classical pass/fail resolution, in a failure nothing would happen and the thief could try again. While in a failing forward the failure could mean the thief broke its lockpick making a echoing sound throuhg the dungeon corridor (and possibly alerting nearby creatures).

So, what you guys and girls think about it? What other games use this concept ? (btw, I heard the new Star Wars rpg uses this through special dice that adds new situations beyond the basic pass/fail. Have someone here tried it ? )


EDIT:

Consider the term used here as applied to both failures and sucesses, and as forward, backwards, upwards, sideways, diagonally, etc. consequences to the situation at hand. In other words: It means that something interesting happen on failure/sucess and the game does not roadblock itself.

[Username17]

Failing forward doesn't make the game interesting, it makes your failures as well as your successes meaningless. Failing Forward was most famously introduced for 4th edition Dungeons & Dragons when they realized that they had fucked up their skill challenge math so hard that they needed the plot to move forward on character failures because characters would fail most of the time.

Having your game fail forward means that you are simultaneously too shitty of a designer to give your players abilities that will allow them to move forward with success, and that you're too much of a pussy to let the game branch off into a genuinely different direction when the players didn't do what you expected them to (whether that is failing on a task or merely making an unexpected choice). Anyone who needs "failing forward" in their quiver is simultaneously a shit designer and an uncreative Mister Cavern.

-Username17

[vagrant]

How is failing forward different than a glitch mechanic (in your example)? The story doesn't progress forward, just new complications are added.

[silva]

Failing Forward was most famously introduced for 4th edition Dungeons & Dragons when they realized that they had fucked up their skill challenge math so hard that they needed the plot to move forward on character failures because characters would fail most of the time.

The problem here seems more like a bad implementation of the "skill challenge" concept, than of the failing forward concept per se. No ?

[silva]

How is failing forward different than a glitch mechanic (in your example)? The story doesn't progress forward, just new complications are added.

There is none. The difference will be one of proportion, I think. While in, say, Shadowrun, you will have a glitch that moves the game forward in only 5% of the times (and a simple fail where nothing happens, say, 40% of times) in a Failing Forward game that will happen everytime the dice thi the table.

Im curious to see how thats handled in the new Star Wars Edge of the Empire.

[Whipstitch]

I'm comfortable declaring Shadowrun glitches a bad idea.

[Chamomile]

It depends on exactly what you mean by "failing forward." Now obviously what Silva means by "failing forward" is that Apocalypse World is the greatest RPG in the history of forever and everything it does is without flaw and we humble mortals are blessed to read such holy scripture, but ignoring the fact that Silva's arguments are once again backed by rampant fanboyism rather than any actual idea of what he's talking about: It is certainly true that if the only penalty for failure is that the gamestate does not change, then there's no reason not to continue trying over and over again until you get it. And thus: Take 20 rules. Which are boring and have no drama to them whatsoever. So it should generally be the case that whether you succeed or fail the gamestate should be changed in some way, and that the gamestate should be changed in ways that are meaningfully different from one another based on whether you succeed or fail.

Frank seems to have taken the tack that "failing forward" means that you succeed even if you fail, I guess? Or that the scenario progresses forward in the same way it would have before. And since failing forward has no actual agreed upon definition, I guess he might be right. But I think what Silva was actually getting at is that failure irrevocably changes the gamestate somehow. Which is not a phenomenon unique to Apocalypse World and has indeed been a staple of good GMing for at least a decade, but I stopped expecting Silva to know what he was talking about a long time ago.

[silva]

Another way to look at it: You encounter a locked door. You try to unlock it. What happens?

1) You unlock it without issue and proceed.
2) You unlock it, but break your lockpicks in the process
3) You unlock it, but make so much noise that guards are alerted to your presence.
4) A combination of 2 and 3
5) 2, 3, and 4 but you fail to unlock the door
6) You don't unlock the door

These are all possibilities, and there's countless more that I haven't thought of. Of these, "normal" pass/fail mechanics only cover 1 and 6 - which are arguably the *least* interesting choices. Choices 2-5 are just as realistic, and a hell of a lot more interesting. (of course, YMMV and all that)

[silva]

Chamo nailed it. What is this "Take 20" rules ?

[hyzmarca]

Frank seems to have taken the tack that "failing forward" means that you succeed even if you fail, I guess? Or that the scenario progresses forward in the same way it would have before. And since failing forward has no actual agreed upon definition, I guess he might be right. But I think what Silva was actually getting at is that failure irrevocably changes the gamestate somehow. Which is not a phenomenon unique to Apocalypse World and has indeed been a staple of good GMing for at least a decade, but I stopped expecting Silva to know what he was talking about a long time ago.

It's probably more precise to have multiple directions of failure.

Failing Forward - The plot advances in a direction that is positive for you if you fail.

Failing Backwards - Failure is a setback. Maybe you have to retreat. Maybe the bad guy just leaves the building and the entire dungeon crawl up to this point is wasted. Either way, failing is objectively bad but it doesn't end the plot.

Failing Sideways - Failing blocks off your current path and requires you to take a new one but doesn't set you back. The classic failing sideways scenario would be a room with two doors. If you fail to open one then you must open the other.

