On Harm, Strain, and Endurance
Consider a situation that will be familiar to anyone who has played long enough. A high-level fighter, faced with a dangerous opponent, jumps from a curtain wall onto the rocks below. The fall damage is rolled. It is significant — the kind of number that would kill a lesser character. The fighter absorbs it, fussily brushes some dust off his pauldron, and walks away, because their hit point total is large enough to treat a thirty-foot drop onto stone as a temporary inconvenience.
These moments are funny when they happen. Worth remembering the humor, because it points directly at the problem.
That moment stayed with us. And while we are not inclined to rename hit points for the sake of it — the term is familiar and changing it serves nobody — we do want to reframe what they represent in the player’s mind. Hit points measure combat endurance. That is what they do well. That is what they were built for. Held in that frame, most of what follows makes more sense. The more we thought about it, the more we recognized that this was not an isolated edge case — it was symptomatic of something more fundamental. A lot of types of harm and damage are poorly modeled if you get hit points involved at all. That is not just wrong — it is a category error. The abstraction was never built for most of what the world can do to a person, and applying it there produces results that are, at best, incoherent and, at worst, actively misleading about what survival in a dangerous world actually involves.
Hit points represent combat survivability — the accumulated capacity to avoid decisive harm in a fight through awareness, positioning, reflexes, and hard experience. They are very good at what they were designed for. They were not designed for a man falling onto rocks. A high hit point total means you are very hard to kill in a fight. It does not mean physics has stopped working.
Once you see this clearly it is difficult to unsee. The curtain wall was the moment we stopped trying to make the abstraction do work it was never designed for, and started asking what a better model would look like. Oath-Bound separates these two things. Combat damage and non-combat harm are not the same system, and the difference between them matters.
What Non-Combat Harm Actually Is
Non-combat harm in Oath-Bound is not a parallel damage system with its own tables and hit point equivalents. It is a qualitative model — the GM determines what happens based on the nature of the harm, the circumstances, and what a person in that situation would actually experience.
A fall from height onto a hard surface has consequences that follow from the fall — not from the character’s level, not from their hit point total, not from their class. A character who is on fire is in a situation that must be resolved. A character who has been traveling hard for three days in cold rain without adequate shelter is accumulating a set of conditions that will eventually demand attention. None of these are combat exchanges. The combat abstraction does not apply to them.
This covers a wide range of circumstances: falls, fire, drowning, exposure, illness, hunger, exhaustion, sustained exertion under load, poison, and any other situation in which the world puts pressure on a character’s body outside of a fight. The PHB handles some of these with tables. Oath-Bound handles them with judgment, because a table produces a result and judgment produces a consequence, and consequences are what the model is built around.
How Harm Accumulates
Non-combat harm presents in different ways depending on its source, and the distinction matters.
Some harm is immediate and potentially decisive. A fall onto stone, a crushing impact, impalement on a fixed object — these are not moderated by the character’s combat capability. The GM determines the outcome based on the circumstances: the height, the surface, the character’s condition, and what a person in that situation would realistically survive. In some cases the outcome is survivable. In others it is not. The character’s hit point total is not the deciding factor.
Some harm is sustained. Fire, drowning, and asphyxiation are ongoing conditions that continue to produce harm until the situation changes. These are not resolved as repeated attacks. They are circumstances that must be escaped or ended. The longer a character remains in them, the worse the outcome becomes.
Some harm is gradual. Fatigue, illness, exposure, and the weight of sustained burden accumulate over time. None of these is dramatic in the moment. Collectively, they reduce the margin available for everything else — including combat, when combat cannot be avoided. A character who arrives at a fight already exhausted, slightly ill, and carrying more than they should is not in the same position as one who arrives fresh. The hit point totals might be identical. The situations are not.
Conditions, Not Repeated Tests
When a non-combat condition is established — illness, significant fatigue, injury from a fall — it is established as a known state and treated as part of the character’s current reality until it resolves. It is not re-tested every hour with a new roll. It shapes what the character can do, how quickly they recover, and what they carry forward into subsequent situations.
This is deliberate, and the reasoning is partly practical. The vanilla approach to non-combat harm — rolling damage, tracking a separate metric, applying table results — created a bookkeeping burden that most GMs quietly abandoned. Not because they were lazy, but because a system that requires tracking just one more number tends to get hand-waved when the session has momentum and the harm probably won’t determine the story’s outcome anyway. The result was that non-combat harm had no consequence at all, which is worse than modeling it badly.
A condition held in the GM’s judgment rather than tracked on a sheet does not disappear. It persists as a known state that shapes what the character can do, how quickly they recover, and what they carry forward. And because it lives in the fiction rather than in a number, it can be leveraged — for texture, for dramatic timing, for the moment when a character who has been pushing through something finally cannot. A metric you track is data. A condition the GM holds is a story modifier. The purpose is not to create a second damage track. It is to make consequence real without making it administrative.
Recovery — Three Gradients
Most harm heals without divine intervention, and it heals in a pattern that anyone who has been genuinely hurt will recognize.
The first phase is slow. In the days immediately following a significant injury, improvement is limited and the body’s reserves are occupied with basic repair. This is not the time to push.
The second phase accelerates. Once the acute phase has passed, recovery picks up — the character noticeably improves from day to day, capability returns, and the condition is visibly resolving. This is the longest part of the curve and the most encouraging.
The third phase slows again. The final resolution — the last stiffness, the injury that is mostly healed but not quite, the stamina that takes a little longer to fully return — drags out beyond what most people expect. The character is functional well before they are fully recovered.
Three gradients. The CRM provides practical reference points for each phase that require no arithmetic more demanding than counting days. The model is realistic without being a calculation.
Divine intervention — healing miracles — compresses the curve rather than replacing it. The gradients still apply; they are simply shorter. The GM determines by how much, based on the nature and band level of the working.
What This Means in Play
The practical implication is that characters who treat the world as a consequence-free environment between fights will find that it is not.
Jumping off a wall to escape a fight has consequences. Traveling hard without rest has consequences. Pushing through illness because the mission requires it has consequences. None of these are necessarily fatal in themselves — but they accumulate, they persist, and they arrive at the worst possible moment. The character who walked away from the curtain wall in the opening example would, in Oath-Bound, be dealing with the effects of that decision for some time afterward. The immediate fight may have been escaped. The fall was not free.
Exceptional actors in Oath-Bound understand this. They manage their condition as a resource, not just their hit points. The combat abstraction handles the fight. The rest of the world handles itself.