User:Vaegrim/Archetype Guidelines

General Rules

Creating a new archetype is a matter of finding play space for a class that doesn’t already exist, but also doesn’t overshadow the other existing options. Where possible, new features should compare easily to extant mechanics. Try to reuse existing structures where you can; spellcasting, superiority dice & maneuvers, concentration, resistance, effects dependent on conditions, and advantage/disadvantage. Don’t write new mechanics just for the sake of doing something interesting, this tends to introduce unnecessary complication and can produce unexpected results interacting with other rules. While thematic coherence is valuable, don’t mistake mechanics for flavor.

When attempting to adapt material of older editions, aim for the “feel” of prior designs rather than attempting to reproduce their mechanics. Many design conceits from older editions don’t translate well into the 5th edition paradigm. Concepts like marking, move actions, encounter powers, caster levels, negative levels, ability damage or bloodied can cause frustrating interactions with existing material that are difficult to resolve when encountered at the table. This is also true when adapting material from other games or fictional settings. Many mechanics from video games just aren't a good fit for collaborative role-playing games. Constraints based on non-magical equipment or meta-game conditions can feel arbitrary and frustrating.

If a feature grants a pool of dice that can be spent to create an effect, any effect that expends those dice should involve rolling the die and using the result. Moreover, be extremely careful with features that use Hit Dice as currency. Unlike many other features, hit dice are accrued by all classes; that means any feature that spends hit dice will continue to improve after multi-classing. Additionally, a wild shaped druid can expend their form’s hit dice freely without depleting their druid hit dice pool.

Barbarians

The centerpiece of any barbarian path is Rage. At least two of the Barbarians features must tie to it, ideally either supplementing the rage state or adding an effect to the start or end. While it isn’t necessary that all the features rely on rage, it should be the driver of combat features. Since rage already includes a rationing mechanic, path features tend to be either tied to it or else constantly or conditionally available. Think twice before you add features with short-rest or long-rest recovery limits. Likewise, if you intend to “spend” rage uses on alternative features, it’s imperative they have similar duration and impact. Quite a few core Barbarian features interact with rage, so an archetype that sidesteps it will perform unevenly. Be aware of the endgame consequences of features powered by strength or constitution; either saving throw DCs or uses per long rest. Since level 20 barbarians can potentially increase these ability scores to 24, the limits here are higher than they’d be for other classes.

The first feature of a barbarian path should always tie to rage, ideally it will help shape the barbarian’s primary combat routine. The BattleRager does this with BattleRager Armor, the focus of which is supplanting Polearm Master's primary effect (bonus action d4 weapon attack). Berserker achieves a similar outcome with the broader but more expensive Frenzy. At this point, damaging features shouldn’t be granting more than a potential 9 extra damage per round without a serious drawback. Zealot’s Divine Fury is a pretty safe example, starting at about 4 damage and scaling up to 13 damage per round without any real cost. The effects of a Totem Warrior’s Totem Spirit are rather varied, but all of them are predicated on rage. While none of these features are direct damage boosts, they generally improve either defense or mobility. Wolf totem and Ancestral Protector share the barbarian’s features with the party, but still conform to these standards.

Based on these comparisons, we can see that Storm Aura is likely under-delivering (most obviously with the Sea option). Compared to BattleRager Armor you shift the die up (+1 damage) but lose out on Strength bonus (3-5) and rage bonus (2-4). While the guarantee of damage is nice, as a saving throw it doesn’t help maintain rage and can’t benefit from Brutal Critical. Unlike Divine Fury or Ancestral Protector, it consumes your bonus action which precludes Pole-arm Master’s d4 weapon attack.

The 6th & 10th level features aren’t as easily defined; this space has the greatest variation in possible designs. Ideally at one of these two levels, you’ll get at least one feature that doesn’t require rage and/or at least one feature that helps non-combat scenes. Intimidating Presence is a fair example of a feature with a combative feel, but maintaining potential in social interaction scenes. Barbarians still tend towards a strong emphasis on the Combat pillar, as demonstrated by the Zealot or the BattleRager’s lack of non-combat features. There’s plenty of room to stretch out thematically here, developing more niche features or ancillary aptitudes. Granting cleric or druid rituals is a proven option, though otherwise barbarians avoid spellcasting.

