by Mike Shea on 18 April 2016
Note: this article has been updated since the original published in April 2013
One of the things that sets Dungeon World apart from d20-based RPGs is the inclusion of a core mechanic with partial successes. Instead of a mechanic with only either success or failure, such as rolling 1d20 + skill modifier and matching it against a difficulty check number, Dungeon World gives a wide middle range of partial successes. In fact, given the 2d6 curve of Dungeon World, the odds are higher that you'll have a partial success than a full success or full failure.
There's no reason we can't steal this idea and add it into our d20-based RPGs like the fifth edition of Dungeons & Dragons or Pathfinder. Though these games have a focus on "roll a d20, add a modifier, check against a DC score", we can change this slightly to "roll a d20, add a modifier, and see what the results are". We don't actually NEED a DC score for many things.
Typically when we run our Dungeons & Dragons game, we choose a difficulty for any particular action. In D&D 5e, this ladder is between 5 (very easy or "why the hell are we rolling for this?") to 30 (very hard or "I'm pissed at you and really don't want you to succeed."). Generally speaking, we're picking numbers between 10 and 20 for this DC. When players roll, add their skill or modifier, and then tell us the result, we check to see if it succeeded or failed.
It doesn't have to be this way. Instead of picking a single DC, we might pick a range of DCs to represent shades of gray. For example, Say a PC is drawing negative energy out of a powerful magical pillar tied to the Shadowfell. We might set the lowest DC for this to 13 and the highest DC at 18. Instead of succeeding or failing, we can describe different results depending on which DC they successfully met. Rolling below a 13 for example, causes tremendous negative energy feedback resulting in necrotic and lightning damage to the PC attempting the check. With a result between 13 and 17, the wizard has successfully deactivated the pillar. On a result of 18 or above, the surge of energy flows into the wizard and gives the wizard advantage on its next attack roll or arcane skill check.
We can set a number of skill DCs for a given situation. Very low results usually do something bad to the PC. Mid-range results are successes but with complications. High results succeed. Very high results give some major advantage to the PC.
What if the PC blows way past whatever DC you had in place? Reward them for it. Show them how much better they are than this pitiful challenge. For story-based skill checks its easy to describe how much better they did.
Consider the scene in Romancing the Stone when our heroes approach the local drug kingpin in his lair. After a tense discussion involving a pistol, the villain looks out the little slot in his door and says "You're Joan Wilder? The novelist??". That's a critical diplomacy check right there. In our game, maybe that cynical guard is actually a huge fan of the PC asking the question. Maybe the PC had saved the villain's cousin six years back and it ends up saving the lives of the PCs. Overachieving should never get a PC in trouble. It should offer something even greater than what they had set out to do in the first place.
Many GMs already do something like this. They don't have a fancy chart giving them a range of successes or failures, but when they hear about a particularly high or low roll, they describe a greater success or terrible failure. Sometimes, when a player sees how low he or she rolled, he or she might describe the critical failure without the GM's help.
In reality, we probably don't need a DC at all. We ask for a check, the player rolls the check, we look at the result and we describe the outcome. No DC needed.
This turns skill checks into an analog gauge rather than a binary "success" or "failure". There's not a lot of preparation required for this sort of change. All that's required is trust between a player and a GM that the GM isn't trying to screw them and is actively using this power to make the game more fun.
We might keep some mechanical tools at our disposal for particularly fantastic rolls. Maybe the character gains advantage on its next attack or skill check. Maybe damage is boosted. Maybe monsters become vulnerable to particular types of attacks. Maybe the skeptical castle guard just became the PC's new best friend. There are lots of ways for us to reward high successes and we should keep them in mind.
A lot of game masters and newer RPG systems have discussed the concept of "failing forward". Instead of a failed skill check resulting in a monologue of disaster from the GM, a failure can instead put PCs in a tough spot. Maybe they fall prone. Maybe they lose a valuable resource. Maybe they are forced into a more desperate situation. Dugneon World refers to these as "hard moves" that we make to put PCs in a tough spot.
Failing forward means putting PCs on a new path, a hard path, that still propels the story forward. It's not easy to do. Saying "no, you fail" is a lot easier than changing the course of a scene. This is why a lot of us haven't done this for most of our GM lives.
Adding levels of success puts a lot of pressure on GMs. Instead of two potential outcomes for a skill check we could have twenty. A good GM can come up with dozens of variants to a situation depending on this role and this, of course, is the advantage of having a human being running these games. All of this depends heavily on good improvisation skills, a key skill of the Lazy Dungeon Master.
Describing partial successes is equally hard. With a partial success, you give something and you take something away (what Dungeon World refers to as a "soft move"). PCs may get what they want, but not through the obvious method or in the way they want it. Complications occur. Maybe they got attention when trying to pick that lock. Maybe they absorbed the magic of the rune and have to pour it out before it devours them. Something good happens and something bad happens. You have to decide what they are. Like many storytelling areas of D&D, this could potentially be a bit of negotiation between you and the player. Stretch your improv muscles with a bit of "yes and" back-and-forth. 13th Age designer, Rob Heinsoo, is oft described as running "lets make a deal" sort of negotiations in these sorts of situations.
As human beings, we don't have to live with zeros and ones, "yes" or "no", or success and failure. We have huge ranges we could work within as players roll skill checks. There are as many possible outcomes for a skill roll as there are numbers on the die and every one of those outcomes could fork our story in a new direction. Step back from the DC ladder and enjoy the shades of gray.