Homebrew has always been a welcome addition at many tables. Even with the thousands of possible combinations on offer (Thirteen classes, each with at least three subclasses, and dozens of races besides), there are some people who always want more. Or, more specifically (otherwise UA would do the job nicely), they want something different. Something that the regular options provided can’t supply. Furthermore, it’s just good fun to create. It’s fun to tackle game design under the limitations provided by Fifth Edition, and a lot easier than designing your own stuff from scratch.

The flipside to this, of course, is that not all homebrew is good. In fact, quite a lot of it is actively rather bad. Homebrew has developed a poor reputation, and this has only increased as Fifth Edition has exploded in popularity. DanDWiki, internet-wide repository for homebrew, has become an instant punchline in discussion groups, and for good reason (poor curation and a userbase whose enthusiasm outstrips their knowledge of balance, primarily). Now, homebrewing isn’t easy, especially homebrewing well, but it’s also not some arcane skill only a few can ever possess. It’s something that pretty much anyone can pick up, assuming they’re willing to put in the time to get it right, and not just charge in, d12 hit die and six attacks by Level 10 blazing.

Now, I’m not going to claim to be some master homebrewer extraordinaire. I’m not Reddit user KibblesTasty, who has made pretty excellent and balanced homebrew classes such as their take on the Artificer, Pison, and Warlord. I’m not AngryGM, inventing systems wholesale for use in the game. My record for ‘successful’ (read: seemingly balanced and occupying a niche) homebrew is seven subclasses (one each for Sorcerer, Cleric, Ranger, Barbarian, Paladin, Warlock, Rogue), a few more ideas scrapped for whatever reason (for one, a very similar UA was released), and several half-written ideas floating around. But those seven have all, with a bit of tweaking and in some cases wholesale changing large parts, been well-received by those who have read them, and I’ve DMed for some of it that felt like it slid in perfectly with the rest of the party balance-wise. I’m still sure that a seasoned homebrewer could look through my stuff and pick out problems, but I feel like I can offer some advice for avoiding, at the very least, the worst pitfalls of homebrew in Fifth Edition. Thus, below, we’ll discuss some fairly basic rules to follow, to ensure your homebrew doesn’t fall at the first hurdle.

Does It Need To Be Homebrewed?

I’m of the mind that anything you add to Fifth Edition should be able to justify its own existence in some easily recognisable way. I apply this to houserules most (‘realism’ is not an acceptable reason), but also to homebrew. You should be able to identify a clear need for the homebrew you create. An obvious gap in pre-existing options. This isn’t to say that it has to be something people are clamouring for, but that you have to find something that can’t already be done with what’s already there. I cannot count the number of times I’ve seen people homebrew ‘Dancer’ classes or subclasses that could easily be done with nearly any sort of Bard, or even a Fighter subclass.

Now, this gap can be mechanical, or it can be flavour-wise (more on that below). But there has to be a gap. For instance, there is no Divine Third-Caster Rogue (or Fighter, for that matter). That’s a clear gap there, and one I, and many other homebrewers, have attempted to fill. There’s also, for instance, no Fighter themed around the Norse Valkyrie, with abilities to suit that flavour. There’s countless others, you just need to look.

But, also to consider, can that gap be filled in a more easy way? If you’re looking to do a homebrew class, can it be a subclass instead? The answer is almost certainly yes, and I would advise turning away from homebrewing classes whenever possible. If you’re looking to create a subclass, could the same thing be achieved with reflavouring an existing subclass? This is less certain, but it often can. This is part of the risk with filling a gap on flavour, in that in many cases, reflavouring can achieve the same goal for much less effort. To go back to the Dancer example, unless there are solid mechanics to back it up, it could be a Swords Bard with some fancy description. There’s no Plague Doctor in Fifth Edition (although the Way of Mercy from Unearthed Arcana is starting to look that way), but a Cleric or even a Thief Rogue could stand in easily and have a mechanical backing. If you’re going to fill in a flavour gap, make sure there are some mechanics to heavily support that flavour in a way no other class can. It’ll make your homebrew better, trust me.

Obey 5e Design Conventions

If there is one piece of advice people take from this article, I honestly hope it’s this one, but I’m trying to go in a sort of order. The designers of Fifth Edition are professionals with a lot of experience. They’re better at this than me. They’re better at this than you. They’re better at this than probably any homebrewer I’ve read, although people on Reddit would passionately disagree. They built design conventions into Fifth Edition, and they did this for a reason.