At 14th, the Path feature should either be another big damage boost (equivalent to an extra attack) or something equally flashy. Features that require rage should be a little stronger than those that don’t. While usage-limited features are valid at 6th or 10th, try to avoid them at 14th. This feature should tend towards something that feels inherent to the barbarian's capabilities, rather than a "finishing move".

Bards

With the addition of three archetypes in XGtE, Colleges have a well defined structure with plenty of examples to design from. One conspicuous trend is the two “features” at 3rd; one altering how Bardic Inspiration dice are used. Bardic Inspiration is intended as a primary driver of bard function; that and the bard spell list establishes the bard’s role as one of support and control. The bard generally can’t inspire themselves and are forced to rely on allies. Whisper’s Psychic Blades are certainly the upper limit of offensive power one should see from a Bardic Inspiration. While Valor also has a damage boosting trick with Combat Inspiration, they’re forced to rely on allies to deliver it. On the other hand we have the Lore bard’s Cutting Words and Glamour Bard’s Mantle of Inspiration with purely defensive power. The College of Sword’s Blade Flourish is an acceptable compromise between these two, trading some of the damage output of Whisper with the defensive options of Lore.

It is important at this level, not to fall into the common trap of bard homebrew: forgetting to roll the Bardic Inspiration die. The Improved Bardic Inspiration die granted at 5th, 10th and 15th are a significant portion of the feature budget of the bard. When you add an option that doesn't actually roll the B.I. die, you undercut these features. Glamour and Whispers manage this by manually inserting those improvement into the proper breakpoints and this is the model to follow if you’ve designed a feature that doesn’t roll dice. Blade Flourish shows that Bardic Inspiration can be used to parallel other features like Combat Superiority, that should provide some guidance for suitable features.

The other feature of a 3rd level bardic colleges can roughly compare to those of the cleric domains. While the bard starts behind the cleric in armor proficiency, the Valor and Sword bard’s Bonus Proficiencies provide a reasonable point of comparison. For a new archetype access to a non-bard cantrip should be a valid choice, or even something esoteric like a Favored Enemy. Words of Terror and Enthralling Performance should be about the upper limit on complexity with these features.

The 6th level feature is an important linchpin for the archetype; necessary for the continued viability of the class. For the Valor and Swords bards this is an Extra Attack, central to any weapon based attacker. For the lore bard, the extra spell granted by Additional Magical Secrets presumes it will be used to directly increase the bard’s personal combat capability. Even when this feature could be used for Interaction or Exploration challenges, it’s crucial that this can pull its weight in combat. While Mantle of Majesty manages this by giving you potentially 10 rounds of free first level spells, it’s single long-rest use makes it more of a panic button. The Whispers bard has an extremely potent 3rd level feature which takes some pressure off Mantle of Whispers. In both cases these aren’t a pattern that should be emulated by novice homebrewers however. The 6th level feature should inform the bard’s typical approach to challenges; whether that’s spellcasting, weapon attacks, or something more exotic.

Rather than adding new capabilities, the 14th level feature should help reinforce the existing approaches of the archetype. The Lore bard’s Peerless Skill emphasizes their overall capability with ability checks, something that was already established with Bonus Proficiencies. The Valor’s Battle Magic ensures weapon attacks stay relevant as the bard grows into higher level spell options, likewise Master’s Flourish ensuring the bard can always use a maneuver. As with other Bard features, try to stay away from big powerful long-rest-recovered features here, focusing instead on features that are either always available or have multiple short-rest uses. Shadow Lore just barely passes this test, only due to the effect’s long duration. Once again, it is not a design to be emulated.

Clerics

Clerics have a staggering 12 printed subclasses (as of GGtR), making Domains one of the most comprehensively demonstrated archetypes in the game. The basic layout is this: One or two 1st level features that inform the basic trajectory of the class, a channel divinity option at 2nd, and either Divine Strike or Potent Spellcasting at 8th, and a juicy capstone at 17th. The expanded spell list of existing domains is about 40% spells on the cleric list, but every domain list should have at least one. Levels 6 and 17 are the most open in terms of design options, with some parity to sorcerer features at these same levels.

