In a typical fifth-edition Dungeons & Dragons adventure, characters will reach every battle with full hit points. Healing comes too easily to enter a battle at less than full health. Above level 10 or so, spells like Aid and Heroes Feast mean parties routinely pass their day with hit point totals above their ordinary maximums.
By the time characters near level 10, few monsters inflict enough damage to seem threatening. Except for a few outliers like giants, foes lack the punch to dent characters at maximum hit points. If round of combat results in a gargoyle hitting a 90-hit-point character 6 damage, then the fight seems like a bookkeeping exercise. “At this rate, I can only survive 14 more rounds!”
I know I sound like a broken record but man, high level #dnd monsters don’t do nearly enough damage. This is true in Mordenkainen’s as well. The Leviathan, a CR 20 monster, does about 62 damage per round when it should do about twice that at least according to the DMG.
— SlyFlourish (@SlyFlourish) May 19, 2018
I think every edition of DnD has had that problem. It’s why authors would tweak 3E monsters with absurd templates and levels. Or why I spent hours picking good monsters in 4E and 5E, plus adding terrain damage, etc.
— Alphastream (@Alphastream) May 19, 2018
The fifth-edition design limits the highest armor classes so weaker monsters can attack stronger characters and still hit on rolls less than a natural 20. This design aims to enable hordes of low-level monsters to challenge high-level characters. In practice, the hits inflict such pitiful damage that the hero would feel less pain than the bookkeeping causes to the player. It’s the pencils that suffer the most.
The obvious fix to high-level creatures and their feeble damage is to make monsters’ attacks hurt more. Mike “Sly Flourish” Shea routinely makes creatures inflict maximum damage on every hit.
But what if the solution doesn’t come from the monsters? What if characters at double-digit levels just have too many hit points?
If high-level characters had fewer hit points, high-level monsters with their puny attacks would suddenly become a bit more threatening. Lower-level monsters could pose more of a threat high-level heroes without becoming too dangerous to low-level characters. High-level PCs would still rip through weak foes, but the survivors could deal enough damage to seem dangerous rather than laughable.
D&D no longer focuses entirely on dungeon crawls where characters judge when to rest based on their remaining store of hit points and spells. The game’s move to storytelling means characters often face just one fight per day. Healing comes cheap and easy, so characters start fights at full hit points. Lower hit points at high levels would suit the reality that characters enter every fight at maximum health. In more battles, foes would seem like credible opponents.
Of course, no one has ever argued that low-level characters sport too many hit points. New characters feel as fragile as soap bubbles. Before level 5, don’t get too attached to your hero. As characters near level 10, they begin to seem stout. They rarely go down in anything short of a slugfest, so they feel like superheroes, but not invulnerable.
But in double-digit levels character hit points keep rising at the same steep rate until DMs resort to letting monsters routinely deal maximum damage. D&D might play better if, somewhere around level 10, characters stopped gaining so many hit points.
When I first considered this notion, I dismissed it as too big a break from the D&D’s conventions. For nearly two decades, characters have gained a full die worth of hit points at every level.
Except for most of D&D’s history, somewhere around level 10, characters stopped gaining so many hit points.
From the original game through second edition, when D&D characters reached level 9 or so, they started gaining hit points at a much slower rate. In Advanced Dungeons & Dragons, fighters rising above 9th level gained 3 hit points per level with no bonus for constitution. Other classed gained even fewer points. Continuing to let characters gain a full hit die plus constitution bonus at every level defies D&D’s origins.
The original limits to hit dice served as co-creator Gary Gygax’s way of putting a soft level cap on D&D. The cap kept the game’s link to the Chainmail mass-combat rules, where the best fighters acted as “superheroes” who could match the power of 8 soldiers. Gary wanted a game where crowds of orcs or goblins could still challenge the heroes.
Admittedly, when I started playing D&D, I disliked how characters’ hit points topped out. Gary and his hit-dice tables seemed to punish players of high-level characters—especially fighters.
Although the soft cap on hit points lasted 25 years, the cap on the other perks of leveling started to disappear as soon as the first Greyhawk supplement reached gamers. While the original box topped out at 6th-level spells, Greyhawk included spells of up to 9th level. Gary never intended player characters to cast the highest-level spells, but that didn’t stop players.
By the time designers started work on third edition, they aimed to deliver perks to every class at every level from 1 to 20. The soft cap on hit points must have seemed vestigial. The designers felt the game’s math could handle a steep rise in hit points past level 10. The design abandoned any aim of making groups of low-level mooks a match for high-level heroes. Besides, a steady rise in HP made the multi-classing rules simpler.