This piece of advice is not only practical for improving your homebrew, but for getting people to actually read it. Pretty much anyone who critiques homebrew will tell you that the fastest way to make people lose interest is to, for instance, have your homebrew class have proficiency in both Dexterity and Constitution saving throws, or to get Rogue subclass features at levels 3, 6, 12, and 14 instead of 3, 9, 13, and 17. It’s like showing up to a job interview with your fly down. You could give the best interview of your life, make the best homebrew subclass known to man, but those fundamental errors at the start are gonna really screw you. 

Luckily, the design conventions of Fifth Edition are not, on the whole, hard to grasp. The very basics can be gotten by a readthrough of existing material, especially for homebrew subclasses. The abilities are gained at certain levels, they tend not to change anything too fundamental (a few subclasses give additional proficiencies, and a few give effective Hit Die changes, but those tend to be restricted to certain classes). Slightly harder to grasp is the general ‘rhythm’ of a subclass (I promise that makes sense), but reading the various subclasses through also helps with this. I especially recommend Xanathar’s Guide to Everything here, as subclass design is more refined and unified, whereas in the Player’s Handbook it’s a bit all over the place.

To give an example of what I mean, Ranger subclasses are some of the most strictly-bound to a ‘rhythm’ out of any, partly due to issues with the Ranger. At 3rd level, they get a damage-dealing feature, some class spells, and sometimes a utility feature. At 7th level, they get a defensive feature (although this can take a variety of forms, from mobility to saving throws to shutting down abilities). At 11th level, they get a conditional multiattack, which is usually dependent on battlefield situations (such as missing an attack, or multiple enemies grouping together, or targeting different creatures). And finally at 14th Level, the ability is a bit more free, but still often defensive in some way. Most classes are less rigid than this, but there’s still usually a rhythm to be understood. Reading the subclasses and, genuinely, reading other homebrew is a great way to try and get a handle on this.

Similar to the rhythm, be aware of the general ‘tiers’ system of the game. Eldritch Blast is the simplest demonstration. At 1st level, it’s one blast. At 5th level, it’s two blasts. At 11th level, it’s three. At 14th level, it’s four. This increase is matched by near enough every class in Fifth Edition (except the Rogue), with only a few variations. If your homebrew is doing three Eldritch Blasts of damage consistently and it’s not 11th Level, you’ve overstepped the mark.

In general, there are hundreds of conventions to follow, but these are some of the more egregious seen. Be mindful of them, and not only will your homebrew improve, but people will be more willing to read it.

Balancing With Only Numbers Is Heartless, and With Only Gut Feeling Brainless

Forgive the title of this section, I’m feeling a bit dramatic. Only a few people would deny that balance is important in homebrew, and for the most part, those are people whose homebrews you don’t want in your game. Balance is important for the sanity of the DM, and the enjoyment of the other people in the party. Xander Harris was a great character in Buffy, but it’s not a fantasy many people would want to live, standing around and watching the Chosen One kill sixteen vampires while they struggle with one. And if you do, that can be achieved with roleplay, rather than with numbers.

But there’s some divide on how you balance. Some people think it’s about the numbers, and other people think it’s about how it feels to play. Now, saying “it’s both” is hardly a revolutionary message, but it’s one that bears repeating. Crunching numbers into a calculator is an important part of homebrewing. It’s a boring part, but it’s worth it. But so is the feeling of your particular homebrew. As it stands, one side of this is far more prevalent than the other. If the numbers are imbalanced, chances are the class is going to feel it, whether it’s underpowered or overpowered. But even in published content, there are some subclasses that are absolutely fine balance-wise, but don’t feel it (some of which you can check out in Five Underrated Subclasses That Are Better Than You Think) to a lot of players, providing an unsatisfying game experience. And this can be as crippling to a homebrew as poor balance – nobody wants to play something that feels bad. Land Druid and Arcane Archer are two examples of this in published content. Somebody balanced those with a calculator, but didn’t pay enough attention to the answer to “how does this subclass feel to play?”

So, once you’ve gotten your subclass written up, and you’re going over it for balance, pay attention to both sides of it. If you put the numbers into a calculator and it’s too much or not enough (and yes, balancing is harder than this due to how many ways there are for something to be powerful in Fifth Edition besides damage), tweak it. If somebody (or better yet, multiple somebodies) plays it and it doesn’t feel right, tweak it. When the two sides reach something of an agreement, then you might have a finished homebrew.