As a level 1 archetype, domains offer some incredible versatility to the playstyle of the cleric class. Like the sorcerer, immediate access to an archetype feature opens up the capacity for significant deviation from standard build incentives. Primarily, it’s important to establish your expectations for the Cleric’s role. The Light domain builds up the Cleric as a ranged spell attacker, where the War domain’s heavy armor and martial weapons position them as a Melee Weapon attacker. Whatever choice you make here, it should be a significant hook that’s reinforced by subsequent features.

Those two dynamics (blaster vs warrior) are the simplest expression of domain roles, but there are some more complicated examples as well. Instead of martial weapons, Nature clerics can get shillelagh from the druid. The improvement of the Forge cleric’s Blessing of the Forge brings their simple weapons up to parity with martial weapons (and count as magical). And we have examples of even more abstract roles in the case of Blessings of Knowledge and its similarities to Expertise. Whatever style you adopt, the 8th level feature should support it as well. Balancing Channel Divinity in a vacuum is quite challenging. As a rough guide, it should sit somewhere around the low end for a 3rd level spell. Channel divinity should stay relevant over the life of the class, damage or healing effects will need to scale accordingly (unlike spells). Since it’s recovered on a short rest, the duration of such effects tend to be instant.

Druids

When you look at Druid archetypes, the 2nd level feature should be a foundation of the play-style of that circle. It's important to establish a strong identity by way of a mechanical hook right at the beginning. This hook should be something that stays relevant along the life of the class. Land druids are THE caster druid, they get more spell slots at every level than any other druid (thanks to natural recovery). Moon druids are THE wild shape druid, with a better CR progression and an easier time entering and maintaining shape. You can even see this philosophy echoed in the Xanathar's Guide circles. In each case the level 2 benefit includes a feature that helps inform the combat role of the druid and stays significant at every level, with either increasing potency or uses.

The last druid circle feature isn't like the Wizard or Paladin, the 14th level feature shouldn’t be considered as a signature or “capstone” power (though Shepherd flirts with this). Instead it should be relatively persistent and reinforce the themes of the circle without granting too much combat potency, it's a support feature that broadens the druids abilities.

Typically, Druids have two pools of character resources to spend from: Wild Shapes and Spell Slots. The dynamic of existing druid archetypes is to divide these resource pools into the "Primary Combat" pool and the "Primary Interaction/Exploration" pool. The foundation of this is the limitation of Wild Shape; it prevents spellcasting during use, making it impractical to use both in a combat routine.

Since the Land Druid lacks a strong wild shape their combat routine comes from good concentration spells. Spells like call lightning, conjure animal and flaming sphere give the druid an effective, scaling, per-round contribution to combat along with combat recovery (either cure wounds or healing word). Often these effects require only a bonus action, leaving the action for cantrip combat via produce flame or frostbite. Natural Recovery ideally ensures the Land Druid has enough spell slots to perform throughout the day, and means his Wild Shape can be used for Interaction and Exploration. Various beast forms get new vision modes, movement types, skill proficiencies and special features (Keen Senses, Sure-Footed, Web Walker). The Land Druid spell list means more spells prepared, giving more room to memorize ritual spells. When you write an archetype modeled on the Land druid (where the primary combat routine involves expending spell slots), you’ll need to weight the versatility of the effects added to the druid list against the Circle Spells. For both the strength and frequency of the effects, compare how many free spell slots you’d get from Natural Recovery.

Dream’s Balm of the Summer Court is the best example of this. Spending a bonus action to heal someone a small amount at range is basically healing word. Now, "fey energy" healing is better than healing word in some significant ways since it isn't actually casting a spell: it can't be countered, you can still cast a non-cantrip with your action, it’s usable in wildshape. The healing is a little less at first (1d4+mod vs 1d6) but the 1 temp hp helps. The big difference is you get a number of dice equal to your druid level, where natural recovery is slot levels equal to half your druid level. That means comparing apples to apples, it's 1d4+Mod vs 2d6+2 which comes out at best favoring BotSC by about 2-3 hit points per druid level. All-in-all it should be slightly better than Natural Recovery since you didn't get to choose how to spent it, you just got healing word.

