# D&D HP Considered Unhealthy, News at 11 ![[Hyperborea shirt.jpg]] tags: #thoughts/dungeons-and-dragons Sometimes you know the response is going to ramble a bit, so you just head straight to feeding the garden. Let's go. ## The Detonator ![D&D-esque HP Recovery](https://x.com/hyperborea576/status/1867303981904765357) I'll include the full text here: > Inspired by a post by [@mikemearls](https://x.com/mikemearls), I'd like to talk about hit point (hp) recovery, and why I feel the system that we have in **[HYPERBOREA](https://www.hyperborea.tv)** is preferable and in better keeping with heroic fiction. Yes, in full disclosure, I am biased. > > The poll Mike presented concerned whether injured player characters (PCs) should be able to recover 100%, 75%, 50%, or 25% of their hp upon a decent night's rest. Of course, in the comments, it didn't take long for traditionalists to submit a preference for a 1-hp-per-night recovery rate, per **[[Dungeons and Dragons|AD&D]]** rules. > > For my own part, I prefer the innovation that I came up with about 15 years ago, when HYPERBOREA was in its earliest development. The concept is that the player rolls the PC's hit die (HD) type for hp recovery following an appropriate rest period (6 to 8 hours). > > With this in mind, let us take the example of a pair of adventuring comrades: > > 6th-level fighter (40 hp) > 6th-level magician (18 hp). > > Let's say the pair got blasted by a fire trap spell that, due to some failed saving throws, inflicted 10 hp damage to each PC. > > The fighter is down to 30 hp (-25%). > The magician is down to 8 hp (-55.5%). > > First off, in my opinion, to allow for 100% recovery simply is not fun. I think it undermines the dangerous business of being an adventurer and removes that edge-of-your-seat feeling that your PC's life is in constant peril. Sure, a 100% nightly reset should be possible, but only at the expenditure of valuable resources; e.g., spells, potions, scrolls. I am a firm believer that resource management breeds greater campaign investment. > > Using rates of 75%, 50%, 25%, or 1 hp are more palatable than 100%, but I find static results in fantasy role-playing games to be less exciting than the wildly variable nature of rolling dice. Furthermore, I don't prefer what static results imply. Let's take the 50% recovery rate as an example. This is the result: > > The 40-hp fighter, who is down to 30 hp, recovers 5 hp, which is 12.5% of his hp total. > > The 18-hp magician is down to 8 hp and recovers 5 hp, which is 27.7% of her hp total. > > Basically, any static sum (even the 1-hp rate of **AD&D**) is going to be more impactful to the magician. In essence, the system is implying that the frail, lowly mage is a better natural healer than the warrior is. For this reason, I prefer our variable HD recovery system, in which the fighter gets to roll a d10, and the magician gets to roll a d4. > > Does it offer the possibility of the magician doing better than her fighter comrade after a single night's rest? Sure! That's part of the fun, that chaos factor that rolling dice provides. The magician might well roll a 3 on the d4, and the fighter might well roll a 1 on the d10. But overall, with the passage of time, the fighter will prove better at shaking off his wounds, which I feel is in better keeping with the type of heroic fiction that I've spent my lifetime enjoying. > > I could get more into the definition of what hp means and the Gygaxian abstraction associated with it, but perhaps that is a topic for another day. > > So, what do you think? Do you like some of the options Mike mentioned, the traditional **AD&D** method, the **HYPERBOREA** method that I prefer, or something else entirely? Lastly, keep an eye on Mike's alternative 5e developments. It's going to be interesting to see where he's going with this stuff. ## Problems I've written about the idea of the [fantasy heartbreaker](http://www.indie-rpgs.com/articles/9/) before, and while the Forge thread that I linked and generally link to whenever this comes up provides some examples of the day, more pop up every 15 minutes like toadstools after a good rain. One of the common traits of the fantasy heartbreaker is that they just can't seem to break away from some of the worst parts of the original **[[Dungeons and Dragons]]** design, and it doesn't matter how much lipstick you put on that pig. Hit points are among the worst design elements of the game for 10,000 different reasons, but among them are that *the abstraction is nailed down at the corners in all the wrong places.* Trying to work around that initially poor design gets us here with some generally messy results. Let's start by breaking down where the traditional idea of HP is pinned. - What HP really represent is never fully explored or explained, and as it decreases there are no ongoing effects until you hit the bottom of the pool. - How many HP you receive at each level is random based on your field of endeavor, not technically inherent to your stats (though affected thereby), so your physical being doesn't actually fully affect how much combat you can take right to the face (which feeds into the problems with the previous point). - Additionally, the degree of variance increases as the level system pushes you up the ladder, generating more HP from your hit die as you go, making it even more plastic and setting up for a lot of runaway situations.