Using Failure as a Narrative Tool instead of a Player Frustration

How often have scenarios like these come up?

Scenario One:

Needing to kill the Goblin Chieftain, but not wanting to fight the entire goblin war camp at once, the party decides to sneak past slumbering goblins to the chieftain’s tent.  The stealthy ranger, sneaky rogue and unarmored wizard all roll well, but the scale mail wear dwarven fighter rolls poorly, eliciting a groan from the other players as now the goblin sentries raise the alarm and the party is stuck fighting the entire war-camp.

Scenario Two:

The party has just defeated a beholder and its cultist allies in a hard-fought battle.  Now catching their breath, they’re able to search the room.  Before the DM can speak, the player of the wizard says “My wizard will look at those arcane carvings in the wall more closely now…shit… I rolled a one; guess my wizard doesn’t know anything, dur dur dur.”

These scenarios and countless others like them, from the mundane, a PC missing on a melee attack to the esoteric knowledge checks.  Failure is an inevitable occurrence in any RPG, but in many D&D games I’ve played in, and yes, many D&D games I’ve run, failure just becomes boring and when it occurs repeatedly in a session due to a player having a run of bad dice luck it can extremely frustrating for a player.

Most versions of the D&D rules set up these checks as binary PASS/FAIL, maybe with a chance of automatic or extraordinary success (natural 20) and automatic or extraordinary failure (natural 1).  This works okay mechanically but in practice it can be boring, and in many cases, frustrating to players who feel like not only is their character not getting to do anything, their character falls out of the narrative of the game largely, as they repeatedly swing and miss during combat or fail perception check after perception check.  It’s a shame because many other games handle failure much better, implementing, explicitly or not, more of a “Fail Forward” approach where the player’s failures at dice rolls (or whatever mechanical randomizer is used) as a chance to continue the narrative as opposed to the way failure often functions in D&D as a narrative top (think of the dull silence that lays in wait between a player sullenly says “I got a 5, I don’t see anything” or “I rolled a 7 on my attack” and the DM picking up the narration).  In particular I’m going to draw on Dungeon World for examples of strategies to draw upon, mostly because I’ve been reading it heavily lately and also because I recently ran Apocalypse World which was the source for the Dungeon World Mechanics, but there are many other games out there that have good “Fail Forward” mechanics that you can draw on.

Now, not every roll needs to be narratively interesting, it’s okay to have the fighter swing at an orc with her axe and miss but it’s much more interesting to have some of the PC failures drive the narrative of the game forward instead of having them feel like a pause for a player, a moment where the player, due to bad dice luck, just doesn’t get to do anything.  Think of an old Final Fantasy style game from the era of NES and SNES, if you played any from that era.  Remember the frustration that would build when you had a streak of bad luck or your party accidentally ventured into an area your characters weren’t high enough level for yet?  The annoyance of watching the clunky animation of your character making an attack and then the pause before the disheartening “MISS” showed on screen and nothing happening and then having to wait for the round to cycle?  That is unfortunately what D&D emulates all too often, so below are some strategies I plan to implement in my games to turn failures from frustrating pauses in the fun for players into narrative tools to drive my games forward.

