Colin recently posted a thorough examination of some of the flaws in the 5e barbarian. I agree with most of his points, and have several of my own that I plan to do a writeup on eventually (spoiler: your choices in character build and combat tactics are extremely limited and boring). Since I’ve built my feelings playing a barbarian in Brandes’ game (in which Colin also plays, so I’m one of the data points in his analysis), I’ve been discussing various options with them for how you could make different house rules to improve the class. This post is less about redesigning the barbarian, and more a very specific deep dive on the core issue: how do you build a rage mechanic, and how can you raise or lower its potency for balance concerns?

What Even Is Rage?

I assume there were various rages populated throughout AD&D, but the first major times I saw it was in the form of a power possessed by Minsc the ranger in Baldur’s Gate and in the Barbarian class as designed for 3e. Later iterations in 4e, Pathfinder, and 5e have kept various spins of the core elements of:

  • The state is triggered at will.
  • It only lasts a limited amount of time and is tiring.
  • You can’t really do things during it an angry person couldn’t do (i.e., cast spells or other actions that require concentration and intellect).
  • You get really strong and hit harder with things strong people hit harder with.
  • You get really tough and can take more punishment.
  • You’re more resistant to mental effects because your mind is so focused/unhinged.

Basically, it’s unabashedly modeling a less grandiose version of the Incredible Hulk, or the fears of what individuals are like when hopped up on amphetamines or otherwise having a violent mental episode (so modeling Bane). While that’s arguably not the most culturally appropriate thing (why are we glorifying uncontrollable rage and assuming that’s the hallmark of all tribal warriors?), you probably have to start from trying to keep them all if you’re looking at keeping something that’s inarguably a “rage” mechanic.

How Do You Time It?

The default expression of rage has been a fixed, short period effect which you get more uses of as you level, and which has a small but non-trivial action cost to activate on your turn. Pathfinder made the currency much more granular (a big pool of rounds rather than a small pool of larger blocks of time). Notably, the push seems to be to balance it against the expected number of fights in an adventuring day such that you can probably use it most of the time but not all of the time. Timers with a limited resource make this simple.

Another way to do it would be to tie it to a trigger that’s somewhat out of the PC’s control but is less resource-based: you rage until some ending trigger. Non-D&D games have been more inclined to make this trigger “everyone is dead, including maybe your friends.” This is more common in WoD games, and it’s probably not appropriate to most heroic D&D games to regularly start making the barbarian make hard checks to calm down before hurting people other than the bad guys. But it’s certainly worth considering that you can set triggers for when rage ends instead of a timer if you can find a trigger that makes sense for the setting you want to evoke.

You can also tie it to a resource other than actions and time. 5e‘s timer is already somewhat superfluous, since the rage ends if you stop attacking enemies or taking damage, and if you’re in a fight where you’ve still got enemies and HP after a minute, it’s kind of an unusually hard fight where it’s annoying your rage gives out anyway. The presence of people to attack/sources of damage is the resource involved. You could simplify it to a straight up damage over time effect: you end the rage on purpose whenever you’re tired of slowly bleeding HP.

My favorite rage mechanics from video games are ones that are additive in combat and decay over time. In City of Heroes, brutes gradually built up a fury meter from attacking and taking damage, and it diminished over time such that if you weren’t fighting, you were losing the cool fury abilities, so you were very inclined to risk rushing into the next fight unprepared rather than bleed off fury. World of Warcraft has a more sedate implementation for their warriors: unlike mana-using classes that start full of resources and gradually spend them through the fight faster than they can recover them, warriors start with little or no rage and get more from attacking and taking damage, and spend it on special abilities. Both of these approaches are tough in D&D lifted in a straightforward way, because you don’t usually experience a lot of combat rounds over the course of an adventuring day compared to an MMO. But I’m very enamored of the idea of rage being tied to a resource that starts small, bleeds off per round, but can be recovered faster than it bleeds by dealing and taking damage.

There’s also an outside chance that you could make the resource you’re using up just actions in your action economy. Depending on the utility of a bonus action for the class in question, it might make sense for rage to be something you can turn on or off at will, whenever you’re willing to burn actions on it. This might be a 1:1 (rage on rounds you spend an action, no rage on ones you don’t), or one action may get you multiple rounds of time/resource so your anger is like a fire you only have to stoke every so often. This is probably only worth investigating if the overall build has a big demand on bonus actions from all sides (e.g., presently, all it does is make using a two-handed weapon even more significantly better than two-weapon fighting).