^[That is, if your hit die is a d10, every level you advance along with the d4 Wizard, you're going to get further and further ahead.] - Time is fixed: a short rest, a long rest, a good night's sleep, both of which have a very specific time that they last and demand. Of all the things to be truly fixed, this may be the silly one because… - Time doesn't usually matter at all, as generally run in **D&D** and **D&D**'s fantasy heartbreakers. It really wouldn't matter much if you stayed in one place for eight hours or camped out in a room for 24 hours. Either something happens in that time or it doesn't, and it's not meaningfully differentiated whether it's two or 22 hours comparatively. Which brings us all the way back around to a place we can actually start breaking this down. My position is that both approaches are kind of dumb. ### D&D Let's start with one hit point per night of rest. From whence does this come? It's an artifact of **D&D**'s previous existence as a revision of man-to-man skirmish warfare in **[[Chainmail]]**. Of course, in the structure of those campaigns, taking a couple of weeks off from the adventure was perfectly sensible because the usual scale involved marching to and fro for months at a time. If your character is out of play for up to four days from a simple poke from a dagger, that's not a big deal. Really, time pressure was even lower than it often is in modern game designs. You need to recover from a real fight down in the dungeon. You drag yourself back to town and hang out for a month, it's no big deal. Not only that, it's the expected norm. As **D&D** continued to differentiate itself, not just in terms of the game rules themselves but the kind of scenarios and situations which characters were thrown into, time pressure became more of a thing. But it was never really a first-class thing. It was, and is, a matter of rigid timelines for most DMs because they can't imagine a different way to handle it, given that rest times are so absolutely fixed. So we get to Mike Mearls' new suggested approach to recovery of hit points over a decent night's rest. And he shifts from an absolute value for recovery to a *percentile, normalized value of recovery.* This is interesting because it begins to approach one of the problems with HP in **D&D** from the back end rather than simply fixing it. HP is way too wildly volatile between classes in terms of variation, which means that every tacked-down corner pulls very differently on different classes in ways that make very little sense within the context of the fiction. But rather than normalize the mechanical representation of health, Mearls is normalizing on the *recovery time* in this context. It's an interesting approach, but it feels the wrong way around, in part because character decisions in construction and development have no real impact. No matter what the answer is, it applies to everyone roughly equally.^[The irony is that while I'm dissecting it here, I actually think this is the superior solution purely within the context of *heading back to town after getting beat down in the dungeon.* In my mind, the problem is the fixity. *"You drag your sorry asses back and are healed up and refreshed in a couple weeks' time"* seems like the obvious solution. You don't get a good night's rest in the dungeon or on the road; you head back or to town to heal up fully and be done with it. Maybe it takes a night if you're bruised, maybe it takes a month if you've had a train run on you by an orc tribe.] This has certain advantages, in particular that you know up front how long you're going to need to rest to be combat effective, which plays into making it easier for the DM to plan out and figure out what's going on within the fiction that they control. And exactly the same on the player side, I get why it might be appealing — it's fast, it's easy, it's a no-brainer, but it is a little bit bland. It also just doesn't fix the real problem. Now let's look at the **Hyperborea** solution. It starts with almost reasonably calling out that an absolute value of healing favors the character with the lowest HP in terms of rate of being back up in combat effective. This is true while simultaneously absolutely *ignoring* that it means that the character with the lowest HP is also vulnerable to being removed from being combat effective more quickly than their high HP counterparts. That is the inherent trade-off with the design as traditional: you go down fast, you get up fast. This is the problem with a narrow variability being plugged into that much randomness. There is no room for swing. Combat and healing are two sides of the same coin in that case. ### Hyperborea And thus onto the **Hyperborea** variable healing rate which would seek to solve the problem by dumping in *another pile of randomness.* Now, in fairness, it is *biased* randomness with the intention of reflecting at least the character class being meaningful within the framework of the healing process. I'll give it points for that. The problem is that it's not narratively interesting in terms of giving the player *more interesting choices* because it sticks to the fixed time frame through which this can occur. We are still talking about the rate of recovery over a good night's sleep. The only "interesting choice" available to the players occurs in the morning when you roll the die and see how much healing you did. Maybe it was a single hit point, maybe it was the max value on your hit die, which means it's possible for the warrior to have recovered a full 10 hit points and the wizard to just have pulled in a single hit point.^[In the latter case there's a 25% chance of that every single night, which doesn't make for a lot of fun for the wizard.] At that point, the following morning, effectively all you can do is look at the numbers and decide if you want to take another day off. Rinse and repeat until the team is combat effective again. There's nothing really you can do to affect the outcome; there's no real interesting choices to be made and it just doesn't feel any more rewarding to pursue than a flat percentage. There's no real advantage to be had there unless what you're at the table for is to throw dice and hear them clack. I have to be honest, that's not generally why *I'm* there. ## Get Normal, Dweeb So let's rewind all the way to the top and talk about the real problem, which is that damage to the body and mind in **D&D** and its close fantasy heartbreakers is tightly *quantized* rather than being *normalized*, or in short: HP sucks as a variable pool. Normalizing and rationalizing hit points fixes a ton of problems both with damage and healing because they are effectively the same thing. What would that look like? Well, this would normally