D&D 5E (2024) Healing Isn’t What You Think: The Value Curve Across a Fight

[Cap'n Kobold]

I think this is a double sided issue, either monsters aren’t dealing enough damage that just killing them is more effective than healing your teammates or they are dealing enough damage and you get complaints that nobody wants to have to be forced into being the party’s medic.

. . . and if the monsters are dealing enough damage, then killing them is more effective than healing your teammate, because you can't outheal them and so need to remove the source of damage in the first place anyway.

 

[Zardnaar]

. . . and if the monsters are dealing enough damage, then killing them is more effective than healing your teammate, because you can't outheal them and so need to remove the source of damage in the first place anyway.

Theyre hard to kill in a hurry unless theyre solos or close to it.

 

[ECMO3]

I think it depends on the characters. Even when I play a Cleric for example I would agree it is very rare that healing, especially early, is worth the action cost for my PC. But my Clerics are not "healers" per se even though they can heal. Other PCs are built around healing specifically and often with weak offensive action options. They are not going to give up as much, especially if their healing can keep a high DPR martial standing for longer.

There is also the action cost. Potions of healing, as well as some abilities like the 6th level Celestial Warlock ability and Paladin Lay on Hands are a bonus action. Those have less opportuninty cost to using them in combat, sometimes no opportunity cost.

 

[Silam]

BTW, how do you folks cost the act of feeding a potion to a fallen teammate?

5.5e introduced Diablo-style sword-in-one-hand-potion-in-the-other-binge-drinking via BA, but it seems unclear if that is only for the self or for others too. It seems logical that drinking one’s own potion requires less opportunity cost than feeding it to someone else who’s prone and unconscious, but as far as I can tell it’s not spelled out in the rules.

If a table plays with an asymmetry in the cost of gulping a potion for one’s self versus someone else, that adds an interesting dimension to the tradeoff.

 

[Blue]

Just trying to talk something out.

Say there's only one foe left. Is it better to heal, or potentially kill them. Killing them removes the cost (the action) from healing.

But while I was wondering that, I applied damage logic to the base premise. Attacks early in the combat that kill an opponent only reduce incoming expected damage by a smaller fraction than when there's only a single foe left in combat. So one early attacks are worth a lot less... except of course they aren't, because that's what's getting us to one foe left.

Which made me question the whole premise. Regardless of what turn it is in the example, healing is trading an action (or bonus action for some healing) and a spell slot for an action of the foe. It's always off by the exact same amount - the spell slot - regardless if otherwise it's a push (action for action) or a win (bonus action for action).

But just like the damage given early in the combat is just as useful as the last hit (rendering damage vs. total foe DPR invalid as a metric), so can healing be useful. Healing's primary goals are to prevent character death and to preserve actions in combat. So a heal on someone that will negate the damage from a foe's attack and prevent them from going down and losing an action, is a worthwhile use of in-combat healing besides standing them up. And it has little to do with the total DPR the foes put out except that has a loose correlation to how much damage any particular character will take before their next action.

And this doesn't even address other at-table details like healing someone to keep them up might keep up concentration on a valuable spell vs. letting them fall, healing a tank with high AC may mean less total DPR from foes as they miss on some attacks that would have hit other characters with lower ACs behind the tank, etc.

So we can state factually that the premise doesn't hold for damage vs. total foe DPR, and that brings into question if it's a valid metric to measure healing vs. total foe DPR. It feels very white room as opposed to dealing with the actual details of how many foes are attacking a specific ally before that ally's next action and other relevant, per-combat, situations.

 

[Tessarael]

BTW, how do you folks cost the act of feeding a potion to a fallen teammate?

5.5e introduced Diablo-style sword-in-one-hand-potion-in-the-other-binge-drinking via BA, but it seems unclear if that is only for the self or for others too. It seems logical that drinking one’s own potion requires less opportunity cost than feeding it to someone else who’s prone and unconscious, but as far as I can tell it’s not spelled out in the rules.

If a table plays with an asymmetry in the cost of gulping a potion for one’s self versus someone else, that adds an interesting dimension to the tradeoff.

Feeding a healing potion to a fallen teammate is a bonus action; otherwise healing or stabilizing them is usually an action, unless you have a bonus action healing spell like Healing Word. So there is some asymmetry in the two, and it is better for someone to heal as a bonus action before they go down and cost someone else an action, or worse end up laying there making death saves while the action continues around them and they get no actions as they are down.

 

[Silam]

Feeding a healing potion to a fallen teammate, or otherwise healing/stabilizing them, is usually an action, unless you have a bonus action healing spell like Healing Word. So yes, there is an asymmetry in the two, and it is better for someone to heal as a bonus action before they go down and cost someone else an action, or worse end up laying there making death saves while the action continues around them and they get no actions as they are down.