[Chamomile]

Chamo nailed it. What is this "Take 20" rules ?

I suspected for a while that Silva's fanboyism was inspired by a total dearth of experience with anything except AW and maybe 4e or something, and here's the proof! Silva: If you want to talk about RPGs in general you should probably play or at least know the rules for more than one of them.

Taking 20 is a not-at-all obscure rule from the most commercially popular RPG system yet to exist, D&D 3.5e, in which you take twenty times as much in-game time and twenty times as much whatever the other costs are, and then immediately assume you got a 20 on the roll. The idea is that instead of actually rolling the die over and over again until you get a twenty, you just say that you're taking 20 and getting basically the same thing: It's assumed that it takes you exactly 20 attempts to get a 20 on the die and then the game moves on.

EDIT: Hyzmarca, you're not getting it. What you call "failing backwards" is failing forward as defined in the OP. You fail and the plot moves forward, because now you've encountered some kind of significant setback that changes the gamestate. "Failing sideways" is also failing forward, because it also changes the gamestate by making future attempts impossible. If you want to argue that "failing forward" is a really stupid and misleading thing to call this concept, you would not see any arguments from me on the subject.

The alternative to failing forward is that you fail and nothing changes which also means that there is absolutely no cost to trying again. And it is definitely a bad idea to ever do that because then that's just a problem you solve by brute forcing it. But that statement is so uncontroversial that I'm baffled that Silva 1) thinks that this is a new idea and 2) thinks that we need to talk about it. It reminds me of Lago's tired old "Fighters suck" threads. The point is so completely uncontroversial that it's obviously just an attempt by the OP to piggy-back their more controversial agenda on something that's easier to argue for. So the mickey that Silva's trying to slip us is that AW=Good GMing, therefore an attack on AW is an attack on Good GMing, in the same way that Lago has in the past tried to convince us that Fighter/Caster Parity=Whatever insane violation of player agency for their own good he's cooked up.

[silva]

Interesting concept.

BTW: nah, D&D was never my thing. Too thematically childish for my tastes.

[silva]

It's probably more precise to have multiple directions of failure.

Failing Forward - The plot advances in a direction that is positive for you if you fail.

Failing Backwards - Failure is a setback. Maybe you have to retreat. Maybe the bad guy just leaves the building and the entire dungeon crawl up to this point is wasted. Either way, failing is objectively bad but it doesn't end the plot.

Failing Sideways - Failing blocks off your current path and requires you to take a new one but doesn't set you back. The classic failing sideways scenario would be a room with two doors. If you fail to open one then you must open the other.

Well, this is actually what I mean with the term. The game can go foward, sideaways, backward, diagonally, etc, etc, etc. both through failures and sucesses.

[echoVanguard]

I think you have a fundamental misunderstanding of how glitches are supposed to work (and, potentially as a side effect, about how Failing Forward works).

Glitches are irrelevant to the success or failure of a roll - they add a complication independently of the outcome of the roll. It is possible to simultaneously glitch and succeed. In your example, the rogue failed to open the lock (desired action did not succeed) and he broke his picks and potentially alerted nearby creatures (1-2 additional negative complications). In a standard glitch-based system, there are four outcomes, not two, to the scenario proposed:

• Success - the thief picks the lock on the door silently.
• Success with Glitch - the thief picks the lock on the door, but incurs a complication, such as breaking a pick, making a noise, or ruining the lock.
• Failure - the thief fails to pick the lock.
• Failure with Glitch - the thief fails to pick the lock and also incurs a complication, such as breaking a pick, making a noise, or ruining the lock.

By contrast, Failing Forward is the idea that the plot moves forward whether you succeed or not, which is different from glitching in that glitches are perpendicular to the advancement of the plot. Frank is correct in his assertion that failing forward trivializes successes - if the plot advances regardless of success or failure, then both success and failure are equivalent from the standpoint of the plot and the PCs have no reason to favor one over the other. In fact, if I get the same rewards from failing that I do from succeeding and failing is easier, players might be tempted to fail on purpose to advance the plot more quickly or easily.

echo

edit - partially ninja'd by silva's 4th post.

[silva]

Echo, you misunderstood what Failing Forward means. It has nothing to do with "plot". Instead it means the addition of situations linked to the roll( be it a sucess of failure). These aditions couldn be actually forward, backward, sideways, upwards, whatever.

(yes, I know the term is actually crap, but it wasnt me who coined it, so.. )


EDIT: could we be talking the same thing here, only using different labels ? ("glitches" and "failing forward" ? )

[Pulsewidth]

I'm comfortable declaring Shadowrun glitches a bad idea.

Glitches serve an important role in discouraging dump stats.

[silva]

What you call "failing backwards" is failing forward as defined in the OP. You fail and the plot moves forward

I think saying the plot moves forward is not the best chosen words here, because it may give the impression of railroading on the part of the GM. I think the game moves forward is a better word.

I'm baffled that Silva .. thinks that this is a new idea

Yup, I really do. The only games Ive seen where the concept used in a meaningful way as to really impact/drive the gameplay (and not as an optional or inconseuquential rule) is Warhammer 3e, Apocalypse World and now Star Wars Edge of the Empire. What other games use the idea ? Im genuinely curious.