Today’s D&D game does a fine job of awarding every class—even fighters—perks at every level. Nobody leveling into the teens gets excited about another helping of hit points. Reverting to smaller hit point advances doesn’t spoil anyone’s fun.
Fifth edition keeps levels and monsters at power levels broadly similar to those in original game. This loose compatibility makes adventures written during D&D’s first 20 years continue to work with the new edition. In theory, a DM can just swap in monster stats from the new game and play. In practice, higher-level characters have more hit points, more healing, and the creatures fail to do enough damage to keep up. Story-centered adventures make the mismatch worse.
Suppose Gary Gygax had hit points right all along. Would D&D play better if characters stopped gaining so many after level 9?
As soon as you started talking about the slog of bookkeeping HP totals at high levels, I wondered if this is where you were going.
The fantastic thing about D&D’s broken high-level math is that, no matter which edition you’re playing (3rd and up), if you follow the old-school rule of only giving player-characters a maximum of 9 hit dice (and just give Con bonuses from level 10 onwards, instead of the weird 3-point cap for warriors) the math holds up better, the characters feel powerful but never invincible, and the game moves along much faster. I think for 4E you might have to mess with monster HP totals as well since they’re so spongy, but other than that it’s a good house rule for instantly “fixing” one of the most consistently broken things in the game.
We’ve continued using the AD&D hp cap all along. It works great for us.
I think switching to pre-3e style hit points without other changes such as high level spells would impact balance and hurt the martial fighty-type classes. I suggest that if you are running a story-based, low combat game in a lowish-fantasy setting, then go “E10” – levels should be capped at 10 not 20, and use the Epic rules in the DMG from 10th level: every 30,000 XP above level 10 a PC either gains +2 attribute (I’d retain cap at 20, maybe allow a 24 in PC’s highest attribute) or if you use Feats they can gain a Feat of their choice.
Interesting idea, but I’m concerned that I’d have to be fiddling with everything in the system that adds or subtracts HP in order to balance. (e.g. Magic, potions, etc).
To me the solution of more damage isn’t difficult, can provide drama with massive hits to a PC, & doesn’t contribute to players feeling nerfed, for the ones that view it that way.
I also find it much more exciting for me to add a customization to monsters for encounters. Things like, you’ve attacked a tribe in it’s turn, the orcs get a legendary action or two each round. Or things like, these trolls have mutated to adapt to fire, but weak to ice, so now you need to cause acid or ice dmg to stop regenerating.
I feel like tweaks like that are fun for me, & widen the feel of a dynamic world.
That being said, Gary’s value of running your table the way that is fun for DM & players, & rules is a big part of why I think D&D is great !
Hi Dave, thanks for another insightful article. I have rhe same concerns regarding the lack of challenge at high levels of play. I run an 5e game in the spirit of OSR (aka “O5R”) and heavily influenced by Robert E Howard and H. P. Lovecraft. I find that a level cap of 12 helps keep the game grounded for pulp fantasy and cosmic horror. A 12th level character is roughly equivalent to King Conan, John Carter, or Thoth Amon. Powerful, but not a superhero by any means.
I’d like to give this a try when I DM 5E. I’ve never been a huge fan of massively inflated hit points (it’s the Traveller player in me).
Would players still earn Hit Dice each level, even if they aren’t gaining the HD in HP? If so, maybe Second Wind should expand to all classes to help handle burst damage. Fighter would still need a second wind kicker or speciality to make it worth a class feature, then.
Even when I was ostensibly running D&D, I never had HP increases. Characters started with 2*Con HP and never got more. Combatants got _harder to hit_ as they advanced. Before AD&D1 almost no one ran rules as written.
I like this idea of returning to the HP cap. It would restore some lost excitement to the game. I would use 0HP + Con bonus for all classes though, and allow negative results! That could have certain wizards start to go backwards in HP – something which I think fits the RP aspects nicely. Of course it will work better if you start the campaign with that expectation from the get go (and it might put a stop to stat dumping Con).
I’m nice. I was thinking of similar but instead they 1HP + Con bonus after 10th level (thus requiring someone to have really dumped constitution to get a drop) for a old school megadungeon crawl campaign I am working on.
I am also thinking of limiting healing to regaining 1HD per long rest; no auto heal at the end of a long rest; and only able to spend HD to heal at the end of a long rest after the HD regain, use of a Healer’s Kit during a short rest, or with potions during combat and high tension scenes.
