Not Failing Forward

I might be a bit old-school.  I started gaming in 1984, give or take. Cut my teeth on Palace of the Silver Princess, picked up Call of Cthulhu as my second ever RPG, and then ended my AD&D experience with Throne of Bloodstone before I moved on to other games in another part of the country.

I understand the idea of "fail forward", and I understand the intent. The GM wants the PCs to make it to the BBEG, and doesn't want the road there to be blocked. The GM doesn't want the players to fall to 'accidents' and crappy dice rolls along the way.  No, I totally get this.

And in some games, sure.  Okay.

But not every game.
Here's the thing for me. If the hazard isn't actually hazardous, then there's no point to it being there. Some GMs may talk about 'tension' or 'dramatic scenery', but I'm fine with doing that as background, not as a 'threat that isn't really a threat don't worry about it'. If you're in a swordfight in a burning building, you're on a time limit and if you're not out in time, you're crisp.  Or you die of smoke inhalation. Or something falls on you. The house is a hazard, treat it as such.

If the PCs run into four goblins as a 1st level party, I don't care if it's a 'random' or 'small' encounter. They're a threat.  They can very well kill the party long before the party gets to the actual dungeon. They're a threat, treat them as such.

"Well, that's not a fun experience" I've been told.
No. Nobody likes to lose - not just 'fail', but lose.
But that is usually part and parcel of the game. If it wasn't, where are there skill checks, saving throws, to-hit rolls, hit points, AC, or whatever the hell indicating that success and failure are things?

If you don't want the chance of failure, and the consequences that come from it, 1) don't put the threat there, and 2) don't ask for a roll.

Look, I want the PCs to win as much as the next guy - I just want that win to be earned through player decision making, the capabilities they picked out and put down on their sheet, and the luck of the dice.  The player, over time, is going to take things to mitigate the luck of the die - the character gets better, and the stuff that might have been a problem before, isn't now.
But it needs to be earned. Handing it to the PCs isn't earning it - it's one reason I don't like GMs fudging.  And sometimes, 'failing forward' feels a lot like fudging.

An example I used a long time ago:  The PCs have run out across a bridge. The bridge is rope and fragile, and below them is a river of magma. Enemies on the other side cut the ropes on the bridge, causing it to fall (the PCs didn't kill them fast enough, tried to race to them, the monsters got a critical hit on the ropes or just did a lot of damage, pick an outcome).  The gap on either side is too far to jump or run (and Mythbusters already proved you can't run / jump from a falling bridge, but whatever).

The GM asks for a check.  Could be Dexterity, Strength, Athletics, whatever. The success allows the PC to hang onto the bridge and not fall into the magma.  Well, then what happens if they fail?

Someone mentioned 'fail forward'.  'You grab the bridge, but...'
So I asked, "How many chances do the PCs get?"
The GM replied, "All of them."

"They fall onto a ledge below."
I don't remember mentioning a ledge, and how did they fall diagonally to get to the ledge if there was one?

"They leap to the edge and hang on."
Already mentioned, too far to jump.

The answers I got, more often than not, was trying to get the characters who failed to survive. Anything but 'you fall into the magma'. But ... if the magma isn't to be a threat, why the hell is it there?

"If you don't want them to die, don't put it there." I've been told.
No. I don't want the characters to die. I'm cheering them on.
But I'm the game master. I play the world. The world is not cheering them on. It isn't for or against them. There are hazards. These hazards exist. These hazards can kill. That's why they're called hazards.

Do I want the PC to fail and fall into the magma? Of course not. But there's still going to be the risk of it happening.  Or the walls closing in and crushing someone who triggered a trap. Or a deadly poison. Or a disease. Or rolling boulders in an avalanche, or exposure when the PCs are out in the winter unprotected.

A goblin gets a nat 20 hitting the 1st level Cleric, the Cleric dies. The PCs don't have the resources for a resurrection. The PCs are down a combatant, and the goblins are still coming.

TPK is on the table. Do I want a TPK?  No.  Do the players?  No.  That doesn't mean it isn't there.

It means the PCs need to be careful. If a group of people are wandering in the middle of the jungles of the Amazon with bow and spear and sword, you'd damn well bet they'd be bringing along some extra precautions.

The PCs should about that paranoid, if not more so - they know there's monsters out there, they should be ready to face those monsters, or run if they need to. The monsters are a Very Real Threat, the characters are well aware of this - but they chose this line of work, anyway.

If you're playing Vampire, you don't thrown down with the Gangrel and not expect him to shred you open with Aggravated Damage, or the Lupine to not tear you in two. If you're facing a dragon, you shouldn't expect to survive. You fight and hope you survive, but it should never be a sure thing (even if only the GM knows it's a sure thing). You don't go on a shadowrun in a high-end top-secret corporate facility and not expect the guards to fill you full of holes if you're not careful.

That's how things work. Sometimes... they don't. And there's no 'failing forward', because that cheapens the entire experience.  Some things should be 100% open for failure - absolute failure. The lock doesn't get picked - maybe if you take a bit more time, maybe you need to find a key, maybe the lockpick breaks. That door isn't opened.  There's no 'the door opens, but...'

It's closed.
Find another way.