Many popular RPG systems measure success (or failure) as a simple binary. For example, by a strict reading of D&D 5e’s rules, either your character is successfully sneaky or they’re not: there’s no middle ground. There’s no benefit for being exceptionally stealthy, and there’s no real penalty for being exceptionally not-stealthy. Thus, there’re no degrees of success or failure. Every test is pass or fail.

This streamlines resolution of tests, and has the benefit of being fast and simple. But it also misses out on some opportunities.

Some systems require a basic target for success and offer optional benefits for exceeding that target. The new Wrath & Glory Warhammer 40k RPG uses this system, where you must achieve a target number of points in order pass challenges, and can spend points in excess of your target number to further improve your results. This gives you degrees of success, but has no degrees of failure. If you do really well, more cool things happen. If you do well but not quite well enough, that’s the same as if you’d failed completely.

While this mechanic rewards players for lucky rolls or for a character’s particular expertise, it doesn’t move the game or story forward very much when a test is failed. Furthermore, it can be frustrating to fall barely short of a target and gain no recognition of how close a character was to success.

Other games use more differentiated resolution systems in which success and failure may be intermixed. This is extremely evident in Apocalypse World, and in other games inspired by it. The relevant mechanical concept from AW is that any test should result in one of three options: unmitigated success, success with complications, or failure.

This, unlike the binary success state of 5e, uses a trinary success state. Luck and character expertise are still rewarded with the possibility of unmitigated success, and there is space between absolute success and absolute failure for PCs to muddle through in the middle. This does sometimes result in slower resolution, as the details of what “success with complications” looks like are hammered out.

I’m not even going to try talking about Fantasy Flight’s Star Wars RPGs right now, though they deserve a mention in this conversation and I may return to them in the future.

It’s worth noting that “success with complications” also often means that progress has been made while still introducing reversible failure. That failure represents a threat or risk to the player characters, but is something they may deal with or attempt to forestall. This introduction of danger which can be ameliorated through quick and skillful action tends to build tension for the players. I usually appreciate that added tension, as it adds a certain zest to most games. As with all sources of tension, however, it is sometimes important for the storyteller to downplay it (or play it up) to better match the game group’s desired experience of play—or to match where the storyteller wants the players’ experience of tension to be at any given moment.

So what if you want to experiment with trinary success states without abandoning your game in a binary success state system?

Justin Alexander has written about system paradigms (A Paradigm for Stealth) which replace the binary success state with a trinary one. While Alexander’s post above focused on stealth-states, with a little tinkering it’s easy enough to add this approach to strictly binary success systems like 5e. If it seems reasonable to say that a given test in 5e wouldn’t be strictly pass or fail (probably best applied to skill and attribute tests rather than things like combat), try the following: if the test result is within +-2 of the test’s target number, give the PC success with complications.

If a PC was trying to sneak, that “success with complications” might be that guards know there’s someone nearby, but don’t know the PC’s precise location (e.g. the “Detected” state from the post above). If a PC was trying to convince a castle’s guards to let them in, “success with complications” could mean that the guards let the PC in but are suspicious—perhaps they insist on a search of the PC and their belongings. The exact details will vary by situation. Furthermore, perhaps “target number +-2” is too narrow (or too broad) a range. Adjust it to fit the feel of the situation.

Remember, RPG systems are ours to tinker with as we choose! Let me know if any of this sparked ideas for you, or if you’re already doing something similar.