Kill Your Darlings, Don’t Slaughter Them Wantonly

It’s an important part of any creative process, killing your darlings. In writing, that chapter you poured your heart and soul into, all about the background of your story, kills the pacing and must go. In music, the twenty-minute guitar solo in the middle of the album might be your pride and joy, your favourite thing about it, but if it doesn’t do well with test audiences, get rid of it. And the same applies in homebrew. Do not get so attached to a part of your homebrew that any criticism of it feels like a personal dig at you, and don’t let there be a part that you won’t change or even get rid of if it will improve the homebrew as a whole.

You may think that your subclass’s ability to paralyse a roomful of enemies by singing at them, without a saving throw, is the best part about it. The entire thing may be built around that ability, and it may be what you picture every time you imagine your class complete. But when people point out that it’s far too powerful, or out of place on a Viking Fighter subclass, tell it to think about the rabbits and then shoot it through the head. It’s the kindest thing to do, and it will improve your homebrew immensely. 

I once published a homebrew subclass, the Witchfinder Rogue, that I loved to pieces. I thought every feature on it was great. I thought it’d get adoration, love, and a job offer from Wizards of the Coast (okay, not really). I posted it, and it got negative reaction. One person called it the most poorly-designed homebrew they’d seen that week. And of course, I was hurt, and I tried to defend it, but with the mounting criticism, I started from scratch, keeping only a few core concepts. You know what? In hindsight, the second draft is so much better, and a lot of the features I loved in the first one really were a bit shit. Kill your darlings, and your work will improve.

But the flipside is also important. Don’t kill them at the word ‘go’. They’re your darlings for a reason, they might still have some value. Keep the gun loaded, but don’t fire it the second someone tells you to. It might be that the person commenting on your feature doesn’t understand it (which can require editing on your part, but not deletion), or it’s their personal bias, or they’re just wrong about it. The reviewer is right more times than the author wants to admit, but they’re not perfect. Always try and defend it, if it’s got merit, and sometimes the other person will come round, or you’ll realise it’s still good (or even just needing a tweak).

With my Betrayed Warlock Patron, I had somebody decry the expanded spell list as confusing and meaningless, until I pointed out it was based around a theme of detecting and escaping danger and betrayal, and then they agreed it was actually pretty good. In another case, my Vanguard Soul Sorcerer’s core mechanic is attacking and then casting cantrips (and later spells) as a bonus action, to allow for a gish that doesn’t rely on the Blade cantrips. Somebody in the comments suggested I just give them more ways to cast the Blade cantrips, that it made far more sense. But that was missing the point of the homebrew, and the mechanic I had in place was actually pretty good (although doubtless not perfect). That darling survived, and the homebrew is probably stronger for it.

So yes, be prepared to change any mechanic based on feedback, even the one you love more than you love your siblings. But don’t change it at the drop of a hat. Defend it, and see if the criticism has merit. Even then, it might be a tweak, not a murder.

Practise and Develop, Develop and Practise

Again, including this on a list about developing a skill isn’t going to set the world on fire. But it’s true. Practise really does make perfect, and homebrew is as much an art as it is a science. The painful truth is that your first draft of your first homebrew will not be very good. Mechanics will be clunky, wording will be poor, the idea at its heart might not even be good. It may well get some criticism (although a lot of homebrewing communities are good about being fair and not overly harsh with criticism), and it’ll be easy to get discouraged when it feels like you’re just getting negativity for something you worked so hard on.

The criticism probably has merit, but so does your idea, even if you’re not good enough to realise that merit yet. Keep at it. Keep developing your homebrew, in response to feedback and in accordance with your own opinions on it. Not only will the homebrew improve, but so will you. You’ll get better at coming up with ideas, better at developing mechanics, better at the wording needed to get those mechanics across clearly, and much better at spotting both balance and the ‘rhythm’ talked about so much above. Your second draft of your first homebrew will be better than the first draft, trust me. Your second draft of your second homebrew will blow that out of the water.

Nothing is perfect the first time, and neither is anybody. Even if you’ve fallen afoul of some of the common blunders in this article, even if your fantastic Dancer homebrew was savaged by a homebrew community, keep at it. You’ll eventually make something so awesome it’ll be played at a table by somebody you’ve never met. And isn’t that the dream?