 Let the Narrative Drive Dice Rolling

• Apocalypse World and Dungeon World have a Principle “Begin and End with the Fiction” and part of that is only making a move when the fiction demands it, which if you just change ‘Move’ to ‘Roll’ can easily be applied to D&D.  If it’s simply and/or doesn't matter to the narrative, there’s no need to roll.  If a character is walking on level ground, not affected by any injury or mind-affecting state, not being chased by any monster or threatened by an oncoming wave of lava, there’s no need to roll, there’s an infinitesimally small chance of failure and any failure that would occur is narratively irrelevant and uninteresting.  So the PC trips on their own feet?  They’d just get up and continue walking.  Rolling for each action becomes prosaic and boring, better to save dice rolls for when actions are consequential, when their outcome is exciting and matters to the fiction of the game.
• Now, that walking scenario is extremely unlikely to come up during a game, so it gets a little trickier for situations that arise during game, but again, take a second as the GM and consider “Is this action consequential to the fiction, would failure at this action be interesting?”  So, climbing a wooden ladder in a dungeon after defeating the orcs in the room?  Probably doesn't need a roll, the danger of the orcs is dealt with, if the ladder is reasonably sturdy and a fall from it wouldn't kill or seriously injure a character, then really, there’s no need to roll.  If, on the other hand, the PCs are fleeing from a swarm of devil frogs and need to climb up the ladder to avoid suffering from a fiery tongue lashing, then it makes sense that you might call for an Athletics check to see how quickly a PC could climb up the ladder.
• The trickier facet of letting the Narrative Drive the Dice Rolling is not letting the players hijack the game with compulsive, impulsive or reflexive dice rolling that aren’t driven by the narrative.  This can be difficult because rolling dice can be a big reason why players enjoy playing D&D, but often times the compulsive, impulsive and reflexive rolls don’t add to the game’s narrative, or worse, distract from it.  
• A very common example of this kind of rolling is the Perception check.  Players often compulsive will say something like “I’ll perceive” or “I’ll make a perception check” when entering a new room or area and sometimes may even roll dice on their own but what is the point, the players are basically saying “my character is using their senses” which I think we can all pretty much agree should be a given.  Another common compulsive dice roll I see often is the Search for Loot roll, often declared as soon as the DM declares that the last adversary is dead or unconscious players will declare “I search” because they resort to the Thieving Murder Hobo mentality of ‘the point of the combat was loot, and if I find it first maybe I can call dibs on something cool’.  
• The problem with the above two examples is not just that they don’t follow from the narrative, they’re purely mechanical declarations instead of being a result of the description of what a character doing (i.e., what is the character trying to perceive: are they watching the weather for a storm, watching for ambushes, looking for traps.  Are they searching the room for secret doors, searching the cult leader for treasure, etc?) but also because they encourage “Me Too-ism” where as soon as one player declares they are making a roll, much of the rest of the party feels like they too should roll so that they are not left out and/or to make sure that the party gets at least one success.
• To combat “Me Too-ism” I've already established in my games a house-rule that I can, at any time, declare a limit on the number of characters who can roll for any one check, using the narrative excuse of too many cooks in the kitchen spoiling the food.  Or to put it more in game terms, if you have more than a couple of characters trying to listen intently at a door, you can imagine the scuffling and whispers of “Be quiet” or “Did you hear that” obscuring any sounds that the characters may have been able to hear on the other side of the door.
• And to combat the compulsive, impulsive and reflexive dice rolling, I plan on imposing a house rule that aside from attack rolls during combat, players can’t roll until I direct them to.  The idea being that instead of saying “I perceive” and then rolling a wisdom check on their own, they can say “I perceive” and I can ask what specifically their character is attempting to perceive and then decide what check is needed.  That will also help me leave aside the binary pass/fail and instead implement a continuum of success and failure as discussed below.

Ditching the Binary PASS/FAIL dichotomy for a Continuum of Success and Failure

• Not every check should be PASS/FAIL, but players often assume they are.  For instance, looking back at the example from the start of the post where the Wizard took a look at the arcane carvings, there’s no reason why the wizard, who spends a great deal of time considering and studying arcane matters, just as a matter of their class, shouldn't be able to tell at least basic information about arcane carvings.
• This is especially true about non-physical checks, for knowledge or perception type checks were it doesn't make sense that a character wouldn't just not know anything or where a character wouldn't at least be able to gain a basic understanding of the tableau before her eyes.
• Again, take the example of the Wizard studying the arcane runes after the party defeats the beholder.  Even if the player rolls poorly, her wizard should be able to determine some basic information about the runes, knowing that they aren’t related to the beholder is a big piece of information that seems intuitive that a wizard would be able to determine, even if she couldn't puzzle out more.  Then if the player had rolled well you can reward her with additional information, like that the runes tell the story of a group of good wizards crafting a weapon to battle beholders.  
• To take another example, imagine a dwarven fighter looking at the construction of a room of an underground dungeon to try and determine when it was constructed and by whom.  Even if the player rolls poorly, his dwarf, being wise in the ways of stonecunning, should still know basic facts like: it’s not natural construction, it wasn't constructed by dwarves, it was constructed over a century ago.  Then if the player had rolled well, you could have rewarded him by letting his dwarven fighter detect the faintest of breezes from a hidden door or revealed that he suspected the room was carved out of the stone by magic.
• Let’s be honest, you shouldn't just punish PCs for failure on poor dice rolls, the failure was already punishment enough.  They should still get a baseline of information for attempting the knowledge check or perception roll and then you can reward success if they are lucky enough to roll well.  Which isn't to say that you can’t still use Botches or have similar bad luck befall a character when a player fails a roll, it just shouldn't be arbitrary punishment but instead should be narratively interesting, as discussed in the next section below.

