For the record I have never been in a game where we spent hours, much less entire sessions, trying to figure out a puzzle. I'm not even a big fan of puzzles where it's solely down to the players to solve. Especially some of the weird math puzzles that we'd just hand to Jeff to solve because he knew how to do them and the rest of us had no clue. In those cases it's not that we wouldn't have incorrect guesses, we wouldn't have any guesses at all. I don't use them, and the last time I encountered one was over a decade ago when I was playing a series of modules for LFR written by the same author.
Even if I had people in my group that enjoyed puzzles like that 99.9% when we talk about failure we aren't talking about literal puzzles. Unless you're
@Lanefan, puzzles that take hours or even entire sessions to solve the percentage of puzzles that take more than a few minutes are approximately 0.
OK, but substitute puzzles for anything else out there. And even if
your game never has any times where the players are stuck for ages trying to get past a specific point, there
are games where that happens.
"But why aren't the players doing something else and looking for a different way around?" Not the point. If every player out there always looked for another way around, and if every GM out there thought to put multiple ways around, we wouldn't be having this discussion.
Well there’s the gap in communication. I’m not saying your implementation of fail forward does that. I’m saying why isn’t that technically/definitionally one implementation of it? I don’t think @Maxperson is right that it has to be something interesting. Most of my blades in the dark successes with consequences weren’t particularly interesting., take some heat or stress or click a clock or something. What each of these things did was interact with the rules in such a way as to mean we were one step closer to the complication building up into a major problem. That the particular thing we were trying to accomplish that invoked this roll was resolved and in the rear view mirror.
Personally, I'd consider "building up into a major problem" to be interesting! Maybe not
immediately interesting, but definitely tension-building because
something is going to happen.
But Blades in the Dark also had outright failure and in those instances it could very much be result in you tried to hit the enemy he blocked it and counterattacked landing a solid blow. Take a trauma. The enemy is still there and needs dealt with. But mechanically the game the game moved one step closer to resolving said conflict either in your favor or the enemies.
Sure, but combat is (nearly) always the exception in RPGs because there's an opponent who fights back, whereas with other skills, it's typically static. Not always, true--that one mirror that attacked on a wrong answer is a rare exception, and there
are lock traps--but most of the time.
This seems to be your definitional requirement, that the consequence failure not drastically reshape the fiction, but that it brings resolving a conflict one way or the other one step closer. This explains why the riddle example doesn’t work for you. It’s not about change per se, but about steps closer to resolution.
But it's not a step closer because it provides no useful information other than "this thing doesn't work." We don't learn
why it doesn't work, nor do we get a clue as to what
would work (at least not according to the description we were given of that riddle). All we learn is that one possibility out of potentially an infinite number of possibilities doesn't work. And even if it were a more reasonable number, it shouldn't be a case of crossing off options on a list until you hit on the right one, because that's a long time in real-time to be spent getting past one point and, quite frankly, boring. OK, combo 5 didn't work, let's try combo 6.
If it's steps towards a resolution, then those steps should be something useful. Whether this is real info the players can use to solve the puzzle, a tick on a clock that's counting down to something bigger, or some other method of getting past gets highlighted ("you don't pick the lock, but you notice a third-story window that seems to be partially opened.") depends, of course, on the task in hand.
Except for d&d combat, which is different than the rest of the game, most d&d actions aren’t going to necessarily bring you a clear step closer to resolution (and even in combat it’s not uncommon to miss entirely and change absolutely nothing on your turn). I think it might be interesting to touch on why d&d doesn’t do that kind of thing. But I think that’s probably best for another post.
True, and that's a major problem with D&D. I didn't play 4e so I don't know how well that edition's skill challenges worked, but I can't imagine it would be difficult to use them in 5e. Level Up has their own countdown system for major tasks, and that's
designed to be ported into regular 5e. And it probably wouldn't be hard to bring either of those into any earlier edition, even if you have to use stats instead of skills in 1e or earlier.