I feel like this would might help bring back some of the more classic pre-3e era feel with the slower natural recovery, but the ability to use spell slots for healing when “on the move”.
I might add in an extended rest option (7) which recovers all Hit Points and half Hit Dice and encourage including downtime elements into play.
This reminds me of my first reaction to reading the 5E PHB and thinking “you have to stop at level 20?” For some reason it bugged me that there was an upper limit. Then I remembered “back when I was young and AD&D was new and we walked to school uphill both ways …” that we never really got characters beyond 9th level much. Once to 12th.
Your post reminds me that “back when” there were inducements to gracefully “retire” a character somewhere around there.
Personally I’ve not liked the “encounter math” of 5E much in any of its facets. The “balance” for CR is an interesting change from earlier editions, but it took me awhile to get used to the tool, and I still rely heavily on Kobold Fight Club for building encounters. But even there, I realized that I would often have to double or triple what the math calls “Deadly” for an encounter to make a party beyond 3rd level break a sweat.
In the end, or at least at this point in my thought on my favorite game, the HP and “combat math” doesn’t much matter to me.
5E seems to me focused more on and played by those who value a more combat oriented, “on line game” approach where it’s all about optimizing stats and getting big numbers to rack up points to get more numbers. As a player, me and my group, “break” that math regularly. CR is based on four *individuals* in a generic “fight” against creatures acting as individuals. Teamwork, tactics, terrain and cunning break the math quickly.
So I ignore it somewhat. I use CR like I used to use “number appearing” as a rough guide to what’s reasonable. (I still like “number appearing” better. It describes the nature of the creatures, not what’s an “ideal and balanced” encounter. More verisimilitude.) I have started to “tweak” and create in most encounters. I stopped worrying about the math. I went back to what drew me to the game in the first place, boiled down: mystery. “I do not know what’s around the corner….”
So I make sure that they don’t when I DM. Not always. Sow doubt. Cast shadows. Leave clues. Keep wondering and they have a sense of wonder.
The math is mechanical, a tool. It’s not the spirit of the thing. I have lots of ways besides maxing damage to give a party gas.
“# Appearing”! Yes! Your comment is spot on – this “trait” spoke more to the creature, rather than building an encounter. A kobold is dangerous not by being CR 1/4, but because they come in packs of 6-30!
Seems like a great idea to me!
I played some oD&D and 1e AD&D way back, some other RPG’s, and recently B/X D&D, and “5e” WD&D, and yeah low levels are still fun with 5e but, hile gaining a level is fun, playing at higher levels is of diminishing funness.
Limiting hit point inflation seems to be on the right track to me.
Hi, David! My continued thanks to you for such an awesome blog!
Relevant to this topic, I’m currently running Tomb of Annihilation. Like you, I tend to run a relatively challenging campaign, as I find it is more fun when the PCs have pressure. Overall, I’ve had to really add a fair bit to compensate for the PCs’ resilience.
Now we’ve come to the Fane. Here I’ve added a twist… I’ve placed the Dungeon of Doom adventure (available free on the Dwarven Forge site) inside the Fane. In a way similar to Tamoachan, the adventure provides its own pressure by limiting short rests and long rests. The change has been dramatic! The team is now regularly starting encounters at less than full hit points, significantly upping the challenge. (It is the Dungeon of _Doom_, after all.)
So, from this limited test, I would say you are right in this and your previous post. Limiting healing, limiting hit points, these would all significantly change the challenge level for PCs.
Is that more fun? It’s hard to say. It is for me, and seems to be fun for my players. But I am fairly sure other groups will want a lower challenge level. And, clearly, others find the default system to work completely well.
It is interesting to me to note that some very experienced DMs struggle with these issues. But, what about new DMs? Here I go back to the encounter design system for 5E. It is both hard to implement (regardless of whether you use the DMG or Xanathar’s) and once utilized the result it creates is unpredictable. I think that makes it particularly tough for new DMs. WotC has recommended only using the rules to see if an encounter is deadly or not, but I don’t think this helps most DMs. Does altering HPs or some other factor get around this issue? I don’t think it does.
I want to go back to another factor: rests. One possible problem with capping hit points is the issue of the “5-minute workday.” Because 5E players have so many ways to heal, they keep going. If we limit their healing or cap hit points, won’t they just rest more often?
Hi alphastream. I thought it went without saying that a DM should limit rests. In a typical session my players get 0-1 short rests because there’s limited time, a risk of wandering monsters, or both. I couldn’t imagine running D&D without some kind of pressure ensuing that rests are carefully considered. 5e is built around resource management, it’s baked into the DNA of the game.
