Failing Forward in D&D (and Some Alternatives!)

Much proverbial ink has been spilt on the subject of running skills checks in D&D, including on this very blog (…discussing adjusting the default Difficulty Classes, using passive skills as a solution for swingy dice, and an early suggestion that certain proficiencies should be considered ‘technical’ proficiencies).

This could be a costly failed Stealth check….

Like most DMs, my methodology for adjudicating such checks has evolved over the years, and by now I can lean on various different techniques to help me smoothly deal with different game situations. In my recent post about running heists, I actually went into a fair bit of detail about some of these methods, in particular on how to deal with failed checks. By now, we’ve probably all heard of the concept of ‘Failing Forward’, but I have a personal preference for a technique I call ‘Failing Sideways’ which I introduced in the heist post, along with some other ways of dealing with failed checks and avoiding scenarios that set up the players for failure.

Anyhow, since my advice on the topic got buried pretty deep in that very long post about running capers, I felt it was worth copy and pasting that advice here, as its own separate post…

Note: While I was tempted to edit this advice for more general scenarios, in the end I decided to CTRL C + V exactly as it appears on the heist post. But if you substitute the word ‘heist’ for ‘adventure’ or ‘mission’, I think you will find it useful counsel for running any D&D scenario in which players are making multiple important skills checks in order to succeed in their mission.

+++++++++++++++++++++++++++++++

Dealing With Failed Checks

Perhaps the one thing that could ruin your running of a D&D heist is using an unforgiving interpretation of the rules for skills checks. In any given heist, the adventurers are going to make a LOT of skills checks, and what we really want to avoid is for the game to implode on a single bad roll – whether that being the first PC to attempt to sneak past the guards, or the only rogue in the party fumbling the all important thieves tools check to pick the vault’s lock.

Bad rolls are going to happen. Most likely they are going to happen often. So if you build a scenario where the party have to individually succeed on even three or four crucial rolls to pull off the heist, then they will fail.

On the other hand, if you repeatedly allow players who fail at skills check to simply try again, the game world loses any sense of reality and consequences. The players will feel they are playing the adventure on cheat mode. And, even if they complete the heist, they won’t get any great sense of satisfaction from having done so.

There are actually a few tricks and techniques we can employ here to prevent failed skill checks derailing the adventure.

1. Failing Forward

Failing forward is a much-fêted technique that every good Dungeon Master should have in their refereeing repertoire. As the name suggests, the players may fail a roll, but the adventure still moves forward. It works by allowing the players to succeed on a skills check they failed, but at a cost, or with an ensuing complication. When the rogue rolls a 2 on the dice, it doesn’t mean the lock is beyond their abilities to pick, it means their lock pick broke in the process, or they injured their hand, or they made a loud noise in doing so. In other words they failed to do a good job, rather than failed outright, and the story continues…

2. Failing Sideways

Ok, I made this term up, but another way of dealing with a failed skills check is to enforce the failure – the door is still locked – but to give the player another go at the challenge. But you just said failure should have consequences? And it should. To acknowledge the failed roll, you can now raise the DC of any subsequent checks, or stress that each failed attempt costs a significant amount of time (obviously this only really works if they are labouring against a ticking clock!). You can also enforce the same type of complications I mentioned in Failing Forward, but this time they haven’t achieved their objective and still have to come up with a good roll to progress. I like Failing Sideways, because it’s a bit grittier than Failing Forwards, which is cool and cinematic but feels a little bit too generous at times.

3. Accumulated Failures Needed to Fail

One of the reasons Blades in the Dark is hailed for its heist-enabled gameplay is its progress clocks mechanic. Progress clocks measure the success of a task over a number of rolls, with final success dictated by achieving X amount of successes before Y amount of failures. This helps mitigate the role of luck, tipping the probability balance of success firmly in the players favour (assuming they are any good at the skills they are attempting!).

This is a bit similar to how group checks officially work in D&D, in that a group check is considered to have succeeded as long as more players succeed than fail (I actually have a house rule for group skills checks that I prefer, if you wish to check it out). But progress clocks offer more flexibility than standard 5e group checks (you can make tasks easier or harder by changing the number of successes needed vs. failures), and they can also be applied to solo endeavours, or applied to a complex task. A sophisticated lock, for example, might require 5 successes before 3 failures to pick, which could also be spun out into a dramatic mini-scene.