While the Moon druid can self-heal with spell slots it's rarely an economical choice (many beasts have a bonus action attack like Pounce or Charge), making wild shapes their primary combat resource. The greatest effect a moon druid can get out of spell slots in combat is a pre-cast concentration spell. This isn't too taxing on spell slots, leaving them for Interaction and Exploration (charm person, darkvision, pass without trace). A druid circle modeled on Moon should have some way of attacking twice by 6th level and needs to improve either his effective HP or AC by 11th. This roughly compares to the animal forms available.

If you let a druid archetype trivially mix spellcasting with wild shape you have two potential problems. The first problem is that you undercut Beast Spells, the level 18 feature. The second problem is that the ability to use both pools at full effect in combat produces one of two outcomes

• The Circle of X druid is just as strong in combat as Moon or Land but has to burn through both pools to keep up; consuming his Out of Combat features to essentially no net gain. Result: weaker than existing archetypes
• The Circle of X druid is stronger in combat than Moon or Land; and is therefore both encouraged to burn through his Out of Combat resources in Combat, and is unbalanced as a combat choice. Result: stronger than existing archetypes

So if you're going to make a Druid Archetype either it's a generally a Wildshape Warrior or a Combat Caster. I think there's room for variants of Moon that grant different forms (like Plants or Swarms or something) or do something new with Wild Shape uses (while still restricting casting). There also room for casters to be more esoteric; give them an animal companion/familiar or conditional metamagic or something.

If you’re following the “Wildshape Warrior” template, the key to keep in mind is the comparison to Moon Druids. If the Druid is relying on Wild Shape to create their combat form, it’s vital that form scaling or combat features continue even past level 14. You have to carefully consider what the new forms can do and if any of the forms you're adding outrun the benefits of a beast at equivalent CR. This can be tricky since CR isn't always balanced with the presumption of that creature operating in a player's control. Beasts are the "safe" creature type in this respect; between conjure animal, polymorph and Wild Shape, they're written with the presumption of reaching player hands. Fey and Elementals are considered less safe, you can see this in comparing conjure animal to conjure woodland beings. Despite using the same chart for quantity/CR, it's a full spell level higher to get fey instead of beasts.

Monstrosities and Aberrations offer options that just flatly outclass beasts at equivalent CRs and aren’t a good choice for Wild Shape. The Giant Strider isn't far behind the Dire Wolf until you add its Fire Burst, which is a simultaneous 4d6 attack AND self-heal. The polar bear is the only CR 2 multi-attacking beast but the centaur, ettercap, and griffon all have comparable damage output while also outclassing it on either HP, AC or both. There are similar dangerous options for Aberrations (the intellect devourer, nothic, spectator). Often more manageable than adding entire creatures types, is simply adding a template to wild shaped forms. My preference is to either devise a template that you can apply (“Circle of the Scale”, applies Half-Dragon template onto your shapes), or create a list of specific forms rather than a type.

Monks

The best way to describe the role of a monk is Front-Line Support. The Monk engages the enemy directly in melee but isn’t expected to be the primary damage dealer. By the time the monk archetype appears, they’ve already got 3 ki points and 3 different ways to spend them. If there is one rule of writing Monk Archetypes it’s this: don’t obsolete monk techniques. The damage dealing Ki technique is Flurry of Blows, so don’t create a more damaging effect for 1 ki. Monk features don’t tend to raise their damage “ceiling” (the maximum damage per round possible), Kensai should be considered the upper limit here. Most commonly the 3rd level feature will extend their existing options in some way, either adding range, control or durability. When adding spell-equivalent features, it’s best to balance the ki cost against spell points (as described in the DMG). Since Ki is a short-rest recovered feature, it’s also important to be careful about which spells to grant in the same way you would a Warlock.