Ultimately, rage in 5e is usable most of the time but not all of the time (unless you’re having fewer fights per day that still don’t go very long). Changing the timer to let you use it closer to all the time makes it a little more powerful, and making it so it’s available less often makes it a little less powerful. In my opinion, varying how you govern staying in rage is actually more about making the mechanic offer interesting choices and tradeoffs to the player.

How Do You Get Really Strong?

The traditional implementation of the strength boost was, well, a strength boost. In 3e and Baldur’s Gate, the increase to strength was meaningful, but increasingly less relevant at higher levels compared to other sources of damage bonus. 5e opted to just grant a couple of the derived effects of higher strength rather than forcing you to recalculate by granting advantage on strength checks and saves and a damage bonus to strength-based melee. The improvement to offense is really small, even compared to 3e (where at least strength improved your attack bonus, the extra damage could be stretched with two-handed weapons, and got multiplied on a crit).

5e‘s mechanisms for representing angry strength are pretty comprehensive, though. There are only so many ways to represent it in the system, and advantage on strength rolls and extra damage on strength attacks are most of them. You could theoretically give advantage on strength attacks, but that eats into a lot of other mechanics. Otherwise, I’m not sure how else you’d model “I’m even stronger right now.”

The amount and style of the damage bonus are your primary ways of raising or lowering the potency of this aspect of rage. There’s limited room to decrease the bonus, because it’s already +2 for much of your career, but you could switch it to something that’s not always available (e.g., if you’re using a resource-based rage, adding damage to an attack could be something that costs rage resources much like adding a superiority die for battlemasters). As Colin notes, it might feel better to switch it to a die instead of a flat add, even if you didn’t increase the average significantly, because that would at least get multiplied on a crit.

Instead of a damage bonus, you could also just give extra attacks, have some damage splash/cleave onto nearby targets, or make attacks do a certain amount of damage even on a miss.

Another option would be to increase the applicability of the strength. To wit, currently rage is a big help if you’re trying to force open doors or climb walls as part of a fight, but even if you wanted to blow a use out of combat the timing of it makes it difficult (e.g., if you’re climbing a cliff that will take more than a round, your rage will end because you’re not hurting someone or getting hurt). You could improve the utility of rage by coming up with some way to use it for out-of-combat strength checks. It’s probably a stretch to have it last long enough to meaningfully interact with encumbrance, though (“Grog… so angry… about carrying all this heavy treasure back to town. Still… just so angry.”).

How Do You Get Really Tough?

The traditional method of being really tough was just a fairly small pool of extra HP which might even go away at the end of the rage (so you really got no benefit from them at all unless you kept acting while you should be dead, then promptly died). 3e did this through a constitution boost, which also meant you were slightly better at concentration-based checks and saves. The 5e method is just to let you take half damage from most weapons (or from basically everything if you follow the bear totem), which made being tough the most significant aspect of rage.

It’s difficult to look at different strengths of resistance, since 5e has really made it on or off. You take full, half, or no damage from things. There isn’t a granular damage reduction like in 3e. If you wanted to keep resistance but scale it down, you could do like Colin suggested and make it start out only affecting one of the three weapon damage types. If you’re using a more granular resource model, you could make the player spend rage-maintaining resources to activate the resistance for a round or an attack (possibly using up your reaction as well).

Going back to a bonus HP model instead of resistance would provide more granularity. On the one hand, temporary HP aren’t usually typed in what they can be spent on, so they’re more versatile than weapon-damage-type resistance. On the other hand, they don’t stack, so a rage with resistance is more useful if you’re already getting temporary HP from other sources. There are basically two ways to award temporary HP: as a big, up-front pool, or as a smaller amount regenerated every round (like with the Heroism spell).

The big pool is likely to be better than resistance in most cases, except in fights where you almost died even with resistance (i.e., resistance can be thought of as a pool of temporary HP equal to how much damage you ultimately took, because you didn’t take half, so if you have less temporary HP than your total HP, there are some times it will be less good). But on fights where you didn’t get hit a lot, you might not even have exhausted the pool, where resistance means you always take at least some damage, because it only stops half.