An advantage of this system is that players can also see when they are doing badly, and perhaps throw some extra resources (different forms of inspiration, guidance etc.) at the problem, or change their approach to the situation.

Use the Suspicion Mechanic. If you bought Keys from the Golden Vault, turn to page 60 to read about the Suspicion Mechanic introduced in the Prisoner 13 adventure. Essentially, the more suspicious shit that goes down, the more alert the guards become and the harder they are to sneak by or deceive. This gives DMs a nice consequence they can apply to failed Stealth checks etc., without having to immediately bring the house down on them on a single botched roll.

[For non-heist scenarios you could create a new mechanic relevant to your scenario. In essence, the suspicion mechanic combines a complication (DCs going up), with a progress clock, something which you can easily replicate in other scenarios].

4. Stacking the Odds

I’ve blogged before that DC 15 shouldn’t be 5th edition’s go-to DC. It seems fine when it’s a Perception check thrown at the whole party and only one PC has to pass, but as soon as players have to individually succeed on multiple tasks heroes quickly become clowns at this difficulty class. The chance of any given character passing three different DC 15 skills checks is usually much less than 50%, and will often be less than 10%. My advice, keep the DCs as low as possible and award advantage as readily as you can, without appearing to go easy on the party. “The wall is DC 10 to climb, but with your grappling hooks and rope you get advantage.”

And while I generally hate rerolls (they undermine the whole point of rolling dice in the first place), I can stomach the characters starting any given heist with 5e’s single point of inspiration. I do also approve of story-earned re-rolls. Maybe have the party’s patron suggest they leave an offering at the Temple of the God of Trickery before attempting their endeavour. If they succeed on a Charisma-based Religion check, while doing so, they get one or more ‘luck‘ points which they can use during the heist.

5. Success Can Go A Long Way

If a character succeeds on a Stealth check, in particular, let them carry that result forward as far as possible, before calling for another one. If players need to succeed on a new Stealth check every single time they pass a guard, it’s inevitable they are going to raise the alarm sooner or later and you’ve essentially given them a 100% chance of failure. If, however, they roll a 15 to get past the first guard, let them keep going further into the fortress until you feel a higher roll would be required to progress, and only then say… “you peep into the courtyard and see four guards practicing their swordplay. If you want to sneak through here, you’re going to need to pass on a new DC 20 check.”

Planning For Failure

Some DMs feel that for D&D to be fun the players have to succeed, and what is decided at the table is not whether they succeed, but how they do so. I am not one of those DMs.

For success to be meaningful, failure has to be possible. And one of the great things about heists is that players can fail and the session can continue. Guards rarely summarily execute thieves on the spot, so if the players get caught during the heist simply have a plan for what the guards do with them. Maybe they bring them before their boss for questioning right away, but most likely they will lock them in some kind of cell (which, for gameplay purposes, needn’t be particularly secure). Depending on the scenario, the thief might get away with being escorted off the premises, perhaps with a mild beating and/or confiscation of some or all of their equipment.

One scenario that you almost want to play out, is just ONE of the players being caught. Now you have a cool complication, but the heist is very much still on, as the party have to raid the vault and rescue their bumbling colleague.

+++++++++++++++++++++++++++++++

Ok, so that concludes my current suite of tools for dealing with failure. Another technique, favoured by many DMs, is not to call for a skills check at all, if you feel that the players are attempting something that falls well within their capabilities (the pros of cons of this technique are discussed in the comments section of my post on creating drama in D&D).

I will finish by saying that bad dice rolls – and the resulting narratives they inspire – often create the most memorable moments on the table. So we’re not trying to get rid of them! But once we learn how to transform these bad rolls into setbacks and complications, instead of story-crushing dead ends, we evolve into better Game Masters.

Sure, sometimes the party will have to slide back a few squares on the proverbial snakes and ladders board in order to try a new approach, and other times they will fail outright in their mission, but the more tools and techniques we have for adjudicating the game the less dependent we are on unmitigated blind luck for determining the party’s success… and the more we can allow player decisions to determine the outcome of their adventures.