Working around existing monk techniques stays a challenge at all levels. Stunning Strike immediately precedes the 6th level feature, a monk benchmark in the same way as Flurry of Blows. The Stunned condition accomplishes two things for you: it delays the target from acting, and makes them easier to hit. Any effect which provides one of these two benefits more cheaply (considering you may need to make multiple Strikes before you stun) runs the risk of discouraging use of Stunning Strike. Shadow Step only grants advantage on the first attack you make for exactly this reason. Hour of Reaping makes an interesting trade, sacrificing all of their damage to both prevent enemies from approaching and causing disadvantage on their attacks. Although Clench of the North Wind causes the preferable Paralyzed condition, it’s more expensive unless the target fails their save at least 3 times (in addition to only affecting humanoids).

When offering mobility or defensive features, be aware of core monk features that you’ll get later. Don’t try to mitigate the Frightened, Charmed or Poisoned conditions because you’ll get Stillness of Mind at 7th and Purity of Body at 10th. Contrarily, advantage on saving throws interacts explosively with Diamond Soul at 14th. Drunkard’s Luck walks a thin line here and probably shouldn’t be emulated. Shadow Step is a good example of a mobility feature, while it’s superior to Step of the Wind when it’s applicable, the Shadow Monk is often challenged to shape those conditions.

It’s also important to be careful of features which can improve unarmed strikes. By level 5, a monk can consistently make 4 attacks per round. That means features like hex that cause extra damage per hit can add up fast. Adding Ki costs to these kind of features can be an effective counterbalance, every ki spent on a damage boosting effect is one less spent on Flurries. Likewise, if the feature consumes bonus actions while active it can prevent multi-attack abuse. Concentration can also be a useful counterweight, but don’t forget that Diamond Soul makes it much easier to maintain.

Finally, try to avoid features that expend more than 6 ki in a single action. Even a max level monk only has 20 ki per short rest (plus 4 from Perfect Self). It’s not uncommon to stretch this between two separate combat encounters, meaning 5 ki is a commitment of half your “encounter budget”. Quivering Palm outright kills its target on a failed save (and 10d10 damage even if they succeed) but only costs 3 ki because it splits the action cost between two separate rounds one of which needs to include a hit. Touch of the Long Death removes this prerequisite, but only deals 10d10 on a successful save with 10 ki. For area attacks Searing Sunburst costs 3 ki to deal 8d6 in a 20 foot radius, where a 6 ki Breath of Winter deals 8d8 in a 60 foot cone.

Rogues

Rogue Archetypes have a relatively consistent pace to their features. While the power of such features can be shuffled a bit; this is a good structure to think about when developing a new archetype. Running behind any rogue archetype should be the question of combat goals: “what’s my angle?”. Sneak attack is one of the simplest combat goals, and generally a rogue archetype will be shaped to provide a path for ensuring a Sneak attack is delivered every round. Cunning action is an important part of the Rogue’s versatility, as well as their Expertise. That means Bonus Actions are more “expensive” for a rogue than level-comparable features of other classes, and be careful about granting advantage on ability checks. Likewise, the expectation is that a rogue’s best defense is avoiding the enemy completely. Between the ease of applying sneak attack from range and the ability to Disengage after a melee attack, rogues are discouraged from remaining in the thick of combat.

Level 3. You pick up two features at this level that i'll refer to as "The Meat" and "The Dressing". The meat is the feature that draws you to the archetype in the first place. Spellcasting, Assassinate, Master of Tactics, Rakish Audacity, Fast Hands are all Meat. In each case, you choose the archetype because the feature supplied is going to be the centerpiece of a build. That's true because as a rogue, you won't see another archetype feature until level 9. The other part of this level is the addon feature called "the Dressing". Dressing features either extend the rogue chassis to make the Meat work properly (like Fancy Footwork, Master of Intrigue or Bonus Proficiencies) or enhance the feeling of the archetype towards their described style (Mage Hand Legerdemain, Second-Story Work). It's not that these features aren't useful, they just tend to either be very niche or not particularly impactful. These aren't ribbons though, as they often have an influence significant ability checks.