Making Failures Interesting

• I think one of the enduring reasons why so many groups play with a Critical Failure or Botch house rule is that house rules of that sort at least make failure interesting.  The tension always ratchets up when a natural 1 is rolled and a critical failure occurs.  
• But things don’t have to be catastrophic to make failure interesting, you can watch out for opportunities, apart from natural 1’s, where instead of just having the roll fail, you, as a DM, take advantage and add something to the narrative.  This is where the GM Moves from Dungeon World (which can be viewed at: http://www.dungeonworldsrd.com/gamemastering#TOC-Moves) are useful, not as mechanics to try and port straight into your D&D game, but as concepts to apply and enrich your Dungeon Mastering.
• Here is a list of the GM moves from Dungeon World, I've highlighted the four that I think are easily adapted to D&D and eminently interesting when done well.
• Use a monster, danger, or location move
• Reveal an unwelcome truth
• Show signs of an approaching threat
• Deal damage
• Use up their resources
• Turn their move back on them
• Separate them
• Give an opportunity that fits a class’ abilities
• Show a downside to their class, race, or equipment
• Offer an opportunity, with or without cost
• Put someone in a spot
• Tell them the requirements or consequences and ask
• Reveal an unwelcome truth/Show Signs of an approaching threat
• While these are two separate moves in Dungeon World, I think they get at the same consequences when applied to D&D.  What you’re doing as the DM when you use either of these moves is showing that the world is dangerous.  
• Perhaps when the Paladin misses the skeleton with the swing of her war hammer, she hits the wall behind the it and causes a crack to run up to the ceiling, foreshadowing a possible collapse of the ceiling in that corner of the room.
• Or when the wizard misses the ogre with his burning ray, the illumination it provides lets the party see the horde of goblins in the dark forest who are coming to the aid of the ogre champion.
• Give an opportunity that fits a class’ abilities
• Cast your mind all the way back to the first example I gave at the beginning of this post, the scenario with the party sneaking through a goblin camp and the dwarven fighter almost predictably failing his stealth check.  Now, it’s easy to have the failure of the one PC cause the entire party to fail, but what if the failure of the one PC gave the other characters a chance to shine?  
• What if, instead of the whole camp waking up, perhaps the Ranger, Rogue or Wizard noticed a goblin guard, or two, coming to investigate the noise of the Fighter and realized that they only had a round or two to silence the guards to salvage the stealth mission?  Not only does this still allow the characters to possibly salvage their plan to sneak to the chieftain’s tent instead of having a single bad roll spoil it, but it gives the rogue and ranger a chance to show off their stealthy high damage style or for the wizard to show the utility of their spells.  And who knows, maybe even the not-so-stealthy fighter would get a chance to knock a guard unconscious before the hue and cry is raised. 
• Tell them the requirements or consequences and ask
• This is a great concept to use for more physical actions, jumping a chasm or swimming back to the side of the boat.  Take an example where a Rogue wants to use a rope to swing across a river canyon to the cliff on the other side where he can hide and snipe at the Stone Giant further down the canyon, the player makes the roll and it’s 5 short of the difficulty you assigned to the action.  Sure you could just say he failed, but isn’t it more interesting to present a choice: (A) stay hidden on the side the rogue is currently on, where he can’t see the Hydra to shoot at it or (B) Swing to the other side successfully, but be stuck out in the open and unable to hide for a turn, exposed, potentially to the Stone Giant’s attacks.  Then the player can decide how bold he wants his character to be.

Since I’ve been thinking about failure in D&D, reading Dungeon World and running Apocalypse World earlier this spring, I’ve been haphazardly implementing some of the strategies above and have been very pleased with the results.  By taking my D&D games away from always having strictly binary PASS/FAIL outcomes and by taking advantage of the narrative possibilities that failures can offer I’ve had more fun running and it seems, to me at least, like my players are having more fun too.