Warlock Subclass Design

From its first appearance in 3e’s Complete Arcane, the warlock is one of those rare D&D concepts that graduated from a supplemental book into the core game. Back in 3e the warlock was cool because its magical abilities were almost entirely at will powers. It was wild in 2004 to see a class that could use magic without any limits. These days, it’s old hat, but at the time it was a radical departure from D&D’s roots.

That original warlock merged with another 3e class that premiered in the Tome of Magic, the binder. The binder introduced the idea of magical abilities derived by a link to a specific entity. 4e combined those two concepts into the root of what became 5e’s warlock.

Not surprisingly, warlocks are one of the most popular classes in the game. They have great flavor, fun mechanics, and enough variation that you can play several warlocks and feel like you have a distinct character each time.

Pact of Game Design

Warlock design starts with the entity that forms a pact with the character. In the 2014 edition of the PHB it comes online at 1st level, but it looks like that is changed to 3rd level for the 2024 revision. In either case, the pact’s flavor is a huge part of the warlock’s appeal. Getting your design in sync with your flavor is critical to success with this class.

At 1st level, a subclass should grant a suite of expanded spells to help flavor the warlock’s casting. It’s best to find spells that feel distinct for an arcane caster and speak to the nature of the pact. Healing is a great choice for good-aligned or angelic patrons, while the druid spell list is a good place to seek out anything related to nature, plants, or animals. Those subtle adjustments to the warlock’s spell list can go a long way to making a character feel distinct.

The 1st level feature is ideally something that comes up a lot in play, as you want the pact to feel distinct. The Fiend pact offers temporary hit points equal to Charisma modifier + warlock level. Temporary hit points are a great benefit to give out on an effectively at-will basis, as they don’t stack.

To stay within the right power level, look for a conditional bonus (the Fiend grants temp hit points when an enemy drops) that fits into about the same total, scaling bonus. Remember, this feature needs to feel relevant at all levels so it needs to scale up.

The ideal trigger speaks to the warlock’s typical actions based on the pact’s nature. Consider looking at skill checks as a good outlet, such as a free teleport with a successful Dexterity (Stealth) check for a pact that focuses on a sneaky or deceptive entity.

At 6th level, the Fiend provides a meaty benefit to counter the effects of a bad die roll. The effect is equivalent to a 2nd or so level spell. For something at-will, consider an effect that is roughly equivalent to a cantrip cast at the warlock’s level.

10th level provides another significant boost, with the Fiend provided resistance to your choice of damage. I’d peg this as equal to about a 3rd level spell.

Finally, at 14th level the Fiend grants a powerful attack equivalent to a 6th level spell.

I think part of the warlock’s strong design is related to the amount of power packed into its subclasses. As you can see, each level provides a benefit in line with casting an additional spell per day. That gives us a lot of space to work with and makes it easy to benchmark your work against existing designs.

Mechanical Invocations

Building a subclass is only half the battle. Warlocks have access to invocations, and these are an overlooked area for design that helps deliver on your subclass’s promise. In my own work, I treat invocations like spells for sorcerers. I see them as a key part of the design. Though they are optional, they give us more space to let players customize their characters.

As a rough rule of thumb, an invocation allows once per day access to a spell equal to the highest-level spell slot available to a wizard of the level that unlocks the invocation. At-will access usually comes in about six or so levels after a spell could be cast once per day.

There is some flexibility here for spells that offer narrow utility, especially weaker spells that don’t see much use in play.

For invocations that don’t grant spells, use that spell guidelines as a starting point to judge the power level of an option. Invocations that alter cantrips are an exception. You can compare those to effects that boost a weapon-user’s attacks, such as the Fighting Style class feature and feats.

What Makes a Good Pact Entity?

I think the real challenge of warlock design lies in creating an interesting and fun patron. I try to think of patrons as NPCs that the DM can bring to the table as needed. Consider the patron as a character that you want the player and DM to look forward to using. It’s great that players like to lean into this element of the class, as it brings a lot of roleplay possibilities to the game.

A good entity should have clear goals for their warlocks. Why do they offer up their power, and what do they expect in return? The idea of making a bargain for power resonates with players. Lean into that by creating an entity with motives, desires, and flaws.

You can think of a pact entity as your mostly absent boss who makes random, sometimes impossible, requests at odd hours. It’s up to the DM to figure out exactly how that fits into the story, but as long as you focus on suggestions and ideas you should be good.

At one point in the 5e design process we added flavorful drawbacks to warlock class features. For instance, invocations (or their equivalent) gave your character quirks or weird little details. Playtesters didn’t like those, but at the time we made them a required element. Making them a suggestion or option could go a long way to inspiring players without forcing them into details they don’t want.

Bonus Topic: Eldritch Blast

Eldritch blast, and its pervasive use by warlocks, has become a long-running meme among D&D players. Looking back, I think we simply did not have the time to find ways to add more variety to warlock cantrips, especially in terms of what we supported with invocations.

These days, I think it’s a good idea to consider building an eldritch blast replacement for any new warlock subclasses. Think about a cantrip that you can custom design for a subclass, balance it against eldritch blast, and then create a few invocations that augment it in flavorful but useful ways. Making a warlock distinct, especially 10 years into an edition’s life, is tricky, but there are plenty of open areas beyond subclass design that you can explore in your work.