be [[Recommended Fantasy RPGs For the Discerning Escapist of Wizards|where I made reference to how a bunch of other systems do health and healing better than both D&D and Hyperborea,]] but I'm not feeling like doing all those citations today. If there's interest, I'll do that as a separate thought entry. Let's talk about normalizing **D&D** HP. I'll start with a simple proposition: What if all characters had 10 hit points? All of them. All the time, regardless of their level, regardless of their stats. Every **D&D** character has ten hit points. No more, no less. What would that really mean to the experience? Well, it means firstly that your damage rolls and your healing rolls are both going to need to be affected by stats and possibly class in some way. But also, weapon damages are going to need to be more reasonably fixed. Instead of reasoning in a weird way trying to figure out what kind of die to stick on each of them, you can simply say, *"How much do I think hitting a normal person who's trying to defend themselves but fails would hurt and bring this person closer to being combat ineffective?"* Maybe it's reasonable to you that a dagger would probably need a few good stabs against someone who is actively defending themselves, so you set its damage to 3. And something like a warhammer, which does massive crushing damage, may be as high as 7. This makes figuring out how much damage something *not* in a table does very fast and easy. Awesome. But it does seem to make people considerably more squishy if you just take it raw. This is absolutely true. Obviously, the damage roll must be affected by something, and that thing is a composite of stats like Constitution, any particular combat or defense skills possessed by the character, and their choice of armor. Armor might be the easiest to deal with right off the top because you can have it simply reduce damage from a weapon on a one-to-one basis (to a minimum of one). Is someone with a dagger likely to take down someone in plate armor in a head-on face-to-face battle? Even with a strike that lands, that's pretty unlikely and probably is better handled with other mechanics. So far, this is working out all right. Let's look at the opposite: healing. If weapon damage can be normalized along with hit points, then it makes sense that healing can be normalized in the same way. As much as it pains me to do so, let's maintain the fixed periods for recovery, the short rest, the long rest, and the good night's sleep, thinking about them as the inverse of weapons. Perhaps a short rest is good for three hit points, a long rest for six, and a good night's rest for a solid nine in line with our understanding of normalized hit points. Just as there is for damage, there should be modifiers based on class, stats, and situation. This is where the warrior's ability to heal may come into play, though I wouldn't make it a blanket effect for all of them. Except insofar as a warrior tends to have a high Constitution and a high Constitution should make it easier for you to heal.^[It's perfectly reasonable to have class mean nothing at all here and make it purely a function of skill and stat, but that'd be making it *entirely* not **D&D**, which would be great but not what we're doing here.] Likewise, a warrior might know how to tend to their own wounds where a wizard would not. That might allow a skill roll in order to increase the amount of recovery they can get from a given rest. This leads to a far more interesting outcome that involves actual player choice and differentiation between characters rather than the warrior getting a significantly better outcome from a long night's rest. The warrior can be far more effective after a short rest than a wizard because of their knowledge of how to treat their own wounds and a likely higher Constitution. That said, a wizard with a high Constitution themselves would gain the same benefits with the same knowledge. Now it's easy to calibrate how long you want a good rest to impact play and it's very easy to adjust. How long do you think it should take for a normal person who is combat ineffective to get back to hearty and whole from good night's rests? Two weeks? That's going to make short and long rests useless for healing, but you can do it. About a week? 1.43 HP a sleep, but doable. Same problem with short and long rests. About 3 days? That puts short rests at 1 HP, long rests at 2 HP, and good night's sleep at 3 HP, which works out in a fairly interesting way. One good night's sleep? You can set it all the way up to 9, and you're good to go. As you probably noticed, the real complication once you start examining this kind of setup is that HP in general, as a concept, treated as a pool, is inherently broken because of the necessary relative magnitude of the fact that you would like to play the game in a reasonable way without utterly breaking the fiction. You want short rests and long rests to be meaningful in terms of bringing your characters up to a certain level of combat effectiveness for the invested time. You also want there to be a reason to go back to town and take a breather when resources get too low. But to solve that problem, you have to make even *larger* changes to the way that health and wholeness is tracked, which is going to have repercussions throughout the entirety of the system. At that point, you might as well just play a better game that doesn't use that abstraction at all. [[Recommended Fantasy RPGs For the Discerning Escapist of Wizards|I won't name names, but I probably wrote an article recently literally about which ones you might want to consider.]] Still, normalized hit points are really interesting within the context of the **D&D** framework that already exists. Adding more randomization doesn't solve the problem, it just adds an entirely different class of problems on top of the ones that already exist, along with fewer decisions, and no ability or expectation to be able to make decisions because you don't have good information. That doesn't make for fun, it just makes for dice rolling. Out.