[hyzmarca]

I'm comfortable declaring Shadowrun glitches a bad idea.

Glitches serve an important role in discouraging dump stats.

No they don't. Glitches just discourage you from attempting to use your dump stats to do anything, which you wouldn't anyway. They're dumpstats because you don't intend to use them.

What gliches attempt to do is make things "interesting" be introducing unpredictable complications.

[echoVanguard]

Echo, you misunderstood what Failing Forward means. It has nothing to do with "plot". Instead it means the addition of situations linked to the roll( be it a sucess of failure). These aditions couldn be actually forward, backward, sideways, upwards, whatever.

(yes, I know the term is actually crap, but it wasnt me who coined it, so.. )

EDIT: could we be talking the same thing here, only using different labels ? ("glitches" and "failing forward" ? )

I am using the term "Glitch" in its Shadowrun 4E implementation (that which is most relevant to discussions of RPG mechanics) and the term "Failing Forward" in its most relevant implementation relative to RPGs. If you want to get technical, Failing Forward has a fairly robust definition independent of RPGs, and it is totally inconsistent with what you're describing. It sounds like the Shadowrun implementation of Glitches are what you really want to discuss, although your original post posited a somewhat chaotic model of outcomes. I definitely second Chamomile's sentiment that you may want to do more research on RPG topics in general if your only exposure to gaming has been Apocalypse World - even if you don't like the thematics of a system, you can learn an awful lot about good design from some of the more successful systems on the market. GURPS, Call of Cthulhu, D&D 2-3.5/4, Shadowrun 3/4E, White Wolf's WoD (and associated systems such as Scion) and FATE/Dresden Files are a nice cross-section of mechanics and approaches. D&D 4E is an excellent example of how to learn from bad implementations - I wouldn't recommend that anyone play it, but I would definitely recommend that everyone study it at least a little.

echo

[silva]

Not necessary. I play rpgs for more than 20 years already - began with Shadowrun 2e and from there to Gurps, Vampire, Runequest, Risus, etc. Enough to avoid 4e only by reading about it.

Im curious to what other games actually use the concept besides the ones I mentioned (Star Wars Edge of the Empire, Warhammer 3e and Apocalypse World and its progeny).

[shadzar]

Falling forward is great, though many people call it walking.

[Whipstitch]

No they don't. Glitches just discourage you from attempting to use your dump stats to do anything, which you wouldn't anyway. They're dumpstats because you don't intend to use them.

What gliches attempt to do is make things "interesting" be introducing unpredictable complications.

Also, in many cases I'd rather default to a pool of 0 and just fail rather than risk ultra-failure on a test that requires more successes than I am capable of rolling with one or two dice anyway. I like the dice pool w/ fixed TN mechanic in theory, but the glitch rules greatly undermined many of its advantages by introducing weird "Odd numbers are better" breakpoints for little real gain. Besides, due to dice bloat you were rather unlikely to glitch on lots of tests anyway as more splats were added to the system. I mean, granted, nobody ever dump statted Perception in Shadowrun, but even with just basic gear and implants it wasn't too hard to get your pools up to 5 or greater. Same deal with Etiquette and emotitoys and social softs--getting negotiating tips from Tickle-Me-Elmo wouldn't be enough to give you leverage over the Fixer and his Moodwyrm, but oftentimes it gave you just enough dice to pull out a single success and thus at least avoid ultimate humiliation even if you dump stat'd CHA.

[Prak]

Frank, out of curiosity, have you actually played any games with Failing Forward type mechanics (__ World, Fiasco, 4E, etc)? By all accounts, the actual players of such games seem to prefer them.

[silva]

From my reading of SR4 Glitches, it doesnt really look a "Move the game Forward" mechanic at all. It looks more like a variant/twist on the critical failure concept of old SR editions (the old "Rule of 1"). Specially if you consider the frequency/probrability of it actually showing up during sessions. If SR4 designers actually intended the Glitches to be a "Moving the game Forward" mechanic, they did a bad job at it.

[Aryxbez]

Not necessary. I play rpgs for more than 20 years already - began with Shadowrun 2e and from there to Gurps, Vampire, Runequest, Risus, etc.

Amount of RPG and time can be quite regardless, it's the understanding you attained in that time, especially of the individual systems themselves. If lack Critical thinking or objectivity for example, very easily get people like grognards, who worship a system they "still" hardly know anything about. Discussing dishonestly about RPG design, and continuing this self destructive paradigm that hates on discussion, knowledge of systems, and otherwise ensuring everyone in a group is on the same page. Given what I've heard from you on here, you don't speak like someone with 20 years of experience, and thus I can see why people think you're very much new to RPG's, and its designs.

On the notions of Glitches/Failing-Forward, I think it would help to have degrees of failure that can be activated. Opposed to Apocalypse World's way that be as minor as cosmetic defect, to: failed to pick the lock NOW EAT THAT BABY. This way, it's bit more consistent, and allows for less unreliably minor or debilitating choices that your average DM might do accidentally.