That is what seemed logical to me, but I didn’t find a clear ruling for it. I could have missed it.

It’s interesting to consider this asymmetry further… it depends on the amount of HPs given. You can cheaply get 10 goodberries and split them among the whole party. In this way, everyone has a defibrilator kit and everyone can essentially become a secondary healer of last resort. This is a case where the asymmetry exists (you can eat a goodberry as a BA or feed it to someone else as an Action) but it’s not that advantageous to burn your BA preventively on eating a goodberry, since it only gives 1 HP and therefore is unlikely to make a difference in whether the next hit downs you or not.

For a potion that grants more HPs (e.g., 2d8+) then it’s more tempting to take it preemptively as a BA to save a teammate’s future Action.

Of course, besides the action cost, there are other costs… the goodberry costs 1/10th or a spell slot, while the potion costs 50 gp or whatever. And the potion lasts indefinitely (barring a nasty DM breaking the clinking bottles in your pack …) while the goodberry has an expiry date.

So many tradeoffs

 

[FrogReaver]

Just trying to talk something out.

Say there's only one foe left. Is it better to heal, or potentially kill them. Killing them removes the cost (the action) from healing.

Agreed.

Attacks early in the combat that kill an opponent only reduce incoming expected damage by a smaller fraction than when there's only a single foe left in combat. So one early attacks are worth a lot less... except of course they aren't, because that's what's getting us to one foe left.

Proportional value only makes sense if you evaluate total value over the entire encounter, not just the turn the enemy dies. You note this yourself when you point out that early attacks are what get us to the last foe. The issue is that once you reject your version of proportional value, nothing in the logic requires the damage‑value curve to become flat over time - but that seems to be the assumption underlying your later conclusions.

Which made me question the whole premise. Regardless of what turn it is in the example, healing is trading an action (or bonus action for some healing) and a spell slot for an action of the foe. It's always off by the exact same amount - the spell slot - regardless if otherwise it's a push (action for action) or a win (bonus action for action).

This is where our models diverge. Because of the damage‑value curve, damage now > damage later. If you disagree with that premise, let me know and we can walk through it, because everything hinges on that concept. Early damage removes more future enemy actions than late damage, which means the value of damage is not static.

But just like the damage given early in the combat is just as useful as the last hit (rendering damage vs. total foe DPR invalid as a metric), so can healing be useful. Healing's primary goals are to prevent character death and to preserve actions in combat. So a heal on someone that will negate the damage from a foe's attack and prevent them from going down and losing an action, is a worthwhile use of in-combat healing besides standing them up. And it has little to do with the total DPR the foes put out except that has a loose correlation to how much damage any particular character will take before their next action.

And this doesn't even address other at-table details like healing someone to keep them up might keep up concentration on a valuable spell vs. letting them fall, healing a tank with high AC may mean less total DPR from foes as they miss on some attacks that would have hit other characters with lower ACs behind the tank, etc.

These both read as additional tactical context for evaluating healing in the moment, not as a contradiction of the general insight. I’ve never argued that healing has no value early - only that, all else being equal, its relative value increases as enemy count decreases.

So we can state factually that the premise doesn't hold for damage vs. total foe DPR, and that brings into question if it's a valid metric to measure healing vs. total foe DPR. It feels very white room as opposed to dealing with the actual details of how many foes are attacking a specific ally before that ally's next action and other relevant, per-combat, situations.

You seem to be reading my OP as "never heal in combat" or "never heal in round 1", but that’s not the claim. The claim is that unless there’s a pressing tactical reason to heal now, you generally get more value from delaying, because the opportunity cost of not removing future enemy actions is highest early and lowest late.

 

[Retros_x]

Are DMs just not dealing enough damage or running easy combats?

OP presented it well. When DM deal more damage and hit harder you need to heal at some point to bring back downed PCs, but thats about it. Better to kill off the strong hitting enemies or bring them under control with spells than trying to compensate their damage with healing.

 

[FrogReaver]

OP presented it well. When DM deal more damage and hit harder you need to heal at some point to bring back downed PCs, but thats about it. Better to kill off the strong hitting enemies or bring them under control with spells than trying to compensate their damage with healing.

Interesting to note, what this model predicts is that you could buff every action healing spell heal to 999 hp and heals would still mostly be delayed to later rounds outside downed PC or nearly downed PCs.

What would change is which spells provide the most value per spell slot. So you'd potentially see less guiding bolt, command and spiritual weapons cast and more cure wounds.