Meanwhile, the smaller, regenerating pool is better if you’re taking a little damage every round (i.e., just enough temporary HP to soak it all up without touching your real HP). But it could be much worse against spike damage. A barbarian that gets missed three rounds in a row then takes 40 damage would much rather have resistance than 5 temporary HP per round, even though on average the math says he took 10 HP per round and mitigated half of it.

Where do you even start on the math involved? It’s obviously highly variable. How many monsters are attacking the barbarian each round, and how much damage do they do when they hit? How long could the barbarian sustain that? Without healing, the average raging barbarian can soak up around 20 HP per level before dropping (assuming 7 + 3 HP for most levels and halving the damage). If you do some rough assumptions that the barbarian on a busy and near-deadly day uses half her hit dice to heal plus gets some miscellaneous healing and has six encounters of at least three rounds each, you can assume a worst-case scenario is that she took as much as 40 damage per level over 20 rounds of fighting, for an average of 2 HP per level per round. As noted, this could be a very bad assumption if the damage actually comes in spikes rather than evenly distributed.

The DMG’s Monster Statistics by Challenge Rating chart paints a much spikier picture of the potential damage output from monsters. How much of the encounter’s monster budget will the barbarian have trying to hit her at any given moment? An 8th level barbarian tanking three characters’ worth of medium budget is facing 2,700 XP of monsters: is that one CR 7 dealing up to 50 damage on the average round it hits, or is that six CR 2s who are much less likely to all hit on the same round, but deal 120 points if they manage it? Either way, does the barbarian need significantly more than 16 HP (8th level x 2 HP) per round to get anywhere near the mitigation provided by straight resistance, or are spikes likely to be weird aberrations and a more conservative number is fine most of the time?

I’d want to playtest the hell out of it, but my gut says 3-4 temporary HP per level per round is likely to be as good or better than resistance under most circumstances. You might want to backstop it with some additional ways to emergency mitigate a spike from a crit or high-rolling spell, but barbarians already do have the best HP totals, so if a spike is so bad it ruins the barbarian’s day even without rage, it would have potentially killed anyone else in the party. The important thing is that if you go with temporary HP on this schedule, it’s pretty easy to reduce them to make rage’s toughness weaker and increase them to make it more powerful.

Finally, you could also show toughness by granting advantage on constitution-based saves, particularly against things like exhaustion and poison. But, like strength, these are more likely to come up when you aren’t in combat, so you’d have to have a way to stretch the effects of rage outside of a fight.

How Do You Model the Enraged Mind?

The biggest fear of the 5e barbarian is anything that requires an intelligence, wisdom, or charisma saving throw. In Baldur’s Gate, Minsc is outright immune to a lot of mind-affecting spells while berserk, and 3e barbarians at least got a small bonus to saves. In 5e, only one primal path gets anything near that benefit, and the tradeoff is that they become exhausted after their frenzy (and that exhaustion is way worse than the short fatigued state that hit 3e barbarians). Also, 5e barbarians can’t cast spells while raging or get any benefit to dexterity or ranged attacks, which prevents many nonstandard builds.

There’s a lot of system tweaking you can do to model what it means to be in a rage, that could make it more or less powerful, or more or less interesting.

Does the rage provide some kind of protection against mind-affecting spells, or do you want that to remain the barbarian’s kryptonite?

Is the rage controlled enough that you’ll allow it to benefit attacks other than strength melee and not lock out spellcasting, or do you like that they’re like bulls seeing red and can do little besides go beat on people? Do you want to balance it by making the barbarian’s tunnel vision even worse and limit her tactical options while in a rage (e.g., “you must attack the nearest enemy”)?

Is exhaustion/fatigue a reasonable cost to add to weaken rage/free up more “budget” to make other parts of it stronger? Is there something you can do to model that without relying on 5e‘s exhaustion track (which is a pretty major limit on how often you could rage, since it’s a short death spiral)?

Are there more interesting psychological aspects of how you want rage to work that suggest mechanics other than the aforementioned?

Putting It Into Practice

Continue through to a fighter college and ranger conclave that provide key barbarian elements in a more interesting class shell.