Level 9/Level 13. The level 9 and 13 features have the most room for variation. They collectively extend the capabilities of the rogue in a unique direction and generally establish them as the master of a particular niche style. Rogues often tend to be front-loaded. Archetypes with less potent level 3 features can add more combat strength here, especially if the level 3 features don’t scale with level. Assassins are the best example of a strong combat based level 3, and Infiltration Expertise and Impostor are exclusively Interaction features as a result. Contrarily, if the level three features didn’t have much Exploration or Interaction impact, this is a great place to expand those capabilities. Misdirection is a very potent feature for both offense and defense, justified by the low offensive power of the Mastermind up to that point.

Level 17. The last feature of a rogue archetype is a true “capstone” feature. This effect should be as dramatic an effect as the 3rd level features, but shouldn’t change the overall combat goals for the archetype. This can be an incredible counter (like Spell Thief or Soul of Deceit) or an overwhelming “trick” (Master Duelist, Thief’s Reflexes, Death Strike). The potency of these features should be tempered by their rarity; an effect more frequent than once per short rest should be dependent on uncommon circumstances.

More than any other archetype, rogue homebrew often tends to lean on specialized nonmagical equipment (poisons, explosives, exotic weapons). While this isn't a problem in of itself, it's important to consider the costs (rogue's shouldn't have to spend more than fighters to stay in combat) and the potential for proliferation. There aren't many good ways to prevent a rogue from stockpiling (the Saboteur builds a bunch of bombs during downtime) or distributing (the Poisoner gives out vials to the whole party) nonmagical equipment so it's important to either balance with these considerations in mind. With exotic weapons like garrotes or traps, proficiency can be a helpful control on proliferation. Lastly, try to avoid features that trade sneak attack dice for improved effects (aka Inferiority Dice, because you spend them to deal less damage). Keeping ideal DPR on rogue archetypes is tricky enough without tempting people into spending it.

Sorcerers

The sorcerer is a primary spellcaster with a “something extra”, like a bard or cleric. Also like the cleric, a sorcerer starts getting archetype features from 1st level. As the sorcerer’s spell list is closer than the clerics to the wizard’s more robust list, the sorcerer’s first features aren’t quite as “defining” as those of a Domain. Additionally, at least half of the origin features should interact with the sorcerer’s spellcasting in some way. Wild Sorcerers do this by establishing a twist on traditional spellcasting with Wild Magic Surge, granting features which improve it’s reliability and encourage it’s use. Both Tempestuous Magic and Heart of the Storm work any time you use spell slots (with no other regard for the spell level). In this way both Origins encourage casting low level spells more frequently, than casting high level spells. Dragon is lighter on spell modification, with only Elemental Affinity pushing them towards elemental damage spells and offering two different ways to spend SP on non-spells options.

One of the first features provided by all three existing origins are defensive. Draconic Resilience, Tempestuous Magic, Tides of Chaos all provide some measure of protection. Dragon is the most conspicuous, granting the equivalent of a permanent mage armor and the hill dwarf’s toughness. While Tides of Chaos can be used for advantage on an attack, the more effective use at low levels will be for saving throws in dangerous circumstances. The storm sorcerer’s Tempestuous Magic can be used as a bonus disengage, removing them from melee combat and letting them complete a ranged spell attack without disadvantage from firing with an adjacent enemy.

The 6th and 14th level features are much less constrained. They should generally keep to the themes established and should maintain parity with other archetype/class features of the same levels. It’s not uncommon to see effects that improve mobility in some way and I’d expect a short ranged teleport or burrowing feature. Don’t obsolete metamagic options, since that can be frustrating to players who didn’t plan their character build appropriately.

A sorcerer’s 18th level feature is very much a “capstone” in the way of a rogue’s. At this point, a sorcerer is a high level full caster, so be careful not to reproduce or obsolete the effects of an existing spell. Draconic Presence walks a thin line here, 5 sorcery points is the equivalent of a 3rd level spell slot, and has superior range to fear. The redeeming factor is that passing the save renders a target immune to further uses of aura, but not to further uses of fear.

While some insist a sorcerer origin must not grant new spells known, doing so won’t necessarily create an unbalanced archetype. It’s important not to undervalue the benefit however, especially when granting access to spells from other lists. It’s important to draw a distinction between adding spells to the sorcerer list and adding additional known spells. Sorcerers have the lowest count of spells known among full casters, disrupting this standard is a significant feature and should be valued for each level spells are learned.