Hey, Brock! Interestingly, the 5E team has said the assumption is that PCs have full opportunities to start any encounter at full HPs and to get rests as needed. There is no attrition or exhaustion of resources expected by the core rules.
It doesn’t surprise me that the 5e team said something ridiculous like that; Wizards presents 5e as being things to all men. In a sense they succeeded- 5e is a wonderfully robust system that lends itself well to tweaks, hacks, and reskins. That being said, resource management has been the heart and soul of D&D from day one and is impossible to ignore without consequences (boring combat being a big one). There are many better systems out there for DMs who want an evening of roleplay capped by a combat or two.
The PHB does say that a character can only benefit from a long rest once per every 24 hours. If you bother tracking time, and your players are in a dangerous environment this limitation can be effectual.
My players are currently in a dangerous wood, having retreated from a dungeon, and have elected to lay low long enough to benefit from a long rest. They could try leaving the wood, but they’re hours in and draw more attention (higher % encounter rate) when travelling. They’re low on resources (HP, spell slots) though so the “stand still” encounter rate is providing some real tension at the moment.
I took a guideline and made it a hard-stop: 2 Short Rests per day. 1 Long Rest per day. You have specific limits. [I also instituted a light “lingering damage” ruleset, where big hits impair your body’s ability to *heal*, while not impairing *function* (to avoid “death spiral”), so even a throwaway encounter where only one PC got hit once might linger on for a few days…]
No mention of THE biggest change to healing rules from AD&D: long rest gives characters ALL their hit points back. This never made logical sense to me. Plus hit dice healing for short rests is significant as well…
One small bit is that you don’t automatically recover all your Hit Dice. Rarely a factor, but it is a constraint in some cases.
I remember that one of my favourite games coming out of 3E being OGL was the d20 Conan rpg by Moongoose. In that game there is a cap at level 10 after which you stop rolling for HP every time you level up and just add CON. It made the game much more grounded and still didn’t feel like the fighting classes where in any way nerfed. However, and this is important, the game made sure that characters would continue improving after lvl 10 by having a different approach to AC. In there, AC depended on the level (essentially, characters would have a level dependant attack bonus and Defence Bonus), and armor only added damage reduction.
I understand that because of the flattened math of 5E, character progression may seem inexistant for fighting classes if there is a cap on HP. For example, a fighter gets its third attack at lvl 11 and does not get a fourth until level 20, which means there are 8 full levels in which the fighter is not getting better at hitting things (except for proficiency going up by a meager 1). If we add to that the fact that AC does not change significantly either, we end up having a character that for 8 full levels has next to no improvements in its offensive and defensive capabilities. Now, add the fact that its hit points do not increase significantly and you will have as a result that a lvl 19 figther is not that much better than a lvl 11 fighter.
In comparison, the power increase of a spellcaster from level 11 to level 20 is significant. If there is to be a HP cap at, say, level 10, I see no point in taking any more levels on that class after that. Every player will automatically try to multiclass into a spellcasting class, which is OK if that goes well with the theme of your game, but sucks if you are playing in a Conan-esque where magic users are tipically the villains.
How would I solve it? Two things. First, I would make it such that the basics of computing AC change after one hits lvl 10. I would include for the fighting classes the following trait. Your level gives a bonus to armor class equal to (lvl-10)/2. i.e. +0 at lvl 10,11; +1 at lvl 12,13; etc. In that way, armor class increases every two levels, making it feel like melee fighters do get more resilient.
Second, saving throws must be bumped up also because this is D&D and a lot of magical damage bypasses AC altoghether. I would make it such that the same bonus is added to all saves for those fighting classes. If this is still not enough, at level 15 I would give all such classes the ability to become proficient at one saving throw they are not proficient in yet.
ACKS does this…. I never thought about why. This article helped provide insights into that. Thanks.
In D&D 3e, every single design decision turned on the ability to multiclass by simply taking the lowest level of a class that you didn’t already have. That’s why feat pre-requisites checked if you were “fighter enough” by looking at your Base Attack Bonus, or magical enough by looking at your Caster Level.
Simplicity demands that you get a hit die at each level, because you don’t want a Wizard 8 (with his 8d4 hp) who decides to pivot into 5 levels of d12-hit-die Barbarian (for some reason) having significantly different HP than a Barbarian 5 who decided to branch out into 8 levels of Wizard. Having non-linear behavior based on total Character Level in theory requires you to keep track of the order you picked up levels.
Aside from the move to use d20 vs. DC (or AC) for everything, multiclassing was the most important design decision in 3E. It colors pretty much everything,
The multiclass complication is probably the most telling point. When “order of operations” is important, it could artificially drive character development by mechanics. However… it already does, as you don’t get all the level proficiencies (armor, weapon, skill) of a new class, so why must we shy away on HP either?
In my experience starting at about lvl 10 , caster/martial balance tips heavily towards the caster types , and this discrepancy also holds for PC’s adversaries. Casters are able to inflict significant amounts of damage whlie enemy martials try to hack into soft and squishy casterPC’s. The problem you cover imo only applies to a low magic setting. Also there is a homerule that sets longrests to be a week long undertaking and shortrestd to be a day long one
Horji, are you saying the proposal to “stop massive hp gain at level 11+” would make casters *more* powerful compared to martials? or less? Other than Power Word: Kill suddenly becoming a functional spell (since PCs will have far fewer hp), where do you think martials will come out worse than the casters?
Stop playing 5e. That was my solution, after two years of trying to “fix it” with an ever-expanding list of house-rules. Go back to AD&D, or B/X, or play something else. I like Basic Fantasy myself, a streamlined version of B/X. It is far easier to house-rule B/X or BF into the game you want, than to try to “fix” 5e. If I were going to run 5e at all, I’d use “milestone leveling” to keep the PCs under 5th level as long as possible. 5th-6th level is where the game starts to break down.
Goshin, this is actually pretty insightful (while being offtopic). I’ve been house-ruling for a year now – 46 pages and climbing! – and one of my players recently said “why don’t we just play Pathfinder 2.0?”
Pingback: Making High-Level D&D Click: Advice from Alan Patrick, the DM Who Has Run More Tier 4 Than Anyone | DMDavid
Dave, this idea seems like a much better solution to me than “scale up the damage” that I’ve seen a lot of recommendations for. Just for starters, the Concentration saves remain reasonable even though the hits are still deadly!
It will require more structure/houseruling, though, depending on how you implement across the classes and multiclass characters. Have you given any thought to a “rule of thumb? My instinct is KISS: “after level 10, you gain bonus hp according to your class with the highest HD” (hurrah for the wizard 9 / barbarian 1); my preference would be slightly more math “… bonus hp according to the type you have the most HD of” (sorry wizard 9/barb8).
Another aspect of early D&D was that leveling up was a doubling function instead of 3e’s “sum 1 to n”; each level required around twice the XP of the one before it.
When you compare the two progressions (Doubling vs Sum 1 to n), Doubling advances slightly faster until around 4th level, at which point it plateaus to where advancement between 4th & 8th is slower. Leveling broke down like this:
Low-level: 1st to 4th.
Medium-level: 5th to 8th.
High-level: 9th to 12th.
Epic-level: 13th & up.
Medium-level (4th to 8th) was my favorite. Characters were studly enough that a couple orcs wouldn’t be a problem, but a warband of over 100 orcs coming over the hill at you (with a couple higher-level leaders) would be. I never understood leveling up fast into double-digit levels; at that point the DM needed to ring in super-monsters and gods, and any low-levels in the line of fire would be scorch marks on the pavement, so how could they ever get to high-level?
When I tried 3e, I was surprised at the four-character parties of “One Fighter, One Mage, One Cleric, One Thief” all rolled up for a closed campaign and levelling up together. None of the OD&D eight-to-twelve-character parties with everything from 1st to 8th level adventuring in a group. Where there was strength in numbers and high-levels had to run interference to keep the low-levels alive if things got too hairy.
“In a typical fifth-edition Dungeons & Dragons adventure, characters will reach every battle with full hit points. Healing comes too easily to enter a battle at less than full health.”
I’m sorry, but if this is happening in your games, you’re doing it incorrectly. You are allowing players to take too many long rest and not using short rests the way they are meant to be used. Players should not reach every battle with full hit points, mine rarely do.
David, you’re discounting spells, potions, feats, and class abilities. My players – level 7 now – generally are “functionally” at max hits when they head to the next encounter. Maybe due to potions, maybe due to Healer feat, maybe due to the warlock’s Celestial Light or the bard’s spells. Maybe down some hits, but “covered” with some Temp HP from Inspiring Leader, or some herbal brews crafted by the Rogue.
Unless *every* encounter is truly Deadly in intent… in which case, they tend to withdraw and retreat. And despite the odd wolf attack or story-driven assassin squad, they begin the next day fine. And that’s with me using “Long Rest: roll your HD to see how much you heal” and “Critical Hits lower your MAX HP by 1; chance to recover once/Long rest” house rules!